Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2007-09-05

00:00:33DefineBytefind out where the config file is (it's hidden by default), attach, find log file, attach x2 v run rbutil attach.
00:00:38 Join jac0b [0] (
00:00:55DefineBytehmm i think i miscounted xD
00:01:02 Quit Matt-W (Remote closed the connection)
00:01:13DefineBytestill the config file is hidden by default
00:01:16 Quit davina (Remote closed the connection)
00:01:23LloreanIn that case, there's a lot of steps to installing RBUtil, creating the log, etc, as well
00:01:33DefineByteyou want a clueless user to delve into the underbelly of RB?
00:01:36LloreanEither way, one requires an additional program to be run or available, the other does not.
00:01:45LloreanDefineByte: How is "copying one file" delving into the underbelly?
00:01:56 Join Rob2222 [0] (
00:02:10DefineBytethe steps for installing RBUtil have already occured. it is the only supported method of installation (in the future)
00:02:27LloreanBut again, there's no guarantee it's on the computer they're using
00:02:44LloreanSomeone's player dies at work, they seek help during lunch, for example
00:02:46rasherThere's no guarantee it *can* run rbutil
00:02:51 Quit BHSPitMonkey (Connection timed out)
00:02:59LloreanThere's no sense in REQUIRING RBUtil if it's not *necessary*, and in this case, it's not.
00:03:19KalthareIncidentally, is it intentional that rbutil isn't included in the tarballs?
00:03:21DefineBytejeez, what is the point of RBUtil then?
00:03:22jac0bI am having trouble building a rb environment on ubuntu
00:03:25LloreanThe whole point of Rockbox is to not tether your player to a specific piece of software, and we shouldn't tether support to that piece of software beyond any files created during installation
00:03:31LloreanDefineByte: It's a bloody INSTALLER
00:03:38jac0bchecking whether the C compiler (gcc ) works... no
00:03:38jac0bconfigure: error: installation or configuration problem: C compiler cannot create executables.
00:03:39DefineBytehaha easy :p
00:03:44LloreanDefineByte: How many more times do you need to be told that, seriously?
00:03:49jac0bthat is the error I am getting
00:03:55DefineByteso why is it called rbutil then and not rbinstaller?
00:03:55 Quit petur ("Zzzzz")
00:04:02obojac0b: have you installed the build-essential package?
00:04:07LloreanDefineByte: Because it can also create voice files, talk clips, install themes...
00:04:27jac0bwhere is that
00:04:32DefineBytehmmm i thought it was just an installer. my mistake :)
00:04:39oboit's a ubuntu package... so apt-get/synatpic etc
00:04:48jac0boh okay
00:04:54LloreanDefineByte: Its purpose is to be an installer.
00:05:04DefineBytebut... :)
00:05:05LloreanDefineByte: But it serves several secondary functions.
00:05:21LloreanEven if it does serve other functions, none of those is "be required for support requests."
00:05:22DefineBytelike formatting support requests? no? why not?
00:05:36DefineByteand why is that?
00:05:43LloreanI've told you, several times
00:05:48LloreanIt may not be available when the request is being made.
00:05:56LloreanExplain to me why, if software is not necessary, we should MAKE it necessary?
00:06:31jac0bthat should be added to the wiki for
00:06:41 Join atsea-68 [0] (i=atsea-@gateway/tor/x-a00f7fdf1a6daefe)
00:06:49 Quit atsea-68 (SendQ exceeded)
00:06:54DefineBytei don't remember saying support should be denied to those who don't follow the formal procedure. one minute, i'll check
00:06:59DefineBytewait, nope :)
00:07:07jac0bobo: thanks that worked
00:07:07 Quit SoapSud ("ChatZilla [Firefox]")
00:07:08delYsidbesides, qt is not accessible, there are people on that planet that can not even use rbutil.
00:07:13LloreanDefineByte: You said the formal support process should include an unnecessary step.
00:07:13Kaltharejac0b: it's specific to Debian and Ubuntu.
00:07:32BagderdelYsid: it is accessible, or at least it can be
00:07:46DefineByteno the FORMAL request procedure should require it. informal requests would obviously...not
00:07:46jac0bmaybe at least a note saying to install that for debian based os's
00:07:56delYsidBagder: no, I tell you, reality is different. Trolltech is doing pr, but there is no working solution yet.
00:08:03Bagderjac0b: it is mentioned in the wiki already
00:08:04LloreanDefineByte: Why should we have to ask "Give me this", find out their log contains outdated cfg settings, then make a second request, rather than requesting exactly what we need up front?
00:08:30LloreanDefineByte: Why should the formal request be made more complicated by the additional chance of unavailable information?
00:08:44DefineBytewhere are you getting this outdated log from? you were the one who started all this paste the log business. :)
00:08:51LloreanDefineByte: You really, really need to stop making things more complicated. Yes, RBUtil doing it might be *slightly* easier, but it also add a degree of complexity that invites failure.
00:08:52rasherBagder: The fact taht FILES need to include FILES is a bit... odd
00:09:00Bagderhehe, yeah
00:09:13LloreanDefineByte: I explicitly said I do NOT WANT IT PASTED. You're the one who mentioned pasting it to the forums.
00:09:24DefineBytehow is running one command more complicated than changing rockbox settings and copying two files neither of which you've never encountered before?
00:09:27LloreanDefineByte: And I don't appreciate you claiming otherwise.
00:09:56rasheramiconn: Are you sure we want to replace usb with U S B?
00:10:08 Join ddalton [0] (
00:10:18DefineByteheh, you are silly. the pasting part is not the important one :)
00:10:21 Quit jac0b ("Leaving")
00:10:21bluebrotherdelYsid: that is not true. It is accessible
00:10:33bluebrotheroops. *scrolls down*
00:10:50rasheramiconn: "Ausblenddauer" -> "A U S Blendmodus"
00:10:56delYsidno, qt is not accessible right now.
00:11:01delYsidi would know if it were.
00:11:04DefineByteyou suggested attaching an old log file, i suggested attaching a newly generated support ticket :)
00:11:07bluebrotherwell, I had rbutil speaking. So ...?
00:11:08 Join atsea-68 [0] (i=atsea-@gateway/tor/x-2d57af93150f4542)
00:11:13amiconnrasher: Hmm, then it should be \bUSB\b
00:11:17ddalton~what with?
00:11:19bluebrotherwhat "PR" are they doing?
00:11:38rasheramiconn: I guess, or just without the /i
00:11:38bluebrotherI tried a demo version of jaws, Thunder and NVDA.
00:11:41delYsidbluebrother: on windows, right?
00:11:46LloreanDefineByte: I suggested the installation log, and the current config. You suggested requiring they use a program that may not be installed.
00:11:52delYsidbut on linux, it is not
00:11:54amiconnrasher: I rather don'
00:12:01amiconnt rely on capitalisation
00:12:03bluebrotherwell, what's spoken on linux anyway?
00:12:14delYsidbluebrother: gtk2
00:12:26bluebrotherand 95% of our users are most likely windows users.
00:12:29ddaltonand how did jaws go? with it?
00:12:32amiconnHmm, or perhaps in this case it's better
00:12:42DefineBytefor the formal procedure, yes, it should be installed :)
00:12:48*amiconn fix
00:12:52LloreanDefineByte: NO
00:12:58LloreanDefineByte: The formal procedure should NOT require extra software.
00:13:01BagderI agree with Llorean
00:13:01bluebrotherwell, I don't know how jaws usually goes. It worked fine as far as I can tell.
00:13:07DefineByteand and reason for the no? :)
00:13:11rasheramiconn: well, \bUSB\b would probaly work better, I suppose. But as it is now, it's broken
00:13:16KaltharedelYsid: I'm not sure I understand. what's the problem with qt on Linux?
00:13:25LloreanDefineByte: I'm going to put my foot down here, and say that as the primary support contact for Rockbox, I will NOT write a formal procedure requiring any software beyond a file manager, a USB 2.0 driver, a FAT32 driver, and Rockbox on a player
00:13:33DefineByteyou seem to think i'm closing other avenues of support
00:14:03LloreanThe informal procedure can use handy tickets generated by another piece of software.
00:14:08bluebrotherDefineByte: what's the intention? Making rbutil a real support tool would be a bunch of work too
00:14:26amiconnrasher: Yeah. Strange thing is - I _did_ notice the A-U-S-Blenddauer, but didn't realise it was my fault...
00:15:06bluebrotherand as I assume rbutil will help decreasing the support questions because it does everything needed for installation why put work in yet another tool?
00:15:11bluebrother(that even limits the user)
00:16:42DefineBytellorean suggests changing rockbox settings and copying and attaching files, all i'm saying is that the fist part if not needed and may be confusing to the user
00:17:04LloreanDefineByte: I never once suggested changing Rockbox settings.
00:17:16bluebrotherchanging settings? In what way? And what's confusing about resetting the settings?
00:17:19 Quit Thundercloud (Remote closed the connection)
00:17:57 Quit DerPapst ("So Long And Thanks For All The Fish!")
00:18:11 Quit atsea- (Remote closed the connection)
00:19:26DefineBytewait, i'm slow for some reason i was seeing things from within rockbox. forget that. :D
00:20:45DefineByte.rockbox == hidden by default of course but...we're not in rockbox
00:21:12DefineBytehow would you prevent users copying outdated logs?
00:21:24LloreanDefineByte: What outdated logs?
00:21:35DefineBytegenerated by rbutil
00:21:54 Quit nicktastic (Read error: 104 (Connection reset by peer))
00:21:56LloreanIs their installation going to magically change SVN revisions without a new install?
00:22:18LloreanThe RBUtil logs should be of the sort that don't become outdated until there's need for a new log anyway.
00:22:25DefineBytehow would you know if the user didn't manually copy a new .rockbox folder over?
00:22:29bluebrotheroutdated logs will only occur if the user installs Rockbox manually
00:22:32DefineByteit's a simple unzip
00:22:39bluebrotherbut an average user will just not do this ;-)
00:22:42LloreanDefineByte: In that case, any log generated by RBUtil would be bunk too.
00:22:46bluebrotherrbutil also downloads etc.
00:23:11LloreanA user can falsify whatever he wants, and/or not provide useful information.
00:23:19DefineBytehow would it be bunk? scan DAP on support ticket generation
00:23:35LloreanDefineByte: Scan for what? If a user manually installed, their files could contain whatever they want them to.
00:23:42 Quit Febs ("ChatZilla [Firefox]")
00:24:02DefineByterockbox-info? the md5s of the files? anything
00:24:03LloreanDefineByte: As you said, the only supported installation method will be RBUtil, in this hypothetical future. If they installed by other means, it's not supported now is it?
00:24:17DefineBytehow would yoiu know that?
00:24:26DefineByteRBUtil would tell you
00:24:32LloreanDefineByte: If they built on their own computer, but it's clean SVN, the MD5s may be different, and rockbox-info could be deleted or not copied over.
00:24:54DefineBytethe md5s would be different
00:25:00DefineByterbutil would check that
00:25:21LloreanBut it'd be a 100% valid build.
00:25:49LloreanIf you build the same SVN source on AMD64 vs x86 versions of arm-elf-gcc, for some reason there's variance.
00:25:49bluebrotherI'm really under the impression we're discussing some rare case here.
00:25:58bluebrotheris it really worth it?
00:26:03DefineBytebut the logs would be inconsistant. RBUtil could tell the support "staff" what build the user was actually running
00:26:22LloreanDefineByte: How do you identify it as *anything* but "Not the current version"?
00:26:33bluebrotherDefineByte: it is absolutely NOT possible to guarantee logfile and installed build to be in sync
00:26:37DefineBytebe checking SVN? :)
00:26:42bluebrotherwon't work.
00:26:42 Quit ddalton ("leaving")
00:27:03DefineByteput the MD5s in FILES? :)
00:27:08LloreanDefineByte: How does that help? Build every previous build on AMD64 and x86 and take all their MD5s?
00:27:08bluebrotherbesides, it isn't worth the work needed to implement such a thing
00:27:34bluebrotherwhat's the problem? If a user has problems just instruct him to update using rbutil.
00:27:40DefineBytei guess if you see support as fun instead of work :D
00:27:41 Quit ender` (" The secret of success is sincerity. Once you can fake that, you've got it made.")
00:27:41bluebrotherthe same as we do now: update first.
00:28:01 Quit elinenbe (Read error: 110 (Connection timed out))
00:28:03bluebrotherafterwards we can be sure log and build match.
00:28:23LloreanDefineByte: Your method doesn't ensure anything, so I don't see what you're talking about.
00:28:54KalthareCan rockbox only be built on x86 hosts, then?
00:29:08LloreanKalthare: Nope, any host you can build the appropriate cross-compilers on.
00:29:17KalthareSo, there's that, too.
00:29:35DefineByteit would assure you know what build the user was running. look if support isn't really something you want to put a lot of work into i understand. this is an OS community after all and i'm not sure anyone else in OS offers something similar
00:29:56LloreanDefineByte: Support is something we put a lot of work into, but I don't see what that ensures.
00:30:02 Join krazykt [0] (n=kkit@
00:30:10LloreanDefineByte: The old log will show a non-current build, at which point we immediately tell them to update.
00:30:24KalthareAnyhow, I've heard gcc has at least one optimization that's slightly random in its results.
00:30:26LloreanDefineByte: So, how exactly does "outdated" logs provide us anything but actually useful information?
00:30:36 Join jac0b [0] (
00:30:37 Quit krazykit (Nick collision from services.)
00:30:41 Nick krazykt is now known as krazykit (n=kkit@
00:31:05DefineByteall it would tell you is the build bootloader rev and the settings in a pre-formatted way. so the logs wouldn't have a date in them? or even a hash of the date? :)
00:31:25bluebrotherand then you un-hash the date?
00:31:30LloreanDefineByte: Um... build revision is more specific than date?
00:31:43jac0bnow I am getting another error building a rb environment on ubuntu
00:31:46jac0bmake[2]: sh-elf-ar: Command not found
00:31:46jac0bmake[2]: *** [libgcc.a] Error 127
00:31:46jac0bmake[2]: Leaving directory `/home/jacob/build/gcc/gcc'
00:31:46DBUGEnqueued KICK jac0b
00:31:46jac0bmake[1]: *** [stmp-multilib] Error 2
00:31:46jac0bmake[1]: Leaving directory `/home/jacob/build/gcc/gcc'
00:31:47***Alert Mode level 1
00:31:47jac0bmake: *** [all-gcc] Error 2
00:31:50bluebrotherwe use svn revision currently.
00:31:51LloreanDefineByte: Why the heck do we need a date if we have a build revision #?
00:31:58bluebrotherjac0b: please use a pastebin for multiline pastes
00:32:00Lloreanjac0b: Please don't spam the channel, see the channel guidelines.
00:32:07bluebrotherhave you build binutils first?
00:32:10Bagderjac0b: are you using ?
00:32:22 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
00:32:25jac0bbluebrother: yes
00:32:36bluebrotherjac0b: also installed and in PATH?
00:32:42bluebrotherDefineByte: yes.
00:32:42jac0bbagder: no
00:32:46LloreanDefineByte: Per-file what?
00:33:00Bagderjac0b: I'd suggest you do, it makes all that automatically for you
00:33:14DefineByteall i'm seeing this as is formatting important information in a user friendly way. that is all
00:33:18bluebrotherwhy don't you just check how it currently works? By storing the svn revision we can also detect (and clean up, but that's not implemented yet) old files
00:33:22DefineByteif it is too much work, fine
00:33:29LloreanDefineByte: Why do you refuse to answer questions.
00:33:47LloreanIt's not too much work to hash a date, it's 100% silly and pointless if the SVN revision is already included.
00:33:55jac0bdamn I should have did that from the start
00:34:00DefineBytewhere are you getting this SVN rev from?
00:34:05LloreanIt's A) Human readable and B) gives information as to the date and time the source code came from.
00:34:05DefineBytethe outdated logs?
00:34:06jac0bbadger: thanks
00:34:09LloreanDefineByte: Exactly.
00:34:15bluebrotherfrom the server. You want me to guess it?
00:34:32LloreanDefineByte: If the SVN revision is not "The current SVN revision" we stop and say "Use RBUtil to update your Rockbox installation"
00:34:36Bagderjac0b: if you move your already downloaded tarballs to the dir is set to use, it won't have to re-download them
00:34:37DefineByteand if the user updates a single file manually?
00:34:57bluebrotherthe average user *won't* do this.
00:34:57JdGordonhey all
00:35:03LloreanDefineByte: How do you update a file to "newer than the current version" without using a custom build?
00:35:06 Join SirFunk [0] (
00:35:08JdGordondid I miss anything interesting re e200r last night?
00:35:27LloreanDefineByte: If a user updates a file to a version other than the current version, they're lying to us when we ask if they're using an official version of Rockbox.
00:35:37BagderJdGordon: this perhaps =>
00:35:43amiconnrasher: Hmm, if the pool becomes rather big, it becomes somewhat slow, at least on windows
00:35:45LloreanDefineByte: Such a user could readily provide a false or edited log from RBUtil anyway, so it's pointless.
00:35:45CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
00:35:45*bluebrother just installed gnopernicus and it segfauls upon startup :/
00:35:50DefineByteand if the user updates his RBUtil installed RB to a custom build?
00:36:10LloreanDefineByte: If they do that, they could update it to provide falsified logs, as they know we don't support custom builds.
00:36:12DefineBytei'm really not seeing users forging support requests here
00:36:44LloreanDefineByte: If they're not falsifying support requests, then the question of "What if they modified a file" is irrelevant: We don't provide support on modified builds
00:37:06DefineByteunless they forgot they altered the file?
00:37:28 Quit miepchen^schlaf ("Verlassend")
00:37:30Bagderthe rev number tells
00:37:35DefineBytedo you rally see users with perfect recall in support requests?
00:37:43LloreanBagder: He's assuming they replace one file on their player, rather than a whole build
00:38:06Bagderusers can always mess up, no matter what we do
00:38:11LloreanDefineByte: In which case unless rbutil does an md5 sum of every official Rockbox file and compares, it's pointless anyway.
00:38:12DefineByteor hell, dleteing a file
00:38:35 Join LycoLoco_ [0] (
00:38:46Bagderbesides, trying to imagine use cases that users might do is pointless imo
00:38:46LloreanDefineByte: If a user deletes a file, we'll know from the error generated.
00:38:48DefineBytethat's what I'm suggesting, yeah
00:38:50DefineBytetoo far?
00:39:07JdGordonBagder: if thats the most recent news... did you miss that I got a stripped e2t version giogn which can load our own code onto the e200 and run it?
00:39:11bluebrotherwell, we could write a usb stack that hides the .rockbox folder except for rbutil.
00:39:13DefineByteand if he opens it in notepad to "fix" it?
00:39:15JdGordonI had a nice flashing wheel last nitght :)
00:39:16LloreanDefineByte: If a user modifies a single file, it's highly unlikely it won't work without updating the whole build, and even then, they're compiling for themselves: They have a clue.
00:39:18bluebrotherbut it's completely pointless
00:39:38BagderJdGordon: that I missed. very nice!
00:39:48LloreanDefineByte: Honestly, at this point you're being silly. A user who opens it in notepad to fix it is probably going to remember that he did that.
00:39:52DefineByteexactly, it won't work and then they come here
00:39:53bluebrotherand btw, missing files could easily detected from the log
00:40:10JdGordonBagder: wasnt really hard and linuxstb should take the credit :)
00:40:12DefineBytewhat log?
00:40:28DefineBytethe one generated at install time?
00:40:29bluebrotherah, and if a user replaces his build with a custom build rockbox-info.txt will tell us
00:40:47bluebrotheryes, that one. You just need to check if all files in the log are still present ...
00:40:52DefineByteso why can't rbutil gather that?
00:41:01amiconnrasher: Btw, if you need to quote some tts options, using double quotes won't work. Single quotes do work
00:41:02DefineByteor modified :)
00:41:18DefineBytecould be filesystem corruption for all we know
00:41:18amiconn(e.g. for sapi: '/voice:Microsoft Mary')
00:41:30bluebrotherhell could freeze as well.
00:41:37DefineByteor a bad USB connection
00:41:42LloreanDefineByte: Why do you think there's some phantom user who will modify or break a file by compiling a single file of his own, then forget he did it, and tell us his build is unmodified in the support request?
00:41:48DefineBytenow you're getting it :D
00:41:49***Alert Mode OFF
00:42:06SoapIf I may step in here for a moment.
00:42:20DefineByteand if the USB connextion is faulty when transfering files over 1MB?
00:42:25DefineBytewho knows :)
00:42:44DefineBytewill rbutil verify the installation?
00:42:51bluebrotherand if hell freezes while the user downloads Rockbox?
00:42:57LloreanDefineByte: Consider this: I've been doing support for Rockbox for 2 and a half years, and have read every post in the forum during that period. I think your idea is unnecessary.
00:43:10SoapIn your scenario, DefineByte, as described by Llorean... So the user does just that - uses a modified build w/o knowing. (S)he will be told to update to the latest version - thus overwriting the covertly modified file.
00:43:26bluebrotherand how should "verifying" be done?
00:43:39DefineBytemd5? )_)
00:43:40LloreanSoap: He's hypothesizing a user that's running the current SVN version of Rockbox, then immediately modifies a file, and then forgets, all within the average of 2.5 hours between revisions
00:43:41bluebrotherwithout great speed penalties and stuff?
00:43:52bluebrotherwell, here you have the penalty.
00:43:57DefineBytei'm just playing with the idea : rally :D
00:44:19*bluebrother starts considering this discussion and "idea" pointless and shuts up
00:44:38bluebrotherwe are writing software for users, not idiots.
00:44:40DefineBytehey, no skin xD
00:44:50DefineBytehave you been in the forums?
00:45:07bluebrotheryes. Not even only the Rockbox ones.
00:45:14JdGordonbluebrother: your forgetting about all the ipod users...
00:45:16LloreanDefineByte: If the whole statistically likely ONE user this happens to in the next four years shows up, I'll happily take his support request for the benefit of enjoying the fact that users don't have to download RBUtil to provide me the necessary information to give support.
00:45:57DefineBytethey already need rbutil to install and presumably update with a fixed build
00:45:58bluebrotherJdGordon: maybe a bit ... but those will use our official install method (i.e. rbutil at some point). And everything else (i.e. those fancy custom builds) are unsupported :)
00:46:00SoapLlorean: On another topic. BigMac's interest in music metadata triggering .cfg file changes reminded me of the older feature request to allow/create a hybrid playlist/config. Either by allowing config "commands" in playlists (as .m3u comments?) or allowing a config file to "launch" a playlist. (or vise versa). Why, if you recall, was this idea dismissed?
00:46:47LloreanDefineByte: We've covered this: RBUtil may not be usable at the time, while downloading an extracting an SVN build might be.
00:46:52bluebrothertoo much possible confusion?
00:47:11BagderSoap: a playlist can hardly launch anything since it is just a list of files and they can be played in shuffled order
00:47:22Bagderthe config could be based on dirs or files though
00:47:34LloreanI like the idea of allowing playlists to be launched by .cfg files
00:47:36LloreanA LOT
00:47:43DefineByteright, so you're getting users to manually update. i thought they wouldn't do this?
00:47:43Bagderbut I don't think we've seen any decent implementation that doesn't spin up the disk on every track change
00:47:45LloreanBut that's just me.
00:47:54LloreanDefineByte: What do you mean they wouldn't do this?
00:48:32SoapBagder: I grok that - I ment the .m3u parser reads specially formatted comment lines directing it to a .cfg file to be used.
00:48:45LloreanSoap: I like the idea of .cfg files being able to launch playlists, and playlists being able to include .cfg files (though absolutely only via the "queue" function, and never in a saved one)
00:48:54DefineByteyou said the average user would do this then you're encouraging them to do it. user enters forum , sees updating manually to be the mrthod for fixing problrms, they update manually, error!
00:49:04LloreanSo you could queue a .cfg that changes the settings and launches a new playlist.
00:49:16BagderI would prefer having cfg files that are dir-specific, song-specific or even metadata-specific
00:49:27SoapThat would allow BigMac to have a different EQ per album without changing the original tracks, etc.
00:49:30DefineByteusers will read other's posts and try to fix the problem themselvs before posing about it
00:49:30Bagderso that all my 422 playlists would use them
00:49:39LloreanDefineByte: Then users will do all kinds of crazy things
00:49:40 Quit freqmod (Read error: 104 (Connection reset by peer))
00:49:50DefineByteand there's te rub
00:49:51 Quit ze (Read error: 104 (Connection reset by peer))
00:49:57LloreanDefineByte: But every post will include "Use RBUtil to update" "I can't install RButil here" "Okay, please manually install by extracting..."
00:50:06DefineBytehow do you recognise a crazy user when they make a support requst?
00:50:31LloreanDefineByte: So, are you suggesting we deny support request to people who can't use RBUtil? Because either your method or mine both require manual installs...
00:50:35LloreanSO I don't see why you even bring this up
00:50:39SoapBagder: You mean as in a .cfg file which says "When playing from directory X use settings set 1 and when playing from directory Y use settings set 2"?
00:51:01 Nick BigBambi is now known as SherlockHolmes (n=alex@
00:51:04***Saving seen data "./dancer.seen"
00:51:04BagderSoap: yes, probably like "dir.cfg" places in the directory, always found and used if present
00:51:13DefineByteI'm suggesting the preferred method. that is all :)
00:51:15bluebrotherwell, you don't need to install rbutil. It's a single binary.
00:51:17 Nick SherlockHolmes is now known as BigBambi (n=alex@
00:51:27bluebrotherbut it might not work for everyone.
00:51:40LloreanDefineByte: Then arguing against my method with a point that's equally true for your method is silly. Next challenge?
00:51:50Bagderand probably a [song].cfg for a cfg for a particular song
00:52:04DefineBytewhat's is your method again?
00:52:13Bagderfor metadata-specific, it would require a special dir for those configs
00:52:17LloreanDefineByte: Not requiring RBUtil.
00:52:25SoapI see, Bagder. Thus the problem of spinning up the drive every track change as a relevant .cfg file is needed to be looked for?
00:52:42Bagderthe cfg files would be checked for, and loaded when the songs are loaded
00:52:55DefineByterequire? nay, just prefer :)
00:53:05SoapI was getting to that.
00:53:17LloreanDefineByte: I consider making it a part of the formal process a "requirement" for formal support.
00:53:27DefineByteif it's outside the scope of RBUtil that's fine :)
00:53:54DefineByteformal procedures can't have optional steps? blimey
00:54:02LloreanDefineByte: As I've said, I will fight tooth and nail to have the formal process involve the use of any program to make the request.
00:54:15LloreanDefineByte: If it's an optional step, why not just leave it out?
00:54:31DefineBytebecause it's easier for the user
00:55:17DefineBytenatually, only if they alrady have rbutil
00:55:23 Quit LycoLoco (Connection timed out)
00:55:27DefineBytewhich they should :D
00:55:33 Join ze [0] (
00:56:03LloreanDefineByte: Running a program, then attaching one file located somewhere on your computer, vs attaching two files located somewhere on your computer are about equally difficult I'd say
00:56:07 Join Xero [0] (
00:56:18XeroDoes Rockbox work with iPod Shuffle?
00:56:19LloreanI'd rather have one set process.
00:56:23XeroOr must you have a screen?
00:56:37LloreanXero: It doesn't require a screen, but the iPod Shuffle is not supported. Nobody has ported to it.
00:56:39DefineBytebut the user would knoe where the one file is but not the two
00:56:44Xero:( :( :(
00:56:47LloreanDefineByte: How do they *not* know where the two are?
00:56:53XeroSadness are upon me.
00:57:05SoapXero: I'll sell you a Nano.
00:57:07BagderXero: sell it and buy a unit rockbox runs on... :-)
00:57:19XeroSoap: How much for a Nano?
00:57:23LloreanDefineByte: Are they magically going to psychically know where the one from RButil is? Or will they have to read it from the screen somewhere? Is it harder for them to read the folder name on their player?
00:57:33 Join RudMan_ [0] (
00:58:01DefineByterbutil can't tell tem? i thought c++ was turing complete (or what ever the language in use is)/
00:58:43LloreanDefineByte: Yes, so they read it on the screen somewhere.
00:58:58LloreanDefineByte: And the place where they find out what the formal support process is, *that* can't tell them where the files are on the device?
00:59:22 Quit RudMan_ (Read error: 104 (Connection reset by peer))
00:59:32 Join RudMan_ [0] (
00:59:38DefineBytei see, you're thinking of them actually reading your guidlines. surely your experiance tells you otherwise?
00:59:58LloreanDefineByte: If they don't read the guidelines, how do they know to even *run* RBUtil to create the ticket?
01:00:05DefineBytethey post request, you say run rbutil, they attach file done.
01:00:13 Quit perplexity (Read error: 113 (No route to host))
01:00:18LloreanThey post request, I link instructions, they attach files, done.
01:00:21SoapI'll pay $10 to the first person, (DefineByte or Llorean) to stop this conversation.
01:00:27LloreanAlright, I'm done.
01:00:37*pixelma gets bored
01:00:45Soapyou want that check, cash, or donation in your name?
01:00:47DefineBytethey read instructions? riiiight. :p
01:00:47 Quit jac0b (Read error: 110 (Connection timed out))
01:01:14LloreanSoap: I'll put that $10 up to definebyte if he spends one year reading every post in our forums and acclimatizing himself with our userbase. Otherwise, keep it, or donate it, your choice. :)
01:01:23DefineBytei guess you could send the leather bound version through the mail
01:01:25 Quit mikeT ("-")
01:01:39scorcheSoap: you are always in here flaunting your money around! ;)
01:02:13DefineBytei know you like to think they read them but...they really don't.
01:02:30DefineBytemost people see the length of te page and skip it
01:03:03scorcheso we should adjust to the lowest common denominator?
01:03:10scorchethis isnt the US school system...
01:03:15DefineBytewhere as one sentence in a post specifically directed at them they're more liekly to heed
01:03:49DefineBytethe idea is to shiled you guys from them :)
01:03:58LloreanDefineByte: I repeat, I've spent two years dealing with our userbase, I know how much they do or don't read. Much better than you have any idea, I assume. It's better to force them to read, as they'll have to at some point anyway.
01:04:14SoapI just made the donation and you start back up.
01:04:21LloreanSoap: Sorry, that's my last comment on it.
01:04:40DefineByteforce? how's that work?
01:05:04DefineBytei guess you could put some hidden works in their and require them and registration
01:05:13LloreanDefineByte: That would be off topic for this channel, it's about Rockbox not about user herding, I humbly request you read the guidelines.
01:05:49scorcheSoap: you might want to make a donation to the profanity jar too =P
01:05:51DefineBytei've read them but they're a bit long to digest
01:06:55DefineBytei'm just following the leader :D
01:07:03LloreanDefineByte: If you don't follow them you'll be asked to leave, and eventually banned. Same goes for the forums, and in both we require the manual being read.
01:07:07chrisjs169one of the leaders is Llorean
01:07:10LloreanI have no problem with not coddling users who refuse to read the manual. :)
01:07:15scorcheDefineByte: ever play lemmings...?
01:07:31LloreanDefineByte: It was a subtle hint as to how we force them to read it: They get no help until they've showed that they have.
01:07:38 Quit JdGordon ("Konversation terminated!")
01:07:41DefineByteyeah, i loved that game. is there a timer on me? ^^
01:07:43 Join Merthsoft [0] (n=Shaun@
01:07:45*chrisjs169 notices Llorean isn't on the access list
01:07:51Lloreanchrisjs169: Nope.
01:07:54LloreanGood thing too.
01:07:54scorcheDefineByte: you miss my point...
01:08:05MerthsoftLlorean: Using sunday's version fixed the database error, btw
01:08:15chrisjs169Llorean, haha
01:08:41LloreanMerthsoft: Good to hear, it looks like there was a bug added in the ID3 handling, or so I hear.
01:08:41DefineBytei didn't think it was important scorche.
01:08:58MerthsoftAh, was it fixed do you know?
01:09:03DefineBytei thought we were talking about rbutil anyway. my mistake
01:09:06scorcheDefineByte: the point is, lemmings always follow the leader...right off cliffs to their death
01:09:08 Quit ompaul ("night night etc")
01:09:18DefineByteno they don't :p
01:09:22LloreanMerthsoft: I believe so.
01:09:48DefineBytejust use a blocker or something
01:10:04pixelmaMerthsoft: not that I know of
01:10:05LloreanMerthsoft: I may be wrong.
01:12:26 Join jhulst [0] (
01:13:40 Quit bluebrother ("leaving")
01:13:46DefineByte"DefineByte: the point is, lemmings always follow the leader...right off cliffs to their death"oh, i get it. you're saying llorean is trying to rile me so i get myself banned. heh
01:14:12chrisjs169notice how nobody here is even op'd..
01:14:42DefineByteif is getting off topic now you guys. tsk
01:15:00chrisjs169so is your quite
01:15:30scorchehow about lets both just hush and wait for a new topic...
01:15:36amiconnFunny... I can now build voice files for all my targets in less time (for 1 language) that a rockbox build for a single target...
01:15:42DefineBytehey, i didn't say i'm a pillar of virtue
01:15:56scorcheDefineByte: hush about the issue
01:15:59Lloreanamiconn: I always could on Linux
01:16:07Lloreanamiconn: Oh, *all* of them at once.
01:16:16DefineBytei'm trying to think of a new topic now
01:16:25scorchethen do it silently
01:16:27LloreanI could build five with use of the pool in less time, on a first run, though
01:17:45DefineByteokay, what could be a reason for a plugin to fail, supposing that there are no coding errors in the plugin itself
01:18:00DefineByteto compile/
01:18:16LloreanDefineByte: If Rockbox compiles fine without the changes, and fails to compile with the changes, there are coding errors.
01:18:35DefineByterockbox compiles just fine
01:18:46DefineByteit's just the plugin that doesn't work
01:18:49 Quit obo ("bye")
01:19:10chrisjs169DefineByte, what plugin
01:19:41DefineBytebasically if i modify the disktidy plugin at all (:p) the plugin compiles fine but won't run and is unusually small
01:20:24 Quit atsea-68 (Remote closed the connection)
01:20:55DefineBytedo any other files need to be modified first?
01:21:12 Join Febs [0] (
01:21:36 Join BHSPitLappy [0] (n=steve-o@
01:22:47LloreanDefineByte: If there's just a name.c file for the plugin, it's rare that you'll need to modify other files unless you're changing the plugin api.
01:23:42 Join atsea-68 [0] (i=atsea-@gateway/tor/x-0708f353f97868eb)
01:23:45DefineByteI'm just trying to modify the disktidy plugin to clean ipod files
01:23:48 Quit atsea-68 (SendQ exceeded)
01:24:27 Join atsea-68 [0] (i=atsea-@gateway/tor/x-1d578ed7e3be62c6)
01:24:53 Join annulus_ [0] (
01:25:37XeroSendQ too low.
01:26:02DefineBytejust to try i modified the TIDY_MAC function to delete iPod_Config instead of .DS_Store and it compiles find but when i try to run it it errors
01:26:39 Join kugel [0] (
01:29:28Soapisn't the folder name ipod_control?
01:29:40SoapI don't have either of my iPods in front of me at this time, or I would check.
01:29:41*XavierGr sighs
01:29:56DefineByteyeah it is, i can assure you the code had the right name :D
01:30:00XavierGranother review of rockbox that stresses the uglyness of the defualt gui
01:31:47DefineByteah the bain of OS.
01:34:34DefineBytewell, i'm stating to fall asleep at the wheel here (starting?). it was fun chatting with oyu. night night.
01:34:38 Part DefineByte
01:35:23KalthareHm. I take it rbutil doesn't support custom builds?
01:36:54 Join DarkStylist [0] (
01:37:18DarkStylisthey guys
01:37:40iambenhey dude
01:37:43DarkStylistanything like rockbox for my technaxx musicman easy player?
01:38:23DarkStylistits one of those neat usb mp3 systems with sd cards
01:38:49 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
01:38:56 Quit _jhMikeS_ (Nick collision from services.)
01:39:00 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
01:40:18 Join austriancoder [0] (n=austrian@rockbox/developer/austriancoder)
01:40:43 Quit spiorf (Remote closed the connection)
01:41:08DarkStylistand 1 aaa battery
01:41:43xDragonforce16xdont think soo....
01:42:06xDragonforce16xim just a geust here so dont take my opinion
01:43:23Soapnot to be dismissive, DarkStylist, but Rockbox doesn't run on it. This being a Rockbox channel there is little chance of learning about things outside Rockbox here. Your best chance is simply to google the player's name/model and perhaps the word "firmware" to see what pops up. That being said, Rockbox is one of a VERY few MP3 player replacement firmwares.
01:43:45BigMacOk, I am planning on working on the idea that was being discussed earlier of being able to load playlists from config files or Bagder's ideas, but I would like to get an opinion on what the developers think the best implementation would be
01:44:06 Quit kugel ("ChatZilla [Firefox]")
01:45:16chrisjs169austriancoder is on :D
01:47:03 Join Rob222241 [0] (
01:47:52 Quit Rob2222 (Read error: 110 (Connection timed out))
01:48:07DarkStylisti noticed there are quite a few usbstick players with same firmware
01:48:21DarkStylistalways same Gui
01:48:39DarkStylisti just never learned the firmware name
01:48:56DarkStylistand googeling it helps not
01:49:37DarkStylisthave any idea of what firmware it is? it always starts with Welcome
01:50:40 Quit jhMikeS (Nick collision from services.)
01:50:46 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
01:50:54BigBambiDarkStylist: Not here please
01:52:39*chrisjs169 wonders how many players start with Welcome
01:52:55*xDragonforce16x has no idea
01:53:41xDragonforce16xalmost all of them, the think the player has to be polite to its user?
01:55:18xDragonforce16xpeace out everybody
01:55:28DarkStylisthowabout these chinese mp4 players?
01:55:33 Join perplexity [0] (
01:55:37 Part xDragonforce16x
01:55:49DarkStylisti got a noname one saying ipod on it from china
01:56:07LloreanDarkStylist: Rockbox only runs on the players explicitly mentioned on the front page of the site.
01:56:27DarkStylistyea but
01:56:50DarkStylistwhat if my model is based on one off of the site?
01:57:32LloreanIf your player has the exact same hardware internally, set up in the exact same way, then in theory it would run. Looking like the same player, though, does not make it the same player.
01:58:22iambenLlorean: yeah but they printed "ipod" on it!
01:59:01LloreanAFAIK, a lot of the knockoffs use the s1mp3 hardware.
01:59:41 Quit austriancoder (Remote closed the connection)
02:00:21 Part Merthsoft ("Leaving")
02:04:07DarkStylistyea s1mp3 os thats it
02:05:09 Part pixelma
02:05:53 Quit Rob222241 (Connection timed out)
02:06:05 Join BrickedE200R [0] (i=82c20d6a@gateway/web/cgi-irc/
02:06:18BrickedE200R'i broke my ipod
02:06:22DarkStylistthat bad?
02:07:05BrickedE200Ri followed the instructions and used e200tool on my nano 2g and now it wont turn on
02:07:47 Nick BrickedE200R is now known as JdGordon (i=82c20d6a@gateway/web/cgi-irc/
02:07:57JdGordonlinuxstb: you still round?
02:08:19*Llorean wonders if there's a way to trigger manufacturing mode on iPods.
02:08:48hcsthe emergency disk mode?
02:08:54JdGordonlooks like there is a limit to how much code we can copy onto the snasa with e2t :(
02:09:29Lloreanhcs: No, on the e200 there's "manufacturing mode" which is much lower level than that.
02:09:37amiconnhi JdGordon, I have a bug for you (in the settings system)
02:10:22amiconnCancelling live settings doesn't work: it displays 'Cancelled' but does actually accept the setting.
02:11:24JdGordoncan you stick it on flyspray? im at uni and giong to forget about it otherwise....
02:17:13JdGordonwhats the easiest way to add to binary size without adding code?
02:17:19JdGordonwill a const array do it?
02:17:36 Join kkurbjun [0] (
02:17:51LloreanAdd to binary size?
02:18:21JdGordontrying to figure out the largest binary we can send to the sansa
02:19:01amiconnlinuxstb: You broke usb on the player
02:19:11amiconnLooks like the logic is reversed
02:19:53amiconnAh, yes:
02:21:42JdGordonamiconn: any ideas how to add bin size easily?
02:21:57JdGordongc is optimizing my unused arrays out
02:23:44JdGordonnup :(
02:23:59LloreanWhy not just use the first element of the array for something?
02:24:07 Join My_Sic [0] (
02:27:54 Join Deibeed [0] (i=63f694d3@gateway/web/cgi-irc/
02:28:21DeibeedHey guys?
02:29:49LloreanIt's often best just to ask a question, if you have one.
02:30:02DeibeedI was just trying out Rockbox on my Ipod Nano 1st Generation, and I got an error while uploading database.
02:30:18LloreanThere is a known problem with creating the database right now.
02:30:25DeibeedIt says: Data abort at 000457E4
02:31:04DeibeedSo, I simply can't upload music onto it?
02:31:23krazykityou can, the database just won't work right now.
02:31:49DeibeedSo how do I do that?
02:31:55krazykitdrag and drop.
02:32:00 Part Llorean
02:32:07 Join Rob2222 [0] (
02:32:19krazykitif you use iTunes to sync normally, however, the original firmware will NOT be able to see the stuff you've simply dragged and dropped on the player
02:32:42KalthareJdGordon: volatile, maybe?
02:33:00JdGordonchanged my tact.. and it seems to be working now
02:33:11*JdGordon giong to break his e200 usb port ...
02:33:18DeibeedHmm. I don't use the iTunes to sync my ipod. Is there some other way?
02:33:35krazykituh, how do you normally put music on your ipod?
02:33:51DeibeedImport to Library, then drag into the Ipod itself.
02:34:13JdGordonjhMikeS: hey, yuo round?
02:34:23krazykitwhich is using itunes to sync the music...
02:34:31DeibeedSo, do it again?
02:35:14DeibeedSo what do I do?
02:35:21krazykitthe way you do it, both the apple firmware and rockbox can use the files. drag-and-drop from explorer or finder will allow rockbox to see it but not the original firmware
02:35:27 Quit perplexity (Read error: 113 (No route to host))
02:35:30 Quit qweru (Read error: 110 (Connection timed out))
02:35:44krazykitDeibeed, you could downgrade to a previous build that doesn't display this problem.
02:35:55DeibeedAll right. Thanks a lot.
02:36:01krazykitthen, you could simply wait til the problem is resolved
02:37:37jhMikeSJdGordon: Hey, yes. I've even been around but I don't talk much when doing complicated work.
02:37:51amiconnJdGordon: I doubt that you will break the port (assuming it's a standard mini-B port)
02:38:28amiconnBack when doing the Ondio port, I plugged/unplugged loadsa times for several weeks
02:38:35JdGordonjhMikeS: how much of the sansa ata driver can be stripped out? we need a working ata driver for the e200r installer, but it seems we are very limited with the aount of code e200tool can send across...
02:38:39JdGordonamiconn: its nto standard...
02:38:54 Join skate3214 [0] (
02:39:14jhMikeSJdGordon: does it require any SD support?
02:39:20amiconnJdGordon: I guess e200tool loads the stuff into iram, probably minus some used area
02:39:38JdGordonjhMikeS: no, just need enough to write to the internal disk
02:39:44jhMikeSwrite only?
02:39:51JdGordon... and read
02:40:11jhMikeSok...lemme have a read over
02:41:10jhMikeSyou don't need the fancy fast read function so the slow version could be used for one
02:41:19amiconnAny hints how to reproduce the crash in the db init?
02:41:42amiconnI tried both on mini G2 and video (G5.5/80), but got no crash (latest svn)
02:42:12 Join AceNik [0] (n=AceNik@
02:42:35jhMikeSJdGordon: I suppose I could just pastebin a stripped-down verions. it shouldn't be much trouble.
02:43:10AceNikguys since svn 146XXX i get data abort everytime i go to initialise database on my h10 [20GB] please help
02:43:22 Quit hcs (Read error: 110 (Connection timed out))
02:43:24JdGordondont worry about it now.. just wanted to know if a stripped one could be dnoe
02:43:27amiconnJdGordon: Did you find the upper limit for the transferable size?
02:43:36JdGordonnot yet
02:44:03amiconnI wonder what e200tool does... I would guess we'd have at least half the iram available
02:44:08amiconn(that'd be 64KB)
02:44:37AceNikamiconn are the latest builds working perfet on ur h10
02:44:54 Quit skate3214 ("ChatZilla [Firefox]")
02:45:08amiconnDepends on the definition of "perfect"
02:45:48JdGordon64K is too large...
02:46:03JdGordonalthough, i thought 10k was too big before also :p
02:46:15 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
02:47:00AceNikdoes database initialise & also d songs play instantly or there are hang up sometimes?
02:47:25JdGordon65535, 65536 doesnt
02:47:35JdGordonso maybe we dont need to be careful with the size
02:48:21AceNiki get data abort everytime i initialise database, & also everytime i start my player with the earlier builds, somewhr cloase to 14577 or so i had my player hang on "loading"
02:48:39amiconnDB init works
02:49:02JdGordonsending a 28118 binary doesnt work... but sending 1100bytes followed by 30k of random data does...
02:49:10AceNikamiconn: then why do i get these data aborts , has the bootloader changed?
02:49:16amiconnWhether it would hang occasionally I don't know. The H10 is not my main target
02:49:30amiconnIt didn't hang in a quick check
02:50:02AceNikamiconn: you using a 20 GB right
02:50:09amiconnnope, 6GB
02:50:43amiconnNone of my 4 PP targets crashes on db init
02:50:46AceNikamiconn: then how do i confirm with someone's 20GB , there is a problem this is definate,
02:51:06***Saving seen data "./dancer.seen"
02:51:15amiconnI don't want confirmation (there already are multiple reports), I want a way to reproduce...
02:51:48AceNiksecondly the usb detection patch was added today, when i connect my usb it shows nothin on the screen of my player , pressing the power button boots into rockbox, disk mode is no where?
02:51:48 Quit JdGordon ("CGI:IRC (EOF)")
02:52:37AceNikamiconn: confirm, data init for the first time, gives data abort at 045XXX
02:52:53amiconnThat doesn't help me
02:52:58AceNikhas the h10 bootloader changed
02:53:19amiconnNot that I know of
02:53:48AceNikwell, ill just wait then
02:54:08amiconnWhat svn revision are you running?
02:54:16AceNikhow does the usb detection thin work, just plug the usb & let it go to disk mode or any button combos?
02:54:32AceNik14604-14614 all the same
02:55:02amiconnYeah, no relevant change concerning this problem
02:55:24AceNik14577 was better , DB init worked , but at random times, while startin up player & playing a track caused the player to hang at "loading..."
02:55:30 Quit Rob2222 (Success)
02:56:00amiconnI wonder what's different here then
02:56:12AceNikno clue
02:56:47amiconnDB init works like a charm (with r14607) on 2nd gen, mini 2nd gen, video 5.5th gen, and iriver H10/6GB
02:57:14AceNikwell, im clearly stuck up with something, no clue why?
02:57:15amiconn3 of those have dircache enabled, one does not (the mini)
02:57:17 Part Deibeed
02:57:42krazykiti'm building r14615 now for sansa, i'll see if i get a data abort too
02:58:05AceNikk thanks a lot, ill take a kip & return c ya
02:59:03 Quit AceNik ("bye guys have fun , enjoy !!!!!")
03:00:07krazykitno, no problem with db init on the sansa either.
03:00:07jhMikeSJdGordon: quicky strip-down. I didn't pull the SDHC cases but I'm sure those could go too.
03:00:30jhMikeSblah, he disappeared
03:05:56 Join bagawk_ [0] (n=lee@unaffiliated/bagawk)
03:08:05 Join sambolina3 [0] (i=48495f23@gateway/web/cgi-irc/
03:08:53 Quit cmvjk (Read error: 110 (Connection timed out))
03:11:06 Quit atsea-68 (Read error: 104 (Connection reset by peer))
03:14:26 Quit sambolina3 ("CGI:IRC (EOF)")
03:16:02 Quit RoC_MasterMind (Read error: 110 (Connection timed out))
03:16:19 Join RoC_MasterMind [0] (
03:16:34 Quit bagawk (Read error: 110 (Connection timed out))
03:22:04 Quit My_Sic (Connection reset by peer)
03:23:16 Join My_Sic [0] (
03:42:01 Quit jhulst (Remote closed the connection)
03:42:05 Quit kkurbjun (Read error: 110 (Connection timed out))
03:55:35 Part DarkStylist
03:57:33 Quit donutman25 ("ChatZilla [Firefox]")
04:02:10 Quit advcomp2019 ("Leaving")
04:06:04 Quit hcs ("Leaving.")
04:07:19 Join advcomp2019 [0] (n=advcomp2@
04:09:13 Quit perrikwp ("CGI:IRC (EOF)")
04:14:09 Join perrikwp [0] (i=9821616e@gateway/web/cgi-irc/
04:22:35 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
04:23:46 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
04:27:19 Join jhulst [0] (
04:30:31 Join gunner [0] (i=446d1c6a@gateway/web/cgi-irc/
04:34:21 Quit Kalthare ("Leaving.")
04:36:35gunnersome1 mind helpin me with a problem on an ipod nano?
04:36:57Soapask the question, please don't ask to ask.
04:37:24Soapthough such shorthand as "some1" is severly frowned upon.
04:37:35gunnerwell i put rockbox on it and it froze during 1st song playback and the reset wont work...
04:37:37Soapsee the channel topic.
04:37:48SoapOn the iPods reset always works.
04:38:03gunnersorry i just need to type quick
04:38:44gunnerits stuck with back light on and minimal information on the screen
04:39:02SoapI don't mean to be blunt, but reset will always work. Toggle the hold switch to the on (red) and back to the off positions.
04:39:30Soapthen Menu + Select, and hold those buttons (without a wiggle in your fingers) for 30 seconds, and it will reset.
04:41:43gunnerstill nothing and i know its supposed to but it wont
04:42:47gunnerwell thank you but i have to go now
04:43:09 Quit gunner ("CGI:IRC (EOF)")
04:44:24 Join puzzlet [0] (n=puzzlet@
04:51:08***Saving seen data "./dancer.seen"
05:05:00 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
05:06:07 Quit bb (Nick collision from services.)
05:06:12 Join bb_ [0] (n=bb@unaffiliated/bb)
05:11:18 Quit RudMan_ (Remote closed the connection)
05:15:22 Quit RoC_MasterMind ("Leaving")
05:21:17 Join RudMan [0] (
05:23:42 Quit perrikwp ("CGI:IRC (EOF)")
05:27:40 Join Llorean [0] (
05:35:31 Join Hellmark [0] (n=hellmark@
05:36:10 Quit Llorean (Remote closed the connection)
05:36:25 Join Llorean [0] (
05:36:50 Quit Llorean (Remote closed the connection)
05:37:04 Join Llorean [0] (
05:38:57 Join atsea-68 [0] (i=atsea-@gateway/tor/x-9403cb6fa024494b)
05:39:47 Quit atsea-68 (SendQ exceeded)
05:43:18 Quit BigMac ("Konversation terminated!")
05:43:27HellmarkQuick question. I am looking at getting a player that supports Rockbox, and two of the ones I am looking at are the iPod Mini and the 5-6gb iRiver H10. Reason for such, is that the microdrives can be swapped out for compact flash, which would help with battery life. On the Mini, under rockbox, you have to patch it if you run a compact flash card, due to it crashing when trying to put the drive into sleep mode. I am curious if this is an issue on the H1
05:43:29Hellmark0 as well
05:47:42LloreanYes, I'm fairly confident it would be, though I don't believe anyone has tried
05:50:52 Join Kalthare [0] (
05:51:20advcomp2019Hellmark, are you looking for a digital audio player that is flash based that can have rockbox on it?
05:52:16 Nick fxb is now known as fxb__ (
05:56:44Hellmarkadvcomp2019, I mostly want one that I can rebuild if needed.
05:57:03Hellmarkthats why I like the mini. pretty much everything is replaceable and easy to get parts for
05:57:12 Join webguest64 [0] (i=4679eaed@gateway/web/cgi-irc/
05:57:24Hellmarkright now i have a Creative Zen NX, screen broke, and I am SOL
05:57:33Hellmarkdoesn't have to be flash based
05:58:04webguest64Found a bug in Gigabeat F latest build. when viewing credits the the player shuts down
05:59:06Hellmarkalso, don't really need any frills, like video and what not.
05:59:56KalthareSay, uh.. any chance someone could give me an executive overview of how Rockbox determines the audio length of an MP3 file?
06:00:20krazykitHellmark, well, most of the targets support video anyway, but you don't have to use it.
06:00:33Hellmarkkrazykit, I know
06:00:50Hellmarkbut if I can forgo that option and save a few bucks, why not
06:00:56krazykithave you considered the sansa at all? i'm not sure how easy it is to get parts, though
06:01:09advcomp2019look at the e200 is a nice player that rockbox works on
06:01:14 Quit Febs (Read error: 110 (Connection timed out))
06:02:00KalthareI've got some MP3s that most software, including Rockbox, vastly overestimates the length of.
06:02:14LloreanKalthare: This usually means bad information in the header
06:02:23 Join cmvjk [0] (
06:02:37Hellmarkkrazykit, from what I can tell, no real replacement of parts.
06:03:10krazykiti mean, i'm sure you can get refurbs and broken ones for dirt cheap on ebay
06:03:14KalthareLlorean: But isn't an MP3 header, like, two bytes long? I don't think there's much in that that could be gotten wrong without breaking playback...
06:03:28advcomp2019yea krazykit
06:04:03webguest64Llorean. when viewing credits in my Gigabeat F.. the player shuts down ..latest build
06:05:53LloreanKalthare: For VBR files, you need another header to be able to get right length. For CBR if you simply know the bitrate and file length, you know.
06:06:16krazykitwebguest64, when does it shut down? i'm displaying the credits just fine here. also, what revision, specifically?
06:06:52KalthareLlorean: Ah, okay, that's probably exactly it, then. Thank you. ...Know of any programs that fix broken VBR headers?
06:07:28LloreanKalthare: Try the "vbrfix" plugin in Rockbox
06:07:52KalthareLlorean: Thanks, I'll try that :)
06:08:09Hellmarkkrazykit, from what I've seen of disassembly photos, most stuff seems to be soldered on
06:09:08 Quit webguest64 ("CGI:IRC (Ping timeout)")
06:09:56advcomp2019Hellmark, here is a site that will fix it for you −−>
06:10:28 Join webguest64 [0] (i=4679eaed@gateway/web/cgi-irc/
06:10:47advcomp2019this monitor is making me see things
06:11:09krazykitwebguest64, what revision are you using? i can run the credits without error on my gigabeat.
06:11:22Hellmarkadvcomp2019, prefer fixing things myself if I can
06:11:32webguest64sept 04 /07 build gives the credits shutdown issue to gigabeat F
06:12:10krazykitwebguest64, look in system -> rockbox info and tell us the r number
06:13:03advcomp2019ok Hellmark
06:13:27 Join atsea-68 [0] (i=atsea-@gateway/tor/x-461923b7290d79e1)
06:14:59webguest64ok now its not doing it.....
06:16:36krazykitthat's not the latest, you're 14 revisions behind ;-)
06:16:37 Nick bagawk_ is now known as bagawk (n=lee@unaffiliated/bagawk)
06:16:39KalthareLlorean: That fixed it perfectly. Thank you again!
06:17:11webguest64have to push power button lightly though to get back to menu
06:17:49 Join bluey- [0] (
06:18:48webguest64how can that be if I just download latest build this morning\
06:19:11Lloreanwebguest64: Because there are constant updates?
06:23:19 Quit webguest64 ("CGI:IRC (EOF)")
06:29:15 Quit idnar (Nick collision from services.)
06:29:18 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
06:30:39 Quit RudMan (Remote closed the connection)
06:32:05 Quit bluey- ("Leaving")
06:32:21 Join ptw419 [0] (
06:48:35 Join Jangari [0] (
06:51:12***Saving seen data "./dancer.seen"
06:57:21 Quit Llorean (Remote closed the connection)
06:57:35 Join Llorean [0] (
07:01:33 Quit jhulst ("Konversation terminated!")
07:04:14 Part rvvs89 ("Cycling channel")
07:07:10 Join rvvs89 [0] (n=rvvs89@unaffiliated/rvvs89)
07:08:02 Join grndslm [0] (
07:08:10 Join grndslm_ [0] (
07:08:12 Quit grndslm_ (Client Quit)
07:20:25 Join davina [0] (
07:28:06 Join eigma [0] (
07:37:41 Join darkless [0] (
07:38:41 Quit Llorean (Remote closed the connection)
07:38:56 Join Llorean [0] (
07:39:29 Quit Llorean (Remote closed the connection)
07:39:45 Join Llorean [0] (
07:39:56 Join AceNik [0] (n=AceNik@
07:41:59AceNikguys bug remains still on the h10[20GB] data abort at 0045XXX after evert data init, this can be observed after svn revision no.14577 , thats the last one i was using, after which i switched to 14614
07:43:49ptw419anyone have any tips on how to get svn merge to work w/ the gigabeat-s branch?
07:44:22LloreanAceNik: Yes, you'll see there haven't been any commits relating to metadata since the bug was identified...
07:44:41amiconnThe bug isn't identified yet
07:45:01 Quit BHSPitLappy (Read error: 110 (Connection timed out))
07:45:04Lloreanamiconn: I thought it was identified... someone said something about and ID3 change of some sort causing it?
07:45:08AceNikllorean:i guess would the id3.c help in anyway, cause i suspect that file only has changed since then ?
07:45:09amiconnIf someone could tell me how to reproduce... it doesn't happen on any of my arm based targets
07:45:13ptw419so far i have : svn merge
07:45:20AceNikya thats what made me suspect it
07:45:39Lloreanamiconn: For most people, it's just "initialize the database and it happens after anywhere from <50 to >1000 songs"
07:45:41AceNikamiconn: somehow please get it to be tried on a 20GB
07:45:48LloreanAceNik: Please don't pester.
07:46:34amiconnAceNik: It is reported on ipod video as well. And I tried on iPod video G5.5/80, Mini G2, H10 6GB and ipod 2nd gen. DB init works on all 4 without crash
07:46:38Lloreanamiconn: The data abort location seems to be fairly consistent too.
07:46:49LloreanIt's also reported on the Sansa
07:47:35 Join tehxed [0] (
07:47:36amiconnI suspect that it might be related to what tagging format(s) are used on the device. All my MP3s are now id3v2 *only*...
07:47:57tehxedIs there any way to turn off ID3 tags?
07:48:03tehxedAnd just get the file name.mp3 to appear?
07:48:07AceNikamiconn: i have my mp3's tagged with both in winamp
07:48:16KalthareI'm trying mine. I've got a horrible motley collection of file formats.
07:48:29amiconntehxed: You cannot "turn off" tags as they are in your files
07:48:50HellmarkI think he wants to not have them be displayed
07:48:52tehxedamiconn: When I say "turn off" I mean stop them from being displayed
07:48:57tehxedYeah, Hellmark got the idea
07:49:02krazykitAceNik, i also can't get the crash on the sansa
07:49:08amiconnYou could either remove all tags from your files if you don't want them at all, or you can make your own wps that ignores tags
07:49:29tehxedamiconn: I've removed the tags, I'm not sure how to go about making my own wps
07:49:38tehxedIs there any tutorial anywhere that could help me out with that?
07:51:34tehxedThank you
07:52:44amiconnLlorean: Do *you* get that crash?
07:53:27KalthareThis is in initializing the tag database, yes?
07:53:41Lloreanamiconn: The only target I have available right now is my H120.
07:53:52amiconnAH, bummer
07:54:28*amiconn tries on archos, as the SH1 is also algnment sensitive
07:55:34amiconnIt would be helpful to know the exact address of the crash and then have the build dir of that build available
07:55:47amiconn(for checking the .map and disassembling the respective .o)
07:55:58KalthareWell, I didn't get a crash on my 5.5G/30GB, and I've got a lot of different types of files.
07:56:13 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
07:56:42AceNikaddress keeps varying in the range 0045XXX
07:57:10AceNikidentify nickel6
07:57:12LloreanAceNik: He asked for an exact one.
07:57:29amiconnJust the address isn't very helpful, even if it's exact
07:58:29amiconn0x45xxx is most probably somewhere in tagcache.o, that's all I can tell without further information
08:00:37amiconnNo crash on archos either
08:01:45 Part rvvs89 ("Cycling channel")
08:01:50 Join perplexity [0] (
08:02:17Lloreanamiconn: Do your tags have comments? The most likely culprit would at least *seem* to be
08:02:41 Join ddalton [0] (
08:02:43amiconn*very* few files have comments iirc
08:03:00ddaltonwho is james83? the svn user name.
08:03:07rockoI gotta say
08:03:16rockorockbox has saved my
08:03:19rockoipod so many times
08:03:35HellmarkTaking battery life issues on some of the players (like the ipod) into account, what players do you prefer?
08:05:14Jangarire: that H10 bug during tag init, it happened to my ipod 5g with yesterday's release as well
08:05:47amiconnLlorean: Eurgh. Many thanks for pointing me there again. Yesterday I looked at that myself, but couldn't see the problem, although it's soo obvious....
08:05:50Jangarisame crash, data abort 045XXX
08:05:58Jangarii went back and found a previous build, probably about a week ago, and it's fine now
08:06:17 Part toffe82_
08:06:18LinusNddalton: James Espinoza
08:06:24amiconnshort* s = (short*) string; <== ouchouchouch on alignment sensitive targets!
08:07:35KalthareHuh. What would that do?
08:07:46amiconnIt works if the COMM tag itself is at an even address in memory, and crashes if not
08:08:16KalthareOhh. Ouch.
08:09:10Jangariis that why?
08:09:17Jangarifucken comments tag
08:09:26amiconnWell, both arm and SH1 don't like unaligned pointers. But a char* (or void* in that case) isn't necessary aligned
08:09:51Jangariit's probably worth using anapod or something to clear the comments box from every tag
08:10:04LinusNbetter fix the code imho
08:12:18amiconnAha, and it can only happen if there are COMM tags in UTF-16. That's why I didn't get any crashes
08:12:18amiconnMy tags are all either iso8859-1 or utf-8
08:12:18Jangarioh well, my week-old build seems fine
08:12:18 Part puzzlet
08:13:10amiconnThe fix should be fairly simple. The cast to short* isn't necessary as the important point here is just to check pairs of bytes.
08:13:22 Part AceNik
08:13:54 Join CaptainSquid [0] (
08:15:35 Join Thundercloud [0] (
08:16:09amiconnunfortunately something like while (*s++ || *s++) { } isn't well defined. It needs a few more lines
08:16:17ddaltonLinusN: are you around?
08:17:11ddaltonDo you think p7682 should be committed?
08:19:38 Join norbusan [0] (
08:19:48LinusNddalton: i think so, but i'd like to investigate the need for the longer syncsplash delay
08:19:48 Part norbusan
08:20:11ddaltonLinusN: do you have an ipod?
08:20:25ddaltondo you have time to change it and test?
08:20:41ddaltonbecause I did this for a friend and I don't have an ipod myself.
08:20:44LinusNi'll have a look ati it
08:20:50ddaltonor a player with a wake up rtc
08:21:48ddaltonalso can you make sure the voice says the right infomation? so when the alarm is set? It is just a bit hard for me because I can't read the screen. But I don't know why it wouldn't
08:23:23amiconnAnyone experiencing the crash on db init who wants to test? Or should I commit right away?
08:25:15LinusNamiconn: commit it
08:25:22Lloreanamiconn: If you're reasonably sure, I'd say go ahead and commit. We seem to be getting at least one bug report every 2 hours, if not shorter periods, from people not searching first.
08:26:21jhMikeSwhat are the details about it and how to make it happen?
08:27:00ddaltonLinusN: If I create patches that call the talk functions will this be ok when the proper fix comes in?
08:27:12LinusNany sansa owner looking into FS #7692?
08:27:12ddaltonso should I be creating patches now that voice the plugins?
08:27:37*jhMikeS will peek over at FS #7692
08:27:42LinusNddalton: i think we should find a solution to the voice file problem first
08:27:53amiconnjhMikeS: I'm pretty sure I identified the problem. You need (1) an alignment sensitive target (2) mp3 file(s) with comment tags using UTF-16
08:27:55LloreanLinusN: I don't think any of the Rockbox Sansa owners here actually experience the garble issue.
08:28:09amiconnLinusN: Committed.
08:28:14LinusNamiconn: goodie
08:28:19ddaltonLinusN: but if I use the functions provide in plugin.h won't it all just work once we modify those functions?
08:28:19jhMikeSamiconn: ah, all my files are UTF-8
08:28:40amiconnMine are either iso8859-1 or utf-8, hence I wasn't able to reproduce
08:29:00LinusNddalton: well, yes, if you don't add plugin specific phrases to english.lang
08:29:18 Join ^pSiOnIc_BeInG^ [0] (
08:29:26 Quit Thundercloud (Remote closed the connection)
08:29:35ddaltonLinusN: so would every plugin file have to be modified or just pluigin.h or .c or talk.c or .h?
08:29:39GodEater_morning everyone
08:30:35^pSiOnIc_BeInG^"unsupported builds" would mean mp3 players that are currently not done yet, right?
08:30:59GodEater_^pSiOnIc_BeInG^: no - it would mean builds of rockbox not using the core code in svn for players which ARE done
08:31:00Llorean^pSiOnIc_BeInG^: No, it means "compiles of Rockbox made by other people"
08:31:11 Quit eigma ()
08:31:17ddaltonLinusN: if I added strings to english.lang could these be copied fairly easily when we figure out the new way?
08:31:34jhMikeSI guess that supercedes FS #7654
08:31:36 Quit ptw419 ()
08:31:49^pSiOnIc_BeInG^ahh, i see thanks; i am just making sure i dont break a rule when asking stuff xD
08:32:19^pSiOnIc_BeInG^will there be more builds for other mp3 players in the future?
08:32:48Jangariwhat do you have, ^pSiOnIc_BeInG^?
08:33:13GodEater_^pSiOnIc_BeInG^: only if someone that owns that player decides to port rockbox for it
08:33:16 Quit CaptainSquid (Remote closed the connection)
08:33:17*jhMikeS will test #7692 and see if it's ok on the older model
08:33:22Llorean^pSiOnIc_BeInG^: No way to predict that. New ports happen when someone who owns the player (or a group of someones) puts in a lot of time and work
08:33:25^pSiOnIc_BeInG^samsung YP-K3, i love it; but wish it did a few other things...
08:33:38^pSiOnIc_BeInG^ahh, i see
08:33:46 Quit BigBambi ("Leaving")
08:34:11 Join ptw419 [0] (
08:34:59^pSiOnIc_BeInG^okay then, is porting it to it difficult? Well i would think it would have to be if it may take a group of people to do, but still..
08:35:12^pSiOnIc_BeInG^you would need to own the player to port it, so what would i need to do to do so
08:35:47JdGordonjhMikeS: if you have some time could you see how small you can strip the ata driver down to please? I was hoping it wouldnt be needed, but my playing this morninng seems to think it might be
08:36:01jhMikeSamiconn: Something interesting for cache coherency between cores seems to be making sure unshared but cacheable data doesn't share the same 16-byte memory segment.
08:36:30jhMikeSJdGordon: I posted a quick strip down but you had disappeared by then. :)
08:36:32ddaltonLinusN: have you had a chance to test that patch?
08:36:44JdGordonoh ok, ill go over the logs then
08:36:57jhMikeSnopaste stays 24-hrs?
08:37:01Llorean^pSiOnIc_BeInG^: There's some information linked to in the New Ports forum stickies
08:37:02LinusNddalton: i'm busy as always, please give me some time
08:37:11ddaltonok then
08:37:14JdGordonjhMikeS: well.. its only been about 5.. so we shuold be safe :p
08:37:22^pSiOnIc_BeInG^(Llorean): alright then, thanks will browse the forums then
08:37:36 Join ender` [0] (
08:38:34 Quit Llorean (Remote closed the connection)
08:38:50 Join Llorean [0] (
08:39:52 Quit ddalton ("leaving")
08:40:17*jhMikeS 's screen looks good with FS #7692 running
08:40:50LinusNjhMikeS: feel free to commit if you think the code looks ok
08:41:08JdGordonjhMikeS: thanks, found the paste :)
08:41:12JdGordonhopefully its small enough :p
08:41:32jhMikeSLinusN: I'll do that. I think it's about as simple and perfect a fix as could be asked for.
08:42:30 Join qweru [0] (
08:42:42 Join B4gder [0] (
08:42:54*jhMikeS double checks to make sure there's no mistake and is running an actual patched build
08:44:52 Join linuxstb_ [0] (i=5343d4aa@rockbox/developer/linuxstb)
08:44:54JdGordonjhMikeS: have you still got a copy of the .c? rafb seems to have died
08:45:16linuxstb_amiconn: Sorry about Player USB typo - thanks for fixing.
08:45:23amiconnBah, just after commit I spot that this stuff can be written more efficient...
08:45:42 Quit Llorean (Remote closed the connection)
08:45:59jhMikeSJdGordon: sure do
08:46:43amiconnjhMikeS: That makes very much sense...
08:47:02*linuxstb_ wonders about austriancoder's commit - it looks like it would break usb detection when HAVE_USBSTACK is defined...
08:47:17 Join CaptainSquid [0] (
08:47:26amiconn...otehrwise core B will overwrite parts of core A's data if it decides to flush the cacheline
08:48:31jhMikeSamiconn: Yes, that is what's going on
08:48:54 Quit qwm ("Reconnecting")
08:48:58jhMikeSI set some things up like that and I get no problems
08:48:59 Join qwm [0] (
08:49:52 Join Llorean [0] (
08:50:06jhMikeSIt's probably not as reduced as it could be. I think I forgot to change some currcard-> to card_info.
08:50:28 Part Llorean
08:51:14***Saving seen data "./dancer.seen"
08:52:30linuxstb_JdGordon: I read your comment about there being maximum size of the binary e200tool can upload - how did you determine that that was the problem, and not something else? Does e200tool give an error message?
08:55:19jhMikeSamiconn: having stacks that are in cached memory strategically aligned seem to stop problems. I also made the maintenence functions as calls, put them in IRAM, and make sure instruction prefect doesn't prefetches into core code.
08:55:43jhMikeS*prefetch doesn't prefetch
08:56:12*jhMikeS whatever :P I think the point will be gotten
08:56:49 Join petur [0] (n=petur@rockbox/developer/petur)
08:57:51jhMikeSJdGordon: we have no need for threading when injecting the code right?
08:59:05amiconnThe first version of my fix was really silly compared to the new one
08:59:12*amiconn points to the delta table :)
08:59:51 Join Zagor [0] (n=bjorn@rockbox/developer/Zagor)
09:00:35linuxstb_amiconn: Did you ever compare LCD speeds on the mini with the framebuffer in and out of IRAM? bluebrother has mentioned that the USB stack doesn't compile for his mini due to a shortage of IRAM, so I was wondering if removing the framebuffer would be an option?
09:01:34amiconnUsing the extra iram on PP5021+ would really be better
09:01:37jhMikeSif I can get this dual-core stuff done, the whole cop stack will be history
09:02:11linuxstb_amiconn: That still leaves the 1st gen mini and (I assume) 4G greyscale with a shortage though.
09:03:51JdGordonjhMikeS: yeah, no need for threading
09:04:19jhMikeSJdGordon: good to hear since I removed all that too
09:04:29JdGordonlinuxstb_: just an uneducated guess because it stops flashing the light once the bin get "large"
09:04:40tehxedHow do I change the font color!?
09:04:54JdGordonalthough, i tested just copying random data across after the smaller bin and that worked up to 64K.. so i dunno
09:05:17 Quit Jangari ()
09:05:39linuxstb_JdGordon: Have you just tried appending random data (e.g. dd if=/dev/random count=100 >> bootloader.bin) to a known working binary?
09:06:09B4gderJdGordon: did you read the e200tool source code and checked that it just isn't a limit in the code?
09:06:31*amiconn was just about to say something similar
09:06:34JdGordonno, I tried increasing the amount of the buffer that gets sent (its read into a 128k buffer, so random garbage at the end of the file)
09:06:50JdGordonI didnt see anything in the e200tool code about that.. but yeah, i might have missed it
09:07:25linuxstb_B4gder: Can you ask MrH where in RAM e200tool is copying arm_code.c to? IIUC, arm_code.c relocates itself to 0x40004000, but that's all I can tell from the source...
09:08:32*jhMikeS wonders about the who gets the credit for FS #7692 :\ 1/2 the lines are from one person, 1/2 from another
09:08:42B4gderit relocates itself?
09:08:57linuxstb_From what I can tell, there shouldn't be a limit to the size - e200tool first sends the length as a 32-bit int, then transfers the data in 64byte chunks.
09:09:21linuxstb_B4gder: Yes, if I understand the startup asm code in arm_code.c correctly.
09:09:46linuxstb_(I could be wrong...)
09:11:38B4gderthe arm_code is linked to address 40004000 according to the header comment
09:11:57B4gderit gives me the feeling that is where it is copied to
09:12:03JdGordon... let me explain why I tihnk the filesize has something to do with it... the first thing the code does is flash the light a fe times, then calls ata_init(), then should flash forever (at a different speed so its easy to tell the difference).... if the ata_init call is commented out the binary is ~1100 and the 2 flashing speeds are very visiable... with it uncommented i get no flashing at all....
09:12:13peturthe buffer of usb is in iram
09:12:17JdGordonuncommented makes the bin 22k or so
09:12:34B4gderright, 40004000 is IRAM
09:12:51linuxstb_B4gder: Can you explain what the code commented as "Make sure we run at the right address" at the start of arm_code.c is doing? IIUC, that's relocating the code to 0x40004000
09:13:18jhMikeSok, both get the credit then
09:13:22JdGordonisnt iram 40000000?
09:13:28 Join Rob2222 [0] (
09:13:32B4gderJdGordon: yes, so 40004000 is iram too...
09:13:33linuxstb_It starts at 0x40000000...
09:13:48JdGordonah, i thought you were saying the start was 40004000...
09:13:56B4gderlinuxstb_: I'll ask him if he knows
09:13:58*JdGordon sshh....
09:14:24linuxstb_B4gder: Thanks.
09:14:35 Join FOAD_ [0] (
09:15:59amiconnJdGordon: It could also be the ata init hanging
09:16:23 Quit ^pSiOnIc_BeInG^ (".UPP.")
09:16:53amiconnWouldn't surprise me, as most probably some hardware isn't initialised in manufacturer mode that is initialised by the sandisk loader in normal mode
09:17:13B4gderyeah, we should expect different inits to be required
09:17:45B4gderdisassembling the BL could offer some interesting ones ;-)
09:17:52amiconnHence my question regarding bidirectional communication yetserday
09:18:12amiconnThis way we can get register dumps from manufacturer mode
09:18:32linuxstb_That should be easy enough to add to MrH's original arm_code.c
09:18:43B4gderthe i2cdump command should be a good inspiration for such tweaks
09:18:47linuxstb_i.e. just implement extra commands...
09:18:49amiconnWell, we could also blink them out via the wheel led
09:19:06B4gderthe i2cdump command already does read out 8K of data from target
09:19:09B4gderto a local file
09:19:13amiconnBut that'd be much slower and error prone
09:21:54JdGordonamiconn: I dont think so, because i dont get the flashing before ata_init...
09:22:06JdGordonthats why i did 2 speeds, so i can see where its at
09:22:49*JdGordon will be in and out for the next few hours
09:23:00 Quit Weiss (Read error: 110 (Connection timed out))
09:24:00 Quit akaias (Read error: 110 (Connection timed out))
09:24:02jhMikeSblinking the wheel light is quite useful indeed for debugging imho when lcd isn't available.
09:26:51amiconnjhMikeS: In case you ever need to read bytes that way, here's the method I used on archos for debugging the self extractor:
09:27:26jhMikeSUse it with a photo diode and a latch? :)
09:27:56amiconnbit == 0 -> short blink, bit == 1 -> long blink. Short pauses between bits, long pause between nibbles, and double long pause between bytes
09:28:23amiconnA morse-alike code.
09:28:40 Join przemhb [0] (i=5035dcaa@gateway/web/cgi-irc/
09:28:49 Quit przemhb (Client Quit)
09:28:55*jhMikeS was considering making the wheel light an optical s/pdif output :P
09:29:33 Join przemhb [0] (i=5035dcaa@gateway/web/cgi-irc/
09:30:26LinusNhmmm, i need to find a way to wait until all queued voice clips have been played
09:31:25 Quit FOAD (Read error: 110 (Connection timed out))
09:31:25 Nick FOAD_ is now known as FOAD (
09:31:50LinusNthe problem is that is_voice_queued() returns false as soon as the last clip is scheduled for playback
09:32:44LinusNi'm trying to make gui_syncsplach() not return until the voice is finished
09:33:00przemhbI would like to tell you that I have added a new codepage (WIN1250 - "Central European") - take a look FS #7712.
09:34:26przemhbCould someone tell me is it possible to implement convlover in rockbox? It would be very nice to have a possibility to convolve impulse response to flatten frequency response of headphones used or add some room acoustics.
09:34:40B4gderand I would like to tell you that we get notified by mail of all new tracker entries... :-)
09:35:00 Quit davina (Read error: 110 (Connection timed out))
09:35:14jhMikeSLinusN: I suppose wait on both the queue and playback stopped conditions would do
09:35:27przemhbBadger: Excuse my mistake - I am first time Rockbox developer.
09:35:36B4gder:-) no worries at all
09:35:40LinusNjhMikeS: but what if music is playing?
09:35:42JdGordonamiconn: i was hoping to be able to use the lcd backlight also, one light for 0 and one for 1 :) but didnt have luck there either
09:36:34jhMikeSLinusN: voice_is_playing shouldn't be false again until just before the thread goes back to wait
09:36:45LinusNjhMikeS: ah, thanks
09:38:04LinusNis_voice_speaking() is broken according to the comments in playback.c :-(
09:39:57 Join davina [0] (
09:40:51 Quit linuxstb_ ("CGI:IRC")
09:40:59jhMikeSis_voice_speaking()? seems like a new addition.
09:41:12LinusNit's called by voice_wait()
09:42:16LinusN...and that function is only present on swcodec platforms... :-(
09:42:17jhMikeSyeah, that's a new one. I would think { is_voice_queue() && voice_is_playing } would be more appropriate a test.
09:42:43LinusNyes, i wonder why the music playback would be involved at all...
09:43:19LinusNhmmm, i wonder what would happen if the music was paused
09:43:26jhMikeSI suppose do up an equivalent for hwcodec ?
09:43:38LinusNi guess it would freeze
09:43:50B4gderlinuxstb: "It is right 0x40004000"
09:43:52jhMikeSyeah, I think it will
09:44:12B4gderThe reason for the relocation code is for the pre-boot usb mode
09:44:28grndslmwhat's the status of rockbox on the sansa e200 players?? ... in terms of sound quality?
09:44:36grndslmis it equal to original firmware yet?
09:44:49B4gdergrndslm: it sounds great
09:44:51jhMikeS... && !paused?
09:44:58linuxstbB4gder: I don't understand that relocation code then...
09:45:21grndslmB4gder: so, it sounds as good as original firmware?? .. or better?
09:45:26B4gderlinuxstb: it is there to make sure the code is at 40004000
09:45:43B4gdergrndslm: yes, although I can't remember the OF anymore... :-)
09:45:52linuxstbB4gder: But MrH said the code is copied to 0x40004000 ?
09:45:56jhMikeSgrndslm: It's quite high on my sound quality list compared to many others
09:46:02B4gderlinuxstb: yes, if it isn't there already
09:46:11B4gderit is there already in the manufacture mode
09:46:14grndslmB4gder: when i tried it a few months with rockbox, it seemed to be a bit scratchy or staticy at some points
09:46:16B4gderbut not in pre-boot mode
09:46:36B4gder"So I just want to support both modes with the same code."
09:46:39jhMikeSgrndslm: I think rockbox tops it. Our codecs seem to be higher quality.
09:47:17linuxstbB4gder: Ah, so in pre-boot mode it's transferred to ???, and in manufacture mode, it's transferred to 0x40004000 ?
09:47:19grndslmjhMikeS: interesting, it must have improved a good bit, i'll have to check it out again... as soon as i can find my charger ;-)
09:47:47B4gderlinuxstb: correctly, in pre-boot it gets transfered to 0x40000000
09:47:55jhMikeSscratchy shouldn't be a problem. that was addressed as well as all the other earlier problems
09:48:06B4gder"In manufacturing mode the i2c contents are at 0x4000000 so the load address is higher."
09:48:22linuxstbB4gder: Thanks. That's clear now, and it could also explain JdGordon's problems...
09:48:42jhMikeSgrndslm: when did you last try rockbox on e200?
09:48:50GodEater_linuxstb: it does ?
09:48:52linuxstbJdGordon: You need to make those two changes (crt0-pp-bl.S and that I originally suggested.
09:49:12grndslmjhMikeS: prolly around 5~6 mo ago
09:49:45jhMikeSgrndslm: yeah, time to try again I'd say
09:51:28JdGordonlinuxstb: ok, ill give then a go soonish...
09:51:41JdGordongood thing i did svn diff > changes before reverting them :)
09:54:30*jhMikeS goes to commit FS #7692
09:54:41 Part przemhb
09:59:20linuxstbGodEater_: The code currently relocates itself from 0x40004000 (the load address) to 0x40000000 (where it is linked to run from). IIUC, that could cause problems if the size is > 0x4000 bytes (16KB), as the code appears to run on both cores - it's before the COP is halted.
09:59:26 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
10:00:07*jhMikeS wonders what's up with the yellow coming out in rec_format_afmt (haven't seen the build table in a bit)
10:00:52*linuxstb wonders why both MrH's arm_code.c and Rockbox's crt0-pp-bl.S both do the relocation before halting the COP.
10:01:35B4gdersounds like an odd choice
10:01:44 Join class37 [0] (
10:02:42class37having trouble when i put my ipod 30 gig vidoe on coffee mug fm transmitter... goes into usb mode
10:02:50class37have most current build installed
10:03:01GodEater_would someone mind telling me what that filename translates to in english ? Obviously I get the "pp" is portalplayer - but I always wondered about the rest of it
10:03:28jhMikeSdoes anyone recall when those warnings started?
10:03:34linuxstbI'm not sure what crt0 actually means - it's the filename always used for the initial startup code. "bl" is bootloader.
10:03:47B4gderjhMikeS: a gcc 4.2 update
10:03:58GodEater_I guessed the bl was bootloader too - but wasn't sure
10:04:04GodEater_I'd love to know where crt0 comes from
10:04:16linuxstb"C runtime" according to wikipedia -
10:04:29linuxstbAnd I guess the zero is because it's the very first code to run
10:05:57GodEater_I imagine you're right
10:08:12jhMikeShmmm...what a useless warning for something that uses documented compiler bahavior to default-fill an array and then override them with specified inits. :\
10:08:45B4gderjhMikeS: yes, they do seem... pointless
10:09:00B4gderobo mentioned a flag that will silence (most of) them
10:09:58 Quit tehxed ()
10:11:09jhMikeSsome of those tables to convert types have gaps in places and they change size with every codec added. the setup now avoids the need to edit them at all. I'd hate to lose the convenience of that and worry about the bugs that will crop up if the method can't be used.
10:15:48B4gderwe'll see more servers going 4.2, so we need to deal with it
10:16:39 Join obo [0] (n=obo@rockbox/developer/obo)
10:17:23jhMikeSI've looked before about suppressing specific useless warnings in gcc and came up empty handed.
10:18:28B4gderI don't share that experience. it really depends on what warnings
10:18:46B4gderthere's no generic disable option
10:18:48jhMikeSOne that bugs me (x is possibly used uninitialized) : { int x; if (cond) { x = something; } ....... if (cond) { something_else = x; }
10:19:25JdGordonthats a valid warning though...
10:19:34jhMikeSnot in that case
10:19:39JdGordonoh, your saying its the same cond?
10:19:40B4gdernot if 'cond' remains the same
10:19:52JdGordonah well... then you might have a point
10:19:57jhMikeSso I'm forced to initialize x for no reason
10:20:23B4gderdoes using -O2 or similar change such a case?
10:20:24 Join ddalton [0] (
10:20:54B4gderI mean, can gcc possibly see this better if optimizing
10:20:54peturyou should init it... if the second cond ever changes, a nasty bug is introduced easily
10:21:24*jhMikeS couldn't even compile an ARM build with any other -O option.
10:21:27B4gderyes, I agree that the warning can be seen as pointing out a sloppy practise that leads to hard-to-maintain code
10:21:53 Quit Hadaka (Read error: 60 (Operation timed out))
10:22:01jhMikeSpetur: I feel like I should be the master of what's needed, not the compiler. :)
10:22:13B4gderit's just a warning
10:23:25jhMikeSheh...just a warning. the cops tell me that all the time :P
10:24:37peturoh... you're _that_ dangerous? :p
10:28:44amiconnLinusN: Why do you want gui_syncsplash() to wait for voice to end? Voice is asynchronous by design, I'd hate that behaviour...
10:28:45*jhMikeS *whistling*
10:29:04*jhMikeS supposes he should close some FS stuff now
10:29:15LinusNamiconn: gui_syncsplash() is not asynchronous, hence the name
10:29:37*jhMikeS thought "sync" was referring to screens
10:29:41amiconnNo., the "sync" just means "all LCDs" in that case
10:30:58LinusNamiconn: ah, my bad, but if the "ticks" parameter is set, it sleeps
10:31:16LinusNand then it should wait until the voice is finished
10:31:30amiconnImho it should not
10:31:37LinusNotherwise the voice is interrupted when the screen returns to the menu
10:31:41amiconnVoice can play *way* longer than the timeout
10:31:52LinusNthat's my point
10:32:18amiconnVoice is only interupted if there's all new stuff to announce, and that's on purpose then
10:32:36LinusNin this case, the alarm screen exits before the "waking up in xx" is spoken, so you don't hear it
10:32:42amiconnIt ensures quick navigation
10:32:55ddaltonyour talking about my patch?
10:33:10amiconnThat's a special case then
10:33:22LinusNyes, perhaps
10:33:26ddaltonhow long does it take to finish talking
10:33:56amiconnddalton: That's not predictable, as it depends on language and voice speed
10:34:05ddaltono yeah...
10:35:13ddaltonwell when exiting the alarm you land on the wake up alarm option why don't we just make that voice not interrupt the alarm wake up message.
10:35:53ddaltonI guess we can't. Because of make_menu
10:36:24ddaltonwhat would happen if you press a button when a gui_syncsplash is displaying something is set at 20 seconds?
10:36:37ddaltonHow does "Loading" work?
10:36:42 Join DefineByte [0] (
10:36:47ddaltonisn't it much the same?
10:37:58amiconnThe button press will be queued like normal
10:38:23ddaltonso what should we do then?
10:38:30ddaltonany ideas?
10:38:40LinusNwe should let the alarm screen wait for the voice before exiting
10:39:10ddaltonthere are different spees the voice file can be
10:41:06DefineBytei've noticed that, on the iPod 5G, when the backlight is off the first keypress is not enacted, it only turns the backlight on. the second keypress does enothing (at least while the backlight is on), the third keypress works as expected. is this known?
10:42:54KalthareI think that's configurable.
10:42:54pixelmaLinusN: could this problem with voicing splashes also be the reason why I sometimes get the "loading" splash voiced twice?
10:43:08GodEater_DefineByte: that's a config option
10:43:11LinusNpixelma: yes, probably
10:43:21pixelmaI see
10:43:41DefineBytedisable second keypress is a config option?
10:43:55linuxstb"first keypress enables backlight" is the option
10:44:01DefineBytei know you can set whether to just turn te backlight on or not on first keypress
10:44:06KalthareOh, the second keypress? no, that'd be a bug then.
10:44:11DefineBytebut the second gets eaten
10:44:19KalthareSorry.. getting late, and I didn't read it right.
10:44:24*jhMikeS knew his dual-core changes weren't the reason for double-talk
10:44:35DefineBytei didn't describe it very well, probably
10:44:47GodEater_DefineByte: ah - not noticed that behaviour myself
10:45:06DefineBytei get it all the time. as in 100% reproducable
10:45:13GodEater_although I have noticed that the ipod is very slow to acknowledge key presses after having the hold switch turned off
10:45:23GodEater_sometimes it takes as many as five goes for it to accept the "pause" perss
10:45:37amiconnI never observed this, but then I don't use "First keypress enables backlight only"
10:45:38DefineByteit may be related to the baclight coming on with the hold switch, i'm not sure
10:46:00DefineBytei always use the hold switch
10:46:08amiconnI do know that sometimes the first keypress after hold is eaten
10:46:22DefineBytethat could be it then. :)
10:46:26amiconn...but then I don't use hold often either
10:46:35 Join Jangari [0] (
10:46:46Jangaridoes anyone use that dos based font converter?
10:46:58DefineBytewell, as long as it's known. not a big issue really. :)
10:47:02B4gderdos based font converter?
10:47:18Jangariit's fucken cool, and very quick and simple
10:47:34B4gdersomehow I doubt it is truly dos based
10:48:05DefineByteprobably just console based
10:48:27Jangariwhatever, you run it from cmd
10:48:35Jangarii aint no nerd
10:48:47DefineBytehey, we need more nerds
10:49:05Jangarii'm a nerd, just not a computer nerd
10:49:13gammyHeh a friend of mine once made a little dos TSR which rotated the bios font. It was wicked :D
10:50:01DefineBytewhat open hardware is out there that doesn't have a rockbox port? anyone know any?
10:50:26B4gderopen hardware such as... ?
10:50:42DefineBytenot open open, just documented i guess
10:50:53B4gdersuch as... ?
10:51:03DefineBytei thought that was my q
10:51:14B4gderI don't understand your question
10:51:16***Saving seen data "./dancer.seen"
10:51:24DefineBytecoldifre is documented right?
10:51:27DefineBytepp is closed
10:51:40GodEater_coldfire is an architecture
10:51:42GodEater_pp is a brand
10:51:57Jangariah, wickid
10:52:00B4gderI'm not sure all coldfires are documented
10:52:04Jangarijust found what i was looking for
10:52:13Jangaria simulator
10:52:26gammyDon't the iHP120-series use coldfire?
10:52:32LinusNDefineByte: what we are trying to say is that there are actually no players around with documented hardware, except for the neuros
10:52:32gammy(Sorry for such a silly question)
10:52:38DefineBytebut is there any DAPs out there with known documentation but no port due to lack of interest?
10:52:38B4gdergammy: they do, yes
10:52:44gammyB4gder: Ok.
10:52:54B4gdergammy: but there are many different coldfire versions
10:53:00DefineBytereally? that's a pretty sorry state of affairs
10:53:02gammyB4gder: Yeah I understand that.
10:53:14B4gderDefineByte: well, if you think of "documented" for the cpu only
10:53:17B4gderthen yes
10:53:18LinusNDefineByte: and ironically, the neuros uses a microcontroller with no public documentation
10:53:21B4gderlike the iAudio m3
10:53:55B4gderbut virtually no players are documented all the way
10:53:56DefineBytewhy do these companies not want their hardware to be used? it's mad
10:55:15 Join skate3214 [0] (
10:55:26jhMikeSDRM, license fees, paranoia, I'm sure
10:55:41B4gdermostly paranoia
10:56:04B4gderand the "we don't want the hobby projects to clog our support"
10:56:13B4gder(basically what the AMS guy told me)
10:56:14jhMikeSthat telechips doc was so huge on IP
10:56:14DefineBytethey probably see their hardware being ripped off by chinese companies etc
10:56:15LinusNwell, rockbox *has* caused an increase in support calls to for example archos
10:56:35DefineBytereally? that's pretty bad
10:56:48DefineBytebut archos isn't the chip maker
10:57:06DefineByteso it wouldn't affect them
10:57:11LinusNthe chip maker paranoia is beyond me
10:57:39DefineByteit's corperate culture i guess
10:57:51DefineByteeverything they do has to be secret or it has no value
10:57:57jhMikeSanything NDA'ed I've seen ... it's just ridiculous. hardly anything, if anything worth hiding
10:59:04DefineBytethey like to feel important it seems. like they have something so valuable that if others found out about it they'd be instantly screwed
10:59:07linuxstbDefineByte: ArchOpen is a project similar to Rockbox - it's an open source replacement OS for the newer Archos devices. So it contains lots of source/info that could aid Rockbox ports to those (and similar) devices. Although those Archos devices are PMPs, rather than strictly DAPs, so Rockbox ports may not be that useful.
10:59:38GodEater_which isn't stopping you trying to port it to one :)
10:59:40B4gderI would also think that customers of these hardwares don't care enough
10:59:48B4gderso there's no push for the manufactorers to change
10:59:50jhMikeSI'm not just getting that from stuff RB related.
10:59:55DefineBytei'd imagine owners of those would be after pretty extensive multimedia support
11:00:59DefineBytewhat about DAB radios? maybe rockbox could work on those (the ones with storage at least)
11:01:01gammyB4gder: Have you done any work for the iHP-series yourself?
11:01:17 Join miepchen^schlaf [0] (
11:01:21jhMikeSYou have to sign NDA just to get service manuals stuff, for instance.
11:01:22gammyB4gder: Neato.
11:01:36B4gderme been poking all over
11:01:56gammyB4gder: Without the rockbox port to that I would have bought a new player many, many yeaes ago. Rockbox keeps it kickin'.
11:02:31DefineBytei guess that's why companies don't like to open hardware then. they want you to upgrade
11:02:34gammyyears even. Silly new keyboard.
11:03:03jhMikeSyes, rockbox does show how underutilized the devices are and most technological "progress" is a sham
11:03:11linuxstbDon't chip makers also sell expensive SDKs with runtime licenses... ?
11:03:33DefineByteforced obsolescence is a terrible thing
11:04:15DefineByteespecially if we're supposed to be more environmentally conscious
11:04:52jhMikeSDid I read correctly here that Sansa thought about actually introducing an enhanced e200 to do what was done with a few lines of code for SDHC?
11:05:03DefineBytehow many SDKs do they sell for 2 year old hardware?
11:05:22B4gderjhMikeS: yes, at least according to some rumour
11:05:24LinusNjhMikeS: doesn't sound surprising at all
11:05:38B4gderthe sdhc support should be displayed with a special logo
11:05:52B4gderwhich the current e200 players don't feature
11:06:04DefineBytepreferable holographic
11:06:24linuxstbB4gder: So we should add the logo ? ;)
11:06:31jhMikeS I guess we just can't make it official then.
11:06:43B4gderyeah, "buy your microsdhc logo sticker here =>"
11:07:21DefineBytewould users really know what it meant?
11:07:26jhMikeSbut it many stickers in one
11:07:39DefineBytewould they even care?
11:08:23DefineBytei was more thinking holographic to make forgery of their precious logo harder
11:08:25jhMikeSmakes me think, it's really the lawyers
11:09:18DefineBytecapitalism is founded on lawyers
11:09:43jhMikeSheadlines: SDHC Sticker Forgery Ring Busted
11:09:59DefineByteit wouldn't surprise me
11:10:16DefineBytethey do it for sony memory sticks
11:10:22B4gder"To ensure compatibility, look for the SDHC logo on cards and host devices (cameras,
11:10:23B4gdercamcorders, etc.)"
11:10:51jhMikeSrockbox is _not_ _not_ compatible with SDHC
11:11:08peturnot not? so it is?
11:11:18*jhMikeS didn't say that
11:12:38jhMikeSisn't it possible "not not" != "is" ? :P
11:13:02Aveis it such a crime if you just say it out loud
11:13:18Avethe emperors bloody new clothes
11:13:44*jhMikeS will wash out those blood stains ;)
11:14:46DefineBytethey may say it's quality control. there's a shit load of fake memory stick duos on ebay don't know about other standards
11:15:17B4gderif they're fake, they surely fake the logo sticker as well
11:15:32DefineByteindeed. it's all bs
11:15:46DefineBytejust trying to appear like they're doing something
11:15:53DefineBytesame with drm
11:16:10DefineByteif i was a share holder i'd be pissed at the amount of omney wated on drm
11:16:31DefineBytebut it keeps up appearances
11:16:42B4gderI believe companies such as Apple has earned quite a few bucks even though they do DRM a lot
11:17:03DefineByteim more thinking of the record labels
11:17:04B4gderbut sure, they may still have wasted money on it
11:17:22DefineBytethe hardware manufactuers are quids in
11:17:26 Part LinusN
11:18:58jhMikeSif the company can be convinced it's to their benefit to help the rockbox project, I don't doubt they would. Just the fact that sansa approached the project seems to indicate the waters are being tested.
11:19:26DefineBytei buy daps purely based on rockbox support now
11:20:22DefineByteno closed solutions offer anything close
11:20:50jhMikeSif they do share info about things such as the AS3514, it's best to be on best behavior and not screw that stuff up.
11:20:51GodEater_they do offer a bit more in terms of reliability mind you
11:21:06GodEater_when *is* playback.c going to be re-written ? ;)
11:21:41DefineBytebut if you're stuck with an unreliable dap you're screwed
11:21:46*GodEater_ suspects it's still a case of "let sleeping dogs lie"
11:21:48jhMikeSwhen the buffer is separated...a big start esp. to getting threading right
11:21:50DefineByteat least with rockbox there's a future
11:22:35*petur wishes linuxstb good luck on getting a response from austriancoder
11:22:45B4gderyeah, does he even read the list?
11:23:01B4gderI think all committers should of course...
11:23:18DefineByteformal procedure and all x)
11:23:40B4gderrockbox is 100% about being formal
11:24:23*jhMikeS gets lazy about forums and mailing lists, etc. when working on a big project.
11:24:28GodEater_yeah, we have specs written in Z and everything
11:24:57 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
11:25:15 Quit kubiixaka ("Miranda IM! Smaller, Faster, Easier.")
11:25:48DefineBytewere you being sarcastic b4dger?
11:25:58jhMikeSdual-core thread.c sort of rivals playback.c in length but actually doesn't crash or deadlock :)
11:26:23DefineBytehow common are rockbox crashes in normal use?
11:27:02ddaltonwhat target?
11:27:14DefineByteall, in general
11:27:21jhMikeSnot uncommon enough thanks to race conditions. cooperative multitasking really doesn't afford any shortcuts to synchronization.
11:27:24DefineBytein non-target specific code
11:27:30ddaltonh300 not many ipods a couple more
11:27:35ddaltonbut over all very good
11:28:07GodEater_DefineByte: there are things you can do to produce reliable crashes in rockbox
11:28:17GodEater_for which the solutions are not at all abvious
11:28:40DefineBytemy useage pattern is pretty basic so i've never seen a crash
11:29:13GodEater_DefineByte: do you have the setting for automatically changing folder turned on ?
11:29:27DefineByteno i don't
11:29:36GodEater_DefineByte: I won't bother demonstrating then ;)
11:29:37DefineBytedoes that crash often?
11:29:53GodEater_it does if you're in the file browser as the lpayback engine tries to change to the next folder
11:29:55B4gderTI releases another DMS320 sibling, sub 10USD with HD-abilities...
11:30:04oboand you have dircache turned on
11:30:16B4gderguess we'll see more TI based DAPs and media players ahead
11:30:18DefineBytethat i do have on
11:30:18GodEater_you have to have dircache on for that one ?
11:30:39oboyup, I could only reproduce it with dircache enabled the TI offices
11:30:57DefineBytehow about porting rockbox to PVRs? i'd love to turn my topfield into a jukebox. 8)
11:31:01B4gder"can encode or decode 720p HD MPEG-4 SP video at 30 frames per second"
11:31:19B4gderat 400mW
11:31:27*jhMikeS hates the day he needs to add AUDIO_SRC_VIDEO :P
11:32:22*B4gder looks forward to a rockbox port for a DMS320 device
11:32:49DefineByteportable video seems a bit of a dead end to me
11:32:51GodEater_obo - how right you are - that doesn't cause the codec failure if you have dircache off
11:34:42obothat is as far as I got tracking it down, I think it'll need a dircache guru to take a looksie
11:34:54*GodEater_ points at jhMikeS
11:35:01*linuxstb updates SansaE200RInstallation and hopes it's clearer...
11:35:40GodEater_linuxstb: that looks much better
11:35:42*jhMikeS stares puzzled
11:36:09GodEater_at what ?
11:36:20*obo points at Slasheri :)
11:36:20jhMikeSwhy you're pointing at me
11:36:28GodEater_thought you were Mr. Dircache :)
11:36:48jhMikeSbleh ... should I be? I need a new hat.
11:36:51GodEater_clearly I was wrong
11:36:55*GodEater_ goes to stand in the corner
11:37:26DefineBytesounds like someone needs to be
11:39:01*jhMikeS can't even remember what file implements it :P
11:39:13oboFS #5797 is the bug report
11:43:50 Join Weiss [0] (
11:45:12 Quit LycoLoco_ (Read error: 110 (Connection timed out))
11:51:49amiconnGodEater: Slasheri is Cacheri (and Flasheri as well) ;)
11:52:17DefineBytelooks like dircache removes the existing cache file before creating the new one. is that the right way to do things?
11:52:28DefineByteor am i getting it wrong?
11:52:44 Join Rob222241 [0] (
11:52:51GodEater_DefineByte: when it initialises you mean ?
11:53:40GodEater_DefineByte: would seem the safest way to do it to me
11:53:46amiconnThere is no cache file, except on a flashed H1x0
11:54:32DefineByteso if thw write fails you just lose the cache? i guess it's easily rebuilt
11:54:39 Join fm2 [0] (i=c27f0814@gateway/web/cgi-irc/
11:54:46*jhMikeS basically gets the dircache hoosgow now
11:55:41fm2I'd like to talk about FS #7232. Would a setting for keymap to use be accepted in SVN?
11:55:51amiconnDefineByte: On all standard targets, the cache only lives in RAM
11:56:13preglowjhMikeS: adding dual core support adds 3k lines to thread.c? +? ??
11:56:15DefineBytedidn't know that. thought it got written to disk eventually
11:57:08amiconnNo, that would be dangerous as the disk content can change without rockbox noticing
11:58:13linuxstbDefineByte: What would be the point in saving dircache? Between boots to Rockbox, anything (and everything) could have changed on the disk, so Rockbox has to do a complete scan anyway.
11:58:57DefineBytecan't it check to see if the FAT has changed (or whatever it's called xD)?
11:59:31B4gderthis may sounds strange but we have given this some thought already
12:00:04DefineBytei'm sure there's a good reason. i just like tothink about it :)
12:00:14 Join kclaf [0] (n=kclaf@
12:01:22 Quit kclaf (Remote closed the connection)
12:01:24 Part fm2 ("... awaiting JdGordon and Llorean to arrive")
12:01:25jhMikeSpreglow: not that much.
12:01:48jhMikeSI did comment it a whole lot
12:01:54preglowdefinitely good
12:02:55DefineByteyou can't save the FAT as well and compare it?
12:03:20jhMikeSit's about 2100 lines fixing all holes, adding comments, richening the threading API and guaranteeing low latencies all about 25-30% faster.
12:03:57dionoeaDefineByte: save the whole disc on the disc ??
12:04:25DefineBytei thought the FAT was just an index pointing to the clusters
12:05:05dionoeaAnyway, comparing would be the same as rebuilding the cache wouldn't it ? (interms of disc access)
12:05:19preglowjhMikeS: got a patch hanging around?
12:05:40DefineBytehow small is the FAT then?
12:05:43jhMikeSpreglow: could make one
12:05:44amiconnDefineByte: Directory contents may change without the FAT being touched. A little bit of thinking reveals that
12:05:52amiconnAnd the FAT can be several MB
12:06:02DefineBytehah, or lack of ignorance :D
12:06:03preglowjhMikeS: don't bother, can't test now anyway
12:06:26ddaltonhow does make_menu work. Like for example if I make an option "music" what function does it call?
12:06:49jhMikeSI _should_ be able get to cracking the problems on pp5020 soon
12:07:40 Join The-Compiler [0] (
12:08:22jhMikeSthe trickiest little bit to implement is to switch a thread's processor but that works nicely now too (it will have practical application).
12:08:36 Quit Rob2222 (Read error: 110 (Connection timed out))
12:08:50DefineBytehow does the directory contents get changed without touching the FAT in valid operation? no doubt a stupid question born out of ignorance of file systems x)
12:09:22 Quit skate3214 ("ChatZilla [Firefox]")
12:09:26DefineBytefeel free to ignore if you find yourself slapping your forehead
12:10:39LinusNDefineByte: the FAT only contains information about *where* on the disk the files/directories are
12:10:58LinusNnot the contents
12:11:21LinusNwhich means that you can, for example, rename a file without altering the FAT
12:11:24preglowjhMikeS: what application?
12:11:40 Quit Rob222241 (Connection timed out)
12:11:44GodEater_LinusN: the file name is not held in the FAT ?
12:11:49DefineBytelearn something new... where are the filenames held then?
12:11:56LinusNGodEater: nope
12:12:17LinusNDefineByte: in the directories
12:12:26GodEater_ah of course
12:12:41DefineBytewhich you scan. okay fair enough :)
12:13:09*ddalton better read his c book!
12:14:04DefineBytemy brain is a load of 0xFFF7 it seems. :)
12:14:33jhMikeSpreglow: one big place where it's safer is to terminate another processor's thread. another will be for codecs that which to choose their processor configuration. another is pcm control which should be in the correct context.
12:16:25 Part Kalthare
12:19:23jhMikeSthere's currently a huge bug in SVN where a terminated cop thread continues to use a stack from an exited plugin until it changes to another thread to run. baaaad. taken care of using idle stacks when needed.
12:22:11 Join seablue [0] (
12:25:34jhMikeSone must mind the stack of the last thread to call switch_thread is the one used for idling, until the next real context switch.
12:25:38amiconnjhMikeS: Eh? How can a terminated thread continue to use something? Shouldn't the cop just idle in cop_main() if there are no other threads?
12:25:52jhMikeSread above :)
12:27:01amiconnHmm, I don't think I understand that fully. Such things happen when too much complexity is introduced :/
12:28:45jhMikeSwhich is sp is the one being used until load_context is called for the next thread? even the old type threading would be subject.
12:31:04amiconnNo it wouldn't, as it always switched context for real
12:31:14 Join austriancoder [0] (n=austrian@rockbox/developer/austriancoder)
12:31:23jhMikeSas long at least one thread was not sleeping
12:31:43preglowamiconn: there's nothing "keep it simple" about dual core in the first place, you can't talk about introducing complexity there
12:31:56austriancoderlinuxstb: I have seen this too that usb detection wasnt working... i am checking it atm
12:32:19jhMikeSgiven that geniune blocking cuts processor activity by some 200 times (from your measurement) I'd say it's worthwhile especially with powerful CPUs that would just spin around horribly making millions of switch thread calls when doing nothing.
12:32:33 Join My_Sic [0] (
12:33:03 Join james_malone [0] (n=James@
12:33:17linuxstbaustriancoder: OK. Something else I wanted to bring to your attention is that the Sansa bootloader runs usb_init() - which (unless you were running your own bootloader) would have explained why the usb detection still worked when you removed that init code from Rockbox.
12:33:29 Join Nick_Brackley [0] (
12:33:49jhMikeSIf archos managed 2400 doing nothing, I don't even want to see what a Gigabeat would rack up over the 10-12 switches it currently does.
12:34:15austriancoderlinuxstb: I think its time build bootloaders :)
12:34:16 Quit james_malone (Client Quit)
12:35:27obolinuxstb: any thoughts about FS #7709?
12:36:38linuxstbobo: Did you read the results of my own attempt to implement "usb bootloader mode" in IRC last night?
12:36:52obono, I missed that
12:37:14linuxstbobo: Basically, calling usb_init prevented both disk mode and the original firmware from working.
12:37:33linuxstbYour method works (IIUC) because you're rebooting.
12:39:31 Nick fxb__ is now known as fxb (
12:39:33oboah, okay. Maybe I ought to add a check for the hold button as well as usb_power... (thinking about having hold on, and connecting a USB charger)
12:40:19 Join aliask [0] (
12:40:19 Quit petur (Read error: 104 (Connection reset by peer))
12:40:30 Join petur [0] (n=petur@rockbox/developer/petur)
12:40:56 Join Spaetzchen [0] (i=57a0e346@gateway/web/cgi-irc/
12:41:10Spaetzchenis someone here who can speaks german?
12:41:31 Join Rob222241 [0] (
12:41:34Spaetzchendu sprichst deutsch und kannst mir helfen? ;)
12:41:42DefineBytethe name kinda gives it away :D
12:41:58austriancoderdeutsch: ja,,, ob ich dir helfen kann... keine Ahnung
12:42:06Spaetzchenhihi, ich versuchs mal ;-)
12:42:12pixelmathis is an english speaking channel though
12:42:13linuxstbobo: That could work... Although it would be nicer to work out how to "uninit" the USB hardware and make the OF and diskmode happy...
12:42:31Spaetzchenich hab gestern rockbox auf meinem sansa e250 installiert
12:42:45austriancoderSpaetzchen: lets move to a privtate chat
12:42:52DefineByteit's all about the games isn't it? >.>
12:42:53Spaetzchenwie geht das?
12:42:54jhMikeStimeout checks are also now done only at the soonest task ready to run instead of every tick, so that's a factor of 10 reduction in checks not to mention lazy removal from that list that speeds thread wakeup a good deal. you can count that a thread will never have to wait till the next tick unless it calls sleep with an arg of 1 or greater.
12:42:58pixelmaSpaetzchen: maybe you'll find more people who can help you when you try to explain your problem in english
12:43:20Spaetzchenmy problem is that I can't speak english very well ;)
12:43:43Spaetzchenso I can't explain it in english, because I even don't know how to explain in german
12:43:46GodEater_Spaetzchen: you speak it a lot better than some of the idiots we get in here who claim it as a first language
12:44:00Spaetzchenthx ;-)
12:44:04jhMikeSthe thing that irks me the most is the scheduler boost control. how a thread blocks has nothing to do with boosting imo.
12:44:05Spaetzchenok, I'll try
12:44:23Spaetzchenyesterday I installed rockbox on my Sansa e250 :-)
12:44:46Spaetzchennow I have the problem that I have no backgroundpictures
12:44:50 Join norbusan [0] (
12:44:50preglowjhMikeS: agree wit that
12:44:55linuxstbjhMikeS: I'm not sure there are many fans of the scheduler boost control feature
12:44:57 Part norbusan
12:45:18Spaetzchenand no sound - wikipedia said that this software for people who has problems with their eyes (blind people)
12:45:35DefineByteah speech
12:45:38Spaetzchenbut I've got many games on it, but I don't need it ;-)
12:45:44Spaetzchenjo, sorry, speech ;-)
12:46:14Spaetzchenso I tried to deinstall - nothing happend - it said that he can't find the system ;-)
12:46:28Spaetzchenso I have no rockbox and not the normal player ;-)
12:46:32jhMikeSI'm wondering if boosts should just fizzle out if not explicitly maintained or are cancelled.
12:46:36Spaetzchenso my question is: what can I do? ;-)
12:47:15jhMikeSthat's another subject. the fix for voice basically reduced it all to the original protocol.
12:48:04preglowjhMikeS: hmm, that might work nicely
12:48:11preglowbut would of course stress placement of boost calls more
12:49:05DefineBytewhat player are you using?
12:49:54DefineByteyes :)
12:50:04SpaetzchenSanDisk Sansa e250
12:51:04B4gderSpaetzchen: what exactly does it say when you start it now?
12:51:17***Saving seen data "./dancer.seen"
12:51:22SpaetzchenFile not found
12:51:31DefineByteanyone with any experiance of recovering a sansa?
12:51:39B4gderSpaetzchen: and you did unzip a rockbox zip on the player?
12:51:40Spaetzchenbecause I deleted it ;-)
12:51:47B4gderdeleted what?
12:51:51Spaetzchenthe file
12:51:54B4gderwhat file?
12:52:12DefineByteyou deleted the /rockbox directory?
12:52:23Spaetzchenyes ;-)
12:52:34Spaetzchenbecause there was no pictures and no speech
12:52:44B4gderand now you have?
12:52:44Spaetzchenso I tried to delete everything that I can try it again
12:53:07SpaetzchenNOW I have the prob that I can't start ;-)
12:53:13B4gderso get into recovery mode and put back .rockbox, or re-install the original sansa OF
12:53:16SpaetzchenI can't deinstall
12:53:21SpaetzchenI can't reinstall
12:53:25B4gdersure you can, just read our guides
12:53:31linuxstbShouldn't it start the OF if you start the Sansa with USB attached?
12:53:42linuxstb(OF = original firmware)
12:53:50B4gderthat's the idea, I don't think it works everywhere
12:54:02linuxstbOr manually press a button to start the original firmware?
12:54:06Spaetzchenit doesn't starts, I don't know how to start in the normal
12:54:11B4gderpressing << starts the OF
12:54:17Spaetzchenok, mom
12:54:24B4gderSpaetzchen: you COULD read the manual you know...
12:54:39B4gdera crazy idea, I know
12:54:43SpaetzchenI read it, but I don't understand it
12:54:55DefineByteis the manual in german?
12:54:56SpaetzchenI did everything what the manual said
12:55:05Spaetzchena little one, yes
12:55:34SpaetzchenI did everything what the manual said - press enter etc. to install ....
12:55:34B4gderwe only have english manuals
12:55:41jhMikeSpreglow: I suppose but I think it wouldn't take too much given a sufficiently long expiration. I think giving cancelling control to the _last_ thread to request boost is all that's needed. not a per-thread record as it is now.
12:55:49B4gderSpaetzchen: I don't think it told you to remove the .rockbox directory
12:55:51Spaetzchenbut there wasn't pictures etc.
12:56:04B4gderSpaetzchen: rockbox is not "pictures"
12:56:09linuxstbSpaetzchen: There will only be a background image if you install one...
12:56:09Spaetzchenno, I removed it because nothing happend
12:56:20SpaetzchenI installed it
12:56:23B4gderreally nothing?
12:56:29linuxstbSpaetzchen: It sounds like you had a perfectly working Rockbox install, but didn't install the themes or voice files.
12:56:36SpaetzchenI think the games are ok ;-)
12:56:44B4gderthe games of nothing you mean then?
12:56:45Spaetzchenbut I can't do anything
12:56:56Spaetzchenso I tried to deinstall, but it doesn't happend
12:57:01*B4gder stops this and hides
12:57:09 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
12:57:13linuxstbYou then half-uninstalled it (not following the uninstall instructions in the manual), and stopped your Sansa from working...
12:57:19DefineBytebring on rbutil xD
12:57:20SpaetzchenI installed it, but I haven't seen and heard something
12:57:33linuxstbSpaetzchen: Did you try playing some music?
12:57:51SpaetzchenI have no music on the player because he is new
12:57:51linuxstbSpaetzchen: Did you download and install the voice files?
12:58:01DefineBytedoes/will rbutil uninstall rockbox?
12:58:33linuxstbDefineByte: Yes (I'm not sure which - the old WX rbutil had an uninstall option, I can't remember if that's in Qt yes, but I expect so)
12:59:01Spaetzchenso, I would like to try it again, but with help
12:59:08Spaetzchenand this is the reason why I
12:59:12SpaetzchenI'm here ;)
12:59:18linuxstbSpaetzchen: But anyway, turn your Sansa off (hold power for about 15 seconds I think), then hold <<, then turn your Sansa back on. It should start the original firmware.
12:59:19DefineByteshould prevent this particular problem from cropping up then. :)
12:59:30SpaetzchenI would like to have my old system or the new system with all things I need ;-)
12:59:46preglowjhMikeS: doens't sound like something that would take too long to try out either
12:59:54Spaetzchenyes, I'm in the original
13:00:22 Quit Rob222241 (Connection timed out)
13:00:25linuxstbSpaetzchen: Then there's no problem - copy .rockbox across and start again.
13:00:39Spaetzchenok, mom
13:01:05linuxstbWhen you downloaded the .voice file, did you correctly rename it?
13:02:06SpaetzchenI don't know what u mean
13:02:12DefineBytechange the name of the file
13:02:13linuxstbChange the name of the file
13:02:20jhMikeSpreglow: probably not...what I said isn't _quite_ it. I wanted to wait to establish this code before working priority inheritance and that boost control thingy.
13:02:26*linuxstb wins for proper capitalisation
13:02:35 Quit The-Compiler (Read error: 104 (Connection reset by peer))
13:02:50Spaetzchenno I didn't do this
13:02:57Spaetzchenand I deleted it
13:03:12Spaetzchencan u give me the link that I can download it again, please?
13:04:03*amiconn really isn't sure anymore whether we should use the 2nd core on PP at all, given the huge problems it causes
13:04:10*linuxstb agrees with amiconn
13:04:19preglowi am
13:04:23preglowthe ui suffers horribly
13:04:34preglowand given the nature of the clickwheel, we need a responsive ui
13:04:38 Join PaulJam [0] (
13:04:39linuxstbApart from in certain restricted situations like mpegplayer - where the app is responsible for obeying the restrictions.
13:04:40amiconnIn what way?
13:04:51*jhMikeS doesn't know which problems it causes
13:04:55linuxstbpreglow: Temporary boost whilst the UI is active would solve that - if code optimisation can't.
13:05:01jhMikeSthreaded programming is easy
13:05:09preglowgiven a sufficiently hungry codec, no it won't
13:05:26Spaetzchenok, and now dezip in the sansa?
13:05:29*petur spots austriancoder and waves
13:05:54amiconnDualcore seems to add the same amount of problems (at least) as preemptive threading would
13:05:58linuxstbjhMikeS: It's changing Rockbox from a simple co-operative multi-tasking kernel that more than one person can understand to something only you understand...
13:06:09preglowamiconn: it does some of them, but not all
13:06:20austriancoderpetur: hi
13:06:22amiconn...good enough for some serious headaches
13:06:30*jhMikeS doesn't see problems with preemtive threading and is quite at ease with it.
13:06:33preglow... on dual core targets
13:06:36SpaetzchenDefineByte what must I do now?
13:06:59 Quit GodEater (Read error: 113 (No route to host))
13:07:10B4gderaustriancoder: I think it would be good of you to respond to Dave's mail on the list since it was posted there
13:07:18preglowi don't think we should keep chanting the "kiss" mantra so much we start closing our eyes on features just because the code gains a certain complexity
13:07:21B4gderfor those who only read the list
13:07:30jhMikeSlinuxstb: I hope to comment it sufficiently and messing with that isn't for everyone. Basically it's either: understand it or leave it alone kind of thing.
13:07:32linuxstbpreglow: Then such a codec could use the COP. Or we run all codecs on the COP - requiring only the codec api to be multi-core safe, not the whole of Rockbox.
13:07:41DefineByteto be honest i've never installed voice files beofre. i'm sure someone else knows better than me
13:08:11B4gderjhMikeS: pre-emptive thread may be easy to do, but it opens a whole new can of worms for us
13:08:40Spaetzchensomeone said that I must rename this file and where I must put this file?
13:08:49preglowwe don't need or want pre-emptive threading
13:09:00amiconnjhMikeS: I really fear that dualcore will cause serious issue in other code than just the scheduler, and I probably have no idea how to deal with them
13:09:06jhMikeSB4gder: depends on the design choices. it can be a non-issue for most code. the code it affects most people don't even look at.
13:09:17*linuxstb is also curious to know the power consumption of running different combinations of COP/CPU at different clock speeds
13:09:23B4gderjhMikeS: I beg to differ on that
13:09:33B4gderthreads come back to byte
13:09:34 Quit ddalton ("leaving")
13:09:36B4gderbite even
13:09:44 Join ddalton [0] (
13:10:24jhMikeShmmm...then my brain is really odd
13:10:56preglowi like the way we can control the execution flow now anyway
13:11:02preglowno need for critical sections or other stuff
13:11:08linuxstbSpaetzchen: Does the manual have instructions for using voice files?
13:11:29preglowapart from disabling interrupts, that is...
13:11:33 Join GodEater [0] (i=c2cbc95c@rockbox/staff/GodEater)
13:11:33jhMikeSbut there is a need, and that's just cooperatively. it's not a crutch for improper locking.
13:11:50 Quit GodEater_ (Read error: 113 (No route to host))
13:11:54Spaetzchenyes, but it's only in english ;-)
13:12:02Spaetzchenso I don't know how to do this ;-)
13:12:07SpaetzchenI try it again ;_)
13:12:31jhMikeSif you don't have complete control over what called code does, it affords no luxury
13:12:53Spaetzchenok, rockbox is again on the player
13:12:57Spaetzchenbut what must I do now?
13:13:12B4gderSpaetzchen: find someone who can translate the manual for you
13:13:31jhMikeSI'd have no intention of getting fancy with every part. just certain subsystems which are rather easy to control.
13:13:55preglowbut i do agree with linuxstb as far as not needing to make every part of rockbox dualcore aware goes
13:14:03Spaetzchenthe problem is
13:14:08DefineBytei guess translating the manual for release would be a goal to strive for
13:14:10preglowbut i suspect a good part of the changes will be needed anyway even for limited support
13:14:11Spaetzchenthere stands only ONE sentense
13:14:28Spaetzchenbut there doesn't stands how to use it
13:14:34B4gderDefineByte: if we find people with enough boredom in their lives, sure
13:14:43Spaetzchenand the background isn't there, too :(
13:14:47linuxstbpreglow: Yes, I suspect that as well...
13:14:50jhMikeSpreglow: that would be overkill even though I can conceive of a construction to make UI multithread safe with no bunches of mutexes.
13:15:02DefineBytethere isn't a background by default
13:15:08DefineByteyou have to add one
13:15:12SpaetzchenI downloaded a theme
13:15:18Spaetzchenand I installed it
13:15:23DefineBytewhere did you put it?
13:15:23Spaetzchenbut I don't see it
13:15:31Spaetzchenin .rockbox
13:15:32preglowbut just closing our eyes on dualcore sure as hell isn't a way to go if we ever want to support ipods properly
13:15:37DefineByteit may need patches
13:15:38jhMikeSthere's certain places where avoiding threading actually makes for nasty workarounds.
13:15:43DefineBytewhat theme was it?
13:15:55pixelmaSpaetzchen: did you look into the "enabling speech support" paragraph in the installing section of the manual?
13:16:10SpaetzchenI did what there stands
13:16:30Spaetzchencopy the file in the .rockbox/langs
13:16:33SpaetzchenI did this
13:16:40B4gderbtw, we only have english voices for download
13:16:48SpaetzchenI know
13:17:08pixelmawhat file?
13:17:29Spaetzchen - here I found german
13:17:33pixelmawhere do you have it from?
13:17:39jhMikeSpreglow: the dual core SPC code runs nicely with echo running on both cores simulataneously and it was just a quicky thing.
13:17:57pixelmaSpaetzchen: those are outdated you can't use them currently
13:18:12B4gder... as is mentioned on that very same page
13:18:19Spaetzchenthis isn't good :(
13:18:21pixelma(with a current build that is)
13:19:04B4gderas I said "we only have english voices for download"
13:19:05jhMikeSand one major factor in properly supporting such threading reliably is a proper kernel synchronization object set.
13:19:14Spaetzchenhm, outdated stands at the end of this page
13:19:32B4gderSpaetzchen: then start over from the top again
13:19:40Spaetzchen-start +stands
13:19:44B4gderespecially this line: "Pick one of these if you are running an older daily build of Rockbox."
13:19:50jhMikeSheck, good threading use even fixes the emulator slowness and it runs just like the real thing
13:20:19 Quit class37 (Read error: 110 (Connection timed out))
13:20:38Spaetzchenmom, I'm searching someone who speaks english and german ;-)
13:20:56preglowproper sync primitives is a must anyway if you're going to use the other core at all
13:21:11 Join webguest94 [0] (i=521d0260@gateway/web/cgi-irc/
13:21:43pixelmaSpaetzchen: did you chose german as the language on your player?
13:21:48DefineByteyou need this voice file
13:22:05Spaetzchenand in the files I found on this site I showed
13:22:08Spaetzchenthere stand
13:22:14jhMikeSone big one is thread_wait, but there's also semaphore and events. these benefit single-core as well by helping get rid of while(cond) yield() stuff that blows up all the time.
13:22:31Spaetzchenand I downloaded this
13:22:35DefineBytethat won't work
13:23:00Spaetzchenonly the englisch one?
13:23:13DefineByteat the moment
13:23:16Spaetzchenand the english one talks only in english or tries he to speak german? ;-)
13:23:21pixelmaSpaetzchen: you can only use the file DefineByte linked you to - only that one is currently available
13:23:35B4gder"we only have english voices for download"
13:23:44B4gderanyone notices the echo?
13:23:50Spaetzchenok, the german one under this means what?
13:23:50DefineByteit's english or nothing currently
13:24:16Spaetzchenso the words are programmed or the letters?
13:24:41pixelmabut if you want to try out voice in general as the start you can have the text shown on german and have it talking in english
13:25:01jhMikeSI'm constantly fixing bugs related to yield patterns that vary from target to target. That's where this cooperative stuff breaks down to preemtive behavior. You may consider a "suprise" yield to be equivalent to preemption.
13:25:34 Quit Nimdae (Read error: 113 (No route to host))
13:25:57delYsidddalton: around?
13:27:59pixelmayou would need to download the english voice file (as it says in the manual), copy it to your player at the appropriate place and rename it to deutsch.voice - then after a reboot and if you're really chose the "deutsch" language, it starts talking (in english) but at least you'll have the menus and so on in german
13:28:16pixelma(as text)
13:28:25Spaetzchenoki, mom, I try it
13:28:53preglowjhMikeS: i've still got a feeling pre-emptive multitasking will hamper us more than help us
13:29:00DefineBytewhat does mom mean by the way?
13:29:41Spaetzchensorry, mom means in german Moment ;-) so I would like to say, wait a moment pls ;-)
13:30:37SpaetzchenI downloaded it what Byte linked, but
13:30:43SpaetzchenI can't use it
13:30:45amiconnjhMikeS: whie(cond) yield(); is nasty anyway, unless the condition is only true for a very short time. Otherwise while(cond) sleep(1); is much much better
13:30:58Spaetzchenmy program said that he can't find a program to use this file
13:30:58 Quit webguest94 ("CGI:IRC (EOF)")
13:31:03jhMikeSpreglow: no worry, it's still cooperative per core
13:32:05DefineBytejust put it in the right place. you dont do anything with it apart from renaming it
13:32:14jhMikeSamiconn: if it's multithreaded code, the condition may go true-false-true and no longer apply by the time the waiting thread runs again.
13:32:22amiconn(or better, sleep(0))
13:32:24delYsidddalton: regarding 7704, what do you think about just committing the stripped down patch you made? It doesnt have any impact on the english voice and/or lang file, just makes the talk functions available through plugin_api and splits out the talk_enum.h header.
13:32:40delYsidthis way, plugins could at least do stuff like talk_number and talk_id if we do not need any additional strings.
13:33:22jhMikeSsleep(0) will equal yield() as anything I've programmed makes a zero delay a simple giving up of remaining CPU burst.
13:35:14ddaltonI think it would be a good idea. Then we could just create a fix and commit that later.
13:35:18jhMikeSof course with these changes I'll evaluate all scheduler use globally.
13:35:49delYsidddalton: great.
13:36:15amiconnsleep(0) should sleep until the end of the current tick, unless a bug got introduced
13:37:03 Join Buschel [0] (
13:37:10ddaltonAre you going to commit?
13:37:11jhMikeSsleep(1) would actually accomplish that since the end of the current tick is when the tick changes.
13:37:32delYsidddalton: talk-plugin.diff looks simple enough anyway, can you commit it?
13:37:34 Join GodEater_ [0] (n=bryan@rockbox/staff/GodEater)
13:37:35Spaetzchenok, he speaks now with me in english ;-)
13:37:52DefineByteyou got there eventually :)
13:38:06ddaltonI don't have svn access
13:38:18delYsidddalton: ah, IC. thought you had.
13:38:34ddaltonhave you got it?
13:38:40delYsidno :-)
13:38:43jhMikeSamiconn: unless you want to round up so that sleep(x) could sleep x+1 ticks
13:38:44delYsidrasher: ? :-)
13:39:04ddaltoncan you code?
13:39:04amiconnjhMikeS: It used to do that...
13:39:11Spaetzchenso now I did the right thing to install the speech? I have the german settings and the english voice for this, is this right now?
13:39:30delYsidddalton: me? I wrote this patch, yes...
13:39:50jhMikeSyeah, not really proper imo. if you want to sleep x+1 ticks, why not write sleep(x+1) ?
13:39:54 Join akaias [0] (
13:39:58DefineByteyes, that's right. it's the best you can do at the mom :)
13:40:18Spaetzchencoooooool, thx a lot
13:40:26Spaetzchenand now to my second problem ;-)
13:40:29Spaetzchenthe themes ;-)
13:40:38ddaltonok can you fix p7705 if you have time? It adds voice for battery bench but I would like to see how you check if voice menus is enabled. I don't really have time to update
13:41:09ddaltondelYsid: so how should the voice file be done do you think?
13:41:14amiconnjhMikeS: It's just nasty to change behaviour in that way if it was different before
13:41:18ddaltonin an if?
13:41:27ddaltonif (what you said before)
13:41:40amiconnI think some of the newer problems might be due to this behaviour change
13:41:42delYsidddalton: yes.
13:41:47ddaltonso if(rb->voice_menus_enabled())
13:41:58jhMikeSthere was no behavior change in SVN
13:42:06ddaltonok how should we do the voice file do you think?
13:42:19delYsidddalton: no idea yet :-(
13:42:51delYsidbut getting 7704 (the tsecond patch) committed would be a first step to get things moving.
13:42:53ddaltonI see what your saying about using strings from english.lang but I think we could just re do them in plugin.voice for example
13:43:02delYsidbut I have no idea yet how to code the secondary voice file thing
13:43:09ddaltonyep and then write the fix
13:43:37jhMikeSI employed the change in a later revision my code. I thing most code assumes sleep(10) means "sleep 100ms" not 110ms.
13:44:00SpaetzchenI tried to take the WVA-theme into my player, but it doesn't work
13:44:01DefineBytewhat's the name of the theme you installed, Spaetzchen?
13:44:29amiconnjhMikeS: There was. At least back in times of the simple round-robin scheduler, sleep(0) slept until the end of the current tick
13:44:42DefineBytewhat's the name of the theme you installed, Spaetzchen?
13:45:04amiconnjhMikeS: 100 or 110ms, both won't be exact
13:45:26delYsidddalton: but I think we will need a secondary voice file and load both (the english.voice and plugin.voice), because there is potentially a lot of duplication. Think about talk_number or talk_value
13:45:29amiconnThe old behaviour let it sleep for 100..110ms, current svn lets it sleep for 90..100ms instead
13:45:30DefineBytedidn't mean to post that twice
13:45:31jhMikeSThis revision is snappy. Thread wakeups never wait. Latencies are cut
13:45:58SpaetzchenI found it there
13:46:04Spaetzchenu asked for the theme ;-)
13:46:06jhMikeSamiconn: I remember seeing tick = current_tick + 1 in current SVN
13:46:22amiconnjhMikeS: I'm referring to the tick granularity
13:46:46jhMikeShow would that have changed?
13:47:09DefineBytethat theme should work on the standard build
13:47:20ddaltonI got to go but maybe plugin.voice and then we do something like lang v2 and if certain players don't have certain plugins there will be less clips.
13:47:27ddaltonProbably to many voice files!
13:48:04ddaltonso just plugin.voice for all targets maybe and then we just add everything in to that for the plugins including the numbers
13:48:07DefineByteSpaetzchen: how did you install it?
13:48:11ddaltonand letres
13:48:12jhMikeS100 should actually be 95-105 ms and average 100 over the long term
13:49:11jhMikeSbtw: round-robin is a term used for preemtive multitasking. the correct term is FCFS (first-come-first-serve). </pedantic>
13:49:12amiconnNo, it's not
13:49:22 Quit ddalton ("leaving")
13:49:37amiconnIt enters sleep state anywhere within a tick period, but always wakes up at a tick period boundary
13:49:54Spaetzchenhm, I can't remember, I think I dezipped it in the sansa
13:50:14amiconnAnd you're right, the +1 is still there so sleep(0) works as I expected
13:50:23DefineBytewhere? in the root?
13:50:36Spaetzchenin .rockbox
13:50:37DefineByteand then you selected it in 'browse themes'?
13:50:46DefineBytenot there
13:51:03 Quit advcomp2019 ("Leaving")
13:51:20DefineByteunzio it to the same place you unzipped rockbox
13:51:46Spaetzchenyes, I unzipped it in J:
13:51:50Spaetzchenthis is my player
13:51:59DefineBytewell that's right then
13:52:07Spaetzchensorry, I'm a little bit confused ;)
13:52:17Spaetzchenbut it doesn't work
13:52:26jhMikeSthe processor may wake at the tick period boundary, but the thread may have threads ahead of it doing something but that's usually a couple ms delay tops for a turnaround.
13:52:30DefineBytedid you select it in 'browse themes' (or the german equivelent)?
13:53:15SpaetzchenI don't know where it is
13:53:24SpaetzchenI think I took it in .rockbox/themes
13:54:01jhMikeSThe scheduler still executes ready to run tasks in FCFS order if priority isn't involved. just thread->next since the linked lists are circular.
13:54:39 Quit PaulJam (".")
13:55:02pixelmaSpaetzchen: what did happen - did at least something change (background, icons)?
13:55:28Spaetzchennot really ;)
13:55:50LinusNSpaetzchen: did you unzip it to J:\ or J:\.rockbox\themes?
13:56:03SpaetzchenI didn't changed the names, I only changed the picture in background - but the same sizes
13:56:31SpaetzchenI don't remember, but I think I unzipped it in J:\.rockbox ;-)
13:56:38LinusNthat is wrong
13:56:51LinusNyou are supposed to unzip it to J:\
13:57:04DefineByteSpaetzchen: do you have WVA.cfg in .rockbox/themes?
13:57:28 Quit GodEater_ (Connection timed out)
13:58:03Spaetzchenyes I have this file in themes
13:59:18DefineBytecan you go to "General Settings > Display > Browse .wps files" and slect it
13:59:32DefineByteselect it
13:59:48Spaetzchenin the player without or with computer?
13:59:59DefineBytein the player
14:00:15DefineBytedisconnected from the computer
14:02:22SpaetzchenI've only iCatcher, rockbox_default, Rockboxed and UniCatcher
14:02:26Spaetzchenwhich one?
14:03:03DefineBytethey're the default themes. you don't want any of those
14:03:44DefineBytefrom the main menu go to browse themes
14:03:50DefineByteis it their?
14:04:04 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
14:04:42Spaetzchenwhere I can find default themes?
14:04:43 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
14:05:10DefineBytecan you go to 'Browse Themes' in the main menu first
14:05:58pixelmaEinstellungen > Zeige Themen ;)
14:06:16Spaetzchenaaaaaah, thx, pixel :-)
14:06:28Spaetzchenok, I found this
14:06:32jhMikeSamiconn: another thing I changed is that packed statearg. timeouts will be the full 32 bits as expected and state kept is a separate byte which all the processors efficiently can load and extend.
14:06:36Spaetzchenand I took WV
14:06:38 Join Genre9mp3 [0] (n=yngwiejo@rockbox/contributor/Genre9mp3)
14:06:43Spaetzchenbut nothing new
14:06:52DefineByteplay a song
14:07:09DefineByteand it should show
14:07:11SpaetzchenI've no songs on it ;-)
14:07:27DefineByteyou'll have to wait then :p
14:07:37Spaetzchenmom, I try ;-)
14:08:03B4gderSpaetzchen: mom is short for mother in english
14:08:18DefineByteamerican english/
14:08:42B4gderwell, we're all americanized outside britain anyway
14:08:44pixelmawell that theme has also a different look in the browsers and menus if it's not broken somehow
14:09:05DefineBytei'm inside britain! x)
14:09:19B4gderDefineByte: you use use 'mom' as short for "one moment" ?
14:09:28pixelmaand I've never used "mom" in german for Moment ...
14:09:33DefineBytei don't
14:09:39B4gderI didn't think so either
14:09:47Spaetzchensorry ;-)
14:09:48jhMikeSamiconn: bringing things back in my code to the usual sleep behavior is just about 3 lines change if that's preferred.
14:09:52SpaetzchenI'm too german ;-)
14:09:59Spaetzchenso I took 3 songs on the player
14:10:06SpaetzchenBUT I don't know where to find
14:10:13B4gderSpaetzchen: pixelma happens to be german too...
14:10:24DefineBytewhere did you put them?
14:10:37pixelmamaybe regional differences ;)
14:10:52B4gderor age? :-)
14:11:02SpaetzchenI put it in music
14:11:04DefineBytehey, maybe he's swiss
14:11:11DefineBytedid he say he was german?
14:11:16SpaetzchenI can understand swiss :)
14:11:18DefineByteanyway... :)
14:11:55DefineBytethen from the main menu go to 'files' (or whatever it is in german) and browse to the files
14:12:02pixelmafor what it's worth, that theme works fine in a Sansa simulator here...
14:12:20DefineBytehey, it's got to be worth something
14:12:37B4gdergiven the track history, I bet Spaetzchen only tell us half of what he actually does
14:12:52B4gder... so this will never work
14:12:55Spaetzchennot he, she ;-)
14:13:12SpaetzchenI have no music in "files"
14:13:12B4gdersorry, didn't spot that! ;-)
14:13:26pixelmaa male would probably not pick that nick ;)
14:13:48B4gderwell, my german is... eh, lacking a bit
14:13:48DefineBytewe'd probably have to speak german to know that xD
14:13:50Spaetzchenbut I can't find the music
14:14:09DefineByteyou didn't put the music in .rockbox did you?
14:14:18Spaetzchenunder root I found it
14:14:33Spaetzchenand there is no theme, no picture, nothing :(
14:14:46*B4gder offers DefineByte the "I have patience of steel" badge to wear for the rest of the day
14:14:52 Join DerPapst [0] (
14:15:01DerPapstgood morning :)
14:15:21DefineBytethanks *attaches badge with pride*
14:15:27*petur yawns
14:15:31Spaetzchenand now?
14:15:57DefineByteone mom (x))
14:16:08Spaetzchenyes I have only one mom ;-)
14:17:08pixelmaSpaetzchen: I don't know I had no problem with that theme - just unzipped it to the root (of the simulator in my case) and selected it through "browse themes" - there it was
14:18:17Spaetzchenok, how can I delete these theme and install it again?
14:18:57 Join Kupopop [0] (
14:18:58DefineBytedo you have a directory called .rockbox IN .rockbox?
14:19:04*B4gder calls for theme support in rbutil...
14:19:29DefineByteit would be great if you could drag a theme onto rbutil and it automatically installs
14:19:50pixelmathe directory you put the music in - is that the "Music" directory that the Sansa original uses (I seem to vaguely remember that this one is marked hidden by default, could be wrong though)
14:19:53Spaetzchenyes for women like me ;-)
14:20:03B4gderDefineByte: I believe it'll even interact with the theme site to allow downloads etc
14:20:04dionoeadoesn't rb util already install themes ?
14:20:21Spaetzchenno there is no rockbox in rockbox
14:20:25tumusansa finds music on any directories
14:20:31tumujust tested it few hours ago
14:20:32jhMikeSpreglow: do you feel it's better to have 1% of the code sophisticated to make 99% of the code simpler or the other way? Sometime I get the impression the other way is preferred here. I also point to queue_send which paid for itself at twice over in bin size when first added.
14:20:40DefineByteno idea where you may have out it then
14:21:39pixelmaSpaetzchen: you have to be accurate: is there a "rockbox" or a ".rockbox" directory there - the . at the beginning is important
14:21:46 Join norbusan [0] (
14:22:17Spaetzchenits called .rockbox with ...
14:22:28 Part norbusan
14:23:05 Join miepchen^schlaf [0] (
14:23:37pixelmahmm... should be alright then - and inside it there are other directories like "themes", "wps", "fonts" and "backdrops"?
14:23:45DefineByteas far as rbutil's theme installation; i'd still like to be able to drag themes onto it. if only to qucikly test my own (outside the sim :))
14:24:25B4gderDefineByte: I'm sure your contributions to bring that feature will be appreciated!
14:24:41DefineBytei'm sure they would be :D
14:24:53pixelmaSpaetzchen: and in "backdrops" for example, is there a "WVA.bmp"?
14:25:00Spaetzchenyes :)
14:25:34DefineByteis WVA.wps and a WVA folder in '.rockbox/wps'?
14:25:42 Nick idnar_ is now known as idnar (i=mithrand@unaffiliated/idnar)
14:25:55DefineBytesee, this is the kind of stuff i want rbutil to generate a lsupport ticket for xD
14:26:07 Quit Hellmark ("Leaving")
14:27:47DefineBytei guess rbutil could verify the directory structure itself.
14:28:43Spaetzchennow it is in there :)
14:28:59DefineBytehow on earth did you install it?
14:29:08DefineBytesounds like you didn't just unzip it
14:29:30SpaetzchenI unzipped it, but only in the player ;-)
14:29:35Spaetzchennot the right way :)
14:29:41Spaetzchenbut now it's there :)
14:29:59DefineByteand all the other files are in the right place too?
14:30:17SpaetzchenI think so
14:30:48DefineBytethere should be 4 files in .rockbox/icons
14:31:17SpaetzchenI'll look :)
14:31:30DBUGEnqueued KICK DefineByte
14:31:30DefineByteand tango_small_viewers.icons
14:32:25 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
14:32:49SpaetzchenI've only one in icons
14:32:57Spaetzchenits called viewers
14:33:30DefineByteunzip again
14:33:41Spaetzchenok, now I have everything in it :)
14:33:54DefineByteunzip .rockbox in into the root directory (j:)
14:35:14SpaetzchenI unzipped it again in J:
14:35:30DefineByteand can you select it?
14:35:51Spaetzchenmom, reboot
14:35:55CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
14:35:55*DerPapst just returned from dinner...
14:35:56Spaetzchensorry ;-)
14:36:04Spaetzchensec, reboot ;-)
14:36:53Spaetzchennow it looks good :)
14:37:09DefineByteso you're all sorted? that's great
14:38:00 Join MarkVeinot [0] (
14:38:22Spaetzchenyes, it lookes very good, thank you all a lot ;-)
14:38:31Spaetzchenbut I've a little question at the end ;-)
14:38:59Spaetzchenmy display goes out after 5 sec
14:39:10DerPapstyou mean the backlight?
14:39:11Spaetzchenwhere can I change this?
14:39:14preglowjhMikeS: i feel it's only natural for a project the size of rockbox to have code not everybody understands at first glance anyway
14:39:24MarkVeinotIs there a location to download the latest builds of the sansa bootloader? I would like to take advantage of the reduced USB timeout that was introduced yesterday.
14:39:27pixelmaSpaetzchen: try to search for the backlight settings in the manual ;)
14:39:30DefineBytesettings>display>lcd i think
14:39:32preglowit's not like anyone here could jump in bugfix the eq in five minutes
14:39:39DerPapsteinstellungen > ... dunno
14:40:17DefineBytebut you want it to turn off or it will use a lot of the battery
14:40:38pixelmaEinstellungen > Grundeinstellungen > Anzeige > LCD-Einstellungen > Beleuchtung
14:40:55Spaetzchenthx :)
14:40:58DefineBytethat's what i said.... ^^
14:41:02Spaetzchennot its perfect
14:41:08jhMikeSpreglow: indeed...I'd include myself in that category...and as a supplement to that, I wouldn't attempt to bugfix the eq without knowing it inside-out first.
14:41:13Spaetzchenthank you a lot
14:41:16linuxstb_preglow: True, but eq is different - it's one feature that nothing else depends on. If the worst came to the worst, we couldjust rmeove it from Rockbox. We couldn't do that with the kernel, (or playback.c, etc)
14:41:26Spaetzchennow I can to to work :-)
14:41:37DefineByteenjoy Rockbox Spaetzchen
14:41:40Spaetzchenthanks a lot and I hope I see you all again :)
14:41:41Spaetzchenthx :)
14:41:46DerPapstyou go to work at 14:50?
14:41:58pixelmaand read the manual a bit
14:41:59Spaetzchenno, at 15:30
14:42:10Spaetzchenand I must go in 20 mins
14:42:12preglowlinuxstb_: true, but we're approaching a point where we _have_ to add some complexity to it in order for rockbox to expand in the same way it has
14:42:19DerPapsthave fun ;)
14:42:20Spaetzchenthx and cu :-)
14:42:23DefineBytethe manual could be a big problem for non-english speakers
14:42:24Spaetzchenthx :)
14:42:25preglowwe can't just count on never seeing another dual core cpu
14:42:30linuxstb_preglow: Also, I think we should learn from past failings - there are lots of bits of Rockbox which are in effect unmaintained. All I'm saying is that we should do whatever we can to avoid that situation getting worse.
14:42:31 Quit Spaetzchen ("CGI:IRC (EOF)")
14:42:36jhMikeSlinuxstb_: but the DSP couldn't be removed for sure. adding the ASM limited the number of possible maintainers but the benefits were outstanding.
14:43:18preglowlinuxstb_: again true, but i don't think keeping everything so dead simple everyone can understand it is a good solution
14:43:21preglowthat way lies horrible hacks
14:43:29 Quit aliask ("ChatZilla [Firefox]")
14:43:49 Quit GodEater ("CGI:IRC")
14:43:52jhMikeSpreglow: precisely, the hacks that add the complexity and bugs to the other 99%
14:43:57 Quit Nico_P (Remote closed the connection)
14:43:59linuxstb_I don't think we disagree - I'm just saying any complexity needs to be justified.
14:44:11preglowand i think you're right in rockbox having a bit too many of them too :/
14:44:31preglowsure, and i don't think you can get any more justified than "good dual core support"
14:44:33 Quit ptw419 ()
14:44:56*jhMikeS say: (K)eep (I)t (S)imple (S)tupid (M)aybe (Y)et (A)s (S)implicity (S)uffices :D
14:45:30preglowa bit on the fancy side of semantics :>
14:48:24preglowbut anyway, i would be surprised if the majority of jhMikeS' patch isn't just stuff that already should be familiar to any os level programmer
14:49:09preglowmy biggest worry is quite frankly that we're going dual core on a chip we have no documentation for
14:50:11jhMikeSthe biggest headache has been the cache actually...but it seems that overlapping cache lines is the reason.
14:50:28linuxstb_jhMikeS: What's the state of your dual-core work? Is the failure of the PP5020 to play fair still holding things up?
14:50:29preglowoverlapping cache lines???
14:50:53jhMikeSlinuxstb_: yes...that's it. it runs and passes the tests on e200
14:51:19 Join Febs [0] (n=chatzill@
14:51:21***Saving seen data "./dancer.seen"
14:51:29linuxstb_jhMikeS: And you still don't have a 5020 to test on?
14:51:50jhMikeSpreglow: when processor A has it's own copy of an address range that B also has a it's own copy of but they've each modified or read that range in their own way at different times.
14:52:01jhMikeSlinuxstb_: one should be coming up.
14:52:16preglowi don't get how the hell you find out what the hell is wrong with stuff like this :P
14:52:31jhMikeSpreglow: guess?
14:52:35preglowi would have scratched my ass for about a week, then given up
14:52:51B4gderjhMikeS is a secret PP employee working under cover here
14:53:12linuxstb_One that's lost the PP5020 datasheet and only has the PP5022?
14:53:12tumutainting rockbox with IP
14:53:24jhMikeSI theorize for weeks and an answer usually pops in my head when I'm outside smoking
14:53:35B4gderluckily, the rest of us are FBI agents under cover trying to find the actual rockbox devs
14:53:51amiconnpreglow: The overlapping cache lines issue is pretty obvious once you found it to be the reason
14:54:01*jhMikeS is CIA and Secret Service with Top Secret clearance
14:54:23preglowthen you can find mrh for us and force him to code for free 24/7? :>
14:54:34jhMikeS...and possesses an alien scanning device found at the Roswell crash site
14:54:46preglowwhat, sansa doesn't support microsdhc ?
14:54:57preglowhow cool :P
14:55:04B4gderhehe, yes
14:55:14jhMikeShehe...but we don't don't either ... because we have no hologram sticker
14:55:26preglowthat's just weird, i would have thought they'd be jumping at coding support for that
14:55:43B4gderpreglow: they probably rather force people to buy a new player for it
14:55:53preglowdo they have any?
14:56:02B4gderyes i think so, at least in the pipe
14:56:40DefineBytehow easy is it to get hold of a machine cabable of producing holograms?
14:56:46jhMikeSif we paint SDHC on the side of the players by the slot, it will be official and commence to operate properly
14:57:22preglowa business idea presents itself! we should demand users buy stickers in order to use feature upgrades!
14:57:25DefineBytemaybe the player can sense whether the sticker is official or not
14:57:26jhMikeSif you can build a scanning/tunneling microscope from legos, I'm sure one could be fabricated in the kitchen (I really think so)
14:57:33preglowa gapless playback sticker, an eq sticker, an mpegplayer sticker...
14:57:52jhMikeSlol...I'm sure rb will need quite a set of them
14:57:52DefineBytecrossfeed sticker please
14:58:55jhMikeShey, it's a way to bring funds in and make player updates official
15:01:37preglowi'm a genious!
15:01:42 Join RudMan [0] (n=RudMan@
15:02:07*DerPapst wants a sticker too
15:02:21 Quit iamben (Read error: 104 (Connection reset by peer))
15:02:47DerPapsta long battery life one please :)
15:02:57DefineBytewith an iPod crossed out
15:03:15preglowa sticker for mp3 players would actually be kind of cool :>
15:03:29DefineBytei'd buy one
15:03:43preglowlike one covering the entire back of this smudged up nano i have
15:04:12DefineBytei'd like one for my portable amp too
15:04:52DefineByterockbox should open a cafepress store or something
15:05:11jhMikeSyou know, I don't get why a multiprocessor config would need multiple caches. using one shared one would always be coherent among all said processors.
15:05:43preglowjhMikeS: yet they almost never are configured like that
15:05:50DefineByteis that official?
15:06:05krazykitread the announcements forum ;-)
15:06:19preglownow all we need is a prettier logo :>
15:06:20DefineBytebah, read?
15:06:52DefineByterun a competition
15:07:00DefineBytewin a gigabeat
15:07:03DefineByteor sansa
15:07:06DefineByteor something
15:07:15amiconnThe current logo is fine imo, I wouldn't want another
15:07:26jhMikeSpreglow: unless they bus snoop, they should be. less conflict would exist. ok, patent application goes out tomorrow. ;)
15:07:28DefineByteokay old man, stuck in his ways
15:07:29KupopopI like the logo
15:07:29DerPapstor a daily rockbox update guarantie
15:08:08 Join iamben [0] (
15:08:15DerPapst*warranty :P
15:08:24linuxstb_"We guarantee this Rockbox was made freshly this morning"
15:08:31preglowjhMikeS: i think the newer multicore amds have a shared level 3 cache, but that's the only shared cache i know of in multicore design
15:08:47DerPapstlinuxstb_: exactly :D
15:09:07DefineByte"We guarantee this Rockbox was probably made freshly this morning"
15:09:12DefineBytebe careful there
15:09:32linuxstb_I forgot to add "... or your money back."
15:09:43Zagorusing only organic whole-grain bits
15:09:52DefineBytehow about 110% refund?
15:09:56DerPapstit should be possible t do at least one commit per day.. even if it's only code policeing ;)
15:10:16jhMikeSpreglow: as long as they're closely packed on the same die I can't imaging it should be much trouble. I wonder if it's easy to mirror-image a circuit on a wafer and have it function identically.
15:11:14JdGordonDerPapst: that means 3 per day... 1 commit and 2 oops ones :)
15:11:19DefineBytethat cafepress link should be on the frontpage. (it isn't, right x))
15:12:14DefineBytethe mug looks quite good.
15:12:16*jhMikeS is getting too far out right now
15:12:31linuxstb_JdGordon: Any progress with your e200tool experiments, or are you doing other things?
15:12:43JdGordonabout to have a nother go..
15:13:00austriancoderJdGordon: what experiments?
15:13:05*JdGordon is more tired and less sober than last night, so doesnt expect much :p
15:13:17*linuxstb_ points austriancoder to all the e200r discussions over the past few days
15:13:17JdGordonaustriancoder: trying to get something safer for e200r
15:13:21jhMikeSpreglow: out of nescessity, the sim threading will function identically to core threading...blocking and all. it could even be used as a simplified reference to the kernel.
15:13:26DefineByteif a beer glass ias added i'm there
15:14:17linuxstb_austriancoder: Basically, JdGordon is working on an application to be uploaded via e200tool in manufacturer mode which will patch the bootloader stored on disk. That part of the disk is hidden via Sansa's UMS driver.
15:14:44austriancoderlinuxstb_: ahh ... i see
15:15:29 Quit Nick_Brackley ("ChatZilla [Firefox]")
15:15:35preglowjhMikeS: hmm, that makes me wonder, shouldn't we really be having more bugs with the sim due to the pre-emptive multitasking?
15:15:48jhMikeSthreads are serialized there
15:16:09preglowexplain :>
15:16:32jhMikeSbasically it simulates cooperative multitasking by running preemptive threads in turn
15:16:50preglowis that so
15:16:56jhMikeSsure is
15:17:21MarkVeinotIs there a location to download the latest builds of the sansa bootloader?
15:17:22JdGordonlinuxstb_: good news... dont know if you read before about my testing method... but with ata_init() i get 2 different flashing speeds which means we might be lucky :)
15:17:50amiconnpreglow: It has always been like that
15:18:06jhMikeSbut one thing in the sim should run preemtively: the disk io. I have a following update for that and the sim runs so much better. It sort of acts like it has a DMA channel.
15:18:53preglowi would welcome the sim running better, i want to use it as my main music player
15:18:56preglowbut it's too sluggish here
15:19:06linuxstb_JdGordon: Is this with my original 2 changes applied?
15:19:35linuxstb_JdGordon: I only briefly looked at the Sansa ata_init(), but can it actually fail?
15:19:37jhMikeSit employs yielding within rockbox threads while an IO does the read/write at the same time. huge improvement.
15:19:46preglowJdGordon: ata_init()? what're you working on?
15:19:55qwmany seeking in mpegplayer yet?
15:19:57JdGordonlinuxstb_: dunno.. but at least its running which is one up on this morning
15:20:15linuxstb_How big is bootloader.bin now?
15:21:00JdGordon40048 bytes
15:21:23jhMikeSpreglow: all the lag during disk access just goes away, mpegplayer plays video smoothly at all times and this whole patch is really just kernel improvement in general
15:21:33linuxstb_JdGordon: Sounds promising. Hopefully we can go all the way up to 112KB (from 40004000 to the end of IRAM) if we need to.
15:21:46jhMikeSbut the IO updates would be committed separately since it's sort of outside the scope
15:21:56DerPapsti wonder.... afaik rockbox doesn't allow anonymous contributions.... what is about the e200tool if JdGordon succeeds?
15:22:16linuxstb_JdGordon: Ideally we would want to append the Rockbox bootloader mi4 to the code you're uploading, so the app can do the sansapatcher stuff as well. But that's definitely for v2.
15:22:45JdGordonDerPapst: its not doing anything special.. its just using libusb so should be easy to reimplement...
15:23:04jhMikeSpreglow: changes to most files are just tiny compatibility changes (like a rename from struct event to struct queue_event to make way for actual event sync primitives)
15:23:09 Part DefineByte
15:23:34DerPapstor MrH decides to pick some random name :P
15:24:55jhMikeSpreglow: the sim could probabaly compile and use the actual rockbox kernel the way it's set up
15:24:58preglowyes, i look forward to the day Eiben Scrood commits code
15:25:19preglowjhMikeS: would be nifty :)
15:25:51jhMikeSthread.c of course not shared - too much SDL stuff
15:27:12*jhMikeS will now shut up since everyone must be bored with threads now :)
15:27:45JdGordonATA works :D
15:27:58*JdGordon just booted rockbox from manufac mode
15:28:25JdGordonwhich promptly rebooted because it found the usb connection :p
15:28:45jhMikeSJdGordon: how much worked as pasted? or how much didn't?
15:28:54 Quit MarkVeinot (Remote closed the connection)
15:28:58JdGordonusing svn ata-e200.c
15:30:02jhMikeSthought it required to be size-reduced
15:30:03preglowwhat're you doing?
15:30:20linuxstb_JdGordon: ;)
15:30:31 Quit CaptainSquid ("Miranda IM!")
15:30:51JdGordonjhMikeS: we were mislead by the asm from e200tool...
15:31:06JdGordonso we dont appear to have it that problem again just yet
15:31:27JdGordonpreglow: trying to get a safe install method for the e200r
15:31:48*jhMikeS puts an SDHC sticker on e200-tool
15:35:09jhMikeSso I haven't been following too close lately. e200r actually can boot and run rb now?
15:35:45tumufrom manuf mode atleast
15:35:54B4gderno, for real too
15:36:00jhMikeShey, the iPhone's been both JTAG'd and cracked, so let's get going :)
15:36:43*DerPapst downloads iphone
15:38:53jhMikeSB4gder: it's replacing the original bootloader with one that makes no sig check?
15:39:16B4gderor rather brings back the sig check flaw
15:39:24jhMikeSI thought you didn't know where it was
15:39:32B4gderwe figured it out
15:39:43B4gder== on the same place as the vanilla ones
15:39:55B4gderthe R models just hide the second partition when shown over usb
15:40:20B4gderhidden for real you could say ;-)
15:41:46*jhMikeS doesn't understand why the companies just don't give up the futile efforts ... The walls of Jericho and all... :)
15:42:16B4gderin this case, they probably try to keep their DRM hidden or safe or something
15:42:19tumuhave to do even lame attempt to use dmca
15:43:50JdGordonanyone want to have a guess as to why the lcd is useless in manufac mode, but once rockbox is loaded from disk it works? (the init sequence is all ifdef BOOTLOADER)...
15:44:23jhMikeSbecause it's actually initialized already but not used and no backlight?
15:45:41JdGordon... but the debug messages at the end of the bootloader should be visiable, but they arent shown untill rockbox is loaded..
15:45:59JdGordonI checked by sticking a while(1); before the return in the bootloader main()
15:46:13jhMikeSframebuffer is still the OF framebuffer and needs to be set to something else probably
15:46:57JdGordonthats done in lcd_init()
15:47:20jhMikeSIt's already called by the time mm starts?
15:47:25 Join webguest54 [0] (i=55678507@gateway/web/cgi-irc/
15:48:09 Quit webguest54 (Client Quit)
15:48:10jhMikeSlcd_init_device does change it over but not until rockbox itself boots
15:48:20 Join nicktastic [0] (n=nick@unaffiliated/nicktastic)
15:50:59 Quit sslashes ("leaving")
15:51:10 Quit hcs ("Leaving.")
15:52:49JdGordonjhMikeS: doesnt line 411 do that?
15:52:55JdGordon412 even
15:54:05jhMikeSno, line 427 for rockbox boot
15:55:08JdGordon? the #else is !BOOTLOADER
15:56:02JdGordoni want to get the lcd actually working in the bootloader...
15:56:50 Join The-Compiler [0] (
15:57:23 Join sslashes [0] (i=sslashes@
15:57:30jhMikeSso the BOOTLOADER part is being called already? not quite following here.
15:58:23tumusomeone should also take a look at the lcd corruption on older (?) sansas
15:58:37JdGordonits going mm -> our bootloader -> rockbox (from disk)... no lcd works untill the rockbox starts running.. but I need lcd in our bootloader... so yeah that code should have run by then
15:58:38B4gdertumu: it would be newer sansas
15:58:42jhMikeStumu: patch committed today
15:58:49LinusNtumu: eh, we fixed that today
15:59:05tumuhave missed it
15:59:22B4gderits been fixed for aaaaages :*)
15:59:34tumuand now you tell me that!
15:59:35*JdGordon opening can of worms... buuuutt.....
15:59:47 Join skate3214 [0] (
15:59:58jhMikeSJdGordon: And the bl is just our plain vanilla one? maybe some other register is only being set by firmware and not bl :\
15:59:59JdGordonthere is a comment "All this magic worked out by Mrh"... how is that OK with our no anon code?
16:00:21B4gderJdGordon: because we use information from other sources than just devs
16:00:50B4gderand yeah, there are some gray areas
16:00:51JdGordonso that wasnt a striaght copy/paste of his code?
16:02:14JdGordonjhMikeS: could it be the resetting all devices in system_init() which is ifndef bootloader?
16:03:26jhMikeSthat really only was needed to take care of audio problems...lcd worked before that was added.
16:03:52JdGordon.. well its not that :p nice and quick to test
16:04:49 Join jhulst [0] (n=jhulst@
16:04:59jhMikeSJdGordon: did I ask? Does it show anything when holding to the light? If the AS3514 isn't initialized, you'll not get any backlight.
16:05:45JdGordonits very black...
16:06:04 Quit qwm ("leaving")
16:07:35 Join bluebrother [0] (i=r1s98BOf@rockbox/staff/bluebrother)
16:07:36jhMikeSit's verified that the BL section is the one being compiled? hrm.
16:08:33JdGordonhmm.. you may be right... im either dreaming.. or there is a very faint numbers on the lcd
16:08:39*JdGordon looks for better light source
16:08:56jhMikeSit of course won't write to lcd_driver_framebuffer if the backlight isn't on
16:09:07 Quit obo (Read error: 110 (Connection timed out))
16:09:29tumuhm, is getting current build enough for the lcd fix?
16:09:45jhMikeStumu: should be
16:10:00tumustill seeing corruption
16:10:07jhMikeSlatest, not daily though
16:10:52tumuyeah, latest
16:10:55 Join elinenbe [0] (
16:12:01jhMikeShmmm...yet another variant?
16:12:20tumui have e280
16:13:04jhMikeSthat's 8 gig?
16:13:36*austriancoder too.. but i dont see any corruption
16:13:39jhMikeSit was tested on two versions of an e280
16:14:24 Join BHSPitMonkey [0] (n=stephen@
16:14:30tumulooking closely, it seems rockbox displays more lines at top than seen during the sandisk boot logo
16:16:40tumuand some pixels change at the top while using the wheel, can't recall that happening before
16:16:57tumupreviously it was just static corruption
16:18:11JdGordonseems the bin size might be about 57000 bytes
16:20:31*jhMikeS starts to develop a pentagram-shaped rash with the words "Portal Player" encircling it
16:20:47tumujhMikeS, i have the euro version without radio if that helps
16:21:04JdGordonjhMikeS: you said the backlight wont work if the as3514 isnt init?
16:21:09 Part LinusN
16:21:17jhMikeStumu: hmmm...I'll have to ask about that
16:21:42*austriancoder has also euro without radio
16:22:02 Join elinenbe_ [0] (
16:22:20jhMikeSJdGordon: most likely not if at least the i2c isn't ready
16:22:40tumuhmm, the corruption disappeared
16:23:02jhMikeStumu: did you do something in particular?
16:23:18tumuno, just regular use
16:23:30tumumight have been the screensaver, testing again
16:23:50JdGordonjhMikeS: just the i2c_init() call in audiohw_init is important? or the whole thing?
16:24:30jhMikeSJdGordon: I don't think the rtc or anything should matter
16:24:40JdGordonbacklight is on
16:25:01jhMikeStumu: did it normalize after backlight off or anything?
16:25:31JdGordon_now_ we can start doing something useful...
16:25:42tumuleft it idle to trigger the backlight off few times but no change yet
16:25:47jhMikeStumu: is it problematic to replace the bootloader? some code is only bootloader executed.
16:27:01jhMikeSnot sure if the SVN bootloader should be tried or not unless it actually does work ok
16:27:18JdGordonsansa svn bootloader works fine atm
16:27:31JdGordonunless something in the last 2 days broke it
16:27:47 Quit skate3214 ("ChatZilla [Firefox]")
16:28:01tumuhmm strange
16:28:16jhMikeSI think that might be the problem since LCD_REG_8 is only in the BOOTLOADER block
16:28:22tumuit fixes after the saver is triggered
16:28:30tumubut not everytime tho
16:28:56tumubut when it fixes, it doesn't corrupt again
16:28:58jhMikeStumu: I guess it's safe to try booloader from the latest build and see
16:29:18jhMikeSyou're only getting half the fix in place without it
16:29:54tumuhow do i try the svn bootloader?
16:31:28jhMikeSJdGordon: sansapatcher? I've never actually changed my bl. :P
16:31:46JdGordonyeah, sansapatcher -a PP5022.mi4
16:32:01jhMikeSJdGordon: to setup an SVN one?
16:32:44jhMikeSI may dump the #ifdef BOOTLOADER because doing so seemed to stop the white screen. I'll check that again.
16:35:09JdGordonare the -D options on gcc not passed to the linker?
16:35:59jhMikeShmmm...nope...still white screen
16:36:15B4gderJdGordon: -D are for defines, the linker doesn't use defines
16:36:47JdGordoni mean, whatever loads the .lds
16:37:01B4gderwell, that's why we preprocess the lds file with the cpp
16:37:28B4gderbefore linking
16:37:57JdGordonso where do I put my new define in the makefile so it works there? I've added it to EXTRA_DEFINES
16:38:26B4gdersorry, I don't have time to dig into that right now
16:38:47 Quit elinenbe (Read error: 113 (No route to host))
16:38:47 Nick elinenbe_ is now known as elinenbe (
16:39:59jhMikeSdo all the sansas with the "gunmetal" gray backs change color to silver when handled alot? mine's almost totally transformed.
16:40:36linuxstb_JdGordon: Am I understanding correctly that both LCD and ATA are working in your manufacturer mode app now?
16:40:39Jangarimm, chamelian
16:40:49JdGordonlinuxstb_: yep
16:41:04linuxstb_JdGordon: Nice... Have you tested ata read/writes ?
16:41:19JdGordonlinuxstb_: only by getting it to load rockbox from the disk
16:41:36 Join BobShield [0] (
16:41:38linuxstb_Ah, so you basically have the Rockbox bootloader running in manufacturing mode?
16:41:49JdGordonalmost unmodified bootloader
16:41:55*jhMikeS wonders of what disk JdGordon speaks
16:42:12JdGordonquiet you!
16:42:47linuxstb_JdGordon: So there should be no problems now... Do you know what to do for the patching?
16:42:54linuxstb_(or is that for another time?)
16:42:56JdGordonnot a clue :p
16:43:10JdGordonits almost 1am so i should be going to bed...
16:43:40 Quit n17ikh|Lappy (Connection timed out)
16:43:41Jangariyou're in australia?
16:43:56linuxstb_Download and compare the modified and original E200R bootloaders - you'll see that only four (consecutive) bytes are different. You'll need to find the disk sector containing that change (in the firmware partition), read it, modify it, then write it back....
16:44:04*jhMikeS will shut up and listen to music on the patched kernel that's been running since he first came on IRC
16:44:21linuxstb_JdGordon: Plus of course all the safety checks you want to implement...
16:44:44JdGordonboth those bootloaders are availeble where? or sansas site?
16:45:13linuxstb_ and
16:45:23JdGordonok, cool
16:45:52B4gderthe R BL is not available from sandisk, only by extracting from an actual target
16:45:59austriancoderlinuxstb_: whats the difference between USB_POWERED and USB_INSERTED ?
16:46:22linuxstb_USB_POWERED is a USB connection to a wall charger. USB_INSERTED is a USB connection to a computer.
16:47:01linuxstb_The existing usb_detect() code is the only way we know to differentiate the two.
16:47:33linuxstb_Do you have a usb charger?
16:47:44austriancoderlinuxstb_: no
16:47:52jhMikeSAS3514 differentiates the two on e200
16:48:44 Join elinenbe_ [0] (
16:48:44 Quit elinenbe (Read error: 104 (Connection reset by peer))
16:48:48 Nick elinenbe_ is now known as elinenbe (
16:50:10 Join fm2 [0] (i=c27f0814@gateway/web/cgi-irc/
16:50:20austriancoderlinuxstb_: I know whats going wrong in usb detection in usb stack, but i am looking why atm
16:50:26 Quit jhulst (Remote closed the connection)
16:50:50fm2How can I see how the USB mode was recognised (charger/comp)?
16:51:06linuxstb_fm2: What device?
16:51:24***Saving seen data "./dancer.seen"
16:51:50*JdGordon going to bed, cyaz
16:51:57 Quit JdGordon ("Konversation terminated!")
16:52:53 Quit B4gder ("It is time to say MOOO")
16:54:09bluebrotherFebs: what do you think about FS #7714? I think it's confusing
16:54:57FebsI've only had a chance to glance at it, but I noted in the comments one issue that I saw immediately.
16:55:16bluebrotherwell, I just wrote a lenghty comment ;-)
16:55:50 Join _jz [0] (
16:56:03bluebrotherhallo. wow.
16:56:23_jzi boot my sensa e200 with the daily build and i got the skin that appears with "Data abort at 00045F80" on top, nothing much
16:56:38_jzonly looong press allows me to shutdown, and same results afterwards
16:57:07amiconnUse the latest. Daily has a database bug
16:57:12pixelmabluebrother, Febs: is that the one which was posted to the forums first?
16:57:13_jzaah ok
16:57:41bluebrotherpixelma: not sure, but I guess so
16:58:05fm2linuxstb: Sansa e280 with FM
17:00:42linuxstb_fm2: I'm not sure - IIUC, the Sansa doesn't reboot into the OF when you insert USB, which is what happens on other PP targets.
17:01:27 Join desowin [0] (
17:02:26fm2linuxstb: I only ask because of the recent change in usb_detect. We want to tell usb charger from PC charger. Is it intended to show this somehow?
17:02:41pixelmabluebrother: there are also 2 mistakes in his new "first contact" section - Rockbox doesn't play ogg vorbis on the Archos devices and dual boot is also not possible everywhere or done differently (that's the first paragraph I read in the forums until I got confused). I also think he changes too much and don't see something that is more logical than before (very personal impression)
17:02:56pixelmahaven't even read further...
17:03:01linuxstb_fm2: On other PP targets, Rockbox reboots into the OF's USB mode when a USB connection is detected - that's how you know...
17:03:51bluebrotheras far as I can tell he is completely sansa-centric and forgets about all others. And presents the user stuff that isn't completely correct and misleading.
17:03:52fm2linuxstb: will it stay the same when the USB stack is implemented?
17:04:13linuxstb_fm2: When the USB stack is working, the Sansa will enter USB mode when it detects a USB connection...
17:04:21bluebrotherbesides, I'm wondering if we really need to rework that chapter completely.
17:04:26 Join toffe82 [0] (
17:05:00fm2linuxstb: and if USB charger is detected? What should happen then?
17:05:10linuxstb_fm2: Then Rockbox charges via USB...
17:05:33pixelmabluebrother: yeah, my thoughts too. There are other bits and pieces that are really wrong currently or outdated (the text viewer chapter for example)
17:05:48fm2linuxstb: but will it enter USB mode? Or will I be able to listen to the music? I.e. as if nothing had happened?
17:06:31fm2linuxstb: like with H120
17:06:31linuxstb_fm2: How can it enter usb mode if it's connected to a charger?
17:06:46bluebrotherpixelma: do you think it would be a good idea to include features.txt for building the manual? It could help with opt-ing somewhat more automatically
17:07:00bluebrotherotoh it could also lead to faster breakage of the manual.
17:08:00pixelmahard to say... I can't really imagine that at the moment
17:08:44fm2linuxstb: I probably don't quite understand what "USB mode" means. Is it a state when RB has recognised that an USB jack has been inserted (which can be either a wall charger or a PC connection)? Or does it only mean a PC connection (and hence the possibility to transfer files)?
17:08:56bluebrothermaybe I should give it a try and see how complex integrating is ... otoh I have enough other stuff on my todo
17:11:03 Join freqmod [0] (
17:12:38pixelmaI'd say your rbutil work is more important :)
17:12:52fm2linuxstb: if it's the latter then am I right in the assumption that we only need to detect a charger in order to properly show it in the status bar (animated battery)?
17:13:37linuxstb_fm2: "usb mode" is when your DAP is acting as an external hard disk connected to your computer.
17:13:48 Quit The-Compiler ("Connection reseted by god")
17:14:02amiconnThe other is usb charging... and it's not only necessary for the charging animation
17:14:37amiconnIt's also necessary to prevent poweroff while usb powered. Otherwise it would reboot every few minutes
17:15:32fm2amiconn: you mean the setting "Auto shut down after xxx mins"?
17:15:41fm2linuxstb: ok, now I understand
17:16:15 Join n17ikh|Lappy [0] (
17:16:48 Quit `sam` (".")
17:18:16fm2Another question about the plans: once the usb stack is in place, what mode would sansa enter when connected to a PC (with power off)? OF or RB one?
17:19:07linuxstb_I would expect the RB one.
17:21:04linuxstb_On other targets (with a hardware USB bridge), the bootloader enters USB mode directly if it detects a usb connection at boot - it doesn't start either RB or the OF.
17:21:27linuxstb_So I would assume we would want to do something similar on PP targets eventually.
17:23:13fm2linuxstb: yes, that'd be the "rescue mode", right?
17:24:11linuxstb_Yes. Although ipods already have a rescue mode that is accessible before the RB bootloader, so on ipods there is no need for a second rescue mode...
17:27:39fm2linuxstb: but it would be possible. What about sansa? Doesn't it also have rescue ("manufacturing") mode? What would the RB rescue mode be good for?
17:28:27 Join Rob2222 [0] (
17:32:02 Quit tchan (Read error: 110 (Connection timed out))
17:32:55jhMikeShow's the data from the USB stack actually going to get written to/read from the ata? nothing in the current svn kernal seems compatible with that.
17:33:22austriancoderjhMikeS: reading sectors with... one moment
17:33:28 Join DefineByte [0] (
17:34:23austriancoderand ata_write_sectors
17:34:26jhMikeSfrom an interrupt handler?
17:34:54austriancoderjhMikeS: nopw
17:35:04austriancoderan own thread will handle it
17:35:13amiconnFrom the USB thread, I'd expect
17:35:29jhMikeShmmm...yes, of course. but how?
17:35:50austriancoderamiconn: nope.. at the moment i am doing this in an seperate therad
17:36:22DefineByteis there any chance that setting the hold switch to on turns off the wheel on the iPod video?
17:36:26austriancoderin the interrupt handler i put events in the queue of the "worker" thread and he does all the work
17:36:54austriancoderamiconn: using existing usb thread is not the best idea.. a lot of generic code gets mixed with usb-stack specific one
17:37:27DefineByteforget that, if it was a hardware problem with key eating i guess the OF would show it. 0_)
17:37:34jhMikeShmmm...damn I need to commit this patch. it's got the stuff to make that more efficient.
17:38:38 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
17:39:02 Part fm2
17:41:28 Join sitwon [0] (
17:41:59jhMikeSwhat about using the whole audio buffer as a huge FIFO? lots of slop space that way.
17:43:24sitwonThere is a typo on the page <>. To copy .rockbox to ~/sansa/mnt/ the correct command is "cp -R .rockbox/ mnt/" without the '-R' flag the cp command will ommit directories.
17:43:55DefineByteit's a wiki
17:44:58 Quit Rob2222 (Connection timed out)
17:46:21linuxstb_sitwon: A better way would be "unzip -d ~/sansa/mnt/"
17:48:12sitwonI recognize that you could just unzip it on the device, I also realize that it's a wiki. But I don't have write access on the wiki and I'm simply pointing out that the way they tell supposed novices to do it is wrong.
17:48:34DefineBytefair enough. good catch :)
17:49:04austriancoderlinuxstb_: usb_enable(true) get call when USB_INSERTED and usb_enable(false) when != USB_INSERTED - or?
17:49:18 Join bluey- [0] (
17:49:30sitwonby the way, I have rockbox working on my brothers e270R now, thanx a LOT!
17:49:35linuxstb_austriancoder: See firmware/usb.c
17:50:11 Quit petur ("stkov")
17:52:34 Quit bluey- (Client Quit)
17:53:00 Quit Zagor ("Client exiting")
17:53:40 Join jane- [0] (
17:54:08DefineByteso, who wants to port rockbox to x86? 8)
17:55:24austriancoderDefineByte: me has it on the todo for some time... adding an sdl based target
17:55:57DefineBytei'd love to build a jukebox with a Via EPIA and use rockbox on it.
17:57:25linuxstb_austriancoder: Have you heard about the "port" to the Motorolo E680 phone? (basically just the UI sim for the ipod video running on phone, which runs Linux).
17:57:44DefineByteis moving rockbox to targets other than DAPs a goal for the future?
17:58:19austriancoderlinuxstb_: never heard of this "port"...
17:58:26linuxstb_I wouldn't go as far as saying that it's a goal, but the idea of "rockbox as application" has been discussed many times before, and lots of people seem to want it.
17:59:04DefineBytethe audio players on linux are pretty feature light
17:59:12 Join jhulst [0] (n=jhulst@
17:59:17linuxstb_austriancoder: (follow the link in the first post)
17:59:47DefineBytethat's pretty neat
18:00:14jhMikeSaustriancoder: I asked about using the audiobuffer since that would leave lots of room for bursting data in and letting ata handle it largly at it's own pace if it doesn't always keep up.
18:00:37linuxstb_jhMikeS: Wouldn't it be dangerous to cache too much data though/
18:00:43 Join Lear [0] (i=chatzill@rockbox/developer/lear)
18:01:23jhMikeSI don't think so since it's on the running device itself, not cached on the pc end
18:01:39austriancoderjhMikeS: at the moment I need a max buffer of 512byte for data transfers... used also for maxsize in endpoint descriptors
18:03:58*austriancoder makes a break#
18:03:59jhMikeSaustriancoder: does ata keep up or do things have to slow down to accomodate it?
18:04:37austriancoderjhMikeS: to eraly to tell...
18:05:53 Quit Jangari ()
18:06:50 Join senab [0] (
18:08:20 Join bluey- [0] (
18:10:54 Nick jane- is now known as in-jane (
18:11:31 Join barrywardell [0] (
18:12:44 Quit jhulst (Read error: 104 (Connection reset by peer))
18:14:34 Join n1s [0] (
18:17:52 Quit senab ("ChatZilla [Firefox]")
18:19:54 Join webguest13 [0] (i=50d81efc@gateway/web/cgi-irc/
18:20:19webguest13austriancoder: how far away is the serial and msd drivers?
18:21:40*n1s guesses on > < this far...
18:22:24linuxstb_barrywardell: Are you here?
18:22:49barrywardellback from holidays
18:23:01 Part sslashes
18:23:24linuxstb_barrywardell: Welcome back. Have you tried connecting a PP target to your Mac with the "charge button" held down? i.e. to avoid entering USB mode?
18:23:41linuxstb_If I do that on my ipod, the Mac's USB stack dies (for that USB port).
18:24:14barrywardellmy Mac's USB stack dies a lot when I connect a PP target to it
18:24:18linuxstb_This happened both before and after the USB detection change I've just committed.
18:24:36barrywardellit also dies other times too
18:25:08 Join jhulst [0] (n=jhulst@
18:25:17linuxstb_Any idea how to fix it?
18:26:01barrywardellit doesn't die when there's a functional usb driver
18:26:16barrywardellie. when I was using the e200tool code in Rockbox it didn't die
18:26:42linuxstb_I also noticed that when connecting to Linux, I sometimes (but not always) get a "new USB device" message before the ipod reboots to disk mode, and then I get a second "new USB device" when it connects in disk mode.
18:27:31linuxstb_So is the problem that "dr_controller_stop" doesn't close the USB connection thoroughly?
18:27:32barrywardellthe 1.2.15 OF for the sansa also cuses the Mac usb stack trouble, so I don't know if it's a problem we can fix too easily
18:28:18barrywardellI don't think it's a dr_controller_stop problem
18:29:05linuxstb_I mean should we be doing more between calling run and stop - to tell the host that we're aborting the connection?
18:29:50barrywardellI'm not sure.
18:30:40linuxstb_Do you have an ipod, or just H10 and Sansa?
18:30:43barrywardellhave you had a look at the linux driver to see what it does?
18:31:00barrywardellI have an ipod with a broken disk, so it's not much use
18:32:30barrywardellI'll have a look at the linux code now and see if we're missing anything
18:35:11 Quit DefineByte ()
18:35:27elinenbequick question... all the USB commits, are those so rockbox can handle PC/device USB connections, or for USB on the go?
18:36:41 Join DefineByte [0] (
18:40:04barrywardelllinuxstb_: looking at the driver, there is some extra stuff. it dumps the ep queues
18:40:32barrywardellbut that requires working with the host, so it wouldn't be easy to add
18:40:36 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
18:42:10austriancoderlinuxstb_: stalling ep0 would be a good idea
18:42:49*linuxstb_ should point out that he knows nothing about usb programming
18:43:22austriancoderme tries something
18:46:07austriancoderlinuxstb_: you could try this: - as i dont have access to a mac today, i cant test it
18:47:42linuxstb_Thanks, I'll try it later this evening - I don't have access to my Mac now either.
18:48:23austriancoderdo we have functions for utf-16 ?
18:49:02 Quit jhulst (Remote closed the connection)
18:49:04 Join spreslijs [0] (i=spreslij@
18:49:16 Part spreslijs
18:49:30 Join BigBambi [0] (n=alex@
18:51:26***Saving seen data "./dancer.seen"
18:52:42 Join Kingstone [0] (i=A@
18:52:46Kingstonedoes anyone know where can i find info about today's apple conference?
18:54:10scorchethat is offtopic...
18:57:23elinenbescorche: not if the new ipods are running rockbox?
18:58:21linuxstb_elinenbe: Are they?
18:58:48DefineByteno, they're running OSX
18:58:55scorchewhich is a rumor...
18:59:08DefineBytean untrue one i think
18:59:29 Join Naked [0] (i=naked@
18:59:37 Nick Naked is now known as Hadaka (i=naked@
19:00:13 Part DefineByte
19:04:25 Quit barrywardell (Remote closed the connection)
19:07:10Nico_Paustriancoder: seen that we have to uplad code to google ?
19:13:39 Join petur [0] (n=petur@rockbox/developer/petur)
19:14:41preglowfinal stages should be ready for uploading now
19:15:28 Quit freqmod (Read error: 104 (Connection reset by peer))
19:16:37 Join fm2 [0] (i=d9b96770@gateway/web/cgi-irc/
19:17:00Nico_Ppreglow: we have to upload code from the 20th of august
19:17:35preglowi know
19:19:06Nico_Ppreglow: what do you call "final stages" then ?
19:19:27 Join Thundercloud [0] (
19:20:56fm2May I ask an OT question? Ist it possible to disable "Database Refresh" in Sansa OF? It's nasty. I know this is the wrong channel, but it's quiet here, so I thought...
19:21:35tumui've wondered that as well
19:21:55linuxstb_I thought the Rockbox bootlaoder could do that, at least for some OF revisions?
19:22:35fm2tumu: well, one way would be to ask the devs to quickly make USB mode working in RB :-)))
19:23:39preglowNico_P: i meant the final stages of preparing for uploads, from googles side
19:23:46preglowNico_P: they didn't have it ready before today
19:23:53Nico_Pah right
19:25:33 Join spiorf [0] (
19:25:43 Join pepie34 [0] (
19:26:07 Part fm2
19:29:01 Join elinenbe_ [0] (
19:29:30 Quit sitwon ("Leaving")
19:30:16linuxstb_Slightly off-topic, but Apple's big news is a replacement for the ipod video - an 80GB model smaller than the current 30GB, and a 160GB smaller than the current 80GB....
19:30:27linuxstb_So Rockbox is no-longer running on any of Apple's current ipods...
19:30:51 Join Hellmark [0] (n=hellmark@
19:34:39n1swow an (GB as entry level hd player is massive... I still haven't filled my h320... I guess the time for portable flac is here :-)
19:34:53linuxstb_Well, portable ALAC...
19:35:18n1swhat, you're not going to port rockbox to it!111!!!! ONE
19:35:45linuxstb_It's inevitably going to be protected in the same way as the 2nd gen Nano.
19:36:53n1sWell one could get one of those 160 GB monsters and rip out the drive and put it in something useful, (and hope it would work) .-)
19:37:22linuxstb_The 160GB was announced as $349
19:37:37linuxstb_80GB is $249
19:37:51*rasher watches that mean €500
19:37:52*nicktastic is ready to buy a 160GB as soon as Rockbox is ported
19:37:57alienbiker99the ipod touch has wifi =)
19:38:02Hellmarktouch I am thinking $400
19:38:19*linuxstb_ points people to #rockbox-community
19:38:37amiconnWell, I think it won't take that long to get my H180 filled. Flac takes quite some space
19:38:56amiconnIf I want the same lossless stuff on my recorder, I'd need a 120GB disk at least
19:39:13amiconn(WAV, as flac isn't possible with that CPU)
19:39:34linuxstb_My collection of studio CDs comes in at about 130GB in flac...
19:39:43Hellmarkhow is FLAC on a 4G iPod or Mini?
19:40:03linuxstb_FLAC works well on all ipods - it easy on the CPU.
19:40:06amiconnIt works... and you can still carry several albums
19:40:14ThundercloudNot on the default firmware though
19:41:23Hellmarkthe default doesnt support it at all though
19:41:38Hellmarktalking Rockbox
19:41:39amiconnThat's totally unimportant
19:41:40 Quit elinenbe (Read error: 113 (No route to host))
19:41:41 Nick elinenbe_ is now known as elinenbe (
19:41:49ThundercloudI thought we were talking about the new ipods?
19:41:58ThundercloudIn relation to rockbox and how they're likely to be protected
19:42:36amiconnI referred to Hellmark's question regarding flac on 4G / mini
19:42:43ThundercloudAh ok
19:42:57ThundercloudBTW, how exactly are second gen nanos protected?
19:43:01linuxstb_And when I talked about "all ipods", I was referring to the ones Rockbox runs on...
19:43:30HellmarkI'm looking at the last of the monochrome iPods, because they're cheaper. Only thing is, do I want more battery life, or more storage space.
19:43:39ThundercloudHellmark: I see.
19:43:46amiconnHmm, I don't remember whether I tested this, but I expect flac to run without boosting even on PP5002
19:43:49ThundercloudI quite like FLAC.
19:44:03ThundercloudI've got a Cowon D2 and put some FLAC stuff on there. Not a lot mind, I don't _have_ a lot of FLAC stuff.
19:44:04linuxstb_amiconn: I expect so too - it's all in IRAM...
19:44:31HellmarkThundercloud, same here.
19:44:49ThundercloudHellmark: What, you have a D2 or you like Flac? :D
19:44:54HellmarkI've only got a handful of albums in FLAC
19:45:01Hellmarkbut wanting to start migrating
19:45:32ThundercloudStill, hang on, the new ipods. Is it JUST the capacity increase and size decrease?
19:45:50 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
19:46:00 Quit amiconn (Nick collision from services.)
19:46:02 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
19:46:07Hellmarkon iPod Classic, more battery, more storage
19:46:08amiconnlinuxstb: Ah, then it definitely should. It's the same arm cores, and IRAM is equal in speed to PP502x
19:46:11amiconn(we're not running dualcore on PP5002 yet though)
19:46:20Hellmarkthen there is the new iPod Touch, and the third gen Nano
19:46:22 Join obo [0] (n=obo@rockbox/developer/obo)
19:46:35ThundercloudHellmark: :o I see. Links?
19:46:35linuxstb_Thundercloud: No. They first announced a new generation Nano, with a new UI, more similar to the iPhone, and video capabilities. Then they announced the new "ipod classic" which they said had the same UI as the new Nano.
19:47:32linuxstb_So reading between the lines, I would say they've finished writing a new OS for the Samsung CPU, and both the new ipods are using that. i.e. it's the end of the line for Portalplayer.
19:52:05 Join Llorean [0] (
19:52:43ThundercloudThing is though, it doesn't give capacities or prices for the touch
19:52:51krazykitit does
19:52:55krazykit8/16 gigs
19:53:03krazykit300 and 400 bucks
19:53:19bluey-geez.. no hdd
19:53:24ThundercloudWhere's that?
19:53:27ThundercloudI didn't see that
19:54:18ThundercloudThat's a pretty good price as far as flash players go
19:54:41 Join spiorf_ [0] (
19:54:50bluey-for flash but for the same price they could offer a 80gig hdd
19:54:59Thundercloudkrazykit: Well, $300 is 150
19:55:03BigBambiguys, the sme conversation is going on in the correct, off-topic channel #rockbox-community
19:55:10ThundercloudI got my 4GB Cowon iAudio D2 for 150
19:55:20ThundercloudMeh, i'm finished
19:56:18 Join vic^2 [0] (
19:56:32 Quit Llorean (Client Quit)
19:56:53vic^2Are the daily builds of the bootloaders saved and available anywhere?
19:57:15krazykitbootloaders are released, not built daily
19:57:19 Quit Lear ("ChatZilla [Firefox]")
19:57:21 Quit ender` (" AAAAAaA: American Association Against Acronym Abuse and Ambiguity.")
19:57:37 Join Llorean [0] (
19:57:50vic^2so there is no way to get the bootloader for the sansa that has the reduced USB delay short of building it myself?
19:58:11*amiconn wonders what HDDs the new "classic" ipod line uses
19:58:32*linuxstb_ too
19:58:34amiconnThe devices itself are, hmm, not very interesting to be honest
19:58:45linuxstb_I agree - but the disks are...
19:59:34amiconnApple again forgot the radio (or to be more up to date, fm + dab radio)
19:59:47linuxstb_Although they seem to have quite an impressive battery life - 40 hours for audio, 7 for video.
19:59:53LloreanIs there somewhere with images of the new iPods?
20:00:05linuxstb_This is a good link -
20:00:07*Llorean just finished wrangling his computer out of an xgl nightmare.
20:00:09Hellmarkamiconn, eh, do most people really want a radio?
20:00:20LloreanHellmark: Quite a lot of us do, yes.
20:00:48HellmarkI mean, myself, and most people I know, use MP3 players TO GET AWAY from the radio, since the radio sucks cock anymore.
20:01:12krazykityeah, but the radio is nice if you want to listen to, say, NPR or something
20:01:18krazykitor the BBC across the pond
20:01:26 Quit petur ("switching")
20:01:35 Join petur [0] (n=petur@rockbox/developer/petur)
20:01:43 Quit DataGhost (Read error: 104 (Connection reset by peer))
20:01:58 Join DataGhost [0] (
20:02:05Hellmarkwell, it must not be that pressing of a concern though. I mean 110 million sold, and not enough of a clamor to add it
20:02:26linuxstb_Hellmark: Quality of radio heavily depends on your location....
20:02:42HellmarkMajority of the US, its horrid.
20:02:49Hellmarkwell, north america.
20:03:02krazykitcountry music stations :-(
20:03:13Hellmark(since many of the stations in canada and mexico are owned by the same company)
20:03:22 Join GodEater [0] (n=bryan@rockbox/staff/GodEater)
20:03:26Hellmarkkrazykit, no
20:03:41*amiconn would like a player the size (and capacity) of the new ipods, but with a greyscale lcd, real buttons, recording, and fm/dab/drm radio
20:03:48gregjwhat you say folks
20:04:06 Quit spiorf (Read error: 110 (Connection timed out))
20:04:07 Join sambolina3 [0] (i=a9f44693@gateway/web/cgi-irc/
20:04:11krazykitso who wants a pool on how long it'll take for requests for rockbox to be ported to the new nano, the classic, and the touch to be put in New Ports?
20:04:12gregjI wonder if the new nano would be hackable for rockbox
20:04:26sambolina3a long long time
20:04:26Thundercloudkrazykit: Can I have "soon"? :P
20:04:27gregjfew seconds ?
20:04:29Hellmarkthere are a handful of companies that own almost all the radio stations in North America, and they've stripped away the DJs choice of music, and just have them play precanned music, of which really is in short supply, since the same songs get repeated over and over again
20:04:42linuxstb_gregj: It's unlikely to be less secure than the current Nano...
20:04:46HellmarkI mean, there are times, where you will hear the same song about once every hour
20:05:10sambolina3agreed ^^^
20:05:19gregjlinuxstb_: is the current nano makes it just hard to change firmware, because of encryption, etc ?
20:05:25amiconnOh, and I forgot spdif in+out
20:05:33gregjI didn't kne
20:05:39gregjthought it is just completly different arch
20:05:46linuxstb_gregj: Yes, that as well...
20:06:05 Join ompaul [0] (n=ompaul@freenode/staff/gnewsense.ompaul)
20:06:08gregjipod is completly useless without rockbox..
20:06:42gregjwhy would they wanna do this ?
20:06:51gregjthere is no problem with the big ipod
20:06:57gregjso why do it for nano
20:07:09sambolina3so they can stop people from playing videos on their nanos
20:07:24Hellmarkwell, I am betting much of the iphone hacking will start bleeding over to the ipods
20:08:56 Join merbanan [0] (n=banan@
20:11:50 Quit sambolina3 ("CGI:IRC")
20:11:57 Join sambolina3 [0] (i=a9f44693@gateway/web/cgi-irc/
20:12:21 Quit sambolina3 (Client Quit)
20:13:07 Quit Nico_P (Remote closed the connection)
20:13:41 Join saratoga [0] (i=98039a9c@gateway/web/cgi-irc/
20:18:18 Quit GodEater (Remote closed the connection)
20:19:12 Quit atsea-68 (Remote closed the connection)
20:20:20 Part Xero ("Ex-Chat")
20:20:59 Join senab [0] (
20:22:48senabhey, i'm currently selling my ipod and looking at getting a sansa e280. i've got a few questions though: 1. is the screen still flickering? 2. can i record radio? 3. what can i expect in battery life?
20:23:38saratogasenab: no, probably, and 12-13 hours (?)
20:23:40RudManI've got a e240. the screen does not flicker
20:24:00 Join O112358 [0] (
20:24:12 Join GodEater [0] (n=bryan@rockbox/staff/GodEater)
20:24:28RudManI've been unable to get recording working at all, I've only looked at it briefly tho. and my sansa only lasts about 5 hours
20:25:01senabthanks, sounds good then. i'm debating whether to pay the extra £20 for the radio version hmmmm
20:25:06preglowsaratoga: i think uploading is ready now
20:25:13senabrudman: whats the radio reception like?
20:25:28O112358how could i get back to the default rockbox theme after changing my theme to something else? when i try and load up Rockbox_Default it doesn't look the same as when i first installed rockbox..
20:25:30RudManhaven't tried the radio at all.
20:26:19O112358and resetting my settings seems a bit extreme
20:26:33in-janeRudMan: have you figured out why it only lasts 5 hours? that sounds weird
20:26:39senabrudman: okay thanks
20:26:48RudManmaybe you made some font changes after the installed it the first time and that's why it looks different
20:27:13RudManin-jane: I've got a refurb unit so I'm suspecting it's a questionable battery
20:27:20senabjust found the e280 for £73 delivered. Bargain!
20:27:48O112358RudMan: i'm sure even the background colour is different, it doesn't start off white does it?
20:28:17 Part bluey- ("Leaving")
20:28:18RudManmost themes don't reset their backdrop if they don't include them
20:28:19O112358RudMan: there must be some way to just reset theme settings without resetting everything..
20:28:30in-janeRudMan: OK. i just got little worried since i ordered my e240 today ;)
20:29:05linuxstb_O112358: You could open up .rockbox/config.cfg in a text editor and delete any lines relating to the theme.
20:29:10BigBambisenab: where?
20:29:11O112358RudMan: that's what i assumed as well.. seeing as the default theme doesn't included that information it doesn't change it
20:29:32senabBigBambi ^
20:29:49 Quit Buschel ()
20:29:52O112358linuxstb_: nice one, seems quite fiddly for something so basic though
20:29:57 Quit pepie34 ("Ex-Chat")
20:30:21linuxstb_O112358: I agree. If there isn't already, we should add a "rockbox_default.cfg" in the themes directory.
20:31:09 Join BigMac [0] (
20:31:33O112358linuxstb_: good idea, also, can you remember if the image that comes up when you connect to the computer is 'graphical' looking with no settings changed?
20:31:40RudManand it should include a null backdrop line like this backdrop:
20:32:08linuxstb_O112358: The USB picture is hard-coded - you can't change it.
20:32:14RudManif your previous theme set a black bmp as a backdrop and the new theme does not set a backdrop but sets a black font, you're hosed
20:33:17 Join jgarvey [0] (
20:33:35linuxstb_RudMan: Yes, that's a well-known issue. Hopefully the new themes site can add a validation to check for a backdrop: line in the .cfg
20:34:05Lloreanlinuxstb_: It was our intent to require a backdrop line, as if I recall even a blank one at least resets it.
20:35:29 Quit vic^2 (Remote closed the connection)
20:36:05O112358i could change the .cfg through rockbox couldn't I?
20:36:45linuxstb_O112358: You mean with the text editor?
20:37:16O112358hehe yeah i'm doing that right now
20:37:55O112358do the iconset parameters relate to the theme?
20:38:50RudManyes, it reflects what icons you see
20:39:36O112358and it's fine to just delete all these lines? it will just make it resort to its default right?
20:41:46O112358and then reset my player?
20:42:13O112358there we go, back to normal
20:42:27linuxstb_Now you can do "save theme .cfg" and keep that safe.
20:43:08O112358but wont it fail to override the background settings etc?
20:44:03O112358oh right i'll be restoring the .cfg file not loading a .wps..
20:44:39 Part _jz
20:45:26 Quit GodEater (Read error: 113 (No route to host))
20:46:56 Quit datachild (Nick collision from services.)
20:46:58 Join datasleep [0] (
20:47:20O112358argh, my computer isn't detecting my player when i plug it in now, but be because i was unplugging it without safely removing hardware
20:48:00O112358when ever i try to remove the device it just says that it can't be removed
20:49:06LloreanjhMikeS: The garbled lines fix for Sansa requires a new bootloader, no?
20:49:08 Quit webguest13 ("CGI:IRC")
20:51:05 Join The-Compiler [0] (
20:51:30***Saving seen data "./dancer.seen"
20:53:10 Quit O112358 ()
20:55:02 Quit senab ("ChatZilla [Firefox]")
21:00:58 Join flo [0] (
21:01:55 Nick datasleep is now known as datachild (
21:03:59flohi. i have a problem with the database on a sansa e260. i removed a file from the player but the database still contains an entry for the directory and the file inside. even when using "initialize now" or "update now".
21:04:31 Quit flo (Client Quit)
21:04:38bluebrotherhave you checked the recycle bin?
21:04:58bluebrotherhmm. Impatient or connection problems?
21:05:56lostlogicw00t, we're gonna get to write a wifi stack!
21:06:44lostlogichmm, only 16gb though :(
21:07:49 Join AceNik [0] (n=AceNik@
21:07:50LloreanIt seems more like a target for Rockbox-as-an-app assuming it's hackable in similar ways to the iPhone
21:08:40linuxstb_How is the iPhone hackable? You mean you can run third-party apps on top of the existing OS?
21:08:51Lloreanlinuxstb_: So I've been lead to believe, at least.
21:09:00lostlogicall I know is that it's been jtag'd
21:09:18 Join atsea-68 [0] (i=atsea-@gateway/tor/x-05c3b94a64ba0b83)
21:09:44*linuxstb_ is starting to doubt the future of Rockbox - no-one is making DAPs any more...
21:09:50AceNikguys what exactly is supposed to happenw hen i connect the usb to my h10 when its off
21:10:13lostlogiclinuxstb_: meaning they're all making "do everything gadgets" instead?
21:10:20AceNiklinuxstb: well there will be a definately "ray of light"
21:10:37amiconn"do everything but nothing properly" :/
21:10:40linuxstb_lostlogic: Yep. Plus the focus on video features.
21:10:49lostlogicthe new iPod classic at 80 and 160 gig looks promising actually. I wonder what it's got under the hood.
21:11:17linuxstb_IMO it's almost certain the same as the 3rd Gen Nano, which is almost certainly the same as the 2nd Gen Nano...
21:11:36Lloreanlinuxstb_: I don't see any reason Rockbox couldn't move into media-playing too eventually
21:11:46linuxstb_But I guess we'll see dissections on the usual websites soon
21:12:14lostlogicnano 2g is samsung SOC right?
21:12:29AceNiklinuxstb:any chance of rockbox entering phones?
21:12:30linuxstb_lostlogic: Yes, plus the as-yet-uncracked encryption.
21:12:56lostlogicyeah, I bet there'll be more interest in it now ;)
21:13:28linuxstb_AceNik: Don't ask me - Rockbox will go whereever people take it...
21:13:29 Join Frazz [0] (
21:13:33LloreanAceNik: Haven't you asked this before, several times?
21:13:38 Join O112358 [0] (
21:14:02AceNikLlorean: sorry , can you tell me about the usb ?
21:14:12O112358god i'm stupid, my computer wasn't detecting my iriver because i was plugging my usb cable into the host socket, didn't know it even fitted in there
21:14:14amiconnAceNik: Tried the latest build already? Shouldn't crash anymore on db init...
21:14:19O112358does rockbox have usb charging?
21:14:55O112358and i just realised that the usb charging image looks differen't to me because i was running a really old build
21:14:58amiconnO112358: The USB host port uses a mini-AB socket that both accomodates mini-A and mini-B plugs
21:15:13AceNikamiconn: yes thanks, it doesn't, although it does crash after a couple of minutes of playing music, it hangs, does playing at higher volumes make a difference
21:15:26O112358amiconn: unfortunate for people like me who don't look
21:15:28linuxstb_lostlogic: Any plans to return to Rockbox hacking?
21:15:28n1sO112358: yes rockbox can usb charge
21:15:56amiconnThis is meant for OTG devices, which can negotiate which device acts as host and which acts as device
21:15:56n1sbut for it to be fast enough to be useful you need to enable a setting somewhere
21:16:00O112358how do you get it to start charging?
21:16:21O112358i have it plugged in but the charging logo isn't animating
21:16:34amiconnAfaik, the very same socket provides H300 MTP mode on US devices (at USB1.1 speed)
21:16:45n1sO112358: are you in USB mode?
21:16:51AceNikamiconn: that means automatic detection isnt possible on h10 yet, right, cause my h10 does nothin on usb being connected
21:17:06O112358AceNik: well i have a giant usb plug on my screen
21:17:29n1sO112358: then unplug it and hold down the rec button when you plug it in again
21:17:51O112358ok so it's not automatic?
21:18:11 Join ficsch [0] (
21:18:22amiconnI'm not sure whether the H300 charges from USB in rockbox
21:18:36O112358n1s: argh, it went into recording mode
21:18:43 Quit The-Compiler (Read error: 104 (Connection reset by peer))
21:18:59Lloreanamiconn: I'm *almost* certain it does.
21:19:04amiconnThe hardware allows it, but powermanagement in rockbox isn't quit what it should be
21:19:09n1samiconn: I think it does, a setting was added for it at least
21:19:18*linuxstb_ was about to say what n1s said
21:19:28 Join My_Sic [0] (
21:19:48peturh3x0 does charge from usb
21:19:48n1sbut as with most of rockbox's features I have never tried it
21:20:14O112358maybe it's charging but i just can't see it
21:20:21ficschhi. my database does not change after moving a file to another directory an removing the source dir. I already did "initialize now" and "update now". I have a sansa e260
21:20:27peturit is a setting
21:20:30n1sO112358: did you enable the option?
21:20:39O112358a little plug icon appears next to the battery meter, i have a feeling it stops the meter from doing the charging animation
21:20:57peturO112358: general settings -> system -> battery
21:20:57O112358oh an option?
21:21:08O112358why would anyone not want the player to charge from the usb?
21:21:12n1sO112358: I told you to do that already
21:21:30n1sO112358: because not all usb ports support it
21:21:34O112358n1s: sorry i thought you were talking to someone else
21:21:36AceNikamiconn: new bug−−>player hangs after 1min or so of playing music
21:22:01RudManfisch: did you reboot?
21:22:18n1sand if you try to draw more current than the port supplies it (should) kill the connection
21:22:57O112358oh, that's a good reason to have a setting then
21:23:11ficschRudMan: is it possible to delete the database from the file system to force a rebuild? Or shall i try debugging rockbox? Maybe it's a bug?
21:23:18Genre9mp3pressing the rec button while inserting the USB cable will make the H300 charge... though it's a bit ugly that while doing so you get to the Recording screen as well
21:23:35O112358hmm, i did a battery bench test yesterday on the first cycle of my brand new battery and it came out with some very odd values
21:23:46Genre9mp3we should probably change this (use the Play button instead or something)
21:24:12linuxstb_Genre9mp3: Is long-press on play currently unused?
21:24:27n1sGenre9mp3: also the rec screen comes up if you hold rec to reset settings on boot :-)
21:24:58linuxstb_Who's idea was it to map the "rec" button to the recording screen? ;)
21:25:11Genre9mp3linuxstb: No... at least not while on file browser
21:25:13O112358it started off detecting a voltage of about 60 mV which went down to 2 mV over the period of an hour, then it jumped up to over 4000 mV (what one would expect) and then acted normally
21:25:38Genre9mp3file browser/menu that is...
21:25:41lostlogiclinuxstb_: yes, it is likely. I've been busy moving 2000 miles, buying a home and starting a new job for the past couple of months, but should finally have internet access at home tomorrow which might just might put rockbox back on my "way to spend time" radar.
21:25:44linuxstb_Genre9mp3: What about the WPS?
21:26:19linuxstb_lostlogic: Good to hear - playback.c has been asking after you...
21:26:36Genre9mp3linuxstb: It's not mapped as well
21:26:38O112358is that an issue with the battery bench plugin? seeing as it's impossible for my battery to be at such a low voltage and for the player still to run
21:26:48RudManfisch: I had some issues with the database recently. it wasn't seeing the tags correctly.
21:26:59RudManfisch: maybe there is something flaky with it
21:27:04linuxstb_Genre9mp3: Then a feature request would seem in order (I don't own an H300)....
21:27:47O112358has anyone else seen strange results like that while running battery bench tests?
21:28:10 Quit linuxstb_ ("Client Exiting")
21:28:13*Genre9mp3 can't remember the shortcut to Pitch screen (from WPS)
21:28:21ficschRudMan: you mean the database does not build correctly due to bad tags of mp3s?
21:28:24Genre9mp3linuxstb: will do
21:30:17O112358is there a posiblity that battery_bench was detecting the voltage wrongly?
21:30:23RudManfisch: that's what it appeared to me.
21:30:45amiconnficsch: I don't think there's a bug in the db rebuild (other than the crash bug I fixed this morning). If you deleted a folder but the files still appear in the database, it's most probably because you didn't delete for real, but the files are now in a recycle bin
21:30:51RudManI rebuilt the tags on the files and the database appeared to work correctly again
21:30:57amiconn(where they will still be found by the database)
21:30:59Genre9mp3ah... it's Play+ up(or down).... you just have to be a bit... quick
21:31:30 Join googlegoogle [0] (
21:31:37googlegooglehey guys
21:31:38 Quit AceNik ("bye guys have fun , enjoy !!!!!")
21:31:45ficschRudMan: i'll try that
21:32:05ficschamiconn: i delete it using rm -f on linux
21:32:08Genre9mp3strange cause Play+AB (ID3 info shortcut) doesn't have the same behavior
21:32:25googlegoogleis there any chance of rockbox going to the cowon D2
21:32:30ficschRudMan: do modify mp3s using windows or linux?
21:32:51RudManI used a program called EZ Tag (I think) on windows
21:33:43RudManbut I didn't have the same problem as you, I had issues that the database wasn't even seeing my files or seeing most of them as untagged
21:34:23ficschisn't it possible to delete the database from the file system to force a rebuild? do you know the file name?
21:35:03RudManthat I don't know
21:35:36amiconnficsch: Sure, but I don't think the result will be different from normal "Initialise now"
21:35:42RudManmove all the files off the player, then run init now. shutdown, restart and see what is in the database
21:39:26O112358yay i got my player to crash
21:40:00n1sO112358: what did you do?
21:40:41O112358n1s: I deleted a file while i was playing it, i then pressed Navi to browse files and it crashed
21:40:51O112358not surprising really
21:41:33n1sO112358: that is a known bug, that has been aroud for ages... unfortunately no one has fixed it yet
21:41:49n1swere there other files in the dir?
21:42:01O112358hmm, why have a delete option when you hold down Navi while a song is playing anyway?
21:42:17O112358the file was in my root
21:42:25O112358it was something i had just recorded
21:43:00n1sO112358: yes but were there other music files in the root then ?
21:43:22O112358erm, no
21:43:56O112358would it have been ok had there been other files for it to play?
21:44:04n1sok, thanks, I guess it is actually two seperate bugs, the one I was thinking about was apparently fixed
21:44:19 Quit jgarvey ("Leaving")
21:44:21n1sYes I think so
21:44:52 Quit DerPapst (Read error: 104 (Connection reset by peer))
21:45:29O112358the file deleted fine, then i assume it just carried on playing since the file was already loaded into the buffer, but when i pressed navi the total song time went to 00:00 and it just froze on that screen
21:46:16O112358n1s: oh, and have you heard of people getting strange voltage values on battery bench tests before?
21:46:34n1snot that I can remember, no
21:47:21O112358i can't see how it can be a problem with my battery, because had it actually been at 60 mV the player wouldn't have been running
21:48:56n1shmm, yep sounds pretty weird... if rockbox would read very low values it should shut down with a battery warning anyway...
21:48:59 Join billytwowilly [0] (
21:49:34O112358hmm, confusing
21:50:02O112358could it be a bug?
21:50:19O112358well it must be a bug
21:50:26amiconnIt's a bug in battery_bench
21:50:34n1sdunno, I havent run battery bench for several months, does it read constantly 60mV ?
21:50:40amiconn..introduced when switching to mV for battery voltages in rockbox
21:50:41O112358amiconn: ahh so it's not just me?
21:50:55amiconnIf voltage > 4.095V, it wraps to zero
21:51:30O112358n1s: no it goes from 60mV to 0mV over an hour and then once it gets to 0mV it jumps up to the proper value
21:51:55n1sO112358: then it sounds like exactly what amiconn said :-)
21:51:57amiconnThis is due to some bits in batt_info.voltage being used for other purposes
21:52:01O112358that seems to obey amiconn's theory :D
21:52:23*amiconn looks at XavierGr
21:52:39O112358yay mystery solved, thanks amiconn, i'll be able to sleep tonight
21:52:43amiconnXavierGr: Is there a special reason for this bit reuse, instead of using some bool members in the struct?
21:53:29O112358seems like my battery is just too much of a beast for rockbox to handle hehe
21:54:04amiconnNo, it's a plain bug
21:54:17amiconn...but in battery_bench only. The rest of rockbox is okay
21:54:37amiconnVoltage >4.095V is expected for fully charged LiIon
21:54:59googlegoogleis there any chance of rockbox going to the cowon D2
21:56:28 Quit rocko (Read error: 110 (Connection timed out))
21:58:56 Join ender` [0] (
21:59:09ficschhmmm i moved all directories from "music" to my harddisk (mv /mnt/mp3/music/* /tmp). then i chose "initialize now" and rebooted. then after choosing "database" i was asked to initilize the database. i chose yes. now there's msg. box "building database ... 512 found (prev to return)". the msg box is there for minutes now.
21:59:23 Join jhulst [0] (
22:00:11*amiconn pings Slasheri
22:00:40amiconnThat's a known bug as well. The db won't finish initialising if there are no music files on the device
22:00:53ficschkk ;)
22:02:00Slasheriargh :D
22:02:08Slasheriindeed, i haven't tried that yet..
22:02:16amiconnSlasheri: Remember? I reported this months ago...
22:02:34 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
22:03:10Slasheriamiconn: yep, but i haven't had much time to do anything atm :/
22:03:19Slasheriamiconn: will add that to the TODO :)
22:07:53Hellmarkdamn, looks like I'm getting a new iPod Nano
22:08:26HellmarkI mention I am looking at a new MP3 player, and they order me one. I was wanting a 4g or a mini to run rockbox on...
22:08:50Hellmarkso, if there is anything I can do to help with the 3G nano, let me know
22:11:48 Quit saratoga ("CGI:IRC (Ping timeout)")
22:12:14 Join Iamnotacrook [0] (
22:12:43 Quit austriancoder (Remote closed the connection)
22:14:07preglowman, that new nano is ugly
22:15:51 Quit elinenbe (Read error: 113 (No route to host))
22:16:01preglownow someone get cracking and figure out how to stuff rockbox on them :>
22:16:08 Join japc [0] (
22:17:05japci would like w access to the wiki
22:17:07Iamnotacrookhey im new to this so please dont flame me for anything (yet) i know that there is nothing for the nano 2nd gen yet but was wondering if there was any work being done on this and if there is any way i might help... an
22:17:17japcwith whom should i talk ?
22:20:13 Quit Iamnotacrook ("ChatZilla [Firefox]")
22:23:17 Join creshal [0] (i=508cf183@gateway/web/cgi-irc/
22:23:28creshal'lo guys.
22:23:36pixelmajapc: there are a few here who could grant you access but only if you tell us your wiki name ;)
22:24:28japcah, right
22:25:37creshalGot a nasty problem with... um, every build for the iPod nano since the new menu structure was introduced (half year ago or sth. like that)... these versions just show a "data abort at (random hex number)". In the forums were no useful answers (or better, no answers)... is there a way to fix this?
22:25:50 Quit ficsch ("[BX] Time wasted: 25 days 3 hours 1 minutes 57 seconds 96 milliseconds")
22:26:22pixelmajapc: you should be able to edit the wiki now
22:27:01 Quit merbanan (Remote closed the connection)
22:27:28 Quit Frazz (Read error: 110 (Connection timed out))
22:27:38japcthank you pixelma
22:27:44japcdinner and hack
22:28:09pixelmacreshal: how old is your bootloader (when did you last update it)?
22:28:11Bagdercreshal: you probably have a too old bootloader
22:28:30creshalah, could be... should be a year old.
22:28:38Bagderthen it is
22:28:48creshalk, thanks.
22:30:12 Join Stalwart [0] (
22:30:22Stalwartis it joke or what? apple released 3g nano?
22:31:03creshalNope, no joke...
22:31:05Bagderthey released it
22:31:12Bagderand two other new ipods
22:31:22Stalwartholy crap
22:31:29Stalwartmy nano is SO obsolete
22:31:36Stalwart(i have 1g)
22:31:59 Join DerPapst [0] (
22:32:04preglowis it just me, or are there no really serious attempts at cracking the nano2g?
22:32:29DerPapstit's not just you ;)
22:32:40*Stalwart cries
22:33:02DerPapstbut that may change with the new ipods.
22:33:17preglowwell, all of them are probably based on the same arch now
22:33:27Stalwartwhen did they released them?
22:33:36DerPapstthe 3rd gen nano and the 6th gen iPod (the classic one) are decrypted as well
22:33:41preglowi dunno, though, i have a feeling apple has done their homework better now...
22:33:56creshalWhat? No encryption? Yay...
22:34:03BagderDerPapst: you mean encrypted?
22:34:18*Stalwart slaps DerPapst with a rubber dildo.
22:34:32Stalwart(my random slapper)
22:34:43StalwartSome one needed the powerstrip, so they pulled the switch plug.
22:34:49Stalwart(my bofh excuses)
22:36:10linuxstbDerPapst: Has the encryption been confirmed for the 3G Nano and 6G, or are you just assuming? (I would agree with that assumption though).
22:36:48creshalNarf... it seems to run, but I can't find my earphones anymore... oO
22:36:56Hellmarklinuxstb, encryption is not confirmed, just likely.
22:37:09DerPapstlinuxstb: more some kind of insider knowledge ;)
22:37:33Stalwartrockbox will need multitouchscreen driver and new gui... oh boy...
22:37:33DerPapstso it's not hjust a guess.
22:37:51 Quit Kingstone ()
22:37:56Stalwartor maybe it's easier to port linux to touch ipod
22:38:23DerPapstor maybe just keep OSX and write apps
22:39:03creshalWell, I'm off... good night ev'ryone. ^^ And thanks again.
22:39:11 Part creshal
22:39:26StalwartDerPapst, you think it will be cracked to run native apps?
22:39:43 Join advcomp2019 [0] (n=advcomp2@
22:39:56DerPapstbut i think it's nearly the same hardware wise like the iphone
22:40:32Stalwartright now i have nokia e61 which is as much pda as i need, so i think about buying subnotebook, and unreleased yet ASUS EeePC 10" seems to fit me 200%
22:40:49preglowwhat cpu does the iphone have again? samsung arm core?
22:41:01BagderI think so
22:41:31StalwartAn analysis of the iPhone's firmware has revealed that the main Samsung chip (designated S5L8900) contains an ARM1176jzf processor, together with a PowerVR MBX 3D graphics co-processor.[67]
22:41:36Stalwart(c) wikipedia
22:42:16Bagderplenty of power to be able to run osx ;-)
22:43:59Stalwartit'd be even more cool with linux
22:44:54Stalwartmultitouch x11 is already at beta stage and multitouch apps should arrive soon (which multitouchscreens becoming reality)
22:45:11Stalwartoh, and multitouch rockbox would rock too!
22:46:21LloreanRockbox doesn't really have much use for multitouch at the moment.
22:46:29linuxstbStalwart: Rockbox is a DAP firmware - I don't see its future as being a multi-purpose OS
22:47:01LloreanWere it ever ported to a touchscreen device, I'd just designate the edges of the screen the directional controls, and quadrants away from the edges four buttons or so.
22:47:21LloreanOr something similar after testing usability
22:47:24tumurockbox doesn't really have much use for a phone at the moment :)
22:47:40Stalwartipod touch != iphone
22:48:03Bagderipod touch == iphone - phone ;-)
22:48:11LloreanBut on anything needing/expecting a full OS, Linux is better suited, and a tailored Rockbox sim (or app) like in theory exists for that one phone (did we ever hear back 'bout that)?
22:48:11linuxstbBagder: Faster than me...
22:48:21linuxstbBagder: Isn't that just "i" ?
22:48:37Bagderyeah, "ipod touch == i" ;-)
22:48:41Stalwarti, robot
22:48:59linuxstbBagder: No, I am "ipod touch"...
22:49:48tumutalking about ports, has there been any about porting to generic dvd player chipsets?
22:50:22tumusort of
22:50:33tumuthey usually come with 8bit micro plus an arm
22:50:50Bagder"usually"? really
22:51:03tumuand lot of integrated processing circuitry for audio and video
22:51:09BagderI know AMD/ATI and Nec are big on that area, they also do lots of MIPS cores
22:51:16tumuas far as i've seen
22:51:28BagderI poked on a Nec one for a DVB-T box
22:51:33***Saving seen data "./dancer.seen"
22:51:38Bagderthey have a whole family of such beasts
22:51:50tumuthe manuf dvd player firmwares really can suck
22:52:01 Join darkraven [0] (
22:52:08 Quit darkraven (Client Quit)
22:52:27 Join freqmod [0] (
22:53:09DerPapstrockbox as dvd player firmware replacement? i guess that's an entire different project :P
22:53:26tumumaybe, maybe not
22:53:31tumuthey have framebuffers etc
22:53:46tumucan output raw audio if necessary
22:55:34DerPapststill you have to rewrite huge parts of rockbox (meaning more than 70% of the code)
22:56:04Bagdertumu: feel free to work on it, let us know how it turns out! ;-)
22:56:15tumugee, thanks :)
22:57:22tumuhave to get my dvd player back from loan and start bricking it :p
22:57:43Bagderthat's the EMMA series from NEC, seems to be fairly well used
22:57:52Bagderno public docs, MIPS multi-cores
22:58:00 Quit perplexity (Read error: 113 (No route to host))
22:58:06 Quit desowin ("use linux")
22:59:50 Quit RudMan (" HydraIRC -> <- IRC has never been so good")
23:01:06DerPapstno public docs are a reason but no barrier :P
23:01:10*DerPapst hides
23:01:18Bagderhaha, indeed
23:01:25 Quit mtoups (Read error: 104 (Connection reset by peer))
23:01:31 Quit seablue ("life, death, life, death")
23:07:08 Join linuxstb_ [0] (i=5343d4aa@rockbox/developer/linuxstb)
23:09:07delYsidwhat does mi4 actually stand for?
23:09:46BagderdelYsid: :-)
23:10:03Bagderbut seriously, there's no known explanation that I know of
23:10:58linuxstb_That link is probably as good as any...
23:12:57n1sSorry for the OT but I am jus shocked about how many of the comments I've read about those new ipods are about how they look must be more than half of all comments are either "fugly!" or "the best lookingngngngn thing evar!1" </OT>
23:15:00japcwho gives a ff about the new ipods? great is the 2 year old hp140 i bought from ebay
23:15:04japcway better
23:15:50n1spersonally I think my h300 is sort of ugly but I like it anyway, it's not like I sit looking at it all day
23:16:21 Quit Febs (Read error: 104 (Connection reset by peer))
23:16:24japcyeah. kind of ugly and bulky
23:16:40japci just try not to make eye contact with him
23:16:51japcother than that is perfection
23:16:55n1syep, if I ever manage to kill it I will get a smalle rplayer
23:16:59LloreanI just replaced the battery in my H120. Don't mind the way it looks at all, and plan to keep it running for as long as it can.
23:17:22linuxstb_n1s: At least the "classic" isn't white any more...
23:17:40japcyou replaced with a stock one or with a bigger ma Llorean ?
23:17:55japclinuxstb_: black is the new white :)
23:18:14 Join webguest34 [0] (i=97c98a9f@gateway/web/cgi-irc/
23:18:29linuxstb_I just want to get my hands on one of those 160GB drives... $349 is a lot to pay for a disk though.
23:18:53Lloreanjapc: Just a little bigger, 1700
23:19:08n1slinuxstb_: true but for all I care my player could be neon-green with purple dots and pink fluff on it as long as it worked ok and sounded good :->
23:19:14chrisjs169just a random thought - does Rockbox adjust the time for DST (for players with RTC)?
23:19:19LloreanThe 1300 was enough for my use, I just needed a new one since it was ancient, and I'm hoping to avoid the misc ballooning problems.
23:19:25linuxstb_n1s: Good luck finding a manufacturer that agrees with that...
23:19:45linuxstb_chrisjs169: No
23:20:03linuxstb_chrisjs169: I think the idea is discussed for about 5 minutes twice a year...
23:20:03chrisjs169linuxstb_, I didn't think so
23:20:09n1sI might try some custom painting as a theft prevention thing :-D
23:21:27*n1s goes to bed, goodnight everyone
23:21:30 Part n1s
23:22:35delYsidhuh? "return !!QUEUE_LEVEL;"
23:22:50delYsidAm I missing something or is this a typo?
23:23:44delYsidoh, IC...
23:25:12 Quit BigBambi ("Leaving")
23:26:26*bluebrother wonders if he should add a "report problems here" to the rbutil thread
23:26:41linuxstb_Add it to rbutil itself...
23:26:50 Join BigBambi [0] (n=alex@
23:27:03bluebrotherin what way? As a "problems?" menu entry in the help menu?
23:27:20linuxstb_"Feedback" ?
23:27:29linuxstb_But no, it's probably a bad idea...
23:28:12linuxstb_Did you see the Sansa bug report in the forums - apparently rbutilqt didn't detect the device, but sansapatcher in the same usb session did?
23:29:49bluebrotherI found a couple of thread which unspecifically mentioned problems with rbutil
23:31:07linuxstb_I just read this one -
23:31:50bluebrotherhmm −− strange.
23:33:00bluebrothertoo bad I don't own a sansa for testing myself.
23:33:13linuxstb_Has anyone successfully installed the sansa bootloader with rbutilqt?
23:33:24bluebrotherI don't know.
23:33:32 Quit ompaul ("night all")
23:33:49 Quit Guile`` ("KVIrc 3.2.0 'Realia'")
23:34:03 Join Guile`` [0] (n=Guile@
23:34:06bluebrothersomeone with a sansa around willing to test? ;-)
23:34:20linuxstb_Maybe some kind of table on the wiki page could be useful - so people can test bootloader and rockbox installs and tick the box if it worked...
23:34:22BagderI can try it out tomorrow
23:34:27bluebrotherwell, if autodetection works it should be fine at least.
23:36:12preglowam i the only one annoyed by the white/gray frame around the tables? :>
23:36:15linuxstb_bluebrother: Any luck with libusb? From the sound of things, jdgordon will soon have a bootloader patcher working for the E200R, but to use that will require libusb working with rbutil.
23:36:21 Quit Guile`` (Client Quit)
23:36:35 Join Guile`` [0] (n=Guile@
23:36:52Stalwartdo you plan to fix - nano bug?
23:37:02linuxstb_Who are you asking?
23:37:03bluebrotherlinuxstb_: well, libusb inclusion in rbutil isn't a problem. But it seems to always require the driver on windows.
23:37:10BagderStalwart: we plan to have no bugs at all
23:37:21bluebrotherStalwart: should I answer "no"?
23:37:21 Part webguest34
23:37:32 Quit Guile`` (Client Quit)
23:37:45 Join Guile`` [0] (n=Guile@
23:37:45Stalwarti'm using last working revision now... but at some point you'll implement some must-have feature... =\
23:37:58preglowStalwart: i might have a look at it some day
23:37:59bluebrotherthere are no must-have features ;-)
23:38:05preglowcan't promise any more than that, my nano is not affected
23:38:10Bagderbluebrother: not even doom?
23:38:21bluebrotherBagder: remove doom! *g*
23:38:33linuxstb_Couldn't we just lower the clock speed until the problem is fixed?
23:38:55bluebrotherI could use libusb to detect the attached devices easily. But for detection only I don't want to require the user to install a driver.
23:39:12bluebrotherit might be necessary to require for e200r installation.
23:39:12preglowlinuxstb_: that'd qualify as a hack
23:39:42DerPapstStalwart: you can always compile rockbox yourself and revert the patch that toubles nano
23:40:15bluebrotherbut I'm not really sure what's the best way.
23:40:52linuxstb_preglow: I didn't say it wasn't...
23:40:57Stalwartif i ever set up toolchain i'll try to fix bug
23:41:15DerPapstsetting up the toolchain is faily easy,
23:41:22DerPapstfixing this bug is not.
23:42:09bluebrotherthere is a pre-setup toolchain hidden in some vmware image ...
23:42:30linuxstb_Or you type in the complicated command "tools/" and go to bed
23:42:46DerPapstbut i would compile myself a build excluding r14004 (or whatever breaks your nano) instead of waiting for someone to fix this bug.