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

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

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

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

#rockbox log for 2007-03-19

00:00:22Nico_P...something I know everything about :D
00:00:22Bagderalthough I'm not sure the benefit is that great
00:00:32XavierGrso in order for the while playing screen to show up this "wps compiler" wouldn't have to run?
00:00:47Nico_PBagder: have you looked at my patch ?
00:00:50Bagderits simple, if the wps already are compiled the byte-code is loaded
00:00:57BagderNico_P: nope
00:01:11XavierGrso each wps must have been precompiled to show up?
00:01:14bluebrotherbut it adds a dependency to an external bytecode compiler
00:01:22Bagderyes, or it would first need to be compiled
00:01:26XavierGrcan't say I like the idea
00:01:28Bagderto be able to "run"
00:01:37XavierGrnot KISS imho
00:01:42bluebrotherok, this could be done with a viewer
00:01:44Moossame here
00:01:49bluebrotherlike a theme installer ...
00:02:01*bluebrother remembers his theme installer patch once more ;-)
00:02:07Mooscurrently we even create a wpas on the fly and use it immediatly
00:02:35***Saving seen data "./dancer.seen"
00:02:37XavierGrmoos: well with the wps compiler you can still do it, the compiler would be a plugin
00:02:54XavierGrbut I think that compiling wps themes sounds weird
00:02:56Bagderwell, personally I'm for a concept that allows "wps-plugins" that runs instead of the normal wps
00:03:03Moosmore laborious that just open text editor
00:03:08 Join sneakums [0] (
00:03:12bluebrotherit's at least better than this "compiled skins" idea
00:03:16LloreanBagder: An alternate interpreter that can allow extended tags, basically?
00:03:34Juice^we should have a WPS editor, that has some pre-made templates for sizes etc.
00:03:40bluebrotherbut moving functionality out and loading it in a plugin way sounds nice
00:04:01bluebrotherJuice^, not before viewports are implemented ;-)
00:04:07LloreanJuice^: Anyone's free to make one.
00:04:11BagderLlorean: no, even more rough: let the plugin do whatever it wants and offer enough functions and info in the plugin API
00:04:20XavierGrplugins are nice to relieve core from binary size from extras, but not for core features
00:04:23 Join Rob2222 [0] (
00:04:25Juice^bluebrother: viewports?? woot :)
00:04:26Moosdidn't we have a WPS creator already?
00:04:49bluebrotherI think there was some VB thingy somewhere in the forums
00:05:01bluebrotherdon't know what happened to that though
00:05:05Moosthat what I remenber too
00:05:21*amiconn has slight problems regarding the player lcd and target tree
00:05:22Bagdercode comes and goes...
00:05:22Nico_PA WPS compiler would allow us to use packaged WPS...
00:05:33bluebrotherIMO the basic wps functionality should be in the core
00:05:43Bagderbluebrother: I agree
00:05:48bluebrotherbut stuff like replaygain tags etc could be external
00:05:48amiconnThe lcd charset is definitely target dependent, but cannot reside in target tree because the sim needs it...
00:06:09bluebrotherthe wps should keep at least conditional images and sublines
00:06:12Nico_Pthe RTC tag should be external !!
00:06:45bluebrotherbtw, do we have a tag for disk activity?
00:06:58Mooswe have
00:07:04bluebrotherit's quite annoying not to see disk activity on the mini as most themes turn of the status bar
00:07:25pixelmacalled "virtual led" %lh - (I believe)
00:07:32bluebrotherbut I guess I need to design my own for the mini ;-)
00:07:41Juice^bluebrother: you could easily modify the theme to include the status bar?
00:07:54bluebrotherJuice^, not if the theme includes graphics
00:08:09Juice^bluebrother: okey
00:08:12bluebrotheras image positions are absolute and would overlap with the status bar
00:08:20bluebrotherwhile the text itself moves down
00:08:23XavierGrpixlema: I did an ondio bb with 900mah batteries, I got 7 hours 40 minutes from it, is that normal?
00:09:02amiconnSounds normal
00:09:17pixelmasounds about right - I get (got :/ ) about the same with my 900mAh rechargeables
00:09:45amiconnSomewhat less than what I got with 900mAh NiMH in a test (9h 3 min)
00:09:59pixelma(just the rechargeables aren't ok anymore)
00:09:59Juice^here, if someone want/need a battery indicator for a wps, its layered in photoshop;
00:10:02XavierGrnice to know that my ondio isn't affected by excessive power drain
00:10:45bluebrotherJuice^, my browser can't display photoshop ...
00:10:49 Quit aliask ("Chatzilla 0.9.77 [Firefox]")
00:11:11Juice^bluebrother: you need to download it .. dont think many browsers can open psd files :p
00:11:48bluebrotherbtw, I had a nice idea for backdrops ... I added the rockbox logo from zezayer in light grey as backdrop to my mini
00:11:57Juice^bluebrother: this is the look:
00:12:25bluebrotherJuice^, thanks
00:13:03bluebrotherbut I mainly was just curious how it looks ;-)
00:13:17Juice^Think i will make others too..
00:13:20*bluebrother prefers gimp
00:13:30Juice^pixel by pixel :)
00:13:53 Join courtc_ [0] (
00:14:32 Quit courtc (Read error: 54 (Connection reset by peer))
00:14:53BagderI've added the web server as a build server again now
00:15:31 Quit muesli__ (Read error: 110 (Connection timed out))
00:17:24 Join midgey [0] (
00:19:41linuxstbamiconn: What's your problem with the lcd charset code? It seems fine where it is now - if it's needed in the sim then it's higher-level than target-tree, so shouldn't go there.
00:20:44amiconnI am separating lcd specific code from generic charcell code and target code
00:21:17amiconn...while rewriting the driver to provide real unicode support
00:21:29Juice^i found the wiki page for this VB app for editing WPS code..
00:22:05pearldiverJuice^ nice battery icon :P
00:22:13pearldiverdo you have any other icons?
00:22:29Juice^pearldiver: thanks.. ehm not at the moment, i will do other tomorrow
00:22:39Juice^im going to sleep now
00:22:40bluebrothermonday approaches ... time for sleep
00:22:42 Quit courtc_ (Read error: 54 (Connection reset by peer))
00:22:46Juice^yeah.. nighty
00:22:46 Quit bluebrother ("...zzzZZZ")
00:22:47MoosJuice: it's probably out of date with some missing tags now
00:22:57pearldiverJuice^ ah night then
00:22:57Juice^Moos: I could guess
00:23:04 Quit Juice^ ("nighty - work to do")
00:23:22 Quit DataGhost (Nick collision from services.)
00:24:58 Join efyx [0] (
00:25:55 Join safetydan [0] (i=cbca159f@rockbox/developer/safetydan)
00:27:00 Join courtc [0] (
00:27:54 Join lukaswayne9 [0] (
00:29:13 Join Shaid` [0] (
00:30:33 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
00:30:42amiconnQuite a number of chars available on the new lcd isn't mapped currently...
00:30:51 Quit Arathis ("Bye, bye")
00:34:11 Join Thundercloud_ [0] (n=thunderc@
00:35:53 Quit JETC- ("Konversation terminated!")
00:36:25 Join Vyrus001 [0] (
00:37:24 Quit pearldiver ("some games are better left unplayed")
00:37:24 Quit My_Sic (Read error: 104 (Connection reset by peer))
00:41:18 Join webguest94 [0] (i=415fb535@gateway/web/cgi-irc/
00:41:36 Join Shaidd [0] (
00:41:46 Quit Shaid (Nick collision from services.)
00:41:52 Nick Shaidd is now known as Shaid (
00:41:57webguest94can i install rockbox on a 5G ver 1.2.1 30gb ipod?
00:42:00 Quit Thundercloud (Read error: 104 (Connection reset by peer))
00:42:43webguest94with the rockbox installer X?
00:43:05scorchethat is not our software, and as such, we will not support it
00:43:08linuxstbNo - that installer is very out of date now.
00:43:23linuxstbThe official install instructions are very simple.
00:43:34webguest94really? i heard they arn't
00:43:42webguest94maybe i will try it
00:44:08scorchethey used to be much "harder", however, it has gotten easier since then
00:44:11linuxstbThey were hard up to a couple of months ago, but have been greatly simplified now.
00:44:26ShaidI didn't find it that hard really...
00:44:46webguest94where can i find how to install
00:44:54scorcheour manual
00:45:01webguest94ok thxs alot
00:45:08 Join Ribs [0] (n=ribs@
00:45:09 Join mirak [0] (
00:46:04 Join SirFunk [0] (
00:47:16webguest94rockbox install X the new one is not out of date
00:47:33webguest94or RIX
00:47:42scorchewe still will not support it....honestly, our instructions are quite simple
00:47:54webguest94just saying
00:48:41scorchefrom that changelog, they are still using an out-of-date bootloader
00:49:24 Quit Shaid` (Read error: 110 (Connection timed out))
00:50:24linuxstbwebguest94: There is also an official installer in development, but you could use it (it installs the most up-to-date version of Rockbox) - (the download links are at the bottom of that page)
00:50:24webguest94oh ok
00:50:45webguest94that looks cool
00:50:58 Quit Moos ("Quitte")
00:53:48 Quit ender` (" Be wary of strong drink. It can make you shoot at tax collectors and miss. -- Robert A. Heinlein")
00:54:38 Quit ompaul ("lights out")
00:54:48 Join bagawk [0] (n=lee@unaffiliated/bagawk)
00:56:31 Join ment [0] (
00:56:46 Quit robin0800 (" robin0800 quit")
00:59:09 Quit obo ("bye")
00:59:18 Quit Ribs ("eh eh ehhhh!")
01:00:46lImbushi ment
01:00:49Soapthe core running bytecode WPSs, wouldn't that allow people to create WPSs with (for example) album art, multiple fonts, etc - and distribute those w/o a need to have modified builds?
01:01:06SoapAs soon as I asked that I realize it wouldn't.
01:01:14 Join clyde [0] (
01:02:06safetydanNice idea though
01:02:31safetydanProbably a wee bit complicated to turn the WPS in to a full "Rockbox" scripting language.
01:02:35 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
01:03:49 Join robin0800 [0] (
01:06:43amiconnsafetydan: Wanna test further speex asm constraint simplifications for arm?
01:07:13amiconnBinary code shouldn't be different, and it compiles cleanly, but I don't have any speex files to test with
01:08:23webguest94that "RBU" thing is great!!!
01:09:41webguest94and the new main menu is so much easier to work with
01:09:46webguest94i love rockbox!!!
01:09:52linuxstbwebguest94: It worked for you then?
01:10:17 Join lee-qid_ [0] (
01:10:49 Join miepchen^schlaf [0] (
01:13:43 Join webguest72 [0] (i=18f44eb5@gateway/web/cgi-irc/
01:14:12webguest72anyone know anything about shell scripts?
01:14:29webguest72I'm trying to make one that will compile rockbox for my target
01:14:37webguest72my question is...
01:15:24webguest72I run the configure file from the script but want the choices to be put in automaticly
01:15:37webguest72which commands do I put in?
01:15:55webguest72into my shell
01:16:00webguest72so I have like
01:16:35lImbusI guess you have to pipe in the answers
01:16:44lImbusbut I'm not shellmaster neither
01:16:46webguest72which works fine but when I run the script it still asks for stuff
01:16:49Nico_Psafetydan: have you seen my WPS tokenizer patch ?
01:17:23lImbuswebguest72, easiest is to modify your configure script to not ask for this stuff but simply do the right thing
01:17:38webguest72lImbus: thats right but how do I do that what command
01:17:47Nico_Pwebguest72: echo "11" | ../tools/configure
01:17:48 Quit Shaid (Read error: 110 (Connection timed out))
01:17:59Nico_Pthat will select the H320
01:18:10webguest94sorry to bug you guys again but, does the iTunes database converter work? or is it out of date
01:18:28safetydanamiconn, I'm not near a compiler unfortunately
01:18:32lImbuswow. I got rbutil to crash
01:18:53safetydanNico_P, saw the patch but I haven't looked at it yet. Have this whole "work" thing to do :)
01:19:00 Join DataGhost [0] (
01:19:12safetydan(well, not near a C compiler anyway)
01:19:16webguest72so I would do echo "15" ../tools/configure
01:19:24webguest72for my target
01:19:32Nico_Psafetydan: it's based off your code, so you might find it interesting
01:19:47linuxstb`webguest72: Just do "(echo 20 ; echo N) | ../tools/configure)"
01:19:48Nico_Pwebguest72: you're forgetting the pipe
01:19:49webguest72would I do echo "15 n" for a normal build?
01:19:59lImbusUnhandled exception at 0x00436ae0 in rbutil.exe: 0xC0000005: Access violation reading location 0x0152f1a4. While trying to "install rbutil on my device"
01:20:21webguest72Ok I will give it a try!
01:21:14linuxstbwebguest94: I think it's out of date - but you shouldn't need it, Rockbox can build the database itself from the tags in your audio files.
01:21:16 Join Shaid [0] (
01:21:35 Quit entheh ("^~")
01:21:38 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon) do i get my itunes songs on rockbox?
01:22:29linuxstblImbus: I'm not sure if that feature has been tested recently - try and catch Domonoky next time he's around, he's Mr RbUtil at the moment.
01:22:59linuxstbwebguest94: What happens when you select Database in the main menu?
01:24:32lImbusIll get the code, compile it and chase it down
01:24:48linuxstblImbus: Even better :)
01:24:50webguest94i don't know yet
01:25:02webguest94i just installed it
01:25:12webguest72linuxstb: "(echo 20 ; echo N) | ../tools/configure)" what is that symbol in the middle? the long line?
01:25:31nlswebguest94: wait till the disk stops spinning and reboot your player
01:25:45webguest94but do you need the music to be in the apple Firmware
01:26:02lImbuswebguest72, thats the pipe symbol
01:26:15webguest72how do you do it?
01:26:21lImbusyou reach it usually with an AltGr combination
01:26:31lImbuson which keyboard layout are you ?
01:26:43webguest72standard american
01:26:46linuxstbwebguest94: No, Rockbox will play anything you copy onto your ipod's disk.
01:26:49webguest72or w/e it's called
01:27:05 Quit lee-qid (Read error: 110 (Connection timed out))
01:27:18Nico_Pwebguest72: you can also copy-paste it
01:27:55lImbustry AltGr and the key left of z (if any)
01:28:00Nico_Pwebguest72: and the right command is actually : (echo 15 ; echo N) | ../tools/configure
01:28:22Nico_Pi just tried rbutil for the first time and it's awesome !
01:28:48 Quit bospaadje (Remote closed the connection)
01:28:55lImbusNico_P, yes, it is
01:28:59linuxstbNico_P: It is. We should try and get it up to release quality and release it.
01:29:17 Join bospaadje [0] (
01:29:26mentIt's possible to estabilish usb connection with computer during audio playback? For example for sending control commands (next-track, prev-track)?
01:29:32webguest72ok I just saved the symbol
01:29:37Nico_POne thing I find annoying is the giant progressbar when downloading a theme preview
01:29:38webguest72in a file
01:29:44lImbusment, no way
01:29:49 Quit webguest72 ("CGI:IRC")
01:29:49 Quit bospaadje (Remote closed the connection)
01:30:01 Join pearldiver [0] (
01:30:09lImbususb takes control over it all. The last possible hack is to continue fm radio while usb
01:30:12linuxstbment: In theory, yes (on some targets at least), but Rockbox doesn't have any USB drivers yet.
01:30:15 Join bospaadje [0] (
01:30:34mentlinuxstb: which ones?
01:30:58 Quit bospaadje (Remote closed the connection)
01:31:16mentor are there any other usable interfaces? (like some sort of ir or serial)
01:31:24linuxstbThe PortalPlayer targets (software USB) and I _think_ the targets with USBOTG - H300, X5, Gigabeat.
01:31:30 Quit clyde ("Leaving")
01:31:32linuxstbWhich player do you have?
01:31:55 Quit efyx ("Quitte")
01:33:08mentactually no one, but i'm looking for some capable of this
01:34:16 Join bospaadje [0] (
01:34:24linuxstbLots of targets have remote controls which I'm sure you could interface to if you're a hardware hacker.
01:34:37lImbuslinuxstb, don't think h300 would do it.
01:35:08lImbussince atm, we are already doing usb-mode on our own whitout having implemented a usb-stack
01:35:11 Quit lee-qid_ ("aufwiederbyebientotsayonara")
01:35:15lImbusits usb-ata-bridge
01:35:17*amiconn found a many years old bug in the player lcd character mapping
01:35:41linuxstblImbus: Yes, the usb-ata bridge can't be used for it, but what about the USBOTG port?
01:36:40linuxstbment: But I'm curious, why do you want to control your mp3 player from a computer, instead of just playing music directly on your computer?
01:36:57 Quit secleinteer (Read error: 104 (Connection reset by peer))
01:37:34lImbususb-on-the-go means to me that the h300 is able to work with an external usb mass storage device. not AS a device to another usb host.
01:37:40mentlinuxstb: one reason is that my notebook has broken sound card
01:38:20lImbusthen get a archos and do remote control via serial
01:38:45lImbusor a usb sound card :)
01:38:45mentlinuxstb: the other one is that it would be nice to have cheap synth with external control made from mp3 player
01:38:59 Join bagawk_ [0] (n=lee@unaffiliated/bagawk)
01:39:03 Quit atsea- (Read error: 104 (Connection reset by peer))
01:39:05 Quit bospaadje (Remote closed the connection)
01:39:16 Join bospaadje [0] (
01:39:46 Quit robin0800 (" HydraIRC -> <- 100,000+ downloads can't be wrong")
01:40:18 Quit bagawk (Nick collision from services.)
01:40:30 Nick bagawk_ is now known as bagawk (n=lee@unaffiliated/bagawk)
01:41:02 Join zeth [0] (n=zeth@sblug/lugmaster/zeth)
01:41:14 Quit nicktastique (".")
01:41:20 Join robin0800 [0] (
01:42:13JdGordonNico_P: hey, just looking at your wps code... looks nice
01:42:26JdGordonbut is
01:42:26JdGordon+#if CONFIG_RTC
01:42:26JdGordon+ case WPS_TOKEN_RTC:
01:42:26DBUGEnqueued KICK JdGordon
01:42:26JdGordon+ return NULL; //rtc_string(buf, buf_size, data->tokens[i].value.s);
01:43:01zethSorry if this is spammy, but I just posted a little article/diary of my move to rockbox on my blog, if anyone reads it and finds that I have made serious errors then let me know cheers
01:43:11Nico_PJdGordon: I got some stack smashing crashes on the sansa whith the RTC tag enabled
01:43:38Nico_Pit worked on the H300 sim, though
01:43:55 Quit robin0800 (Client Quit)
01:44:03Nico_Pbut for now i just disabled it to focus on more important issues
01:45:03JdGordonI see you have a fwe others commented out also... so its not quite ready then?
01:45:03 Join robin0800 [0] (
01:45:10Nico_PJdGordon: no
01:45:16Nico_Pnot quite
01:45:25Nico_Pwhich ones are the others ?
01:45:48JdGordonin the table in wps.h
01:46:19Nico_Pah yes. those are handled by wps_data_preload_tags
01:46:27Nico_Pbut I wanted to keep them available
01:46:34Nico_Pin case i decided to change it
01:46:44JdGordonah ok :)
01:47:34Nico_PFor easier reading you should patch it in... almost all the interesting parts are in wps.*
01:48:10 Quit Bawitdaba (Read error: 104 (Connection reset by peer))
01:48:37JdGordonits faily easy to read the patch, seen as its all one big blob of +'s :)
01:49:00 Join efyx [0] (
01:49:39 Quit robin0800 (Client Quit)
01:49:39JdGordonits unlikly that it would be allowed.. but would it be difficult to turn the parser into a rock? might save a fiar amount of space on the recorders....
01:49:58Nico_Pthere was talk about this a bit earlier
01:50:07Nico_Pit could be a good idea
01:50:08 Join robin0800 [0] (
01:50:16JdGordonon the ml i saw.. here aswell?
01:51:24Nico_PXavierGr started it at 23:56, ie a little less than 2 hours ago
01:51:58LloreanJdGordon: It's not _that_ unlikely that it would be allowed.
01:52:26 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
01:52:31 Join miepchen^schlaf [0] (
01:52:32Nico_PIt wouls shave a fair amount off the binsize
01:52:40JdGordonwell, I aslways thought everyone was against moving core parts to plugins...
01:52:46linuxstbMy understanding of the conversation was that it was unlikely...
01:53:17Nico_Pwith my code, if the plugin could acces the WPS data, all the parsing could be in a rock
01:53:54Nico_Pafter that, the WPS data struct contains all the necessary info... i could even get rid of the original source
01:53:54 Join z0de [0] (
01:53:58 Join Bawitdaba [0] (
01:54:16JdGordondump that strcut to a file so it doesnt need to be reparsed every boot.. just read in
01:54:43Nico_PJdGordon: that was the idea yes... byte code... I'm not sure how to do it yes
01:54:51 Quit webguest94 ("CGI:IRC (EOF)")
01:54:52Nico_Ps/yes/though... i'm very tired
01:55:22Nico_Phow is a struct dumped to a file ?
01:55:37JdGordonopen the file.. then write(fd, source, len)
01:55:50JdGordonand make sure the return == len
01:55:59Nico_Pjust write the memory to the file ?
01:56:13 Quit robin0800 (" HydraIRC -> <- IRC with a difference")
01:56:16JdGordonthen read it back with read(fd, dest, len)
01:56:20JdGordonvery easy
01:56:36JdGordonunless you have stuff like pointers which wont work...
01:56:47 Join robin0800 [0] (
01:57:08Nico_Phmm.. i do but i could recompute them
01:57:49JdGordonand it would probably need a header with the usual version and target id or youll have problems with endiendness if ppl swap these dumps
01:57:59safetydanI always thought the parser could be a plugin like the properties plugin. When you select a WPS, the plugin gets called, which parses the file into the memory format the core needs.
01:58:33JdGordonIt could (should even on certain targets), but if that rck isnt there we have problems...
01:58:56 Quit Bawitdaba ()
01:59:20JdGordonthe bytecode for the default wps needs to be hardcoded...
01:59:30safetydanNo worse than it is now.
01:59:31JdGordon... into the core that is
01:59:32amiconnBasic wps functionality should always be in the core
01:59:48pixelmaI wouldn't like to sacrifice main features for bin size even though I'd like to have rombox
02:00:06JdGordonyou wouldnt be.. thats the thing :)
02:00:09Nico_Ppixelma: the features would be identical
02:00:25JdGordonthe penalty would be slightly slower wps parsing, which would only be done once tho
02:00:34Nico_Pamiconn: is WPS parsing basic WPS functionality according to you ?
02:00:54amiconnLoading a wps shouldn't require a plugin
02:00:56Nico_Pbasic enough to need to be in the core though
02:01:27amiconnRendering a wps might be different - later, with very special wps features
02:02:40***Saving seen data "./dancer.seen"
02:02:41 Quit robin0800 (" HydraIRC -> <- IRC with a difference")
02:03:01Nico_Pamiconn: what about compiling the WPS to bytecode to avoid having to parse it again ?
02:03:04Soapwhat becomes the definition of "basic"? Text only?
02:03:07JdGordonNico_P: could we have a text only parser in the core (so basic features only) and then the full blown one in a rock? or would that not give enough bin saving?
02:03:21 Join Thundercloud [0] (n=thunderc@
02:03:22amiconnSounds like overkill to me
02:03:45Nico_PJdGordon: I agree with amiconn on this
02:03:46 Join robin0800 [0] (
02:03:59CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
02:03:59*JdGordon thinks so too :p
02:04:06amiconnThat was related to compiling
02:06:05Nico_Pwell by compiling i just mean parsing and then saving the struct to a file
02:06:25Nico_Pit would avoid having to parse each time
02:06:59safetydan"each time" should only be once per reboot and once per WPS load. Not sure it's worth optimising that for speed.
02:07:38JdGordonwell yeah, it would be pointless unless the parser was in a rock...
02:07:46Nico_Ptrue. that's what I had in mind while coding anyway
02:07:49linuxstbAnd then you have the caching issues to deal with - i.e. bugs caused by not detecting changes in the WPS correctly.
02:08:09*JdGordon gone
02:08:25safetydanI can't see that putting the parser in a plugin would be that much slower. You've already spun the disk to load the WPS file, so loading the .rock isn't going to slow things down.
02:10:04Nico_Pamiconn: your objection about a plugin is that it wouldn't be possible to have a WPS with only the core file anymore ?
02:10:19 Quit robin0800 (" HydraIRC -> <- State of the art IRC")
02:14:58Nico_Psorry :) In fact I don't really understand your objection against having the parser as a rock
02:14:58amiconnProblems if wps/parser would be a plugin: you couldn't run another plugin while playing
02:14:58amiconnThat also and especially goes for tsr plugins
02:14:58 Join webguest29 [0] (i=415fb535@gateway/web/cgi-irc/
02:14:58Nico_Pamiconn: i'm only speaking of the parser. the rendering would stay in the core
02:14:58webguest29hi (again :p)
02:14:58Nico_PTSR plugins... right
02:14:58pixelmaand as JdGordon already mentioned you rely on a plugin - if it's missing you couldn't load a new wps
02:14:58webguest29about the database, whats it for?
02:14:58safetydanamiconn, I'm talking about a plugin that's run once to parse the file into the WPS structs.
02:14:58amiconnAnd wps is really core functionality imho, and more important functionality than esoterics like .cue support
02:14:58DBUGEnqueued KICK Nico_P
02:14:58Nico_Ppixelma: I could say the same about a codec...
02:14:58safetydanpixelma, that's a problem, yes. But how often does that happen?
02:14:58pixelmadon't know about codecs :P
02:14:58DBUGEnqueued KICK amiconn
02:14:58amiconnIt happens every boot at least
02:14:58 Join robin0800 [0] (
02:14:58amiconnCompiling a wps is a weird idea
02:14:58***Alert Mode level 1
02:14:58amiconnHow long does parsing take? 10ms? 100ms? you wouldn't notice...
02:14:58safetydanI think we're talking across each other. I'm not talking about compiling the WPS.
02:14:58***Alert Mode level 2
02:14:58amiconnIt's done once on load
02:15:06Nico_Pamiconn: about TSR... the WPS would be parsed at boot time (no problem with that). then if the user runs a tsr plugin and then tries to load a new WPS, we'll tell him he has to choose... it's not a very likely case anyway
02:15:30amiconnAnd how would it load the wps if the plugin is missing?
02:15:38linuxstbI don't think compiling is the main suggestion - it's moving the parsing code (once there is a tokenising parser) into a plugin.
02:15:46Nico_Pamiconn: i'm not talking about compiling anymore... the rock would write the WPS data to the memory and exit
02:15:59safetydanwhat they said :)
02:16:13amiconnlinuxstb: I understand that. Imho both of these ideas are weird
02:16:14linuxstbAnd if plugins/codecs are missing, Rockbox isn't much use... Although hwcodec is different (at the moment)
02:16:23pixelmasorry, can't follow
02:16:42Nico_Pamiconn: if the plugin is missing, we load a basic hardcoded WPS
02:16:52Nico_Pbut it's the same as if the codec were missing
02:16:54 Quit jhulst (Read error: 104 (Connection reset by peer))
02:16:56CtcpIgnored 2 channel CTCP requests in 0 seconds at the last flood
02:16:56*linuxstb is against moving code to plugins, but can't think of a logical argument against this suggestion...
02:17:08webguest29what does the database thing on the mainmenu do?
02:17:24Lloreanwebguest29: Have you looked in the manual?
02:17:29Nico_PI agree with linuxstb. it wasn't my idea but I don't see anything angainst it
02:17:34safetydanlinuxstb, well it's been done once already with the properties plugin which is where I go the idea.
02:17:45webguest29i don't really understand still
02:17:59linuxstbsafetydan: But that's not core functionality - it's just a special viewer plugin.
02:18:03Lloreanwebguest29: What part of what's in the manual don't you understand?
02:18:35Lloreanlinuxstb: The question could be "Is a _customizable_ WPS necessarily something that should be considered core"?
02:18:50safetydanlinuxstb, true. But it's a possible direction for making things more "modular". I have no strong feelings either way.
02:18:50amiconnsafetydan: Unlike wps, file properties aren't core functionality
02:19:01webguest29oh nvm
02:19:07webguest29i understand now
02:19:12LloreanThe core could always have a hardcoded, functional but nongraphical WPS that works no matter what, with customizability dependent upon a plugin.
02:19:37webguest29i thought mine stoped working because the disk kept spinning
02:20:11linuxstbLlorean: You could say the same about the settings, or even the database/file browsers...
02:20:17JdGordonLlorean: no, thats a different discussion....
02:20:25Soapwould offloading "eyecandy" such as graphical WPSs to plugins allow a "soft forking" of HW and SWCODEC targets?
02:21:00JdGordonthe core would still draw the wps so every token would be handled by the core... the thing thats being moved to a rock is just the parsing code to save bin space
02:21:03amiconnhwcodec does have graphical wps
02:21:30pixelmagah... now that forking idea again - thought it should be unified more
02:21:33amiconnAnd I am firmly against a fork (and I guess at least linuxstb thinks the same)
02:21:58Lloreanlinuxstb: Well it's just a point along what's already being said. WPS is definitely core. Graphical/customizable WPSes, not necessarily. So why not put the minimum amount of code in the core to allow customizable WPSes, and the parseing and anything else in the plugin?
02:22:02linuxstbYes, I've always been against a fork (even before the av300)
02:22:43Soapthat's why I said "soft fork" - let the 16bit screens have their eyecandy - but let them have it partitioned off outside core.
02:22:49pixelmaI don't know dividing that code into two parts sounds complicated to me
02:23:11Soap(or even people with 16 bit targets can choose not to have such bloat in their core)
02:23:14safetydanpixelma, this would all be "post WPS tokenizer" which should simplify things greatly
02:23:14 Quit webguest29 ("CGI:IRC (EOF)")
02:23:36 Quit robin0800 (" HydraIRC -> <- The future of IRC")
02:23:53 Quit Thundercloud_ (Read error: 110 (Connection timed out))
02:23:57Nico_Ppixelma: safetydan is right. the current wps code isn't suited for what we are suggesting. it all relies on the tokenizer
02:24:06amiconnSoap: The lcd drawing will also be more unified than it is atm
02:24:13 Join robin0800 [0] (
02:24:59***Alert Mode OFF
02:25:15*amiconn doesn't understand why people even think that forking would be useful
02:25:33amiconnOne of the big advantages of rockbox is that it is multiplatform
02:25:37SoapI wasn't trying to argue for it - I was mearly trying to prod the conversation so that I might learn.
02:27:00Soapeveryone is free to say "Soap, shutup. The grownups are talking."
02:27:09Lloreanamiconn: Multiplatform is good, definitely. But we need to start figuring out some way so that the restrictions on some won't limit others.
02:27:33amiconnYes, and that should be done by making the code more flexible instead of forking
02:27:33 Join atsea- [0] (i=ariel@gateway/tor/x-678484f2258392e4)
02:27:43amiconnI am working on this right now
02:27:48LloreanThe 2mb RAM devices are providing a lot of stumbling blocks, and while they definitely should still be supported and benefit from advancement, something needs to be come up with.
02:28:13*amiconn again reminds of the iFP7xx
02:28:24LloreanThat one's even worse. :)
02:28:32LloreanI'm actually thinking of Voice right now.
02:28:59LloreanThe act of even adding more strings makes voice harder to fit on any of the low RAM targets, right?
02:29:05amiconnlangv2 rework is being worked on
02:29:16pixelmavoice will definitely get better with langV2 and target specific voice files
02:29:24amiconnThen every target will only have those voice strings it actually uses
02:29:34LloreanBut that's just a stopgap if Rockbox keeps expanding, right?
02:29:41amiconnNo it's not
02:30:08amiconn*many* voice strings from swcodec don't apply to hwcodec, and never will
02:30:45amiconneq, crossfade, crossfeed, ...
02:30:45LloreanBut that doesn't mean that the HWCodec won't hit its voice file limit again at some point. It just means it's a good distance off again.
02:31:06pixelmaand many voice strings for color screens won't apply to mono screens etc.
02:31:19amiconnDo you think there will be new features for hwcodec which will require that many new clips?
02:31:58amiconnI *can* think of a couple of features, but those will require only few clips, or none at all
02:32:32LloreanI just don't think there's a guarantee that the voice files on hwcodec will stay below a certain size.
02:32:34LloreanOr on the iFP
02:33:23amiconnLlorean: One problem on low-ram targets atm is that adding language strings and voice clips expands all targets (both core and voice file), including those which will never be used on certain targets
02:33:37mentlinuxstb: another idea - do you think it's possible to steal a button from keyboard controller and use it as a sort of very lowspeed serial line? (eg multiplexer for a larger keyboard)
02:33:48Lloreanamiconn: Yes, and LangV2 fixes should resolve that part, right?
02:34:03 Quit lukaswayne9 ("Ex-Chat")
02:34:59 Quit robin0800 (" HydraIRC -> <- Go on, try it!")
02:35:48linuxstbment: I guess so. But a lot of targets already have remote control interfaces.
02:36:05*linuxstb sleeps
02:36:09 Quit linuxstb ("CGI:IRC")
02:37:02safetydanWell if all those measure will be sufficient, then hooray! No complicated hacks required :)
02:38:14z0dewhere can i find info on what visual files the x5 can play with rockbox?
02:38:21z0dewiki is hard to navigate :(
02:38:32Nico_Pso i can increase the codesize without second thoughts ? (only a bit, don't worry)
02:39:25safetydanz0de, what do you mean "visual" files? Pictures and movies?
02:40:23z0debest equivlent word to audio i could think off
02:40:24safetydanWell for movies it's MPEG-1 and MPEG-2 only. See
02:40:33Nico_Pzeth: nice article :)
02:40:59z0dethanks. i'll have a read of that
02:41:00safetydanDoes the JPEG viewer really not support progressive scan?
02:41:16Lloreansafetydan: Not "progressive scan", progressive jpegs.
02:41:36Lloreansafetydan: The ones that go from blurry to sharp as more data loads over the internet.
02:42:20safetydanLlorean, mistype on my part :) I'm familiar with progressive jpegs, just surprised it's not support.
02:42:53amiconnIt requires a rather different decoder
02:43:09safetydanNow someone just needs to port dcraw to Rockbox and you can treat it as your photo storage system.
02:44:55pixelmaNico_P: even though the langV2 cleanup decreases bin size a lot, it doesn't help completely for _rombox_ but will also help for the future. Also I think a second thought would always be worth it in this case (maybe even high mem targets could benefit too) ;)
02:46:38Nico_Ppixelma: i agree fully. I was actually very disappointed when I saw my WPS code was becoming a bit bigger than the current one, because I was hoping to reclaim at least a small part of what I had added with cuesheet support
02:46:38 Join Bawitdaba [0] (
02:47:13Nico_Pespecially as it's not an unanimously loved feature ;)
02:48:34pixelmamay I ask what "a bit" is?
02:48:59Nico_Ppixelma: do you have a target in mind ? i'll check now
02:49:47Soap1/8th of a byte ;)
02:49:52 Join chrisjs169 [0] (n=Jack@unaffiliated/chrisjs169)
02:50:08 Quit funky ("leaving")
02:50:36pixelmais that different from target to target? I'd be interested in the OndioFM of course
02:52:00pixelma(the impact of your patch for this target) ...
02:52:13 Quit lImbus (" (fork(bed))")
02:52:25 Join JPMaximilian [0] (
02:53:47Nico_Ppixelma: i did a build on the recorder... i'll check on the ondio fm now. I got -141 on the ucl and +172 on the uncompressed image
02:54:03 Quit JPMaximilian (Client Quit)
02:56:05Nico_P+109 on the ondio fm
02:57:25 Join Thundercloud_ [0] (n=thunderc@
02:57:37pixelmathe rockbox.bin?
02:57:42Nico_Ppixelma: and yes it's different from target to target, because of all the different ifdefs for the different tahs
02:57:57Nico_Ppixelma: err, rockbox.ucl
02:59:29pixelmaNico_P: generally I could understand that but the difference in the WPS for recorders/Ondios is only the rtc features
03:00:08 Join robin0800 [0] (
03:00:28 Join JPMaximilian [0] (
03:00:40Nico_Ppixelma: the impact is probably very similar across targets, as they keep the same tags, but I haven't done enough tests to really know
03:00:44 Quit JPMaximilian (Read error: 104 (Connection reset by peer))
03:01:18Nico_Pbut some targets have more additional code than others
03:02:36pixelmasure, thanks for the info... just had the impression the difference (by number) was a bit high (but you can also only compare the uncompressed bin size)
03:02:52 Quit robin0800 (Client Quit)
03:03:00 Join wrobbie [0] (
03:03:33Nico_PI didn't get the uncompressed size for the ondio fm
03:04:12amiconnNico_P: check apps/rockbox.bin
03:04:20amiconn(in the build dir)
03:04:38amiconnFor Ondio FM, checking ajbrec.ajz also works
03:05:09amiconnThat works for all archos targets except the recorder v1, where ajbrec.ajz is already self-extracting
03:05:26amiconnrockbox.ucl is compressed, and its size not *that* interesting
03:06:40z0dehmm, i downloaded the movie samples for thwe x5 but cant get them to play on in
03:06:43z0deon it*
03:06:51Nico_P+652 on rockbox.bin :(
03:06:56z0deoh wait needs plug in
03:08:47 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
03:09:06 Join miepchen^schlaf [0] (
03:11:07Nico_Ptime for me to go to bed
03:11:23 Quit Nico_P (Remote closed the connection)
03:13:33 Quit billytwowilly (Remote closed the connection)
03:14:56 Join goffa [0] (n=goffa@
03:18:22menthas rockbox any problems with 6gb ipod? (third generation, maybe?)
03:19:06 Join letmomsleep [0] (
03:19:17letmomsleepso is the build for the h10 player full-featured?
03:21:36 Quit Thundercloud (Read error: 110 (Connection timed out))
03:22:21Lloreanletmomsleep: Depends on your definition of 'full-featured'
03:22:29LloreanMany people want things that Rockbox simply doesn't offer at this time
03:22:51letmomsleepwell, i currently have a h340
03:23:01letmomsleepwith rockbox, of course
03:23:14letmomsleepdo they share the same featureset?
03:23:18LloreanWell, I don't believe the radio works on the H10 yet.
03:23:34LloreanAnd Rockbox doesn't have its own USB mode, you have to depend on the iRiver software for that right now.
03:23:41LloreanBut it's almost entirely the same, really
03:24:16letmomsleepdo you think its worth it to upgrade?
03:24:30LloreanI don't consider an H10 an upgrade from an H340
03:24:57 Quit netmasta10bt (Read error: 110 (Connection timed out))
03:25:20letmomsleepis rockbox currently working on the ipod 5g?
03:25:44letmomsleepmy h340 is gigantic, and i'm looking for something smaller
03:25:48Lloreanletmomsleep: The 5G yes, and the 30gb 5.5G, but not the 80gb 5.5G
03:26:01letmomsleepi've been out of the game for a while
03:26:15safetydanz0de, it shouldn't need anything else. The MPEG player plugin ships as part of Rockbox
03:26:16LloreanDo you use recording or FM Radio?
03:26:25Lloreanletmomsleep: Gigabeat F40
03:26:48LloreanWell, it still may be big for your tastes, but it's smaller than the H120, which is smaller than the H340.
03:27:27 Quit johnzorn (Connection timed out)
03:27:38letmomsleepi'll look into it!
03:27:55letmomsleepthe gigabeat s30 is out of the question, though?
03:28:33LloreanFor now, yes.
03:29:18LloreanBut all of the PortalPlayers shouldn't be viewed as comparable to the H300 series. That's the iPods, the H10, and the Sansa.
03:29:42LloreanBecause of the lack of hardware documentation, we've got a lot of stumbling blocks. Rockbox is a bit slow on them, and the battery life is generally poor.
03:30:20letmomsleepi see
03:35:59letmomsleepit's frustrating because all of the competing players lack the ipod's size
03:36:44LloreanI'd rather my player be shorter or less wide than have less depth, honestly.
03:36:56 Quit efyx (Remote closed the connection)
03:37:30letmomsleepi'm exactly the opposite
03:39:31LloreanMaking it thinner makes it fit in a pocket better, sure, fine.
03:39:50LloreanBut my pockets fit the a TI-89 with plenty of room to spare. A modern MP3 player is nothing.
03:40:18letmomsleepah well
03:40:34LloreanI like being my MP3 player to have a little heft to it, but also to be able to easily fit in a cup holder, or hand hold, or wherever I need to wedge it.
03:40:42letmomsleepwhile i'm here, i might as well ask about the slow navigation on my h340
03:41:25LloreanThat's a vague question. Is this "Navigation suddenly started being slow" or "It's always slow but I've never asked" or even "It's slow in the filetree, but not in the menus"?
03:41:48letmomsleepit's only slow in the filetree
03:41:51letmomsleepit didn't used to happen
03:42:01letmomsleepi believe i upgraded to a new version
03:42:34LloreanSlow while scrolling up and down lists, or delays when you try to go into folders?
03:43:14letmomsleepdelays when moving in and out of folders (ie pressing left and right)
03:44:20LloreanCheck up on the Dircache option
03:44:32LloreanIt should solve that problem, and if you've used it previously, it may have become disabled somehow
03:44:42letmomsleepi think i have it on
03:45:02LloreanTry disabling it, shutting down, starting back up again, enabling it, and doing the same.
03:47:55letmomsleepdoes the f40 build happen to support video?
03:48:54LloreanAs much as any other build does
03:49:11letmomsleepoh! that's the issue!
03:49:21letmomsleepi didn't have dircache on, i had database on
03:49:47letmomsleepthe new menus threw me off
03:50:09 Join BigMac [0] (
03:50:24letmomsleephow do you know when the database update is finished?
03:50:40BigMacHey can someone tell me what is happening with my make?
03:50:51Lloreanletmomsleep: When the disk stops spinning.
03:51:40BigMacI am not sure if that means it compiled right or what
03:51:51BigMachaven't ever used vmware for compiling
03:52:30BigMacIt is the entire session but check out near the bottom
03:53:17nlsBigMac: all is in order
03:53:22letmomsleepthis video says video is fully implemented?
03:53:36nlsthen it is wrong :_)
03:53:40Lloreanletmomsleep: It's very, very wrong
03:53:49BigMacBigMac: So when I extract, it should be fine?
03:54:14nlsanyway I should have gone to bed about 4 hours ago...
03:54:20 Part nls
03:54:33BigMacand is there anyway to access \\debian without having vmware open
03:54:43BigMacbecause I have to close everything to run vmware
03:55:04LloreanIf VMWare isn't running, \\Debian doesn't exist
03:58:05BigMacis there anyway to have it make a local copy on my harddrive?
03:58:49 Quit BigMac ("Leaving")
04:00:51letmomsleepis there a simple solution for cover art?
04:01:37LloreanThe simplest solution is "don't bother with it"
04:02:43***Saving seen data "./dancer.seen"
04:04:46letmomsleepso what's next for rockbox?
04:06:20LloreanPrimarily bug fixes and optimizations
04:06:46 Quit BHSPitMonkey (Connection timed out)
04:07:52 Join BHSPitMonkey [0] (
04:09:41 Join webguest88 [0] (i=415fb535@gateway/web/cgi-irc/
04:09:47 Join BigMac [0] (
04:09:57 Quit letmomsleep (Read error: 104 (Connection reset by peer))
04:10:03BigMacYah I etracted and it did not work...
04:10:09webguest88its me again
04:10:27webguest88ROCKBOX ROCKS!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
04:10:58 Quit webguest88 (Client Quit)
04:11:32BigMacIt seems to not have updated at all
04:11:54BigMacIf I delte .rockbox and rockbox.ipod off of root
04:12:00BigMacthen copy my files over
04:12:31BigMacwill it work?
04:12:42BigMacor will I have to reinstall
04:13:57 Join i6zl [0] (
04:14:21i6zlhey guys.
04:14:36BigMacdid something bad happen in the build during compiling (see previous link)
04:14:47 Part i6zl
04:15:11 Join i6zl [0] (
04:18:53 Join FOAD [0] (
04:20:11 Quit Bawitdaba (Read error: 104 (Connection reset by peer))
04:22:10 Join netmasta10bt [0] (
04:22:31 Join bonbonthejon [0] (n=jon@
04:22:54 Quit chrisjs169 (Read error: 145 (Connection timed out))
04:22:56 Join billytwowilly [0] (
04:26:34safetydanBigMac, that compile output looks fine.
04:27:24 Join Bawitdaba [0] (
04:27:38 Join toffe82 [0] (n=chatzill@
04:30:55 Quit merwin (Read error: 104 (Connection reset by peer))
04:31:11 Quit FOAD_ (Read error: 110 (Connection timed out))
04:35:14 Join n00binator [0] (i=42520934@gateway/web/cgi-irc/
04:35:56 Quit n00binator (Client Quit)
04:35:58 Join n00binator [0] (i=42520934@gateway/web/cgi-irc/
04:36:45 Quit n00binator (Client Quit)
04:43:37 Join webguest34 [0] (i=42520934@gateway/web/cgi-irc/
04:44:10 Quit webguest34 (Client Quit)
04:46:23 Join djdtl [0] (
04:46:38 Part djdtl
04:47:17 Quit pearldiver (Read error: 60 (Operation timed out))
04:52:03 Join EnCoRe [0] (i=48b8a339@gateway/web/cgi-irc/
04:52:30EnCoReOk So i just ordered an Iriver h10 and i wanted to know about hacking it... I read some forums and every1 is just happy that this rockbox program works on it
04:52:39EnCoRewhat exactly is the benefit of me hacking it to rockbox?
04:52:58EnCoRene one?
04:54:33scorchegot it off woot? =P
04:58:31EnCoReyes scorche :-)
04:58:33EnCoRethanks bud
04:59:57 Join TCK [0] (
05:00:49 Quit EnCoRe ("CGI:IRC (EOF)")
05:09:13 Quit i6zl ()
05:10:49 Join merwi1 [0] (n=merwin@
05:14:21 Quit TCK (Client Quit)
05:27:36 Part pixelma
05:28:05 Quit Thundercloud_ (Remote closed the connection)
05:30:21 Join webguest52 [0] (i=4738955e@gateway/web/cgi-irc/
05:30:41webguest52just a quick one: checkout: for an H10
05:31:30 Join webguest00 [0] (i=469c49b4@gateway/web/cgi-irc/
05:31:42 Quit webguest00 (Client Quit)
05:32:43 Join Ayle [0] (
05:34:07 Join perplexity [0] (
05:34:28Aylei got a data abort error and i have no idea on how to reset my deviece....
05:34:39Ayleanyone got an idea?
05:34:52safetydanAyle, what device is it?
05:36:03Aylesansa e250
05:36:19 Join printfXh4 [0] (
05:37:05safetydanAyle, I'm not familiar with that one. Try searching the forums or waiting for someone who knows more about Sansas.
05:37:27 Quit webguest52 ("CGI:IRC")
05:37:29 Quit netmasta10bt (Read error: 110 (Connection timed out))
05:38:36safetydanAyle, if you do a google for "sansa reset" the first link is the manual for your Sansa. It suggests holding the power button for at least 8 seconds
05:38:39safetydanmaybe 15 seconds.
05:39:34Aylealready done that
05:39:44Aylei'll try longer
05:40:12Aylegreat it worked
05:40:35billytwowillyoh hey, how's the sansa devel going?
05:45:47BHSPitMonkeyhmm... that's a pretty nice deal on woot
05:47:28 Quit rotator ()
05:47:30 Quit Ayle ("Chatzilla 0.9.77 [Firefox]")
05:50:05billytwowillydon't taunt me. woot won't ship to canada.
05:50:40LloreanBHSPitMonkey: It's refurbished, those iRivers seem to have a decent failure rate, and the H10 has all the same drawbacks as an iPod. Even at $100 for 20gb, I'm not sure I'd go for that deal. :)
05:51:30BHSPitMonkeywell I have a nice lil' iPod, so whatev'
05:51:45LloreanIt's basically an iPod mini with FM radio and a slightly larger color screen.
06:01:22 Quit wrobbie (Read error: 110 (Connection timed out))
06:02:46***Saving seen data "./dancer.seen"
06:04:34 Quit perplexity (Read error: 110 (Connection timed out))
06:07:37 Part Llorean
06:10:15 Nick joshin is now known as josin_ (n=joshin@unaffiliated/joshin)
06:10:19 Nick josin_ is now known as joshin_ (n=joshin@VDSL-130-13-11-75.PHNX.QWEST.NET)
06:19:54 Part safetydan
06:31:54 Join pearldiver [0] (
06:41:33 Part merwi1
06:42:09 Quit SirFunk (Read error: 110 (Connection timed out))
06:54:31 Quit Aikon_ (Read error: 110 (Connection timed out))
06:54:58 Join Aikon_ [0] (
06:59:19Bawitdabaanyone know if the the toshiba gigabeat s is rockboxable or in the works
06:59:29Bawitdabai think only the f series is?
06:59:39scorchein the workas
07:00:08Bawitdabaim thinking about getting one
07:00:14Bawitdabai was gonna get an ipod 80gb
07:00:19Bawitdabashit is insanely pricey
07:00:27Bawitdabathey dont even give u a charger either
07:00:31Bawitdabawhat kind of crap is that
07:01:19 Join daniel1234 [0] (
07:03:01Alias777why the hell wouldnt you get an creative vision M
07:03:07Alias77760GB, cheap as hell
07:03:20Bawitdabawanted to rockbox it
07:03:24Alias777o :(
07:03:28Bawitdabaand creative m only does mtp
07:03:29daniel1234hey I have got the rockbox vmware image thing installed I found a linux screen reader will I be able to install the screen reader
07:03:33Bawitdabagap'd playback
07:03:34Alias777why do you want rockbox
07:03:37Alias777main reason?
07:03:47Alias777CVM has near gapless
07:03:55Alias777usually less than half a second
07:04:06Alias777CVM has themes you can download and built in
07:04:08Bawitdabamostly for mass storage device
07:04:13Bawitdabaso i can keep backup of mp3s
07:04:14Alias777all the colors and backgrounds is chanbgeable
07:04:21Bawitdabaand easily sync with out 3rd party stuff
07:04:23Alias777you partition CVM into a UMS device
07:04:36Alias777and part of it works with the player part of it is like a mass
07:04:45Bawitdabai know that
07:04:50Bawitdabai mean i want it as a mass device
07:04:52Bawitdabafor mp3s
07:04:56Bawitdabaso i can just drag drop
07:05:03Bawitdabastore coverart
07:05:06Alias777yeah, thats a bad thing for thr CVM :(
07:05:26Alias777dont succumb to the fad and buy an ipod just for that though :(
07:05:27Bawitdabadunno if cvm does coverart
07:05:33Bawitdabaand if it did id have to id3 it
07:05:33Alias777it does
07:05:37Bawitdabamaking mp3s a lot bigger
07:05:42Alias777no? lol
07:05:47Alias777id3s take up a few bytes dude
07:05:57Alias777dunno what youve been smoking
07:06:09Alias777one or two bytex x60gb equals maybe a few MB
07:06:15daniel1234does rockbox run on the creative zen or any creativies
07:06:26Bawitdabaif u put 1 picture per mp3 cd
07:06:27Alias777not creative zen
07:06:32Bawitdabavs 1 picture for every mp3 in a mp3cd
07:06:43Bawitdabaits a lot more costly to have it in every mp3 as id3 tag
07:06:51Bawitdabaand its more then just a few bytes
07:07:05daniel1234is there any port to the creative yet
07:07:08Bawitdabaprolly 50-100 mb atleast
07:07:16Bawitdabanaw not yet
07:07:24Alias777but with 60gb
07:07:26Bawitdabathere is a good sized thread about it
07:07:28Alias777doesnt really matter that much
07:07:35Bawitdabadude 60gb isnt big enuf for me
07:07:40Bawitdabai have like over 70gb
07:07:43Bawitdabaid have to trim it down
07:07:56daniel1234r the creativies cheaper than iriver
07:07:57Bawitdabawas my main reason for wanted an ipod
07:08:11 Part toffe82
07:08:26Bawitdabai really have a hard time deciding what to get
07:08:30Alias777me too
07:08:33Alias777im choosing the CVM
07:08:34Bawitdabai guess cost is a pretty good factor
07:08:42Alias777CVM is muuuuuch cheaper
07:08:44Bawitdabaipod is just way to fawking expensive
07:08:45Alias777than every iPod
07:08:52Alias777CVM 30GB = 159
07:08:53Bawitdabalike $400 by the time u get a charger and shit
07:09:01Alias777ipod 30GB = 250 without charger and eveyrthing
07:09:14Alias777the choice is clear my brothers
07:09:46Bawitdabawell i like creative cuz ive had sblive and audigys
07:09:54Bawitdabathey do amazing sound quality
07:10:57Alias777$175 without rebate
07:11:06Alias777trusted retailer
07:11:09Alias777free shipping
07:11:15Bawitdabalol 30gb
07:11:26Alias777i have
07:11:28Alias7772GB in music
07:11:33Alias777and im getting 30GB
07:11:53Alias777this 2GB has lasted me 3 years
07:11:56Bawitdabatoo bad it doesnt play dvds
07:12:03Bawitdabathen id need tb players
07:12:07Alias777how do you get that much music
07:12:08Alias777tell me
07:12:17Bawitdabadownload it?
07:12:37Alias777from where? how do you download so much?
07:12:44Alias777more than 30GB? how is that humanly possible?
07:12:58Bawitdabaive seen ppl with way more then that
07:13:14Alias777lol but how
07:13:18Alias777i know they exist
07:13:23Alias777i just dont understand how
07:13:37Alias777i use for mainstream and limewire for anything less than mainstream
07:13:40Bawitdabai see
07:13:52Alias777i want more than 2GB of music :(
07:14:46Bawitdabaid like to just copy and paste that, would be much simplar then mtp stuff
07:16:30 Join perplexity [0] (n=joust@
07:17:32Alias777can you like
07:17:36Alias777mail me your hard drive
07:18:03Bawitdabawhich 1
07:18:04Alias777hows vista treating you btw?
07:18:11Alias777are you rich?
07:18:13Bawitdabalol its ok
07:18:18Bawitdababeen runnin it for a few months now
07:18:20Alias777i dont have enough money to buy all this stuff :(
07:18:21Bawitdabasucks ram real good
07:18:32Alias777how old are you
07:18:39Bawitdabahorrible for playin games
07:19:15Alias777how do you get the money to buy your multiple hard drives and Vista and such
07:19:17scorchemove it to PMs
07:19:36Alias777thought you were an op
07:19:38Alias777but you arent\
07:19:41Alias777so no :)
07:19:43Bawitdabahe can be
07:19:45scorchewhat does it matter?
07:19:46Bawitdabawatch magic
07:20:03Alias777scorche: yes, because if you're an op i have to listen to you
07:20:24Alias777but i don't. if you don't want to listen to us retarded people talk, then minimize your IRC
07:20:27Alias777not that difficult
07:20:40*Bawitdaba gets popcorn
07:20:44Mode"#rockbox +o scorche " by ChanServ (ChanServ@services.)
07:20:45Kick(#rockbox Alias777 :listen to doesnt matter if i am an op) by scorche!i=ScorchE@rockbox/administrator/scorche
07:20:55 Join Alias777 [0] (
07:21:02Alias777sorry mr. op :(
07:21:02BawitdabaI am currently oped in 22/67 channels on < 17 > different networks. I control < 643 > inferior mortals
07:21:40Alias777now i must sleep
07:21:42scorcheit odesnt matter if i am an are breaking the rules...this is not a social channel
07:21:45Alias777g'night all
07:21:48Alias777and sorry scorche :D
07:21:53 Nick Alias777 is now known as Alias777-zzz (
07:22:13Bawitdabayeah im out too i gotta get some sleep
07:22:21Bawitdabaand i still got no idea what mp3 player to get
07:22:41Bawitdabaeh maybe
07:23:00scorchewe also ask that you dont change your nick for away
07:23:06Alias777-zzzoh, kk
07:23:11 Nick Alias777-zzz is now known as Alias777 (
07:23:29Bawitdabaand u listened to him
07:23:52*Alias777 sets +b
07:24:20scorchelast warning...stop spamming the channel with useless, off-topic messages
07:24:30Alias777jesus okay good night mr. mean op
07:24:48Alias777actually, bad night to you
07:24:55Alias777good night Bawitdaba
07:24:57scorchea mean op would have banned you by now
07:25:04Alias777probably true
07:25:08Bawitdabaso ontopic gigabeat-s/creative zvm60gb/ipod 80
07:25:11scorcheas i said...this is not a social channel
07:25:13Bawitdabawhat should i buy
07:25:22Alias777lol zvm
07:25:28Alias777farewell TOODLES
07:26:35 Quit daniel1234 (Read error: 60 (Operation timed out))
07:26:57Mode"#rockbox -o scorche " by ChanServ (ChanServ@services.)
07:37:17 Join aliask [0] (
07:49:07Mode"#rockbox +o scorche " by ChanServ (ChanServ@services.)
07:49:24Mode"#rockbox +b *!n=Alias777@* " by scorche (i=ScorchE@rockbox/administrator/scorche)
07:49:24Kick(#rockbox Alias777 :I thought about it and changed my mind.. see you later. much later...) by scorche!i=ScorchE@rockbox/administrator/scorche
07:49:40Mode"#rockbox -o scorche " by ChanServ (ChanServ@services.)
07:53:31*aliask goes to read the logs for some entertainment.
07:55:02 Join _Veseliq_ [0] (
07:58:32 Join Nico_P [0] (
08:02:33 Join bug [0] (n=bug@
08:02:46bugDoes RockBox support S1MP3?
08:02:49***Saving seen data "./dancer.seen"
08:04:21bugThank you. Have a nice day.
08:06:48 Join Stalwart^ [0] (
08:09:52 Join Rob222241 [0] (
08:09:59 Join tucoz [0] (n=tucoz@rockbox/staff/tucoz)
08:12:58 Part billytwowilly ("Konversation terminated!")
08:13:04 Join billytwowilly [0] (
08:20:33 Join daniel1234 [0] (
08:21:39 Quit Nico_P (Remote closed the connection)
08:21:56 Quit cynicalliberal (Read error: 110 (Connection timed out))
08:22:18daniel1234where do I put the rockbox sorce when I am using vmware to compile and stuff
08:22:43daniel1234what directory
08:23:29 Quit Stalwart (Read error: 110 (Connection timed out))
08:25:10 Join ender` [0] (
08:25:18scorchein a folder in your home dir
08:26:02daniel1234which is the home dir the folder that rockbox.vmx is located in is the home dir in the vmware image
08:27:20 Join LinusN [0] (i=linus@gateway/web/cgi-irc/
08:27:39daniel1234where is that isn't there only 1 folder
08:28:05 Quit Rob2222 (Read error: 110 (Connection timed out))
08:33:40 Part daniel1234
08:35:50 Join B4gder [0] (
08:36:41 Join kaaloo [0] (
08:37:46 Part kaaloo
08:40:09 Join GodEater_ [0] (
08:43:22 Join Stalwart_ [0] (
08:46:13 Join bluebrother [0] (i=C8ljdhED@rockbox/staff/bluebrother)
08:46:28 Quit GodEater (Read error: 60 (Operation timed out))
08:47:44 Part ootput
08:48:51 Join GodEater [0] (
08:49:20 Part tucoz
08:51:55 Quit GodEater_ (Read error: 60 (Operation timed out))
08:55:10 Quit Stalwart^ (Read error: 110 (Connection timed out))
08:56:08 Join Arathis [0] (
09:00:08B4gderanyone knows about the sansapatch status?
09:00:22B4gderI brought my sansa to work today and thought I'd try it out
09:01:04B4gdersansapatcher even
09:01:46*LinusN figures B4gder will soon try one of the recovery modes
09:02:14B4gderhaha, yeah it is about time I get to face them really good myself
09:02:43JdGordonsansapatcher wprks reat
09:02:57JdGordonbut OF detection in the hidden parition isnt perfect
09:03:04JdGordonso make sure you have an OF.mi4/bin
09:03:33JdGordondoes e200tool give us access to the hard disk without the OF?
09:03:49B4gderthere is no hard disk ;-)
09:04:03*JdGordon wants to figure out what is being set by the OF that it knows to rebuild its db
09:04:10JdGordonbah, memory....
09:04:35B4gderhehe, but no the e200tool only fiddles with ram memory. I think.
09:05:05JdGordoncan I use rb to dump the hidden partition?
09:05:36B4gderuse dd under linux
09:06:05JdGordonno good... can only get to usb once the OF finishes rebuilding its db...
09:06:11JdGordonso no way to see what changes
09:06:35B4gderbut the usb mode is made in software
09:06:46B4gderit could easily just set a flag in ram whenver something is written to it
09:07:20JdGordonin nvram?
09:07:23B4gderI think the key here is rather to not run the OF ;-)
09:07:40JdGordonof course.... but the only way to get usb atm is through it
09:08:16B4gderyou tried if uploading music with different extensions also make it re-index?
09:08:38B4gdernot that it would help much anyway
09:08:41JdGordonIm pretty sure it will rebuild every time usb is disconnected
09:09:34B4gderyeah, it probably just remembers that anything has been modified on the nand
09:10:30 Join wrobbie [0] (
09:10:40*JdGordon does not like it not going verbose mode when booting the OF
09:10:59B4gderwasn't that fixed?
09:11:17JdGordonI emailed barry about that and he disagreed
09:11:30JdGordonjust mounting the drive is enough to trigger a rebuild.. no nand changes
09:13:15JdGordoncan the nvram be dumped in rb?
09:13:38B4gderyou can of course write a plugin to do whatever you want
09:14:06JdGordondo you know what address that ram is at?
09:14:22B4gderwhat memory are you referring to when you say nvram?
09:14:51JdGordonany ram that isnt the nand..?
09:15:00JdGordonthe 2gb..
09:15:17 Join petur [0] (i=d4efd6a6@gateway/web/cgi-irc/
09:15:22JdGordonthe OF probably would but its settings in the hidden partition tho..
09:15:35B4gderthe 2GB is the nand flash
09:15:48B4gderand we access that as a fat file system, the same way the OF does
09:15:53JdGordonyeah.. is there other non volatire ram in it?
09:16:13B4gderthere is another (nor) flash
09:16:33B4gderbut I'm not sure at what address it is
09:16:55B4gderI would expect the pre-bootloader to be in that flash
09:17:40B4gdersandisk chips even, and it shows how much SanDisk "cooperates" with us when they can't even give us docs on them...
09:19:01B4gderwould you happen to know what the problem is with the "detection" of the OF in the hidden partition?
09:19:13JdGordonchecksum error
09:19:21JdGordonit doesnt happen everytime..
09:19:27JdGordonand it usually works the 2nd time...
09:19:35B4gderthat sounds weird
09:20:02 Join GodEater_ [0] (i=c2cbc95c@gateway/web/cgi-irc/
09:20:22B4gdercoffee coffee coffee
09:20:48GodEater_is that a play on "Badger Badger Badger" ? ;)
09:20:49JdGordonthe database is in the hidden parititoin!
09:21:16scorcheGodEater: were you there back when i did my photoshopping? =P
09:21:25GodEater_can't say I was...
09:21:30JdGordonuncidoe chars are 2 bytes correct?
09:22:30GodEater_invaluable reading ;)
09:22:56JdGordondont really care about it.. just see all the text in the parition is |t.u.r.b.e.d...a.| ...
09:23:04scorcheGodEater: read from this onward:
09:23:54GodEater_JdGordon: In that case I think it's safe to say that in most cases Unicode chars are 2 bytes - though this is not always true
09:24:07JdGordonah ok
09:24:42GodEater_scorche: heheh - cool!
09:25:08JdGordonwhats 5mb in hex?
09:26:21scorchei never found out why LinusN edited B4gder's post though
09:26:53*JdGordon loves google calc
09:27:05LinusNscorche: which one?
09:27:21 Join daniel1234 [0] (
09:27:31scorchethe shopped B4gder's face onto the bagder bagder bagder image
09:27:34scorchelinked above
09:27:49daniel1234where is the home dir for the vmware image
09:28:50LinusNscorche: i don't see any post from Bagder on that url
09:28:56scorchedaniel1234: it is the place where you are when you firstboot the image
09:29:05scorcheLinusN: next page
09:29:14jhMikeSoy found a bug where queue_wait_w_tmo(&voice_queue, INT_MAX) is always timing out...
09:29:48LinusNah, i changed "Speachless" to "Speechless"
09:30:02daniel1234if I boot the image from a folder called rockbox image where would my home dir be and that folder is on C:\
09:30:39scorchedaniel1234: as i said before, the home dir is inside the file structure of the image itself
09:30:43*JdGordon found the OF hidden in the hidden partition :)
09:31:03daniel1234so how would I find my compiled build
09:31:13scorcheyou can access it in windows by typing \\debian\user in windows explorer
09:31:24scorchethe VMware wiki page describes this
09:33:18*jhMikeS is getting sick of everything turning up threading bugs :P
09:35:14daniel1234so when I type make zip can I find the zipped compiled rockbox in windows explorer somewhere
09:35:34scorchein the network samba share
09:36:04scorchein otherwords, "\\debian\user\build directory"
09:37:46daniel1234I type it in to run and I get an error that it can't find it
09:38:02JdGordonis a 262Kb buffer too large for the bootloader?
09:38:03LinusNjhMikeS: that's what we get when trying to make the scheduler smart and introducing thread priorities. If you want even more problems, add dynamic memory to the mix. :-)
09:38:14JdGordoni mean doing a char buffer[512*512]; ?
09:38:36LinusNthe bootloader can use as much memory as it wants to
09:39:38jhMikeSLinusN: rrrr...changed code to depend on that being correct and spent hours wondering why plugin iram was being corrupted :\
09:39:51LinusNwhat was the bug?
09:40:50 Join Entasis [0] (n=Jarred@
09:40:50jhMikeShaven't gotten to fixing it, just ranting about it. queue_wait_w_tmo(queue, INT_MAX) is just timing out instantly, not after INT_MAX ticks
09:41:43jhMikeSwhich means after some changes, the voice codec was being reentered instead of waiting in voice_request_buffer_callback
09:42:45 Join thewho [0] (i=c27f0812@gateway/web/cgi-irc/
09:43:09thewhoJdGordon: ping
09:43:44JdGordonbut dinner... so back in 30
09:46:19thewhoJdGordon: could you please elaborate on the intended magic with the same values for USB_SYS_CONNECTED and GO_TO_ROOT?
09:46:32LinusNjhMikeS: wow, the new scheduler can only use 29 bits for the timeout
09:46:51jhMikeSI'm not sure why the threads next scheduled run time is combined with it's state flags either...that will mess up the tick comparision
09:46:53 Quit GodEater_ ("CGI:IRC (Ping timeout)")
09:47:00jhMikeSit should be a separated variable
09:47:16jhMikeSseparate even
09:48:03jhMikeSthat should reduce the code size without all that masking too
09:48:38 Join austriancoder [0] (n=austrian@
09:48:55SlasherijhMikeS: please don't change the scheduler code atm, a big fix is coming soon :)
09:50:26jhMikeSSlasheri: you kept the sync_queue stuff light, right? it's meant to have practically no overhead for queues that don't need it...and even for ones that do.
09:51:07jhMikeSI suppose I can just make that timeout reasonable that filter the timout messages for the moment
09:51:16jhMikeSand filter*
09:51:50JdGordonthewho: its more good luck that they are the same value... either way the old behavious was that after usb you would be put back in the file browser.. so going to root is the same
09:52:12*JdGordon wonders why its taking so long to read 50 sectors of nand! ?
09:52:30SlasherijhMikeS: yes, synchronous queues are now core-wise thread safe
09:52:45Slasheriand it's even lighter as it was =)
09:52:46*JdGordon also realised he put his code before the OF part which means im fucked :p
09:52:48jhMikeSbut no new overhead for single core?
09:54:55 Join Ribs [0] (n=ribs@
09:55:04jhMikeSThere's also that thing with the voice thread not unboosting during playback...when it waits for the lock in swap_codec. Threads should probably boost cancel if they wait too long or something to that effect.
09:55:40austriancoderandrewg867: ping
09:56:53thewhoJdGordon: but I think it would be better to handle that explicitly rather than rely on such a fragile definition.
09:57:07 Nick Stalwart_ is now known as Stalwart (
09:57:37LinusNjhMikeS: that's the reasoning behind using wait_w_tmo instead
09:58:21LinusNif you use wait_w_tmo, <some large number 3 bits less than intmax>
10:00:11jhMikeSI know, but you can't do that with mutex
10:00:29LinusNsilly me
10:00:56jhMikeSThat wait must be forever and still unboost eventually
10:02:17LinusNi think the cpu boost handling in the kernel needs to be documented in RockboxKernel
10:02:52***Saving seen data "./dancer.seen"
10:03:24jhMikeSSlasheri: will the scheduler allow me to squeeze in counted semaphores? I'm fairly sure I'll need a super fast object with a signaled count that keeps it's count when waited on and blocks when the count is zero.
10:03:26 Quit miepchen^schlaf (Read error: 113 (No route to host))
10:03:31 Join BigBambi [0] (
10:04:14daniel1234do I just type \\debian\user from the run dialog in windows to find my home dir for the vmware image and what drive is it on
10:05:38jhMikeSShasheri: and what of playback.c? I've rewritten a bunch of stuff to tighten things up, and of course having only one swap buffer. you said you changed stuff there.
10:07:09*JdGordon is an idiot :'(
10:07:35JdGordonnever put untested code before the bootloader checking :p
10:09:03JdGordoni hope not :'(
10:10:20 Quit daniel1234 ("Freedom Chat - Your Home Away From Home | | tIRC script by the Freedom Chat Leets")
10:11:59 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
10:13:49 Join daniel1234 [0] (
10:14:34daniel1234where do I extract the daily sorce to if I want to compile it from the vmware image
10:15:25scorcheas i said before, in a folder in your home dir
10:15:40daniel1234what drive is that on C
10:16:07scorche<scorche> daniel1234: as i said before, the home dir is inside the file structure of the image itself
10:16:19scorcheplease stop repeating questions
10:16:38daniel1234in the vmx file
10:17:00 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
10:18:21 Quit thewho ("CGI:IRC")
10:19:03B4gderdaniel1234: run the vmware, extract the source
10:19:16linuxstbB4gder: sansapatcher itself seems to be working fine, the remaining problems are with the new version of the bootloader - barrywardell is working on it.
10:20:19linuxstbAre there still problems with the flash driver on the Sansa (thinking about why the CRC check sometimes fails) ? I seem to remember there being some in the past.
10:20:27daniel1234I have vmware player installed but don't understand where score needs to go
10:20:32 Join Rondom [0] (
10:20:44daniel1234sorce I mean
10:21:39B4gderdaniel1234: run it
10:21:49B4gderthen you have a linux running
10:21:54B4gderuse that
10:22:11B4gderextract the source exactly whereever you feel it fits inside that
10:22:46 Join Redbreva [0] (
10:22:57B4gderOT: I do notice this distribution of vmware images for development turning really popular these days
10:23:08daniel1234yes but I need to put the sorce in my home dir and can't seem to find this in windows so is it only able to be able to be accessed through linux
10:24:18B4gderthat's beyond me
10:24:33B4gderbut then I don't understand why anyone would like to use the windows side ;-)
10:25:04B4gderusing windows just makes me upset
10:25:09scorcheB4gder: when crosscompilers are involved, it makes things very handy
10:25:30B4gderwhy is windows handy with cross-compilers?
10:25:55scorchei was referring to just the popularity of images
10:26:00B4gderah yes
10:26:09B4gderI got this funny image from Xilinx
10:26:20B4gderfull linux, including kde, mozilla and more
10:27:11scorcheaye..that is why i keep the image the way it is...there is even a few things i could probably remove
10:27:20bluebrotherand how much RAM does it require to run nicely? ;-)
10:27:36B4gderI don't know, I have 2gb ram on this box
10:27:37scorchewith VMware tools, things are much better
10:27:56B4gderusing CentOS out of all distros
10:30:27linuxstbdaniel1234: Do you have vmware player running?
10:31:51 Join Redbreva_ [0] (
10:31:55daniel1234I open the rockbox.vmx in the vmware player but don't know how to be able to see the compiled build in windows
10:32:27scorchedaniel1234: that is all explained in the VMware wiki page
10:33:01linuxstbdaniel1234: Forget copying files for now, the first thing you have to do is actually get it running, log into Linux and open up a terminal window.
10:33:16bluebrotherCentOS is nice
10:33:34B4gder740 subscribers of the users list
10:33:52B4gderand that _after_ the bounce of the century
10:34:02daniel1234i have log in made a dir downloaded rockbox I think compiled it but didn't zip it just only problem is can't find anything not even the sorce
10:34:26linuxstbSo you downloaded Rockbox with the "svn" command in Linux?
10:34:47daniel1234yes I can go and tell u the command if u want but yes I did
10:35:24linuxstbSo if you browse (in Windows) to "\\debian\user" (as the wiki page describes), what happens?
10:35:48daniel1234should I type that into run
10:36:14linuxstbNo, but I don't know where (I don't use Windows)
10:36:25*linuxstb looks around for a Windows user
10:37:05 Quit Siku (Read error: 60 (Operation timed out))
10:37:26austriancoderdaniel1234: yes type it into run
10:37:36linuxstbYou can run a directory?
10:37:46daniel1234I get an error
10:38:28GodEaterlinuxstb: the run dialog box will launch an explorer window if you only give it a directory
10:38:54daniel1234the error i get is the network path can not be found
10:39:10GodEaterdoes the vmware image start samba by default ?
10:39:37daniel1234if I open rockbox.vmx up it opens in the vmware image
10:39:56linuxstbdaniel1234: Is Linux running when you try to open \\debian\user?
10:40:03Redbreva_you do have the vm image 'running' when you typed the command?
10:40:17daniel1234know windows is i type it in the run dialog
10:40:35linuxstbYou need to leave the vmware player running.
10:41:17daniel1234ok I ok what dir should I be in
10:41:26linuxstbIt doesn't matter.
10:42:01daniel1234how do I get in to windows and leve the vmware image running
10:42:30linuxstbdaniel1234: vmware is a virtual computer - think of it as having two computers on your desk - one running Windows, and another running Linux. The only way one of those computers can access the files on the other is via the network, and for that to work, both computers need to be running.
10:42:55GodEaterdaniel1234: hold down ctrl+alt in the vmware window to give mouse control back to windows
10:43:25daniel1234ok I will just see if I can get some sighted help to do that
10:45:15GodEaterI find it amusing that Llorean so stricly enforces the "no abbrevations" rule on the forums for the benefit of our non-sighted users, and then I see daniel1234 using the "u" contraction right here ;)
10:47:41 Quit Redbreva (Read error: 110 (Connection timed out))
10:49:56daniel1234can i be on the screen that says press control g to input to the vertual computer or what ever it is
10:50:54linuxstbGodEater: I think that rule is mainly for non-English speakers using translation software.
10:51:06amiconnLinusN: around?
10:53:55daniel1234I am on the screen that says To direct input to this virtual machine, press Ctrl+G can I be on that screen then go to run and type the dir in
10:54:47amiconnLinusN: I stumbled about a slight derivation from the ata protocol in ata.c that might cause more power consumption when the drive is supposed to sleep (on targets with no hw ata poweroff)
10:55:04LinusNamiconn: oh?
10:55:30amiconnWould be nice if you could check and verify
10:55:53amiconnata.c: ata_perform_sleep() (line 571 ff:9
10:56:11 Join kaaloo [0] (
10:56:17GodEaterdaniel1234: that should work I think, yes.
10:56:25amiconn...and the description of the ata SLEEP command
10:57:13daniel1234I tried it it said \\debian\user
10:57:14daniel1234The network path was not found.
10:57:15amiconnWe don't do the step that says 'read status register to clear INTRQ'
10:57:31 Quit bun-bun ()
10:58:30LinusNamiconn: does it work better if you add it?
10:58:38GodEaterscorche: I'm stumped on daniel1234's problem - you any ideas ?
10:59:06scorchesamba can be finiky...rebooting the image helps or rebooting the host
11:01:45 Join borisyeltsin [0] (
11:01:59linuxstbSlightly off-topic, but my girlfriend runs Linux in vmware, and I found that exporting the Windows drive via the Cygwin NFS server to be a much more reliable way of sharing files than samba. This also means the files live on your Windows drive, so are available without vmware running.
11:02:26 Quit billytwowilly (Read error: 104 (Connection reset by peer))
11:02:50daniel1234how do i do that
11:02:52linuxstb(but this will of course slow down Rockbox builds to cygwin levels, so isn't helpful in a Rockbox context)
11:03:02amiconnThe specs then say: "Because some host systems may not read the Status register and clear the interrupt pending, a device may automatically release INTRQ and enter Sleep mode after a vendor specific time period of not less than 2 s."
11:03:16amiconn*may*, not *should* or even *must*
11:03:16linuxstbdaniel1234: Sorry, I wasn't suggesting it as a solution for you, just a general comment about using vmware.
11:03:50amiconnLinusN: Dunno, can't really test. The best test case would probably be an ipod with 1.8" hdd
11:04:20amiconnThe coldfire targets all do hw ata poweroff, and I suspect the mini's microdrive auto-sleeps
11:06:54amiconnThis can only be tested easily by measuring the current consumption of the target
11:08:05daniel1234anyone know what I should do can't seem to access the dir could I copy the files I compile to a folder that windows can see
11:08:50LinusNamiconn: i think we should follow the specification
11:09:54 Join mechanor [0] (
11:10:12mechanorhmm.. my h340 doesn't seem to play anything anymore!
11:10:30mechanorthe disk spins for a moment when i choose a song, but then stops, leaving only the tag information onscreen
11:10:35mechanorthe song never plays
11:10:46 Join barrywardell [0] (
11:11:24daniel1234how do you copy a file in linux
11:11:46mechanornever mind! i guess it just needed a reset
11:14:02barrywardellJdGordon: around?
11:15:20austriancoderbarrywardell: do you have an sansa with fm?
11:15:55 Join _Amblin [0] (
11:16:09austriancoderbarrywardell: and we have the problem how tho power the fm chip?
11:16:38barrywardellI don't know. I haven't heard anything about the fm chip. I think someone was working on it here though
11:16:49B4gderyes, andrewg867 is
11:17:48barrywardellbtw. I figured out how to disable that annoying database refresh in the OF :)
11:18:09B4gdertell us!
11:18:21*_Amblin stands at attention.
11:18:29barrywardellthere's a NVPARAMS section in the hidden partition
11:18:38austriancoderandrewg867: ping
11:19:56barrywardellbyte 0x7810e1 of the hidden partition determines whether a database update is done or not
11:20:14barrywardell1=update, 0=don't update
11:22:16daniel1234anyway i could copy the zip file i make in linux of my compiled build to a folder on my hard drive
11:24:31 Quit daniel1234 ("Freedom Chat - Your Home Away From Home | | tIRC script by the Freedom Chat Leets")
11:28:42 Join Amblin [0] (
11:28:46 Quit perplexity (Read error: 60 (Operation timed out))
11:29:19 Quit BigBambi (Read error: 110 (Connection timed out))
11:31:06bluebrotheranother idea regarding the vmware image: I guess ssh is running, so copying files should also work using scp
11:31:23*B4gder loves sshfs
11:33:27B4gderbut that would then require the sshd to run in windows
11:38:56BigMacany reason I would get these errors when compiling?
11:39:09BigMacdown in the last threee lines
11:39:40BigMacI extract it over my current build and it does not seem to recognize
11:39:55LinusNBigMac: those are not errors
11:40:11 Quit _Amblin (Read error: 110 (Connection timed out))
11:40:22BigMacOk then what would cause me not to be able to extract over my current build?
11:40:37BigMacshould I extract outof \\debian
11:40:41BigMacthen copy over?
11:41:43*barrywardell creates SansaE200FirmwarePartition
11:41:43BigMacany ideas?
11:42:22LinusNi believe you might want to update the bootloader, and also remove the old rockbox.whatever before extracting the zip
11:43:02BigMacok but I have the black background bootloader already
11:43:08BigMacI will re add though
11:45:30 Quit BigMac ("Leaving")
11:46:32barrywardelllinuxstb: I think I know why the usb detection is a bit buggy. it also could explain JdGordon's problems with the checksum in the bootloader
11:47:08barrywardelllinuxstb: when you connect the USB cable, it resets UDC_ENDPOINTLISTADDR to 0x0
11:47:35barrywardellthis is a pointer to the start of the usb endpoint data structures (which we don't have yet)
11:47:53barrywardellmy guess is that the usb chip then overwrites some of this data
11:47:55 Join _Amblin [0] (
11:49:31B4gderhey, we need a few commits to see if we can detect any build speed diffs! ;-)
11:49:47 Join funky [0] (n=repulse@unaffiliated/funky)
11:52:19 Quit Amblin (Read error: 110 (Connection timed out))
11:52:42barrywardellB4gder: do you know why the sprintf functions in Rockbox don't support the .precision format?
11:53:03 Quit Nibbier ("bei uns hat noch niemals nicht koana koan durscht ned ghabt")
11:53:04B4gderjust because it is a very simple implementation
11:53:37barrywardellwould it be desired to have it? i added it like this:
11:53:53barrywardellfor strings only
11:54:49B4gderI'm fine with that
11:55:31barrywardellcool. I'll have a commit for you now in that case ;)
11:55:55bluebrotherB4gder: no, my idea was to run the sshd on the vmware side
11:56:05B4gderbluebrother: I realized that
11:56:21B4gderbut then you can't use sshfs! ;-)
11:57:41*bluebrother should check out sshfs ;)
11:59:33JdGordonbarrywardell: back
11:59:51barrywardelli figured out a couple of things
12:00:28*JdGordon too
12:00:40JdGordonnever do anything untested before the OF decision :'(
12:00:49*JdGordon reading e200unbrick :'(
12:01:09barrywardellwhat happened?
12:01:09JdGordonhow do you get into recovery mode?
12:01:24B4gderit is mentioned on the install page
12:01:28barrywardellswitch hold on and hold rec while turning on
12:01:30JdGordonI tried dumping the hidden partion.. but it seems to have frozen
12:01:40 Join toni1 [0] (i=d9b95eb5@gateway/web/cgi-irc/
12:01:44toni1hi all
12:01:51barrywardellit took quite a long time when I dumped the hidden partition last night
12:02:01toni1barrywardell: I experimented with the sansa lcd controller a little.
12:02:12barrywardellalthough I did one sector at a time, so maybe that's part of the problem
12:02:13JdGordonbarrywardell: woot! it connects
12:02:20toni1barrywardell: The only solution to the display problems I have currently: Use another frame buffer
12:02:20linuxstbbarrywardell: Is the UDC_ENDPOINTLISTADDR problem easy to fix?
12:02:34toni1barrywardell: and copy the content at each lcd_update, which takes 3ms/update (25MB/sec DRAM speed)
12:02:37barrywardellJdGordon: make sure you only put a mi4 on. other files tend to cause bricking!
12:02:54***Saving seen data "./dancer.seen"
12:03:10barrywardelllinuxstb: i guess the easiest thing would be to set it to a dummy endpoint descriptor
12:03:54JdGordonunbricked :)
12:04:18barrywardelltoni1: ah, so triple buffering? i've heard the gigabeat does similar.
12:04:41barrywardelltoni1: it would be nicer if we could figure out when the lcd has finished its dma transfer
12:04:43toni1barrywardell: Yes, it solves the problems. What do you think?
12:05:17toni1barrywardell: I checked the LCD_REG_5 but with no result.
12:05:50toni1barrywardell: If we could trigger the dma start, but how to do?
12:05:51JdGordonLCD_REG_5 is the ony one that changes at all during lcd_update()
12:06:25toni1JDGordon: Yes, at least on the LCD_REG_xx
12:06:32JdGordonbarrywardell: removing the lcd_init works fine, but causes a nasty "freeze" when booting the OF which is why i tihnk we should enable verbose mode tere
12:07:07barrywardelltoni1: have you asked dan_a about that? he wrote the lcd driver. there's also the docs on Bagder's site
12:07:14markunbarrywardell: the Gigabeat LCD controller generates an interrupt when a new frame is being transfered I believe, maybe we can use that.
12:07:44 Quit _Amblin (Read error: 110 (Connection timed out))
12:07:51JdGordonbarrywardell: you said you dumped the hidden partition, have you got it before and after a usb connection?
12:07:51barrywardellJdGordon: i was able to remove the lcd_init without any problems. what sort or freeze did you have?
12:08:07JdGordonwhen you boot OF it appears to freeze for a second or two...
12:08:23JdGordonit doesnt really.. but the user might think it has
12:08:28 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
12:08:53barrywardellah, yes. so you want an indication that it's actually booting?
12:09:23barrywardellmarkun: yes, possibly. when does it generate the interrupt? at the end of the transfer?
12:09:45JdGordonbarrywardell: have you still got your parition dumping code?
12:09:47pixelmathe numbers in the build table for Ipod Video and OndioFM look suspicious
12:09:56barrywardellyes, you want a patch?
12:10:14JdGordonplease :)
12:10:53pixelmafor the last commit and the size table... I meat
12:11:04barrywardellit takes quite a while (maybe 1 minute)
12:11:06JdGordonbarrywardell: do you think the way usb is detected in my patch is bad?
12:11:19JdGordon1min is better than the 10years i was scared mine was going to take :D
12:11:39barrywardellIt's only as bad as the usb detection in rockbox
12:11:47barrywardellwhich I believe is quite bad currently
12:12:36JdGordonwhy only the first 40960 sectors?
12:12:44barrywardellthat's 20MB
12:13:39JdGordonhaha... that sprobably my idiotic mistake then... I took the sector numbers from the wrong partiiton in the fdisk output :'(
12:14:33markunbarrywardell: the LCD controller is still a bit unclear to me. Maybe you can have a look?
12:15:18barrywardellJdGordon: haha, so you were dumping several gigabytes instead!
12:15:54JdGordonno, it must have had other problems also becuase it never even wrote the first sectors
12:16:07JdGordonbut yeah, it would have eventually :p
12:16:26barrywardellmarkun: I'll have a look
12:16:38 Quit Redbreva_ ("Chatzilla 0.9.77 [Firefox]")
12:17:32barrywardellJdGordon: did you see I created SansaE200FirmwarePartition
12:17:42barrywardellwould be interesting to see if it matches what you get
12:17:58 Quit wrobbie (Read error: 110 (Connection timed out))
12:18:18 Quit YouCeyE (Read error: 110 (Connection timed out))
12:18:23JdGordonill let you know in a few min :p
12:19:06JdGordoncant sansapatcher uninstall an old rockbox booloader?
12:19:19 Quit Shaid ("I will see you, in the end. And I will laugh at your pain...")
12:19:39 Join Snowborne [0] (
12:20:50Snowbornewhats the actual improvements you get with rockbox for sansa e200 ?
12:20:59JdGordonnot much atm...
12:21:07barrywardellJdGordon: no, but it can upgrade a current one in place
12:21:30barrywardellJdGordon: just run the install and it will replace the current one an move things around properly
12:22:06JdGordon 0x7810e0 Database rebuild. If this is set to 1, the database is rebuild on next reboot <- so you know the address? can sansapatcher write a 1 here every time it upgrades?
12:22:57Snowbornefigure I'll wait a couple of months longer and check again :P
12:23:25SnowborneDont wanna brick it with an uncomplete build that in reality wont change much at all
12:23:39 Part toni1
12:23:47barrywardellJdGordon: yes. it could do, although I'm not sure at what stage the OF writes to the address. it's probably after disconnecting the usb so too late for sansapatcher.
12:24:06barrywardellJdGordon: i have a patch for disabling it in the bl if you want
12:24:25JdGordonthats even better :)
12:24:48JdGordonthats bassically what all this dumpin was for.. I assumed noone had checked for that yet :p
12:25:25barrywardellJdGordon: I only checked it last night after reading your email
12:25:49JdGordonhave a nice st pats day btw?
12:26:09JdGordondoes the mi4 filename matter in recovery mode? or just the extension?
12:26:23barrywardellhaha, yeah
12:26:50barrywardellit depends on your firmware version. it should be either PP5022.mi4 or firmware.mi4 IIRC
12:27:19barrywardelldid you see that I think I figured out the problem you were having with usb and the checksum?
12:27:36JdGordonno, was it a big problem?
12:28:05JdGordonI mean, hard to find?
12:28:10barrywardellI haven't solved it yet.
12:28:26JdGordonoh, but you think you found it? :)
12:28:35barrywardellbasically what happens is UDC_ENDPOINTLISTADDR gets reset to 0x0 when you plug in usb
12:29:01barrywardellthat should be the address of the usb endpoint data structure
12:29:09barrywardellso the usb controller probably overwrites it
12:30:13JdGordondo we really need to check the checksum? worst case is the image is stuffed so you need to reboot?
12:30:45linuxstbI think it's good to tell the user that the image is corrupt.
12:31:08barrywardelland also we'll be running random code - who knows what that could do!
12:31:14linuxstbObviously, any bugs causing false-positives need to be fixed.
12:31:43barrywardelllinuxstb: i don't know if it is actually a false positive though
12:32:02barrywardellif my theore is correct, then the image really is different
12:32:21linuxstbYou mean the image in RAM, rather than the image on disk?
12:33:14JdGordonthe db rebuild hack didnt work :(
12:33:25JdGordonbut the OF booted from the partiion first go
12:33:34JdGordonfirst go after rb tho...
12:34:15barrywardelllinuxstb: yesh
12:34:45barrywardellthe image on disk is fine. i'm just checking now if the image in ram is different...
12:35:20barrywardellJdGordon: no?
12:35:32JdGordonyeah, didnt work...
12:35:47barrywardellJdGordon: also, I think if you just add a delay before loading the OF it should boot from the partition first go
12:35:50JdGordondumping the partition now... then gonna do usb, then redump
12:36:05pixelmaBagder: do you have an idea what's going on in the size table? Seems related to the new client *shrug*
12:36:22 Quit Rondom (Operation timed out)
12:36:29barrywardellJdGordon: maybe a different firmware version? I was using 1.01.11A from Bagder's site
12:37:58JdGordon1.00.12 is the one i used
12:38:50 Join secleinteer [0] (
12:41:34*JdGordon having no problem with crc errors now...
12:41:51B4gderpixelma: really weird...
12:41:54 Join netmasta10bt [0] (
12:41:54barrywardellis that wit a delay? it only happens when the usb cable is connected
12:42:12JdGordonnope, no delay, no usb
12:42:27JdGordonoh, but with my usb detection code... so yeah, delay
12:43:40barrywardellB4gder: can I safely ignore those gigabeat build errorsthen? i doubt they're related to my commit
12:44:24 Join himitsu [0] (n=himitsu@
12:45:41JdGordon diff between my part.bin's before and after usb...
12:46:30 Quit aliask ("Chatzilla 0.9.77 [Firefox]")
12:48:36barrywardelldoes your part.bin have NVPARAMS at 780000?
12:50:16JdGordon00608000 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
12:50:16JdGordon00780000 4e 56 50 41 52 41 4d 53 00 00 00 00 00 00 00 00 |NVPARAMS........|
12:50:16JdGordon00780010 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
12:50:44*B4gder is a bit scared to tell what the problem was
12:50:54LinusNB4gder is lame
12:51:14barrywardellJdGordon: my guess is that you need to change 78b808 to be 0
12:51:18B4gderthe eternal problem!
12:51:21markunB4gder: just do it..
12:51:33B4gderwhen you're born with a lame brain, you get to suffer a lot
12:51:45B4gderit still only had CVS
12:51:50B4gderno svn checkout on the server
12:52:16bluebrotherso we have that much size change since the move?
12:52:43bluebrotherhmm, longer size comparisons might be interesting ...
12:53:06B4gdershould be fixed now
12:53:21markungraphs! :)
12:53:37bluebrother5000 bytes less than at the move sounds nice
12:53:53bluebrotherwho wasted the most bytes? ;)
12:53:57B4gderthe size table has been very beneficial for that
12:54:13markunfunny, more and more replies to questions in are "upgrade to rockbox!"
12:54:24bluebrotherhow good size isn't an issue for the manual :P
12:54:32pixelmaI guess the cue sheet support is a good candidate
12:54:44pixelma(added about 3k IIRC)
12:54:46bluebrotherfor blaming someone?
12:54:48JdGordon5kb in 2 months isnt that bad...
12:55:07JdGordonSlasheri: had the big one with the endieness stuff iirc
12:55:39pixelmayeah, that too - the both were added within a day
12:55:46 Join Nibbier [0] (
12:56:08markunwasn't the endieness stuff reverted?
12:56:24JdGordonbarrywardell: I assume your upload speed is faster than mine.. is your dump file somewhere I cna download to compare?
12:56:42barrywardelli'll upload it now...
12:56:53JdGordoncheers :)
12:57:47barrywardellJdGordon:"> and
12:58:09*JdGordon wonders if we can upgrade the bootloader from inside rockbox
12:58:13barrywardellthey're with and without database rebuilding
12:58:25pixelmamarkun: afair no - just a bit of optimisation afterwards...
12:59:00*JdGordon goes into recovery mode for the 3rd time tonight :'(
13:00:37barrywardelloops, wrong url, should be">
13:01:14*JdGordon is jelous of your upload speed :'(
13:02:20barrywardellit seems to be only limited by my computer's 100BASE-T card :)
13:03:44JdGordonsorry, my brain sucks tonight,,, how did you go from 0x7810e0 to pinfo->start + 0x3c08 ?
13:06:10*JdGordon is out of ram.. too many 20mb files open :p
13:06:23barrywardell0x7810e0 is the address in bytes
13:06:42barrywardell0x3c08 is the sector (address/512)
13:07:43 Quit petur ("CGI:IRC (EOF)")
13:08:52JdGordonlooks like we can find the needed sector at runtime... your bootloader first bytes... 50 50 42 4c 78 01 06, mine50 50 42 4c 98 01 06
13:09:01JdGordonthe 5th set changes
13:09:30JdGordonoh, hmm.. //me should read first.. that probbaly wont work
13:09:59 Quit Snowborne (" HydraIRC -> <- Try something fresh")
13:10:43barrywardellthe last three bytes are just the bootloader size...
13:13:07JdGordongoogle calc doesnt like devision in hex
13:13:27JdGordonwhich sector do I need to read? :'(
13:14:45barrywardellyou want address 0x78b808?
13:15:27*B4gder uses bc for calculations
13:15:49*linuxstb uses his calculator for calculations
13:15:59barrywardellthat will give you data starting at 0x78b800
13:16:31*barrywardell uses the "programmer mode" on his apple os calculater
13:16:53barrywardellit does bit shifting, AND, OR, ...
13:18:25B4gderbut... bc can show the full precision of 500^500! ;-P
13:18:36 Quit himitsu ("Leaving")
13:19:22 Join miepchen^schlaf [0] (
13:19:59barrywardellyou win...500^500 in the apple calculator = "Infinity"
13:20:18B4gderit is quite a large number
13:20:29*barrywardell goes to lunch
13:21:01B4gderhm, even 500^30000 works
13:21:15*B4gder stops this and tries to work instead
13:22:48 Quit smolyn ("KVIrc 3.2.5 Anomalies")
13:23:23linuxstb500^30000 is too big for my terminal's scrollback...
13:23:44*linuxstb thinks he can live without knowing the answer to that
13:24:48B4gder83351 digits
13:24:59B4gderwell almost
13:26:15pixelmawindows calsulator can keep up to 500^16091 ;)
13:26:22pixelma*calculator too
13:28:46B4gderto me, I think it being available from the command line is why I use it
13:28:51 Join ctaf [0] (
13:29:18pixelmathough maybe not with full precision...
13:29:35B4gderI can't say I use that feature very much
13:30:17JdGordonwoot! no more db rebuilding :)
13:30:18pixelmathat surprises me... ;)
13:36:17preglowlinuxstb: around now
13:38:19 Part kaaloo
13:40:16 Join wrobbie [0] (
13:40:30linuxstbpreglow: That's OK - I just wanted to know if you knew why CONFIG_RTC was #ifdef'd out for the ipod bootloaders. Checking the CVS logs, it turned out that I did it...
13:41:14linuxstbI've removed that #ifdef now, and the bootloader still seems to work fine - including loading the OF.
13:43:18 Join perplexity [0] (
13:47:04 Quit bug (Nick collision from services.)
13:47:39 Join Bug2000 [0] (n=bug@
13:52:09 Join marc| [0] (
14:00:43JdGordonlinuxstb: are the checksums for the various sansa OF's written down anywhere?
14:01:58linuxstbI don't believe so. They also vary depending on whether the firmware is encrypted or not.
14:02:30B4gderwhy would they be written down?
14:02:48linuxstbWhy is JdGordon interested in them?
14:02:56***Saving seen data "./dancer.seen"
14:03:02JdGordonIs there a better way to check which OF version is installed than by that?
14:03:24markunI was wondering if blind users would be happy with morse output.
14:03:26*JdGordon is adding database rebuild removal autodetection thingamy :p
14:03:59linuxstbI suppose a checksum database could be used for that - so is the DB magic different for different OF versions?
14:04:15 Join PaulJam [0] (
14:04:37linuxstbJdGordon: Are you adding it to Rockbox or sansapatcher?
14:05:50JdGordonI was going to read it from the mi4 header, but apparently my math sux and the value isnt what I was expecting :p
14:07:57 Join SirFunk [0] (
14:08:08JdGordon:) it worked
14:09:30JdGordonalthough, this is going to cause problems for people who actually want to use the OF for music
14:11:51linuxstbMaybe enable database building in the bootloader if the user manually starts the OF, but disable it if Rockbox is re-entering the OF due to USB insertion.
14:12:41*linuxstb doesn't know how the "reboot to OF" works on the sansa for usb mode
14:14:08JdGordongood idea :) easy to do
14:16:54JdGordonbarrywardell: ping when your back?
14:18:16B4gdereuh, it snows...
14:20:30 Quit qwm (
14:20:30 Quit preglow (
14:20:56NJoinqwm [0] (
14:20:56NJoinpreglow [0] (n=thomjoha@rockbox/developer/preglow)
14:22:49JdGordonWhat do we do about the ~3sec freeze while the OF boots when usb plugs in... there is no way the user knws if its frozen because of usb, or because of a crash :(
14:23:01JdGordonenabling the lcd doesnt look so good
14:23:22 Join Nico_P [0] (
14:28:27linuxstbDon't you just learn to wait 3 seconds to see if it's worked?
14:29:24linuxstbAnd why would it crash?
14:29:53JdGordoni dunno... I find it annoying.. but meh :p
14:30:34 Join ackbahr [0] (
14:33:41JdGordonanyone wanna try out my patch?
14:36:00 Quit Ribs ("eh eh ehhhh!")
14:36:12Nico_PJdGordon: what patch is it ?
14:36:29JdGordonfor the sansa bootloader
14:38:54 Join perldiver [0] (
14:38:54 Quit pearldiver (Read error: 104 (Connection reset by peer))
14:40:46 Quit netmasta10bt (Read error: 110 (Connection timed out))
14:44:09 Quit midgey ()
14:46:18 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
14:49:07 Join Thundercloud [0] (n=thunderc@
14:50:05 Quit bonbonthejon (Remote closed the connection)
15:01:00barrywardellJdGordon: back
15:01:50JdGordoncheck your email :)
15:01:55barrywardelljust looking at your patch now ;)
15:02:04JdGordoni was about to goto bed, but ill stick around for a bit longer
15:03:23JdGordonone thing I dont like about this hack is that its a bit annoying if you actually want it to rebuild your database
15:04:27barrywardellIs it necessary to check the usb 5 times?
15:04:40JdGordonyou mean in the loop?
15:04:56barrywardellshouldn't it only need to be done once?
15:05:19barrywardelland what do you think about enabling verbose boot whenever any button is pressed?
15:05:24JdGordonit needs a pause there for some reason, and I didnt know what it was.. checking 3 times failed, and 10 was a bit long
15:05:44barrywardelldoes it work if you just add sleep() instead?
15:05:55JdGordonhow long a sleep tho?
15:06:24barrywardellI'm not sure.
15:06:31linuxstbbarrywardell: I like that idea - there's a similar problem on the ipods, with the apple_os being loaded from either the partition of a file, so it's nice to tell users which one is happening.
15:06:55linuxstb^partition OR a file
15:06:58 Join Redbreva [0] (
15:07:40barrywardellJdGordon: I think the usb_detect() code needs some changes. ideally we should only have to do it once
15:07:42JdGordonbarrywardell: I did the loop so as soon as its ready it will go... although, half a second isnt going to make much difference realy is it?
15:08:25barrywardellJdGordon: I guess I could handle half a second. It just seems like there's a better way
15:08:39*barrywardell wishes he could get a working usb driver
15:11:19barrywardellJdGordon: so disabling database rebuild only happens when USB is connected?
15:11:29barrywardellsounds reasonable to me
15:11:35JdGordonlinuxstb's idea
15:12:19linuxstbDoes Rockbox detect USB on the Sansa and reboot to the OF?
15:12:21JdGordononly if the checksum passes tho which is irritating
15:12:27JdGordonnot in svn
15:13:50barrywardellJdGordon: there are also different checksums for the encrypted mi4's
15:14:21JdGordonso we need to get a list of them.. unless there is a better way?
15:15:17 Join Siku [0] (
15:15:36JdGordonusb isnt detected even with a HZ/2 sleep :(
15:15:40barrywardellmaybe there is a version in the PPPS header
15:16:27barrywardellJdGordon: it probably needs to run twice, once before and once after. It's because of the way usb_detect() is done
15:16:38 Join lee-qid [0] (
15:18:53barrywardellJdGordon: yep, looks like you can detect the firmware from the PPPS header :)
15:19:28barrywardellaround 0x780200 or so
15:20:17JdGordon00 02 ff 3f in mine
15:20:45 Join B4gd3r [0] (
15:20:57JdGordon00 02 ff 3f in yours :p
15:20:59JdGordonno good
15:21:23*JdGordon idiot
15:21:30JdGordonin the string shortly after that header :p
15:21:39barrywardellyeah :)
15:21:49 Quit B4gder (Nick collision from services.)
15:21:54 Nick B4gd3r is now known as B4gder (
15:22:23JdGordonfiddling with usb_detect and sleeps I havnt got it working once...
15:22:29JdGordonso I think the loop is ok
15:23:41barrywardelli guess it's ok until we have a better usb_detect()
15:24:14barrywardellhmmm... if there's an error, lcd_init() never gets run
15:26:52JdGordonwe probably need to run lcd_init on startup and lcd_update with either a all black screen, or the rb logo....
15:28:00barrywardellyeah, or even better figure out why lcd_init is updating the screen
15:28:37JdGordonbecause it calls lcd_clear_display() ?
15:29:43JdGordonanywho... im going to bed.. cya
15:29:52 Quit JdGordon ("Konversation terminated!")
15:34:06 Part LinusN
15:34:46 Nick joshin_ is now known as joshin (n=joshin@unaffiliated/joshin)
15:38:58 Quit wrobbie (Read error: 110 (Connection timed out))
15:41:55 Join perl|work [0] (
15:42:00 Quit PaulJam (".")
15:42:02 Quit kclaf (Remote closed the connection)
15:43:39 Join kclaf [0] (
15:44:38perl|workfeatures Toshiba's new high-quality, 1-bit TC94A82XBG DAC
15:45:11 Quit Seed (Nick collision from services.)
15:45:17 Join Seedy [0] (
15:46:25markunperl|work: what about it?
15:46:33 Quit barrywardell ()
15:49:20perl|workjust wondering how "high quality" it is
15:51:26 Join Domonoky [0] (
15:51:37markunyou doubt it's quality because it only uses 1 bit?
15:53:21 Join toffe82 [0] (
15:54:20perl|workhaha no
15:55:35markunok :)
15:58:24 Join pondlife [0] (
16:00:01amiconnmarkun: Wanna test a patch that further simplifies arm libspeex asm constraints?
16:00:35amiconnCode *should* be identical, but I can't test atm - no testfiles
16:01:29markunamiconn: ok, I can test it
16:02:26 Join barrywardell [0] (
16:03:00***Saving seen data "./dancer.seen"">
16:18:47 Quit atsea- (Read error: 104 (Connection reset by peer))
16:21:18 Join netmasta10bt [0] (
16:22:23 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
16:25:58 Part austriancoder
16:26:13jhMikeSnow logf is whining all over the place since the printf checks :\
16:29:32 Join Llorean [0] (n=Llorean@rockbox/administrator/Llorean)
16:30:26jhMikeSstill whining I should say
16:30:47LloreanWould anyone mind if I added something like "When writing a post, try to assume it will be read literally and make sure to say what you mean, as we cannot know when your word mean what the dictionary says they do, and when they mean what you intend them to instead." to the forum posting guidelines?
16:30:59LloreanAnd an equivalent statement about reading posts?
16:32:34barrywardelljhMikeS: are you talking about my commit earlier?
16:36:01 Quit linuxstb (Read error: 110 (Connection timed out))
16:37:09jhMikeSbarrywardell: there's still some in places that give warnings: playback.c, aac.c; I didn't do a full rebuild so don't know of others yet
16:39:06barrywardelljhMikeS: Ah, I meant was it my commit earlier that caused the warnings. I don't think it was
16:39:34Lloreanbarrywardell: Do we have a documented official dual boot process for the H10 somewhere?
16:39:36jhMikeSthe precision stuff, no. Some were fixed
16:40:45 Join web-taz [0] (
16:40:49barrywardellIriverH10InstallationBeta, although that looks like it was copied from the manual
16:40:50LloreanOr is it a case of "RBUtil should handle it soon" or even "The bootloader should be able to load an encrypted original firmware soon" so the mi4code process isn't really that important if people will just wait?
16:41:12 Join TheSphinX^ [0] (
16:41:30Lloreanbarrywardell: Man, I overlooked it in the manual. Sigh.
16:41:56LloreanSorry for the interruption
16:42:47barrywardellLlorean: it should be very easy soon. hopefully it will eventually be: rename /System/H10.mi4 to /System/OF.mi4. download H10.mi4 from and put it in /System
16:43:47Domonokybarrywardell, that would be good, and also easy to support with rbutil :-)
16:44:37Lloreanbarrywardell: So, it's "hopefully soon the bootloader can decrypt" :)
16:45:25barrywardellyeah, exactly. hopefully very soon
16:46:54 Quit B4gder (Read error: 110 (Connection timed out))
16:47:35 Quit ackbahr ("Chatzilla 0.9.75 [Firefox]")
16:48:42 Quit barrywardell ()
16:55:09 Quit perplexity (Read error: 110 (Connection timed out))
16:55:15 Join barrywardell [0] (
16:58:36 Quit web-taz (Read error: 104 (Connection reset by peer))
16:58:51toffe82I have one question about the dircache on rockbox, I validate it and in the manual it says : "Thereafter, the directory cache will work in the background"but each time I switch on and off, I have to wait that it finish scanning the disk ?also when I disconnect from the usb
16:58:52toffe82is this normal ?
16:59:09 Join web-taz [0] (
17:00:05Lloreantoffe82: Yes. Most of the work is in the background, but at boot it can't know if you used USB outside of Rockbox (booting the original firmware for example), and in USB, since it's not software controlled, it can't know what files you put on until it scans.
17:00:41markunamiconn: sorry, can't test your patch right now. Here are some sample files if you want:
17:01:16 Join nls [0] (
17:04:20toffe82Llorean: ok, but it is long to boot when dircache is enable. Wouldn't be good to have a dircache on boot option ? to have a faster boot if you never use the original firmware
17:06:54toffe82and just let the dircach on disconnect usb
17:07:11linuxstb_toffe82: It works like that if Rockbox has replaced the original firmware in flash - motivation for you.... :)
17:07:36 Join midgey [0] (
17:08:23amiconnDircahce updates in the background anyway except on first boot after enabling
17:08:55toffe82linuxstb_: ;)
17:09:49 Quit pondlife (Read error: 104 (Connection reset by peer))
17:10:11 Join cynicalliberal [0] (
17:10:17Lloreantoffe82: Is it doing a full, slow scan every boot?
17:10:35toffe82Llorean: yes
17:10:39LloreanOh, that's not right.
17:11:06toffe82I don't know if it is full but it take at least 5 seconds
17:11:21toffe82on the gigabeat F40
17:11:32toffe82perhaps more
17:11:42LloreanI have an F40 too, and don't get any delay in booting.
17:11:47GodEatertoffe82: that's definitely not right - I have it switched on on my F40 too - and have no such symptoms
17:12:07nlsamiconn: would doing just a lcd_bitmap_transparent_part be noticably faster than doing "set drawmode, lcd_bitmap_part, set drawmode, lcd_bitmap_mono" ? with the transparent bitmap equal size in pixels as the mono but slightly larger than the native used in ordinary bitmap_part?
17:12:07GodEatertoffe82: which build are you on ?
17:12:13toffe82you don't have a screen saying scanning disk at the boot ?
17:12:17Lloreantoffe82: No.
17:12:34Lloreantoffe82: Have you checked in the debug menu to make sure your dircache is actually active and healthy?
17:12:52toffe82I have to check
17:13:27amiconnnls: ??
17:13:34nlstoffe82: there's a bug that will (or used to?) cause that if there were a very large amount of files on the disk
17:13:47nlsamiconn like the drawing in the bubbles plugin
17:13:54toffe82I don't think I have too much files
17:14:10GodEaterI've got 7.5K files on mine - seems to work fine
17:14:28nlsfirst a color bubble is drawn with drmode_solid, then a ring around it as a mono bitmap with drmode_fg
17:14:34GodEaterit takes about 25 seconds to do the scan - but it's *definitely* happening in the background, you can only see it if you go into the debug menu
17:15:58nlsamiconn: I wonder if combining the two bitmaps into one and only doing one draw as well as removing the need to change drawmode will be faster if a transparent bitmap is used
17:17:01toffe82GodEater: correction : I have a lot of files (83000), I use it as an exteranl and Rockbox is only 2720 files
17:17:22Llorean toffe82: Well, dircache is *all* files, not just music/rockbox files.
17:17:25GodEater83,000 is quite a few file I'd say :)
17:18:09*linuxstb_ wonders if toffe82 has any audio buffer left
17:18:31toffe82linuxstb_: no pb with the audio
17:18:32 Quit web-taz ("( :: NoNameScript 4.02 :: )")
17:18:32XavierGrwith that ammount of files probably the dircache file would be insanely big
17:18:49LloreanXavierGr: Or simply failing.
17:18:51linuxstb_toffe82: If you go into the Rockbox info screen, how big is your audio buffer?
17:19:01nlslinuxstb_: I think that's the problem, the dircashe is limited to about 6 megs and when the cashe doesn't fit, it will scan on every boot
17:19:12toffe82I have to check, cannot do it now
17:19:30amiconnnls: Ah, that one. I'm not so sure lcd_transparent_bitmap will be faster, because lcd_bitmap(_part) can use optimisations which the transparent version cannot do
17:20:19toffe82so why don't we organize rockbox in folders : one for rockbox, one for music and one for video so we don't check outside thes folder
17:20:23amiconnIt would be easier to read, but since greyscale targets don't have lcd_transparent_bitmap (yet), it would mean more ifdefing
17:20:23nlsamiconn: ah, ok, I will not change it anyway because the greyscale targets don't seem to be able to draw native bitmaps with transparency...
17:20:52 Part Llorean
17:20:57amiconnSame goes for mono targets. This will change soon(ish) though
17:21:03linuxstb_toffe82: You could repartition...
17:21:24GodEateryeah, re-partitioning is a much better idea
17:21:26toffe82linuxstb_: rockbox see only one partition ?
17:21:35GodEaterotherwise we'd have to re-educate an awful lot of users
17:21:38amiconnYes, unless you enable multivolume support
17:21:47nlsamiconn: but mono tagets can draw mono bitmaps with white as transparent, can't they?
17:22:07 Join atsea- [0] (i=ariel@gateway/tor/x-31be85636afeaac1)
17:22:29amiconnnls: Yes, but that's still different from real transparency (and the reason why the animation in jewels doesn't look as good as it could
17:22:50nlsamiconn: I'll leave it alone then :-)
17:25:26toffe82survey : how is organized your hd ?
17:26:12GodEatertoffe82: .rockbox/, Music/, Videos/
17:26:27GodEaterMusic is broken down into Artist, then Album, then Tracks
17:26:29toffe82mine too
17:26:42toffe82the same
17:26:58toffe82is this a general setting for everybody ?
17:27:00GodEatergreat minds think alike it seems ;)
17:27:13GodEatertoffe82: very much not the case on iPods I would imagine
17:27:24GodEatera lot of our users continue to use iTunes
17:29:12GodEatertoffe82: also, for some reason, we have some users who think Music should go under .rockbox. Not our recommended way, but they still do it
17:29:28*amiconn doesn't use Music/
17:29:51GodEateramiconn: you have no video capable target though do you ?
17:30:09GodEateramiconn: I don't use that folder stucture on my H140 either - all artists are off root
17:30:56amiconnI do have video capable targets
17:31:09amiconn..and for special uses like video I do use subfolders with appropriate names
17:31:13GodEatermy mistake then
17:31:18*linuxstb_ has /Music, /Comedy, /Radio, /Videos, /Bootlegs etc
17:31:35amiconnBut musical genres go in the root, as music is the main purpose of a dap
17:31:58GodEaterlinuxstb_: interesting fetish ;)
17:33:29toffe82am I the only one to use his player as an external hd ?
17:34:08toffe82how many files do you have ? do you use dircache ?
17:34:08amiconnI do use (some) players as external hdd, but I have the files in a hidden folder
17:34:09perl|worktoffe82 it gets warm like crazy though
17:34:16amiconnNo need to show these files in rockbox
17:34:35*TheSphinX^ organized /Movies /Music/<Genre> /Data
17:34:36amiconnA few dozen files at max. Mainly installation crapped, zipped or self-extracting installers
17:34:43perl|workwhen usb connected and hdd spinning
17:34:44 Quit netmasta10bt (Read error: 60 (Operation timed out))
17:34:54amiconnAnd I don't use dircache
17:35:24amiconnTotal files count is usually less than 4000
17:35:39TheSphinX^toffe82 i'm using it as external hdd 11000 files for music and 12000 files for work
17:36:26 Join oKtosiTe [0] (n=oKtosiTe@unaffiliated/oKtosiTe)
17:38:43oKtosiTeI have an iAudio X5, and am very happy with rockox sofar, but I was wondering if rockbox allows me to use the USB host feature, or if not, if it's a planned feature.
17:39:20TheSphinX^USB OTG isn't yet implemented, but you can use an dualbootloader
17:39:37oKtosiTeTheSphinX^: I can?
17:39:51GodEateroKtosiTe: it is a planned feature, it's on our list of ideas for Google Summer Of Code in fact
17:40:25toffe82I have pratically all my programs runnig from the player, email, internet, grahics, office... so I don't have to open my laptop when I am at home, just connect the player ..
17:40:41toffe82to my home computer
17:41:16GodEaterperhaps we should include some way of telling dircache to skip a given directory hierarchy ?
17:41:56toffe82or perhaps I should use my player as a player ;)
17:42:07GodEaterwell, you could consider that at least ;)
17:42:16oKtosiTeTheSphinX^: Where can I find this dualbootloader? It would be ever so useful, albeit only for the USB host feature...
17:42:44TheSphinX^1 moment
17:42:55toffe82it is only to don't have two device doing the same thing in my pocket ;)
17:43:11 Join atsea-142 [0] (i=ariel@gateway/tor/x-98c96bc2ca6a58a3)
17:43:49 Quit midgey ()
17:44:46toffe82perl|work: when it is on usb it doesn't get so warm, but when it is in the cradle, yes it is
17:49:09perl|workreally? quite opposite for me
17:50:45 Quit linuxstb_ (Read error: 113 (No route to host))
17:52:57 Join mattzz [0] (
17:53:35oKtosiTeTheSphinX^: What I just read on the mailinglists doesn't make it sound too attractive...
17:54:35*TheSphinX^ actually my installed version of dualboot works fine for me, the only thing is that the OF everytime have to recreate it's dircache
17:54:40 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
17:54:45*TheSphinX^ oKtosiTe
17:55:23oKtosiTeTheSphinX^: Every time you switch or everytime it loads?
17:56:10oKtosiTeTheSphinX^: The former would be quite acceptable to me...
17:56:18*TheSphinX^ oKtosiTe: everytime i runned rockbox before (or used usb, like without rockbox)
18:01:41*nls wonders what's going on with the bin sizes in the current build page.
18:03:01***Saving seen data "./dancer.seen"
18:04:16 Join My_Sic [0] (
18:04:50oKtosiTeI was kind of hoping to find a binary dual-bootloader, without all the hassle...
18:04:51*TheSphinX^ oKtosiTe tested?
18:05:10TheSphinX^short power rockbox, long OF
18:07:54oKtosiTeTheSphinX^: I haven't found the dual-bootloader yet...
18:08:20*TheSphinX^ oKtosiTe
18:09:28 Quit SirFunk (Remote closed the connection)
18:10:09oKtosiTeAwesome, I'll go get my cable upstairs. One more thing, to upgrade daily build, should I just unpack and replace? I have installed rockdoom, customized a lot of settings and wouldn't like to do that all again after every update...
18:11:20*TheSphinX^ oKtosiTe it only replaces the bootloader everything else stays as it is, you can use every build you like
18:11:22nlsoKtosiTe: an update doesn't overwrite your settings, gamefiles or other custom stuff, but if you changed your tagnavi.config file rename it
18:11:36 Quit Arathis ("Bye, bye")
18:12:13 Join Arathis [0] (
18:12:37 Quit FOAD ("I'll be back")
18:12:45oKtosiTenls: what about files that possibly don't exist in a new version? They may pile up after a while...
18:12:50oKtosiTeok brb
18:13:01 Quit lini (Read error: 110 (Connection timed out))
18:13:25nlsoKtosiTe: yeah, but that is another thing, and it doesn't change that much anyway
18:13:54andrewg867would anyone happen to have austriancoder's email address?
18:13:55 Quit My_Sic (Read error: 104 (Connection reset by peer))
18:14:31linuxstbandrewg867: You could probably find it through the wiki
18:17:04 Join lini [0] (i=pugsley@
18:19:25 Join FOAD [0] (
18:22:11 Join webguest83 [0] (i=18f44eb5@gateway/web/cgi-irc/
18:22:26oKtosiTeD'oh, forgot to bring the power adapter, have to go back upstairs to try this bootloader...
18:22:42webguest83how do I set a variable to the current directory on a shell script?
18:23:21webguest83so I have rockbox=/mnt/other/rockbox
18:23:52webguest83but how do I make it rockbox=(current dir)
18:24:48amiconnset rockbox=`pwd` ? just a guess
18:25:13webguest83isn't pwd short for password?
18:25:26amiconnPrint Working Directory
18:26:00amiconnSorry export, not set
18:26:20bluebrotherexport variable=value
18:26:41linuxstbYou only need to export it if you want child shells to see it.
18:26:46bluebrotherbut only if oyu want to export it
18:26:52*bluebrother is too slow
18:27:06webguest83I use the variable farther in the script
18:27:21linuxstbThen just "rockbox=`pwd`" should be enough.
18:27:24bluebrotheras long as its the same script there is no need to export it
18:28:01mattzzanybody with a fully charged sansa online?
18:28:44webguest83didn't work
18:29:16webguest83should be 'pwd' or "pwd" ?
18:29:41linuxstbThey need to be backticks
18:29:59webguest83oh i see
18:30:09webguest83lol thought it was a quote mark
18:30:22linuxstbIt means that the pwd command is executed, and the result of running that command is put in the variable
18:30:41bluebrotherbut in you case using "." could be sufficient too.
18:30:44 Quit mattzz (Read error: 104 (Connection reset by peer))
18:30:54bluebrotheras long as your script doesn't change directories
18:31:32mentbtw is ipod remote supported in devel branch?
18:32:24menti can't find it anywhere in the sources
18:32:50 Join mattzz [0] (
18:33:08linuxstbment: There are no devel branches in Rockbox.
18:36:03 Quit Redbreva ("Chatzilla 0.9.77 [Firefox]")
18:39:03 Join Llorean [0] (n=Llorean@rockbox/administrator/Llorean)
18:39:18mattzzno sansa users online?
18:39:29Lloreanmattzz: Did you have a specific question?
18:40:15mattzzLlorean: yeah, I am looking for the battery voltage of a fully charge sansa as I am currently fixing powermgmt.c
18:40:15 Quit webguest83 ("CGI:IRC (EOF)")
18:40:51mattzzbatterybench did not work
18:41:01LloreanIt didn't work?
18:41:03LloreanEmpty log?
18:41:25andrewg867mattzz: I have a E260
18:41:49mattzzone is emtpy and another try results in entries: −−File already present. Resuming Benchmark−−
18:41:52mattzzbut no numbers
18:41:59mattzzandrewg867: is it fully charged?
18:42:11andrewg867its almost dead
18:42:15 Join xandl12 [0] (i=3eb244f7@gateway/web/cgi-irc/
18:42:23andrewg867I can charge it if you will be around later
18:42:25mattzzso is mine... chaaaaarging...
18:42:25Lloreanmattzz: It probably needs to be fixed the same way the iPod Nano was.
18:42:49Lloreanmattzz: Once yours is charged you can find the current voltage in the Debug screen under the Battery Info (or was it View Battery) option.
18:43:01mattzzandrewg867: sure, that would be nice. I expect the e260 to have the same battery (3,7/750)
18:43:03xandl12I have a full charged sansa ... what can I do?
18:43:13amiconnmattzz: Then the sansa disk driver doesn't handle the ata callback properly
18:43:26amiconn(and/or 'disk spinning' emulation)
18:43:40mattzzamiconn: yeah, seems like that
18:43:48amiconnThat should be fixed then
18:43:59mattzzxandl12: check out settings/debug/battery
18:44:33mattzzfor the archive: shutdown voltage is 2,4V (surprise)
18:44:42XavierGrthat low?
18:44:58mattzzbut the bootloader handles that
18:45:01XavierGrhow many hours this thing lasts?
18:45:22andrewg867my battery is currently at 3.96v
18:45:25mattzzhard to tell. I put it in a drawer (with the backlight on)
18:45:28LloreanWas the advertised 16 in the original firmware?
18:45:51Lloreanmattzz: Backlight on to drain the power more quickly intentionally? Backlights can make a HUGE difference on some targets.
18:46:13Lloreanmattzz: My battery's at 4.22, but it's not quite full.
18:46:20andrewg867I cant exit out of the View Battery menu
18:46:26xandl12ok 4,16-4,29V
18:47:26mattzzthere are two readings in the debugscreen - one on the page with the chart and one on the other page
18:47:56mattzzI think one is the battery voltage and the other one behind the buck up converter, right?
18:48:07andrewg867the other page has the charger indication right?
18:48:22mattzzandrewg867: IIRC yes
18:48:46preglowhmm, think i like the wps preprocessor plugin idea a bit :>
18:48:50mattzzLlorean: you would not recommend backlight for faster power draining?
18:48:54andrewg867the graph one for me says 3.95v, and the other is 3.92v
18:50:12mattzzI had 2,4V on the one screen and 3,1V on the graphical screen just befor low-power-shutdown
18:50:42mattzzxandl12: thanks for the numbers
18:51:14barrywardellthe I/O ports debug screen gives the actual value read from the adc
18:51:20amiconn2,4V sounds really low for Li batteries
18:51:32amiconnIs that LiIon or LiPo?
18:51:39 Join My_Sic [0] (
18:51:52 Join bluey- [0] (
18:52:33amiconnSeems somewhat unusual today
18:52:34mattzzThe toshiba gigabeat has 3,4V for low power shutdown
18:52:43amiconnThe archos fm/v2 had LiIon
18:52:48barrywardellthe value from the adc is the reading passed through a 1/5 voltage and scaled by 2.5
18:52:55amiconnLiIon can obviously go lower than LiPo
18:53:12barrywardellLiIon is 3.7V when fully charged
18:53:24markunI thought 4.2
18:53:33amiconnH1x0, H300 X5, M5 and ipods can only go down to 3.3 ... 3.5V
18:53:34barrywardellLiPoly is 4.2
18:53:38 Join OgMaciel [0] (n=omaciel@ubuntu/member/gnukemist)
18:53:41mattzzThere are 4.1 and 4.2V charging types
18:53:44amiconnLiIon should be 4.1V fully charged
18:53:58amiconnLiPo is usually 4.2V
18:54:01OgMacielcan someone help me with restoring bootpartition.bin?
18:54:07OgMacielI received Input file too big for buffer
18:54:14amiconnNominal voltage (for both) is 3.6V or 3.7V
18:54:40markuncan you charge both batteries with the same charger?
18:55:12barrywardellthe sansa battery is labelled as 3.7V. maybe that is referring to the nominal volatge
18:55:19amiconnit is
18:55:22LloreanOgMaciel: Our uninstall instructions don't involve bootpartition.bin. What are you trying to do?
18:55:56mattzzFrom the numbers xandl12 gave, I would think it is a 4,2V type
18:56:04OgMacielLlorean: I backed up the boot partition with ipodpatcher
18:56:14OgMacielLlorean: and am now trying to put it back
18:56:20mattzzamiconn: Are you sure that it is always 4,1V for LiIon?
18:56:29 Join desowin [0] (
18:56:38OgMacielLlorean: my ipod seems to have a bad block somewhere
18:57:16 Quit TheSphinX^ ("Miranda IM! Smaller, Faster, Easier.")
18:57:37mattzzbarrywardell: I didn't get the ADC scaling thing. 1/5 scaled by 2.5?
18:58:01LloreanOgMaciel: Are you using the same version of iPodPatcher that you backed it up with? You haven't downloaded a newer one since then?
18:58:36OgMacielLlorean: I actually tried both approaches
18:58:41OgMacielLlorean: original and new
18:58:51barrywardellmattzz: the battery source is fed into a 1/5 voltage divider and then amplified by 2.5
18:59:21LloreanOgMaciel: What do you mean by "original?" What version number?
18:59:40barrywardellso I think the values read by the adc (eg. in the i/o ports debug screen) are 2.5/5 * actual voltage
19:00:03OgMacielLlorean: I meant the original version of the ipodpatcher I had used when creating the partition
19:00:06mattzzbarrywardell: which is pretty much /2
19:00:09LloreanOgMaciel: I've never had a problem with writing a bootpartition.bin back to my Nano, though I'll admit the very single latest version hasn't been tested by me.
19:00:19OgMacielLlorean: but I believe I ran a newer version by mistake
19:00:30*OgMaciel checks
19:00:32barrywardellmattzz: yep
19:00:38linuxstb OgMaciel: What's the actual problem you're trying to solve?
19:01:03OgMaciellinuxstb: just trying to reset my ipod back to the "factory" settings
19:01:12linuxstbWhat operating system are you using?
19:01:28linuxstbAnd what did you do to necessitate a restore?
19:01:40OgMaciellinuxstb: foresight linux
19:01:53OgMaciellinuxstb: I wanted to remove rockbox for the time being
19:02:16linuxstbThen just running ipodpatcher and selecting "u" for uninstall would do it.
19:02:30mattzzbarrywardell: so ADC_BVDD reading 0x3f means what?
19:02:39OgMaciellinuxstb: I did remove the rockbox loader... that was not the problem
19:02:56OgMaciellinuxstb: I wanted to restore the backed up boot partition
19:03:14OgMaciellinuxstb: whih I just successfully did now ;)
19:04:12*mattzz thinks debug_menu.c has to be fixed
19:06:45barrywardellmattzz: 0x3f would be 0.31 volts
19:06:45barrywardellthat doesn't sound right though
19:06:48*amiconn compiles a H300 build without disk poweroff in order to verify his theory about ata_sleep
19:07:04mattzzbarrywardell: sorry I meant 0x3ff
19:07:29barrywardellthat's what I was thinking. 0x3ff=5.12V
19:07:47xandl12i opened up my sansa and on the battery "Li-ion Battery 3,7V" ...if that helps :-)
19:08:14barrywardellthe adc reading goes in 5mV steps from 0-5.12V
19:09:04mattzzbarrywardell: I we say that the charging circuit stop charging when 4,2V are reached
19:09:30 Join obo [0] (
19:11:00 Quit preglow (
19:11:00 Quit qwm (
19:11:14NJoinqwm [0] (
19:11:14NJoinpreglow [0] (n=thomjoha@rockbox/developer/preglow)
19:12:03mattzzbarrywardell: that would give us a scaling factor of 1219
19:12:21 Nick datachild is now known as krebb (
19:12:22 Join nathan [0] (
19:12:22 Quit FOAD ("I'll be back")
19:12:26 Nick krebb is now known as datachild (
19:12:59XavierGramiconn: will you measure amperage or just do a battery_bench with it to verify your theory?
19:13:18amiconn"just" battery_bench
19:13:20nathananyone here does compiling for the H10?
19:13:58amiconnI don't want to solder something in my H340
19:15:17nathanno one?
19:15:27 Quit zylche ("-")
19:15:38 Quit nathan (Client Quit)
19:15:57 Join zylche [0] (
19:16:17 Join nathan [0] (
19:16:32barrywardellnathan: i do
19:16:54nathanI made a shell script
19:17:11nathanto compile the builds and/or bootloaders
19:17:19nathanwant to try it?
19:17:32barrywardellyeah, sure
19:17:42nathank its on my website here
19:17:54nathanat the top
19:18:20nathantell me if it is good or not
19:18:37nathanmaybe it would be somthing that we could impliment into all the builds
19:18:50nathancertainly easier for noobs
19:21:13nathanyou have to make it executable
19:21:17nathanby the way
19:21:51 Join MonkeyTamer [0] (
19:23:17 Join FOAD [0] (
19:23:34 Quit My_Sic (Read error: 110 (Connection timed out))
19:23:51Lloreannathan: I think that at the point where you actually have reason to compile a build, the current process of using configure is still pretty simple.
19:24:26nathantrue but why not make it simpler
19:24:42nathanand I aim to make it so that you can cue up multiple builds
19:24:49LloreanHaving a whole bunch of shell scripts for every possible combination of configure seems more complicated to me.
19:25:01nathanwould just be one script
19:25:08LloreanThen... why not improve configure?
19:25:35nathanbecause I've only started to code
19:25:52nathanthat was actually my first script/code/anything
19:26:44 Join My_Sic [0] (
19:27:56LloreanThe combination of configure, and then make, is somewhat expected when creating a build of something.
19:28:01barrywardellit would be handy to have a script to build all the targets and check for compile errors warnings before doing a commit
19:28:33Lloreanbarrywardell: The idea of creating an "all," or other form of commit test, build target has been bandied around, I think.
19:29:00 Join Lear [0] (i=chatzill@rockbox/developer/lear)
19:29:28nathannot bad for my first ever code
19:29:48nlsdoing a build all in cygwin would take about a week to finnish tho :-)
19:29:55barrywardellnathan: it would be helpful for automating building for a load of targets
19:30:03 Nick datachild is now known as krebb (
19:30:19nathanYa i was thinking about making the ability to que up the builds
19:30:45nathanthen just hit 4,2 6,3 or w/e and then hit build
19:30:53 Quit My_Sic (Read error: 104 (Connection reset by peer))
19:30:56nathanand then it will work away for a while
19:31:03nathansaves having to sit there
19:31:34Lloreannls: Nah, even at 15 minutes per build, it'd only take about 15 hours. :-P
19:31:39 Quit nathan (Nick collision from services.)
19:31:45Lloreannls: Well, for a commit test.
19:32:12 Join nathan [0] (
19:32:30nathandid I miss somthing?
19:32:59nlsLlorean: but you need to make all the sims and bootloaders too, to be safe :-)
19:33:52Lloreannls: That's what my commit test number included, all the columns in the build table, under the assumption of a average time of 15 minutes per build (which is much too high, I think)
19:34:26nlsah, ok that isn't too much then :-P
19:34:54LloreanStill, a realistic time on a commit test is still "too long"
19:34:54linuxstbI think I would rather just commit and suffer the yellow/red...
19:35:16nlslinuxstb: I agree :-)
19:35:26LloreanYeah, 7 minutes of compiling, 5 minutes of red while fixing, another 7 minutes of compiling, is still far less time.
19:35:47LloreanAssuming it's a quick fix / stupid mistake style red.
19:36:20amiconnThis hang-in-usb-screen on ipod gets annoying
19:36:32 Quit nathan (Nick collision from services.)
19:36:42linuxstbHopefully barrywardell's theory was right, and that would fix that.
19:37:08 Join yyyggg [0] (
19:37:20yyygggman I keep getting booted
19:37:30yyygggfirst nathan is taken
19:37:37yyygggthen It isn't now it is
19:37:43yyygggnow i'm yyyggg
19:37:49 Quit jaebird (Read error: 110 (Connection timed out))
19:38:19yyyggganyways so barry wardel? should I continue to work on a que for that script?
19:38:50 Join jaebird [0] (
19:40:12 Join raphi [0] (
19:40:13barrywardellamiconn: did you see my theory about usb problems?
19:40:33yyygggbarrywardell: you didn't answer my Q
19:40:42yyygggabout the shell script I wrote
19:40:50barrywardellyyyggg: i think work on it if it useful to you
19:41:03 Join My_Sic [0] (
19:41:13barrywardellamiconn: we need to set UDC_ENDPOINTLISTADDR to a region of memory that isn't in use
19:41:23yyyggggood way to learn the code anywyas
19:41:32barrywardelland maybe even put a vaild data structure there
19:41:35 Quit yyyggg (Client Quit)
19:41:50amiconnbarrywardell: Hmm, in fact I saw that one, but didn't really understand it
19:41:58amiconnWhy does it only happen sometimes?
19:42:02 Part raphi
19:42:43barrywardelli'm not sure, why it's random, but UDC_ENDPOINTLISTADDR gets reset to 0x0 when usb is connected
19:42:57 Join jhulst [0] (
19:43:09barrywardellso it's possibly overwriting some data around the start of DRAM
19:43:27amiconnYes, but what does that do? Is the usb controller able to write directly into memory?
19:44:52barrywardelli guess so
19:45:43barrywardellwhen you connect the usb it will receive a setup packet from the computer
19:46:49 Join Seed [0] (
19:47:15 Join Thundercloud_ [0] (n=thunderc@
19:47:19barrywardellI'm fairly sure something is happening when usb is connected that makes it corrupt part of DRAM.
19:49:50 Quit MonkeyTamer ("Chatzilla 0.9.77 [Firefox]")
19:51:23 Join Thundercloud__ [0] (n=thunderc@
19:51:26 Quit FOAD ("Lost terminal")
19:51:30mattzzhm, button_STD_CANCEL is not in button_context_settings for sansa.
19:52:24 Quit atsea-142 (Read error: 104 (Connection reset by peer))
19:52:24 Quit atsea- (Read error: 104 (Connection reset by peer))
19:52:59 Quit My_Sic (Read error: 104 (Connection reset by peer))
19:53:43 Quit xandl12 ("CGI:IRC")
19:54:40 Join JavaMan22 [0] (
19:54:56 Join atsea-142 [0] (i=ariel@gateway/tor/x-a4f51a26b4dfe0e8)
19:55:04 Join FOAD [0] (
19:55:51 Join My_Sic [0] (
19:55:54 Join webguest28 [0] (i=54382824@gateway/web/cgi-irc/
19:56:20JavaMan22svn up svn:// rockbox <<<< is that how to update my rockbox files?
19:56:35 Join atsea- [0] (i=ariel@gateway/tor/x-393f12450f80a5f6)
19:56:39 Quit Domonoky (Read error: 54 (Connection reset by peer))
19:56:49barrywardellyou can just run svn up when you're in the rockbox directory
19:57:12 Quit Thundercloud (Read error: 60 (Operation timed out))
19:57:39 Join SirFunk [0] (
19:57:39JavaMan22U means update what does A mean
19:58:17JavaMan22it doesnt show the files that stay the same right?
19:58:33linuxstbbarrywardell: What I don't understand either about your USB theory is that if you write 0x0 there, then it should overwrite the exception vectors immediately and crash Rockbox. But it doesn't seem to happen every time, and seems to have got worse since that code was first committed.
19:59:21amiconnlinuxstb: I wouldn't expect the usb controller to always write there, only if it detects a connection
20:00:04 Quit FOAD ("I'll be back")
20:00:06amiconnToo bad we don't have a debug interface
20:00:31barrywardelllinuxstb: it only ever writes things if the run/stop bit is set to 1
20:00:44 Join Thundercloud [0] (n=thunderc@
20:00:56 Quit Thundercloud_ (Read error: 60 (Operation timed out))
20:01:25linuxstbamiconn: I'm pretty sure a lot of the ipods have a JTAG interface... Not sure about your mini though.
20:01:38barrywardelland UDC_ENDPOINTLISTADDR is just the queue head of a linked list of transfer descriptors
20:02:03barrywardellthe H10 has a JTAG interface, so the ipods probably do too
20:02:23JavaMan22if i use a bunch of printf("Hello\n");
20:02:32JavaMan22will the screen update?
20:02:37JavaMan22when i reach the bottom
20:02:39 Join FOAD [0] (
20:02:54amiconnMuch #if 0 in arcotg_udc.c ...
20:03:03***Saving seen data "./dancer.seen"
20:03:19barrywardellit's based off the linux driver
20:03:21linuxstbJavaMan22: Rockbox doesn't have printf, unless you're talking about the bootloaders.
20:03:26 Quit Seedy (Read error: 110 (Connection timed out))
20:03:56 Nick krebb is now known as datachild (
20:04:15amiconnSeems the screen capture loop is back
20:04:35amiconnMy mini just finished playing, and now splashes "Nothing to resume" over and over
20:04:47linuxstbJavaMan22: Welcome to the world of embedded programming...
20:05:23 Quit My_Sic (Read error: 104 (Connection reset by peer))
20:05:46JavaMan22embedded let me look that up
20:07:03 Quit Thundercloud__ (Read error: 60 (Operation timed out))
20:07:11amiconnHmm, now it worked.... strange
20:10:09mattzzbarrywardell: I have a question regarding keymap-e200.c - Which keymap context should be used in debug_menu.c? There is a nice mixture an some menus do not operate well in sansa
20:11:01mattzzbarrywardell: shouldnt ACTION_STD_CANCEL be BUTTON_LEFT in context settings?
20:11:06 Quit FOAD ("I'll be back")
20:11:40barrywardellmattzz: that is a common problem with a lot of targets.nobody has decided on the best way to fix it
20:12:55mattzzbarrywardell: ok, I see.
20:15:02 Join FOAD [0] (
20:16:14JavaMan22is there something wrong with this : rb->splash(HZ*3, true, "Movement Test 1.0");
20:16:16mattzzbarrywardell: do you know the difference between ACTION_STD_PREV and ACTION_SETTINGS_DEC in the context settings?
20:16:20 Join NickG [0] (i=551994f2@gateway/web/cgi-irc/
20:16:27mattzzJavaMan22: remove the "true" parameter
20:16:42NickGhi all!
20:16:46NickGhi all!
20:16:47JavaMan22what is the true for
20:16:58barrywardellmattzz: no, sorry. JdGordon is the guy to ask about that sort of thing
20:17:46mattzzbarrywardell: OK, I will stop bugging you then ;-)
20:18:13JavaMan22mattzz what the true for? and how come it compiled before?
20:18:15mattzzJavaMan22: used to be a parameter for center orientation on the screen. Has been removed
20:18:41mattzzJavaMan22: because it has nearly never been used otherwise than "true" ;-)
20:19:26JavaMan22whats wrong with this: if (key == BUTTON_OFF) {
20:19:47JavaMan22it says its undeclared
20:20:25NickGI am going to install RockBox on Kingstone KPEX 100 (it has Telechips TCC 767 core) who can help me to port it?
20:21:35JavaMan22but isnt it already in plugin.h?
20:21:57linuxstbNickG: Only other people with that player can really help - but I'm sure people here can offer advice and encouragement.
20:22:33 Join My_Sic [0] (
20:22:35linuxstbNickG: Have you made any progress yet? I assume you've read the "NewPort" wiki page?
20:23:13 Join petur [0] (n=petur@rockbox/developer/petur)
20:23:42linuxstbJavaMan22: Which player do you have?
20:24:01JavaMan22ipod video 5.5
20:24:26linuxstbThere is no BUTTON_OFF on the ipods...
20:24:57JavaMan22what about the hold switch?
20:25:10bluebrotheramiconn: I had this "nothing to resume" a couple of days ago, both on h120 and mini :(
20:25:14 Quit bluey- ("Leaving")
20:25:26 Join toni1 [0] (i=d9b95efe@gateway/web/cgi-irc/
20:25:28NickGin progress......
20:25:33bluebrotherso it's an old issue?
20:25:38linuxstbThat's not a "BUTTON_" define - you call the button_hold() function to find out if the hold switch is on or off.
20:25:59toni1barrywardell: I think I know the reason for the bootup problem on sansa.
20:26:07linuxstbAnd if the hold switch is on, Rockbox can't read any of the buttons...
20:26:45barrywardelltoni1: great! what is it?
20:26:49toni1barrywardell: If the dma is active during the setting of the lcd base address, it might crash.
20:27:16toni1barrywardell: I modified the bootloader and since then no crash
20:27:19linuxstbNickG: Have you started a wiki page to document your findings?
20:27:20barrywardellah, that makes sense
20:27:39 Join Moos [0] (
20:27:39barrywardellhave you got a patch?
20:27:41 Join chrisjs169 [0] (n=Jack@unaffiliated/chrisjs169)
20:27:52 Quit funky ("leaving")
20:28:00toni1barrywardell: Not yet, but I can put one to the tracker.
20:28:25barrywardellyes, please :)
20:28:52toni1barrywardell: Also the colours change to white on black as the bootloader code implies.
20:29:14NickGI have topic in the forum - (RU)
20:29:36linuxstbNickG: Have you got as far as running your own code on the device?
20:29:58JavaMan22why doesnt rockbox have a contrast setting?
20:30:16bluebrotherit does.
20:30:30Moosamiconn, bluebrother: I encontered this strange bug too, here I didn't can stop playback x5, when I stoped it directly resume, tried a couple of times but anyway to stop playback directly, I had to stop playback in activating FM radio for not shutdown, and this issue happened here few times, last few hours ago
20:30:32JavaMan22where ??
20:30:34pixelmabut not on colour targets
20:30:40barrywardelltoni1: even better! so no more flash of blue first?
20:30:57NickGno - only GNEX ( VM) code
20:31:13bluebrotherMoos: had the unstoppable playback issue a few times on the mini too, but I could stop it from the menu
20:31:15 Quit miepchen^schlaf (Read error: 113 (No route to host))
20:31:18Moosamiconn, bluebrother: IIRC this bug came with the new way to save seeting
20:31:23amiconnpixelma: Not completely correct. *some* colour targets have a contrast setting (e.g. X5)
20:31:31 Join miepchen^schlaf [0] (
20:31:37JavaMan22what about ipod video 5.5
20:31:54Moosbluebrother: did you remenber when that began for you?
20:31:58bluebrotherJavaMan22: grep the source?
20:32:20bluebrotherMoos: I noticed it only a couple of days ago. Mentioned it here but seems nobody noticed ;-)
20:32:28Mooshehe :)
20:32:29amiconnBut contrast doesn't really make sense for colour displays. Setting a non-default contrast just makes the colours look wrong
20:32:48JavaMan22id rather have that than to lose battery quickly
20:33:06bluebrothercontrast doesn't change the battery consumption ...
20:33:06 Quit NickG ("CGI:IRC (EOF)")
20:33:10amiconnLevels are much more stable on colour displays than on mono/greyscale displays. That's why the latter do need a contrast setting
20:33:14LloreanJavaMan22: You're confusing contrast with backlight brightness, they're not the same thing.
20:33:18Moosbluebrother: very annoying bug, and don't know who it occure
20:33:19bluebrotherI guess you mean backlight brightness
20:33:36JavaMan22is there an option for backlight brightness?
20:33:37bluebrotherMoos: indeed.
20:33:50bluebrotherjust turn of the backlight ;-)
20:33:51 Join NickG [0] (i=551994f2@gateway/web/cgi-irc/
20:33:55Moosbluebrother: couldn't be linked to save resue position?
20:34:43bluebrotherMoos: linked to save resume?
20:34:51JavaMan22yeah but i cant see in the dark :(
20:35:34bluebrotheruse a short backlight timeout?
20:35:50desowinbacklight on iPods is too bright even on original firmware IMO
20:36:00bluebrotherI have a quite short timeout and a long fade time.
20:36:01JavaMan22i agree :)
20:36:08Moosbluebrother: s/linked/caused, I don't know how the resume position is saved with the JdGordon's rework
20:36:28bluebrotherresume position should be saved in nvram.bin
20:36:46bluebrotherbut it sounds like amiconn knows the problem?
20:37:12Mooshave you got log time please? :)
20:37:13amiconnI don't
20:37:22JavaMan22there needs to be a screen brightness option :D
20:37:34 Quit My_Sic (Read error: 104 (Connection reset by peer))
20:37:45amiconnMy guess was that it's related to the root menu and its state machine
20:38:14amiconnBut if that bug existed before (didn't experience it with playback until today) this can't be true
20:38:19 Quit NickG (Client Quit)
20:38:55amiconnJavaMan22: Sure, if someone finds out how to control backlight brightness, adding it would be simple
20:39:02bluebrotherhmm. Maybe I should try the last build with the old main menu for a while
20:39:08Moosamiconn: I don't remenber well if I began to encountered this annoying bug after save settings rework or root menu thingy
20:39:23JavaMan22but dont we have it in the fade option
20:39:27JavaMan22or is that contrast
20:39:46amiconnThe fade in/out is software, so it causes extra cpu load and shouldn't be used for permanent dimming
20:40:00JavaMan22oh :io
20:40:17JavaMan22rb->lcd_setfont (FONT_UI); << that returns to your theme font?
20:40:26amiconnAfaik the G5 does have hardware brightness control, but we don't know how to control it
20:42:17desowinJavaMan22: it does what name of function says
20:42:28JavaMan22font_ui is?
20:42:46desowinit sets default font used in your plugin
20:43:26 Join My_Sic [0] (
20:44:28nlsJavaMan22: yes FONT_UI is the theme font, FONT_SYSFIXED is the built in sysfont
20:44:38 Join funky [0] (n=repulse@unaffiliated/funky)
20:44:40JavaMan22oh :)
20:45:35toni1barrywardell: sansa booting patch done:
20:47:29 Part toni1
20:47:51 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
20:47:54JavaMan22i found a bug
20:48:24JavaMan22if you plug in your adapter when you are in a plugin
20:48:33JavaMan22and then you quit the plugin it wont go to disk mode
20:48:49JavaMan22it should check for a adapter after you quit or something :)
20:49:20nlsJavaMan22: No, it should quit the plugin
20:49:30JavaMan22well it didnt
20:49:38JavaMan22not for my plugin hehe
20:50:48nlsthen you have a bug, check for handling of SYS_USB_CONNECTED in other plugins
20:51:10JavaMan22oh :)
20:52:31 Quit barrywardell ()
20:52:39 Join BigMac [0] (
20:53:49desowinyeah, almost every plugin has check for SYS_USB_CONNECTED
20:55:01BigMacHey which gigabeat model performs better currently or are they the same between the f and x
20:55:36 Join |Rincewind| [0] (
20:55:59 Join Lethanialist [0] (
20:56:07nlsBigMac: in terms of performance they should be identical
20:56:28 Join nathan [0] (
20:56:30 Quit nathan (Nick collision from services.)
20:56:32nlsthe x i believe have slightly worse battery life, is smaller and has a bigger screen
20:56:41Lethanialistanyone have problems with ipod mini 1g restarting randomly?
20:56:55nlsalthough it is the same resolution
20:57:00BigMacalright thanks nls
20:57:01hcsit restarts precisely when it means to
20:57:12nlshcs: haha :-)
20:57:20 Join webguest77 [0] (
20:57:25 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
20:57:46Lethanialistheh :P its just a little annoying to be listening to music and have it restart, then freeze on the apple logo until i restart it yet again
20:57:56webguest77How do I execute a command in a different directory in a shell script?
20:58:29nlswebguest77 like you would from the command line?
20:58:42webguest77say i'm in rockbox dir
20:58:54webguest77and want to run make in some other dir from here
20:59:04 Join kaaloo [0] (
20:59:28nls../someotherdir/command should do it afaik
20:59:37linuxstb_webguest77: You can use the -C option to make - "make -C /path/to/other/dir"
20:59:45 Part kaaloo
21:00:08 Join Llorea1 [0] (
21:00:25nlsoh, you want to run it IN that directory, then don't do waht I said :-)
21:00:28 Quit Llorean (Nick collision from services.)
21:00:30 Nick Llorea1 is now known as Llorean (
21:00:48Nico_Pjust thinking about something... what about making bug/feature request require approval ?
21:01:18Nico_Pi don't know if flyspray supports it but it could be nice
21:01:45 Quit linuxstb (Read error: 113 (No route to host))
21:01:53 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
21:01:58JavaMan22i just made a nifty little program
21:02:11JavaMan22you are this sqaure and you move around :)
21:02:11linuxstbNico_P: Isn't that what the close button does?
21:02:13Nico_PJavaMan22: let me guess... in java ?
21:02:22JavaMan22no in c
21:02:27JavaMan22for rockbox
21:03:11 Quit borisyeltsin (Remote closed the connection)
21:03:44 Join Thundercloud_ [0] (n=thunderc@
21:04:17Nico_Plinuxstb: even when we close a redundant/useless task, it's still created. Also, maybe if users knew their requests were submitted for approval, they'd think twice about posting them
21:05:26Nico_Pthe "improve rockboy" feature request made me have this idea
21:05:28 Join elborak [0] (
21:05:41linuxstbNico_P: I just think we need better (and clearly linked to) guidance for bug reporters.
21:05:42 Join mots_ [0] (
21:05:51linuxstbNico_P: At the moment there is nothing...
21:05:58mots_is there any way to charge my h300 using usb?
21:06:12Nico_Plinuxstb: ask Llorean whether he thinks people read guidelines
21:06:26mots_it always charges till 3.35V, then boots
21:06:27peturmots_: you have to enable it in the settings
21:06:30Nico_Plinuxstb: but i agree they should be on the tracker
21:07:28 Quit perl|work (Read error: 110 (Connection timed out))
21:07:28mots_petur, where exactly?
21:08:24bluebrotherNico_P: add a "yes, I have read the tracker guidelines" checkbox :)
21:08:26peturmots_: settings_>general->system->battery->charge during... you have to boot into rb before connecting.
21:08:27mots_oh, found it
21:08:35bluebrotherand require it to be ticked before a task gets filed
21:08:46mots_how do I find out my battery capacity?
21:09:11 Quit funky ("leaving")
21:09:15peturmots_: did you swap it?
21:09:34mots_well, i bought it on ebay
21:09:58peturbattery or h300?
21:10:17mots_so i have no idea if the stock battery is in there
21:10:39linuxstbbluebrother: I like the "For Reporters" section on the FlysprayHowto page, but IMO there is far too much other text on that page, so that could easily be skipped. I think a page with just that "For Reporters" section would be nice.
21:10:46peturmots_: it's only used to estimate how much time is left
21:11:16 Quit Thundercloud (Read error: 60 (Operation timed out))
21:11:22mots_k, nvm then
21:11:58Lethanialistif i run an ipod nano simulator on the vmware image off the wiki, will things run exactly like the nano or differently because of memory issues? im trying to see how bad mpegplayer runs on a nano but i dont know if itll be the same as an actual nano...
21:12:33Nico_Pbluebrother: maybe several checkboxes : "i've read the guidelines", "i've made sure i'm not reporting an already existing task", "i've asked myself if the feature i'm asking for is useful/doable"
21:12:35linuxstbLethanialist: Don't you mean "how well mpegplayer runs on a nano" ?
21:12:55LloreanLethanialist: It won't be. It's just simulating the user interface, it doesn't emulate the hardware, so there's no reason for performance to even be remotely similar.
21:13:01Nico_Pbluebrother: a whole page of checkboxes :)
21:13:02linuxstbLethanialist: But using the UI sim for that is pointless - mpegplayer will run as fast as your PC's processor will allow.
21:13:16JavaMan22how come my splash thing wont go away?
21:13:31mots_another question: my old mp3-player used to play till the battery was REALLY and then running out of power without a warning
21:13:31 Quit webguest28 ("CGI:IRC (EOF)")
21:13:35linuxstbJavaMan22: splashes don't erase themselves - you have to draw over them.
21:13:35JavaMan22even when i do rb->lcd_clear_display();
21:13:51mots_is there any way to get that behaviour on my rockbox'd r300?
21:13:53JavaMan22it wont go away after that tho
21:13:54linuxstbJavaMan22: Are you doing lcd_update() after the lcd_clear_display() ?
21:14:06JavaMan22maybe i ought to do that :)
21:15:57 Join funky [0] (n=repulse@unaffiliated/funky)
21:17:29 Part JavaMan22
21:20:12 Quit Arathis ("Bye, bye")
21:20:26 Quit webguest77 ("using sirc version 2.211+KSIRC/1.3.12")
21:20:39 Nick mots_ is now known as mots (
21:21:13 Join raphi [0] (
21:21:22linuxstbnls: Is there a reason the manual generally has links to a directory on the download server, rather than the individual files?
21:22:39*bluebrother thinks for a closing reason for FS #6858
21:22:42 Join Thundercloud [0] (n=thunderc@
21:23:04bluebrotherlinuxstb: I think I did that to require the user having a bit more knowledge than simply clicking links
21:23:25bluebrother(and to be independent of future name changes)
21:23:53linuxstbBut shouldn't we be making it as error-proof as possible?
21:24:52bluebrotherdo you think it's error prone linking to a folder on the server? There aren't much options ...
21:24:54nlslinuxstb: it might have been because there are two different bootloaders for the x5s
21:25:19 Join midgey [0] (
21:25:32linuxstbIt just seems a unnecessary extra step for the user to have to do - why not just provide the full URL to the file the user needs to download?
21:25:56linuxstbFor the bootloaders, the names are generally fixed by the native bootloader.
21:27:30bluebrotheryes, but for ipodpatcher (and later sansapatcher) we would need to add a couple of links to the manual for each platform we have binaries
21:27:41bluebrotherIMO this somewhat crowds the manual and makes it harder to read.
21:27:58 Quit Thundercloud_ (Read error: 60 (Operation timed out))
21:31:43 Quit FOAD ("I'll be back")
21:32:49 Join demian79 [0] (i=c87ed463@gateway/web/cgi-irc/
21:33:44 Join FOAD [0] (
21:34:40demian79Hi, is my first time here and I need help
21:34:53nlsdemian79: just ask your question
21:35:28demian79I've got a TDK MOJO and I've read somewhere that I can find a firmware in rockbox
21:35:39nlsit's not supported
21:35:40demian79but I cant find it
21:35:52bluebrotherthe supported players are listed on the frontpage
21:36:00bluebrotherand that list is up to date.
21:36:10demian79do you know of any other open firmware that might help me
21:36:15bluebrotheryou might want to look in the "new ports" forums
21:36:27bluebrotherbut I highly doubt there is anything about it.
21:36:34demian79ok, I'll do that
21:36:46GodEaterand as far as I'm aware, we're the only alternative firmware out there, other than archopen
21:36:46demian79any other suggestion?
21:36:50bluebrotherbtw, where did you read that you can "find a firmware in rockbox" for that player?
21:37:20bluebrother(rockbox is a firmware, you can't find one "in" Rockbox ;-)
21:37:21demian79in a forum, im not sure
21:37:24demian79let me check
21:37:54demian79i cant find where
21:38:38demian79no, I understand, I meant if it was supported
21:38:41demian79thanx anyway
21:39:10nlsGodEater: doesn't ipl count? :-P
21:39:19GodEaternls: no :)
21:40:26demian79what is ipl?
21:41:03bluebrotheranyone knows off the head what ata error -32 is on h100?
21:41:31*GodEater greps the source
21:41:33*nls googled "TDK MOJO" and found some crazy flashplayer-built-into-headphones and a cd/mp3 portable...
21:42:49GodEaterbluebrother: something in check_registers ?
21:43:16bluebrotherI have no idea −− just stumbled upon a forum post and was wondering.
21:43:50GodEaterbluebrother: yeah - looks like check_registers, it's second check has failed
21:43:56 Join donutman25 [0] (n=chatzill@
21:44:06GodEaterwhich is reading / writing to the address registers
21:44:12GodEaterthat's failed it seems
21:44:15linuxstbbluebrother: The manual already links directly to ipodpatcher for the supported platforms - there are platform-specific instructions on how to use it.
21:45:59bluebrotherhmm. So where does it link to a folder?
21:47:11*bluebrother is a bit out of the current state of the manual
21:47:43linuxstbnls's most recent commit made me notice that the iaudio bootloaders are not linked to directly, but maybe there are other cases.
21:48:34linuxstbIn the PDF, I think we need the full URLs, but do you know if it's possible to make the html version of the manual have links without the full URL displayed?
21:48:53GodEaterwhy in one but not the other ?
21:50:15bluebrotherhmm. Should be possible.
21:50:50linuxstbGodEater: The PDF may be printed, or the PDF viewer may not support URL linking.
21:51:08 Quit donutman25 ("Chatzilla 0.9.77 [Firefox]")
21:51:20*GodEater hadn't considered the latter
21:51:34GodEaterI can't imagine someone typing in the URL from a printed copy though
21:51:53 Quit datachild (
21:51:53 Quit Hadaka (
21:51:56 Quit desowin ("use linux")
21:52:18 Quit Lear ("Chatzilla 0.9.77 [Firefox]")
21:52:51linuxstbGodEater: But maybe not even the PDF needs the full URL in the text - it could be a footnote.
21:53:12 Part raphi
21:54:48GodEaterlinuxstb: that's a good idea
21:55:15 Join Redbreva [0] (
21:55:21 Join Thundercloud_ [0] (n=thunderc@
21:56:24bluebrotherhmm, looks like the url package doesn't support different text and href :(
21:57:07 Join kwbr [0] (
21:57:22GodEaterbluebrother: that's a shame, I'd have thought that was a commonly required eature =/
21:58:03 Quit demian79 ("CGI:IRC (EOF)")
21:59:20Bagdernice to see near the top of the build servers
21:59:38LloreanGodEater: I type full URLs from printed things all the time, and I was planning on printing the PDF manual for my brother when I handed him his Rockbox'ed Recorder back.
21:59:52 Join Thundercloud__ [0] (n=thunderc@
22:00:06nlsBagder: and no rebuild on manual commits any more :_)
22:00:30bluebrotheroh ... my score on manual commits ...
22:00:37Bagdernls: the exclusion is even controlled by the little script in tools
22:03:04***Saving seen data "./dancer.seen"
22:03:47LloreanThe SansaE200InstallBeta instructions are invalid now.
22:03:55LloreanHas Sansapatcher been removed for a specific reason?
22:04:03 Quit Thundercloud (Read error: 60 (Operation timed out))
22:04:39bluebrotherGodEater: seems \href can do it but I don't know how that handles line breaks
22:05:30pixelmareminds me of the talk about bin size growth since moving to svn... IIRC there is still the plus of ~1000bytes of Slasheri's scroll code on targets that don't need it
22:06:09 Join Id2ndR [0] (
22:07:38acathlais it normal on a sansa e200 and rockbox that i can't see my music folder?
22:07:53bluebrothermaybe it's hidden?
22:08:01bluebrothertried setting filetypes to "all"?
22:08:27acathlaon rockbox itself?
22:08:57acathlai'm on the old firmware, i'll change the rights
22:10:04linuxstbLlorean: The links have never worked on the SansaE200InstallBeta page - it's not "live" yet.
22:11:11linuxstbLlorean: The only binaries are the ones I've been hosting. I think we're now waiting for Barry to finish his work on the bootloader.
22:11:31 Quit kwbr ("Leaving")
22:12:16acathlausb didn't work :(
22:12:42acathlahaaa, msc
22:13:36 Join billytwowilly [0] (
22:13:59preglowhmm, with the wps preparser plugin, bitmap files could also be packed together to a single file in the correct order
22:17:19linuxstbDidn't Slasheri implement bitmap caching once, which was subsequently removed?
22:17:27linuxstbThat sounds like the same thing.
22:17:52bluebrotherit was like 3 days in cvs. I think somewhat about a year ago
22:19:28linuxstbSome discussion is here - starting around 19.35 -
22:19:28 Quit ompaul ("enjoy")
22:19:35preglowlinuxstb: yes it did, but a wps preparser plugin would need some way to detect a modified wps anyway
22:19:40 Quit Thundercloud_ (Connection timed out)
22:20:14preglowlinuxstb: i think i started lobbying for the removal of that patch about 10 minutes after it was commited
22:20:44Lloreanlinuxstb: Ah, thanks.
22:21:42 Join Thundercloud [0] (n=thunderc@
22:22:28Lloreanpreglow: Why not have the plugin parse every time a .wps is loaded, and give parsed WPSes a different extension? Is there some reason we'd want to save parsed WPSes outside of the "Current WPS" and any pre-parsed WPSes we might provide with Rockbox?
22:23:16preglownone that i can think of
22:23:41preglowbut i really think we should consider the bitmap issue at the same time if we're gonna do anything like pre-parsing wpses
22:23:52Nico_PI agree
22:24:09Nico_PLlorean: i answered to your mail, it should arrive soon
22:24:17LloreanBut at that point all you need to do is say "If a WPS is modified, it must be reloaded"
22:24:17preglowand i pretty much think pre-parsing wpses is a good idea, as long as it can be done without the user/wps developer being too hassled by it
22:24:21 Quit Lethanialist ()
22:24:29Nico_Ppreglow: what was bitmap cache about ?
22:24:42bluebrotherthy not use a packaged file format initially like tar?
22:24:44preglowNico_P: it just dumped the loaded bitmaps to one huge files when loaded
22:24:46LloreanSince the post-parse WPS is stored as a separate file for quick loading on boot, you shouldn't ever get errors in the images since the WPS won't change at all until reloaded.
22:24:53preglowNico_P: but it didn't notice if bitmaps were changed
22:25:14preglowLlorean: how do we easily detect if a wps is changed, though?
22:25:19Lloreanpreglow: We don't.
22:25:31bluebrotherthen we could simply save indexes for the bitmaps
22:25:31preglowLlorean: so we force manual reload?
22:25:43preglowLlorean: not a big problem, i suppose
22:25:56Lloreanpreglow: I don't see forcing a manual reload if they consciously changed the WPS that big of a deal.
22:26:08linuxstbWhat about WPSs distributed with Rockbox?
22:26:13Lloreanlinuxstb: pre-parsed.
22:26:21LloreanThat way they can be loaded by the core, no plugin.
22:26:24preglowLlorean: me neither, i suppose. mainly just seeing this from a wps creator's perspective, but even there it shouldn't be much of a bother, i guess
22:26:43Lloreanlinuxstb: We can include the source .wps file and .bmp images in the wiki and of course they'll be in the source.
22:27:11Lloreanpreglow: Since WPSes and/or themes can be in the root anyway, they can make it as little as two button presses from boot to re-load the .wps
22:28:29 Join web-taz [0] (
22:28:38LloreanNico_P: I think preparsed files solves the "The core shouldn't depend on a plugin for WPS loading"
22:28:47 Quit Thundercloud__ (Success)
22:29:10preglowi really don't see anything wrong in core depending on a plugin for wps parsing anyway
22:29:18preglowthe disk access is already there, loading the plugin should be fast
22:29:22BagderI'm starting to like the pre-parsed/byte-compiled idea
22:29:34preglowBagder: it does grow on you ;)
22:29:38Lloreanpreglow: It's the idea of Rockbox not being able to use ANY WPSes if the plugin is absent or outdated that is the problem.
22:29:45LloreanPlugin API changes don't coincide with codec changes.
22:30:03preglowif you have an outdated wps parser plugin, you haven't installed rockbox correctly
22:30:25Lloreanpreglow: Yes, but you could still have a Rockbox that can play music. Someone made a stupid mistake before a long trip, for example, and got lucky.
22:30:38preglowsure, more robustness is good
22:30:47preglowbut storing them preparsed makes it a non-issue anyway
22:31:06linuxstbpreglow: My only worry is that if more and more core functionality is in plugins, or uses the plugin buffer, then we're limiting the flexibility of Rockbox. e.g. TSR plugins will no longer be as useful.
22:31:10LloreanThat was the point of that idea. :) I know there's at least someone who feels strongly about the aforementioned dependency on the plugin. ;)
22:31:21preglowlinuxstb: yeah, the last especially is one of the only concerns i can see
22:31:39preglowlinuxstb: but in this case, at least, the plugin shouldn't be called too often
22:31:45preglowletting tsr plugs do their thing
22:33:04Nico_Pif the wps are preparsed files, the plugin shouldn't need to be used that often anyway
22:33:20linuxstbAnd rbutil could parse them...
22:33:22preglowpeople should distribute them preparsed
22:33:40preglowi don't think creating one preparsed format for all targets should be impossible
22:33:59Nico_Ppreglow: i'm pretty sure it's not
22:34:06 Quit _Veseliq_ (Remote closed the connection)
22:34:16LloreanThe parser plugin would really only exist for easy use while designing WPSes, or for people who wanted to distribute WPSes that can be modified on-player
22:34:18bluebrotherhow should this binary format look like?
22:34:18Nico_Pah, endianness... forgot about that for a sec
22:34:38linuxstbpreglow: Bitmaps should be converted to the native LCD format, so there is one difference (e.g. ipod greyscale and H1x0 LCDs are same size/depth but different format)
22:34:40preglowNico_P: do we need values higher than 256? :>
22:34:47BagderNico_P: we could use a fixed endianess for the format
22:35:01preglowlinuxstb: hrm, true, but not much of a problem anyway, if you ask me. wpses are already target specific
22:35:23Bagderah yes
22:35:26Nico_Ppreglow: not quite sure... would the bin file be a simple memory dump of the wps_data struct ?
22:35:41preglowNico_P: can't help you there, i haven't ever touched the wps code
22:35:46linuxstbBut that was my example - a WPS in the current form can be distributed to work on all targets with same lcd size/depth, but a binary format would extend that to LCD framebuffer format.
22:35:47preglowi just enjoy talking about it, apparently...
22:35:57Lloreanlinuxstb: Converting the bitmaps doesn't have to be part of the parsing. They could just be concatenated in the file, and converted by the final load still.
22:36:18Nico_Ppreglow: i'm comfortable with the wps code, just not so much with the bin file concept
22:37:07 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
22:37:24linuxstbLlorean: I think we should go all the way and store bitmaps in the native LCD format, even if that means very slightly less portability. As preglow says, WPSs are almost target-specific already.
22:37:57 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
22:38:02bluebrotherwhat is the upside of such reduces portability?
22:38:07bluebrotherfaster loading?
22:38:08nlslinuxstb: and wpss could still be distributed in the current format, keeping the portability
22:38:18Lloreanlinuxstb: As it is, I really don't think any but the included WPSes NEED to be preparsed.
22:38:34linuxstbSo basically the suggestion is to make Rockbox core only load the pre-parsed binary format, and offer parsing capabilities everywhere (Rockbox build system, rbutil and on-target plugin).
22:38:44Nico_PI agree with Llorean. it's a one time thing anyway
22:38:56preglowthe parsing code should be simple to compile outside of rockbox anyway
22:38:57Lloreanlinuxstb: If we have an official gallery of WPSes, like the twiki or whatnot, I think they could be provided as text still (or better yet, included parsed versions for each compatible target there, at least)
22:39:02preglowcan't see why that should use too many rockbox functions
22:39:10Nico_Ppreglow: yes it is
22:39:20Lloreanlinuxstb: I think so.
22:39:29 Quit Entasis ("Leaving")
22:39:30linuxstbLlorean: Now that rbutil can grab WPSs from, it can parse them before installing.
22:39:30bluebrotherhaving some extracting tool would be nice tool
22:39:43bluebrotherso one can adjust binary themes
22:39:54Lloreanlinuxstb: Indeed, that would be excellent. And if a user manually downloads them, they just have a slightly longer load time since they depend on the pre-parser viewer.
22:40:18nlsbluebrother: why not keep the original text one and just reload it when it's changed?
22:40:32nlstext==human readable
22:40:54Lloreanbluebrother: The original text one should always be around, somewhere. Either at the theme site where the pre-parsed one was downloaded, or in the Rockbox source tarball, at the very least.
22:41:09 Join Stefan___ [0] (n=chatzill@
22:41:10Nico_Pnls: i agree it sound like overkill but i think the parser could very easily be inverted :)
22:41:19Lloreanbluebrother: If an old-style WPS is loaded on target, it shouldn't be deleted, just temporarily converted.
22:41:21Bagderwell, with binary formats, people will start spreading binary-only files
22:41:33Bagderthink MR
22:41:35LloreanA reverser would be good though.
22:41:47linuxstbBagder: I've thought the same thing - we don't want "closed-sourced" WPSs...
22:41:48Stefan___i have a question
22:41:48nlsBagder: we eill have to make the files self-destruc when they leave the dap then :-)
22:41:50LloreanIt _should_ be unnecessary, but yeah, I guess you can't really depend on people.
22:42:07Nico_Pa decompiler would be quite easy i think
22:42:12Stefan___can u "hear" me?
22:42:27LloreanStefan___: I don't think theres a "u" in the room.
22:42:30LloreanStefan___: But the rest of us can.
22:42:31linuxstbStefan___: No, but I can read you.
22:42:35preglowNico_P: if you've made a parser with output that can't be easily deparsed, you've made a very bad parser :P
22:42:55Stefan___that's why i used quotes :-l
22:42:59Nico_Pwow this thread is full of funny people :)
22:43:05Stefan___in the plugin api
22:43:09Bagderwell, perhaps the binary format could include the original text at the end...
22:43:10Nico_Pthis channel i meant
22:43:39Stefan___is there a function to get the colors of a pixel?
22:43:47Nico_Ppreglow: i take that as a compliment to my parser :)
22:43:50linuxstbBagder: Yes. And I guess at least the filenames of the bitmaps...
22:43:58Nico_Peven though it's not quite mine
22:44:02linuxstb(assuming we're incorporating the bitmaps in this format...)
22:44:39linuxstbStefan___: I don't think so.
22:44:41LloreanSo the text of the WPS, and the filenames of the bitmaps, then just have the extractor copy the text, and divide the bitmap portion back up to the appropriate files?
22:45:20Nico_PI'm not sure the original text is even needed...
22:45:29Stefan___damn.. it'll be hard to implement semi-transparency..
22:45:31LloreanNico_P: Comments.
22:45:32Nico_Pbut it doesn't hurt
22:45:38Stefan___thanx, bye
22:45:41BagderStefan___: you can read it fromt he framebuffer
22:45:59Stefan___easy to say
22:46:01Stefan___how do i do that? :))
22:46:07bluebrotherhow about a binary format like that:
22:46:08BagderNico_P: and as preglow said, you can make a better binary format if you don't need to decompile it fine
22:46:24bluebrotheruse a standard tar file. Put in the binary wps as first file
22:46:31bluebrotherthen add the bitmap data.
22:46:42bluebrotherand all other files after that
22:47:03linuxstbStefan___: You also have to deal with any backdrop - that's stored in a separate buffer.
22:47:18Nico_PBagder: could you have a look at the current format and tell me what you think ?
22:47:22*bluebrother sees loads of underscores on this page
22:47:42Nico_PBagder: I mean the current one in my patch...
22:47:45BagderNico_P: I'll try to get some time to check it out
22:47:49bluebrotherso if someone wants to decompose the binary theme he just needs to untar everything
22:48:13bluebrotherbtw, I think we should also include theme configuration in that files, not only wps
22:48:15linuxstbtar = "theme archive" ....
22:48:22*Nico_P should probably post the patch to the tracker now
22:48:29bluebrothertab = theme binary
22:48:33Stefan___too much for me :)) i'll just fill the screen with a more or less random color and i'll make an array to store pixel info
22:48:45Stefan___will it fit?
22:49:34nlsStefan___: fti where?
22:49:38nlsfit even
22:49:39linuxstbStefan___: What are you trying to do? Are you doing this in a plugin or core Rockbox?
22:50:37 Join My_Sic [0] (
22:50:51Stefan___core rockbox is too much for me
22:50:56 Join lavid [0] (
22:51:00preglowyes, it is quite fattening
22:51:04linuxstbThen (without stopping audio playback), you have 512KB for your code and data.
22:51:09JdGordonNico_P: I dont think the user should ever see the proccessed wps... so no distroubting them instead of the text wps in the core...
22:51:34Stefan___so is a 320*240 int array too large?
22:52:01Bagderints are lovely oversized for 16 bit color...
22:52:07linuxstb320*240*4 = 300KB
22:52:26Stefan___short in
22:52:29Stefan___short int
22:52:38Stefan___150 KB
22:52:50Stefan___so, again, is it to large?
22:52:52bluebrotherJdGordon: that would be on the fly binary converting ...
22:53:10*bluebrother likes the C99 fixed size types
22:53:18JdGordonor stored to a file...
22:53:27JdGordoneither way the user would always load the .wps
22:53:36 Quit amiconn (Read error: 110 (Connection timed out))
22:53:36 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
22:53:42bluebrotherbut then the parsing itself would still be in the core
22:53:47JdGordonthe core would then check for the_wps.wps_compiled or whatever
22:54:00linuxstbJdGordon: My feeling is that hiding things from the user always ends in tears... Rockbox should be open about what's going on.
22:54:28bluebrotherI had a tar theme loader some time ago. As that was quite slow I added caching of an index for the files
22:54:43bluebrotherbut the improvement wasn't too big
22:56:01 Quit My_Sic (Read error: 131 (Connection reset by peer))
22:56:20bluebrotherbut what is the main idea behind a binary theme format? Code size or loading speed?
22:56:22 Quit FOAD ("I'll be back")
22:56:43Nico_Pbluebrother: a bit of both maybe
22:57:20Stefan___can u spend some time teaching a noob how to use the framebuffer?
22:57:34bluebrotherloading time could be adressed in a different way ... like loading the bitmaps from a tar file and putting it into memory in the order they appear in the file (and adjusting the wps dynamically to the changed order)
22:57:56 Join Thundercloud_ [0] (n=thunderc@
22:58:15JdGordonI would think this preparsing the wps would affect bin size much more than loading speed...
22:58:29bluebrotherhehe ... my mini just plays "Gimme just a little more time" :)
22:58:32Nico_Pbluebrother: not sure that's really useful... but the parser could have a basic tar extractor
22:58:55bluebrotherbasic tar extracting is easy −− I already did that.
22:59:25bluebrothernot sure about the code size −− we would still need to check for some tokens, won't we?
22:59:30mattzzJdGordon: mind to have a look a ?
23:00:23bluebrotherJdGordon: as I'm just thinking of it ... there is a "nothing to resume" splash hang, amiconn suspeced the new main menu to cause it. Any ideas?
23:00:37bluebrotherit's quite annoying as the player locks up completely
23:00:58Stefan___can you tell me how i use the frame buffer, or should i leave you alone?
23:01:11linuxstbStefan___: It's simply an array of 320*240 shorts (but defined as fb_data, not short).
23:01:50bluebrotherthe only advantage I see is that the data could be packed more tightly −− but otoh we need to make sure it's utf8 safe
23:01:55linuxstbSo get_pixel(x,y) is rb->lcd_framebuffer[y*LCD_WIDTH+x]
23:02:14 Join Juice^ [0] (n=Juice@
23:03:44 Quit OgMaciel ("mv OgMaciel $HOME")
23:03:45JdGordonmattzz: I dont think thats the way to go.... power is the menu button... id be more inclined to use the rec button then...
23:03:56JdGordonor _always_ use rec for std_cancel
23:04:04JdGordonbluebrother: no idea... the code is sound :p
23:04:39LloreanJdGordon: I think "Rec" on the Sansa is a good cancel button.
23:04:52JdGordonmattzz: left shuold never be std_previous tho... the wheel should be
23:05:28 Join FOAD [0] (
23:05:31Juice^The sansa doesnt have a "stop" button.. perhaps the rec button could be that when you're in WPS
23:05:43Juice^or is there already one i dont know of
23:05:49JdGordonanyway... im off to uni... ill be back on in 2 hours or so, so leave a message :p
23:05:51linuxstbJdGordon: Why not? IMO, wheels should be used when movement over a lot of range is needed, and the left/right/up/down (if present) are more suitable for small movement ranges.
23:05:52 Quit midkay ("Leaving")
23:06:34JdGordonhmm.. maybe... i dunno...
23:06:37mattzzHm, on other targets I like the left button to quit a settings dialog
23:06:46 Quit web-taz ("( :: NoNameScript 4.02 :: )")
23:08:46 Join netmasta10bt [0] (
23:10:33mattzzJdGordon: in the settings context the power button quitting a setting is the same as calling the menu, right=
23:10:41 Join lnux [0] (
23:10:46 Part lnux
23:11:28mattzz(the same is used on the recorder)
23:11:32 Quit Stefan___ ("Chatzilla 0.9.74 [Firefox]")
23:11:47 Quit lee-qid ("aufwiederbyebientotsayonara")
23:11:51 Quit midgey ()
23:13:04 Join bospaadje_ [0] (
23:13:31 Quit bospaadje (Read error: 54 (Connection reset by peer))
23:13:48 Nick bospaadje_ is now known as bospaadje (
23:13:53Nico_Phas anyone tried the tokenizer patch ?
23:14:04 Quit Thundercloud (Read error: 110 (Connection timed out))
23:16:12bluebrothernot yet
23:16:58Nico_Pit still needs some testing, although it has worked fine in all my tests
23:20:46 Quit JdGordon ("Konversation terminated!")
23:20:55 Join Domonoky [0] (
23:25:41 Part muszek ("Ex-Chat")
23:26:15bluebrotherhow's the code size aspect?
23:26:32MoosBagder: wiki question, is there any reason why we no more have the wiki page with the stats, don't remenber how this page was named?
23:28:35 Quit mattzz ("Leaving")
23:30:13pixelmabluebrother: he said yesterday that it would add ~650 bytes or something on my Ondio... hope there is some optimisation potential in there..
23:31:40bluebrotherI'm not sure if I like this binary wps idea
23:31:49 Quit printfXh4 (Read error: 148 (No route to host))
23:32:37linuxstbbluebrother: Why?
23:32:40Nico_Pbluebrother: as pixelma said, it does increase the code size a bit
23:32:48Nico_Por a little more than a bit
23:33:28preglowjhMikeS: so, your arm asm career has started? :>
23:33:46LloreanNico_P: That's not the binary WPS idea.
23:33:54*bluebrother wants to code Rockbox instead of java :(
23:34:03linuxstbpreglow: Sadly he's at the wrong end (Gigabeat) of the scale...
23:34:05LloreanNico_P: That patch is just the tokenizing of the WPS. Makes the WPS code a lot easier to read and work with, is my understanding.
23:34:18LloreanNico_P: Oops, that was meant for bluebrother
23:34:31bluebrotherhehe ;-)
23:34:48Nico_PLlorean: that's it
23:35:01bluebrotherLlorean: I know that it's not the binary thing.
23:35:04Nico_Pif we offload the parsing to a plugin, code size will be positively affected too
23:35:26bluebrotherstill, I'm not sure if I like the binary idea
23:35:35preglowlinuxstb: yes :/
23:35:53bluebrotherI would rather try the tar attempt in a more sophisticated way again
23:36:14Lloreanbluebrother: The binary idea would allow the code size decrease.
23:36:33 Part lavid
23:36:42bluebrotherLlorean: I doubt it would make a huge difference
23:37:02LloreanHow much code is currently dedicated to the parsing?
23:37:05bluebrotherwe still need loading of bmp files (for backdrops etc)
23:37:16 Join midgey [0] (
23:37:20Lloreanbluebrother: Not if they're stored sequentially in the post-parsed format.
23:37:23linuxstbbluebrother: The tar files will always be slower than just reading a dump of the wps image struct from a file...
23:37:48LloreanIf you can just load the WPS struct and the image cache from a file, that's a lot less code than the active parsing of the WPS into that struct.
23:38:00bluebrotheryes, if we use the native lcd format. But I dislike the portability broken
23:38:04Bawitdabamaybe a bit off topic but, when u buy an ipod it comes with a dock 'connector' can u charge the ipod with this or do u have to buy the un included adapter
23:38:22Lloreanbluebrother: Portability is only broken if people distribute the binary files. They don't have to, and shouldn't be encouraged to.
23:38:40linuxstbBawitdaba: ipods come with a "dock-to-usb" cable - you can charge with that using a computer as the power source.
23:39:00Bawitdabaoh word so u dont need a charger then
23:39:03Bawitdabau can just use usb
23:39:08bluebrotherI'm pretty sure you know better than me that users will do so ;-)
23:39:16Bawitdabaguess ill grab an ipod then
23:39:37Lloreanbluebrother: Yes, but that's okay, people can make WPSes that are target specific anyway. Any iPod 4G WPS with a clock on it, for example.
23:39:42Lloreanbluebrother: Or the million WPSes that require patches.
23:39:48linuxstbbluebrother: Themes are hardly very portable at the moment. The number of targets with the same size/depth LCDs is small.
23:39:58LloreanAt least this way we can have a small header that can pop up a splash saying "This WPS is incompatible"
23:40:02 Join webguest30 [0] (i=04146273@gateway/web/cgi-irc/
23:40:07pixelmaLlorean: why? ... they would work on M5/X5 too (with the clock)
23:40:14*petur imagines the mess when the internal wps struct changes and the binary files no longer load. Hello version numbers and automatic re-parse?
23:40:25pixelmaand rtc modded H1x0...
23:40:31Lloreanpixelma: They won't work on an H120, despite the fact that our WPS gallery is labelled as H1xx/4G iPod
23:40:40webguest30anybody know where I can find parts for an iRiver iHP-120? My hard drive is good, but the board is shot I think.
23:41:07bluebrotherand if I want to run a 4G wps on the h120? Currently it works except for RTC tags (which are rarely used anyway)
23:41:08Lloreanpetur: Again, only a problem if people distribute their WPSes as binary
23:41:22LloreanMy original idea was only for the WPSes in the to be distributed as post-parse.
23:41:36peturstill, you'll need to (auto)re-parse
23:41:43Lloreanpetur: Why?
23:41:58peturif the format changes?
23:42:04linuxstbbluebrother: The idea is still for WPSs to be distributed in their current format, but for them to be converted to binary by either rbutil if the user installs that way, or automagically via a Rockbox plugin.
23:42:11LloreanIf the format changes, the included WPSes in the .zip should be in the new format.
23:42:14 Quit BigMac ("Leaving")
23:42:35LloreanThe only incompatible parsed WPS will be the "current" WPS which will have to be reloaded on the first boot after installing a new build.
23:42:50LloreanThat can be automated, if the parsed WPS version doesn't match the current struct version.
23:42:52bluebrotherwhat is the main bottleneck when loading bitmaps? Decoding?
23:43:00Bawitdabayeah screw it ill get an ipod 80gb
23:43:03MoosBagder: any thought? vs
23:43:08*Bawitdaba sucker
23:43:14linuxstbbluebrother: I would guess disk seeking and reading.
23:43:15LloreanBawitdaba: Rockbox doesn't support the 80gb one yet.
23:43:15bluebrotherBawitdaba: rockbox doesn't work on the 80GB yet
23:43:24Bawitdabait will
23:43:29bluebrotherlinuxstb: but that could get adressed differently.
23:43:38bluebrotherno need for some binary structure
23:43:42Bawitdabai assume it will anyway lol
23:43:45Bawitdabafrom the post i read
23:44:02Lloreanlinuxstb, bluebrother: I wouldn't think it'd be that bad, then if the bitmaps were stored in the binary in a known format instead of the framebuffer format, and converted on load?
23:44:11LloreanBawitdaba: It will be eventually, yes.
23:44:37 Join smolyn [0] (
23:44:47bluebrotherthat's another conversion ...
23:44:49linuxstbI just don't see the problem, if the standard distribution format will stay the same - i.e. a text file and .bmps.
23:44:51Lloreanlinuxstb, bluebrother: It's not like we can remove the bmp->framebuffer code from the core anyway?
23:44:51Bawitdabacant wait to get one so i can be un zombie like at work
23:45:07Lloreanbluebrother: A conversion that takes place with the current WPS code anyway.
23:45:08bluebrotherLlorean: how should we deal bmp backdrops then?
23:45:19Nico_PLlorean: to answer the question you asked I removed most of the loading code and compiled (it worked)... it took 5688 bytes off the gigabeat bin file
23:45:20Lloreanbluebrother: I don't understand the question.
23:45:40Bawitdabajust minor off topic again, should i get white or black i hear black scratches easy?
23:45:41*bluebrother just is converting data ... and converting data ... and loves programming it :(
23:45:55linuxstbBawitdaba: Do you want to use Rockbox?
23:46:10linuxstbThen the ipods are not the best choice.
23:46:17bluebrotherLlorean: currently I can select any file from the backdrops folder and set it as backdrop. If conversion from bmp->framebuffer is removed, how should that work?
23:46:20Bawitdabawhys that
23:46:30Lloreanbluebrother: I said we cannot remove that conversion.
23:46:44bluebrotherLlorean: ok, misread you.
23:46:48linuxstbRockbox performs better on most other platforms - mainly because the ipods have very little public documentation about the hardware.
23:46:59andrewg867hey, does anyone know about the sansa emulator (the windows one)?
23:47:10markunNico_P: nice, but then you need a separate tool to compile the WPS?
23:47:17andrewg867I can't get it to stop crasing
23:47:21Bawitdabado u know of a better 80gb player that is or will be rockboxable then?
23:47:27Nico_Pmarkun: it could be a plugin
23:47:27bluebrotherbut what major code parts would be superseded if we have another binary blob?
23:47:33Lloreanbluebrother: I basically think the binary format should be [wps struct][concatenated bitmap images in the order they appear in the wps]
23:47:36LloreanOr something similar
23:47:40LloreanRather than framebuffer data.
23:47:46LloreanBut framebuffer data seems popular.
23:47:54*Llorean never quite caught why exactly
23:48:00 Quit webguest30 ("CGI:IRC (EOF)")
23:48:19bluebrotherLlorean: that's mostly the attempt I'm thinking of, but I think of using tar as a container
23:48:33LloreanBut you're still advocating using the old text format for the WPS
23:48:35 Quit mots ("Ping-Pong Time's up!")
23:48:36bluebrotherthat way one could easily decompose such a thing
23:48:44linuxstbBawitdaba: The ipod is the only player sold with an 80GB disk, but you could upgrade a lot of the players Rockbox runs on to that size.
23:48:48Nico_Pand to those whith concerns about portability... i'm pretty sure it could be identical to what we have with the text files, at least as far as the tokens are cocnerned
23:48:50bluebrotherI don't see any advantage in a binary format for wps
23:48:52Lloreanbluebrother: I don't think we *have* to support decomposing them.
23:49:07Lloreanbluebrother: How is "smaller core code size" not an advantage?
23:49:14bluebrotherof course we don't *need* to do so
23:49:30bluebrotherbut how much will the advantage be? I really doubt it will be that much
23:49:47Llorean5k on ARM according to Nico_P's test he just attempted.
23:49:52bluebrotherfor the wps: you still need to parse some format. So why is a binary format better?
23:50:01LloreanIn a binary format you DON'T have to parse.
23:50:13LloreanYou just copy the data into memory.
23:50:17Nico_Pbluebrother: with a bin file you read the contents directly to the memory
23:50:18bluebrotherNico_P: you already did some tests with a binary format?
23:50:20Nico_Pand voilĂ 
23:50:34Lloreanbluebrother: He just removed the parsing, as if it were using a binary format.
23:50:49bluebrotheryes, but how is deciding what to display working then?
23:50:55Nico_Pbluebrother: I already have a WPS tokenizer (see my mailing list posts). I removed the code related to parsong and loading the WPS
23:50:59bluebrotheryou still need to do something
23:51:07Lloreanbluebrother: That's a different series of steps entirely
23:51:19Nico_Pbluebrother: the displaying code was still there
23:51:32LloreanParsing is "Going from .wps to the struct in memory", and isn't related to drawing it.
23:51:50Nico_Pparsing and displaying are totally independant. I could compile the core after having removed all the parsing and loading code
23:52:15Bawitdabalinuxstb i dont suppose they use ata-6 2.5" hds
23:52:29linuxstbBawitdaba: No, 1.8" disks.
23:52:30Bawitdabasweet deals on 160gb notebook hds
23:52:34bluebrotherhmm. And what form has the data after tokenizing?
23:52:52Nico_PLlorean: I'm pleased to see you understand and explain the whole thing well :)
23:52:59Nico_Pyou explain a lot better than me
23:53:14linuxstbBawitdaba: Unless you wanted an old Archos device, they take 2.5" drives, but can only play MP3 files, nothing else.
23:53:30LloreanNico_P: Unfortunately his most recent one, I have no clue. I haven't looked at the patch yet. :)
23:53:39MoosNico_P: native english user surely help ;)
23:53:50Nico_Pbluebrother: you'd have to look at the patch, but basically it's an array of "struct wps_token"
23:53:56*bluebrother isn't native english
23:54:06 Quit Juice^ ("Leaving")
23:54:12*Moos thought to Lorean
23:54:13Nico_PMoos: i'm native english but I suck at explaining things
23:54:15Lloreanbluebrother: The tokenization is mainly to make the source code more readable You still have the WPS data in memory.
23:54:27bluebrotherI know.
23:54:36Nico_Psorry for the paste, but here is the struct :
23:54:40Nico_Pstruct wps_token {
23:54:40Lloreanbluebrother: Basically, the idea of preparsing the WPS is that you convert the WPS from whatever on-disk format it is, to its in-memory format, in a plugin instead of the core.
23:54:44Nico_P enum wps_token_type type;
23:54:45bluebrotherbut if I don't have the wps data in memory anymore, how is it supposed to work then?
23:54:48Nico_P bool next;
23:54:49linuxstbBawitdaba: See this page for the different devices Rockbox supports -
23:54:52Nico_P unsigned char flag;
23:54:52MoosNicco_P: hehe :) , don't know why I thought you was french too
23:54:56Nico_P union {
23:54:56Nico_P char *s;
23:54:56Nico_P char c;
23:54:56DBUGEnqueued KICK Nico_P
23:54:56Nico_P int i;
23:54:56Nico_P } value;
23:54:57***Alert Mode level 1
23:55:05Nico_PMoos: I'm both french and english ;)
23:55:07bluebrotherNico_P: found that just myself ;-)
23:55:10Lloreanbluebrother: That way Rockbox can load it directly in-memory without having to have any of the code necessary to interpret the on-disk format in the core, just the code necessary to take the in-memory format and use it to draw
23:55:34MoosNico_P: that make sense then:-)
23:55:47bluebrotherhow does it work currently? It parses the wps text data every time?
23:55:54Nico_Pbluebrother: yes
23:56:14Lloreanbluebrother: The WPS is currently "Convert from File->RAM format" then "Draw based on RAM format", you're splitting that and moving the first one into a plugin. The "binary" format, is identical to the in-RAM data and doesn't need to be converted at all during loading.
23:56:21CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
23:56:21*bluebrother has a strong dislike to binary blobs
23:56:23Nico_Pincluding conditonal stepping and all sorts of useless things
23:56:32Lloreanbluebrother: Ideally, nobody should ever interact with the binary blobs.
23:56:44bluebrotherstill it's a binary blob
23:56:54 Join Frode_ [0] (n=Frode@
23:57:13Nico_Pbluebrother: it could have the original code at the end ;)
23:57:17markunbluebrother: make a patch to keep text parsing after Nico_P's stuff has been committed :)
23:57:18Lloreanbluebrother: So's the Rockbox binary...
23:57:24MoosMe have the feeling that bluebrother will be hard to convince ;P
23:57:27Lloreanbluebrother: We don't distribute Rockbox as source, we just make it available.
23:57:42LloreanText parsing will always be in "Rockbox" just not in the core.
23:57:47bluebrotherLlorean: you can get the source easily ;-)
23:57:54 Join Hal9000 [0] (
23:57:55Lloreanbluebrother: Same with all the WPSes we will provide.
23:58:05bluebrotherand a binary executable is somewhat different ;-)
23:58:19LloreanIf anyone's going to distribute binary-only WPSes, I will fight tooth and nail to prevent them from being on our wiki, or advertised in our forum.
23:58:43LloreanFrom a user perspective, clicking on an old-style text.wps will work exactly the same as always
23:58:54LloreanThe only difference is where the actions take place, half of them take place in a plugin.

Previous day | Next day