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 2007-11-02

00:01:21 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
00:01:40Buschelamiconn: looks all very interesting... for non-full screen you do intermediate dest-address updates for each line?
00:02:31linuxstblostlogic: Do you have an old 5g?
00:02:41amiconnBuschel: Yes, I have to...
00:03:28webguest57Okay, i'm fine with just dual-booting
00:03:31amiconnThe update window is always the complete screen, otherwise I couldn't write behind an update...
00:03:31 Quit webguest57 ("CGI:IRC")
00:04:14Buschelamiconn: can't wait to test :o)
00:04:42lostlogiclinuxstb: I'm not sure what an "old" 5g is ... I'ave had my ipod g5 30g since before devcon '06
00:04:52linuxstbI mean it's not a 5.5g?
00:05:11 Join vladsinger [0] (
00:05:20lostlogicamiconn: I've turne don peakmeter and am running the patch
00:05:21linuxstbOK, I'll test as well anyway...
00:05:51linuxstblostlogic: amiconn asked me this question - did you ever upgrade the Apple firmware?
00:06:09lostlogiclinuxstb: nope, nevar
00:06:23lostlogiclinuxstb: I'm not entirely sure I've ever booted into the OF
00:06:24linuxstbThat's good - so you're probably a better tester than me.
00:06:45 Quit Gnu47 (Nick collision from services.)
00:06:48Buschelamiconn: yep, understood. you'll need to always use the same setup and let the update write to the correct addresses within the fullscreen rectangle
00:08:00amiconnIt's just an internal buffer in he bcm we're writing to
00:08:14amiconnI also had to learn that the bcm doesn't like single-word writes
00:08:54amiconnSo I round to even x and width, and the plus is that I could drop the single-word leader and trailer checks from lcd_write_data()
00:09:42Buschelamiconn: this is what i did with my last patch (which can be closed now as it is not compatible)
00:11:07lostlogicit's been a long time since I turned on peakmeter... has always been too CPU intensive on ipv... I assume this patch is part of the effort to make it not so?
00:11:09 Quit Arathis ("Bye, bye")
00:11:40amiconnWell, peakmeter itself was never really cpu intensive, but lcd update intensive
00:11:46amiconn(20 updates per second)
00:12:22 Quit scorche|w (K-lined)
00:12:52amiconnNow those 20 updates don't do much harm (since they're partial - only the rectangle containing the peakmeter)
00:13:20lostlogicah, sweet!
00:14:16 Join donutman25 [0] (n=chatzill@
00:15:40 Join saratoga [0] (i=9803c6dd@gateway/web/cgi-irc/
00:15:51lostlogicamiconn: other than listening to music and seeing that it works and does not freeze anything I should do to abuse this patch?
00:16:22amiconnDo something else that's graphical intensive, e.g. running plasma at 80MHz, or cube at 80MHz and h=1
00:16:44*linuxstb plays bubbles
00:16:58amiconnOr try mandelbrot and check if there are graphical glitches (like columns not updated when scrolling around and/or zooming)
00:17:23amiconnSpeaking about games - stars now needs a longer delay for the transition effect...
00:17:24saratogai'm noticing a lot of issues on the Sansa since MOB was comitted
00:17:28saratogahave these been reported?
00:17:51amiconn...and also for the ball movement :)
00:17:53lostlogicsaratoga: most of them have, some have been resolved... nanok has one which is the most nagging afaik
00:17:57 Quit barrywardell (Remote closed the connection)
00:19:28*pixelma had one playback stall today on c200 with just mp3s but that was with r15381 (had to search for it on the info screen :\ )
00:19:43linuxstbamiconn: Everything seems fine on my 5g.
00:19:52amiconnNice :)
00:19:57lostlogicyeah, fine here too −− abusing mandelbrot soundly
00:23:13*amiconn adjusts star
00:23:33linuxstbbubbles is still too slow at 30MHz, but a bit too fast at 80MHz...
00:23:57lostlogiclinuxstb: I consider playing music while playing bubbles part of the challenge of the game
00:24:10lostlogiccan I play without causing skipping and still do well :_P
00:24:24preglowlinuxstb: seriously, we're still using tuning constants?
00:24:33pixelmaI always wondered why you see cpu boosted or not so obviously in bubbles
00:24:34preglowit's not that hard to adapt something to the framerate dynamically
00:24:42linuxstbI don't think bubbles even uses that - it's very inconsistent.
00:24:46 Quit MajorC ()
00:25:12 Join alienbiker99 [0] (
00:25:55pixelmalostlogic: there is (was?) a patch in the tracker that "fixes" the issue by letting bubbles constantly boost the cpu... :D
00:26:14 Quit ender` (Read error: 110 (Connection timed out))
00:26:46linuxstbI think we should probably reject that - someone needs to fix it properly...
00:27:01pixelmait's still open: fs #6849 in case someonwe wants to close something...
00:27:46pixelmaI thought "maybe leave it open for the proper fix" but don't know
00:30:15 Join ender` [0] (
00:30:41 Quit saratoga ("CGI:IRC (Ping timeout)")
00:31:36preglowproper frame rate independent drawing isn't any worse than just calculating position offsets based on time differences
00:32:12Buschelamicoon: good night, will test the patch tomorrow :o)
00:32:15 Quit Buschel ()
00:32:22preglowif you're moving a bubble from gun to destination and know how many pixels per second it's supposed to move, just calculate the position based on time
00:33:53preglowthe tick timer ought to be sufficent
00:34:45*lostlogic wnoders if Nico is working on a special file type for atomic-read audio
00:34:47DerPapstnow that you're nearly done with coding that in your mind...
00:42:10 Quit ompaul (Client Quit)
00:44:10amiconnlinuxstb: Do you think it's ready for commit?
00:44:13*amiconn thinks so
00:44:23 Join bytie [0] (
00:44:46 Quit bytie (Read error: 104 (Connection reset by peer))
00:45:40preglowbubbles rocks
00:46:59 Join Llorean [0] (
00:47:33 Quit TotallyInfected ()
00:47:37linuxstbamiconn: Seems good to me. I've just tested mpegplayer - starts at 19.2fps for widescreen, going down to about 17fps during the high motion scene near the start.
00:47:54amiconnYeah, I did test the same thing too
00:47:56linuxstbSeems to be settling around 17.3fps
00:48:09amiconnTry the 160x128 version, it runs at ~45fps on average :)
00:48:27 Join TotallyInfected [0] (
00:48:46amiconnThe beginning is >50fps
00:49:08linuxstbI only have the 176x96 version on this ipod - around 55fps...
00:50:48linuxstbSeems the Nano can only display that 176x96 video at 43fps (according to the wiki)
00:50:54preglowwould anyone happen to know around how many frames per second a good mpeg2 encoder can do for something around 640x480?
00:51:14linuxstbThat might depend on your PC.... ;)
00:51:20 Join advcomp2019_ [0] (n=advcomp2@unaffiliated/advcomp2019)
00:51:24preglowsure, something averga, i just want a figure
00:52:25*amiconn just realised that there is a potential to save a bit of power in the bcm with partial updates
00:52:26 Quit advcomp2019 (Nick collision from services.)
00:52:39 Nick advcomp2019_ is now known as advcomp2019 (n=advcomp2@unaffiliated/advcomp2019)
00:52:40amiconnHmm, or not...
00:52:45linuxstbpreglow: I never really pay much attention to encoding times, I just set it going and do something else... It's been a while since I encoded videos as well.
00:53:30 Quit Zagor ("Client exiting")
00:53:38amiconnIn fact I could, but that's the next step
00:53:45 Join jmworx [0] (
00:54:27 Quit TotallyInfected (Remote closed the connection)
00:54:36amiconnRight now the driver fixes the update rectangle to the whole screen, in order to keep memory layout the same and allow writing in the background
00:54:51amiconnBut at least the y end address could be dynamic
00:55:05 Join TotallyInfected [0] (
00:55:27amiconnThing is, the driver needs to keep track of the maximum, and only reset it on a successful update
00:57:34 Quit mf0102 ("Verlassend")
00:58:19 Quit kugel ("ChatZilla [Firefox]")
00:58:57DerPapstNight all :)
00:59:01 Quit DerPapst ("So Long And Thanks For All The Fish!")
00:59:01linuxstbamiconn: Am I right in thinking the Ondio can't charge batteries, but can be powered via USB?
00:59:48linuxstbThat's handy - exactly the same as my DAB/MP3 player.
01:00:41amiconnIt's designed to run from primary cells
01:01:29 Quit jepler (Read error: 104 (Connection reset by peer))
01:01:34 Join jepler [0] (n=jepler@emc/developer/jepler)
01:02:26linuxstbRight, so charging would be a bad thing...
01:03:44 Join thegeek_ [0] (
01:08:24*preglow looks at mpegplayer.c and wonders if jhMikeS is sane
01:09:11 Quit homielowe (Read error: 110 (Connection timed out))
01:13:17linuxstbamiconn: Nice commit ;) We've come a long way since the original 8fps driver...
01:13:24***Saving seen data "./dancer.seen"
01:15:07pixelmaexcept a little bit of red...
01:15:36amiconneek, forgot that bootloaders are single core...
01:15:38*amiconn fix
01:17:07preglowlinuxstb: is that where ipodlinux is still at?
01:17:23preglow8fps, that is
01:17:32amiconnWouldn't surprise me...
01:17:33preglowi remember they guaranteed us we'd only get 1fps, with luck :P
01:17:39Dark_Apostrophepreglow: Any chance mpegplayer will get seek support in the foreseeable future?
01:17:48preglowDark_Apostrophe: don't ask mee
01:17:59Dark_ApostropheWho, then?
01:17:59linuxstbamiconn: Does it matter that there are no interrupts in the bootloader?
01:18:04preglowjhMikeS is working on it, i think
01:18:36 Quit ender` (" Eagles may soar, but weasels are seldom sucked into jet engines.")
01:19:49linuxstbYes, once he finishes his pre-emptive multi-core kernel...
01:20:18preglowhe's working on mpegplayer right now
01:20:37amiconnlinuxstb: Hmm, good question...
01:20:41*pixelma thought that was a joke...
01:20:47linuxstbSo did I...
01:20:51*jhMikeS watches in the shadows ;)
01:21:03 Quit thegeek (Read error: 110 (Connection timed out))
01:22:31linuxstbjhMikeS: What are you working on in the shadows? ;)
01:22:32midkayamiconn: o gawd.. how major is "major speedup of ipod video LCD driver"? :)
01:22:52amiconn2.5x for full screen, more for partial updates
01:23:05 Join homielowe [0] (
01:23:06midkayholy.. friggin' cow..
01:23:06pixelmasomething for "MajorChanges" then? :)
01:23:07amiconnLess for yuv blitting, but still quite noticeable
01:23:12midkayi gotta see this..
01:24:16jhMikeSlinuxstb: mpegplayer stuff...and some of this is the most boring programming I've ever done but I just focus on the possibilities.
01:25:22jhMikeSthe preemptive kernel is finished, I just have to type it in :p
01:25:52 Join Calcipher [0] (
01:26:00amiconnlinuxstb: I think for bootloader we need to resort to the good old wait-until-ready-and-update strategy, but even that will be faster than before (due to writing behind first, and then checking for the bcm to be ready and kicking it
01:26:31linuxstbamiconn: It's not an issue, as the bootloader only does a single lcd update IIUC.
01:26:41preglowis it just me or is mencoder a glitchy piece of shit?
01:26:51preglowi can't make anything it spits out play without glitches
01:27:15linuxstbI've found it generally OK, but haven't really used it much.
01:27:25linuxstbWhat are you using to play the files?
01:27:36amiconnmidkay: Look at , and then look again in a few minutes :)
01:28:08Calcipherhello all
01:28:22midkayamiconn: got it :)
01:29:21preglowlinuxstb: vlc
01:29:27linuxstbI blame vlc...
01:29:34preglowlinuxstb: so both are ffmpeg, really
01:29:46CalcipherI was wondering if there is any progress in the possible landscape mode for rockbox on the e200 series players
01:29:49linuxstbI didn't think vlc used ffmpeg, at least not for everything...
01:30:01LloreanCalcipher: Nobody's working on it.
01:30:23preglowlinuxstb: mpegplayer chokes on the files too, though
01:30:24jhMikeSpreglow: so...any closer to us getting a builtin voice codec?
01:30:58preglowjhMikeS: kinda, i guess, the main integration work is what'll be the most work, i'm just massaging the codec
01:31:50Calcipheroh, I see, is there anyone specific I can ask for help on this feature? I was thinking maybe the same process that was used in making the landscape mode patch for the Gigabeat could be used, but I don't know any of the technical details.
01:31:51jhMikeSpreglow: you think so? that's doesn't seem to difficult actually imo.
01:32:22LloreanCalcipher: The Gigabeat landscape patch is pretty much a hack unacceptable for inclusion, as it requires the use of two separate builds.
01:32:23preglowjhMikeS: i wouldn't know, really
01:32:29preglowjhMikeS: linus said he'd have a look at that part
01:32:58Calcipherah ok, thanks Llorean
01:32:59preglowjhMikeS: but anyway, the codec itself is really ready to go, i can always make it smaller and stuff, but nothing much will change
01:33:28linuxstbSo will the voice clips be raw speex, i.e. no Ogg?
01:33:30CalcipherI thought it was more of a officially supported feature
01:33:36preglowlinuxstb: i see no reason to use ogg
01:33:40linuxstbNor me.
01:33:49preglowthen we save some kilobytes there
01:34:03preglowand the only remaining source of mallocs...
01:34:06jhMikeSpreglow: The playback.c stuff isn't bad. The voice building changes...I have little knowledge about setting that up.
01:34:26preglowjhMikeS: if it's any worse than just changing from lame to speexenc, i'll be surprised
01:35:04CalcipherI'm also interested in the voice support functionality, so what exactly are the functions your all discussing atm?
01:35:27amiconnpreglow: It is. We cannot just switch, we need to provide both methods
01:35:28preglowdecent quality voice style encoded files are 458% realtime on pp now, so at least it should be fast enough
01:35:44preglowamiconn: both codec swapping _AND_ static codec?
01:35:49linuxstbCalcipher: You can read the IRC logs here -
01:35:53jhMikeSpreglow: 1) remove swapping 2) ?? 3) profit :p
01:36:10amiconnpreglow: No, selecting either lame or speexenc in the voice file build process
01:36:19preglowamiconn: oh, sure
01:36:35preglowthere's still the archoses
01:36:54preglowbut still doesn't sound like a huge problem to me, but then, i have never made a voice file
01:36:56Calcipherlinuxstb I'll take a look through the logs and see if I can figure out what your all up to.
01:39:23linuxstbHow does voice work on swcodec - is the voice file loaded permanently into RAM?
01:39:40linuxstbSo I guess it's not working on iFP...
01:39:45jhMikeSwell, not really
01:40:05preglowwell, not really? i thought it was loaded if it was found, voice ui enabled or not
01:40:36linuxstbAnd so voice won't work on my new player... (total 2MB RAM)
01:40:56 Quit hcs ("Leaving.")
01:41:01jhMikeSit's registered as "required".
01:41:14amiconnlinuxstb: It could be made to work like on archos on low-mem swcodec targets
01:41:31amiconnI.e. load the voice file into the main buffer while playback is stopped
01:41:42jhMikeSfrankly, the method is antiquated and a voice file could be dropped in and loaded without reboot without much modification.
01:41:51linuxstbAlthough couldn't we take advantage of the fact that low-mem swcodec are flash-based?
01:41:52amiconnThen voice will only be available without playback, but still better than no voice at all
01:42:12amiconnThat would be another option
01:42:18Dark_ApostropheWould any of the players that Rockbox runs on be fast enough to do speech synthesis?
01:42:29 Join JdGordon [0] (
01:42:31Dark_ApostropheIt would be nice, for listening to ebooks, articles etc.
01:42:35jmworxpreglow: do you know about liboggz?
01:42:52linuxstbDark_Apostrophe: There's a patch on flyspray to do text-to-speech.
01:42:55preglowjmworx: vaguely
01:42:56amiconnThe Ondio does something like this (loading clips as needed, but keeping the memory layout as if the whole file would be loaded at once)
01:42:56jmworxIt's another implementation of Ogg (very different API). It may be smaller
01:43:02Dark_Apostrophelinuxstb: Flyspray?
01:43:14linuxstbDark_Apostrophe: The patch tracker.
01:43:18amiconnThis is because Ondio has no seek or spinup penalty, but reads rather slow
01:43:29preglowjmworx: i'll check it out
01:43:30Dark_Apostrophelinuxstb: Does it work well? And how is it applied?
01:43:41linuxstbI've never tried it.
01:43:43amiconnLoading the whole voice file at once would take ~5 seconds, during which the device appears very unresponsive
01:46:00preglowDark_Apostrophe: voice files will be around anyway, since you can use better speech synthesizers on computers
01:46:09preglowor even, if you're quite mad, record you own voice clips
01:46:31Dark_ApostropheI'd just like to use it to get through some ebooks
01:46:39preglowDark_Apostrophe: well, you already can
01:46:44preglowthere is an espeak plugin
01:46:52preglowthat will read stuff aloud for you
01:46:55Dark_Apostrophethe Ludwig von Mises Institute almost weekly posts some 50-page article
01:46:56jhMikeSpreglow: I would never want to listen to myself speaking rockbox menus
01:47:02preglowjhMikeS: hell no
01:47:19preglowjhMikeS: but more importantly, i would never want to do the recording and editing
01:47:20Dark_Apostrophepreglow: Ooh, is it included in the default build of Rockbox?
01:47:23preglowDark_Apostrophe: nay
01:47:35Dark_ApostropheDoes it do PDFs?
01:47:39preglowDark_Apostrophe: nay
01:47:47preglowmaking it do pdfs would be excessively hard
01:47:48linuxstbStrike 2...
01:47:48Dark_ApostrophePlaintext, then?
01:47:55preglowpdfs are not made to be read
01:48:01preglownot by computers, anyway
01:48:09preglowDark_Apostrophe: it does .txt
01:48:13jhMikeSpreglow: tedius maximus
01:48:27Dark_Apostrophepreglow: good, would work for me :P
01:48:38Dark_Apostrophepreglow: I just copy/paste the text out of KPDF and into an editor
01:48:48Calcipherwow if actuall tts synthesis was implemented, it would be possible to have the wps info of current track read at a button press or something similar
01:48:53preglowjust watch out for funny glyphs
01:49:31*jhMikeS wonders why phonetic strings aren't used for the voice string so pronounciation is always correct
01:49:45preglowjhMikeS: please have a look at filters_arm4.S and recode the clipping stage of qmf_synth() for me, i also want those four trailing strh's made into an stmia :V
01:49:49linuxstbIIUC, that depends on the tts engine
01:49:57*preglow wishes he had code monkeys
01:50:46*linuxstb sleeps
01:50:48jhMikeSpreglow: mask+store is one thing. clip+mask+store I had real trouble making faster.
01:50:49 Quit J3TC- (Read error: 110 (Connection timed out))
01:50:50*Dark_Apostrophe attacks preglow with his arsenal of Æ, Ø, Å, œ, ë, and Ü
01:50:58CalcipherI was curious about the performance specs on some of those low end gps nav systems which use TTS engines
01:51:59Calcipherjust to have a better Idea of what might be possible in the TTS aspect when dealing with some of the daps
01:53:16preglowjhMikeS: it's easy making the clipped values end up pre-masked
01:53:24preglowbut that's doesn't really help when the majority are clipped, of course...
01:53:32CalcipherI'm googling some devices now, see if I can come up with any numbers
01:55:25jhMikeSpreglow: the best trick is orr all the samples together and clamp only if the result is out of range. YUV blitting does this and it does make a big difference
01:56:05jhMikeSIf the majority are clipped, to heck with it. Fix your file or turn your gain down.
01:56:12Calcipherouch, the tom tom 510 gps device, which I know uses a linux based TTS engine, has a 400mhz cpu, and 64mb of RAN
01:56:13preglowthe majority are not clipped
01:56:36jhMikeSpreglow: I'm saying that I hope not
01:56:46Calcipherso thats not very hopeful at all, when thinking of 90mhz daps
01:56:55 Join toffe82 [0] (
01:56:56Dark_ApostropheWhich Gameboy versions does Rockboy support? Gameboy Color? Advance?
01:57:23preglowjhMikeS: orring samples together aren't easy when they're sign-extended all the way up, and if i mess around too much to fix it, i lose all the speed advantage
01:57:30preglowjhMikeS: anyway, it's not a big deal, it's not even innerloop
01:57:36preglowi just hate the way it looks :>
01:57:48jhMikeSyou wont, you have r9 for scratch
01:58:06preglowi have tons of scratch
01:58:20preglowr5, r6, r7, r8, r9
02:00:08 Quit petur ("Zzzzz")
02:00:16safetydanDark_Apostrophe: that'll be in the manual
02:00:30safetydanCalcipher: there's also dedicated voice synthesis chips
02:00:40jhMikeSdoes it matter if it's 100% symmetrical, -0x7fff to 0x7fff vs. -0x8000 to 0x7fff
02:00:54Calcipheroh, had no idea thats how that was dealt with
02:00:59preglowjhMikeS: in that stage, no
02:01:06amiconnlinuxstb: I now have a version of the driver suitable for bootloader
02:01:18amiconnMaxes out at ~57fps, but no interrupts needed
02:01:21preglowjhMikeS: coldfire does it assymetrical for free
02:02:01jhMikeSok, I'm asking because then shifts can be used instead of cmp with +/-0x7fff
02:02:27Calciphersafetydan so in rb, having voice synthesis, does it seem possible with the kind of hardware we're dealing with?
02:02:44preglowwe're already doing it in the espeak plugin...
02:02:53safetydanCalcipher: well it's already possible as the espeak plugin demonstrates
02:02:59safetydanjust not likely to be high quality
02:03:27Calcipherdamn, I actually thought those were recorded clips, not actually synthesized
02:03:43safetydanI've no idea if anyone actually uses synthesis chips these days
02:03:46*amiconn ponders installing his shiny new bootloader on the G5.5
02:03:47Calcipherpretty amazing what you already have working
02:04:41Dark_Apostrophesafetydan: Finding anything in the wiki is quite a challenge
02:04:59Dark_ApostropheAs I've said before, it's nowhere near as easy to use as MediaWiki. :/
02:05:08safetydanDark_Apostrophe: the manual is not the wiki
02:05:25Dark_ApostropheOh, I thought you meant a Rockboy manual
02:05:55safetydanDark_Apostrophe: no, everything that comes with Rockbox is documented in The Manual, including plugins like Rockboy.
02:06:05Dark_ApostropheI see
02:06:16pixelmaDark_Apostrophe: the hard thing in searching the wiki is to find out that the box there is a "go" box and search is behind the "search" link...
02:06:22 Quit spiorf (Read error: 104 (Connection reset by peer))
02:06:31jhMikeSpreglow: why are the outputs shifted down first?
02:06:33Calcipherwould anyone here be willing to give me a little run through of how to use the rb debian VMware environment on windows XP, to transfer some linux files and test a TTS engine on the debian install?
02:06:51Dark_Apostrophepixelma: I don't know why I find it to be difficult, but I just do.
02:07:03 Join spiorf [0] (
02:07:32pixelmaDark_Apostrophe: I mean the wiki search link at the top, not the one in the left hand side menu
02:07:45CalcipherI was looking through the set up and can't find a way to access my files to do the test
02:07:53Dark_ApostropheMaybe that explains it
02:08:19preglowjhMikeS: fixed point format
02:08:31jhMikeSpreglow: that they're shifted down restricts the possible overshoot a great deal.
02:08:46 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
02:08:47preglowwell, yes
02:09:15jhMikeSsomething I'll take advantage of actually...hmmm
02:09:25preglowi was wondering about that
02:09:29preglowbut my brain just stopped working
02:09:32preglowit often does
02:10:41jhMikeShaha...I usually just don't try coding at those moments. I'm sure I'd get a good deal more done without them since that's alot of downtime for me.
02:11:09preglowi sometimes can't stop :>
02:11:39*preglow hates arm for being load-store
02:15:34jhMikeSmaybe in the _next_ version they'll fix that :)
02:16:08preglowyeah, give all the arithmetic instructions a _really_ flexible third operand :P
02:16:20preglowshift, load, const or free beer
02:19:04 Quit spiorf (Remote closed the connection)
02:19:35 Part pixelma
02:21:47preglowclosing in on bedtime
02:22:13preglowjhMikeS: if you end up with something cool, just leave me a message in the logs
02:22:30 Quit jhMikeS (Nick collision from services.)
02:22:36 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
02:22:44preglowshall i take that as a no? :>
02:23:30jhMikeShuh? free beer...yes. I did just get a disconnect though...logs...
02:24:18jhMikeSfree beer is definite "yes"
02:24:28preglowlet's see if arm goes that way, then
02:25:24jhMikeSoh, I'll leave a msg in the
02:25:32preglowi should recode those iir_mem16 as df1 asap...
02:25:40*preglow tears hands from keyboard
02:27:10 Join kkurbjun [0] (
02:29:32 Join Zom_ [0] (
02:30:08 Quit Zom (Nick collision from services.)
02:30:14 Nick Zom_ is now known as Zom (
02:31:32 Join makuseru [0] (
02:33:47makuserui recently put some video on my Toshiba Gigabeat, and some of the video runs fine, but some of the other ones play at like x2 or x4 speed, what causes this?
02:41:27TTThomasmakuseru: did you convert all the files the same way?
02:42:00TTThomasI've done dozens on my f40 with no issues
02:42:21TTThomasbut all with vlc
02:42:40makuserui didnt convert them
02:42:46makuserui just put them on there
02:45:06TTThomasthey have to be a certain aspect ratio and video/audio bitrates for the best playback. that link has the settings you want and links to different programs to do it with.
02:45:29 Quit HellDragon (Client Quit)
02:46:49 Join HellDragon [0] (n=Nocebo@unaffiliated/helldragon)
02:47:05makuserualso, i have all my album art saved with amarok, how can i get that to show on the Gigabeat?
02:47:40TTThomasyou need a patched version for album art, plus a theme that supports album art
02:47:44TTThomas1 sec
02:47:53makuseruoh, ok
02:48:32 Join eigma [0] (n=cat@
02:48:40eigmakkurbjun: hey
02:48:44kkurbjunhey eigma
02:48:46TTThomasthat is a list of gigabeat themes.
02:49:00eigmaI should be able to repro with SVN?
02:49:08kkurbjunyeah, SVN will work
02:49:23eigmanormal build, anything else I should do?
02:49:32TTThomasyou'll have to readup on patching the build for album art, I've never done it
02:49:42kkurbjunI have some other changes, but I don't think they will effect this, normal build is fine and then just load up the elf
02:49:51 Quit kubrick ("leaving")
02:51:04kkurbjunthe setting I typically check is to go to settings->sound settings->volume
02:51:12eigmainput device.. remote or ts?
02:51:26kkurbjunand it should show a list of volumes in decibles, doesn't matter either will display it
02:51:48eigmahow do I use ts? do I just point at an item or left are is "left key" etc?
02:51:51eigma*left area
02:52:01kkurbjunjust click on an option
02:52:06eigmawow, nice
02:52:20kkurbjun:), jdgordon did a great job with it
02:53:08kkurbjunof course the existing ts code you guys made helped quite a bit too ;)
02:53:59eigmabuild/apps/rockbox.elf, right?
02:54:02TTThomasmakuseru: the above link is the wiki on album art
02:54:07kkurbjunyep, that's the one
02:55:04eigmayeesh, it really takes a long time to load via jtag
02:55:18kkurbjunJdGordon: he's loading via jtag with gdb
02:55:29JdGordonok then
02:55:55eigmahere goes nothin'.. "c"
02:56:48eigmaI see a scrollbar on the left
02:56:50eigmano items
02:56:59eigmabut I can scroll down, and I can see the scrollbar go down
02:57:09kkurbjunok, so that means that it's probably not an initialization problem
02:57:09eigmaso I'm guessing the drawing/clipping area for the items is screwed up
02:57:16JdGordonthats the proble we need fixed..
02:57:43eigmaI'll try to familiarize myself with the drawing system quickly, try to live debug this for you
02:58:15kkurbjunI have some more pointers here in a second
02:58:50kkurbjunso I found a way to make it work on my player, it doesn't work for JdGordon
02:59:00 Join psycho_maniac [0] (
02:59:40psycho_maniacdo i need to update my bootloader on my ipod video for any of the recent svn changes?
02:59:50JdGordon... and really, even if it did fix, I wouldnt want to comit it.. its not ery nice :p
03:00:00kkurbjunif you modify list.c and take the buffer on line 317 and move it to line 171 it seems to work
03:00:12 Join kubrick [0] (n=repulse@unaffiliated/funky)
03:00:12kkurbjunwhich takes that buffer off the stack
03:00:25eigmafull path please?
03:00:40 Join oblib [0] (
03:00:51kkurbjunso I thought it was a stack overflow problem
03:01:27eigmaon entry to gui_list_draw_smart, sp is 0x2a84
03:01:35eigmalet's see how much that is....
03:01:53kkurbjunbut I've tried reducing that buffer size and that didn't help. I also looked at the stack memory and it appears that there's more than enough deadbeef
03:02:40kkurbjunstack should be defined in
03:03:14kkurbjunthat seems like a strange location though..
03:03:25kkurbjunit's 0x2000 bytes
03:03:27eigmayeah.... :S
03:03:56kkurbjuntotal size, and I think it's head is lower than that
03:03:58eigmahow do I get the current thread_entry?
03:04:39eigmabah, I have to recompile with -g
03:04:51eigmathat'll take another eternity to load
03:05:15JdGordonthreads[0] should be the main one i think
03:05:29eigmacores[0].running is what I came up with
03:05:29JdGordoni.e the one doing the drawing...
03:06:32JdGordonthread_stack_usage(&threads[0]) should give the stack usage of main
03:06:56JdGordonthreads[0].name should be "main" or your looking at the wrong one
03:11:29eigmasure enough, cores[0].running->name is "main"
03:11:53eigmastack_size = 8192 and stack = 0xc44
03:12:53kkurbjunstack size looks right
03:13:03JdGordonI dont think its a stack overflow though... why would the title, icons and scroll bar all draw fine but the text not draw?
03:13:25***Saving seen data "./dancer.seen"
03:13:33eigmaso it's just text? I don't see any icons?
03:13:55kkurbjunthe menu you're in I don't think shows icons
03:14:04JdGordonyeah, the main menu has icons
03:14:05 Join teenPSPjunkie [0] (
03:14:07kkurbjunif you go to the file browser it should show them
03:14:18eigmaoh, I'm in volume
03:14:25JdGordonyeah, no icons in volume
03:14:29oblibHi, I'm trying to resync my svn after a couple of months, and I'm getting some compile errors: pcm-pp.c:(.text+0x18): multiple definition of `pcm_apply_settings'. It looks to me like there are two definitions of pcm_apply_settings, one in plugin.c and one in pcm-pp.c. What do I have wrong?
03:14:30eigmaok lemme step by step this silly thing
03:17:49kkurbjunactually I guess 2a84 makes sense
03:18:03kkurbjunit should be starting around 2c44
03:18:03 Quit makuseru (Read error: 104 (Connection reset by peer))
03:18:59 Join darkapostrophe [0] (
03:19:55kkurbjunI have to go to dinner guys, I'll try and stop back in later, sorry I have to leave in the middle of this :(. I really appreciate the help Cat - we've been stuck on this for a while now.
03:20:49eigmaenjoy dinner
03:21:58 Join J3TC- [0] (
03:23:39 Nick fxb is now known as fxb__ (
03:23:51*JdGordon back
03:24:30eigmaP2STR seems to be broken
03:24:59JdGordoncant be...
03:25:09eigmas is (unsigned char *) 0x2984 "6 dB", while entry_name is (unsigned char *) 0xb63cb63c ""
03:25:10 Part vladsinger
03:25:33 Quit animeloe ("This computer has gone to sleep")
03:25:55 Join bal_zac [0] (
03:26:44JdGordonthat doesnt really make sense though... ALL the text from the menus goes through that correctly
03:27:09eigmayeah, entry_name definitely has garbage
03:27:26eigmawhere is P2STR defined?
03:27:35JdGordonumm.. ill find it
03:29:17eigmais this architecture the only one where there's RAM at 0x00000000?
03:29:49eigmathe VIRT_PTR convention assumes there will be no strings on the stack (more generally, in IRAM)
03:30:07 Quit BigBambi (Read error: 110 (Connection timed out))
03:30:26eigmayou see what I'm saying?
03:30:39JdGordonyeah, i tihnk so
03:30:52JdGordonso VIRT_PTR needs moving?
03:31:08eigmapossibly. I haven't thought out all the side effects of that though.
03:31:15eigmaI can "just try it" heh
03:31:48JdGordongoing by, it looks like yeah this is the only one where iram is at 00
03:32:53eigmaI'll try VIRT_PTR at 0x4000
03:34:04jhMikeSodlib: I'd expect somehow you still have pcm_playback.c around or something which is way gone
03:34:15eigmaJdGordon: but there's plats that have RAM there..
03:34:24eigmaahh, this is such an ugly hack
03:34:50JdGordonwell we have 64mb ram to play with right?
03:34:55 Quit bal_zac ("Leaving")
03:35:03eigma?I don't
03:35:11JdGordonisnt it 64mb?
03:35:21eigmayeah but what does that have to do with VIRT_PTR
03:35:22oblibjhMikeS, thanks, I needed a build clean
03:35:39JdGordonanyway i was tihnking we can fudge virtptr like its done for the sim.. give it an actual buffer...
03:36:09 Quit Dark_Apostrophe (Connection timed out)
03:36:09 Nick darkapostrophe is now known as Dark_Apostrophe (
03:36:16eigmawell, then we're special-casing for DM320. we might as well just find an unmapped region for DM320 if we're gonna special case
03:36:17 Quit kclaf (SendQ exceeded)
03:36:43eigmaanyway, I'll try 0x4000 to confirm that this is hte problem at least
03:37:20JdGordonok, well i dont think moving VIRT_PTR is a bad hack if it fixes it...
03:38:15eigmaI think a good cross-plat value for VIRT_PTR is 0x8000000.. but I have a hunch it will gratly increase code size because ARM-based archs will need more than 4 bytes to load an immediate into a registry
03:38:36jmworxCalcipher, preglow: Have you considered doing minimal TTS based on Speex? That is, use "in context" word concatenation.
03:38:50eigma*0x80000000 (i.e. using the MSB as a flag)
03:39:14JdGordonna, it doesnt really _have_ to be the same value for all tagrte
03:39:31eigmarockbox is pretty littered with #ifdef's as it is..
03:40:27eigmaahh jtag loading takes sooo long :(
03:40:37eigmaI think I might get one of those USB JTAG accelerators
03:41:27 Join BigBambi_ [0] (n=Alex@rockbox/staff/BigBambi)
03:42:05teenPSPjunkieHey I'm new to this whole rockbox thing. I have a 30gb Ipod 5gen and I jsut installed rockbox using the ultility but I cant seem to get rockbox ot start when I restart my Ipod it jsut goes to the apple firmware
03:42:38eigmait works :)
03:42:52psycho_maniacteenPSPjunkie: you install the bootloader AND the .rockbox folder?
03:43:21Galoisthe ipod is not so easy to turn off
03:43:25JdGordonall hail eigma!
03:43:37Galoisif you just turn it off normally, it goes into sleep mode
03:44:14teenPSPjunkieUmm I thought the ulitily installed both if it does then yes I have
03:44:21eigmahmmm, I wonder if we can remap the IRAM somewhere else and make the first page be unmapped like everyone expects
03:44:52JdGordondoes it really matter?
03:45:09JdGordonas long as there is no data after 0x4000 for 0xffff bytes
03:45:11eigmait would be "nice"... :)
03:45:13psycho_maniacwell it installs both but at two different times.
03:45:22eigmaanyway, I'll check in the fix in a sec
03:45:56teenPSPjunkieI used it and it said my boot loader was already up to date so I jsut went onto the second part
03:46:37psycho_maniacwas this your first time installing rockbox?
03:47:04JdGordoneigma: had any luck with the backlight ?
03:47:21eigmahaven't touched it
03:47:25eigmamaybe I'll play with it a bit tonight
03:48:29psycho_maniacare you going to be using this player for rockbox mainly?
03:49:19JdGordonok.. looks like im going to have to get a move on the touch screen again seen as its now useable :)
03:50:01oblibDoes anyone know off the top of their head where in the code the "feature" of clearing settings during power-on is? I hate it when I hit hold too late.
03:50:16JdGordonoblib: apps/main.c
03:50:26teenPSPjunkieI jsut wanted ot see if rockbox was better than the original firmware and if I like it then ya it will be mainly what im using
03:50:35eigmamy question is, why didn't this bug give us a data abort?
03:50:46psycho_maniactry installing it again maybe?
03:51:18JdGordoneigma: why would it? didnt you say 0 was ram anyway?
03:51:43teenPSPjunkieOk ill try it and Ill tell you if it works this time
03:52:05 Quit kubrick (Read error: 113 (No route to host))
03:53:53teenPSPjunkieI did not skip the bootloader this time and it says unsupported install method
03:55:03eigmaJdGordon: yeah but that's before P2STR. P2STR was indexing into a bogus table and returning a pointer from outer space
03:55:14psycho_maniactry installing it with this app. same one just older and seems to work. last person i gave this link to that was having problems got this to work
03:56:59 Quit john (Read error: 113 (No route to host))
03:59:17eigmaJdGordon: what would be the best way for me to make a really small build that just takes input from the remote and can run certain functions, so I develop a BL driver?
03:59:34JdGordondo a bootloader build
03:59:47JdGordonand modify bootloader/mrobe500.c
04:00:05JdGordonthats bassically our testing program
04:02:24teenPSPjunkiethanks psycho It works now
04:03:02psycho_maniacI RULE !
04:06:26 Quit SirFunk (Remote closed the connection)
04:08:20 Join Cliffer [0] (
04:09:17 Quit teenPSPjunkie (Read error: 104 (Connection reset by peer))
04:10:51eigmapsycho: lol
04:12:15 Quit Cliffer (Client Quit)
04:12:35 Join z35 [0] (
04:14:49 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
04:15:21psycho_maniacanybody use the theme x glow?
04:16:47eigmaJdGordon: in firmware/target/arm/tms*/spi-dm320.c, is there any reason to keep the set/clr reg as "reg(0x0594)" or can I change them to symbolic names?
04:19:18 Quit Spliffer (Read error: 110 (Connection timed out))
04:21:33Calcipherjmworx I'm not familiar with TTS enough, I would need to read up on it
04:21:55 Quit Thundercloud (Remote closed the connection)
04:22:07 Quit animeloe ("Leaving")
04:22:56 Quit donutman25 ("ChatZilla [Firefox]")
04:24:21 Quit hcs ("Leaving.")
04:27:33 Quit midgey ()
04:33:49 Quit karashata ("Leaving.")
04:34:32 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
04:40:20JdGordoneigma: symbolic names are prob better
04:40:45eigmaBL is incredibly simple :)
04:41:15oblibI'm trying to do some string work in playlist.c, but sprintf and itoa don't seem to be included. My C skills are failing me: how do I get an int to a string without those two functions?
04:41:26 Quit BigBambi_ (Read error: 113 (No route to host))
04:41:28JdGordoneigma: BL = backlight or bootloader?
04:41:37eigmachip select is GIO 29, active low.. assert it, then spi write 0xa400, 0x__bb where bb is between 00h and 55h to adjust the level
04:41:48eigma*where __
04:41:55JdGordonoblib: snprintf is there
04:42:07safetydanoblib: snprintf
04:44:22 Join sarixe [0] (
04:44:28oblibgreat, thanks. Much easier than array manipulations!
04:44:44eigmawhat's the range of the brightness argument to __backlight_set_brightness?
04:45:20eigmaoh! I guess we get to define it
04:52:17 Join BigBambi_ [0] (n=Alex@rockbox/staff/BigBambi)
04:52:31 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
04:59:27 Quit sarixe ("(EE) Failed to load "quit" module")
05:01:15JdGordoneigma: yeah, in config-target.h i tihnk
05:04:52 Quit keanu ("Leaving")
05:07:45 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
05:07:55 Join miepchen^schlaf [0] (
05:12:29 Join localhero [0] (n=dinzdale@
05:13:28***Saving seen data "./dancer.seen"
05:14:54 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
05:16:15 Join bb__ [0] (
05:25:14 Join ptw419 [0] (
05:28:23 Quit bb (Read error: 110 (Connection timed out))
05:28:44 Quit oblib ("Leaving")
05:31:48eigmaJdGordon: it works!!!!
05:46:09eigmaJdGordon: pin
05:46:15eigmaJdGordon: ping
05:52:12kkurbjunhey eigma
05:52:32eigmaI got the BL working.. but it required that I fiddle with the SPI settings
05:52:38eigmait *looks like* TS is still working
05:52:49eigmabut I'm really not sure, I was gonna get you to try it too
05:52:52eigmacan you load bootloader builds?
05:53:14kkurbjunyou figured out that bug?, that's great. yeah, I can load a bootloader build, you want to commit it and I'll try it
05:53:25eigmaokay, one sec
05:53:30eigmaoh the menu bug, yes
05:53:35eigmaI committed a fix for that
05:54:13kkurbjunyeah, I saw that.. I wonder how we could make it cleaner.. I wonder why that bug doesn't show on more targets
05:54:32eigmahehe yeah see I tought the #ifdef was ugly too
05:54:53eigmano other archs have RAM below 0xffff, do they?
05:55:08kkurbjunthe gigabeat remaps everything to 0x0
05:55:22kkurbjunand compiles as if it's all running there
05:55:31kkurbjunthe mmu does the remap
05:56:37*JdGordon back
05:56:43JdGordonand still doesnt thinkk that ifdef is bad
05:56:49 Quit BigBambi_ (Read error: 113 (No route to host))
05:56:49kkurbjunI don't mind ifdef's too much if they are in a central location, that way you can just go to one file and say ahh, that's what all I need to do for this target
05:57:11eigmathe DM320 can remap the AIM to 0xffff0000, right?
05:57:24eigmathe datasheet was a bit cryptic, I was hoping you'd have a better handle on this
05:57:45kkurbjunembedded ifdefs are awful. it can be remaped but I think it depends on what some pins are driven to
05:59:10kkurbjunand I think it's only a mapping of 0x4000 or 0x0
05:59:27kkurbjunwe can also remap it with the mmu
05:59:54kkurbjun:), the mmu is enabled now
05:59:56eigmaI would rather fight my conscience about an #ifdef than the ARM reference manual on the MMU
06:00:20kkurbjunyou can't cache data without it enabled
06:00:29eigmaI didn't know that :S
06:00:34kkurbjuneverything is mapped on itself
06:01:04kkurbjunI didn't know it either.. I just happened to come across it when I was looking through the coprocessor registers
06:01:38kkurbjunnice thing is the mmu is the same on arm v4 and v5 so the gigabeat and mrobe can share it
06:02:00 Quit hcs ("Leaving.")
06:02:12 Quit localhero ("goodnight!")
06:02:42kkurbjunwe can only remap 1 meg segments right now with the mmu driver at the moment
06:02:53kkurbjunwhich is fine for our needs
06:03:05kkurbjunI would rather avoid level2 and 3 lookups
06:05:23JdGordonkkurbjun: have you done anything else about making it portait>?
06:05:33eigmathat would be reeeeeeally nice
06:05:50JdGordonand by portait i mean landscape :p
06:06:01 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
06:06:04eigmaI can't believe they wired up the LCD the way they did
06:06:11kkurbjun:) yeah, I have some code for that
06:06:18kkurbjunI need to commit my changes
06:06:26kkurbjunI have a bunch more plugins building too
06:06:52JdGordoneither of you know anything about
06:07:12JdGordoni had a look at that whole bit of the datasheet is beyond me :p
06:07:28eigmaas in, digital to analog?
06:07:50JdGordonthe thingy so we can get sound going :p
06:08:18eigmait's not too hard.. set the sampling rate and input format, set the volume, and start shovelling samples
06:08:29eigmaall that frequency characteristics crap can largely be ignored imho
06:09:00eigmakkurbjun: 'up' and lemme know if bootloader build still has ts functionality for you.. also, volume up/down should let you choose between 2 backlight settings
06:09:14 Join Bryant [0] (
06:09:15 Join BigBambi_ [0] (n=Alex@rockbox/staff/BigBambi)
06:11:14kkurbjun:), doom runs now
06:11:22JdGordonhaha cool
06:11:27JdGordonat decent frame rate?
06:11:54kkurbjunno, it's pretty slow, looks like I'll have to spend some time getting it to run at reasonable speeds again
06:12:14kkurbjunit's playable, just a bit choppier than I'd like
06:12:50JdGordoneigma: it works.. but i get a high pitch squeel if i change to the low brightness
06:12:57eigmasame happens with OF
06:13:12eigmacan you confirm?
06:13:27eigmait should be equivalent to the lowest setting in OF
06:13:28JdGordonill have a lookl
06:14:12JdGordonyep.. its in the OF
06:14:19JdGordonthats pretty crap
06:14:38eigmawon't be an issue in most cases, I think
06:14:50eigmaunless it leaks into the headphone signal, which I don't think it did for me
06:16:48JdGordoni dont think it is here either
06:17:25kkurbjuneigma, nice work backlight works great and the touch screen is working well too
06:17:39eigmanow look at what I changed in spi
06:21:47kkurbjunjust synced my tree too
06:21:58 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
06:22:26kkurbjundoom builds and it can run with a hack, but I need to fix the user timer so that it works properly
06:23:08 Join amin [0] (
06:24:07eigmawow, you can get the BL so dim that it flickers
06:24:35kkurbjunI'm going to bed, I'll talk to you guys later :) nice
06:24:57amini am confused.. can some one please clear my confusions. Regardless of going through many articles I still have a Question: What's the diff b/w rockbox and linux ? how do I install rockbox in accordance with the firmware ? I have no idea what I have to do with the firmware in order to get rockbox on my ipod30Gb video.. and how do I exactly know the firmware of my iPOD ? infact what firmware actually is ?
06:25:37scorchefor all besides the first question, please read our fine manual
06:26:03scorchefor the first, see
06:27:32aminso RockBox is an OS for ipod ?
06:27:32lostlogicamin: linux is a kernel, written originally by Linus Torvalds and designed to run originally on personal computer hardware. As a kernel, it doesn't do much, just talks to the hardware. Rockbox includes the Rockbox kernel and all of the application software needed to run a Digital Audio Player. The Rockbox Kernel was written by some crazy swedes to run on digital audio player hardware originally.
06:27:36lostlogicRockbox is an OS for many digital audio players
06:28:04 Quit Mouser_X ("CGI:IRC (EOF)")
06:28:04lostlogicwhen talking about portable electronics firmware is often used synonymously with OS
06:28:22eigma"firmware" is preferred, I think?
06:28:40lostlogiceigma: dunno, it's more common for sure...
06:28:42aminso lostlogic are you saying firmware=OS ?
06:28:43hcsSummary: Rockbox - written by crazy Swedes ; Linux - written by crazy Finn
06:28:56lostlogichcs: that's my thought h:)
06:29:04JdGordonthere is obviously someting in the water up there...
06:29:12lostlogicJdGordon: I've often wondered.
06:29:28 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
06:29:30 Join ddalton [0] (
06:30:03 Part ddalton
06:30:08amindoes ROCKBOX plays music vdos of all format or particularly of mp4 formats ? and can I play games on rockbox for free ? Can I change themes too ?
06:30:08safetydanamin: essentially
06:30:29Lloreanamin: Those questions are all answered in the manual
06:30:42aminLlorean can I please have the link of the manual ?
06:30:59Lloreanamin: If you go to, it's in the bar on the left...
06:31:02 Quit Mouser_X (Client Quit)
06:31:36eigmaJdGordon: you think set_backlight_brightness should provide the full swing of values that the hardware can do, or only those values that actually give meaningful, visible, brightness variations (though, I suppose that's relative depending on ambient light levels, etc.)
06:31:59JdGordongo the full set... there is no real difference...
06:32:04JdGordonwhat was the range again?
06:32:08eigmayay, that's easier for me :)
06:32:27 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
06:32:28eigmafull range is 0 - 127.. meaningful is about 30 levels centered aroud 50
06:33:21aminI am still confused.. shall I go for rockbox or linux ? which one has more games/themes ?
06:33:21 Quit homielowe (Read error: 104 (Connection reset by peer))
06:34:08JdGordonamin: if you want games go linux, if you want to use it for music go rockbox
06:34:24aminJdGordon I want both ?
06:34:31psycho_maniacbut rockbox has games?
06:34:31Mouser_XI guess I got the wrong OS then!
06:34:49eigmacopy half the files for Linux and half the files for rockbox.. you might get lucky
06:35:00 Join homielowe [0] (
06:35:18 Quit linuxstb (Read error: 113 (No route to host))
06:35:51 Nick amin is now known as amyn (
06:35:59 Nick amyn is now known as ameeen (
06:36:23lostlogiceigma: play nice
06:39:34ameeen<ameeen> dude pls help me before I do something stupid with my ipod
06:39:34ameeen<ameeen> isn't linux good at music ? I mean what problems can I face while using linux ?
06:39:51 Quit Calcipher ("I-n-v-i-s-i-o-n 2.0 Build 3515 with A Pack Fix By")
06:40:43 Quit hannesd (Read error: 104 (Connection reset by peer))
06:40:43eigmaI don't own an ipod, so i've never put Linux or rockbox on it −− but I would suspect (hope?) that it's a relatively safe and reversible procedure. try each one and see which one you like the most.
06:41:54safetydanameeen: either Rockbox or iPod Linux should be reasonably safe to use on an iPod. If you really can't decide, you can install both and triple-boot between the Apple firmware, Linux, and Rockbox
06:41:57ameeendo they have manuals for linux as-well ?
06:42:12homielowe ameen: this is a rockbox irc channel, for linux please go to #ipodlinux
06:42:15 Join hannesd [0] (
06:45:10ameeenhow to know which gen of ipod I own ?
06:45:29 Quit Mouser_X ("CGI:IRC 0.5.9 (2006/06/06)")
06:46:06eigmaare there any menus in the UI to control the brightness of the screen?
06:46:41 Quit Bryant ("ChatZilla [Firefox]")
06:46:59 Quit psycho_maniac (" HydraIRC -> <- Wibbly Wobbly IRC")
06:47:43JdGordoneigma: yes
06:47:50JdGordonsettings > lcd settings > brightness
06:48:01eigmaI just finished writing backlight-mr500.c
06:48:05eigmahopefully it works :)
06:48:27JdGordonas long as the defines are set correctly it should work
06:48:40eigmaany defines other than DEFAULT_BRIGHTNESS et al.?
06:48:48 Join Spiritsoulx [0] (n=eyes_of_@
06:48:55eigmaare the HAVE_'s set properly, do you know?
06:50:26JdGordoni doubt it
06:51:08JdGordonthey are defined in config-mr500.h as expected
06:51:15JdGordonbut you probably need ot change those values
06:52:45eigmaUGH "region PLUGIN_RAM is full"
06:54:45 Quit Spiritsoulx ()
06:57:21ameeenafter installing rockbox/linux would I be able to access all the music/images/videos that I could access from my Apple OS ?
06:58:33safetydanameeen: yes though Rockbox will not be able to play files you've bought from iTunes
06:59:49 Quit jhulst ("Konversation terminated!")
07:06:21 Quit ameeen (Read error: 104 (Connection reset by peer))
07:06:27 Quit BigBambi_ (Read error: 113 (No route to host))
07:12:23eigmaJdGordon: yep, backlight works great. even the 5s timeout worked right away.
07:13:32***Saving seen data "./dancer.seen"
07:13:51eigmachecked in
07:13:55eigmaaaaaaand I'm out for the night
07:14:13eigmasorry I haven't helped out much lately.. maybe I'll get some time and get back into the spirit soon :)
07:14:28JdGordonwe forgive you :D
07:15:08 Quit eigma ()
07:17:19 Join det [0] (
07:18:18 Join advcomp2019_ [0] (n=advcomp2@unaffiliated/advcomp2019)
07:18:26 Quit advcomp2019 (Nick collision from services.)
07:18:32 Nick advcomp2019_ is now known as advcomp2019 (n=advcomp2@unaffiliated/advcomp2019)
07:20:09 Join BigBambi_ [0] (n=Alex@rockbox/staff/BigBambi)
07:26:48lostlogichcs: around?
07:30:01lostlogicanyone else want to test a playback patch?
07:33:04 Part safetydan
07:38:15 Part toffe82
07:40:12 Nick fxb__ is now known as fxb (
07:41:09 Nick fxb is now known as fxb__ (
07:51:34 Join darkapostrophe [0] (
07:54:32amiconnkkurbjun: Red builds!
07:54:39 Quit ptw419 ()
08:01:42lostlogiclinuxstb_: Nico_P: better buffer type handling, what yous think?
08:07:12 Join karashata [0] (
08:08:37 Quit Dark_Apostrophe (Connection timed out)
08:08:37 Nick darkapostrophe is now known as Dark_Apostrophe (
08:10:31 Quit BigBambi_ (Read error: 110 (Connection timed out))
08:20:03 Quit freqmod__nx (Remote closed the connection)
08:21:23 Quit atsea-39 (Remote closed the connection)
08:21:37 Join BigBambi_ [0] (n=Alex@rockbox/staff/BigBambi)
08:22:17 Join Rob222241 [0] (
08:22:32 Join atsea-39 [0] (i=atsea-@gateway/tor/x-5eaeca7b36562695)
08:26:13 Join nicksola [0] (
08:27:45nicksolaI just upgraded to the daily build for my 1g ipod mini. It was working fine until I did this. Is it known to stop the ipod working?
08:28:48 Quit JdGordon ("Konversation terminated!")
08:35:28nicksolaOh, you have this new utility now.
08:35:33nicksolaI'll try that.
08:35:38nicksolaThanks for all your help.
08:35:40 Part nicksola
08:37:13 Join daurn [0] (i=daurn@unaffiliated/daurnimator)
08:37:58 Quit jhMikeS (Nick collision from services.)
08:38:00 Join _jhMikeS_ [0] (n=jethead7@rockbox/developer/jhMikeS)
08:39:58 Quit Rob2222 (Read error: 110 (Connection timed out))
08:44:51 Join ddalton [0] (
08:45:11ddaltonscorche: around?
08:45:35scorcheyes, but i was just off to bed
08:46:02ddaltonok where were you saying you could get a sansa from? you said $170 is a bit much ?
08:46:08ddaltongot a link?
08:46:22ddaltoni've tried google
08:46:29scorchewell, this is a bit offtopic, but it was at
08:47:26ddaltonoh ok and can that do shipping?
08:47:37 Quit in-jane (Read error: 110 (Connection timed out))
08:47:50ddaltonwhys it OT? it is releated to mp3 players
08:47:57 Quit daurnimator (Connection timed out)
08:48:05scorcheit isnt related to rockbox
08:48:18ddaltonthe player will run rb though. But whatever
08:49:05scorcheit is an online store...i dont see how they could work without doing shipping
08:49:12ddaltoncan they post it to me
08:50:50ddaltonoh free shipping
08:50:58ddaltonand whats it in us $?
08:51:15scorchethey dont ship abroad
08:51:34ddaltonthey can't ship to aust?
08:51:57scorchewhy are you asking me this?
08:52:02scorcheask them
08:52:14ddaltonok sorry didn't think before I wrote it
08:55:36ddaltonwow a bit much:
09:02:19ddaltonis the m240 4 gb? anyone think this would be a good player for rb? would it take much work to port rb to the m200?
09:02:30ddaltonor is it not like the other sansa
09:10:35 Join ender` [0] (
09:12:04ddaltonscorche: how does this sound in aust $? e250 $123, e260 $ 155 and e280 $ 201. Not sure about the e270
09:13:34***Saving seen data "./dancer.seen"
09:16:45 Part ddalton
09:19:23 Join petur [0] (n=petur@rockbox/developer/petur)
09:20:04 Quit BigBambi_ (Read error: 113 (No route to host))
09:23:27 Quit karashata ("Leaving.")
09:28:09 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
09:28:39 Join davina [0] (
09:30:14linuxstb_lostlogic: Regarding your patch, you don't think it's a good idea for the bufread function to block?
09:30:19 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
09:31:59 Join Buschel [0] (
09:32:40 Join BigBambi_ [0] (n=Alex@rockbox/staff/BigBambi)
09:34:00 Join Zagor [0] (
09:35:29*Zagor is tired of kernel oopses
09:35:59linuxstbYes, it seems the Linux kernel isn't very tolerant of USB devices misbehaving...
09:36:43Zagorthe good news is they're chasing the bug
09:42:05 Quit TotallyInfected (Remote closed the connection)
09:42:46 Join TotallyInfected [0] (
10:00:09*linuxstb tries the mrobe sim for the first time and didn't expect it to be portrait
10:00:24GodEater_do you even own one ? :)
10:00:45linuxstbI own a sim ;)
10:01:19linuxstbBut based on this page I assumed it would be landscape -
10:01:35linuxstbGodEater_: Where are you?
10:02:04 Quit nanok (Read error: 104 (Connection reset by peer))
10:02:18*linuxstb also thought the m:robe sim used mouse input, but can't make his mouse do anything
10:02:48 Join nanok [0] (n=nanok@
10:03:52ZagorI too thought it would be landscape
10:04:36Zagorbtw does anyone know what the strange m:robe name is supposed to mean/convey?
10:05:23linuxstb"The name m:robe is a contraction of Music wardROBE."
10:05:34GodEater_linuxstb: back at work now
10:05:38linuxstb(according to wikipedia at least...)
10:05:44 Join spiorf [0] (
10:06:23linuxstbI wish I didn't know what m:robe meant now...
10:06:35Zagorlinuxstb: music wardrobe. that's just ... odd
10:06:59*linuxstb wonders where Olympus are from
10:07:07GodEater_Japan aren't they ?
10:07:20 Join pondlife [0] (n=Steve@rockbox/developer/pondlife)
10:07:50*homielowe accepts japanese for who they are, even if he thinks they have odd naming policies
10:08:42 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
10:14:21 Join karashata [0] (
10:15:39GodEater_goodness me - I have a lot of forum reading to do to catch up =/
10:15:51linuxstbOr not... Just click on "mark all as read" ;)
10:16:37*GodEater_ does that on a forum by forum basis ;)
10:23:54 Quit advcomp2019 ("Leaving")
10:28:15 Join advcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
10:29:43 Quit homielowe (Read error: 110 (Connection timed out))
10:31:13GodEater_any news on MoB being commited yet ?
10:31:49linuxstbWasn't that committed before you left?
10:32:11 Join in-jane [0] (i=jane@
10:32:34linuxstbIt's been in SVN for what seems like a couple of weeks.
10:36:52 Join advcomp2019_ [0] (n=advcomp2@unaffiliated/advcomp2019)
10:38:00 Quit advcomp2019 (Nick collision from services.)
10:38:08 Nick advcomp2019_ is now known as advcomp2019 (n=advcomp2@unaffiliated/advcomp2019)
10:38:09 Quit billytwowilly (Read error: 104 (Connection reset by peer))
10:38:24 Join billytwowilly [0] (
10:40:20 Quit TotallyInfected ()
10:42:02pixelmathat's a bit exaggerated - MoB was committed last thursday
10:43:08linuxstbIt seems longer ;)
10:43:41pixelmaof course :)
10:47:58 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
10:48:48 Join B4gder [0] (n=daniel@rockbox/developer/bagder)
10:49:05 Quit TMM (Read error: 110 (Connection timed out))
10:50:32 Quit BigBambi_ (Read error: 113 (No route to host))
10:53:35GodEater_I missed that then - I flew on Thursday last week
10:53:49GodEater_has the crossfade issue been sorted then ?
10:55:00 Join jumpatrain [0] (i=tabac@gateway/gpg-tor/key-0xB9002659)
10:55:09linuxstbI don't know.
10:55:38 Join kubrick [0] (n=repulse@unaffiliated/funky)
10:59:22pixelmaurgh... it looks like the hack to get battery_bench logging on the sansas somehow sneaked into svn
10:59:42GodEater_svn blame ?
11:00:25pixelmawith JdGordons "plugin for m:robe" commit
11:00:40pixelmathe initial one 9 days ago
11:01:27GodEater_any progress on your gcc multilib thingy linuxstb ? That been put into the build system yet ?
11:02:11pixelmajust noticed because of the latest wiki edits and the mentioned bug report...
11:03:46 Join homielowe [0] (
11:03:55linuxstbGodEater_: It's committed, but not 100% correct. i.e. it works for the armv5 CPU in my device, but not the armv5 CPU in the m:robe. So don't bother updating yet.
11:04:15 Join desowin [0] (
11:05:42GodEater_which device is your one then ?
11:06:12linuxstbMy DAB/mp3 player
11:06:32linuxstbWhich is now also in SVN.
11:07:17 Quit homielowe (Client Quit)
11:07:57GodEater_oh the logick thing
11:08:07GodEater_talking of which...
11:08:11linuxstbYes, that beauty ;)
11:08:23GodEater_when you want to come get nicely soldered one ?
11:13:35***Saving seen data "./dancer.seen"
11:19:03 Quit Soap (Read error: 104 (Connection reset by peer))
11:29:07pixelmahmm... maybe it's not that (about the battery_bench), I can't see it in the diffs (sorry JdGordon). But battery_bench didn't log anything when I first tried on my c200 (beginning of last week or the one before) and now it logs too much. My impression was that it corresponds to what is described in the wiki about the hack (writes everytime the voltage changes)... Second guess: could MoB have changed something here?
11:32:53GodEater_it's possible of course, but I would have thought exceedingly unlikely
11:34:33pixelmamaybe I should try a pre MoB and MoB build...
11:34:48 Quit petur ("hw install")
11:39:01 Nick bb__ is now known as bb (
11:46:34pixelmacan someone else connect to svn?
11:46:39 Join petur [0] (n=petur@rockbox/developer/petur)
11:46:49GodEater_right now ?
11:47:19GodEater_pixelma: I just did - seems fine
11:48:05pixelmayes, at the moment I can't checkout something - neither latest svn nor the older revision. Hmm... maybe it's my connection or my tree then :|
11:52:50GodEater_anyone else heard of "MusicStation" ?
11:58:58 Quit spiorf (Remote closed the connection)
12:01:59linuxstb"the future of mobile music!"
12:02:21GodEater_a friend of mine just started working at omnifone (doesn't that sound Robocop?)
12:02:29GodEater_and for a moment started claiming it was DRMless music
12:02:37GodEater_I shot him down nice and quick though :)
12:02:54GodEater_£2 a week for unlimited DRMless music sounded far too good to be true
12:05:57 Join darkapostrophe [0] (
12:10:55 Quit GodEater (Remote closed the connection)
12:12:01 Join crashd__ [0] (
12:12:18 Quit crashd (Read error: 104 (Connection reset by peer))
12:12:46 Join TMM [0] (
12:19:21pixelmaor maybe it's jhMikeS's multicore changes that's also in the time range between "doesn't log anything and logs too much)
12:19:34 Join midgey [0] (
12:21:18pixelmabut I have to investigate more, currently that's all speculation in the hope that someone with a clue gets an idea
12:22:57 Quit Dark_Apostrophe (Read error: 110 (Connection timed out))
12:22:57 Nick darkapostrophe is now known as Dark_Apostrophe (
12:24:22 Quit midgey (Client Quit)
12:25:27 Join bent [0] (i=57c1ad96@gateway/web/cgi-irc/
12:26:03 Quit Zagor ("Client exiting")
12:32:35 Quit bent ("CGI:IRC (EOF)")
12:32:46 Join BigBambi_ [0] (n=Alex@rockbox/staff/BigBambi)
12:35:23 Quit BigBambi (Nick collision from services.)
12:35:33 Nick BigBambi_ is now known as BigBambi (n=Alex@rockbox/staff/BigBambi)
12:49:11 Join Thundercloud [0] (
12:55:27GodEater_note to B4gder: my build server is ill - need to rebuild it tonight I think. I realise the build system handles this in any case - I'm just telling you as a courtesy ;)
12:57:46 Quit Thundercloud (Remote closed the connection)
12:58:45 Quit markun (Remote closed the connection)
13:00:09MrRichardHowdy irc
13:00:57MrRichardI'm attempting to compile some patches with very little know-how. I've read the guides and am still lost. Are there any kind souls willing to help me out?
13:04:27preglowlinuxstb: doesn't that logik thing have any storage?
13:04:46 Join PaulJam [0] (
13:04:53linuxstbIt has 1GB of flash
13:05:56preglowhave you got any docs for the tuner part?
13:13:36***Saving seen data "./dancer.seen"
13:19:18linuxstbMrRichard: Which guide are you following (probably best to choose one), and then we can help you with specific issues.
13:20:48linuxstbMrRichard: And you should also forget patches for now - just concentrate on compiling an unmodified Rockbox build.
13:21:48MrRichardlinuxstb: but getting the patches together is the whole point I'm doing it. I followed TheSImpleGuideToCompiling, and am now going through WMWarDevelopmentPlatform.
13:22:14MrRichardBut the VMwar guide, which seems to be the one most applicable to me, says nothing about patches
13:22:28linuxstbMrRichard: I just said "for now" - take one step at a time...
13:23:08MrRichardYeah, makes sense.
13:23:46MrRichardso by following this guide: , what exactly am I achieving?
13:24:05MrRichardIs that the basics to creating a build?
13:24:51linuxstbYes, that page tells you how setup the vmware development environment, and compile an unmodified build.
13:26:31MrRichardOkay, I'll go through this first.
13:26:34MrRichardThanks so far ;)
13:28:18*linuxstb wonders why the vmware page instructs people to do "chmod 777" in the source tree.
13:29:17 Join ramon8 [0] (
13:35:13 Quit BigBambi (Read error: 104 (Connection reset by peer))
13:35:42 Join BigBambi [0] (n=alex@rockbox/staff/BigBambi)
13:36:33 Join qwm [0] (
13:38:23 Quit RaRe (Read error: 113 (No route to host))
13:38:49 Join kclaf [0] (i=kclaf@
13:40:53 Join japc [0] (
13:40:55XavierGrlinuxstb: I thought it specifically said "not to"
13:41:13 Quit B4gder ("It is time to say moo")
13:41:52 Quit _jhMikeS_ (Read error: 104 (Connection reset by peer))
13:41:59linuxstbXavierGr: Under the "Compiling" section
13:42:14XavierGrah my bad, I wrote this after I checked out the code from the root account
13:42:33XavierGrit doesn't brake anything and there is a warning to not do it in the filesystem root
13:42:45XavierGr(Which I did on my first installation of linux which broke it)
13:43:38GodEater_why would you chmod 777 it in any case ?
13:43:47linuxstbA better fix would be to do chown
13:44:03linuxstbOr just delete the tree and check it out as the regular user.
13:44:15GodEater_quicker to chown
13:44:50linuxstbYes, but I think we need to try and keep the wiki instructions as simple as possible
13:45:07*XavierGr is a linuxnewbie
13:45:10GodEater_so mention neither then, and make sure the user is told to not do anything as root
13:45:13XavierGrat that time I thought it would work
13:45:32linuxstbBut I can imagine it's a common error.
13:45:34 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
13:46:45 Join darkapostrophe [0] (
13:48:09 Join moos [0] (
13:55:24 Join RaRe [0] (
14:01:30 Quit desowin (Read error: 113 (No route to host))
14:03:55 Quit Dark_Apostrophe (Read error: 110 (Connection timed out))
14:03:55 Nick darkapostrophe is now known as Dark_Apostrophe (
14:05:34MrRichardUsing the compiling example "svn co svn:// rockbox
14:05:34MrRichardcd rockbox
14:05:34MrRichardmkdir build_h120
14:05:57MrRichardbecause I'm using an ivirver H10, do I replace build_h120 with h10?
14:06:14XavierGrMrRichard: you can choose any name you like
14:06:22MrRichardThank you, i was unsure of it's importance.
14:06:58MrRichardI hit enter after svn co svn:// rockbox and now a lot of stuff is scrolling passed. Is this meant to happen?
14:07:11MrRichardThank you.
14:07:21MrRichardI'm sorry for all the silly questions guys...
14:07:36GodEater_so you actually want the "development bootloader" to be flashed linuxstb ?
14:07:55GodEater_is that how the current code works that aliask and co have done ?
14:07:57linuxstbMy idea is that we write a sort of "development bootloader" for the Gigabeat S which is flashed once, and then this is used to load further test code from disk, which we transfer via MTP.
14:08:28GodEater_does anyone know why the attempts to load rockbox.gigabeat have failed so far ?
14:08:34linuxstbMy understanding is that currently the only way to run code, is to install it (i.e. flash it?) via the hacked Gigabeat V firmware updater.
14:08:49BigBambiyeah, as I understand it that flashes it
14:08:53GodEater_it does ?
14:09:00GodEater_hmm - I thought it just wrote stuff to the disk
14:09:04linuxstbHence the risk of brickness
14:09:18BigBambi"Remove the usb cable when prompted, and the gigabeat will flash the new firmware, then run it."
14:09:23GodEater_I didn't think that was a risk ? I thought the "recovery" mode always worked
14:09:35BigBambifrom step 3 of the code inserting part
14:09:36GodEater_up and until it doesn't of course :)
14:09:56GodEater_but aliask didn't indicate it was because he'd uploaded bad code
14:10:25linuxstbNo, but from what I've read, it seems repeated flashing isn't a good idea. It's also very inconvenient, especially for Linux users...
14:10:36linuxstbThe latter issue is the most important.
14:10:41GodEater_until we write that linux updater....
14:10:44linuxstbAnd is probably holding up progress...
14:10:47BigBambiIndeed, the update tool is windows only
14:10:57GodEater_and entirely in Japanese
14:11:04BigBambihah, really?
14:11:06linuxstbNice, I didn't know that ;)
14:11:08BigBambiI haven't tried it yet
14:11:16GodEater_don't - unless you have to
14:11:23GodEater_I only did it to get another usb log
14:11:26BigBambiI shan't
14:12:31MrRichardafter "svn co svn:// rockbox" and the scrolling has ended, the guide says to put in "cd rockbox" but all i get is "no such file or directory"
14:12:39GodEater_so your bootloader will work a bit similar to the original ipod you did I guess linuxstb? I.e. hold a button whilst you power up and then that'll go look for some code on disk somewhere ?
14:12:44BigBambiMrRichard: what does ls tell you?
14:13:01GodEater_GodEater_: otherwise boot OF ?
14:13:08*GodEater_ speaks to himself for some reason
14:13:40BigBambiA habit of the old, they always chose the wisest person present to talk to
14:13:50MrRichardBigBambi: -bash: cd : rockbox : No such file or directory
14:14:16GodEater_MrRichard: just type "ls" and hit enter
14:14:25BigBambiMrRichard: If you type ls what do you see? (please put the results in
14:15:02MrRichardI know I'm going to sound like such an idiot... but...
14:15:08GodEater_MrRichard: it's a website
14:15:20BigBambiIt is a webiste where you can paste text so as not to clog up the channel
14:15:36MrRichardoh okay
14:15:38*GodEater_ just noticed that BigBambi implied he was old
14:15:48MrRichardWell ill go put it there. I got a lot of blue text'
14:15:48*GodEater_ waves his walking stick at the young whippersnapper
14:15:49BigBambiBut wise also
14:15:52linuxstbGodEater_: Yes, we would need to make it dual-boot.
14:16:08BigBambiMrRichard: Give us the link when you have it
14:17:08 Join MethoS- [0] (
14:17:11GodEater_linuxstb: do we have any idea how to do that ?
14:17:34 Quit qwm (Remote closed the connection)
14:17:48 Join qwm [0] (
14:18:29GodEater_MrRichard: You're already inside the rockbox directory there
14:18:34GodEater_so just do the "mkdir" bit
14:18:35BigBambitype cd ..
14:18:40BigBambiscrap that
14:18:43MrRichardOh okay.
14:18:44 Join Thundercloud [0] (
14:18:47MrRichardI had a hunch that was the case.
14:18:47GodEater_do what I said
14:18:49BigBambitou want to be in rockbox :)
14:18:55MrRichardI'm just so precautious with all this :(
14:20:11MrRichardJust a question, purely out of interest. but what exactly does "../tools/configure" do?
14:20:18MrRichardspecificy the location or something?
14:20:20BigBambiIt configures the make file
14:20:36BigBambiSo when you type make it tries to build a build for the correct player etc.
14:20:42linuxstbGodEater_: I have no idea how to do anything on the S at the moment...
14:21:25linuxstbMrRichard: "../tools/configure" runs the script called "configure" which is located in a directory (folder) called "tools", one level higher than where you are.
14:21:41GodEater_linuxstb: I didn't mean you personally ;)
14:21:57linuxstbGodEater_: I know, but I can only answer for myself... ;)
14:22:16GodEater_of course, but you might have heard mention it was at least possible
14:22:37MrRichardwhen prompted, what sort of build to i select. (N)?
14:22:48linuxstbNo, but it should be straightforward to hook into the boot process somewhere, check for a keypress, and then either continue the OF booting, or branch to our own code.
14:23:48linuxstbMrRichard: Yes.
14:23:51GodEater_MrRichard: N is the best choice ;)
14:24:49 Quit japc (Read error: 104 (Connection reset by peer))
14:25:07MrRichardIm gradually gaining an understanding for all of this. which is excellent
14:26:16GodEater_linuxstb: I notice that the guys say they've got no interrupts working yet - is that going to be much of an issue ?
14:30:40linuxstbI don't think so - the ipod bootloader doesn't enable interrupts, and maybe other targets.
14:30:40GodEater_I wonder why they've been unable to load something else...
14:30:40BigBambiIt doesn't say much on he wiki page really
14:30:40MrRichardthe guides have so many references to this C:\Rockbox\home\... etc etc folder. And I can't seem to find mine anywhere. Whereabouts is it created?
14:30:40GodEater_the gigabeat-s.c bootloader is incredibly short :)
14:30:40BigBambiMrRichard: You are using VMWare right?
14:30:40BigBambiMrRichard: Which guide?
14:30:40MrRichardBigBambi: correct
14:30:50BigBambiwhich section does it say this?
14:31:20GodEater_ah yes the "SimpleGuideToCompiling" which is 900 pages longer than the original GuideToCompiling, and is, therefore in my opinion, not at all simple
14:31:40BigBambiIt does spell it out a bit more though
14:31:50BigBambiSimple doesn't always equal short :)
14:32:07MrRichardThe one I'm currently looking at is Adding Patches: Preparation, and it is provided as an example just before the bold 'patching'
14:32:26MrRichardI realise it's just an example and won't be in the same palce on my computer, but I just can't seem to find the folder at all
14:32:37BigBambiMrRichard: I guess that comes from cygwin
14:32:51GodEater_MrRichard: it does only say "e.g." as well
14:32:55BigBambiput the patch in the rockbox folder in your VMWare image
14:33:07GodEater_MrRichard: if you type "pwd" it'll tell you which folder you're currently in
14:33:20Buschelamiconn: you there?
14:33:27MrRichardOh okay. So I should just put the patches in the folder stipulated by pwd?
14:33:38BigBambiput it in the rockbox folder
14:33:51BigBambipwd just tells you where you currently are
14:34:11GodEater_which should be your rockbox-folder/build-h120
14:34:18GodEater_or wherever you were when you ran configure
14:34:29MrRichardI have three rockbox folders though.
14:34:33GodEater_linuxstb: neat - kate understands #if 0 is a block comment :)
14:34:45GodEater_MrRichard: but you're only working in one of them though
14:35:08BigBambiput it in the one you are working in, the one with apps, firmware etc
14:35:27MrRichard .vmdk.lck, .vmem.lck, .vmx.lck. And they all just contain a single .lck file
14:35:40GodEater_that's your windows file system
14:35:47GodEater_not the one that the vmware image is using
14:35:56BigBambiYou should be in VMWare here, not my computer
14:36:02MrRichardSo I have to find the folder within VNware?
14:36:10MrRichardOh man that really clears things up...
14:36:45amiconnBuschel: sort of
14:37:27MrRichardNow comes the issue of how to find folders in VNware...
14:37:59Buschelamiconn: first -> congrats! your LCD-rework does just fine here:o) and: I've found a way to speed lcd_update up a little at least for non-full screen (+2%)
14:38:30Buschelamiconn: performs at 399fps @80MHz now
14:39:55MrRichardHow do I use a zip if I can't find it in explorer?
14:40:21MrRichardSorry. I'm new to this and really don't know how to correctly phrase things.
14:40:35MrRichardI followed the basics, and created a .zip.
14:40:42MrRichardBut where does this .zip even exist?
14:41:06BigBambiIn the VMWare file system, not explorer
14:41:12BigBambito transfer it you can use samba
14:41:24BigBambiI think the VMWare page tells you how to do it
14:41:24MrRichardTransfer into explorer?
14:41:34MrRichardYes there is info on it, I'll look into it.
14:42:40MrRichardHmm. The page mentions it, but doesn't explain how to use it
14:42:45 Join spiorf [0] (
14:42:46BigBambiThe VMWare image is a linux system contained (as windows sees it) within a single file. When you run VMware it reads the linux system from that file and runs it. It translates between it and windows. The linux filesystem as far as windows is concerned stays within the single file
14:43:05BigBambiTherefore you cannot see any of the linux filesystem from explorer
14:43:34MrRichardNow I understand.
14:43:37BigBambiClick start, the run, then type \\debian\user
14:43:59 Join MethoS-- [0] (
14:44:41 Join sdoyon [0] (
14:45:42MrRichardExcellent. So I am free to copy files from here?
14:46:35MrRichardNow that I understand this I might aswell gives patching a try. Can I just create another build eeven though I already have one in there?
14:46:52 Join nicktastic [0] (n=nick@unaffiliated/nicktastic)
14:47:24GodEater_MrRichard: just run "make clean" before you do configure again
14:47:39BigBambitry to apply the patch first, then run make clean before re-configuring
14:47:48BigBambiwhat GodEater_ said :)
14:47:53MrRichardtry to apply patches onto the build I just made?
14:48:02BigBambino, you apply patches to the source code
14:48:14BigBambiThe build is generated from the source code when you type make
14:48:32MrRichardOh okay.
14:49:02MrRichardUmmm. I just don't know where to go from here.
14:49:04lostlogiclinuxstb: hmm, hadn't thought about changing the semantics that much, it would give the buffering.c code more control over things in general to remove the concept of a data_not_ready return
14:50:02BigBambiMrRichard: you want to try and apply a patch
14:50:09MrRichardI followed to guide up to where I enter 'make'
14:50:18MrRicharddo I start again or work with what I've got?
14:50:25GodEater_work with what you've got
14:50:35GodEater_you need to return to your Rockbox directory
14:50:39GodEater_and apply the patch there
14:50:39BigBambiwhen you typed make you generated a build
14:50:44GodEater_and then return to your build directory
14:50:47GodEater_and re-run configure and make
14:51:02MrRichardWhat do I put in to return to rockbox
14:51:08BigBambiNow you can patch the source code and then geneate another build, which will be different as you changed the source code
14:51:17BigBambi"cd .."
14:51:21 Join darkapostrophe [0] (
14:51:44MrRichardokay I'm back in rockbox.
14:51:49MrRichardNow I'll look at how to apply patches...
14:52:38GodEater_BigBambi and linuxstb: I just modified the GigabeatSInfo page to include the HD layout from the thread
14:52:44MrRichardWhereabouts do I put my .patch files?
14:52:54BigBambiMrRichard: in the rockbox directory
14:53:00GodEater_shame it's in French =/
14:53:11BigBambiI'm getting better!
14:53:12*GodEater_ wonders if pixelma would oblige with an english translation
14:53:29 Nick fxb__ is now known as fxb (
14:53:34linuxstblostlogic: I can't remember what the codec API read() did in the past - i.e. whether it blocked or not... I can't remember seeing any code that checked for short reads.
14:53:49BigBambiGodEater_: I'll have a go
14:53:54MrRichardBigBambi: drage and drop into the root Rockbox folder?
14:53:54sdoyonThe info screen has code to display "Battery: Charging" instead of the battery level when the charger is plugged in and the battery is charging. But that code is #if CONFIG_CHARGING == CHARGING_CONTROL. Yet some of the 3 charging states apply to CHARGING_MONITOR and CHARGING_SIMPLE. Anyone knows this stuff?
14:53:55GodEater_or you ;)
14:54:04BigBambiMrRichard: yep
14:54:07pixelmaGodEater_: huh? I don't speak french... at all
14:54:09MrRichardThank you.
14:54:22GodEater_pixelma: ooh - my mistake - thought you did - sorry!
14:54:26lostlogiclinuxstb: so I was wrong the other day when I said that the code shoudl loop for short reads... it should just loop for data_not_readys, it is still all or nothing
14:54:40BigBambiGodEater_: You probably mean Nico_P - him being french and all
14:54:46BigBambiAnyway, I'll do it
14:55:47MrRichardI have put all my patch files into '\\debian\user\rockbox', is this where they should be?
14:55:48lostlogiclinuxstb: I am wondering why Nico made the buffering api reads non-blocking but then made clients call a blocking buf_request_buffer_handle which is blocking.
14:55:56GodEater_MrRichard: correct
14:56:26MrRichardGodEater: thank you :)
14:57:09 Quit J3TC- (".UPP.")
14:57:13MrRichardTo "# In the cmd window, go to the directory containing the patch" would I just ignore that, as I'm already in cd .. ?
14:58:20GodEater_MrRichard: you realise ".." merely means "the directory that is the parent of this one" ?
15:00:54MrRichardOh man D:
15:01:05MrRichardsee I thought it was just the overall root directory.
15:01:15MrRichardThat would explain why the patch command didnt work
15:01:21GodEater_no - the "root" directory is "/"
15:01:51 Quit Dark_Apostrophe (Read error: 104 (Connection reset by peer))
15:01:51 Nick darkapostrophe is now known as Dark_Apostrophe (
15:01:52 Join kugel [0] (i=kugel@unaffiliated/kugel)
15:02:03amiconnBuschel: how?
15:02:03MrRichardso if my .patches are in the root directory, i type cd / and begin patching>?
15:02:34GodEater_are you sure you put them in your root ?
15:02:54lostlogiclinuxstb: I'm inclined to commit said patch though and then separately start working on changing the api more. For the first second time only since MoB my player played all night w/o stopping playback.
15:02:58Buschelamiconn: including the looping via height and the needed address calculations in the asm-routine
15:02:58MrRichardhow do i get out of this loop of patching im stuck in?
15:03:32 Join J3TC- [0] (
15:03:42MrRichardIt keeps asking "File to patch" "Skip this patch [y]?" and I can't get out
15:03:48lostlogicfirst _or_ second that is.
15:04:04amiconnBuschel: Hmm. I'm not sure whether this is a good idea. Too much asm makes the update function hard to understand. And it's not that we'd have a performance problem now :)
15:04:13GodEater_MrRichard: Press "Ctrl" and C together
15:04:26linuxstblostlogic: Sure. I don't see a problem committing the patch.
15:04:41MrRichardIs skipping through all them patches just now going to affect anything?
15:04:48GodEater_MrRichard: and btw - \\debian\user\rockbox is NOT the root directory
15:04:58GodEater_MrRichard: no - you'll be fine if you just ctrl-c
15:04:58Buschelamiconn: but squeezing some more fps is fun :o) and it makes the lcd_update-function itself more "tight"
15:05:09MrRichardGodEater: excellent, thank you very much
15:05:44 Quit MethoS- (Read error: 110 (Connection timed out))
15:05:55linuxstbBuschel: So your patch increases 1/4 screen updates from 391.5% to 399% at 80MHz?
15:06:50GodEater_MrRichard: that error you're getting is because you're not using the right number after -p
15:07:00GodEater_did you use -p0, or -p1 ?
15:07:00MrRichardOh okay.
15:07:04MrRichardI used p0
15:07:05Buschellinuxstb: yes
15:07:07GodEater_try p1
15:07:10GodEater_that should work
15:07:24MrRichardBut I wrote the patch name immediately after the "<", with no space. Will that have affected it?
15:07:33GodEater_no - that won't matter
15:07:46GodEater_at least it shouldn't
15:08:00MrRichardAlso I appear to be in the wrong place, I hit cd / where in actuality I want to be in "\\debian\user\rockbox" what "cd..." should I do to get there?
15:08:27GodEater_MrRichard: try "cd /home/user/rockbox"
15:08:44GodEater_MrRichard: actually do "cd ~/rockbox"
15:08:48GodEater_that'll work better
15:08:55MrRichardFirst one worked
15:12:45MrRichardnow Ill try again, using p1
15:12:45linuxstbBuschel: I would tend to agree with amiconn...
15:12:45GodEater_good luck
15:12:45 Quit Rick (Read error: 104 (Connection reset by peer))
15:12:45linuxstbMrRichard: If you type "cd" by itself (no directory name), it will take you back to your home directory - i.e. /home/user/ That's useful if you're lost ;)
15:12:45 Join Rick [0] (
15:12:45MrRichardThanks heaps linuxstb
15:12:45*GodEater_ wishes windows would behave like that
15:12:45MrRichardI did the patch command and got a series of errors.
15:12:45Buschellinuxstb/amiconn: well, i'll just submit the patch and you may decide.
15:12:45GodEater_MrRichard: paste them to again
15:12:45BigBambiGodEater_: There you go
15:12:45GodEater_BigBambi: thanks ;)
15:12:45DBUGEnqueued KICK MrRichard
15:12:45MrRichardGodEater: What's the easiest way to cut/paste out of VMware?
15:12:45DBUGEnqueued KICK GodEater_
15:12:45GodEater_just selecting the text should be enough to copy it I think
15:12:59CtcpIgnored 2 channel CTCP requests in 0 seconds at the last flood
15:12:59*GodEater_ doesn't use VMWare though and may be wrong
15:12:59nanokGodEater_: not if you lack "vmware tools", iirc
15:13:06GodEater_nanok: ah
15:13:13GodEater_have we not installed them in our image ?
15:13:39***Saving seen data "./dancer.seen"
15:13:45MrRichardselecting did not work. Is there an alternative?
15:14:06GodEater_does our image include a web browser ?
15:15:00BigBambiNo idea
15:15:01GodEater_MrRichard: check the menus in the VMware image for something that looks like a webbrowser
15:15:10MrRichardChecking nwo.
15:15:20GodEater_firefox / epiphany / dillo / opera
15:15:22GodEater_any of those
15:15:24GodEater_or galeon
15:15:34GodEater_or that
15:15:45linuxstbOr iceweasel...
15:15:49GodEater_although I suspect that the GNOME / KDE ones aren't there
15:16:00GodEater_too many supporting libs required
15:16:08GodEater_would make the image too large
15:16:17MrRichardNo browser
15:16:23GodEater_ok - try this
15:16:28GodEater_"sudo apt-get install dillo"
15:16:40*GodEater_ crosses his fingers
15:16:40linuxstbYou could open up a text editor, paste the text into there, then save it as a file, which you can then open in Windows.
15:16:47GodEater_yeah or do that
15:16:47MrRichardIll try that
15:17:09nanokGodEater_: links/lynx/w3m ?
15:17:13GodEater_of course, it'll come out minus line feeds...
15:17:13MrRichardWhats the shortcut to paste what ive selected?
15:17:22GodEater_middle mouse button
15:17:26nanokMrRichard: middle click, shift+insert
15:21:03GodEater_or wheel button
15:21:03GodEater_nanok: I prefer not to walk a newbie through using a text only browser ;)
15:21:03GodEater_I still get lost in them myself
15:21:03*GodEater_ prefers elinks anyway ;)
15:21:03nanokGodEater_: okay, fair enough. allthough i find them the easyest to understand and "digest"
15:21:03MrRichardx) how do I save?
15:21:03linuxstbFrom where?
15:21:03GodEater_MrRichard: which text editor did you use ?
15:21:03MrRichardthe text editor
15:21:03***Alert Mode level 1
15:21:03***Alert Mode level 2
15:21:03GodEater_I think
15:21:03MrRichardthats search
15:21:03***Alert Mode level 3
15:21:03MrRichardby the looks of it
15:21:03***Alert Mode level 4
15:21:03GodEater_no - Ctrl-O it seems
15:21:03*nanok would advise "cat > somefile", paste, ctrl+d
15:21:03nanokmuch easyer to understand, and platform independent :)
15:21:03***Alert Mode level 5
15:21:03GodEater_nanok: possibly so...
15:21:03*petur recalls that nano prints the keys at the bottom - so why ask here?
15:21:03***Alert Mode level 6
15:21:03GodEater_petur: because "WriteOut" is a peculiar thing to call "Save"
15:21:03linuxstbI guess "WriteOut" isn't obvious...
15:21:03nanokpetur: "write-out"? hm...
15:21:03DBUGEnqueued KICK nanok
15:21:03nanoklinuxstb: :)
15:21:03***Alert Mode level 7
15:21:03MrRichardSee here I was thinking it simpyl didnt list the save shortcut
15:21:03***Alert Mode level 8
15:21:03GodEater_stupid Nano developers
15:21:03lostlogicgood, I was afraid I'd increase code size by a few bytes, but I got a -7
15:21:03***Alert Mode level 9
15:21:03MrRichardBut silly me didn't think about write-out
15:21:03***Alert Mode level 10
15:21:03MrRichardI saved it, whereabouts will it be?
15:21:03DBUGEnqueued KICK linuxstb
15:21:03linuxstbThat depends where you saved it ;)
15:21:06GodEater_probably in your \\debian\user directory
15:21:12linuxstbWhich folder were you in when you typed "nano" ?
15:21:23GodEater_I suspect he launched it from the menus
15:21:48linuxstbI thought Nano was a console app - they're also in the menus?
15:21:58GodEater_xterm -e
15:22:11GodEater_or whatever you do to launch xterm with an app already running in it
15:22:11MrRichardOkay, saving to pastebin now. But it's horrible and ugly with no breaks.
15:22:19GodEater_unix2dos is first
15:23:39linuxstbWhat did you use to open it in Windows? Anything apart from Notepad should work fine.
15:23:39MrRichardall good.
15:23:39GodEater_even wordpad should be ok
15:23:39***Alert Mode level 11
15:23:39GodEater_why does that guide insist on using −−binary ?
15:23:39***Alert Mode level 12
15:23:39GodEater_MrRichard: okay - all those errors mean your patch is out of sync with the current rockbox code base
15:23:39***Alert Mode level 13
15:23:39GodEater_you'll have to ask the patch author to resync it
15:23:41GodEater_or learn how to do it yourself
15:24:05GodEater_might be quite a job, since the filename seems to indicate it was last sync'd in june
15:24:47linuxstbThere seems to be an up to date version here - (Oct 2nd)
15:25:49MrRichardall the links to the build I need are dead, so my only option was to compile the patches manually. But that's starting to look progressively less possible.
15:28:16Buschelamiconn/Linuxstb: submitted patch to FS #8041. make up your mind :)
15:28:16 Join scorche|w [0] (i=8dc5049d@gateway/web/cgi-irc/
15:28:16GodEater_MrRichard: look at the link linuxstb posted...
15:28:16MrRichardYeah, I looked
15:28:16MrRichardI downloaded the patch and I'll try that in a moment.
15:28:16MrRichardIt's just the thing is I need to do several other patches too, which will most likely have the same issue.
15:29:10linuxstbMrRichard: You can search for other patches - if they're popular, they are very likely to be available in up to date versions.
15:29:45MrRichardI'll see how this one goes now.
15:29:53 Join MsMacka [0] (
15:30:02MrRichardBecause of all the errors and such that I had with the other one, is that going to mess anyting up?
15:30:46MrRichardI got the exact same errors when I tried that patch
15:31:49linuxstbYou should revert your copy of the source back to its unmodified state - "cd ~/rockbox" followed by "svn revert -R ." (don't forget the dot)
15:32:41linuxstbI also like the "−−dry-run" option to patch - it doesn't actually modify any files, but displays the result as if it did - i.e. "patch -p0 −−dry-run < patch"
15:33:05linuxstbIf that doesn't give any errors, you can then apply it for real.
15:33:05MrRichardso if you do that, see the results, then do the patch command?
15:33:11MrRichardI see.
15:33:40***Alert Mode OFF
15:33:47MrRichardI did the revert
15:33:49lostlogichcs: around now?
15:33:53MrRichardtry the new patch again?>
15:33:59 Quit sdoyon ("ircII EPIC4-2.4 -- Are we there yet?")
15:34:05linuxstbMrRichard: Yes.
15:35:11MrRichardHmm, same errors.
15:35:18linuxstbLet me try it...
15:35:40MrRichardI'd rather not keep wasting your time, I think I'll just give up and start searching for a working link to the build.
15:35:56 Quit MsMacka ("CGI:IRC (Ping timeout)")
15:36:06MrRichardI just want this theme to work.
15:36:13linuxstb"patch -p1 −−dry-run < ymargin-scrollinfo_20071002.patch" works fine for me - no errors.
15:36:36MrRichardman I mustve done something heinous =/
15:37:16 Part scorche|w
15:37:48 Join scorche|w [0] (n=8dc5049d@rockbox/administrator/scorche)
15:37:53CtcpIgnored 2 channel CTCP requests in 0 seconds at the last flood
15:37:53*nanok is using the plain-text theme, ..with the peak meters removed
15:38:11MrRichardI like having a nice theme.
15:38:21MrRichardIt just means so much to me :)
15:38:25MrRichardthat and a working database.
15:38:31MrRichardWhich is also an issue right now.
15:38:33 Join toffe82 [0] (
15:39:17nanokMrRichard: but plaintext is a beautyfull theme..
15:39:46nanokMrRichard: i assume you need albumart, that's why you need to patch?
15:39:58MrRichardThat and other things.
15:40:11MrRichard i really want to use the redeye theme
15:40:34nanokMrRichard: or is it just that the theme you happen to like has album art, and needs the patch? maybe it would be easyer to just remove the album art from the theme, instead, if you don't care about it
15:41:00nanokit would also give you the advantage of being able to update to current builds without any problems
15:41:25MrRichardIm happy without album art, although it is a plus, it is mostly superfluous.
15:41:32MrRichardIt's just the theme itself is nice.
15:41:40MrRichardAnd apparantly needs more than just album art patch
15:42:24karashataMrRichard: most of the other patches are probably used to adjust the rest of the text stuff to not overlap the album art picture
15:42:33nanokMrRichard: x5?
15:42:56karashataif you get rid of the album art, you can get rid of the rest of the text alteration patches as well
15:43:01nanokMrRichard: your player
15:43:12karashatananok: he has a 20GB H10, same as mine
15:43:17MrRichardoh. iriver 20gb h10
15:43:21MrRichardKarashata is a clever one ;D
15:43:27nanokkarashata: aahm, okay. you lucky bastards :)
15:43:41MrRichardMine was free. It makes it all that much better
15:44:01karashataMrRichard: I made a few themes you may want to look into if you're looking for something that looks nice and doesn't need all sorts of patches...
15:44:05MrRichardI went to pickup my broken 5g and the shop owner said "these arent selling, have one"
15:44:07 Quit MethoS-- (Read error: 110 (Connection timed out))
15:44:17MrRichardkarashata: Could you provide some links?
15:44:23 Join MethoS-- [0] (
15:44:33karashatasure, just a sec
15:44:40MrRichardthank you very much
15:44:58MrRichardI guess my only remaining problem is the fact that my database doesn't work.
15:45:10nanokMrRichard: HUH?! where is that shop? :-P
15:45:29MrRichardIf a song, whether it be the file tiself or it's tag, is corrupting my database, will simply deleting the song and re-initialising the database fix it?
15:45:36nanokMrRichard: btw, have you tryed that theme without the patches?
15:46:11MrRichardnanok: the guy felt sorry that he couldnt fix my 5gb. And yes, I have. the WPS screen is default, the only difference is the font and background
15:46:19MrRichardkarashata: thank you
15:46:42karashatafeel free to change the fonts if you don't like the defaults I set on them
15:47:02karashataif you decide to check them out, that is
15:47:27MrRichardthey look nice to me. I might try a few out and see how I like them
15:47:42*karashata nods slightly
15:48:00karashataif you have any suggestions for improvements drop me a line
15:48:08MrRichardsure thing.
15:48:10karashataor ideas for a new theme
15:48:21MrRichardThey seem to have all the good core elements
15:48:26karashatajust, don't ask me to try to do stuff with album art or anything
15:48:44MrRichardSure ;D
15:48:53MrRichardAs for this theme I've been struggling with...
15:48:55 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
15:49:05MrRichardHow do I remove the album art feature and just use the theme without it?>
15:50:02karashatayou'd need to know the WPS code and remove any tags that are introduced by the patches you need to use it with the album art
15:50:43MrRichardwell, i just spent the last few hours being babied through patching. I think I'll give this one a miss ;P
15:51:22MrRichardfor now I'll just try to fix my database.
15:51:36karashatawhich version of that are you looking to use
15:51:48MrRichardversion of the theme?
15:51:57karashatait says there are two versions
15:52:05MrRichardI was planning on using version 2
15:52:09karashataahh, okay
15:52:10MrRichardalthough either one seems fine
15:52:40karashataI'll grab it and fix the WPS code to work without the album art and the other patches in a bit for ya, kayÉ
15:52:53karashataeek, hold on a sec, I need to restart Pidgin...
15:52:58MrRichardThanks a heap man.
15:53:10MrRichardYou know I really need to get around to putting linux on this box =/
15:53:18 Quit karashata ("Leaving.")
15:53:48 Join karashata [0] (
15:53:49nanokMrRichard: nothing can be easyer ;)
15:54:07MrRichardi used to mess around with botting off of a CD
15:54:16MrRichardBut I think I mightaswell just install it now.
15:54:21nanokkarashata: if you do that, i think it would be very interesting to have it uploaded as "themename_anybuild"
15:54:36karashatananok: I'll do that, then
15:54:52nanokas i said, the nice feature about these "honest" themes is that you can upgrade anytime to a current build, so i think many people would be interested
15:55:15MrRichardit does make a lot of sense
15:55:55nanokMrRichard: use a kanotix, or knoppix cd, recent one. it will probably be the easyest desktop installation you have ever experienced ;)
15:56:21MrRichardOnce my internet resets that's exactly what I'll do
15:56:31 Join jgarvey [0] (
15:57:07MrRichardRegarding linux, how easy is it to grab files off of explorer and access them>
15:58:15nanokMrRichard: i'm afraid i don;t follow
15:59:05MrRichardOnce again, sorry about poor phrasing. But back at TAFE we used to dualboot linux, but had problems getting files that we had always been able to use on windows onto linux.
15:59:13MrRichardAs in most of the time they just never appeared.
15:59:40linuxstbThat shouldn't be a problem - Linux can access Windows partitions, both FAT and NTFS
15:59:47lostlogichas anyone _run_ a sim since my most recent two commits?
15:59:50MrRichardoh, fantastic
15:59:52nanokshit. i realised i like rockbox so much i was seriously thinking of not going by car to work anymore, so i can listen to my headphones (bike or walk)
16:00:03nanokthis sounds...somehow wrong
16:00:11MrRichardNanok: I get like that all the time
16:00:26LinusNi connect my player to the car stereo
16:00:30MrRichardI usually prefer to walk or bus, simply so I can listen to music.
16:00:55nanokMrRichard: where did you try to get them from? samba share? filesystem? normally there should be no such problem
16:00:58MrRichardHeadphones feel better than stereo. Stereo is good, but after a while just. makes me feel ill?
16:01:01 Join Domonoky [0] (
16:01:25MrRichardnanok: it was definately a while ago, and i definately didn't know a thing about linux then.
16:01:25nanokMrRichard: it also depends on the type of files (sometimes you may not have anything to open them with)
16:02:18nanokwell, honestly i don't like headphones, not even with that surround shit. but since i discovered the "simple" concept of crossfeed, i am hooked: it;s the best of both worlds
16:02:44MrRichardI've never been able to wrap my head around crossfeed.
16:02:53MrRichardI understand it times the L and R differently?
16:03:02MrRichardWhat does it achieve? other than a stereo effect
16:03:15nanokLinusN: my car stereo is far from matching my px100's :).
16:03:38nanokMrRichard: just try it, with the default settings
16:03:51MrRichardsure thing. I'll see how it feels.
16:04:30nanokMrRichard: it is not just a stereo effect, it achieves the difusion of the separation line between left and right. with headphones, the line is very clear, and annoying imho (unnatural)
16:04:47nanokwith crossfeed, it feels more like a real listening experience
16:05:01MrRichardAhh, I'm sarting to understand the appeal now
16:05:44nanokand ofcourse, this being rockbox, you can tune it to suith your needs
16:06:13nanok*just be carefull, it can do bad things to the sound, like cliping, unless you are carefull)
16:06:27MrRichardahh okay. Guess it'll just take a little fine tuning
16:09:44nanoki am proud of myself. i was able to get through most of the day with no rockbox. but now it's time .. ican't stand it much longer
16:10:34MrRichardpicking up my iriver has actually been quite depressing lately. It's just not the same without the database...
16:10:43MrRichardand I cant for the life of me work out whats doing it
16:10:47scorche|wnanok: well, that isnt rockbox so much as it is an equalizer that trusts you when you say "i want to do this" ;)
16:10:59pixelmaNico_P: playback just "froze" in the wps on my c200, revision is 15397. I'm just playing mp3s, the pcm buffer bar in the "view buffering thread" was empty. I could stop and resume without a problem (already had one occasion of this yesterday with an earlier revision)
16:11:00scorche|w(re: clipping)
16:12:24 Join MethoS--- [0] (
16:12:32nanokscorche|w: ;). it is more than that. it is also the knowledge that my player will do what i want. and if i ever get the fancy idea to configure it in a particular way, i can do it
16:12:57MrRichardWell guys, i need to nap. Work in a few hours.
16:12:58nanokthat's essential to me. gives me the psychological confort i need to listen to music :)
16:13:26nanokMrRichard: yeah, go. last time i heard some customers ended up in the hospital :-P
16:13:46MrRichardTheir fault for not noticing the bacon was raw.
16:13:56 Nick fxb is now known as fxb__ (
16:13:56nanokcleaning products instead of mustard don;t go down well..
16:14:00MrRichardCatch you guys later, adn thanks so much for the help. I'm a terrible newbie, I know.
16:14:06nanokMrRichard: :)
16:17:54 Quit J3TC- (Read error: 113 (No route to host))
16:18:17 Join J3TC- [0] (
16:19:08 Quit toffe82 (Remote closed the connection)
16:23:19pondlifelostlogic: Your codecs.h change, you should probably update the API version...
16:25:21 Join Rondom [0] (
16:25:58 Join localhero [0] (n=dinzdale@
16:29:00 Join mf0102 [0] (n=michi@
16:29:47 Quit MethoS-- (Read error: 110 (Connection timed out))
16:29:53 Quit J3TC- (Read error: 113 (No route to host))
16:34:34lostlogicpondlife: that function isn't exported to the API
16:34:46pondlifeAh, ok
16:34:58pondlifeSorry, move along, etc... :)
16:36:23amiconnlostlogic: What about audio that is large and cannot be streamed?
16:36:31amiconn(e.g. IT modules)?
16:42:36 Quit PaulJam (".")
16:44:02kugelwere are all these GO_TO_ functions defined?
16:44:20kugelI search GO_TO_PREVIOUS_MUSIC but I can't find it
16:46:09GodEater_tags:GO_TO_PREVIOUS apps/root_menu.h /^ GO_TO_PREVIOUS = -2,$/;" e
16:46:09GodEater_tags:GO_TO_PREVIOUS_BROWSER apps/root_menu.h /^ GO_TO_PREVIOUS_BROWSER = -3,$/;" e
16:46:09GodEater_tags:GO_TO_PREVIOUS_MUSIC apps/root_menu.h /^ GO_TO_PREVIOUS_MUSIC = -4,$/;" e
16:48:06kugelIt's not defined there, is it?
16:48:55kugelI just see what happens when it's true
16:49:06kugelor not false
16:49:41GodEater_eh ?
16:50:04kugeloops, I was in root_menu.c
16:50:35GodEater_your grep-fu needs to be stronger
16:51:55nanokkugel: grep is your friend :)
16:54:26kugelI didn't know about grep...
17:00:57 Join MethoS-- [0] (
17:01:56kugelI allways wondered if there is something like that
17:05:00 Part LinusN
17:06:06 Quit petur ("connection reset by BEER")
17:13:44***Saving seen data "./dancer.seen"
17:16:00 Join japc [0] (
17:17:41 Quit MethoS--- (Read error: 110 (Connection timed out))
17:17:46 Join J3TC- [0] (
17:18:15GodEater_Llorean: ping ?
17:19:22lostlogicamiconn: IT? Keep it off my buffer? I mean if it can't be streamed it can't wrap the buffer in the current world...
17:20:15amiconnIT = ImpulseTracker
17:20:59amiconnOne of the many tracker module formats. Can be several MB in size, but must not wrap
17:23:51 Quit jgarvey (Connection timed out)
17:25:26lostlogicamiconn: I guess it could be called TYPE_AUDIO_ATOMIC, but I might want to make TYPE_AUDIO_ATOMIC not relocatable then.
17:27:57 Join toffe82 [0] (
17:29:50 Quit scorche|w ("CGI:IRC")
17:30:01 Join scorche|w [0] (n=8dc5049d@rockbox/administrator/scorche)
17:30:07 Quit TMM ("Ik ga weg")
17:33:00 Join webguest43 [0] (i=58a10615@gateway/web/cgi-irc/
17:33:11 Quit webguest43 (Client Quit)
17:33:13 Quit J3TC- (".UPP.")
17:33:16 Join webguest43 [0] (i=58a10615@gateway/web/cgi-irc/
17:33:31webguest43I'm a bit lost with e200tool
17:34:11webguest43I followed procedure but it always finish with a "Searching for device 6666:e200 ... 9 8 7 6 5 4 3 2 1 0 not found!"
17:34:28webguest43did someone sucessfully used it ?
17:35:23Domonokywebguest43: and your sansa is connected and in the right mode ?
17:35:53 Join mirak [0] (
17:35:59pixelmaso it's been told but 1) why do you have to use that and 2) I think it only works correctly under linux atm
17:36:12 Join princeofdark [0] (i=prince@
17:36:27GodEater_pixelma: it can be run under windows too apparently - but you have to use some dark rite involving chickens and a hedgehog to make it work
17:36:30 Part princeofdark
17:36:44scorche|wi thought it was a badger
17:36:47pixelmaaha... :)
17:37:02GodEater_depends if it's a tuesday or not
17:37:59Domonoky"the dark rite chicken" is a libusb driver file with correct pid and vid :-)
17:38:48webguest43Domonoky: not sure, with "hold" activated and "rec" pressed, nothing appears on the screen
17:39:19webguest43i followed these steps :
17:39:26linuxstbwebguest43: Are you using Linux?
17:39:45webguest43why, god, why di i put a 1.03 firmware ???
17:40:00webguest43linuxstb: yes , a ubuntu live one
17:40:46linuxstbThere shouldn't be any need to change the i2c bin file - that's a _very_ last resort.
17:41:08Domonokywebguest43: the other question is: why do you want to use e200tool ? :-)
17:41:36webguest43Domonoky: can't get in recovery mode
17:41:36linuxstbwebguest43: You can try our instructions -
17:41:54linuxstbBut it might be helpful if you explain the original problem.
17:42:00webguest43im french and have difficulties whith en glish
17:43:12hcslostlogic: I saw my name a ways back, still want me?
17:43:43lostlogichcs: just wanted to make sure spc still works on with my latest commit
17:43:49lostlogicbecause I made my previous commit less hacktastic
17:44:17linuxstbwebguest43: Do you have a file called "BL_SD_boardSupportSD.rom" ?
17:45:04 Quit webguest43 ("CGI:IRC (EOF)")
17:45:12 Join webguest43 [0] (i=58a10615@gateway/web/cgi-irc/
17:45:24webguest43linuxstb: ido have one
17:45:46webguest43ill the your page today, thanks
17:46:04webguest43i saw there was a mistake in dormrf page
17:49:12 Join illissius- [0] (
17:58:48webguest43it was not written to use sansa in "manufacturing mode"
17:59:01hcslostlogic: seems to be back to the original situation
17:59:52lostlogichcs: so it's freezing again :( crap, I'll fight with it again this evening, I forgot my USB cable this morning
18:01:05hcslostlogic: ok, thanks
18:02:23linuxstblostlogic: What's the issue?
18:02:35*linuxstb has only been half-following this saga
18:02:54lostlogichcs: I don't know −− I had added a 'force' flag that says "when a file is explicitly requested, buffer the whole damn thing"
18:03:00lostlogicer linuxstb
18:03:44pondlifehcs: You using a Sansa, or a flash device?
18:04:18 Quit localhero ("Leaving.")
18:04:25lostlogicso now I changed it to just not allow buffering to be interrupted by a buffer call on files that are 'atomic' ie codecs and atomic_audio
18:04:34hcspondlife: a flash-based Sansa, which I'm fairly certain is the issue. not having to wait for the disk to spin up probably tips a race condition in the wrong direction
18:05:09pondlifelostlogic: Sounds like the issue Sansa owners are reporting with MP3 on the forum...
18:05:25lostlogicpondlife: likely related
18:05:51pondlifehcs: You able to pop a logf up on Flyspray?
18:06:08linuxstbJust so I'm clear, the bufread and bufgetdata functions are non-blocking at the moment?
18:06:30pondlifeHuh? Surely they must block?
18:06:43 Quit illissius` (Read error: 110 (Connection timed out))
18:06:43 Quit webguest43 ("CGI:IRC (EOF)")
18:06:52*pondlife hasn't looked at the code...
18:06:53hcspondlife: I could build one if you like. Do I need to put a #define LOGF_ENABLE at the top of anything but playback.c?
18:07:05pondlifebuffering.c too
18:07:21lostlogiccorrect, nonblocking
18:07:22linuxstbpondlife: I don't think so - which is why there's the ERR_DATA_NOT_READY result code
18:07:28pondlifeAh, ok
18:07:32Nico_Phi everyone
18:07:35pixelmaI also got occasional freezes on my Sansa (c200) with mp3s, latest build I had on (and experienced this) was 15397. Will try latest as soon as it is charged again
18:07:43pondlifeI'm out of date, I always envisaged them as strictly blocking
18:07:52lostlogicNico_P: <3, I rebroke spc playback, but I think the changes I made are correct nonetheless :(
18:08:11linuxstbBut looking at codec_request_buffer() (the functions codecs call), that does seem to block.
18:08:22lostlogiclinuxstb: yeah, that does
18:08:34lostlogiclinuxstb: the problem hcs gets is that it blocks forever (at least that was the problem before recent changes)
18:08:58Nico_Plostlogic: maybe bufread and bufgetdata could become blocking
18:09:15linuxstbAlso, if bufgetdata() returns zero, then codec_request_buffer() would appear to return zero.
18:09:29pondlifeNico_P: You seen the forum reports of Sansa/mp3 issues too?
18:09:41Nico_Ppondlife: no, I hardly ever read the forums
18:09:59pondlifeOK,heads up then... I asked them to Flyspray it, but don't know if they have.
18:10:02lostlogicNico_P: yeah, maybe they should −− I'm not sure if it would make the current problems worse or better
18:10:29pondlifeWould it be simpler? I'd think so...
18:10:42lostlogicpondlife: I think it woudl certainly put logic more in the 'right' placse.
18:11:02linuxstbAssuming there are no cases where we would want them to be non-blocking?
18:11:11 Quit MethoS-- (Read error: 110 (Connection timed out))
18:11:18pondlifeI originally thought of bufopen/bufread/bufclose as being like open/read/close...
18:11:28 Join MethoS-- [0] (
18:11:51linuxstbYes, but POSIX speaking read() can return less data than requested.
18:12:14linuxstbBut I agree, a blocking read would seem to make sense to me in this context.
18:12:20lostlogiclinuxstb: which was what _I_ thought it did, but we were both wrong ;)
18:12:31Nico_Pthat's my fault :)
18:12:45 Quit ramon8 (Read error: 104 (Connection reset by peer))
18:12:47*linuxstb mutters something about documentation ;)
18:13:07linuxstbe.g. buffering.h could describe the semantics of the functions in more detail.
18:13:48pondlifelinuxstb: bufread could also return less data than requested, in my original thinking..
18:14:01lostlogicpondlife: would force all codecs to have loops
18:14:18linuxstbNico_P: Another issue which doesn't seem to be documented, is that bufgetdata() is only guaranteed to return up to GUARDBUF_SIZE bytes (because of the buffer wrap).
18:14:18pondlifeNot a biggie, just saying.
18:14:36linuxstb(or at least, that was the old way it worked)
18:14:48Nico_Plinuxstb: you're right
18:15:13Nico_Pactually it can return more even in case of a wrap
18:15:45lostlogicNico_P: worst case guarantee is guardbuf
18:16:39Nico_Plostlogic: I see you changed the codec loading logic
18:16:53lostlogicNico_P: yes, needed done, was needlessly complex
18:18:14Nico_Plooks good, yeah
18:18:36lostlogicNico_P: I kinda like that codec_load_buf anc coded_load_file are now parallel logic
18:19:01hcspondlife: is there an existing bug to add this to?
18:19:09pondlifeNot as far as I know.
18:19:22pondlifeThis one seems to be Sansa (or flash-based) only.
18:19:37pondlifeI asked the forum reporters to log one, no idea if they have.
18:19:49Nico_Plostlogic: shall I make bufread and bufgetdata blocking?
18:20:10lostlogicNico_P: assuming there is no current usecase that requires the nonblocking behavior
18:20:20Nico_PI don't think so
18:20:27lostlogiccool, then do it :)
18:20:42lostlogiclike I said to hcs, I'm non-dev until this evening since I forgot my cable
18:20:58Nico_Pooh segfault in my sim
18:21:15lostlogicNico_P: I probably broke sim_codec_load_ram, I don't know sim code.
18:21:30Nico_Pthat's exactly where the segfault happened
18:21:31lostlogicNico_P: (because I've literally only used the sim like once)
18:21:54*Nico_P loves the sim
18:24:46Nico_Plostlogic: one thing we would take care of is that codecs are discarded after being loaded which makes it necessary to spin the disk to reload them on a backwards skip, even when the audio is in the buffer
18:24:48pondlifehcs: Could you maybe also attach an example SPC file?
18:24:55 Join localhero [0] (n=dinzdale@
18:24:59pondlifeIf you have a suitable one, I mean
18:25:30hcspondlife: sure, it just needs a few (a single one doesn't exhibit this bug), so I'll attach a zip of a few
18:25:31pondlifeNico_P: Only if using a different codec?
18:25:55pondlifeI mean, a backwards MP3 -> MP3 wouldn't need a spinup, right?
18:26:02Nico_Ppondlife: yeah, say you have an ogg and an mp3, once the mp3 starts playing, you can't go back to the ogg without spinning the disk
18:26:07Nico_Ppondlife: no
18:26:26pondlifeThat's reasonable - it's also true of forward skip, right?
18:26:26lostlogicNico_P: yeah, but I'm not sure how to handle that because we'd need another way to reap the codec then
18:26:35pondlifei.e. one codec at a time..
18:26:39lostlogicpondlife: forward skip would have the codec on buffer
18:26:46Nico_Ppondlife: no, for forwrd skip the codec is in the buffer
18:26:47pondlifeAh, right
18:26:56pondlifeOK. maybe not so good then!
18:27:10pondlifeBut better than pre-MoB either way.
18:27:13Nico_Pso it would make sense to keep the codec on the buffer
18:27:21lostlogicpondlife: codecs were on buffer pre-mob
18:27:35pondlifeYes, but previous track data was cleared faster.
18:27:53hcspondlife: ok, attached
18:27:55pondlifeSkip back nearly always needed a spinup
18:27:56lostlogicpondlife: I think in that regard they were about the same... unless I"M misremembering
18:28:19lostlogicthe buffer wasn't immediately invalidated
18:28:40pondlifeNico_P will know for sure, but I thought that MoB was better at dealing with reuse of previous tracks.
18:29:05Nico_Phonetsly I don't know for sure but I had a feeling MoB was better
18:29:23Nico_Pit seemed to me that pre-MoB would need rebuffering more often
18:31:12pondlifelostlogic: Did you see my cheeky bug-report:
18:31:25pondlifeMore of an FR though
18:31:51lostlogicpondlife: yeah, I saw that and rolled my eyes
18:32:07pondlifeHey, it's already buffered!
18:32:20pondlifeOne for the future, though
18:32:27lostlogicbut playback resets its damn self
18:32:48pondlifeThe other interesting (and probably codec-related one) is
18:32:53lostlogicit explicitly calls close on all the handles
18:33:31lostlogichmm, I wonder why MoB would impact gapless :-\
18:33:42Nico_Plostlogic: what should I deduce from the chunk you removed from codec_load_ram? the wrapping case doesn't need treating there anymore?
18:33:57pondlifeBut does closing a handle invalidate the handle? Couldn't it just mark it as available when needed?
18:34:13lostlogicNico_P: it's moved to codec_load_buf by using the bufread call which handles the wrap-and-copy internally
18:34:13pondlifeI mean, internally to buffering, of course.
18:34:24lostlogicpondlife: that's a feature I'd love to see.
18:34:33pondlifeMe too
18:34:53lostlogicthen clients can 'close' a handle and get instant reopen, would slso solve the skip-back codec spin
18:34:55Nico_Pwhy can't playback.c handle that? it already kinda does
18:35:12pondlifeIt's not a playback problem, it's a buffering thing
18:35:21pondlifeMight apply to other buffer users too
18:35:27Nico_Phmm yeah maybe
18:35:34lostlogicbufopen would hafta search the buffer for an extant copy of that fd.
18:35:49Nico_Plostlogic: I was thinking of multiple bufopens of the same file returning the same handle id
18:36:09lostlogicNico_P: that would solve my long standing complaint that a repeat playlist of 1 file buffers)
18:36:23pondlifeSMARTDRV - now on Rockbox!
18:36:23lostlogicNico_P: but we'd hafta track the 'set of file handles' so that the bufadvance works on each copy separately
18:36:31Nico_Pa usage counter would be needed, and closing would change semantics
18:37:05Nico_Pdo we really want to manage position on all copies?
18:37:06lostlogicI think the first step is to let a _closed_ handle be reopened with the old hid
18:37:18 Join A_M [0] (
18:37:30lostlogicNico_P: pragmatically, not really, but to offer guaranteed semantics to each client we would
18:37:36Nico_Plostlogic: then bufclose doesn't call rm_handle anymore and we have a garbage collector
18:37:51lostlogicright, if refcount == 0, reap it when needed
18:38:03 Join bertrik [0] (
18:38:22Nico_Pcould also make things easier in playback.c
18:38:52pondlifeMight need to physically close files though, we have a limited number of opens allowed..
18:39:04Nico_Pthat's already done
18:39:17Nico_PI learned the limit the hard way...
18:39:19lostlogicpondlife: once the physical fd is done buffering it closes
18:39:35lostlogicthe bufapi will automatically reopen if needed to seek earlier than offset
18:41:29pondlifePlayback shouldn't care,. it just asks for data..
18:42:16 Join saratoga [0] (i=9803c6dd@gateway/web/cgi-irc/
18:43:12saratogaNico_P: Regarding Sansa + Mp3 playback, every couple tracks my sansa simply stops playing and dumps me back to the file view
18:43:25saratogaif i click on the track that was playing, it resumes from the beginning
18:43:43Nico_Plostlogic: committed the sim fix
18:43:45saratogathis is with a build from yesterday (i updated when I comitted the sansa battery life fix)
18:43:49Nico_Psaratoga: oh, that's bad
18:44:04lostlogicsaratoga: try the build today, JinC my changes change your situation
18:44:08Nico_Plostlogic: could that problem be solved by your latest commits?
18:44:09lostlogicNico_P: thanks :)
18:44:24Nico_Ppixelma also mentioned something similar
18:45:58*pondlife thinks ata_disk_is_active() should be put into a good_time_to_buffer() which always returns true for flash targets..
18:46:01saratogai also have very abnormal pausing when the sansa first powers up
18:46:19saratogait seems read from the flash for 10-15 seconds when the device first enters the file or database view
18:46:23saratogaafterwards it works normally
18:46:38saratogaduring this time backlight works, but the UI is frozen until it is done
18:46:41Nico_Pthat doesn't sound MoB-related
18:47:08saratogai assumed it was since it started with the MOB builds, but I could be wrong
18:47:24pondlifeHmm, gotta go...
18:47:27 Part pondlife ("Gone")
18:47:45lostlogicsaratoga: if audio isn't playing at the time then it can't be MoB...?
18:48:11saratogawell i assume its trying to buffer something
18:48:16saratogawhat else could it be reading from the disk
18:48:48Nico_Pmaybe the database files?
18:48:49lostlogicsaratoga: but nothing other than playback uses the buffering api at this time, sounds more like a dircache issue or something
18:54:18pixelmathe problems I saw are not the same as saratoga's. For me the WPS just freezes and I'm not dropped back to the browser, I also don't have a lag when entering database or file browser
18:55:25 Part toffe82
18:57:43 Join toffe82 [0] (
18:58:12 Quit MethoS-- (Read error: 110 (Connection timed out))
18:58:34 Join MethoS-- [0] (
18:58:36lostlogicpixelma: wps freezes, or playback stops? Can you browse to the buffering thread screen when it stops to give us debugging data?
18:58:52pixelmawith r15409 I see "disk" activity every 10 seconds
18:59:14pixelmaor about
19:00:18pixelmalostlogic: [16:11:02] <pixelma> Nico_P: playback just "froze" in the wps on my c200, revision is 15397. I'm just playing mp3s, the pcm buffer bar in the "view buffering thread" was empty. I could stop and resume without a problem (already had one occasion of this yesterday with an earlier revision)
19:04:57lostlogicpixelma: can you watch the buffering thread on 15409 (or later) and see what is causing said 'disk' activity?
19:08:04pixelmawow... it even got worse for me now. In the middle of the second track I'm playing now (first was an .ogg now an mp3) it started to constantly hit the disk according to the status bar and then after a while (maybe a minute?) there is a "dropout" but it continues playing
19:10:17 Join jgarvey [0] (
19:10:22lostlogicpixelma: can you try with... 1s, lemme find the rev...
19:11:18lostlogic15389? I think you'll find same behavior there
19:11:29pixelmalostlogic: there's a bit of restriction for me because some things were disabled for the c200 buffering thread screen because not everything fit but sure I can try. Anything specific I should have an eye on?
19:11:53pixelmawill do
19:11:55lostlogicI think that for any non-sid and non-nsf codec the two should behave the same.
19:12:19lostlogicback later
19:12:58 Join bluey [0] (
19:13:23 Join desowin [0] (
19:13:49***Saving seen data "./dancer.seen"
19:17:15 Join petur [0] (n=petur@rockbox/developer/petur)
19:20:04 Join perrikwp [0] (i=98212cf1@gateway/web/cgi-irc/
19:21:23 Join bluey_ [0] (
19:22:02 Quit bluey_ (Client Quit)
19:22:54 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
19:23:32 Quit thegeek_ (Read error: 110 (Connection timed out))
19:25:56 Join midgey [0] (
19:26:00 Quit A_M ()
19:26:51 Join donsdw [0] (
19:28:36 Quit Guile`` ("KVIrc 3.2.0 'Realia'")
19:30:18 Join thegeek [0] (
19:30:31 Join zicho [0] (
19:32:54 Join perrikwp_ [0] (
19:33:05 Quit perrikwp ("CGI:IRC (EOF)")
19:33:05 Nick perrikwp_ is now known as perrikwp (
19:38:19 Quit bluey (Read error: 110 (Connection timed out))
19:45:39pixelmalostlogic: yes, r15389 shows the same problem. The "fun" thing is that I can break this cycle if I leave the wps... and that reminds me of something I read (not sure if it has to do with it)
19:47:00Nico_Ppixelma: what about r15390?
19:47:39Nico_Ppixelma: uh, disregard that, it should be that same as r15397 buffering-wise
19:48:33 Quit jhulst ("Konversation terminated!")
19:50:45 Quit Galois (Remote closed the connection)
19:51:04 Join Galois [0] (
19:54:46 Quit MethoS-- (Read error: 110 (Connection timed out))
19:56:55 Part donsdw
19:57:24saratogaNico_P: with the current build it seems to run down the buffer on my sansa and never rebuffer
19:57:48saratogaso i can start the first track in a directory and it plays for 29MB worth of the album and just stops
19:58:44 Join tchan1 [0] (
19:58:48Nico_Psaratoga: have you looked at the buffering thread debug screen while this was occuring?
19:59:06Nico_PI'd like to know if the "usefl" value goes down to zero
20:00:07Nico_Pif you haven't looked, you can speed up the process by skipping to the last buffered track and letting audio play until the buffer empties
20:03:53 Quit zicho (Remote closed the connection)
20:04:09 Join zicho [0] (
20:04:38 Quit linuxstb (Remote closed the connection)
20:04:49 Join kkurbjun_ [0] (n=kkurbjun@alamode.Mines.EDU)
20:04:50 Quit tchan1 (SendQ exceeded)
20:04:59pixelmait looks like I get the disk activity every 10 seconds if it has to (re)buffer during an .ogg and the almost constant disk activity during mp3s both can be interrupted by leaving the wps (at least in r15389 where rebuffering also works - will switch back to latest and see if it's any different)
20:05:16 Join linuxstb [0] (n=chatzill@rockbox/developer/linuxstb)
20:05:51 Join tchan1 [0] (
20:05:55 Quit GodEater_ (Read error: 113 (No route to host))
20:06:16 Join obo [0] (n=obo@rockbox/developer/obo)
20:06:28*Nico_P needs a shortcut to the buffering debug screen :p
20:06:35 Join markun_ [0] (n=markun@rockbox/developer/markun)
20:08:36 Quit tchan1 (SendQ exceeded)
20:08:55 Quit tchan (No route to host)
20:09:11pixelmaNico_P: you said you don't use some callbacks anymore which were used before? Maybe it has anything to do with that? It's a wild wild guess but I remember that something had to be adjusted there once on the Ondio where it also showed constant "disk activity" - I hope amiconn remembers...
20:09:12 Quit hcs ("Leaving.")
20:09:12 Join tchan1 [0] (
20:10:08Nico_Ppixelma: I was previously using the ata_disk_is_active callback, which was disabled (or rather: not used) on targets with less than 8MB RAM.
20:10:33 Quit zicho (Remote closed the connection)
20:10:40Nico_Phow much RAM does the c200 have?
20:12:03Nico_POTOH, on the c200, ata_disk_is_active is always false so another condition is used...
20:12:23Nico_Pbtw, the callback I'm not using anymore isn't ata_disk_is_active, it's ata_idle_callback
20:13:01 Join zicho [0] (
20:13:25pixelmaok, I have no idea how that works just wanted to throw something in to get brainstorming for you going :)
20:13:35 Join ameeen [0] (
20:13:47ameeenhi can someone tell me how to copy photos from ipod to my pc ?
20:14:22saratogaNico_P: yes that buffer counts down to basically zero
20:14:29saratogait ends up having only a few KB left in it
20:14:33Domonokyameeen: just copy n paste... at least for rockbox :-)
20:15:07Nico_Psaratoga: and I assume there is still data remaining to buffer? ("data_rem" IIRC)
20:15:13 Quit tchan1 (Connection reset by peer)
20:15:31saratogaNico_P: yes, 29MB worth
20:15:38 Join TotallyInfected [0] (
20:15:43 Join tchan1 [0] (
20:15:52Nico_Psaratoga: are you sure that's the data remaining to buffer?
20:15:59Nico_Pnot the data in the buffer
20:16:15ameeenDomonoky I am going to install linux .. cuz rockbox isn't good for games
20:16:23saratogadata_rem: 29790761
20:16:40ameeenand themes/graphics are also available for linux, I guess
20:16:56saratogaNico_P: the sansa has a weird ATA setup
20:17:05Domonokyameeen: only rockbox stuff in this channel please... for linux you are on your own..
20:17:07saratogawhere they hacked the flash to look like an ATA disk or something bizzare
20:17:23Nico_Pin that case the ata setup shouldn't matter.
20:17:25saratogait screws up some of the plugins too, are you assuming anything about the flash on the sansa?
20:17:51saratogaif so you might want to ask Bagder or one of the sansa dev's about it
20:20:18pixelmaduring that "disk activity" even volume changes in the wps cause "dropouts" in playback
20:21:00pixelmaI'll go watching the buffering thread screen.
20:21:19Nico_Ppixelma: that screen is your friend ;)
20:21:27*Nico_P is starting to think he needs a sansa
20:21:27 Quit bertrik (Connection timed out)
20:22:01 Join dandin1 [0] (
20:22:35pixelmaNico_P: should data_rem be updated from time to time?
20:23:16Nico_Ppixelma: not often. it's that amount of data from the files that isn't in the buffer yet
20:23:27Nico_Pso once the buffer is full, it shouldn't change until the next rebuffer
20:24:06Nico_Psaratoga: the last track you have on buffer is big, right?
20:24:30pixelmait currently is only 12MB for me but then I broke the disk activity cycle by leaving the wps before
20:26:53 Nick tchan1 is now known as tchan (
20:28:24saratogaNico_P: no its actually fairly small
20:28:39saratogathough only 10 seconds worth of it fit in the buffer
20:28:42ameeenDomonoky does rockbox can play games on ipod ? and can u breifly tell me how rockbox is better than ipodlinux ? please
20:30:26Nico_Psaratoga: isn't it 29 MB?
20:31:07saratogaNico_P: No 6.4MB
20:31:09Nico_Ponce the buffer is full, data_rem should be the amount of data left to buffer for the last file
20:31:20saratogai assume 29MB is the sum of the remaining tracks in the album?
20:31:30saratogahmm no its not
20:31:31pixelmaNico_P: should data_rem and usefl be somehow related?
20:31:41saratogai don't know what the 29MB refered to then
20:31:42Nico_Ppixelma: no
20:31:44 Quit nicktastic ("Leaving")
20:32:46pixelmacould you explain them shortly? Or at least what the more important figures for you are there...
20:34:17Nico_Pusefl is the amount of unplayed data left in the buffer (that's why it decreases constantly)... data_rem is the amount of data left to buffer for the files that are partly buffered
20:34:45Nico_Palloc is the amount of data reserved, real is the amount of data actually on buffer
20:35:50 Join TMM [0] (
20:35:50 Quit scorche|w ("CGI:IRC (Ping timeout)")
20:41:03lostlogicseriously, why does sansa have all these problems that other targets don't... we have other flash targets that aren't having the trouble, don't we?
20:41:33amiconnNico_P: The ata callback exists on all targets, and should be used for the purpose of hooking into a spinup cycle for buffering purposes
20:42:08pixelmaah, see a bit clearer now, thank you. Then I can report a weird progress already... I entered the screen shortly before I expected a rebuffer which it did, usefl only climbed to 4MB first then fell again and emptied again, the following rebuffer filled it completely (or almost)
20:42:30pixelmaat Nico_P of course
20:42:41amiconnUsing ata_disk_is_active is not recommended because of its varying behaviour depending on target. Some even have a plain wrong implementation which always returns false (the Sasnas, obviously)
20:43:37amiconnEven on hdd targets, monitoring ata_disk_is_active is an unreliable method
20:43:49 Quit desowin ("use linux")
20:44:51 Quit Toxicity999 (Read error: 104 (Connection reset by peer))
20:45:41pixelmaNico_P: if that matters, my playlist contains usual music files (length varying between 3 and 9 minutes maybe, mainly mp3 sprinkled with ogg, bitrate higher than 192 kbps)
20:45:50amiconnIn addition, this method is bad because it could intercept the disk access which originally triggered the spinup
20:46:16amiconn...while the callback activates *after* the original disk access finished
20:46:16saratogaamiconn: since teh ata_disk_is_active seems to cause problems, could we just make it return true?
20:46:43saratoga(on the sansa)
20:47:01 Quit japc (Remote closed the connection)
20:47:28amiconnIt should be implemented correctly, but still *not* be used at the app layer like in buffering. Maybe it should even be removed, after checking carefully what is currently relying on it
20:48:02amiconnReturning 'true' all the time causes about the same amount of trouble as returning false all the time
20:48:22 Join Arathis [0] (
20:48:41amiconnI don't understand why the sansa implementation is borked; on Ondio, ata_disk_is_active() works like it should - and it's a flash target
20:49:57 Join cooz [0] (
20:49:58amiconnIf the callack causes problems with thread synchronisation, we should rather fix this, e.g. by implementing a method to 'register' for sysevents which then must be replied before the originating thread continues
20:50:43amiconnRather similar to how usb mode works today, just that threads which don't care about certain (most, or even all) events don't have to handle them
20:50:59pixelmaso do I understand correctly that MoB "just" revealed a different quirk?
20:51:01 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
20:51:21pixelmaor a bit of both
20:51:32amiconnThis method is needed for proper shutdown handling too, and a precondition for implementing suspend properly
20:55:03 Quit spiorf (Read error: 104 (Connection reset by peer))
20:56:38 Quit obo ("bye")
20:57:06amiconnpixelma: MoB revealed a bug in the sansa's ata_disk_is_active() implementation - by using it for something it shouldn't be used for though.
20:58:05 Join spiorf [0] (
20:59:41 Join bertrik [0] (
21:01:06 Quit mf0102 ("Verlassend")
21:01:50 Join bluey [0] (
21:02:21 Quit TMM (Connection timed out)
21:02:47 Quit midgey ()
21:04:19 Join obo [0] (n=obo@rockbox/developer/obo)
21:05:37 Quit BigBambi (Remote closed the connection)
21:05:50 Quit krazykit ("brb, installing more memory")
21:09:12 Join merbanan [0] (n=banan@
21:11:00 Join perrikwp_ [0] (
21:13:52***Saving seen data "./dancer.seen"
21:15:34 Quit spiorf (Remote closed the connection)
21:17:18 Quit cooz ()
21:18:19Nico_Psorry, I was off for dinner. I'll work on a solution without using ata_disk_is_active... thing is I'm not very satisfied with how late the ata idle callbacks are called
21:18:33Nico_Pbut I can probably work around that
21:20:43Nico_Panyway, I'm off now
21:20:50 Quit Nico_P (Remote closed the connection)
21:21:44 Quit saratoga ("CGI:IRC (EOF)")
21:21:55 Join J3TC- [0] (
21:23:17 Quit ameeen (Read error: 110 (Connection timed out))
21:24:25 Quit perrikwp (Read error: 110 (Connection timed out))
21:25:55 Quit amiconn (Nick collision from services.)
21:26:03 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
21:28:40 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
21:31:23 Join Veinor [0] (
21:32:09VeinorI just installed rockbox on my iPod, and I tried synching a playlist from amarok, but it creapted a bunch of folders in the root directory. So do I remove all the iTunes-related directories from my iPod?
21:32:55 Join krazykit [0] (
21:35:15 Quit bertrik (Remote closed the connection)
21:35:37linuxstbRockbox doesn't need any PC apps to work - you can just drag and drop files onto your ipod's disk. But you can use amarok or itunes if you wish...
21:36:25 Join bertrik [0] (
21:37:32Veinorbasically, my question is: if I'm not going to be using iTunes anymore, is removing all the old iTunes files and redoing my collection under however amarok or whatever organizes it a good or bad idea?
21:37:46Domonokygood idea
21:38:42Veinorso I can basically remove everytihng except .rockbox and it'll be OK?
21:40:29 Join scorche|w [0] (n=8dc5049d@rockbox/administrator/scorche)
21:42:56 Quit perrikwp_ ("ChatZilla [Firefox]")
21:43:19 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
21:45:26 Join webguest61 [0] (i=485afe2b@gateway/web/cgi-irc/
21:45:40webguest61is there any hacks for ipod nano 2nd gen
21:45:53kugelhcs: I don't get the message
21:46:18scorche|wwebguest61: this is #rockbox...for talking about rockbox
21:46:32webguest61is there rockbox for 2nd gen nano
21:46:48hcswebguest61: weird, someone just asked me this on AIM, asking the same questions
21:46:58webguest61that was me
21:47:01 Join karim_ [0] (
21:47:14webguest61maskedtomato right
21:47:16hcswebguest61: well my answers were true, believe it
21:47:23webguest61i know
21:47:28webguest61just wanted to make sure
21:47:50hcskugel: that's odd, you rebuilt it? and you're sure the DEBUGFs are being executed?
21:47:53webguest61is there rockbox for samsung yp-t9
21:48:06kugelI used printf
21:48:10scorche|wwebguest61: all the players rockbox supports are ont he front page
21:48:10hcswebguest61: there is rockbox for the systems listed on
21:48:11krazykitwebguest61, read the front page to see what is supported
21:48:17Domonokywebguest61: read the webpage, all supportet players are listed..
21:48:23BigBambiwebguest61: Check the frontpage
21:48:25*scorche|w sees himself, but wonders who won
21:48:33krazykitscorche|w, looks like you, here
21:48:35webguest61what should i click on when i go to
21:48:36BigBambiscorche: you
21:48:44*Domonoky nearly lost... :-)
21:48:45BigBambiwebguest61: It is on the front page
21:48:52scorche|wwebguest61: we didnt say a link....we said ON the front page
21:48:58hcskugel: try switching to DEBUGF, but if it links alright then printf should have worked
21:48:59*BigBambi just joined in for the hell of it
21:49:39webguest61do you know anything about samsung y-t9
21:49:48scorche|wwebguest61: this is #rockbox...for talking about rockbox
21:50:02BigBambiwebguest61: Any info we have about a possible port will be in the wiki
21:50:07scorche|wif there has been ...bah
21:50:36kugelhcs: just DEBUGF("string") ?
21:51:07hcskugel: yes. I'm trying to recall if there's anything else you have to do but I can't recall any.
21:51:13hcshopefully I'm not mistaken
21:51:19Veinordoes anyone in here know how to get amarok to sync playlists with rockbox?
21:51:45 Nick markun_ is now known as markun (n=markun@rockbox/developer/markun)
21:51:54scorche|wVeinor: well, that is more of an amarok question
21:52:05Veinorjust curious.
21:52:06*amiconn found something rather nice :)
21:52:11scorche|wtry something like #kde
21:52:14kugeloh wait
21:52:17 Join w1ll14m [0] (i=d55d29c4@gateway/web/cgi-irc/
21:52:19w1ll14mhi all
21:52:24kugelI did get the message
21:52:25amiconnThe mini G2 rom contains the same Hd on/off test as the G5.5 rom
21:52:38w1ll14mi have a problem ...
21:52:40kugelbut not until I exited the sim
21:52:46amiconnHence, hdd power handling on all PP502x iPods is most probably the same
21:52:49w1ll14mi've installed rockbox on a nano 1g
21:53:16w1ll14mwhen i try to boot rockbox it says no partition found. insert blablabla
21:53:31w1ll14mfilesystem is fat32
21:53:37 Join bertrik_ [0] (
21:53:38pixelmahcs: still about the probably mob related problems on e200?
21:53:44kugelhcs: did you read?
21:53:52scorche|ww1ll14m: what did you do to the device?....format it? install ipl?
21:54:11w1ll14mscorche it's brand new, installed rockbox bootloader
21:54:17w1ll14mand installed rockbox it self
21:54:19hcskugel: yes, but I can't understand what's wrong. it should display when the DEBUGF is executed. are you sure it is hitting that statement when you think it is?
21:54:34scorche|wnever heard of that happening with a brand new device...
21:54:40scorche|wdo you have access to itunes?
21:54:50w1ll14mwhen i remove the rockbox files from the ipod
21:54:51kugelhcs: It did display the message, but not until I exited the sim
21:54:53hcspixelma: hmm? as far as I know those problems still exist, what do you mean? that's not what I'm talking about now
21:55:00BigBambiamiconn: very nice
21:55:03pixelmaw1ll14m: a nano 1st gen brand new?
21:55:04w1ll14mthe rockbox bootloader says i have to install rockbox files
21:55:12scorche|was it should..
21:55:12 Quit webguest61 ("CGI:IRC (EOF)")
21:55:13kugelhcs: still using printf, I might add
21:55:14w1ll14mpixelma: yes 1st gen new
21:55:16 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
21:55:46w1ll14mscorche: indeed, and when i install the files i can see the rockbox boot logl, then it says the error mentioned above
21:55:47pixelmahcs: yes they still exist, sorry misunderstood
21:56:03w1ll14mboot logo**
21:56:21hcskugel: I don't know what could cause that
21:56:31scorche|ww1ll14m: as i said, i have never seen that happen with a brand new you have access to itunes?
21:56:52w1ll14mscorche: yes i do, i've restored it allready once
21:56:55w1ll14mi can still boot OF
21:57:19w1ll14mi've restored it twice ;)
21:57:46 Join oblib [0] (
21:57:47linuxstbSounds like the usual Nano disk access bug?
21:57:49scorche|wand it did this each time?
21:58:03pixelmascorche|w: weren't a very few people with affected Nanos reporting the same (IIRC there were more and less heavy cases...)
21:58:31w1ll14mscorche: yes it did
21:58:36oblibwhat are you talking about? I've had problems with my nano
21:58:39w1ll14mlinuxstb: which nano disk bug ?
21:58:50 Nick Veinor is now known as Veinor|away (
21:58:51scorche|wpixelma: that is what i am beginning to suspect, but i am not all on the up-and-up on it
21:59:00 Part Veinor|away
21:59:07pixelmahmmm... is the forum down?
21:59:20krazykitpixelma, appears so
21:59:37w1ll14mit looks like it
22:01:08 Join bistouri [0] (i=58a10615@gateway/web/cgi-irc/
22:01:32kugelhcs: DEBUGF is working better, thanks for the hint
22:01:41bistouridumbest question of the day : is there a specifical virus for Rockbox ?
22:01:54pixelmaw1ll14m: then (there was a thread about it in the forums I wanted to point you to)
22:02:12w1ll14mlet me check this one ;)
22:02:43Domonokybistouri: not until now.. but have good idea.. *hehe*
22:02:59 Quit n17ikh|Lappy ()
22:03:06hcsbistouri: not that I've heard of. it spans many architectures, and changes frequently enough that there really wouldn't be anything stable to target just yet
22:03:27 Quit dandin1 ()
22:03:35 Quit bertrik (Connection timed out)
22:03:52kugelhcs: I think a virus corrupting the LCD driver would do damage over a good period of time
22:04:01Domonokywe should build one into it.. if you use doom too much.. *buumm* *hehe*
22:04:11pixelmaw1ll14m: as I already said - people get different kind of problems. The relation to a certain OF revision was the first theory but that doesn't seem to be true, next idea was that it is hardware related somehow
22:04:18bistourinote im not interested in creating one, of course, just curious :)
22:04:30 Join Soader03 [0] (i=18c9c5e4@gateway/web/cgi-irc/
22:04:49bistouriAs i noticed small OS'es have their own as MenuetOS
22:05:21Soader03Is the forum down or it is just my computer?
22:06:15pixelmasame here
22:06:22karashataSoader03: apparently the forums are down
22:06:32karashataat least, for now
22:06:42w1ll14mpixelma: ok, i'm gonna create my own build
22:06:55w1ll14mis it usefull to downgrade my OF firmware ?
22:06:59 Join n17ikh|Lappy [0] (
22:07:02Soader03ok thanks
22:07:15 Quit moos (Read error: 110 (Connection timed out))
22:07:24 Join Domonoky_ [0] (n=Domonoky@
22:07:52pixelmaw1ll14m: don't think so, but I'm not sure
22:08:18pixelmaIIRC you were from the nederlands?
22:08:43w1ll14mi'm from the netherlands indeed
22:08:46Soader03Can someone redirect me to the download page of EvilG build? (Not in the forum)
22:09:14w1ll14mpixelma: i'm gonna try a custom build, after that i'm gonna try to overwrite my firmware partition with the older firmware
22:09:31karashataSoader03: <−− that what you're looking for?
22:09:57Soader03yes! thanks a lot
22:10:02pixelmaw1ll14m: was just asking in case your Nano suffers from the same and you would be willing to lend it to an interested developer ;)
22:10:26 Join SirFunk [0] (
22:10:44w1ll14mpixelma: well, i would but my sister won't like it i think ...
22:11:03pixelmahehe, ok. Understandable :)
22:11:13w1ll14mwhahahah i thought so ;)
22:12:16w1ll14mpixelma: you where form the netherlands too right ?
22:12:19kugelwhat means implicit declaration of function 'DEBUGF'?
22:12:38pixelmano, germany
22:12:49w1ll14mwell, close ;)
22:12:53*bertrik_ is dutch
22:13:03 Nick bertrik_ is now known as bertrik (
22:13:21 Join Chronon [0] (
22:13:23w1ll14mwhere do you life bertrik?
22:13:46 Join moos [0] (
22:13:59w1ll14mahh, i'm from alkmaar....
22:14:48hcskugel: it means there wasn't a prototype for it, which I suppose means that there isn't actually a macro for it as I thought
22:15:38kugelnono, in tree.c this worked fine.
22:18:07w1ll14mpixelma: pre 1.3.1 nano's don't have those problems /
22:18:50Chrononw1ll14m: I don't think that's the case.
22:18:50pixelmaas I said, that was the first theory but it didn't hold true
22:18:59pixelmaw1ll14m: no I remember, weren't you the one who overclocked his Ipod by... far?
22:18:59kugelhcs: Do I need to include a specefic file to make it work?
22:19:18w1ll14mpixelma: i still do (ipod video 5g 60GB)
22:19:30Chrononhow fast?
22:19:35 Quit karim_ (Remote closed the connection)
22:19:36w1ll14mrunnong normal @ 60MHz max 92 MHz
22:19:50pixelmaweird that it doesn't seem to cause major problems there
22:20:01 Quit Soader03 ("CGI:IRC (Ping timeout)")
22:20:18w1ll14mi also have the 30 GB ipod 5g ... that one is running at 105 MHz :D
22:20:32w1ll14mif i go higher i get data aborts
22:20:52lostlogicbertrik: are you on sansa?
22:20:53ChrononSo, it doesn't seem to be inherently a CPU issue.
22:21:10hcskugel: it looks like it is defined in firmware/export/debug.h
22:21:42w1ll14mChronon, i'm talking about a video 5g 30GB and 60GB, those differ alot from the nano's
22:21:45kugelyea, I looked at a file were it works, and there was debug.h included
22:23:23pixelmaw1ll14m: you aware of recent enhancements e.g. in lcd updates and so on? :)
22:23:38 Quit Domonoky (Read error: 110 (Connection timed out))
22:23:42w1ll14mlet me check my last build.....
22:23:53Chrononw1ll14m: yeah, but mostly in peripheral hardware, right? Are the cores actually different?
22:24:14w1ll14mChronon, every core is different....
22:24:23w1ll14mi don't have to try to let my 60GB run at 105 MHz...
22:24:25Chrononas in each instance?
22:24:26w1ll14mit won;t even boot
22:24:35w1ll14mbut my 30 GB will ;)
22:24:47w1ll14mrun at 105 MHz pretty stable and very fast
22:24:50Chrononsure. Manufacturing cannot make any two devices to be identical
22:25:03ChrononI see what you're getting at
22:26:01w1ll14mbut those cores are made to run at 80MHz max (most cores can be overclocked a bit, some more)
22:26:25 Quit bistouri ("CGI:IRC (EOF)")
22:26:26w1ll14mmy last build was 01-11-2007
22:26:53Chrononw1ll14m: Are we confident of that spec? I see that tossed around, but I have also seen a 100MHz figure tossed around as an upper limit
22:27:07bertriklostlogic: yes, i have a sansa e260
22:27:21w1ll14mi've read the docs for the pp5020
22:27:27lostlogicbertrik: just cataloging that pretty much all of the MoB bugs out there are (sadly) sansa
22:27:34w1ll14mas far as i can tell, they are designed to run at 80Mhz
22:27:37 Join The-Compiler [0] (
22:27:38lostlogicsomeone should send Nico (or me) a sansa to play with ofr a week or 3
22:27:44pixelmaw1ll14m: see commit of rev 15397 (November 2nd)
22:27:46Chrononw1ll14m: Okay. I'll yield to your expertise on that
22:27:52 Quit The-Compiler (Read error: 104 (Connection reset by peer))
22:28:08scorche|wlostlogic: ask rasher to loan you his brother's while it is still int he US ;)
22:28:19*scorche|w runs
22:28:19bertriklostlogic: sorry, you can't have mine, but I won't mind helping to debug stuff
22:28:47w1ll14mpixelma: which commit was it ?
22:29:17pixelma"Major speedup of the iPod Video LCD driver."
22:30:02w1ll14myes i've seen it this morning, i was very excited to test this one :d
22:30:25pixelma;rev2=38 for reference ;)
22:30:53bertriklostlogic: I can probe around a bit in the code on my own, if you have some pointers
22:32:01w1ll14mpixelma: looks very promising
22:32:56 Nick fxb__ is now known as fxb (
22:33:40w1ll14mi'm gonna try my build of yesterday for the nano
22:33:46lostlogicbertrik: nothing in particular at the moment, I tend to be very touchy feely with debugging (I play with it until I have a feel for what is broken and then sniff at the code until I find something that smells the same to poke iwth a stick)
22:34:44hcssounds like an auto mechanic extracting a dead squirrel from an engine
22:35:01w1ll14mpixelma: do you still remember why i overclocked my ipod ?
22:36:35pixelmalostlogic: did you see the conclusion earlier that the problems on Sansa are probably a combination of mob and a wrong implementation?
22:36:47bertriklostlogic: is it ok if I ask you a few questions about playback.c?
22:36:57pixelmaw1ll14m: no, I don't remember that
22:38:27 Quit karashata ("Leaving.")
22:38:36lostlogicbertrik: ask away, I'll do my best not to be wrong
22:38:58bertrikfor example, function audio_clear_track_entries looks like it is supposed to clear tracks between the write ptr and the read ptr
22:38:59 Join Toxicity999 [0] (n=bryan@unaffiliated/Toxicity999)
22:39:07bertrikI wonder if this shouldn't be the other way around
22:39:09lostlogicpixelma: yeah, but that only sufficiently answers the slow/inefficient buffering, it doesn't sufficiently answer the stoppages
22:39:18w1ll14mwell, i'm using software and hardware eq verry intensive
22:39:46w1ll14mso, my ipod is verry slow and my nice 64MB buffer is filled upto 5 or 9 MB's
22:40:03bertrikalso in three places, this function is called with track_widx == track_ridx, resulting in unbuffering *all* tracks instead of just the ones between track_widx and track_ridx
22:40:16bertrikI wonder if that's intentional
22:40:25lostlogicbertrik: the ones where it's called with them equal are intended to do exactly that.
22:40:42w1ll14mso when i use 60 MHz as normal speed and 92 as max speed the buffer fils to max and ipod is functioning normal
22:40:50lostlogicthe 'normal' case is to use it to unbuffer=clear all tracks other than the playing one IIRC
22:41:03lostlogicthe unusual case is to use it to 'free' played tracks
22:41:18lostlogic(calling the unbuffer callbacks as it goes)
22:41:29bertrikok, it wasn't immediately obvious to me
22:42:03 Join karashata [0] (
22:42:25lostlogicbut I'd hafta reread to code to know for sure (in case you weren't aware, I've been remarkably inactive on rockbox between last summer and the commit of MoB so my memory is flawed in places)
22:43:00pixelmaw1ll14m: well if you want to burn the already little battery life like that... I would probably just not use the equalizers or just a bit. In fact for myself I just use a little more bass and treble, that's all.
22:43:27w1ll14mpixelma: my battery life is pretty nice for the settings i use
22:43:34w1ll14moverall i'm pretty satisfied
22:43:35bertrikanother question: is there a reason why track_ridx is volatile and track_widx is not?
22:43:39pixelmaI wouldn't expect a mix of equalizers to sound any better
22:43:54w1ll14mif i won't use 92 Mhz the buffer won't fill to the max....
22:44:03w1ll14mand if you have these
22:44:36lostlogicbertrik: track_widx is only used on one thread, audio_thread, track_ridx is used on multiple threads
22:44:36w1ll14mit's verry nice to tune the eq
22:44:50lostlogicbertrik: tryt o prefix stuff to me with my name so I don't miss ;)
22:45:25Buscheloops, could anyone change FS #8075 from bug to patch?
22:45:28w1ll14mi use the softeq for very specific settings of the low frequencies and the hardware eq to compensate
22:45:29 Join darkapostrophe [0] (
22:45:41w1ll14mthe high freqiencies
22:45:43pixelmaw1ll14m: to me this sounds the same as the problem of people getting more and more pills one to "cure" the things others cause. But: to each his own... :)
22:46:06w1ll14mindeed :) that's why i love rockbox so much
22:46:26w1ll14mi can't life a day without it
22:46:44w1ll14mit's to bad my 60GB disk crashed
22:47:04w1ll14mso now i have a ipod 60GB (with a 30GB harddisk)
22:47:14moosBuschel: done!
22:47:33Buschelmoos: thanks!
22:47:41amiconnAhahaha: There's yet another magic constant in the apple of: 0x1337BEEF
22:47:48 Quit jgarvey ("Leaving")
22:48:16w1ll14mamiconn: that's a nice magic number
22:48:46pixelmahow 1337! :)
22:48:52w1ll14mindeed lol
22:49:02w1ll14mleet beef ;)
22:49:17*scorche|w wonders if amiconn is joking...
22:49:35amiconnIt's there; found it in the mini G2 rom
22:49:38 Join karim_ [0] (
22:49:56w1ll14mwhat does it do ?
22:50:07w1ll14midk ?
22:50:16 Quit karim_ (SendQ exceeded)
22:50:22amiconnCurrently I'm just kinda brute-force-disassembling the rom
22:50:39bertriklostlogic: the wrap variable near the end of function audio_check_new_track looks fishy to me
22:50:47bertrikvery confusing piece of code
22:50:47scorche|wamiconn: that cant be a coincidence ;)
22:50:48w1ll14mi just hate these fing apple guys....
22:51:10 Join karim_ [0] (
22:51:12 Part bluey ("Leaving")
22:51:19w1ll14mwould it be nice ti have the ipod touch with rockbox....
22:51:24amiconnw1ll14m: idk == I don't know
22:51:37scorche|ww1ll14m: sure
22:51:42w1ll14mamiconn: lol i'm not familiar with that one ;)
22:52:04w1ll14mis there some info about the specs of those 'new'ipods ?
22:52:43scorche|wnot really
22:52:44krazykitw1ll14m, not really.
22:52:50krazykit...ffs scorche.
22:52:55*amiconn is hunting for 2 things in the roms atm
22:53:08w1ll14mhmmm ... to bad, i was wondering what cores they use right now ....
22:53:11amiconnG5 backlight brightness and hdd power (all PP502x models)
22:53:15w1ll14mamiconn: what are you hunting for ?
22:53:25scorche|ww1ll14m: feel free to go and hunt
22:54:09w1ll14mbadger: kickass thanx!
22:54:42Bagderand then this
22:54:46lostlogicbertrik: yes, confusing code, but I believe it's correct −− when skipping backwards, a wrap has occured if the new track_ridx is > the old, I do see a problem that wrap is never cleared once cur_idx has made the wrap... hmm.
22:54:52Bagderbut not a lot to see there really
22:54:54lostlogicmight cause some unnecessary rebuffers.
22:56:46bertrikbut the track_ridx pointer may already wrap around by itselves, so you can't just compare it with an old track_ridx pointer I think
22:57:01 Quit Buschel ()
22:57:49 Join resky [0] (
22:58:03pixelmaBagder: do you perhaps know more about the forums being down? Don't expect you to but it doesn't hurt to ask ;)
22:58:17Bagdernope I don't
22:58:24lostlogicbertrik: not sure I follow −− as long as the skip is < MAX_TRACKS the movement from old_track_ridx to track_ridx can be said to have wrapped if track_ridx > old_track-ridx
22:59:40bertriklostlogic: not the other way around, track_ridx < old_track_ridx?
22:59:42 Quit Dark_Apostrophe (Connection timed out)
22:59:53 Join darkapostrophe__ [0] (
23:00:11bertrik(oh, that part is only invoked when skipping backwards)
23:00:26 Quit Toxicity999 (Remote closed the connection)
23:01:17bertrikthis feels like reverse engineering
23:01:19lostlogicbut I think inserting if (!cur_track) wrap=false; above the complex conditional is correct.
23:01:32lostlogicbertrik: we (I) should have commented that function better.
23:02:07bertrikplease don't feel offended :)
23:02:32 Join Toxicity999 [0] (n=bryan@unaffiliated/Toxicity999)
23:03:44 Join matsl [0] (
23:03:54 Quit w1ll14m ("CGI:IRC (Ping timeout)")
23:05:04 Join cooz [0] (
23:06:10lostlogicbertrik: <−− does that patch make it make sense?
23:06:24lostlogic(I haven't tested the minor change, but it appears correct)
23:08:05*bertrik is looking
23:10:28 Quit kkurbjun_ ("leaving")
23:11:50 Quit karim_ ("Ex-Chat")
23:12:12bertrikyes, it makes sense, but I wonder if this piece of code can be simplified further, without the need for a wrap variable
23:13:53***Saving seen data "./dancer.seen"
23:14:06lostlogicbertrik: go for it, submit a patch :)
23:14:15lostlogicbertrik: playback.c always needs more hackers
23:14:39toffe82why not rockbox on this : :)
23:15:03bertriklostlogic: ok, I'll keeping looking at it
23:16:33 Quit darkapostrophe (Connection timed out)
23:17:28 Quit kubrick ("leaving")
23:17:39jhMikeSamiconn: does gray lib automatically "paint" the overlay when gray_show(true) is called or does it need a gray_ub_gray_bitmap_part call afterwords to show it?
23:18:04amiconnIt 'paints' automatically
23:18:41amiconnAnd whether you can use gray_ub_gray_bitmap_part depends on which mode you use. Said function is for unbuffered mode *only*
23:20:23jhMikeSthat's what video out is using
23:21:11bertriklostlogic: playback just stopped, is it any use to tell you now what I'm seeing in debug menus?
23:21:22 Quit karashata ("Leaving.")
23:21:26jhMikeSbut no screen clearing when hidden or just a redraw of what's behind?
23:21:28 Quit Toxicity999 ("Leaving")
23:21:32 Join Toxicity999 [0] (n=bryan@unaffiliated/Toxicity999)
23:21:35lostlogicbertrik: yes
23:21:46 Join karashata [0] (
23:21:50lostlogicbertrik: in the buffering thread screen
23:21:59amiconnjhMikeS: gray_show(false) automatically calls lcd_update()
23:22:35bertrikpcm=0, alloc=26691728, real=1598976, usefl=16355
23:22:37amiconnThe grayscale lib doesn't touch lcd_framebuffer, so that's effectively a restore of what was there before
23:23:05 Join cocobomb33 [0] (i=414998bf@gateway/web/cgi-irc/
23:23:32lostlogicbertrik: usefl is 16k, what codec was it playing? also, is there a data_left or some such or not?
23:24:25bertrikmp3 was playing an album of 20 songs or so, data_rem=29647998, track count=8, handle count=16, pcmbufdesc=0/21
23:24:30 Join Daniel_S [0] (i=57b0ff4e@gateway/web/cgi-irc/
23:25:20bertrikI'm running SVN 15412 (latest)
23:26:07 Quit cocobomb33 (Client Quit)
23:28:32 Quit jhMikeS (Nick collision from services.)
23:28:38 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
23:30:11bertriklostlogic: I also have logf enabled, I'll dump it
23:30:44lostlogicbertrik: awesome, bor playback and buffering? that would be _very_ helpful!
23:31:05bertrikoh I only have logf enabled in playback.c
23:31:07 Join Ebert [0] (
23:31:17 Join ColinBeamer [0] (i=414998bf@gateway/web/cgi-irc/
23:31:20lostlogicbertrik: aww, add it to buffering.c too for next test :)
23:31:24bertriklostlogic: shall I enable it in buffering too and wait until it happens again?
23:32:17 Quit ColinBeamer (Client Quit)
23:32:25bertrikWhen I get a nice set of files, I'll put it on the bug tracker
23:32:41 Join Buschel [0] (
23:32:52 Quit Rob222241 ()
23:33:19 Quit Arathis ("Bye, bye")
23:33:32 Quit merbanan (Remote closed the connection)
23:34:22*lostlogic is bad about paying attention to flyspray
23:37:45 Quit Buschel ()
23:39:07 Join Rob2222 [0] (
23:39:24oblibSo I recently updated my build (been since May) and my Nano is busted. Is there any more recent info on this than the bug listed above (FS #7510)?
23:39:25 Nick darkapostrophe__ is now known as Dark_Apostrophe (
23:39:31 Join zivan56 [0] (
23:40:10pixelmado you know how old your bootloader is?
23:40:20karashataoblib: did you make sure your bootloader is up to date?
23:40:40oblibyeah, looks like it is September 2007?
23:40:50zivan56Just found some info on the Sansa View. It is portalplayer based (Nvidia) and the firmware can be decoded with mi4code keyscan/decrypt. Seems to follow the same basic bootloader/firmware design as the e200/c200
23:41:17zivan56wanted to post it to the forum, but it seems to be down...
23:41:28jhMikeSamiconn: I know why the frames slow does when not limiting FPS on the sim. It's just the video thread has gone fully around the buffer and is waiting for the audio thread to progress before more data can be buffered.
23:41:29oblibkarashata, got it from
23:41:32Bagderzivan56: cool!
23:41:50 Join resky_ [0] (
23:42:14zivan56Bagder: yep. Now if someone could test decrypting/re-encrypting the firmware
23:42:17Bagderzivan56: but given its video capability, it's still left to see how the display is controlled
23:42:43zivan56yep, but at least a binary could potentially be run on it
23:42:46karashataoblib: exactly what do you mean by "busted"?
23:42:51pixelmaoblib: and make sure that there is no rockbox.ipod file in the root of your player
23:42:54Bagderzivan56: you found a firmware image somewhere?
23:43:06zivan5670e19bda 0c69ea7d 2b8b1ad1 e9767ced (v 1.00.03) is the key. FIrmware is here:
23:43:10zivan56yep, someone posted it
23:43:14oblibpixelma, there is not a rockbox.ipd file
23:43:51amiconnjhMikeS: hmm....
23:43:54oblibkarashata, I have been messing with NSF files, and their various tracks get off sync after a short time. When I try .ogg files, it usually crashes post haste
23:44:33karashataI can't help you there, unfortunately
23:44:34amiconnHaha, try to read: "lppAoC etupm ,re.cnI"
23:44:37oblibI tried applying nas's fix (change back to 75 MHz) but that didn't help anything
23:45:30karashataI'm running the most recent build on my H10 and I'm not having issues myself, so...
23:46:18oblibYeah, it seems to only apply to "lucky" nanos, not all
23:46:41oblibTemperature sensitive too, I guess I could just get a fridge in the car...
23:47:33karashatawell, you could always find an older daily build that doesn't have those issues and use that until things clear up a bit more
23:47:49 Quit scorche|w ("CGI:IRC")
23:47:50oblibhow do I roll back my svn? I'm not familiar with the command
23:48:03 Join scorche|w [0] (n=8dc5049d@rockbox/administrator/scorche)
23:48:59 Join kubrick [0] (n=repulse@unaffiliated/funky)
23:49:00bertriksvn update -r XXXXX
23:49:33zivan56Bagder: some strings in the bootloader suggest PP5022 (maybe same PP5022+AS3514) on the previous targets? Too bad nobody has disassembled one yet
23:49:43oblibthanks bertrik
23:49:46 Quit jhulst (Connection timed out)
23:50:02zivan56although NV6100 is referenced as well :/
23:50:04Bagderzivan56: the e200 image says PP5022 as well
23:50:34BagderI'll inform mrh about it
23:50:42bertrikoblib: svn update tries to keep your patches, if you want a complete rollback you should use svn revert
23:50:47zivan56I guess the new nvidia hardware is still built upon the pp5022
23:51:03zivan56nice, hopefully he can shed some light on the hardware :)
23:51:21oblibbertrik, yeah, I just deleted the files I had changed and it restored them. I'll remember revert in the future.
23:51:35amiconnBagder: Yes, and the PP5024 identifies itself as PP5022
23:52:20zivan56I guess the Sansa View is built using this SoC:
23:52:36zivan56Dedicated high performance 250MHz ARM1176JZ-S core :)
23:53:48resky_how do i listen to wma files in my ipod?
23:53:57advcomp2019zivan56, that is what i have heard too
23:54:10zivan56I'll post an overview on the forum I guess
23:54:13jhMikeSamiconn: Apple Computer, Inc. :p
23:54:18resky_i installed rockbox and it works fine with mp3.
23:55:08amiconnjhMikeS: Yeah, silly little endian...
23:55:26 Join WalkGood [0] (i=WalkGood@unaffiliated/walkgood)
23:55:34 Quit resky (Read error: 110 (Connection timed out))
23:56:46amiconnjhMikeS: Btw, 0x60006044 seems to be another DEV_ENx
23:57:10jhMikeSDidn't strike me right off. I guess it is endian flipped.
23:57:34jhMikeSamiconn: DEV_EN3 ?
23:58:06amiconnThere are functions using a big jumptable which selects one of a whole bunch of code snippets, which set various bits in either DEV_EN, DEV_EN2 or 0x60006044
23:58:52jhMikeS0x60006044 is currently labelled as a cache register

Previous day | Next day