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 2009-06-29

00:00:16Horschtiusualy, you only try a few themes the first time you install rockbox, then you stick with one you realy like.
00:00:28LloreanHorschti: Not true. Some people change *very* often
00:00:48LloreanBut even then, you don't need to know what you're currently using. The more often you change, the more recently you selected one, *and* the more likely you're looking for a different one anyway
00:00:52Horschtithat's why i said "usualy" and not "always" ;)
00:02:07 Join Ziltro [0] (
00:02:21HorschtiI, as a user, don't think the current behaviour needs to be changed at all.
00:03:57 Part domonoky
00:06:58ZiltroHi there, I was wondering if there is a way in Rockbox when playing a list of songs to pause before playing each one? ie. Song 1 plays, finishes, it skips to Song 2 and waits for me to press Play.
00:07:19ZiltroI have Rockbox running on my Sandisk Sansa e260, and it works well. :)
00:08:01Horschtiwhy would you want to do that?
00:08:28Ziltroplaying music at a ... well I'm not sure what you call it. like a concert.
00:08:42Ziltrothere's an intro, they sing to a song, there's another intro, they sing to another song, etc.
00:08:56Mikachulike repeat 1, but without the repeat? :)
00:08:59Ziltroit would save me pressing pause at teh end each time. :)
00:08:59bertrikZiltro, as far as I know, that's currently not possible, but maybe someone else here knows a way
00:09:16ZiltroI would expect it to be in Playback Settings, but couldn't see anything.
00:09:33Ziltroalso no music player software in Linux that I've tried has the option either!
00:09:43pixelmaI believe there is an old patch in the tracker which does this but I don't know if it still applies
00:09:45Mikachufor a in *.mp3; do mplayer $a; read; done
00:09:53Ziltrolike a playlist with manual starting of each song. :)
00:10:11ZiltroMikachu: that would work until they accidentally read the wrong song off of their list. ;)
00:10:17Horschtithat's simple... it's realy an... unconventional usage of a DAP
00:10:30 Join ForumJunkie [0] (
00:10:31MikachuZiltro: "read" just waits for you to press enter
00:10:41pixelma"pause between tracks" it was called
00:10:46MikachuZiltro: ah, you want to be able to move around in the list as well
00:11:10ZiltroMikachu: If I can't move around I know they'll accidentally pick the wrong song at some point. Sod's law and all that.
00:11:50Ziltro"Pause between tracks" sounds good. I wouldn't mind being able to disable ff/rew/next/prev too.
00:12:34kugelyou can disable next/prev at least
00:12:38Ziltrousing a DAP saves lugging yet another laptop up into the loft.
00:12:55Ziltrokugel: you can? oh I hadn't looked into that at all actually, thanks. :)
00:13:13kugelit's called "Prevent track skipping"
00:13:26LloreanZiltro: Couldn't you just play a single song and let it stop after, then play the next and let it stop?
00:13:36LloreanRather than trying to deal with a single playlist and pausing
00:13:51ZiltroLlorean: how? each one in a seperate directory?
00:14:17Ziltrokugel: ah I noticed that just now, wondered what it meant. :)
00:14:20pixelmahmm... what's with "Party Mode" then? I never used it, just know that it disabled some input but not which exactly?
00:14:21MikachuZiltro: context menu the song and explicitly insert the one entry, it won't add the whole dir then
00:14:29LloreanOr a folder full of single-song playlists, with repeat set to off
00:14:37LloreanSo you don't need to do anything when one ends except press "down" and "select
00:14:47Ziltropixelma: imagine not being able to stop the music playback ever!
00:14:50 Join tvelocity [0] (
00:14:53Lloreanpixelma: Party Mode prevents stopping / pausing / interrupting. The opposite of what he wants. :-P
00:15:14 Nick tessarakt is now known as [PIRATEN]tessara (
00:15:21kugelZiltro: I thought the naming was pretty clear. But we also have our fine manual :)
00:16:15 Join funman [0] (n=fun@rockbox/developer/funman)
00:17:08ForumJunkieAnyone know what MD5Sum does? It created a file called everything.md5sum. No idea what it is, can I delete it?
00:17:09Ziltrokugel: when I read it I thought of skipping like how a record or CD can skip.
00:17:43kugelfunman: I don't quite understand your latest response
00:18:30LloreanForumJunkie: Google has a lot of info about md5 sums. They're not rockbox specific
00:19:35funmankugel: i just answered on the task, the datasheeet mentions the Note (1) for both 1.05 and 1.10 core voltage settinsg
00:19:43 Quit Rondom ("Ex-Chat")
00:19:54 Quit ender` (" I contend that we are both atheists. I just believe in one fewer god than you do. When you understand why you dismiss all t")
00:20:12kugelfunman: ah I see, you were going by the "Typ" column
00:20:18kugelI was going by the min one :)
00:21:03*kugel wonders why the "Typ: 1.10V" is not allowed
00:21:17funmantheir hardware is not precise enough ? ;)
00:21:24kugeland why the "Typ: 1.15" gives higher MAX than the 1.20 one
00:21:54funmanyes that's weird : this is why i ask for a new battery bench
00:22:25kugelI really don't think the Typ 1.1V one will give problems though
00:23:12kugelthe min is only 0.04V under specifications, but it looks like the way better setting in general
00:23:15funmanwell you have no proof of that, so I would like to respect the requirements
00:23:24kugelsure, just saying
00:23:29Ziltrohmm "Prevent Track Skipping" does stop prev/next, however when you press one you get a silent pause for about 500ms...
00:23:37Ziltroalso ff/rew still work.
00:23:51Ziltronot quite what I'd expected... :)
00:24:38kugelyea, ff/rew still work
00:25:00 Nick [PIRATEN]tessara is now known as tessarakt (
00:25:27funmankugel: did you have a look at recording?
00:25:43kugelZiltro: I don't get such a pause
00:25:50kugelfunman: no
00:25:56kugeland I don't think I'll have one
00:26:43kugelfunman: I think we should focus on releasing soon, actually (i.e. fixing the sd bug)
00:26:58kugelif it really exists, I yet have to experience it
00:27:16funmanI have that on my TODO list, I would like to have an idea of what happens
00:27:27funmankugel: here i experience deadlocks on startup (stuck on rockbox logo)
00:27:36funmanhow often do you install a newer build of rockbox on your fuze?
00:27:45kugelnot much recently
00:27:54kugeldoes that matter?
00:28:02kugelI mean, I use it, without problems
00:28:09funmana particular build does deadlock, or it doesn't
00:28:14funmanno matter how much times you will boot it
00:28:43AlexPReminds me of a problem that used to exist on another player
00:29:01Ziltrohmm "Skip length" could include "None" or "Disabled". If you set Skip Length to a time then it still works even with Prevent Track Skipping turned on.
00:29:07Ziltrojust thinking out loud. :)
00:29:14AlexPThat would or wouldn't work seemingly (although not) at random
00:29:30AlexPi.e. One build worked, another didin't
00:29:37AlexPCan't remember which player now
00:29:45kugelZiltro: "prevent track skipping" is really only to prevent skipping tracks (accidentally)
00:30:15Mikachuhow do you do it on purpose with the setting on?
00:30:17funmanby particular build, i mean that adding an instruction to completely unrelated code will change the behaviour
00:30:30kugelMikachu: you don't :)
00:30:33AlexPfunman: Yes, this was the same IIRC
00:30:43Ziltrokugel: mmm, but if you set Skip Length to anything other than Skip Track then it has no function?
00:30:47AlexPfunman: I think this might also have depended on what build server built it
00:31:06kugelZiltro: no, it still won't let you skip over the current track
00:31:32kugelwell, I said deactivating next/prev, but that's probably not too accurate
00:31:57Ziltrokugel: ah, I see, if you have skip set to 10s and you are 9s from the end it won't skip.
00:32:10kugelfunman: that's not really what the people reported, is it?
00:32:48 Join funman_ [0] (n=fun@rockbox/developer/funman)
00:33:06 Quit funman (Read error: 60 (Operation timed out))
00:33:20funman_AlexP: perhaps all build servers do not generate the same code?
00:33:40AlexPI think it depended on versions of various things, or something like that
00:33:41funman_kugel: i have read reports of corruption, I had experienced these some time ago but not recently
00:33:45*AlexP is being very precise
00:34:12kugelfunman_: have other people actually reported your problem?
00:34:13funman_AlexP: well you're as precise as all of us ..
00:34:22funman_kugel: dunno
00:34:25AlexPWhat does that say :)
00:35:10kugelI'll try updating more often, but I have the suspicion that this is your problem (whatever it's caused by)
00:35:11funman_what i have seen is that the driver is stuck in an infinite loop because all transfers fail with DATA TIMEOUT
00:35:46kugelhave you made other changes to the code or are you trying unmodified svn?
00:36:01 Join ze [0] (i=ze@
00:36:10 Join mc2739 [0] (
00:36:38kugelI really haven't seen people having this problem, and there's quite a few people here dropping by to install rockbox on their AMS sansas
00:37:25funman_yeah I asked them on the forum to STFU, to not clutter the thread with dozens of reports saying "doesn't work"
00:37:50mc2739kugel: I have also seen the same issue as funman_
00:38:14mc2739boot stops on Rockbox logo
00:38:41mc2739rebuilding does not fix it, unless you make a code change
00:38:49kugelfunman_: probably not such a good idea in a pre-release phase
00:39:14 Quit aaron424 ("ChatZilla 0.9.85 [Firefox 3.0.11/2009060308]")
00:39:21funman_kugel: well we know it's broken, don't we?
00:39:28kugelI apparently didn't
00:39:43*funman_ invites kugel to read the forum thread more frequently ^^
00:39:50kugelI read every post
00:40:02kugelbecause I'm subsribed to it
00:40:25kugellatest svn boots fine
00:41:22 Nick funman_ is now known as funman (n=fun@rockbox/developer/funman)
00:42:22kugelthat post is two weeks old, I've long forgotten about it :p
00:42:37 Join mt [0] (n=mt@rockbox/developer/mt)
00:43:41funmanthe status didn't change unfortunately :/
00:45:48kugelfunman: maybe powering/clocking (sd_enable())on doesn't work reliably anymore and needs a delay?
00:46:50funmanat least we get to the point of working SD commands (since we can start a transfer), so i am not sure
00:47:02kugelnot sure why changing unrelated code alters the behavior though
00:47:06funmanI will look at the OF again to see if there is some delays there
00:47:37funmannot sure either ..
00:48:14bertriknotlistening, have you looked at the radio of your ams sansa player recently? it should now work properly (fixed during devcon)
00:49:14***Saving seen data "./dancer.seen"
00:52:50 Quit funman ("free(random());")
00:52:56notlisteningbertrik, yeah it is working ok, a little bit buggy but i am trying to reproduce the issues i have seen
00:53:56bertrikI know of some bugs where it hangs when the radio is 'paused'
00:54:42bertrikI think it also hangs when modifying radio settings when the radio itself is not active (e.g. set the region through the FM radio context menu in the main menu)
00:56:22kugelbertrik: indeed
00:56:30kugeljust tried that, and it crashed
00:56:39notlisteningi have not been able to modify the volume before pressing themenu button at times
00:57:00kugelwell, it doesn't crash actually, buttonlight still works (so the backlight thread is still working I think)
00:57:48bertrikkugel, I think a region change cause the radio to re-tune, but at that point the radio hardware is powered down, so the tune loop hangs
00:58:06notlisteningand i have seen corruption on my player recently, if that was bought up before?
00:58:06 Quit SUSaiyan (Read error: 113 (No route to host))
00:58:07kugelshould be fixable
00:59:17kugelJdGordon: I think I need to fix up statusbars a bit before customlist goes in
00:59:26JdGordonhow so?
01:00:06JdGordoncan you put up your latest patch? I want to do some hacking tonight so i might be able to look at it
01:00:09kugelbecause it's buggy in SVN, but is apparently more obvious with my changes
01:00:33*JdGordon doesnt think thats a major concern... it can be fixed
01:00:54JdGordonanyway, im heading off for a while, back later
01:00:57kugelyes, but it needs to be done
01:01:16 Quit matsl_ (Read error: 110 (Connection timed out))
01:02:05 Join matsl_ [0] (
01:03:56 Quit bertrik ("De groeten")
01:05:12 Join ReKleSS [0] (
01:06:42 Quit AndyI (Read error: 60 (Operation timed out))
01:07:43 Join AndyI [0] (i=AndyI@
01:14:53 Join SUSaiyan [0] (
01:22:06 Quit notlistening ("Leaving")
01:24:54 Quit mirak ("Ex-Chat")
01:27:30 Quit Thundercloud (Remote closed the connection)
01:27:55 Quit tvelocity ("Αποχώρησε")
01:30:21 Join tvelocity [0] (
01:35:32 Quit mt (Remote closed the connection)
01:36:40 Quit kugel (Remote closed the connection)
01:41:20 Quit petur ("Zzzzzz")
01:45:41 Quit robin0800 (Remote closed the connection)
01:45:50 Join cmwslw [0] (
01:49:00rasherHow isolated is the jpeg-decoding part of the jpeg viewer?
01:49:13 Part cmwslw ("Ex-Chat")
01:50:32 Quit tessarakt ("Client exiting")
01:50:40 Quit barrywardell ()
01:51:10Mikachurasher: isn't jpeg decoding in core already?
01:51:23rasherDoes the jpeg viewer use that decoder?
01:51:28rasherI'm not up to speed with that
01:51:29Mikachuno idea :)
01:51:45rasherThinking of FS #9493 and how reasonable it is to require it to be integrated with the jpeg viewer
01:52:17Lloreanrasher: While I'd agree that it'd be nice for it to become "ImageViewer" don't we need to try to keep the plugin size down (to allow loading as large of images as possible without stopping playback)?
01:52:22 Quit matsl_ (Read error: 110 (Connection timed out))
01:52:22LloreanThat seems like something a bit more difficult to address.
01:52:59LloreanI'd love to require it to be integrated, but I'm not sure it's reasonable to require that up front.
01:53:03pixelmawould be nice for the Archoses and other lowmem targets as well
01:53:11rasherTrue. It just seems a bit wasteful having to keep two interfaces up to date
01:53:27LloreanVery true.
01:53:29rasherPlus, the slideshow would be able to show both. Not possible if they're seperate
01:53:55LloreanWould it be possible to compile the plugin twice, using different decoder libraries, for the time being? So you get two plugins with the same interface code?
01:54:11rasherThat would probably be possible
01:54:12 Quit n17ikh (Read error: 104 (Connection reset by peer))
01:54:16LloreanThen, when possible combine them to dynamically load the decoder, or similar.
01:54:39pixelmaimage "codec"?
01:54:39 Join Guest26583 [0] (
01:54:42*Llorean isn't really sure how much easier it'd be, at that point.
01:54:49Lloreanpixelma: More or less. :)
01:55:01rasherLlorean: It would be better than a plugin at twice the size
01:55:20LloreanStill doesn't solve the slide show issue until the final step, but it's probably a good compromise.
01:55:21 Nick Guest26583 is now known as n17ikh (
01:59:05 Quit n17ikh (Read error: 104 (Connection reset by peer))
01:59:42 Join n17ikh [0] (
02:00:06 Quit n17ikh (Client Quit)
02:01:15 Part toffe82
02:01:17 Join n17ikh [0] (
02:02:36 Quit n17ikh (Read error: 104 (Connection reset by peer))
02:03:02 Join notlistening [0] (
02:03:54dzis there any way to get the sansa e200 bootloader to not boot the original firmware when a dock connector is plugged in?
02:04:16 Quit faemir (Remote closed the connection)
02:04:50 Join n17ikh [0] (
02:05:22Lloreandz: There's new bootloaders in flyspray that shouldn't.
02:06:36Blue_DudeI'm trying to #include a file that's higher up in the directory structure. Anybody know how to reference it?
02:12:50rasher#include "../foo.h"?
02:16:25Blue_DudeI tried that. It doesn't work.
02:26:25 Join sbhsu [0] (
02:28:37notlisteningrasher, which file was I looking in again for the voicing script
02:29:35rashernotlistening: tools/ does the grunt work
02:29:45*rasher hands notlistening a torch and a helmet
02:29:46rasherGood luck!
02:30:08notlisteningcool might be back with others in a min ;)
02:33:33notlisteningrasher, urgh!!
02:35:58 Join rayj [0] (
02:36:22rayjhello all
02:36:42rayjim in the market for a new stereo in my car
02:37:07rayji see all these new setups with usb capabale
02:37:33rayjdo any of these new usb stereos do gapless playback
02:39:27LloreanThis is a channel for questions about a specific software
02:40:41rashernotlistening: Well, it should be more or less straight forward. It's just not very pretty
02:40:52rayjok is the a port for rockbox on any carstereos
02:41:07LloreanThe full list of supported players is on the front page of the site.
02:41:12rashernotlistening: I'd just pick one of the existing TTS that works in the same way, and copy that
02:41:41rasherThat is, copying all code related to that TTS
02:41:46UnhelpfulBlue_Dude: that should work, or if the directory containing the file is in the include path, that's another way...
02:41:49rasherAnd modifying it to fit your new method
02:42:32rashernotlistening: Probably easy to add it to configure first, so you can attempt to do test builds
02:42:40Blue_DudeUnhelpful: It should work but it doesn't. I don't know why. Relative paths only seem to work going down.
02:42:40rashernotlistening: V=1 make voice will help
02:42:59 Quit rayj (Remote closed the connection)
02:43:05 Nick zitune is now known as zitune[afk] (n=zitune@bearstech/zitune)
02:44:03notlisteningrasher, got it generating voice files already just need to add in the gubbings
02:47:35 Join cool_walking_ [0] (i=cb3b81c3@gateway/web/freenode/x-5b063cfc4fb42562)
02:49:16***Saving seen data "./dancer.seen"
02:56:05ReKleSSI'm looking over the iriver h1x0 boot process, how does the code get from flash to the normal address space?
02:56:08ReKleSSis it memory mapped?
02:56:36rashernotlistening: sounds good
03:07:57 Quit lilltiger (Remote closed the connection)
03:10:35 Quit mc2739 ("ChatZilla 0.9.85 [Firefox 3.0.11/2009060215]")
03:18:37 Join aaron424 [0] (
03:18:43 Join HBK [0] (
03:23:03 Quit fred_2 ("bye")
03:23:08 Join fred_2 [0] (
03:28:14 Quit aaron424 (Read error: 60 (Operation timed out))
03:31:10 Quit ForumJunkie ()
03:31:52 Quit _lifeless (Remote closed the connection)
03:31:59 Join _lifeless [0] (n=lifeless@
03:32:38 Join daurn [0] (n=daurnima@unaffiliated/daurnimator)
03:33:21notlisteningrasher, where does you script look for the files to encode them?
03:33:32notlisteningyour ;)
03:33:51notlisteningthe build directory?
03:34:27UnhelpfulBlue_Dude: very odd... relative includes work fine for apps/plugins/lib/pluginlib_albumart.c, and pluginlib_bmp.c uses an include that's relative to apps (includes "recorder/bmp.c"
03:35:39 Quit daurn| (Read error: 60 (Operation timed out))
03:37:20rashernotlistening: I believe so. Or in the dir pointed to by the POOL environment variable. I don't believe you should even have to worry about that?
03:39:44notlisteningthe server is placing the files in the wrong place at this time that all just need to work out a way to make it put them in the right place
03:40:45rasherIt should be given a filename/path
03:40:48rasherin the script
03:45:56notlisteningahh no path as it is I am using $output
03:46:11notlisteningand that gives a filename and thats all
03:46:18rasherWell, that means "current dir" then :)
03:50:55 Quit amiconn (Nick collision from services.)
03:50:59 Join amiconn_ [50] (n=jens@rockbox/developer/amiconn)
03:51:08 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
03:51:15 Quit pixelma (Nick collision from services.)
03:51:17 Join pixelma_ [50] (n=pixelma@rockbox/staff/pixelma)
03:51:26 Quit Llorean (Read error: 104 (Connection reset by peer))
03:51:35 Nick pixelma_ is now known as pixelma (n=pixelma@rockbox/staff/pixelma)
03:51:52 Join Llorean [0] (
03:55:35 Quit notlistening (Read error: 54 (Connection reset by peer))
03:56:57 Join tom_ [0] (
03:58:38tom_right pulled the poer lead out of the PC opps :P
04:09:52 Quit tom_ ("Leaving")
04:10:43 Join notlistening [0] (
04:24:16 Quit n17ikh (Read error: 104 (Connection reset by peer))
04:25:00 Join n17ikh [0] (
04:28:39 Quit n17ikh (Read error: 104 (Connection reset by peer))
04:29:19 Join n17ikh [0] (
04:32:51 Join lilltiger [0] (n=lilltige@
04:36:23Unhelpfulwhat're you doing, exactly, notlistening?
04:36:58 Join Llorean1 [0] (
04:37:17 Quit Llorean (Nick collision from services.)
04:37:20 Nick Llorean1 is now known as Llorean (
04:38:52notlisteningi was learning how to add in command on my xchat client a bit unsucessfully to beingin with to identify myself when i login to use my nickname ;)
04:39:18 Quit BHSPitLappy (Remote closed the connection)
04:39:44notlisteningUnhelpful, was that exact enough?
04:42:15notlisteningI have just managed to getmore corruption on my e200v2 to the point where i had to format it using the OF, is there a way to now rename the partition to its original tag as from linux i can't see the partition with parted?
04:45:05 Quit perrikwp ("Leaving.")
04:45:35 Quit Blue_Dude ("ChatZilla 0.9.85 [Firefox 3.0.11/2009060215]")
04:45:53lilltigertry cfdisk
04:46:39lilltigercfdisk seems to fail for me hehe
04:48:23lilltigerahh had to be root doh :)
04:48:37lilltigeri think he means set the label
04:49:14Unhelpfuli believe that's actually part of the filesystem, not the partition table.
04:49:19lilltigerhmm no label support in cfdisk it seem bah
04:49:22***Saving seen data "./dancer.seen"
04:49:40Unhelpfuldosfslabel would do the job, i think?
04:49:40lilltigerit said free disk space!
04:50:21lilltigerwhat kinda strange fs are they using!
04:53:04notlisteningwhen you format the disk using their inbuilt untility it does strange things
04:53:23Mikachucfdisk does display the label that is set on most partition types, but i don't think it can change them
04:53:50notlisteningyeah not an option i can see Mikachu
04:54:29lilltigerMikachu: it shows my fuze fs as empty :)
04:56:13lilltigerhmm how do i get it to rebuild the database it dident do so upon restart after adding new songs
04:57:13notlisteningin the database menu under general settung you need to auto update option?
04:57:37lilltigerthanks :)
04:57:53lilltigerdident check under general hehe
04:58:21notlisteningUnhelpful, dosfslabel worked a treat thanks
04:59:19lilltigernotlistening: the cfdisk say that your devices where fat or did they say they where free space?
05:01:31notlisteningfree space
05:01:56notlisteningparted reported no parition there
05:02:59 Quit moos ("Rockbox rules the DAP world")
05:03:51lilltigerhmm told ut to update the database but it dosent seem to do it.. hehe
05:14:30 Quit Riku (Read error: 104 (Connection reset by peer))
05:17:50 Join Riku [0] (
05:22:17 Quit notlistening (Remote closed the connection)
05:31:30lilltigerhmm I think i just ran into the data corruption issue :(
05:32:46 Quit dlenski (Remote closed the connection)
05:34:34lilltigerno wonder the database dident want to update when all the newly added data looks currupted
05:35:47lilltigerhmm but when connected to the computer it looks correct hmm
05:38:03 Quit Horschti ("Verlassend")
05:48:31lilltigerok this is realy wierd, the file browser in rockbox is all messed up, it has the dirnames correct but the filenames is just strange symbols
05:49:16lilltigerbut when starting the player into the org sansa firmware there is nothing wrong at all with the sungs
05:51:49LloreanThe original sansa firmware displays information from the tags, not the filenames, right?
05:51:57safetydansounds about right
05:52:22safetydanwasn't that why the Sansa OF was slow to boot after USB disconnect because it was rebuilding the database?
05:52:32LloreanWell these are V2 sansas in question
05:52:36lilltigerbut the filenames is correct when using it in usbmode
05:52:44lilltigerit¨s the v1
05:53:01LloreanI thought you were asking questions about an e200v2 earlier.
05:53:15LloreanAnd you said Fuze up there
05:53:16lilltigeri havent
05:53:19LloreanWhich is an AMS Sansa
05:53:24lilltigerfuze v1
05:53:38LloreanSorry, we refer to all the AMS Sansas generally as "v2" a lot of the time
05:53:39lilltigerso yes ams sansa
05:54:01lilltigerconfusing :)
05:54:13LloreanWell, for quite a period the only AMS Sansas were v2s of other hardware
05:54:31LloreanThen the line extended, but there weren't any v2s of the AMS ones, so "SansaV2" made a general sort of sense still - it was v2 of their product line.
05:54:52LloreanIt's just now recently confusing again, hence the attempt to get in the habit of calling them AMS
05:55:22LloreanEither way, you've already said you're having filesystem troubles.
05:55:43lilltigerI thought i had, from looking the the rockbox filebrowser
05:55:59LloreanWeren't you having partition issues just a few minutes before that?
05:55:59lilltigerbut when looing at the files when connected as usb device theu all seem fine
05:56:10lilltigerand the sansa firmware has no u´issue with em
05:56:20LloreanThe sansa firmware is showing you the database.
05:56:23lilltigernope, it was just labling :)
05:56:34lilltigerand just checked it in cfdisk
05:56:37LloreanYou said it said your partition was free disk space.
05:56:49lilltigerin cfdisk yes
05:56:55lilltigerbut so was his
05:57:15lilltigerprobly cos it's formated from the sansa firmware
05:57:23LloreanPlease, use real words.
05:57:26Llorean"cos" is not one.
05:57:31 Join cmwslw [0] (
05:57:33LloreanNor is "probly"
05:57:55lilltigerOkej, det kan jag väl om du pratar svenska
05:58:03LloreanThis is also an English language channel
05:58:22 Part cmwslw ("Ex-Chat")
05:58:34lilltigerI said that I will start talking proper english when you start speking swedish
05:58:55LloreanNo, you'll start typing proper English when you want permission to continue using this channel, as per the publicly posted guidelines.
05:59:00 Join ruckus [0] (
05:59:18LloreanI tried to be polite about it, now I'm going to be clear: It is a guideline for this channel, and there are those who will enforce it.
05:59:24lilltigerohh ok, if me english isent good enought ill go
05:59:32lilltigerbye then
05:59:36 Part lilltiger ("Konversation terminated!")
06:02:45ruckushey folks, I am getting an "ATA error: -11"
06:03:13ruckusi'd like to know how can i send a log file or anything that may help fix this problem
06:04:22Lloreanruckus: Maybe you could give a little more information. What player, what build, etc.
06:04:57ruckuslets see. i got gigabeat f10, but i replaced the hdd with a 30gb compatible model
06:05:44ruckusi haven't updated the build in about a week, but this problem has been here for a loooong time. i'll update it in a moment, but i doubt it
06:06:19LloreanIt's probably a problem with your disk then
06:07:08ruckusthe way i get the error is a bit tricky to explain
06:07:32LloreanDid you ever get the error with the old disk?
06:08:13ruckusLlorean: i only get it when the bootloader is in a USB mode, and i remove/detach the USB cable.
06:08:21ruckus*safely remove
06:08:46ruckusi don't remember if i was having the problem with the old disk
06:09:32Lloreanruckus: What bootloader version are you using?
06:09:49ruckushow can i tell?
06:10:12LloreanWhen did you install it?
06:10:49ruckuslemme see here
06:11:13LloreanPlease, try to use real words rather than slang like "lemme" so that our logs are able to be read with translation tools more easily.
06:11:47ruckusalright, it's easier to type them, but i'll try
06:12:51ruckusi use the bootloader from this page
06:12:55LloreanWhen, though
06:12:56zehumanish thingings, conforminate to the transbawts! mwahaha
06:13:04Lloreanze: I'm serious.
06:13:10LloreanReal words.
06:13:20ruckusbut it hasn't been changed since 2007 ?!
06:13:21ze(i'm sorry, i know, i just couldn't resist that opportunity)
06:13:24Lloreanruckus: ATA error: -11 is often associated with hardware issues, if you still have your old drive, I'd see if it performs better.
06:14:03Lloreanruckus: No, that date is actually outdated, the file linked is newer than that.
06:14:03ruckusLlorean: ok, well tell me. is the bootloader constantly updated ?
06:14:07LloreanNo, the bootloader is not
06:14:14ruckusoh really?
06:14:21LloreanIt's updated when something important changes, and a new bootloader can be tested. We try to update them as rarely as possible.
06:14:33LloreanThey're really the part that makes sure even if you get a bad build of Rockbox, you can fix things easily.
06:14:54ruckusso the wiki is wrong about the bootloader's date?
06:15:14ruckusnot sure if i am following you here
06:15:21LloreanYes, the wiki's wrong
06:15:54LloreanIf you check the actual file date on the download server it's from Oct 2008
06:17:48ruckusi see
06:17:58 Join timc [0] (n=aoeu@
06:18:59ruckusLlorean: is changing the disk the only thing i can do to see what's up?
06:19:45Lloreanruckus: You could make sure everything's updated to the latest, but if the problem continues it's very likely a hardware issue. That error, I believe, usually indicates one.
06:20:29ruckuseverything's updated = bootloader and build, right?
06:21:28ruckusok. testing now
06:23:08ruckusLlorean: yeah nothing has changed. i tried to plug in the power cord as well, but nothing
06:24:36ruckusbut i do keep the old drive, so i'll give it a try
06:24:41ruckusthanks anyway
06:25:41 Quit ruckus ("Leaving.")
06:36:35 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
06:49:24***Saving seen data "./dancer.seen"
06:53:26 Join ruckus [0] (n=Earworm@
06:55:08ruckushey Llorean. i have one more question. i've been wanting to make my browsing font bigger, but whenever i change it, that also changes the font on the "now playing screen". is there a way to use 2 different fonts?
06:57:07ruckusor anyone else?
06:57:29LloreanNot really, no.
06:57:57LloreanThere are two fonts. The builtin system font, and whichever user font is loaded. You can design a WPS to use the system font, while you use the loaded font for menus, but beyond that there's not much you can do at the moment
06:59:22ruckusbut the built in font is so tiny... i am trying to make the browsing text larger, while keeping a reasonably small text on the now playing screen
07:00:21LloreanHey, I'm just telling you what your option is.
07:00:44LloreanUnless you feel up to fixing the multifont patch so that it's acceptable for inclusion.
07:01:42ruckusright. i'm just stating the predicament, with the hope that it will taken under consideration
07:02:30ruckusthough i with i was able to code, i would've whipped something out right away
07:02:57ruckus* though if i was able to code, i would've whipped something out right away (sorry)
07:04:28 Nick fxb__ is now known as fxb (
07:07:19LloreanIt's not an easy to solve problem. Or rather, a solution that doesn't come packaged with a variety of new drawbacks is difficult to come by.
07:11:07ruckusi am looking at the cfg file. how does this sound: if the "font" variable is split into font1 (for the browsing screen), and font2 (for the now playing screen) would that be a problematic to implement and make rockbox recognize each font variable? i'm trying to form a good suggestion, so i can post a forum topic or something
07:11:29LloreanThe problem is "under the hood"
07:12:02LloreanThere's a lot more to font use and caching to avoid spinning the disk and efficient memory use. Fonts can be quite huge, so we don't want to keep the whole thing in memory.
07:12:17LloreanEveryone agrees multifont would be good, it just needs a lot of work to be done in a way that works well internally
07:13:18ruckusLlorean: I see, so this idea is already on the long term to do list? or maybe short term?! :)
07:13:44LloreanThere's no "term"
07:13:53LloreanIt will happen when a volunteer works on it, just like everything else.
07:17:40ruckusLlorean: cool.
07:19:15ruckusoh no, i got stuck on the golden quotes page...
07:19:57JdGordondoes anyone actually think the buttonbar is useful?
07:22:30LloreanI don't know. How often do those buttons' functions change?
07:22:48JdGordonthey are pretty static I think
07:23:03JdGordonI would really really love to kill it, but I cant see me being allowed to
07:23:21LloreanWell, I mean, we don't really have constant visual cues as to what every other button does.
07:23:37LloreanAnd buttons like "A-B" or "Windows Logo" aren't really obvious.
07:24:15LloreanIf they're not *too* context sensitive (as in, not too hard to learn) I wouldn't be against it.
07:24:31LloreanIt's another one of those cases where a user would *need* the manual though.
07:25:00LloreanBut they're all "extra" features right? Like, normal use of the player can be had without using them (similar to the quickscreen)?
07:25:28JdGordonhonestly I dont know how those buttons are used enough....
07:25:51JdGordonbut yes, the argument is that once oyu learn them (from the manual or screen) you would disable thm to reclaim the screen space
07:27:30LloreanWell, I guess it depends on how hard it is to learn them. What conditions can cause them to do different things and all.
07:28:23JdGordonits bassically the main screens each set them slightly differently (menus, browsers, wps, rec/fm, or none?)
07:29:11LloreanThat's probably not too bad then.
07:29:27LloreanMy worry is that they might change based on other things.
07:30:10JdGordonI'm 99% sure that when they are set they are static
07:30:13LloreanFor example, I could imagine them showing different things in the tree depending on whether or not playback was happening or something. Dynamic content, which is bad if there's no indicators. But if they're static like that, it really boils down to "are they necessary, or extra"
07:30:35JdGordonI'm also not sure if they are trnaslated either...
07:30:35LloreanI don't seem to remember ever using them on my Recorder when playing with it, except when flashing it.
07:35:39 Join FlynDice [0] (
07:38:46JdGordonhmm... my mini really doesnt like plugging into usb while its off... usb doesnt connect and disconnecting the cable data aborts
07:40:37LloreanThat sounds like behaviour we had before USB was enabled.
07:40:43 Part ruckus ("Leaving.")
07:41:24JdGordonrecentish build... 2 weeks ago maybe
07:47:50 Join matsl_ [0] (
07:57:22 Nick fxb is now known as fxb__ (
08:01:30 Quit JdGordon ("Leaving.")
08:20:16 Join ender` [0] (
08:37:58 Join Rob2223 [0] (
08:43:01 Join petur [50] (n=petur@rockbox/developer/petur)
08:48:01*amiconn pokes markun a bit regarding font caching
08:49:09markunamiconn: I could take a look at my code later this week
08:49:25***Saving seen data "./dancer.seen"
08:56:33 Quit Rob2222 (Read error: 110 (Connection timed out))
08:58:10 Join bertrik [0] (
09:21:24 Quit liiwi (Read error: 110 (Connection timed out))
09:25:24 Join Thundercloud [0] (
09:32:19 Join kugel [0] (i=kugel@rockbox/developer/kugel)
09:33:33kugelmarkun: ah font caching is your work? so we can bug you with multifont? :>
09:34:53 Quit _Auron_ ("Infinity repeatedly denies rumours of plotting with zero to bring down the Universe.")
09:36:25markunkugel: I wanted to make it my work :) and yes, multifont was also planned in the rewrite.
09:37:08kugelgood that you gave bertrik the meizu then, more time for multifont :P
09:37:34markunactually it was gevaerts' meizu M3
09:37:58 Quit kachna|lappy (Read error: 113 (No route to host))
09:42:39 Join flydutch [0] (
09:42:43 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
09:42:50 Quit bertrik (Read error: 113 (No route to host))
09:49:28 Join kugel_ [0] (i=kugel@
09:51:01 Quit Thundercloud (Remote closed the connection)
09:51:07 Join kugel__ [0] (i=kugel@
09:51:14 Join _Auron_ [0] (
09:51:14 Quit _Auron_ (Remote closed the connection)
09:51:18 Join _Auron_ [0] (
09:51:35kugel__markun: grml, then you should give gevaerts yours :p
09:51:47 Quit kugel (Nick collision from services.)
09:51:52 Nick kugel__ is now known as kugel (i=kugel@
09:53:03gevaertskugel: markun doesn't have an m3
09:53:38 Quit kugel (Read error: 104 (Connection reset by peer))
09:53:47 Join kugel [0] (i=kugel@rockbox/developer/kugel)
09:58:12 Nick zitune[afk] is now known as zitune (n=zitune@bearstech/zitune)
09:59:50kugelgevaerts: I was talking about meizu, not the m3 specifically
10:00:24gevaertskugel: I don't need two m6sls or m6sps :)
10:06:51 Join kugel__ [0] (i=kugel@
10:08:10 Quit safetydan ("Leaving.")
10:11:11 Quit kugel_ (Read error: 110 (Connection timed out))
10:13:13 Quit cg_ (Read error: 104 (Connection reset by peer))
10:14:59 Join kugel_ [0] (i=kugel@
10:19:52 Join einhirn [0] (
10:20:05 Join cg_ [0] (
10:21:28 Quit kugel__ (Read error: 60 (Operation timed out))
10:26:22 Join kugel__ [0] (i=kugel@
10:31:11 Quit kugel (Read error: 110 (Connection timed out))
10:31:20 Quit kugel_ (Read error: 60 (Operation timed out))
10:38:55 Quit BHSPitMonkey (Read error: 54 (Connection reset by peer))
10:44:29 Join robin0800 [0] (
10:49:28***Saving seen data "./dancer.seen"
10:52:26 Quit cool_walking_ ("Page closed")
10:56:02 Join kugel [0] (i=kugel@rockbox/developer/kugel)
10:57:09 Join kugel_ [0] (i=kugel@
10:59:28 Join kachna|lappy [0] (
11:04:08 Join mt [0] (n=mt@rockbox/developer/mt)
11:05:05 Quit kugel (Read error: 60 (Operation timed out))
11:10:53 Join kugel [0] (i=kugel@rockbox/developer/kugel)
11:13:00 Quit kugel__ (Read error: 110 (Connection timed out))
11:24:01 Quit at0m ("after years of apt-get dist-upgrading, time for fresh install to aptitude safe-upgrade!")
11:25:10 Quit jordan`` (Read error: 104 (Connection reset by peer))
11:25:14 Join jordan` [0] (i=gromit@
11:27:43 Quit kugel_ (Read error: 110 (Connection timed out))
11:37:16 Join kugel_ [0] (i=kugel@
11:41:02 Join kugel__ [0] (i=kugel@
11:45:11 Quit J-23 (Remote closed the connection)
11:45:41 Join J-23 [0] (
11:47:01 Nick fxb__ is now known as fxb (
11:56:46 Quit kugel (Read error: 110 (Connection timed out))
12:01:30 Quit kugel_ (Read error: 110 (Connection timed out))
12:05:35 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
12:06:55 Quit kugel__ (Read error: 110 (Connection timed out))
12:13:50 Quit tchan (Remote closed the connection)
12:14:28 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
12:19:34 Quit vedlith ("unhandled exception")
12:22:41 Join kugel [0] (i=kugel@rockbox/developer/kugel)
12:23:34mtIs there anything special in using lseek() in a metadata parser ?
12:23:50 Join barrywardell [0] (
12:24:32mtlinuxstb: What I've noticed so far is that a 'Data abort' occurs after the first call to lseek(). (although lseek's return value would be correct)
12:26:35mthere's the output :
12:26:39 Join vedlith [0] (
12:28:38gevaertsmt: Data abort usually occurs when you're doing an unaligned memory access. Maybe that's the case here?
12:31:14mtgevaerts: Unaligned accesses are allowed in the sim right ? Because if so, thise could be the case yes.
12:31:38Torneyah, the sim doesn't check
12:31:58linuxstbOr rather, it's not an issue on x86.
12:32:07 Nick J-23 is now known as tw-bocian (
12:32:14 Nick tw-bocian is now known as J-23 (
12:32:55Tornei guess. run the sim on arm? :)
12:36:22mtIsn't 1 word = 4 bytes on arm ?
12:36:30 Join kugel_ [0] (i=kugel@
12:37:00 Join ForumJunkie [0] (
12:37:05 Quit bmbl (Read error: 110 (Connection timed out))
12:37:48mtand starting at byte_offset 68 in a file is equivalent to a word_offset of 17
12:38:37Torneit's not the offset in the file that needs to be aligned
12:39:24Torneyou can read() from any offset you like, but if the buffer you read into isn't word aligned and you do word accesses to it, it will die
12:42:32 Quit kugel (Read error: 60 (Operation timed out))
12:43:25 Join Metalstar [0] (
12:43:33mtTorne : So if reading a half-word then a word into the buffer, the half-word will need to padded so that the following reading of word would be aligned, right ?
12:44:07Torneno, the read(0 takes care of alignment iirc
12:44:19Torneit's when your code accesses the data in the buffer that it needs to be aligned
12:45:26 Join Metalstar- [0] (
12:45:35 Nick Metalstar- is now known as Edwards80 (
12:46:26 Nick Edwards80 is now known as EdwardsEighty (
12:46:41 Nick EdwardsEighty is now known as Edwards80 (
12:46:58 Join martian67 [0] (i=martian6@about/linux/regular/martian67)
12:47:30 Quit martian67 (SendQ exceeded)
12:47:31 Quit Edwards80 (Read error: 104 (Connection reset by peer))
12:47:45 Join Edwards80 [0] (
12:48:43 Quit Metalstar (Read error: 104 (Connection reset by peer))
12:49:27 Join xplagu3 [0] (
12:49:31***Saving seen data "./dancer.seen"
12:50:30xplagu3hi all. how do i tell my ipod 4g to sleep
12:50:55Tornehold down play/pause for some seconds to turn it off
12:51:00Tornerockbox does not support sleep mode
12:51:25xplagu3ah cool. thx
12:52:06 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
12:52:12Tornealso, see the manual :)
12:53:31 Quit martian67 (SendQ exceeded)
12:54:07xplagu3will do. catch
12:54:09 Quit xplagu3 ("Leaving")
12:54:47 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
12:55:13 Quit agaffney (Read error: 104 (Connection reset by peer))
12:55:43 Quit gevaerts (Read error: 110 (Connection timed out))
12:55:56 Join agaffney [0] (n=agaffney@gentoo/developer/agaffney)
12:56:33ForumJunkieUm, I'm trying to compile a RB build, how do I delete the source code that it's received from SVN?
12:57:18ForumJunkieI deleted the "rockbox" folder that I thought had all my modifications, and now when I go to patch it says that there is two of the same file (one is "working copy
12:57:33ForumJunkie" and the other is "revision xxxxxx")
13:07:59linuxstbForumJunkie: Just deleting the folder containing the source code should do it...
13:09:06 Join DataGhost [0] (i=dataghos@unaffiliated/dataghost)
13:15:25ForumJunkieI can't patch it though, still..
13:15:32ForumJunkieIt tells me I have two different codes..
13:16:20ForumJunkieI deleted the .../home/.../rockbox folder in Cygwin, but for some reason it doesn't patch after I obtained a new source code from svn co svn://etc.
13:23:02TorneReKleSS: offtopic stuff should go in #rockbox-community :)
13:23:27ReKleSSahh crap, sorry
13:23:29ReKleSSwrong channel
13:23:45ReKleSSI normally only have one irc window open...
13:24:01linuxstbForumJunkie: What is the precise error message you're getting from "patch" ? Paste it to somewhere like if it's more than a couple of lines.
13:25:23 Quit kugel_ (Read error: 113 (No route to host))
13:30:33 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
13:36:35 Join Rondom [0] (
13:45:30 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
13:49:47 Join notlistening [0] (
13:50:26 Quit robin0800 (Remote closed the connection)
13:54:14 Join kugel [0] (i=kugel@rockbox/developer/kugel)
13:56:54CIA-71New commit by 03funman (r21550): Sansa AMS: correct sd_select_bank() in SD driver ...
13:58:04 Quit Rondom (Remote closed the connection)
14:06:57 Quit Edwards80 (Read error: 104 (Connection reset by peer))
14:08:19 Quit Rob2223 ()
14:08:43 Join Rondom [0] (
14:09:39 Join Rob2222 [0] (n=Miranda@
14:10:43 Join aaron424 [0] (n=chatzill@
14:15:01mtok, now the parser finishes cleanly. Have to figure out where it crashes now.
14:18:01mtCodecs have no access to logf ?
14:21:28 Nick zitune is now known as zitune[afk] (n=zitune@bearstech/zitune)
14:21:47 Quit FlynDice (Remote closed the connection)
14:26:17mtlinuxstb: ^ ?
14:28:31linuxstbmt: They should do, but via ci->logf(). Or if not, you could add it to the codec API.
14:29:53 Join LambdaCalculus37 [0] (i=44a0430d@gateway/web/freenode/x-cad32bd0bb09d9ad)
14:36:05mtlinuxstb: They do but I should have put "#define LOGF_ENABLE" and "#include "logf.h" " before including the codeclib. :)
14:37:09kugelcodec lib should include logf.h? Wouldn't it give a warning otherwise?
14:39:02 Join zhufeng [0] (n=7ae126c2@gateway/web/cgi-irc/
14:39:36zhufenghey guys
14:42:46notlisteningfunman, trying r21550 for you now
14:44:01 Quit Sajber^ ("Leaving.")
14:44:04mtkugel: Yes, I should define LOGF_ENABLE only.
14:46:17 Join Sajber^ [0] (
14:48:33 Join robin0800 [0] (
14:49:33***Saving seen data "./dancer.seen"
14:50:25 Quit zhufeng ("CGI:IRC")
14:50:47 Join zhufeng [0] (n=7ae126c2@gateway/web/cgi-irc/
14:51:02mtAnd now I'm getting no (Data abort)'s, Metadata tags are correctly displayed, but rockbox freezes.
14:52:12 Join KBH [0] (
14:55:06 Quit zhufeng (Client Quit)
14:55:30 Join zhufeng [0] (n=7ae126c2@
14:56:23 Quit zhufeng (Client Quit)
14:56:36 Join zhufeng [0] (n=7ae126c2@gateway/web/cgi-irc/
15:00:39 Quit zhufeng (Client Quit)
15:00:57 Join zhufeng [0] (n=7ae126c2@gateway/web/cgi-irc/
15:05:41mtlinuxstb: After the metadata parsing is done (metadata/rm) the codec is immediatly called (in this case codecs/cook) right ?
15:09:40 Quit HBK (Read error: 110 (Connection timed out))
15:19:23 Join fyre^OS [0] (
15:19:41 Quit fyrestorm (Read error: 104 (Connection reset by peer))
15:25:30notlisteningfunman, r21550 is still giving me corruption issues see the forum for more
15:26:20 Quit KBH (Success)
15:27:12 Quit zhufeng ("CGI:IRC")
15:27:26linuxstbmt: I'm not sure exactly when it's called. It may even be started before the metadata parsing is done - hence the while() loop at the start of the codec's main() function to check for ci->taginfo_ready.
15:28:32 Join n1s [0] (n=n1s@rockbox/developer/n1s)
15:32:01mtlinuxstb: hmm.. the problem is, I've thrown a couple of logf's in cook.c just to make sure the buffering thread actually reaches the codec, but the never get displayed, so I'm now not sure where exactly it freezes.
15:32:18 Join Jaykay [0] (
15:32:22mts/the never/they never..
15:40:01Jaykaypixelma: do you have some time for the german translation? i want FS #9754 to be closed but not without discussing the translations you left out :)
15:43:09pixelmadon't have much time now, probably better after Wednesday or even Thursday
15:49:13 Quit _lifeless (Read error: 110 (Connection timed out))
15:52:08 Quit robin0800 (Remote closed the connection)
15:52:28 Join evilnick [0] (i=0c140464@gateway/web/freenode/x-fc28cae71de71766)
15:54:19 Quit n1s (Read error: 110 (Connection timed out))
15:55:01 Join DarkDefender [0] (
15:56:20evilnickI've noticed a potential bug in the View Playlist screen; if you are Moving one song when the track changes then the song reverts to it's initial place in the playlist. Is that expected behaviour?
15:57:50 Quit Rondom (Remote closed the connection)
15:57:53 Quit kugel (Read error: 110 (Connection timed out))
16:00:50 Join kugel [0] (i=kugel@rockbox/developer/kugel)
16:04:17 Join fml [0] (n=4fd3c6e9@gateway/web/cgi-irc/
16:05:34fmlHello. I'd like to commit FS #10392 (don't modify the file name) and FS #10390 (correctly simulate get_time on non-RTC platforms). Any objections?
16:05:42 Join mt_ [0] (n=mt@
16:06:20 Quit mt (Read error: 113 (No route to host))
16:06:32kugelfml: I think rasher can commit himself
16:06:54 Nick mt_ is now known as mt (n=mt@
16:07:55rasherIf no one can see a problem with the FS #10392 feel free, or I'll do it some time later
16:09:50fmlIt's a good tone to not to commit own patches :-)
16:10:05fmlrasher: can you assit me with FS #10390 ? :-)
16:10:48 Join Blue_Dude [0] (
16:11:50Blue_DudeOK, I'm at my wit's end here. I'm trying to create a library file that's reference by both apps and plugins. But no matter where I put the .h and .c files, there's a linker error someplace.
16:12:13 Join funman [0] (n=fun@rockbox/developer/funman)
16:12:45funmanhi, i want to commit fs#10274 after I have tested build on some targets, any objection?
16:13:01fmlrasher: I don't see any problems with FS #10392, but I didn't test it. I assume you did. Actually, strchr was used there to convert from "const char*" to "char*"
16:13:19linuxstbBlue_Dude: Is this lib always going to be used by the core? Or is it a case of some targets using it in the core (and plugins), and other targets just using it in plugins?
16:14:01funmani will keep the 9 commits separated
16:14:02LambdaCalculus37funman: What's FS #10274?
16:14:04*LambdaCalculus37 goes to look
16:14:13Blue_DudeIt ought to be used most of the time by the core. It is used occasionally according to plugin's requirements.
16:14:19funman FS #10274 - Move timer code in target tree
16:14:31linuxstbBlue_Dude: What do you mean by "most of the time"? "For most targets" ?
16:15:15LambdaCalculus37funman: What targets did you try it on so far?
16:15:23Blue_DudeThe library is for fixed point math. Right now code fragments are scattered all over the place. I'm trying to consolidate them into a single library.
16:15:34funmanLambdaCalculus37: the one i have, as3525
16:15:56rasherfml: I tested it in combination with FS #10391, which fixed the bug that cropped up (trying to use the filename after the call to set_file)
16:16:04Blue_Dudelinuxstb: Maybe not all targets will need it (SWCODEC only maybe), but most should.
16:16:08funmani didn't change the code, only the place where it is, so I expect no regression except maybe a red build
16:16:14kugelfunman: sure, it's a good time to commit "difficult" patches
16:16:21rasherfml: I don't think it causes any bugs currently, but it's probably only a matter of time
16:16:24LambdaCalculus37funman: We should probably test with other targets a little beforehand.
16:16:40LambdaCalculus37Just to make sure it doesn't bork any other target.
16:16:46fmlrasher: but I think FS #10392 is good on itself. If not the strchr call, the compiler had complained.
16:16:50kugelLambdaCalculus37: we can make sure afterwards
16:17:06funmanyeah it's only copy/pasting
16:17:45rasherfml: Yeah, it's a fix on its own, which is why I opened a task for that rather than making it part of FS #10391
16:18:54LambdaCalculus37kugel: Okay.
16:19:28 Join jgarvey [0] (
16:20:45fmlrasher: I'll commit it now. I've also changed the types of fptr and ptr to "const char*". OK?
16:21:03Blue_Dudelinuxstb: A codec needs it too. So, yeah, all builds need to see it.
16:21:50rasherfml: No idea about that, honestly.
16:22:40fmlrasher: just to make sure (through compiler) that nothing is changed through these pointers (the same function).
16:23:26 Quit flydutch ("/* empty */")
16:23:37evilnickEr, is it worth me putting the possible playlist bug into Flyspray?
16:24:10kugelOf course
16:24:23 Quit dmb (Read error: 113 (No route to host))
16:25:47CIA-71New commit by 03alle (r21551): Do not modify the passed in parameter string (FS #10392 by Jonas Häggqvist)
16:27:34 Join cmwslw [0] (n=cmwslw@
16:28:02 Part cmwslw ("Ex-Chat")
16:28:30Blue_DudeOnce a module is compiled in one place, is it inaccessible to another? For instance, if it's compiled into the main app, can it be linked to by a plugin or codec?
16:28:53CIA-71New commit by 03funman (r21552): Move Sansa AMS timer code in the target tree
16:28:59CIA-71New commit by 03funman (r21553): Rename TIMER_REGISTER macro to TIMER_START and TIMER_UNREGISTER to TIMER_STOP to reflect what they does exactly. ...
16:29:05kugelBlue_Dude: just add the functions to the plugin/codec api
16:29:05CIA-71New commit by 03funman (r21554): Move PNX0101 timer code in the target tree
16:29:10CIA-71New commit by 03funman (r21555): Move coldfire timer code in the target tree
16:29:20CIA-71New commit by 03funman (r21556): Move SH7034 timer code in the target tree ...
16:29:40CIA-71New commit by 03funman (r21557): Move PP (last target) timer code in target tree ...
16:29:46 Quit ReKleSS ("Leaving")
16:29:51CIA-71New commit by 03funman (r21558): Remove int_prio argument from timer_register, and move the only use for it into alpine_cdc plugin, since this plugin is only built on SH7034 ...
16:30:01CIA-71New commit by 03funman (r21559): Remove the TIMER_* macros and declare target-specific functions in timer.h
16:30:17Blue_Dudekugel: I could. The problem is that the same code is needed in the main app also. But when I try to use it, I get linker errors. I'm thinking I might need to force compilation more than once.
16:30:20CIA-71New commit by 03funman (r21560): Put TIMER_FREQ definition in CPU-specific config, and remove timer-target.h ...
16:31:08kugelBlue_Dude: weird, what's the linker problem?
16:31:42kugelyou need to append ci-> and rb-> respectively to the functions
16:31:47Blue_Dudekugel: plain vanilla. "Undefined reference to " such and such.
16:31:50CIA-71New commit by 03alle (r21561): Make the get_time on non-RTC sims behave like target, i.e. always return a constant time (FS #10390)
16:32:17kugelBlue_Dude: such as rb->foo() if you call the core function foo() from within a plugin
16:33:42GodEaterkugel: you mean "prepend", not "append" :)
16:33:50kugelyea, sorry
16:33:55Blue_Dudekugel: OK, then, how do I go about setting up the pointer to those functions within the rb structure?
16:33:59 Join kasimir [0] (n=kasimir@
16:34:33kasimirHello all
16:34:57GodEaterBlue_Dude: if it's not already present in the plugin API you add it to plugin.[ch]
16:34:58kasimirIt seems that I consistently get the error "No partition found" when I try to install rockbox
16:35:01Blue_Dudekugel: I'm trying to simplify code, but it might not be worth the effort going in and rewriting everything that needs it.
16:35:13GodEaterBlue_Dude: assuming you're trying to use it in a plugin
16:35:14funmanhum aren't new builds started if a commit happens while a build is in progress ?
16:35:30GodEaterkasimir: helps if you tell us what player you're using
16:35:54kasimirApple 5G 30GB
16:36:00kasimirit seems that it should be supported
16:36:13GodEaterkasimir: is it currently HFS+ formated ?
16:36:14kasimirIPOD version: 0x000B0011
16:36:22kasimirit is FAT32 formatted
16:36:32Blue_DudeI'm trying to put basic math routines in one place. But if I put it where apps can see it, it fails linking with the plugins and codecs. And vice versa.
16:36:47fmlGodEater: do you use the setting "Start screen = WPS"? If yes, could you please comment on the mailing list about how it should behave?
16:36:55kasimirGodEater: but the formatting could be rather off... I got this iPod from a friend, and he dd if=/dev/zero of=/dev/sdb the whole drive
16:36:55gevaertsBlue_Dude: won't most of the bits that need this be performance-sensitive?
16:37:02Blue_DudeI'm not trying to call core routines as such.
16:37:05TorneBlue_Dude: yes. codecs and plugins can't link directly to core functions
16:37:14funmanis there a way to force a rebuild ?
16:37:26kasimirshould I try HFS+ ?
16:37:29GodEaterkasimir: "a bit off" ? If he did that to it, there's no partitions or formatting of any kind
16:37:38kasimirwell, I added what I could find online
16:37:43kasimiran "empty" partition
16:37:46kugelfunman: no need
16:37:47kasimirto 5 cylinders
16:37:48Blue_Dudegevaerts: They're already called as functions in scattered code bits. I'm not trying to pull them out of inline code.
16:37:58GodEaterfml: you mean the discussion on the dev list ?
16:38:03Tornekasimir: you need to either put a copy of the apple firmware back on there by hand or just restore it with itunes
16:38:13Tornekasimir: rockbox can't be installed on an empty disk
16:38:18kasimirTorne: I did put on the latest Apple firmware
16:38:27kasimirTorne: and it plays music/video/etc just fine
16:38:37kasimirTorne and GodEater: Thanks, btw, for the helping me :)
16:39:03fmlGodEater: yes
16:39:08GodEaterkasimir: what did you use to put the apple firmware on with then ?
16:39:09kasimirTorne: the "empty" partition is the partition label in fdisk, partition label type 0
16:39:13Tornekasimir: you did ?
16:39:24kasimirnope, I didn't know about that!
16:39:40Torneit's possible ipodpatcher is more strict about partitoin types etc than the apple firmware is :)
16:39:47gevaertskasimir: what type is the fat partition in fdisk?
16:39:53GodEaterfml: I don't feel qualified to wade in, it's not my use case
16:40:12kasimirwhat should my partition map format be?
16:40:16kasimirI'm using DOS right now
16:40:28GodEaterit should be using a DOS MBR
16:40:34kasimirokay, that's good
16:40:35GodEaterwith two primary partitions
16:40:40GodEaterthe first should be type 0
16:40:54GodEaterthe second should be whatever it says on that wiki page
16:41:01TorneDOS fdisk won't be able to set partition type 0, btw
16:41:13Torneyou will need a smarter partitioning tool :)
16:41:19GodEaterah crap it doesn't say there
16:41:39Tornethe other partition should be whatever the type for fat32 lba is
16:41:48Tornei forget the hex value
16:41:48kasimirTorne: ah, thanks!
16:41:54kasimirTorne: I was using "b", which was without lba
16:42:03kasimirthanks all, I will try it again and see what happens!
16:42:03Tornei think, anyway
16:42:08Torneit doesn't actually matter for dos/windows
16:42:12GodEaterI thought b was the right one to use
16:42:24TorneGodEater: it shouldn't matter unless apple's firmware checks, really
16:42:30Tornethe FAT partition types are maximums
16:42:38Tornei.e. it's ok to have a fat16 partitoin of type fat32
16:42:48fmlGodEater: you don't use the WPS as the start screen? Or you do use it but never were in the situation described there?
16:42:53linuxstbRockbox doesn't check partition types anyway...
16:43:00GodEaterTorne: you really do need to finishing writing the fat32 support into ipodpatcher ;)
16:43:07gevaertslinuxstb: it does
16:43:11Tornewhat? :)
16:43:19GodEaterfml: I do use it, but never in the manner described in the thread
16:43:21linuxstbgevaerts: It does? When did that change?
16:43:23Torneoh, you mean when i implied i might be willing to write the formatting code
16:43:33GodEaterTorne: implied? :)
16:43:39GodEateryou said you could, if you could be bothered :)
16:43:47Tornewell that's a big caveat! :0
16:44:00linuxstbipodpatcher already has formatting code, just not with support for sectors other than 512 bytes.
16:44:13GodEaterlinuxstb: that's what Torne said he could fix
16:44:20gevaertslinuxstb: see disk.c, lines 51 and 171
16:44:57 Nick zitune[afk] is now known as zitune (n=zitune@bearstech/zitune)
16:45:06linuxstbgevaerts: Hmm, r17530, not recent...
16:45:09GodEaterlinuxstb: that's why we need to convince him to do it :)
16:45:29fmlGodEater: but would you expect the playback to start after power on if the last time you switched off the player the playlist has played to the end and stopped?
16:45:34*GodEater tries to remember what Torne was saying about RuggedFAT too
16:45:43Torneoh, yah
16:45:44linuxstbkasimir: But anyway, "b" is what ipods come with from the factory., so that's fine.
16:45:47GodEaterfml: I have no opinion one way or the other
16:45:53kasimirlinuxstb: hm...
16:46:16kasimirlinuxstb: it could be that the FAT partition wasn't formatted for 2048 size sectors
16:46:16linuxstbkasimir: Are you sure it's a "5g" (512 byte sectors) and not a "5.5g" (2048 byte sectors)?
16:46:18Tornethere's a way to order disk writes on FAT such that powering off mid-operation only causes at most one filesystem error which can always be fixed :)
16:46:35GodEaterthat was it
16:46:36Tornebut if i actually wrote that for you then i'd probably be in trouble :)
16:46:38fmlGodEater: ok. So it's not a theorem, it's an axiom (i.e. should just be defined) :-)
16:46:39kasimirlinuxstb: I was told it was a "5g" by my friend, but it seems that parted complains about 2048 byte sectors
16:46:41GodEaterthere was something about NDAs too ;)
16:46:52TorneGodEater: the fact that it exists is documented in the book
16:46:56Torneso i can tell you :)
16:46:58GodEaterfml: and I care not which way you care to define it ;)
16:47:22GodEaterI've never been in a situation where my playlist is finished, and I doubt I ever will be
16:47:24linuxstbkasimir: But anyway, that wiki page you've been pointed to has the correct instrucitons - they should work...
16:47:35 Quit Jaykay (Read error: 110 (Connection timed out))
16:47:42kasimirlinuxstb: yes, thank you very much. I am re-going through them now :)
16:48:57Torneso hm,if ipodpatcher can already format FAT just not for large sectors..
16:49:08fmlGodEater: so our only hope for a qualified opinion was in vain! :-)
16:49:11Tornecan it do the rest of the manual restore process to? :)
16:49:24GodEaterfml: why would my opinion be qualified ?
16:49:36linuxstbTorne: Not a restore, but it can (I think...) do a "−−convert" to convert from HFS.
16:49:37***Saving seen data "./dancer.seen"
16:49:52linuxstbThe complication with a restore is that ipodpatcher doesn't know what ipod it is...
16:49:58Blue_DudeNow I'm trying separate header files and multiple SOURCE includes. Making now to find out.
16:50:16GodEaterlinuxstb: and by "convert" you mean "reformat" ? :)
16:50:18linuxstbAnd it's also a function that can be done in various other ways, and I fight feature bloat...
16:50:34linuxstbGodEater: Yes and no...
16:50:49fmlOh, much yellow and even red!
16:50:58GodEaterlinuxstb: it keeps the content ?
16:51:14Tornelinuxstb: i just mean from the POV of not having to mess around with dd
16:51:22linuxstbGodEater: No.
16:51:25Tornelinuxstb: could still expect the user to go get the firmware etc.
16:51:26*LambdaCalculus37 sees red everywhere on the build table!
16:51:29Tornei dunno. maybe not worth it.
16:51:30 Quit matsl_ (Read error: 110 (Connection timed out))
16:52:15 Quit antil33t ()
16:52:19*GodEater still thinks he'd prefer FAT expertise directed at TFAT on the beast
16:52:30linuxstbGodEater: The Apple partition map is converted to a DOS MBR, the firmware partition is left intact in the same physical location, and the main data partition is formatted to FAT32
16:52:32GodEaterto see if we can prevent random reformats
16:52:35 Join Strife89 [0] (n=nds@
16:52:41pixelmathe Ondios don't have an RTC in case that info helps
16:52:48 Join jfc^3 [0] (
16:52:49GodEaterlinuxstb: that's pretty much what I envisioned
16:52:54CIA-71New commit by 03funman (r21562): Build archos timer code for ONDIO (fix red)
16:53:22TorneGodEater: I don't know anything about TFAT and don't have a beast, so hey :)
16:53:23 Join toffe82 [0] (n=chatzill@
16:53:33linuxstbTorne: I could lend you mine...
16:53:49linuxstbAnd you are clearly keen on learning about TFAT...
16:54:03GodEaterlook into his eyes. Not around the eyes, just into his eyes.
16:54:06Tornebut playing with filesystems is kinda fun i guess.
16:54:21Tornei have a horrible python script somewhere that does hardlinks on FAT
16:54:29Tornefor the old mp3 player i used to have that had no tag db
16:54:55Torneyou can just crosslink the same files into lots of directories to pretend you have tag-based indexing :)
16:55:00Torneas long as you never plan on running chkdsk ever again
16:55:04*GodEater shudders
16:55:50kugellol :)
16:56:15Tornehey, it works pretty well on players that have no ability to edit/delete files anyway
16:56:49GodEaterinnovative approach I suppose
16:57:06funmankugel: the binsize increase can come from the fact that the functions are de-inlined. I won't check since it's a small increase
16:57:47kugelfunman: there's a py script for that :)
16:58:12CIA-71New commit by 03alle (r21563): Fix yellow introduced in r21561
16:58:26kugelthe linux kernel also has such a script
16:58:59 Join _lifeless [0] (n=lifeless@
17:03:22kasimirthank you guys for helping me getting the partition read! however, now, after trying to reinstall/install rockbox several times, it fails with "Can't load rockbox.ipod: File not found"
17:03:39kasimirwould you guys be able to point me in another direction? :)
17:04:04Tornepresumably it means what it says, no? :)
17:04:14Tornethough the bootloader is not very specific about why it can't open files
17:05:26 Join ocean_ [0] (i=d59c23f7@gateway/web/freenode/x-d589fce7ecc2e0df)
17:06:03kasimirTorne: where should the file be?
17:06:09kasimirI have a .rockbox directory
17:06:33Torneit should be in there. :)
17:06:44funmanJdGordon: could you comment on FS #10283 ? (r12907 introduce a very weird casting of an integer to a pointer)
17:07:03kasimirTorne: hm... could it be a permissions issue?
17:07:06kasimirthe file is definitely in there
17:07:09Tornethere are no permissions on FAT
17:07:13Torneso no
17:07:39kasimirthen I'm not sure why it wouldn't be reading it
17:07:49kasimiris there some way I could get more information from the bootloader?
17:08:08kasimirI ran mkfs.vfat -F 32 -S 2048 /dev/sdb
17:08:10Tornenot really
17:08:11kasimirto format the partition
17:08:20Torneit says "file not found" if there is *any* error opening the file.
17:08:32kasimirand then I used the rbutil program install these files
17:08:33Torneit only says other things if it's the wrong firmware or similar
17:08:45Torneso it could be various kinds of filesystem corruption..
17:08:57 Quit jfc^2 (Read error: 110 (Connection timed out))
17:08:57funmankugel: for a fuze build, the binsize difference is -8 !
17:09:07Tornenot sure what the bootloader does if it can't mount the partition?
17:09:13funmanbut the elf file is bigger?
17:09:24linuxstbTorne: It should say "no partition found" - kasimir's earlier issue
17:09:38Torneah yes :)
17:09:58linuxstbkasimir: What command did you use to format?
17:10:09Tornekasimir: i hope you don't really mean /dev/sdb above
17:10:11kasimirlinuxstb: mkfs.vfat -F 32 -S 2048 /dev/sdb
17:10:16kasimirI don't
17:10:18Tornegood :)
17:10:20funmanI think the output is reversed, if that's right the binsize increase come from deinlining
17:10:48linuxstbkasimir: You could try mtools - early tests said that only that worked, but then someone else claimed mkfs.vfat was OK, so the instructions were changed.
17:10:59kasimirokay, I will try that then
17:11:28Torneif that doesn't help one optoin is to stuff rockbox in osos just to get it to boot and see if rockbox can browse the filesystem ok :)
17:11:59GodEaterdid I miss where we found it's definitely a 5.5G ?
17:12:15Tornethat's a good point also :)
17:12:25fmlHe-he, the build table looks like a big tetris game!
17:12:56kasimirGodEater: yes, sorry
17:13:16kasimirGodEater: I think that whomever I got it from was a bit confused about that part, that solved my first issue, and thanks guys!
17:13:32GodEaterah yes, I see where you said parted was complaining now
17:13:46mtlinuxstb: I'm kind of stuck :/ .. after metadata parsing the player freezes without displaying any errors. Not sure where to look next.
17:14:05 Join gregzx [0] (
17:14:09linuxstbGodEater: He said that something complained about 2048 byte sectors
17:14:21GodEaterlinuxstb: yes, I noticed
17:14:33linuxstbAnd I've just noticed you noticing...
17:14:42 Join bambu [0] (
17:14:43*linuxstb makes note to read _all_ backlog before replying...
17:15:04GodEateryes, mtools would be my next choice then
17:15:16GodEaterI'm still not convinced of mkfs.vfat
17:15:58linuxstbMaybe mkfs.vfat works if you've previously used mtools, or something odd like that...
17:16:07funmankugel: binsize increase is 50% less now, so it only comes from deinlining
17:16:10bambuhi. i have a problem with my rockbox for (iaudio x5 player). when i look in the music folder, i don't see any files except a playlist. when i look at database, the files are there and i can play them, but i can't see them in the folder or select them from there. can anyone help?
17:16:25linuxstbbambu: Sounds like you set the "show files" setting to "playlists"
17:17:27bambuwow i feel dumb
17:17:35Torneis it just me or do people get that one a lot? :)
17:17:35bambuthanks linuxstb
17:17:45Blue_DudeWell, that worked. I had to put the foo.c file in two different SOURCE files and two copies of the foo.h file. But it all works.
17:17:50Tornei assume it's because it's on the quickscreen and thus somewhat possible to press without realising?
17:17:50bambui dont ever remember changing that setting so i didnt even think that was an option
17:17:52linuxstbTorne: Yes, they do. I think it's easy to change accidentally in the quickscreen
17:18:10Tornemaybe it shouldn't be on the quickscreen? :)
17:18:12 Quit Sajber^ (Read error: 104 (Connection reset by peer))
17:18:21 Quit fml ("CGI:IRC 0.5.9 (2006/06/06)")
17:18:26 Quit kugel (No route to host)
17:18:39*GodEater likes it just fine on the quickscreen
17:19:11TorneGodEater: i'm just saying. :)
17:19:11bambuanyone here have x5 iaudio?
17:19:19 Quit funman ("free(random());")
17:20:04GodEaterTorne: if we removed it, people would stop changing it, and we'd lose our "quick win" support scenario. People would ask us hard questions instead!
17:20:11LambdaCalculus37bambu: A couple of people do. Do you have a question?
17:20:26*GodEater is not fond of hard questions
17:20:35TorneGodEater: Hehe. Srsly though, is it used for its intended purpose by a significantly larger proportion of people who change it by accident? :)
17:20:52Tornei know *i* use it but i would presume that a lot o fusers have never changed it on purpose
17:20:54GodEaterno idea
17:20:55*rasher agrees that it might be worth reconsidering
17:21:19Tornethe other two things on the quickscreen are obvious playback settings that show up o nthe statusbar
17:21:42GodEaterno-one's ever come up with a good way of surveying user usage pattens
17:21:52bambuis there a way to scroll down faster through folders/ files than just holding down with "joystick" thing? it takes so long...ipod can scroll really fast or slow
17:22:20kasimirlinuxstb: Thank you! it worked with mtools
17:22:25TorneGodEater: that does indeed seem difficult to do
17:22:25*GodEater suggests RBUtil be upgraded so that it downloadeds config.cfg everytime someone upgrades, returns it to, from where it is statistically analysed.
17:22:30kasimirTorne: GodEater: thanks as well :)
17:22:37kasimiralso, one more question...
17:22:39TorneGodEater: that's a pretty good idea, i think
17:22:44Tornelike popularity-contest
17:22:46GodEaterkasimir: you're welcome
17:22:51kasimirwill a 64bit version of the new install util come out soon?
17:22:54kasimiror at least the source?
17:22:57GodEaterTorne: then at least we could see which settings get changed a lot
17:22:57kasimir(for theme support)
17:23:09GodEater64bit on what OS ? :)
17:23:26GodEaterwhen we find someone to build it for us, then it'll come out :)
17:23:33TorneGodEater: yeah, that sounds worth doing as an opt-in thing at least, if someone oculd be bothered to write the code fo rit. :)
17:23:47*GodEater adds it to Mr. Someone's todo list
17:23:48rasherkasimir: the source is available in SVN as always
17:23:55kasimirah, okay
17:24:04GodEaterkasimir: building your own is easier
17:24:08kasimirSo I guess the website just hasn't been updated with the latest tarball
17:24:15kasimirthanks, I will give that a shot :)
17:24:17GodEaterbuilding one for distribution is tougher
17:24:23kasimirthank you guys greatly for the support! you have gained a solid fan :)
17:24:30GodEaterI've failed several times to get the right magic to get a statically linked version
17:24:31kasimiryeah, that makes a good deal of sense
17:24:51GodEatergcc takes -static, and thoroughly ignores it
17:24:53GodEatervery very useful
17:25:39 Join alienrider [0] (n=4e566e20@gateway/web/cgi-irc/
17:25:45bambuwhat is the weird headphone jack thing on x5 iaudio? it has regular 1/8 inch jack and also a little flat slot thing
17:25:58GodEaterbambu: for the remote ?
17:26:20*GodEater guesses having never seen an X5 up close.
17:26:42bambuoh. maybe
17:26:53bambuwhat kinda headphones have this jack?
17:27:11 Join n1s [0] (n=n1s@rockbox/developer/n1s)
17:27:13linuxstbmt: Sorry, I've no ideas at the moment... Although looking at the main function of your codec, you should put the variable declarations at the start, before any code. I don't know if that causes a problem, but it's the Rockbox style.
17:27:45linuxstbbambu: Probably just the dedicated iaudio remote(s).
17:27:54*pixelma hands GodEater some points for guessing correctly
17:28:08*GodEater jots them down on his scorecard
17:28:09bambui have the same kinda jack on my sony mp3 cd player
17:28:10Tornedoes anyone know what use cases the CURRENT_* #defines in config-*.h are supposed to represent?
17:28:25Tornethe current values for ipodvideo seem a bit off, probably because of lcd shutdown
17:28:45pixelmaI only know of one type of remote (unlike the Irivers)
17:28:51LambdaCalculus37bambu: There's a remote for the X5 that plugs into that slot and the headphone jack.
17:31:09*GodEater suspects his config.cfg will prove disappointingly close to the default settings
17:31:40alienriderHello. On a Iriver H3xx should the folder structure /Music/2/3/4/5/6/my.mp3 cause a *Panic* Stkov tagcache error ? This is with the r21546-090628 build. Thank you.
17:32:13*Torne looks at mr someone's todo list and tries very hard not to go and attempt to build with gcc 4.4 :)
17:32:36bambuLambdaCalculus37: k thx. dont really want one of those that bad
17:33:49bambuwhat is the "usb host" jack thing for?
17:34:37GodEateralienrider: *should* it? No. *Will* it? Maybe, but if so, it's probably a bug.
17:34:44GodEateralienrider: why is your structure so deep out of interest ?
17:35:16evilnickalienrider: How many characters are in the full path? i.e. /music/albums/blah/blah/blah/01-blah.mp3 would be 43 (not sure if the / counts as a character or not)
17:35:56 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
17:35:56 Quit ocean_ (Ping timeout: 180 seconds)
17:36:51pixelmaalienrider: is the real complete path quite long as well? (as I think that there was a change to allow deeper folder structures - and that there is also a limit an overall path length - but don't know numbers)
17:36:56alienriderGodeater: I separate out codec types and separate cd extracted files from online purchaes and bootlegs. My nesting is odd but It kinda of grew that way by accident. : )
17:37:53alienriderevilnick: That directory structure will cause the error /Music/2/3/4/5/6/my.mp3 when moved this one won't cause the error /Music/2/3/4/5/my.mp3.
17:38:23evilnickalienrider: And that's a specific real-word example, using those names?
17:38:35alienriderevilnick: Yes.
17:41:51Unhelpfulpixelma: overall path length limit is 260 (MAX_PATH in firmware/include/file.h)
17:42:13evilnickUnhelpful: Isn't there a limit in FAT32 before that?
17:42:32Unhelpfulevilnick: i've no idea if there is
17:42:55alienriderevilnic: Maybe I misread your question. The example I give is a real test result I did to ttry to track down why a DB could not be built using a v3.3 build. I don't have my music laid ouy like that. The test was done like that to rule out a character length and see iff it depended on folder depth.
17:42:55Tornenot in FAT32 itself
17:43:01Tornebut Windows also has a 260 character path limit
17:43:12Tornethough windows means "256 characters plus D:\ and a trailing NUL"
17:43:36alienriderI have completed rearranged my music folder structure to fit into the above result and the DB now builds successfully.
17:43:39Tornesince we don't have drive letters we should perhaps not allow quite so many :0
17:43:40evilnickAh, I had a vague memory of 256 using Windows, so that must be what I was thinking of
17:44:08mtlinuxstb: Interestingly enough, taking those declarations out of codec_main made a significant change !
17:44:26bambuwhat is a good dpi/resolution for bmp files to view on x5 iaudio?
17:44:29Tornewe should probably only allow 258, tbh, by that logic :)
17:44:43 Join kugel [0] (n=kugel@rockbox/developer/kugel)
17:44:52GodEaterisn't there some odd limit on the number of entries in the root directory too ?
17:44:57Tornenot on FAT32
17:45:04TorneFAT12/16 have a statically sized root directory
17:45:05GodEaterwhich got halved when they introduced VFAT
17:45:09linuxstbmt: Hmm, how big is RMContext and RMPacket? The stack is only 8KB...
17:45:14Tornebut on FAT32 it's the same format as a subfolder
17:45:22Torneand it can be a lot less than half on VFAT
17:45:24linuxstbmt: So it could simply be a stack overflow...
17:45:27*GodEater hands Torne that TFAT task again
17:45:34*linuxstb packs his S
17:45:40*Torne grumbles :)
17:45:49Tornelook i just remember this stuff ok
17:46:33linuxstbAnd we'll be very grateful ;)
17:46:33 Join robin0800 [0] (
17:46:48evilnickalienrider: Just tried a file on the Sansa microSD card as: F:\1\2\3\4\5\6\7\01 - So much for love.mp3 which works fine with the database?!
17:47:11GodEaterevilnick: possible stack size difference between coldfire and arm ?
17:47:17gevaertsstack overflows are going to be arch dependent
17:47:26*GodEater thought so
17:47:35alienriderevilnick: Interesting. I have a H3xx machine and it happens everytime.
17:47:48GodEateralienrider: which is a different architecture to the sansa
17:47:49UnhelpfulGodEater: even if the stack size doesn't change, the size of the things on it will...
17:48:09*GodEater files this wisdom away somewhere he can promptly forget it
17:48:11alienriderGodeater: I understand. : )
17:48:29evilnickalienrider: Sorry I couldn't be of more help
17:48:38pixelmaworth checking on my M5? Same arch but other differences...
17:48:48alienriderevilnick: Thank you for your efforts nonetheless.
17:48:49GodEaterI only have one coldfire target, and it's at home
17:49:00GodEaternot an H3xx mind you
17:49:05GodEaterbut pretty close
17:49:37Strife89How does one actually perform timestretching?
17:50:13 Join hillshum [0] (n=chatzill@unaffiliated/hillshum)
17:51:02linuxstbStrife89: One follows the instructions in the manual... (I hope...)
17:51:12 Quit timc ("Leaving")
17:51:18 Nick hillshum is now known as Hillshum (n=chatzill@unaffiliated/hillshum)
17:51:28Tornei think he means "what is the dsp algorithM" :)
17:51:30GodEateralthough I think those instructions could do with improving
17:51:53*Strife89 looks for said instructions once more.
17:51:59Torneoh. maybe not.
17:52:14GodEaterStrife89: assuminging you've turned the setting on, it's done from the pitch screen
17:52:28Strife89GodEater: Ah.
17:52:28GodEateryou need to hit the "Mode" button a couple of times to change the pitchscreen to timestretching
17:53:04Strife89Okay, I see it.
17:53:34*GodEater wonders why the H100 "online" manual only has three sections
17:53:42pixelmaoha, get the stack overflow on my M5 too
17:55:02pixelmaGodEater: then it's broken (PDF will be an older one), if it's the H100 only
17:55:39pixelmait could have to do wth AlexP's changes but didn't follow if anything else was committed
17:55:43 Join BryanJacobs [0] (n=bryanjac@
17:55:51GodEaterI told him he was breaking the manuals!
17:55:55HillshumI've put some thought into the Clip keymap, and I'm wondering, Since LEFT/PREV is always "go back in the menu hierarchy, is having power do that as well needed?
17:55:57 Quit J-23 (Remote closed the connection)
17:56:29n1sgotta love those silently failing manual builds...
17:56:33 Join J-23 [0] (
17:57:08pixelmaweird, the PDF looks fine though
17:57:42n1spixelma: is it the right date?
17:58:38pixelmayes, and with three column button tables, the HTML breaks on the first button table. Maybe the new button table macro needs some specific definitions for HTML?
17:58:52GodEaterStrife89: got it working ?
17:59:18HillshumLambdaCalculus37: I realized my powermac is a g4 not g3
17:59:21Strife89GodEater: Yeah, it's quite neat. :)
17:59:36n1spixelma: aha, are other remote enabled manuals fine?
17:59:51pixelmaH100 is the first one
17:59:53GodEaterStrife89: :)
18:00:14rasherLooks like rbutil 1.2.2 doesn't build from a clean checkout
18:00:18rasherhad to create tools/rbspeex/build manually
18:00:24bambucan you view pdf's with rockbox?
18:00:25n1sah, i'm a bit out of the loop
18:00:38n1sbambu: no
18:00:38Hillshumbambu: no
18:00:45GodEaterrasher: bluebrother is aware of that one
18:01:42 Quit petur ("work->home")
18:02:31*rasher adds it to the tarball he's building
18:02:33LambdaCalculus37Hillshum: I still find it weird that you're having that issue with yours. I tried to set up a Rockbox dev environment again on my old PowerBook G4 and had no problems at all.
18:02:56 Quit gregzx ("ChatZilla 0.9.85 [Firefox 3.0.11/2009060215]")
18:03:02LambdaCalculus37Hillshum: What version of OS X is this with? If it's older than 10.4.X then that may be a reason why.
18:04:02HillshumMy g4 has Leapord
18:04:10bambuwhat is a good dpi for bmp to view on x5 iaudio?
18:05:34HillshumLambdaCalculus37: I also have a g3 iBook running tiger that I havn't found a good native gcc binary for yet, but I'll be putting Linux on it if I end up keeping it
18:08:10 Part kasimir
18:09:13pixelmabambu: dpi is not an important measure just the overall size in pixels - the display of the X5 is 160x128 pixels (width x hide) so that would be the largest possible for viewing in the WPS. If your album art is in JPG you can also look at it in the JPG viewer and could zoom around in it. You decide but larger files mean more data and potentially more work for the scaler etc.
18:09:26alienriderpixelma: Am I correct in believing you got the stack overflow as well on your M5 ?
18:09:41pixelmaalienrider: yes
18:10:22alienriderpixelma: Would it be advisable to raise a flyspray for this ?
18:10:26 Join bertrik [0] (
18:10:28bambuwell i have this pdf to bmp app so i can view pdfs on my x5 iaudio. the only setting for resolution is dpi
18:11:00n1salienrider: please do
18:11:17alienridern1s: I will do so.
18:11:48alienriderThank you for your time. Goodbye.
18:12:07pixelmaalienrider: probably, although I'd like to know if there is a known limit for how deep the folder structure is allowed to be (I think I remember some commits/fixes there)
18:12:12 Quit alienrider ("CGI:IRC (EOF)")
18:12:51n1spixelma: even if there is a limit a stkov is hardly the correct behaviour ;)
18:13:04 Join n00b81 [0] (n=taylor@unaffiliated/n00b81)
18:13:04pixelmatrue that :)
18:13:28 Part n00b81 ("Leaving")
18:13:38BryanJacobsbambu: the output image size is determined by the DPI
18:13:43notlisteningfunman, did you see my post on the forum?
18:13:52BryanJacobsbecause PDF is a display format designed to scale
18:14:08bambuBryanJacobs: what dpi should i set it to for x5 iaudio?
18:14:14Hillshumnotlistening: he'
18:14:21Hillshums not here right now
18:14:43BryanJacobsso the way the PDF is converted into an image is by rendering it at the DPI you specify; a larger DPI means a larger output image. You should first convert to a bitmap and then scale the resulting output down to less than 160x128
18:15:01notlisteningahh missed him again ;)
18:15:11BryanJacobsyou will probably also have to trim unless the PDF has no borders
18:15:23bambudam thats complicated
18:15:38BryanJacobsnot really... if you have imagemagick you can do it in one command
18:16:16BryanJacobsand anyhow anything that converts from a presentation format to an image format is going to have some level of complexity; PDF isn't meant for art!
18:16:55bambuso with imagemagick i can convert pdf's to bmp?
18:17:22mtlinuxstb: Ah that's it. RMPacket is quite big.
18:17:24BryanJacobsyes, although it might use ghostscript for the ps reading, I don't remember
18:17:28*mt goes fixing !
18:17:37pixelmaif you really want to watch PDF pages on your X5 I'd probably convert to JPG
18:17:39notlisteningbambu, do the pdfs have images and diagrams or just text?
18:17:50bambuimages and diagrams
18:18:15bambui have a shit ton of pdfs i'd like to have access to with my x5
18:18:39BryanJacobsis that more than a metric ton?
18:19:00bambuyes. i weighed it the other day
18:19:22pixelmaI doubt you can recognise anything on 160x128 size (and fulll screen is the limit for Rockpaint to show BMPs, if I remember correctly)
18:19:46bambuok thx for info
18:19:49*BryanJacobs agrees with pixelma - a Kindle's the thing if you want to view stuff portably
18:19:52bambuwont bother with this
18:20:08BryanJacobsor plain old paper, of course
18:20:08Hillshumkindle dx
18:20:23BryanJacobsHillshum: actually, I like the 2 better, but off topic anyhow
18:20:33bambubah im not trying to buy more things im to poe
18:20:50bambujust lug my laptop around if i wanna have my pdfs
18:21:08pixelmathat's also getting off-topic-y ;)
18:23:20bambuif i want maximum speed and efficiency, should i not have a special theme instead of the plain default for my x5 rockbox or is it not that significant?
18:24:01BryanJacobsI just tested the USB charging patch for iPod and it works with the Kindle wall-to-USB adaptor
18:24:05Hillshumbambu: the complexity of the WPS will affect battery life
18:24:21bambukk. thanks
18:25:16bambui love my rockbox w/ x5! i wish i could scroll faster like the ipod wheel does though
18:25:30 Nick fxb is now known as fxb__ (
18:30:55 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:32:44 Join Sajber^ [0] (
18:34:38 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
18:37:28AlexPpixelma, GodEater: Huh, interesting. I didn't check the html, just the pdf :(
18:37:31 Quit Strife89 ("Lunchtime!")
18:37:45AlexPI'll have a look now and see if I can work something out
18:41:20 Join {phoenix} [0] (
18:44:59 Join Linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
18:48:09kugelshouldn't mkamsboot show the OF version?
18:48:24kugeli just see "[INFO] Original firmware MD5 checksum match - Fuze"
18:48:53kugelhm, that probably changed in 1.0
18:48:57*kugel dislikes the change
18:49:38***Saving seen data "./dancer.seen"
18:50:58BryanJacobsquestion: if the iPod OF can go into a sleep mode with low battery consumption and not require a full boot when waking up, why can't rockbox?
18:51:25AlexPbecause nobody hasd implemented it
18:51:42 Quit kachna|lappy (Read error: 110 (Connection timed out))
18:51:43BryanJacobswhat does our power draw look like with the LCD off and playback stopped?
18:52:00BryanJacobsoops, got to go, asked to lunch
18:52:03Hillshum25mA is what I read somewhere
18:52:08BryanJacobsbe back later
18:55:15 Quit domonoky ("Leaving.")
18:55:51 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:56:19 Join tessarakt [0] (
19:02:09 Join stripwax [0] (
19:03:09 Quit stripwax (Client Quit)
19:04:34 Join stripwax [0] (
19:06:47 Quit stripwax (Client Quit)
19:08:03 Join FlynDice [0] (n=FlynDice@
19:08:22 Join stripwax [0] (
19:09:57 Quit FlynDice (Remote closed the connection)
19:13:47 Join stripwax5443 [0] (
19:23:56 Quit Linuxstb_ (Read error: 110 (Connection timed out))
19:25:01 Join HBK [0] (
19:25:30 Join Linuxstb|bus [0] (
19:26:44 Nick zitune is now known as zitune[afk] (n=zitune@bearstech/zitune)
19:31:10 Quit stripwax (Read error: 110 (Connection timed out))
19:34:22 Join JdGordon| [0] (n=Miranda@rockbox/developer/JdGordon)
19:42:08 Join stripwax [0] (
19:45:49 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
19:46:15notlisteningCan i use a linux based format tool to format my e200v2? Or wukk that cause issues for the files that are on there already?
19:47:12 Quit stripwax ("")
19:47:38 Join stripwax [0] (
19:47:49 Quit stripwax (Client Quit)
19:48:06 Join stripwax [0] (
19:48:47 Quit stripwax5443 (Read error: 110 (Connection timed out))
19:49:23 Join funman [0] (n=fun@rockbox/developer/funman)
19:50:18funmannotlistening: mkfs.vfat (it will obviously delete all your files), or just use the format option in the OF
19:51:36notlisteningUmm, kugel has suggested the OF format function is causing issues, maybe
19:52:03kugelI didn't
19:52:15 Quit Linuxstb|bus (Read error: 110 (Connection timed out))
19:52:24kugelI suggested it was a hardware issue as I see on my clip after using it 5 mintes
19:52:32notlisteningI've similar issues on my clip (the FS is totally broken after *very* few writes), but they're not related to rockbox (it's broken after using the OF too). So it might be as well a hardware issue as with my clip.
19:52:39 Join Febs [0] (n=chatzill@
19:52:52notlisteningah my misread there
19:53:22notlisteningmyy too) look like tool with my eyesight sorry
19:53:33LambdaCalculus37notlistening: Strange... my Clip has never had a problem with the FS.
19:53:52funmanLambdaCalculus37: he was quoting kugel, who has a broken clip
19:54:09kugelfunman: it boots at least, I just installed it, so I can't make any more statements, but I'm going to copy a bit. Though, I only have a 2GB version (w.r.t to bank switching)
19:55:06funmankugel: i have only 2GB (clip) and 4GB (fuze), but sd_select_bank() is always called once at initialisation to enable access to the whole capacity of 1st bank and not only the 1st GB
19:55:25AlexPpixelma: Well I've found the problem with the html output
19:55:43AlexPIt isn't listening to the \opt in preamble.tex even though the pdf manual is
19:55:54kugelfunman: I don't quite understand the cache align business in your commit. a) 512 is already 32byte aligned, b) we have a CACHE_ALIGN macro to not need such cryptic code
19:56:07AlexPpixelma: If I remove the opts and just use the version for three columns it is fine
19:56:22funmankugel: 512 bytes starting from 0x12345678 are not aligned
19:56:22notlisteningMine is an e200V2 but i will continue to test anything i can check funman that might give some good feedback?
19:56:32 Quit Hillshum (Read error: 110 (Connection timed out))
19:56:43AlexPpixelma: And if I put the whole btnmap define inside an opt it complains about an undefined command
19:56:45funmanCACHE_ALIGN is for allocating data in bss/data segments not on the stack
19:56:47kugelah, I see, right
19:57:33kugelCACHE_ALIGN only expands to "__attribute__((align(X))" or something, where the stuff is shouldn't matter?
19:57:56*JdGordon| interupts
19:57:57funmani am not sure stack variables can be aligned, i should check
19:58:02kugelputting things that rely on alignement on the stack also seems weird to me
19:58:04JdGordon|kugel: see my email re viewportmanager?
19:58:10funmanJdGordon|: could you comment on fs#10283 please?
19:58:16funmankugel: why?
19:58:23 Quit barrywardell (Remote closed the connection)
19:58:46amiconnfunman: Needs to be aligned at runtime -> extra code
19:58:47kugelfunman: because of a) what you said, and because b) a good deal of stack could be wasted if it actually works
19:58:47JdGordon|funman: I had a very quick look on the bus, and yeah casting to int is bad...I'll look properly tonight if i get a chance
19:58:57amiconnI wonder whether gcc even does this at all
19:59:29funmanamiconn: allocated at build time -> extra binsize
19:59:32kugelJdGordon: yes
19:59:39funmankugel: the stack is unwinded (deallocated) when the function returns
19:59:45AlexPpixelma: Now this is weird - if I do the \opt on h100 it works
19:59:56 Join flydutch [0] (
19:59:58kugelI know, but not while the function is running
20:00:26kugelI'd rather made it global
20:00:27amiconnfunman: Wrong, as it's bss
20:01:02amiconnI mean when statically allocaed. Stack allocation with alignment needs alignment code
20:01:10funmanbut the bss size would increase?
20:01:20kugelI don't think we have to worry about 512bytes binsize, especially not when we can be sure that the stuff we *rely* on works
20:01:30AlexPpixelma: It seems not to work on things defined in the platform file
20:01:36funmankugel: well the current code works
20:01:43pixelmaAlexP: maybe it's a "timing" issue? the h100 opt comes from the features.txt but the other is only in the platform file, right? Maybe the html processing doesn't "look into" the platform files...
20:01:56pixelmaquick enough :)
20:02:17AlexPI wonder why it is OK for the pdf output though
20:02:25kugelfunman: we got a bug report, how are you so sure already?
20:02:32kugelnot even amiconn is sure
20:02:33funmankugel: i mean the alignement works
20:02:47AlexPAnd at the beginning of make-html it lists the defines in platform in the "using text marked as..." spiel
20:03:23funmanamiconn discusses of the 'how' to get aligned memory efficiently, not "if it's aligned"
20:03:47kugel"I wonder whether gcc even does this at all"
20:04:10funmanalign stack variables
20:04:21funman.. with __attribute__((aligned))
20:05:37 Quit n1s (Read error: 104 (Connection reset by peer))
20:05:43 Quit mt (Read error: 113 (No route to host))
20:05:51 Join n1s [0] (n=n1s@rockbox/developer/n1s)
20:05:53kugelyou wasn't sure yourself few minutes ago..
20:06:07funman>< what do you mean?
20:06:22kugeln1s: hey, what's about the strlcpy patch?
20:06:22 Join mt [0] (n=mt@
20:06:34funmani am not sure stack variables can be aligned with __attribute__((aligned)), and this is why i didn't use it
20:06:58kugelapparently you were both not clear enough
20:07:00funmanbut a runtime alignement
20:07:02amiconnI'm not sure either. Alignment will work in bss of course - we're using this in various places
20:07:13n1skugel: right, i should commit that...
20:07:18kugelyou should!
20:07:19 Quit stripwax (Read error: 104 (Connection reset by peer))
20:07:27*n1s has no build env set up though
20:07:35kugelwhat the...?
20:07:41amiconnAlignment on stack can only be done at runtime, and I'm not sure whether gcc does this.
20:08:19funmanit doesn't seem to do it here, just silently accepts __attribute__((aligned))
20:08:49saratogawe should change the default battery current in rockbox to 1 mA so that if people forget to put it in they get a nonsense run time estimation
20:09:06saratogahaving it default to a vaguely reasonable value only encourages people to not put in the actual value
20:09:24saratogaby people i mean we developers
20:09:49n1ssaratoga: default for new ports?
20:10:39saratogai mean the #else case
20:10:55saratogathat it falls through to if no one has defined a current
20:11:35n1sso, for new ports, and yes that might be a good idea
20:11:52saratogawell new ports and some old ones too ;)
20:12:07saratogai think we just got around to putting in an entry for the c200v1 for instance
20:12:11saratogai'm still not sure all the ipods have one
20:12:56*rasher says go for it
20:14:22saratogaha the H10 doesn't even have one
20:14:30saratogaits been like 3 years
20:14:31CIA-71New commit by 03funman (r21564): Sansa AMS : use the aligned buffer in bss for all DMA transfers ...
20:16:04saratogai can fix the ones have I have numbers for, but will anyone mind if I give (different) nonsense values for the other targets?
20:16:14saratogai'm hoping this will encourage someone to fix it
20:16:22kugelfunman: that's of course a nice way too :p
20:16:47kugelthough, CACHE_ALIGN macro is still not used :)
20:17:20 Quit martian67 (Connection timed out)
20:17:26funmankugel: CACHEALIGN depends on PROC_NEEDS_CACHEALIGN which isn't defined on AMS
20:17:43AlexPThat is just bloody stupd
20:17:51AlexPpixelma: It breaks on _
20:17:57mt*ahem* .. Mono RM playback on-target ..
20:18:10AlexPkugel, funman: Sorry, I was talking about something else
20:18:12mtdefinitely yay ! :P
20:18:18funmanmt: congrats ;)
20:18:19kugelAlexP: I know, but it came good :)
20:18:29AlexPpixelma: If I make the define HAVEREMOTEKEYMAP it works...
20:18:39AlexPmt: congrats :)
20:18:48saratogamt: congrats
20:18:49AlexPkugel: hehe :)
20:18:55mtThanks everyone :)
20:18:59 Join Thundercloud [0] (
20:19:04kugelalexbobp: I believe I saw _ in opts? in the backlight setting section
20:19:08saratogais it intentional the the player and recorder v1/v2 all use the same runtime current defines?
20:19:35mtStill getting 'data abort's on calling pcmbuf_insert for stereo data though ..
20:19:39AlexPkugel: They are all over the place and work fine, *except* for in preamble.tex when used for html output
20:20:00AlexPkugel: They work there for pdf output, and they work for html output everywhere else...
20:20:18AlexPkugel: Hence my bloody stupid remark :)
20:20:27funmanmt: did you look at the ASM code at this address?
20:20:42saratogaalso, can the player record?
20:20:49AlexPhmmm, maybe not
20:20:59funmanmore important: can the recorder play?
20:21:18mtfunman : No I didn't. Will check ..
20:21:33AlexPno, that is right
20:21:51funmanmt: look at r21548
20:21:52kugelfunman: fsck reports problems, on the empty folders created by the OF though
20:22:13AlexPI was just wondering why I had no output at all then noticed I had done make veryclean after the build by mistake :)
20:22:16funmankugel: on the clip ?
20:22:47kugelno, fuze
20:22:51pixelmaAlexP: weird thing
20:23:06AlexPpixelma: I don't understand at all
20:23:15AlexPpixelma: But with no underscores it works fine
20:23:31saratogado any archos players have remotes?
20:23:47pixelmayes, the Recorders, a non-lcd one
20:24:07saratogaok so it won't use any power
20:24:18saratogado all archos players record?
20:24:35pixelmaAlexP: not the first thing I don't understand there, but I guess I'd just change it
20:24:43AlexPpixelma: I guess I'll just change it to be without the underscores
20:24:49AlexPhehe :)
20:24:49pixelmasaratoga: no, the Player and the OndioSP doesn't
20:24:51domonokyno the archos player at least can not recorder :-)
20:25:46mtfunman : Nice .. thanks for pointing this out. I'll check it later because I have to go out for a while.
20:26:05 Join efyx [0] (
20:26:07 Join petur [0] (
20:26:59AlexPpixelma: Do you think I can assume that all remotes have vol up/down skip back/forward and play/pause?
20:28:32kugelfunman: should we worry about that? I never touched the folders created by the OF
20:28:55pixelmaAlexP: I believe the Archos remote only has 4 buttons (at least that's what I drew) and there is a non-lcd remote for the Irivers but I'm not sure whether we'll explain that one. Don't know what the M:Robes remote has
20:29:06AlexPOK, I won't for now
20:29:11funmankugel: no idea, perhaps rockbox opens them when the database refreshes / file view is selected
20:29:15*bertrik wonders what to work on next for the meizus, nand and USB both look quite complicated
20:29:26kugelfunman: only read-only
20:29:26pixelmaAlexP: the non-lcd Iriver one also has less buttons
20:29:32funmanbertrik: have a rest, come back on Sansas !
20:29:36amiconnsaratoga: The Player and Recorder v1 have a remote (the same). The Recorder fm/v2 doesn't support it as-is, but it's possible to make it support the remote with a little soldering (by trading s/pdif)
20:30:06saratogaamiconn: but its a passive remote that draws no additional power for a screen right?
20:30:12AlexPpixelma: cheers
20:30:33amiconnCorrect. It draws no power from the main device at all
20:31:07amiconnIt's basically unidirectional rs-232. It has a built-in lithium battery, good for a few years
20:31:24saratogai'm going to commit this, if I've messed up the archos current defines (which i'm not even sure are correct) would you mind correcting them in their headers?
20:31:26pixelmasaratoga: do you ask this for some battery runtime estimation?
20:31:45saratogaimproved estimates are always welcome :)
20:32:00kugelfunman: let's not worry :P Release party soon? :D
20:32:08kugelgood job on finding that issue!
20:32:13saratogatime for AMS bootloaders then?
20:32:15funmankugel: well i'm not sure if that's fixed
20:32:26kugelit was an issue anyway
20:32:37amiconnAll archos current defines are correct in svn
20:32:47CIA-71New commit by 03saratoga (r21565): Move Archos battery runtime estimation defines into config files instead of powermgmt.h. Replace old default values (which were actually the Archos ...
20:32:52funmanlet's wait a bit, we have the time
20:33:39kugelfunman: I just notice it's the same what notlistening reported
20:33:48CIA-71New commit by 03alex (r21566): Fix html manuals for targets with remote keymaps. For some reason using an \opt in preamble.tex with an underscore in the name breaks html manuals ...
20:33:53saratogaanyone care to estimate the runtime furrent for the Iaudio players?
20:34:03saratogaor is it more or less the same as the H300?
20:37:02bertrikgevaerts, would it be possible the rockbox usb stack from the bootloader?
20:37:33gevaertsbertrik: yes, if the bootloader has threads and interrupts. The gigabeat S does that
20:38:02bertrikhm, I'm thinking about the meizu, it doesn't have threads yet AFAIK
20:39:06rasherdomonoky: RockboxUtilityDevelopment references a script - where can that be found?
20:39:35domonokyin the rbutilqt folder in svn.
20:40:21rasherWas it added after 1.2.2?
20:40:58pixelmasaratoga: what's with the Ondios? Shouldn't there be a change in their config files too?
20:41:17notlisteningsaratoga, my X5 used to run for about 8-10 hours
20:41:45saratogapixelma: they already define it there
20:41:59pixelmaah hmm
20:42:06notlisteningor was that not the question :P
20:42:16saratogai need the current draw in mA
20:42:46saratogayou can calculate that from runtime, but only if you have a very good idea how many mA-hours your battery holds
20:42:49amiconnYou need an estimation function if you want to do it right
20:43:07amiconnMost newer targets aren't simply basic current + backlight + disk
20:43:19amiconnThere's backlight brightness, buttonlights etc etc
20:43:26saratogayes estimation works less well on SWCODEC in general
20:43:33rasher$ ldd rbutilqt/rbutilqt |grep -ic qt
20:43:47rasherDoes this mean I managed to build an rbutil with static qt?
20:43:51 Quit jgarvey ("Leaving")
20:44:10amiconnImo this estimation function should go into target tree
20:44:36BryanJacobsrasher: some dynamic libraries pull in others when they're loaded, and there's also the dlopen() call
20:44:38saratogahow did that break the iaudios
20:44:40funmanrasher: Qt ?
20:44:45funmanoops -i
20:45:00BryanJacobsfunman: -i is "ignore case"
20:45:19 Quit n1s (Read error: 110 (Connection timed out))
20:45:47funmanrasher: if it runs, it's likely :)
20:45:47rasherI don't see anything that should pull in Qt, but it's certainly far from self-contained. Will probably work in a lot of places I guess
20:45:57BryanJacobsrasher: surest way to tell is to look at the link line; if there's -Bstatic or "libqt.a" instead of "-lqt", it should work
20:46:30saratogai changed the value of a define and that gives a a redefine error
20:46:45BryanJacobsplus you have to watch out for glibc version incompatibilities - right now it's impossible to build a truly portable executable without bundling your own dynamic libs
20:47:38*BryanJacobs has been working a lot with static linking lately and is not liking the state of the landscape... glibc people basically tell people wanting a static link to buzz off
20:48:12agaffneyalthough, you're find if you don't want to do any DNS resolution :P
20:48:13kugelglibc people tell many people for many reasons to just gtfo
20:48:20CIA-71New commit by 03saratoga (r21567): Protect CURRENT_REMOTE define with an ifndef.
20:48:25*BryanJacobs cites
20:48:36*funman points people to #rockbox-community : it's this way ->
20:49:40***Saving seen data "./dancer.seen"
20:57:35amiconnsaratoga: Hmm, does this mean CURRENT_REMOTE now gets defined as "1" for all targets without lcd remote?
20:57:55 Join PaulJam [0] (
20:59:54 Join stripwax [0] (
21:00:29saratogaamiconn: the #ifdef HAVE_REMOTE_LCD should protect against that I think
21:00:54amiconnAh, of course
21:01:23amiconnYou could combine the nested #ifdefs
21:01:55saratogayes perhaps I should have
21:01:58 Join einhirn [0] (
21:02:11saratogayou think its worth a commit
21:02:28kugelcommit count++ always is
21:02:57 Join matsl_ [0] (
21:03:30 Join n00b81 [0] (n=taylor@unaffiliated/n00b81)
21:05:36 Join faemir [0] (n=faemir@
21:08:06CIA-71New commit by 03saratoga (r21568): Clean up nested defines in powermngt.h while I'm at it.
21:10:36 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
21:11:23 Quit martian67 (SendQ exceeded)
21:12:46 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
21:13:02 Quit Thundercloud (Remote closed the connection)
21:13:59CIA-71New commit by 03kugel (r21569): FS #10377 - "PCM buffer remainder is played under too many conditions" by Jeffrey Goode. ...
21:14:23 Join linuxstb|train [0] (
21:14:50 Part n00b81 ("Leaving")
21:17:46 Join domonoky1 [0] (
21:18:12 Quit domonoky (Read error: 104 (Connection reset by peer))
21:20:42kugelfunman: I'm having a patch for mkamsboot here
21:20:57kugelto change the output a bit (I'm a bit disappointed that the version isn't shown anymore)
21:22:30saratogakugel: does Jeffrey Goode's patch do anything about the delay between when you hit stop and when audio stops playing on some targets?
21:23:06kugelwas there a delay? either way: yes :)
21:23:23 Join bluebrother [0] (n=dom@rockbox/developer/bluebrother)
21:23:44 Join martian67_ [0] (
21:24:48 Quit martian67_ (SendQ exceeded)
21:24:49saratogahmm maybe i'm imaging things
21:24:53CIA-71New commit by 03kugel (r21570): Bugfix for FS #10157 - "inaccurate seeking while paused" by Jeffrey Goode. ...
21:25:04funmankugel: nice
21:25:15saratogaeither way hurray that someones interested in fixing up playback
21:25:26funmankugel: we also need to do something for mkamsboot integration in rbutil: parse a downloaded file containing tested OF versions
21:25:29 Quit martian67 (Remote closed the connection)
21:25:41 Join martian67_ [0] (
21:25:43kugelfunman: I think I have a controversial part in the patch though :)
21:26:30notlisteningrasher, you have a minute for soem questions on the configure script?
21:26:59rashernotlistening: I'll try
21:27:04kugelsaratoga: I assume you're not talking about the fade on pause/stop delay?
21:27:10funmankugel: then better put it on flyspray so we can flame on the FS entry :)
21:27:22saratogayeah not the fade out, i may be thinking of some old bug before we had fade out
21:27:49rashersaratoga: fade out predates swcodec
21:28:00kugelI basically decided to change fw_version to give info about the v* of each player (v4 for m200, v2 for e200,c200, v1 or v2 for clip,fuze)
21:28:34funmankugel: hum.. and where is fw_version information (firmware format version 1 or 2) provided?
21:28:42saratogathen i have no idea what i'm thinking of
21:29:06kugelsaratoga: well, the pcmbuf_play_remainder indeed made it playing a little longer, but I asked on the tracker about a possible delay, and I was told it would be next to unnoticeable
21:29:38saratogamaybe i'm thinking of the delay on skipping tracks
21:31:15kugelfunman: hehe, I didn't plan that you ask that :P
21:31:29notlisteningrasher, when the configure script lists the currently available speech engines that are installed on the system where does it get last information from " I can see findtool but the actual output is not working even though i am selecting the enginne manually
21:32:23 Join tux03 [0] (n=4dc82ef7@gateway/web/cgi-irc/
21:32:45notlisteningrasher, the the putput i mean slect your engine message
21:33:12notlistening*by the output
21:33:22tux03It is possible to instal rockbox on a ipod 6 g ?
21:33:46rashertux03: the list on the front page is accurate
21:34:09rashernotlistening: Well it checks for executables in $PATH
21:34:36funmankugel: oh i see, fw_version is 2 for fuzev2 and clipv2, and 1 elsewhere, but it doesn't represent "nth hardware revision of this Sansa AMS model"
21:35:39tux03It's been ppossible in the future ?
21:35:47notlisteningrasher, yeah and that exists but i have missed something on the output of that message somewhere
21:36:02Blue_Dudesaratoga: I don't think the PCM buffer patch will do much about decreasing stop delay. The difference was only a few milliseconds at best. That patch was cleanup and potentially stops future bugs.
21:36:34AlexPtux03: Probably not
21:36:56AlexPtux03: Encrypted firmware, new undocumented hardware...
21:37:17rashernotlistening: line 529 outputs the question, the various variables are supposed to hold something like "(S)API "
21:37:19tux03(i'm french i not understand everyone)
21:37:55 Join kachna|lappy [0] (
21:38:09tux03Why ?
21:38:20LambdaCalculus37tux03: The new iPods are different on the inside.
21:38:28tux03It's not good
21:38:40LambdaCalculus37tux03: Indeed not.
21:38:55tux03I would like rockbox on my ipod !!!
21:39:02LambdaCalculus37tux03: Buy an older iPod.
21:40:35kugelfunman: right
21:40:50BryanJacobstux03: iPod 5.5G is nice
21:40:58kugelfunman: that's my patch:
21:41:05kugelI added hw_revision
21:41:32kugelalso, I changed load_of_file to take a md5sum pointer instead of all members extra
21:43:02 Quit Horscht ("Verlassend")
21:43:49kugelfunman: the output is now
21:44:08 Join Lazatar [0] (
21:44:22 Quit tux03 ("CGI:IRC (Ping timeout)")
21:44:30bertrikIIS on the meizus doesn't look so complicated
21:44:47funmankugel: would it be possible to pack model, hw_version, fw_version in the same structure?
21:44:57funmanbecause they aren't going to change for a given model
21:45:19kugelfunman: weill, they're all member of md5sum already
21:46:00kugelmd5sum is a bad name for the class
21:46:10bertrikI could start by feeding the IIS fifo with simple triangle waves and if that works, use DMA
21:46:34funmankugel: yeah perhaps ->model->model ->model->fw_version etc..
21:47:03 Quit stripwax (Read error: 110 (Connection timed out))
21:47:40funmanstruct md5sums { struct model model; char *version; char *md5 }; and struct model { int model; int fw_version; int hw_version; };
21:47:55funmana lot of "model" :/
21:47:58kugelI understand, but nested structs are ugly :(
21:49:31kugeldoes that really help actually? I mean, wouldn't we still use this array there?
21:49:57funmanwell the md5sum struct would be simpler (2 less members)
21:50:55kugelonly indirectly
21:51:24funmanyeah but the list of OFs would be simpler to read
21:51:36funmannot the actual internal storage
21:52:09funmanjust { model_e200v2, "v3.01.02" , "12345678abcdef"
21:53:02kugelI see
21:53:12 Quit martian67_ (Connection timed out)
21:53:24 Quit evilnick (Ping timeout: 180 seconds)
21:54:23funmankugel: also if you are willing to think about an implementation of downloadable tested-OFs list, you could just put all these infos in such a file and embed the file with the binary?
21:54:28notlisteningrasher, got it, I was using a "." in my executable name and that stopped it from finding it ;)
21:54:47kugelfunman: I'll have a look
21:55:14 Quit linuxstb|train (Connection timed out)
21:55:18funmanthanks :)
21:55:38rashernotlistening: that sounds strange
21:55:46JdGordon|funman: struct md5sums { struct model model; char *version; char *md5 }; and struct model { int model; int fw_version; int hw_version; } <- for the love of all that is holy please dont do that!
21:56:07JdGordon|nested structs are fine, but dont reuse variable names like that
21:56:12kugelfunman: I'm not sure if it's mkamsboot's job to download that
21:56:22funmanJdGordon|: yeah i suggested model_name
21:56:40funmankugel: not download but parse
21:56:50BryanJacobsI have read and it does not give me confidence. Were I to install a 240GB drive in my iPod, would I be able to use that much space with Rockbox 3.3?
21:57:37funmanthe file would be provided by load_of_file() or another function, called by rbutil; or called by main() with a pointer to the embedded OF list for standalone mkamsboot
21:58:59kugelwhat's the function rbutil will call anyway?
21:59:14notlisteningrasher, give is a go in your script rename espeak espeak.linux and it will fail, or it was for me
21:59:20 Join Thundercloud [0] (
21:59:21 Quit LambdaCalculus37 ()
22:00:00domonoky1kugel: rbutil will just do the same thing as the main() does.. plus it downloads the bootloader and maybe md5sums.
22:01:44kugelso the integration boils down to renaming main and compiling it as a lib?
22:02:12 Join funman_ [0] (n=fun@rockbox/developer/funman)
22:02:29 Join Hillshum [0] (
22:02:45funman_kugel: this is already done
22:03:14funman_what we need is a way to provide an up to date tested OFs list even with an older rbutil build
22:03:23kugelI think the struct thing can be done later
22:03:34funman_yes it's only cosmetics
22:03:43 Quit robin0800 (Remote closed the connection)
22:03:56funman_if someone works on a file format it could just go away
22:03:57funmanBryanJacobs: yes, provided you build your own rockbox (with the 3.3 source if you want)
22:03:57 Quit funman ("Reconnecting")
22:04:05kugelthat's mostly cosmetic, but I do think the output should be somewhat more verbose (printing the OF version also gives a better feeling about the success)
22:04:09funman_(no, i don't lag)
22:04:33*Hillshum has put some thought into the Clip keymap
22:04:44 Nick Lazatar is now known as lol3izer (
22:04:48funman_kugel: well feel free to improve it (i was not talking about your patch)
22:04:48 Nick lol3izer is now known as martian67 (
22:04:55bluebrotherit might be helpful to check how the iriver bootloader patcher does the md5sum stuff. It also uses the md5sums to identify known firmware version
22:04:59 Join lilltiger [0] (n=lilltige@
22:04:59funman_Hillshum: feel free to improve it as well (with a patch..)
22:05:23HillshumDoes power need to be "go back" when left/prev do that as well?
22:05:36funman_bluebrother: it has a built in list
22:05:40bertrikI think the clip keymap is quite OK already (at least for general use, don't know about the plugins)
22:05:53 Nick funman_ is now known as funman (n=fun@rockbox/developer/funman)
22:05:57LloreanHillshum: It shouldn't really be a unique function, since it's hard to use normally
22:06:04bluebrotherfunman: I know :)
22:06:06bertrikthe function of the HOME button is a bit fuzzy though
22:06:22lilltigerHow do i clear the disk catche/force my sansaAMS to rebuild it?
22:06:44AlexPIf you mean dircache then it is rebuolt on boot
22:06:46 Quit {phoenix} (Read error: 104 (Connection reset by peer))
22:06:57AlexPIf you mean database, chose initialise now
22:06:58HillshumLlorean: ?
22:07:14 Quit Blue_Dude ("ChatZilla 0.9.85 [Firefox 3.0.11/2009060215]")
22:07:19AlexPIf you want to be sure before, delete *.tcd from .rockbox (database again)
22:07:37LloreanHillshum: It shouldn't really do something else.
22:08:21Hillshumhow about back to main menu?
22:08:49bertrikI would like the home button to switch between WPS and main menu, and between radio and menu menu if radio is playing
22:08:56kugelbluebrother: we want a downloadable list
22:09:22lilltigerAlexP: hmm ok, then the issue isent the dircache. When i Use the filebrowser within rockbox the filnames are boxes and other strange characters(unreadable) and thoes songs arent added to the database. But in the org. sansa firmware the songs looks all correct
22:09:27kugelthey're still in production, so we expect firmware updates (at least for clip and fuze)
22:09:38kugelI know, we're not familiar with that yet :)
22:11:03Lloreanlilltiger: Remember last night, I asked you to type whole, real words. That also applies to things like "org" as compared to "original." The idea is to make machine translation as reliable as possible by using things as they'd show up in common dictionaries.
22:11:18LloreanPlease, try.
22:12:13 Quit BryanJacobs ("Java user signed off")
22:12:21*kugel still no since3.2 link grml
22:12:48rasherkugel: 3.3?
22:13:14kugelthat's too fresh :)
22:14:07Hillshumfunman: What files do I need to tweak?
22:14:24 Join BeChris [0] (i=5c8976c2@gateway/web/freenode/x-f2fc935b68a4cf0a)
22:14:29kugelfunman, bertrik, saratoga, domonoky1: I'll put up the mkamsboot patch to fs so you can test it on your devices, it shouldn't break anything
22:15:00BeChrisStill trying to figure out how to profile my png plugin.
22:15:08BeChrisSomeone has done that recently ?
22:15:37funmanHillshum: apps/keymaps/keymap-clip.c
22:15:54 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
22:16:05Hillshumfunman: thanks
22:16:10funmanBeChris: did you look at the profile option in tools/configure?
22:16:27saratogaBeChris: the easiest way is to just comment out bits of code and time how long decode takes without them
22:16:49BeChrisfunman : yes of course but when I select it it adds -finstrument-functions to gcc and all my plugins freeze at startup
22:17:28BeChriseven if I don't use profile_thread() / profstop()
22:18:22funmansaratoga: didn't you use profiling for codecs?
22:18:23lilltigerLlorean: insted of being unfriendly you could help insted, behaving like you is against the freenode guidlines actually and thoes takes present above the rockbox guidelines. So if you want others to follow the guidelines, you should start fallowing them yourself first.
22:18:44BeChrissaratoga : I though profile might come in plugin to isolate exactly bottlenecks in my code. Adding manually time measurement is quite broing I must admit.
22:18:45saratogafunman: a little, but other people figured it out for me, and mostly i use stripwax's tests
22:18:51kugelI think we just need to have lines like "model_number$version$md5sum" ($ being a suitable seperator), then malloc a struct md5sum for each line and fill it
22:19:01saratogaBeChris: surely you have a pretty good idea what parts are slow though?
22:19:08funmanlilltiger: i don't know what "org" means, instead you can use "OF" which is a documented abbreviation
22:19:39Lloreanlilltiger: Go and complain to freenode then that we politely ask you to follow our channel guidelines. But until such time as freenode tells us they need changed, they are what they are.
22:19:47BeChrissaratoga : almost sure but wanted to have confirmation by a tool
22:19:47funmanlilltiger: don't expect anyone here owes you any answer, being unhelpful is not against the freenode guidelines in any way
22:20:18lilltigerfunman: no but being an elitist is
22:20:29kugellilltiger: We also have guidelines, and those include to use proper English
22:20:30LloreanThe fact that I do not know the answer to your question does not preclude me the ability to remind you this channel is logged and has guidelines in place because of that logging.
22:20:46funmankugel: yeah looks nice
22:21:31kugelI'm fearing my search&replace mixed up fw_revision with hw_revision somwhere
22:22:14Mikachugiven those two names only, i would maybe call them firmware_rev and hardware_rev instead :)
22:22:31 Quit bluebrother ("leaving")
22:23:03lilltigerkugel: yes but being eletist about them and refuse to help people cos they dont use words that you personaly find incorrect is against the freenode policy
22:23:28Lloreanlilltiger: "cos" was a word I mentioned to you less than a day ago.
22:23:44lilltigerLlorean: yes and it's an common used word
22:23:46saratogalilltiger: if you dislike the rules, feel free to get out
22:23:53kugellilltiger: "cos" is one of the words we don't like
22:23:58Lloreanlilltiger: Refusing to help people because they're disorderly in your channel is allowed by freenode rules. They even instituted a mute ban for dealing with such people.
22:24:15Lloreanlilltiger: I explained to you the broad category of acceptable words. You know why 'cos' is not one of them.
22:24:51AlexPlilltiger: I didn't actually see anyone refusing to help you?
22:24:54rasherlilltiger: It's not about "personally finding words incorrect". The fact is that you use words that are not understood by non-native speakers and screen readers. Which is against the rules. Please refrain from doing so.
22:24:59kugellilltiger: We're an international channel, we have people here that have no plans about which words are common somewhere, they probably have only google or something to translate, and using colloquial language makes it hard for those to understand what we're saying
22:25:19lilltigerLlorean: But how am i supposed to know wich words are okej to use and wich are not, i do not have an masters degreee in english, nor am i a native english speaker.
22:25:35saratogawe have a link explaining which are ok
22:25:35funmanlilltiger: just avoid abbreviations and it will be ok
22:25:36Lloreanlilltiger: Don't use slang or abbreviations.
22:25:42LloreanIt's fairly simple.
22:25:47LloreanI told you this yesterday.
22:25:59rasherlilltiger: It's also a good bet that words we *specifically tell you* are not okay, are not okay.
22:26:14rasherSuch as "cos". Do not use that again. Seriously.
22:26:15saratogacan we resume on topic dicussion now?
22:26:17lilltigerLlorean: just like "org." in this case, I did not know that it isent an correct english short form. it's commonly used
22:26:35rashersaratoga: please go on
22:26:42Bagderso was there a question? ;-)
22:26:42Hillshumcommonly used where?
22:26:50Lloreanlilltiger: Don't use short forms then. Type "original." It's a few more characters, but it's much safer.
22:26:53lilltigerfunman: I learnt most my english from the internet, how should I know wich are slangs and not.
22:26:54AlexPBagder: Indeed there was :)
22:27:02AlexPBagder: Back in the mists of time
22:27:10*Bagder tries to fight us back to the topic
22:27:20lilltigerLlorean: but I did not know the correct spelling of it, and you did understand what i ment.
22:27:27saratogawhy fight we have a mute function
22:27:36saratogawhich i've been dying to try
22:27:44AlexPFor all that want to see the question: "When i Use the filebrowser within rockbox the filnames are boxes and other strange characters(unreadable) and thoes songs arent added to the database. But in the org. sansa firmware the songs looks all correct"
22:27:51rasherlilltiger: You've been explained this a million times over. Stop making the same point time and again.
22:27:55 Join stripwax [0] (
22:28:00Lloreanlilltiger: Then look up the correct spelling. You're on the internet. Your justifications do not matter.
22:28:38rasherlilltiger: As an answer to your question, you have filesystem/filenames corruption.
22:29:09rasherlilltiger: Possibly you're using a weird way of mounting the drive, setting the wrong character set.
22:29:10kugelfunman: see, I already found where I mixed it up
22:29:38lilltigerrasher: Wouldent that show up when I connect the device to the computer?
22:30:04rasherlilltiger: Not if it's mounted in the same weird way.
22:30:19lilltigerrasher: so it might be wrong character encoding?
22:30:24 Join evilnick [0] (i=0c140464@gateway/web/freenode/x-464d89026b6c1ead)
22:30:25rasherIt almost certainly is
22:30:58rasherWhat do you connect it to?
22:31:49lilltigeran linux box, i connect it when the device is of so it uses the sansa firmware
22:32:20rasherAnd what are the mount options?
22:33:09funmankugel: where?
22:33:38lilltigerrasher: (rw,nosuid,nodev,uhelper=hal,uid=1000,utf8,shortname=mixed)
22:33:45kugelline ~429 with the patch (already updated on the tracker)
22:34:32kugeldomonoky1, funman: if we're using that ini format, it should be rbutils job to parse the list, then just pass an a struct md5sum vektor to mkamsboot
22:34:34*rasher shrugs
22:34:53domonoky1kugel: yes, that was my idea with the ini list.
22:35:23funmankugel: then we must froze the struct md5sum
22:35:24domonoky1so we just need a function to set the md5sum list in mkamsboot. (and if it isnt set, it uses the builtin list, for standalone).
22:35:59domonoky1funman: you dont really need to freeze it, you just need to change it in two places :-)
22:36:00Lloreanrasher: Could his problem instead be the codepage set in Rockbox? If that somehow got set / changed wrong?
22:36:08kugelfunman: no, we can just check for NULL, and use the built-in in that case (main() will pass NULL, rbutil not)
22:36:08funmancan we be sure the struct is the same when built by gcc and g++ ?
22:36:22rasherLlorean: Maybe.
22:36:24funmankugel: i mean we must use the same binary format between mkamsboot and rbutil
22:36:29 Quit matsl_ (Read error: 60 (Operation timed out))
22:36:37rasherlilltiger: What's the codepage option in Rockbox set to? (I don't remember the location)
22:37:02domonoky1funman: should be no problem.
22:37:09kugeland, doesn't that happen automatically (mkamsboot is statically linked into rbutil)
22:37:22 Join martian67_ [0] (
22:38:02funmankugel: the offsets of each member in the struct must be the same for obvious reasons
22:38:13lilltigerrasher: latin1 it seems
22:38:17 Quit martian67 (Connection timed out)
22:38:20funmanand it's not automatic if the mkamsboot library and rbutil aren't build by the same compiler
22:38:34rasherlilltiger: Maybe try setting it to utf8.. No idea if that's related.
22:38:50kugelmkamsboot should be buildable with g++ and very few changes
22:39:24 Quit stripwax ("")
22:39:36kugelalso, we're thinking too much of rbutil again, we don't need it for release :)
22:39:37funmanwell we buidl the library with gcc, perhaps we should use the packed attribute to be sure of the members offsets?
22:39:44domonoky1funman: the only potential problem is different packing. so declare the struct as packed,and it should be never a problem.
22:40:13lilltigerrasher: I just tried, dident help. And it's just some songs that gets wrong file names while the folders are correctly named.
22:40:50rasherAnd you're using the file browser?
22:40:59rasherNot the Database.
22:41:01Hillshum"dident" isn't really a word either...
22:41:08lilltigerYes, the database dosent find thoes files at all.
22:41:24rasherTry running fsck. Your filesystem sounds troubled
22:42:15AlexPHillshum: No need to pick up on typos
22:42:22kugelfunman: so let's agree on leaving all work for the rbutil guys! They want it that way with the ini list :P
22:42:42funmannot really, i'm the one wanting that
22:42:50funmanto support OFs tested after a rbutil release
22:43:08funmanbut yeah it's not a requirement, but a MrSomeone job :)
22:43:10kugelbut rbutil isn't critical
22:43:26funmandomonoky1 already wrote a patch anyway
22:43:45kugelfor the ini thing?
22:43:56funmannope, just rbutil support with the current built in list
22:44:00domonoky1kugel: rbutil isnt critical, but practically ready. it just needs the function to set the md5sum list in mkamsboot.
22:44:13kugelI know, we talked about it :)
22:44:20funmandomonoky1: well it could be committed as-is
22:44:34funmanwe should check mkamsboot returns a proper error message if the OF was untested (I think it does)
22:44:38domonoky1funman: not really, we still need the released bootloaders :-)
22:45:17 Join FOAD_ [0] (
22:46:01 Join matsl [0] (
22:47:13rasherdomonoky1: I have a shellscript to create a rbutil source tarball. Where should I put it?
22:47:27domonoky1into svn ?
22:48:12rasherAny preferred location? It's completely self-contained anyway, so it could be pretty much anywhere (does a full checkout and shaves off everything unneeded)
22:48:40Bagdertools/release perhaps?
22:48:49 Join robin0800 [0] (
22:48:57Bagderor perhaps in the rbutil tree
22:49:00domonoky1i would just put it into the rbutilqt folder. Maybe one day we can combine all those release scripts for rbutil :-)
22:49:03 Quit flydutch ("/* empty */")
22:49:14 Join funman_ [0] (n=fun@rockbox/developer/funman)
22:49:15kugelwell, you need to checkout to get the tool which does the checkout?
22:49:42***Saving seen data "./dancer.seen"
22:49:46Bagderyou _can_ download it from the svnweb
22:49:55 Quit Hillshum ("ChatZilla 0.9.85 [Firefox 3.0.5/2008120121]")
22:50:01kugelit still sounds funny :)
22:50:03funman_domonoky1: so we only need to release bootloaders / or do you have something left for rbutil integration ?
22:50:34rasherI think it belongs in tools/release
22:51:10domonoky1the other rbutil release scripts are under rbutil/rbutilqt ( and )
22:51:30CIA-71New commit by 03rasher (r21571): A script to create a source tarball of rbutil. Does a complete checkout and removes everything that isn't needed.
22:51:38rasherdomonoky1: Maybe they should be moved :)
22:53:17domonoky1funman: not much, maybe the resulting file naming in rbutil needs to be improved. (at moment they have one name per AMS target, but there are different endings in AMS OF names). Maybe it should use the input OF name.
22:53:48domonoky1rasher: if you move them, make sure they still work :-)
22:55:13domonoky1building rbutil is also one thing which could be improved. *Automated rbutil builds would be sooo nice* :-)
22:55:47kugelI'd propose that we push the release of e200v2 and fuze sort of in 2 weeks if we don't notice a serious bug
22:56:06gevaertskugel: "release"?
22:56:33AlexPmake supported I guess
22:56:52Mode"#rockbox +o funman_ " by ChanServ (ChanServ@services.)
22:56:59AlexPAnd then include in next release
22:57:04Kick(#rockbox funman :no more tab completion problems) by funman_!n=fun@rockbox/developer/funman
22:57:09Mode"#rockbox -o funman_ " by ChanServ (ChanServ@services.)
22:58:01kugelI can't remember how we exactly called the new group (it's not listed on the wiki page too), but of course the "lower" status
22:58:14AlexPeither way, not release :)
22:58:28gevaertsdid we come up with a name?
22:58:31funman_what is the difference with 'released' models?
22:58:42AlexPThere is a Rockbox release
22:58:49AlexPSo next one will be 3.4
22:59:01saratogathe idea being to wait and see if theres any more disk bugs?
22:59:13AlexPIf the AMS are ready they can be in it, but they can't be "released" before that :)
22:59:30 Quit ender` (" If God had intended us to go around naked, He would have made us that way.")
22:59:41kugelreleasing a target != having the first release for a target, IMO :)
22:59:54Lloreankugel: We should try to only use the word "release" to refer to releases
23:00:14domonoky1we should search for better names for our target groupings :-)
23:00:18LloreanOtherwise people reading the logs and reposting them elsewhere (which happens a lot) will get things wrong and it's frustrating to correct later.
23:00:26kugelalright, I'm fine with it
23:00:35kugellet's foo the ams sansas soon!
23:00:38LloreanThe two categories are what again?
23:00:45LloreanI know the "lower" one is "Can be used in day to day playback"
23:00:55*domonoky1 suggest. "the lower" "the higher" "the released ones" :-)
23:01:01 Quit FOAD (Read error: 110 (Connection timed out))
23:01:01 Nick FOAD_ is now known as FOAD (
23:01:12LloreanWhat's the requirement for the upper one - RBUtil, Manual, and "no feature of Rockbox is known not to be working" or what?
23:01:38bertrikI liked how gstreamer came up with categories "the good" "the bad" and "the ugly"
23:01:48domonoky1dont know the exact requirements, but something like this. ie the full Rockbox feeling.
23:02:01kugelI'd say rbutil, manual and all apps/ code (plugins etc), I don't think we really need recording and usb for a real release
23:02:31Lloreankugel: If they're going to be officially supported anyway, why not have the upper one have high standards?
23:02:34 Quit funman_ (Read error: 60 (Operation timed out))
23:03:30kugelI don't know, isn't upping it in the "supported hierarchy" just a ideological thing?
23:04:29domonoky1we should make a contest for this names :-)
23:04:31kugelalso, some hardware features might be next to impossible to implement, due to undocumented hardware or so
23:04:41rasherMaybe "if usb and recording doesn't work after a long time, accept that it might never, and move it up anyway"
23:04:43LloreanIt seems to me if we're going to split them, the top one being "very, very complete" makes sense since people looking can then say "Well, I just want one I can listen to music on" or "I want one that doesn't mean I need the OF for anything"
23:04:54Lloreanrasher: Why do we need to move it up?
23:04:58saratogaofficially supported means we're going to hand out builds, and that should happen as soon as we can get away with it so that we get more feedback and attract more attention to the build
23:05:07LloreanWe can have "Targets that are supported" and then "Targets where you never need to boot the OF"
23:05:30 Nick fxb__ is now known as fxb (
23:05:34Lloreankugel: That doesn't explain why things can't simply stay in the lower "supported" category forever
23:05:36*rasher shrugs
23:05:39saratogai don't think we need to make it that complicated
23:05:41rasherSeems a bit too picky to me
23:05:43domonoky1yes, the lowest bar should be substancially lower the now... AMS would be long ago already in the lowest category.
23:05:47saratogawe have the very nice device chart that explains what works
23:05:49LloreanI thought the whole point of the lower category was so that we could have things that *might* never elevate still be supported
23:05:51kugelthat seems like a good distinction. Effectively, the category doesn't matter anyway
23:05:59LloreanSaying now "but we should keep moving them up" means we didn't need the lower category in teh first place.
23:06:20 Join funman [0] (n=fun@rockbox/developer/funman)
23:06:23kugelsaratoga: Llorean proposal is very uncomplicated I think
23:06:33domonoky1We can even introduce a new layer above "released", like the "golden" ones :-)
23:06:39saratogai think we should just let whoever is working on a target agree when they want to make it supported
23:06:48kugeland the platin :)
23:07:11kugeland if we start dropping hwcodec (or at least charcell), those come into the hall of fame :p
23:07:23domonoky1so low: runs, not rick of bricking, installation is possible. released: its working fine, but few things may be missing: Gold: everything.
23:07:46kugelwe don't need more than 2 categories
23:07:53Lloreandomonoky1: There's only two.
23:07:57Llorean"Supported" and "Gold"
23:08:19kugelWe have "Rockbox runs", the only difference you can make is "You might need to of for some features" and "You can completely replace the OF without losing any features"
23:08:21gevaertsdomonoky1: how many "gold" targets would we have in your system?
23:08:24domonoky1why not also released in between ?
23:08:50LloreanSupported - you can use Rockbox on it day to day, install without high risk of brick. Gold - Nothing the OF does can't be done here (though some examples, like format support, don't count - we treat MPEG2 as equal to h.264, since they're both "plays video")
23:09:16Lloreangevaerts: iPods, SansaV1, H100 series, Archoses, maybe H10?
23:09:22*gevaerts thinks that "supported but not released" should mean a very incomplete manual, or there are real annoyances
23:09:37krazykitnot the gigabeat F/X?
23:09:41*domonoky1 wouldnt put the ipods to gold.. there is still some charging issue ?
23:09:41LloreanWell, "Release" as in release-numbered, like 3.4, is a different category
23:09:44Lloreankrazykit: Ah, those too
23:09:56Lloreandomonoky1: Well, mentally I consider that fixed. I'm confident it will be done by 3.4. Sorry
23:10:03kugelLlorean: most ipods don't have USB in the releases
23:10:14gevaertsLlorean: domonoky1 said "everything", so ipod and gigabeat would be disqualified
23:10:22Lloreangevaerts: Gigabeat would?
23:10:28gevaertsLlorean: the OF has USB host
23:10:29krazykitno usb-host
23:10:32domonoky1yes, i would put them to "released" for now :-)
23:10:38kugelwe just don't degrade once it has it's status
23:10:45Llorean"Release" is a third category, but not one we really distinguish on the front page.
23:10:58LloreanIt's only distinguished by whether 3.3 (or in 3 months, 3.4) downloads are available for it.
23:11:04gevaertsThat's why I don't see the need for that extra-gold status. There really aren't that many targets in it...
23:11:06kugelthey were before this decision. We changed the laws after them, hence they can't be touched by them
23:11:13LloreanThat will have the requirements of Manual and RBUtil support, but not necessarily "Gold" status
23:11:18domonoky1why not 3 categorys ? The name is up for discusson ofcourse..
23:11:33saratogafunman, kugel: for whats its worth current SVN does not even boot for me on the fuze
23:11:44funmansaratoga: stuck on rockbox logo?
23:11:54saratogadivide by zero
23:12:02Lloreangevaerts: One goal of Rockbox has been to be a "replacement" firmware. Being able to be a replacement should be a goal, and marking which ones actually achieve this is worthwhile to users I think.
23:12:11funmansaratoga: can you send me your rockbox.elf ?
23:12:17kugeldid you reinstall bootloader and reformatted?
23:12:27Lloreangevaerts: I think there's not significant benefit to the upper category if targets in it can perform vastly differently
23:12:30rasherDo we have any but the archoses and H120?
23:12:36funmankugel: no need to format since there is a checksum
23:12:39saratogai compiled the bootloader just now using up to date svn, but i did not reformat
23:12:46Lloreangevaerts: The "lower" category should be defined by minimum usable requirements, the upper one by "maximum" reasonable goals.
23:12:54saratogasomeone on ABI reported the same error and address after reforting, so i assume its not the problem
23:12:55kugelfunman: huh?
23:13:12funmankugel: rockbox.sansa is checked to be valid after the dma transfer
23:13:30Lloreangevaerts: Maybe instead of "all OF features supported", "All Rockbox features the hardware is capable of are supported"?
23:13:34kugelthe checksum is simple, and rockbox.sansa isn't big, it might not cover fs issues
23:13:38gevaertsLlorean: I guess I'm mainly afraid of overdoing this
23:14:20Lloreangevaerts: To me, as long as we make the lower category readily and clearly available, the upper one can be moderately strict. Basically there for the people who are _really_ nervous.
23:14:25 Join fml [0] (n=4fd3c6e9@gateway/web/cgi-irc/
23:14:31gevaertsLlorean: that would make targets move out of it again. e.g. now h300 would qualify, but if someone implements USB host on the beast, it would drop out
23:14:33kugellet's just have "gold", and "wood", where gold can replace the OF completely
23:14:43kugelwe do not need to make more differences
23:14:45domonoky1we can also leave the "gold" status for later, and only hve two. one very low, and one at the hight of current "released" status.
23:14:54kugelwe tread all the same anyway
23:14:57funmansaratoga: or check what code is at this address
23:15:17Lloreangevaerts: That's true.
23:15:22*rasher doesn't see the need for more than "release-worthy"
23:15:28kugelit's, if at all, a user indication. it's just that "wood" can differ vastly, so either we make more categories, or we don't
23:15:29LloreanSupported and Released are certainly enough for now.
23:15:37LloreanWell, "Supported and Releasable"
23:15:47fmlHello. When saving theme, recording, sound etc. settings, the produced file is always named "config" (probably, followed by a number). Wouldn't it be better to have a "specific" name base, e.g. "theme" for theme settings?
23:15:51domonoky1oki, but the names for those two could be improved ?
23:16:04kugelsupported and released are fine
23:16:09amiconnOn which target(s) apart from the archoses does rockbox replace the OF *completely*?
23:16:17*amiconn thinks there isn't one
23:16:19domonoky1i think the names are very confusing.
23:16:21rasheramiconn: doesn't h120 manage?
23:16:42kugelsansa e200/c200
23:16:51amiconnHmm, probably, although still not officially
23:16:52rasherkugel: accessories
23:16:52kugelprobably ipod nano too, and the iriviers
23:16:54Lloreanamiconn: "Replacing" in functionality, not overwriting.
23:17:12rasherkugel: usb host
23:17:12funmankugel: but the OF plays other video codecs?
23:17:15fmlIn the code, the cases are diffrentiated, the folder is set depending on what is being saved. We could also give more specific names (as default, the user can still edit them)
23:17:24kugelfunman: codecs don't count
23:17:29rasherfml: seems reasonable I think
23:17:30amiconnkugel: h300: lack of usb host support in rockbox. And all except the archoses and h1x0 cannot be flashed with rockbox
23:17:31 Quit Thundercloud (Read error: 104 (Connection reset by peer))
23:17:36gevaertsfunman: the nano OF doesn't do video
23:17:51kugelbut as I said, I don't think the new categories apply to the existing targets. You can't be sued for laws that came after the crime
23:17:52Lloreangevaerts: The e200/c200 do MJPEG though
23:17:55fmlrasher: I'll provide a patch
23:18:00amiconnSo we're always relying on the OF loader in flash -> not a complete replacement
23:18:09gevaertsLlorean: I'd argue that mpeg2 is better :)
23:18:21LloreanBut I do agree "Supported" and "Release targets" (not "released" targets) are good enough for now.
23:18:23saratogafunman: sorry someone came in and talked to me:
23:18:26kugelamiconn: on which target?
23:18:30Llorean"Release" indicating we will release for them going forward.
23:18:38amiconnkugel: ?
23:18:41kugelIn fact, I completely replaced the OF on my e200
23:19:03amiconnSo the rockbox bootloader is in flash?
23:19:37 Join |niko| [0] (
23:19:57kugel./sansapatcher -a. or something like that
23:20:04rasheramiconn: that's not the meaning of "replacement" I think is useful
23:20:05funmansaratoga: in battery_status_update(), so I think it's your fault ;)
23:20:17|niko|I am trying to install Rockbox with the Qt Rockbox utility on Debian and it fails installing 3.3
23:20:18amiconnsansapatcher writes to the fw partition, not to the flash rom. It can't
23:20:33|niko|I have Rockbox already on there, and everything else installs fine, the bootloader, fonts, themes, etc.
23:20:40rasheramiconn: rather "you never have to boot the OF, because Rockbox does all it does (and more)"
23:20:43funman|niko|: how does it fail?
23:20:52|niko|"Extracting failed: unable to create or open file..."
23:20:57|niko|funman: That error ^
23:21:02saratogacurrent = CURRENT_NORMAL*100 / 122;
23:21:07saratogafixed point for the win
23:21:08kugelamiconn: before that there's only the PP bootloader. I don't have the recovery mode for example
23:21:18kugel(that's how I understood it)
23:21:22 Join dmb [0] (n=dmb@unaffiliated/dmb)
23:21:23|niko|Which confuses the heck out of me because everything else installs fine.
23:21:45funman|niko|: double check you have write permission to the mount point and .rockbox/ if it exists already
23:21:48amiconnThe PP bootloader is what I mean.
23:22:06kugelthat one has nothing to do with the actual OF?
23:22:35amiconnYou can't replace it with sansapatcher alone. We'd need flashing support in rockbox, i.e. a plugin that flashes our own loader into the flash rom
23:22:48domonoky1|niko|: and this error happens on which install function ?
23:22:54kugelyes, but how's that not replacing the OF?
23:23:13amiconnThe bootloader is part of the OF as I see it
23:23:44kugelthe PP bootloader comes from PP, the OF from sandisk, I replaced all sandisk bits. That means to me I replaced the OF
23:24:03amiconnOn the iaudios we don't have dual boot, yet we don't run 100% replacement code as we rely on the OF loader for starting us, for flashing, and for emergency disk access
23:24:09*rasher fails to see where this is going
23:24:22gevaertsamiconn: on the gigabeat F you can flash the bootloader. It's just not in svn yet
23:24:30amiconnOn the archoses in turn we do run our own loader (if it's a flashable archos)
23:24:37saratogasomeone remind karl to commit that
23:24:38saratogait works fine
23:25:00kugelI don't consider the PP loader as part of the OF
23:25:07saratogafunman: Unhandled IRQ 00: INT_WATCHDOG
23:25:12saratogastill my bug :)
23:25:19kugelthat's like the BIOS being part of windows
23:25:23LloreanWhat does whether or not you can physically replace the OF matter?
23:25:31amiconnOk, then the gigabeat F is the 8th target where we have a 100% replacement (after the 6 archoses and the h1x0)
23:25:41LloreanI think I'm the one who brought up "replacing the OF" and I was just referring to featureset.
23:25:42Bagderit doesn't matter _at all_
23:25:50funmansaratoga: yeah still this int 0 is weird, I think it means an interrupt occurs without source defined in the VIC register
23:25:57amiconnLlorean: We're still *relying* on OF parts for rockbox to work, on most targets
23:26:02rasheramiconn: the h1x0 as released doesn't run from flash.
23:26:19amiconnrasher: Yes, and so doesn't the gigabeat f
23:26:20Unhelpfulwhat exactly would need to be done to get the S to that point? i'm assuming "find out where the flash loader lives", at least? make sure we know how to do any init that it does?
23:26:25Lloreanamiconn: Yes, but why does that matter? What topic depends on that?
23:26:39*kugel fails to understand amiconn reasoning to not consider the e200 and c200 being able to completely replace the of
23:26:40*rasher is also unsure where amiconn is going with this
23:26:53Bagderkugel: I would assume he means the flashed bootloader
23:26:53kugelbut, for the rest of the evening, I stick to agreeing ti Bagder :)
23:26:57saratogafunman: don't we disable the watch dog anyway?
23:27:09amiconnkugel: The PP bootloader is part of the OF imo, even if it is provided by PP and not by sandisk
23:27:15kugelsaratoga: we actually use it for reboot
23:27:17amiconnIn fact the OF
23:27:25funmansaratoga: yes, the code just doesn't handle this situation and assumes interrupt source is 0 (watchdog)
23:27:32saratogaoh ok
23:27:38saratogawell i rebooted and its working fine now
23:27:43amiconn's usually contain several components which were made by other companies than the OF itself
23:30:45funmani still do'nt know how to read volume/rec/hold buttons on the c200v2
23:31:01kugelhave you looked at dbop?
23:31:15funmanyes only 5 bits are read, the 5 we know already about
23:31:32kugelso much?
23:31:51kugelnot any on gpio?
23:32:02 Quit bmbl ("Bye!")
23:32:04funmannot as far as i can see
23:32:06 Quit Febs (Read error: 104 (Connection reset by peer))
23:32:07kugelI thought RockRabbit found already like 4-5 buttons
23:32:11kugelon GPIO
23:32:16funmanexcept power which we know already
23:32:18 Join FlynDice [0] (n=FlynDice@
23:32:25funmanyes he read them on gpioc, but reading dbop is more reliable
23:32:50*kugel can't believe that
23:32:58kugeldid you try delays etc?
23:33:05kugelor just changing the order?
23:33:29kugelthat's how we got it to work on the fuze/e200v2 :p
23:33:42funmanthe OF only read 5 bits from dbop
23:33:50funmanso i won't experiment further
23:34:04kugela delay after clearing the bits and setting the afsel, then have a proper (read: random, but better working) order
23:34:27funmanwhen reading over dbop no need for clearign/delays
23:34:40kugelI'm talking about gpio
23:34:53kugeloh, and delays are needed
23:35:05kugelremember the backlight issues on the fuze?
23:35:21funmani see no GPIO pins changing in the I/O debug screen
23:35:26 Join perrikwp [0] (
23:35:42funmanyes, but we copied the delay present in the OF while the c200v2 OF has none
23:36:10kugelbut the OF also doesn't run at 62MHz
23:36:33funmansure, but button reading works fine so no need for tweaking
23:37:04FlynDicefunman: kugel: did either of ou try the last core voltage patch or did you just read the code? I'm getting a white lcd screen on boot caused by the ascodec_read
23:37:16kugelyou can't see gpio's changing in the debug menu if you don't change the afsel
23:37:17funmanFlynDice: only read
23:37:27funmankugel: hum ok
23:37:35kugeland clear/set depending on what value the button gives
23:38:09|niko|funman, domonoky1: The installer still fails even though I run it as root
23:38:39|niko|It detects the ipod's mount point and can install the boot loader, fonts, doom stuff, themes, etc.
23:38:56|niko|but rockbox itself it cannot seem to ungrade
23:39:16|niko|I tried chmod 777 .rockbox -R
23:39:23kugelFlynDice: No, I didn't yet
23:39:32kugelbut I should do!
23:39:55kugel|niko|: what does dmesg say?
23:40:04domonoky1|niko|: strange. But you can always just download and unzip the new rockbox version manually.
23:40:07funmanwhere is rbutil debug log?
23:40:07CIA-71New commit by 03saratoga (r21572): 1 mA gets rounded to 0 during computation, so use 2 mA as the default.
23:40:25|niko|[64844.288650] FAT: Invalid FSINFO signature: 0xc161d252, 0xe141f272 (sector = 1)
23:40:26domonoky1funman: unfortunatly still in flyspray.
23:40:38kugelsaratoga: was that the divide by 0 thing?
23:41:14funmankugel: i see no changes on gpio pins :/
23:41:15domonoky1|niko|: try checking the disk. maybe there are filesytem errors ?
23:41:25 Quit einhirn (Read error: 104 (Connection reset by peer))
23:42:10kugelfunman: do you change all afsels on GPIOB with some delay between resetting? That's basically how fuze manages to read them
23:42:16kugelon GPIOB and C
23:42:26funmankugel: yes (int x = 1000; while(x−−) ;)
23:42:56funmanalso I don't see the OF read something else than DBOP or GPIOA_PIN3
23:43:25kugelI'd guess you can find the rest on GPIOB
23:43:42|niko|kugel, domonoky1, there is a dmesg output indicating.,..?
23:44:09domonoky1indicating errors, but i am not sure what errors :-)
23:44:10kugelyes, many errors
23:44:26kugelthat probably also means that your dap is remounted in read-only mode
23:44:43funman|niko|: check the cable / connexion
23:44:44kugells -l in the dap's root should show you
23:44:58 Quit aaron424 (Remote closed the connection)
23:45:05funman[63539.334923] lost page write due to I/O error on sdd2
23:45:13|niko|It is working fine
23:45:15funmanprobably means you unplugged it without unmounting it first?
23:45:42 Join safetydan [0] (n=deverton@rockbox/developer/safetydan)
23:45:42 Quit fml ("CGI:IRC 0.5.9 (2006/06/06)")
23:45:54|niko|as a USB mass storage device it is working fine thus far
23:48:05kugelwhat else should it be?
23:48:24kugelare you trying to update in mtp or something?
23:51:08kugelFlynDice: I don't quite understand why you removed the CPUFREQ_MAX check
23:52:18FlynDiceTrying to simplify, since it looks like a max frequency < 200MHz is a quickly dying idea it seems...
23:52:21 Quit robin0800 (Remote closed the connection)
23:52:42kugelI can't follow that reasoning
23:52:56kugelI think we are prepated for changing frequencies as we like?
23:53:19saratogasetting the max frequency less then 200MHz would make sense but i guess we'd need a way to set it higher again for things like mpegplayer
23:53:37saratogawe've talked about a more complicated boost system before with respect to the beast
23:55:21funmanFlynDice: i think the check can stay, at least it gives useful information (only needed for > 200MHz freqs)
23:55:54FlynDiceThe way it was set up was that if CPUFREQ_MAX was less than 200 MHz then the voltage was _statically_ lowered to 1.05 ie no switching
23:56:37saratogaso we think the file system issues on the fuze are fixed?
23:56:54FlynDiceand if it was > 200 MHz the voltage was raised & lowered in tandem with boosting
23:57:29funmansaratoga: no, we hope ^^
23:57:48funmanFlynDice: still getting a white screen at boot ?
23:58:03saratogawell i built the database and played around with my fuze a good while and couldn't get anything to go wrong, aside from that one interrupt problem on the first boot
23:58:25FlynDiceworking on it right now, I need to add a ascodec_write to tell the adc which input to read...
23:58:58*kugel thinks we finally need scrollwheel support in doom

Previous day | Next day