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).

#rockbox log for 2008-05-15

00:06:38 Quit stripwax (Read error: 104 (Connection reset by peer))
00:09:08 Quit ompaul (Client Quit)
00:10:09 Quit mf0102 ("Ex-Chat")
00:12:39 Join magmaniac [0] (
00:12:45 Part magmaniac ("Leaving")
00:15:06 Join curtmack [0] (
00:16:07 Join anon1532_ [0] (
00:16:14 Join stripwax [0] (
00:16:17curtmackOkay, before asking, I realize there are limitations to what can be done on an MP3 player. That said, an mpeg that I encoded tends to skip a bit on my iriver H10 20G; at certain points it'll freeze, and then later catch up. What options should I tweak to help ease this?
00:16:34curtmacklike, encoding options
00:16:47 Quit ender` (" Top reason why compilers are like women: Miss a period and they go crazy")
00:16:49 Part Anon2733
00:17:06n1scurtmack: bitrate
00:17:14curtmackthat's what I figured
00:18:05 Quit stripwax (Client Quit)
00:18:12n1salso it might be worthwhile to try different encoders, they can give quite different results in decoding speed and quality
00:18:35 Join stripwax [0] (
00:18:41curtmackwell I'll try lowering the bitrate first
00:18:46curtmackdidn't realize I had it set at 500 :P
00:23:13curtmackoh, and I've found that the harddup option for mencoder causes audio desynch, is it a mortal sin to remove that?
00:23:55 Nick anon1532_ is now known as anon1532 (
00:24:22 Nick anon1532 is now known as anon1532_ (
00:28:57 Nick anon1532_ is now known as anon1532 (
00:30:53curtmacklowered the bitrate, and now it doesn't freeze any of the two places where it used to, but does freeze in an entirely different spot
00:30:57 Join Nate263 [0] (
00:31:40 Quit gregzx ("ChatZilla 0.9.82 [Firefox]")
00:32:21curtmackwell, I'll lower it down to 150, that's as low as I want to go
00:33:03curtmackanything lower and I may as well be watching that stuff that you used to get when you'd watch a payperview movie you hadn't paid for
00:33:10curtmackback when payperview was based on actual channels
00:33:23 Join Shaid [0] (
00:33:33Nate263I screwed up my iriver H10 when I installed rockbox. I'm getting the system files missing error and I connected it to iriver plus like it said and I initialized it and supposedly reinstalled the firmware but I'm still getting the error.
00:34:05curtmackYou installed the right version, right?
00:34:09 Quit Mathiasdm ("Yuuw!")
00:34:17curtmacki.e. the 20G version for the 20G model, the 5/6G version for the 5/6G model
00:35:13Nate263Yeah, I did. And it doesn't seem like even if I did it would matter since when I initialized it it deleted all of those files and reinstalled the original firmware.
00:35:38curtmackare you using mIRC?
00:35:44Nate263No, XChat
00:35:59curtmackdoes XChat support DCC file sending?
00:36:06Nate263I think so
00:36:17curtmacksend me a snapshot of your /System folder
00:36:23*scorche|sh wonders what needs DCC-ing..
00:36:51curtmackbetter still
00:37:06 Join tvelocity [0] (
00:37:16curtmackls /media/H10/System > ~/out.txt
00:37:24curtmackthen put out.txt onto pastebin :P
00:37:49curtmackI'm stupid
00:38:06Nate263Okay, give me a sec
00:38:14*rexbron notes that only works for *nix os's
00:38:20scorche|sh....assuming he is using *nix..
00:38:24scorche|shbah...too slow
00:39:03 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
00:39:05Nate263I don't have an out.txt
00:39:21rexbronNate263: what os are you running
00:39:33Nate263Windows XP
00:39:38curtmackokay then
00:39:50Nate263Sorry, I hate myself too.
00:40:00*rexbron is rusty on his MS-DOS batch commands
00:40:08curtmackdir N:\System\ > C:\out.txt
00:40:17curtmackreplace N with whatever drive letter your H10 is on
00:40:46rexbronNate263: ^
00:41:05Nate263Okay, 1 sec
00:41:55 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
00:42:56Nate263Oh crap
00:43:47Nate263It kept saying parameter incorrect or w/e and I just checked my iriver. It is completely empty.
00:43:59 Quit herrwaldo ("Konversation terminated!")
00:43:59Nate263No system file, nothing
00:44:26curtmackis it a 20 gig model?
00:44:34Nate2635 gig
00:44:42curtmackif it were a 20 gig I could send you my system files :P
00:45:23 Join magmaniac [0] (
00:45:29 Part magmaniac ("Leaving")
00:48:37Nate263I just found a recovery tool for the 5 gig, perhaps this could save me...
00:48:45curtmackgood idea
00:49:40curtmackwhile it's blank you might also reformat the drive, just run format N: /FS:FAT32 (or whatever drive letter)
00:50:16curtmackprobably isn't the issue but it can't hurt if it's already blank
00:52:56 Quit Genre9mp3 ()
00:58:18 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
00:58:54 Quit austriancoder (Remote closed the connection)
01:00:14Nate263It seems to be working!
01:00:55 Nick fxb is now known as fxb__ (
01:01:09Nate263Dang it
01:05:45 Quit leox (Read error: 104 (Connection reset by peer))
01:06:22 Join leox [0] (
01:08:39 Quit Nate263 ("Leaving")
01:11:38 Quit tvelocity (Remote closed the connection)
01:18:26 Quit dan_a (Read error: 110 (Connection timed out))
01:36:24 Quit robin0800 (" HydraIRC -> <- Like it? Visit #hydrairc on EFNet")
01:37:08 Quit Nico_P (Remote closed the connection)
01:37:34 Quit curtmack (Read error: 110 (Connection timed out))
01:38:05 Part toffe82
01:40:46*pixelma spots yet another audio commit by austriancoder...
01:40:54saratogaanyone around who can test a patch on the X5 or M5?
01:41:26***Saving seen data "./dancer.seen"
01:41:53pixelmawasn't there still something fishy with the first changes? And didn't he himself agree on asking for testers in the mailing list before committing something new there?
01:42:45 Quit Rick ("I don't need to be here.")
01:42:46pixelmasaratoga: I would if it wasn't so late... what does the patch do?
01:43:18saratogapixelma: speeds up WMA playback
01:43:29JdGordontumbleweeds in the recording forum :/
01:44:45pixelmaJdGordon: I'll try out the trigger screen patch tomorrow (well after sleep) here if you remind me
01:45:20JdGordonis ther user ml email rockbox @
01:45:34pixelmaI haven't used it much before but at least I tried out a bit before
01:47:03pixelmasaratoga: sounds nice but I should really get some sleep (getting 2 a.m. here)
01:49:50saratogasomething has to be wrong with MAD on ARM
01:49:56saratogatheres no way it should be so slow
01:50:20saratogai just noticed its barely faster then faad, even though half of it is written in ASM, and inspite of FAAD being crap
01:51:21 Join webguest99 [0] (n=413cc33c@gateway/web/cgi-irc/
01:51:26 Part pixelma
01:52:29webguest99I have a quick question with the installation of rockbox on a 30GB iPod Video, does rockboy (or the rockbox version of gameboy) come already installed when you install rockbox?
01:52:55preglowsaratoga: probably because of the slow mac unit
01:53:08preglowsaratoga: mp3 decoding loves using those, and they're shit slow
01:53:31preglowsaratoga: did you manage to profile anything, btw?
01:53:38saratogapreglow: yes and no
01:53:49saratogai can get the profile data, but haven't gotten the parser to work
01:53:56preglowah, damn, taht's right
01:53:58preglowi forgot about the parse
01:54:05preglowwell, making it work would be hugely appreciated
01:54:08saratogathe map file addresses dont' seem to match the ones in the profile
01:54:14saratogaand i have no idea why
01:54:26saratogathough its generally not too hard to figure out whats what just by digging at it
01:54:41saratogaalso, I realize arm has a slow multiplier, but this still makes no sense
01:54:50preglowwell, it makes it make more sense
01:54:54saratogaMP3 should need fewer multiplies then AAC
01:54:59preglowhell no
01:55:05preglowaac has no subband processing
01:55:13preglowand that gobbles up macs
01:55:27saratogayeah but it has slower IMDCTs instead
01:55:28preglowcheck the entire synth.c thing
01:55:51saratogathe point of a hybrid subband-transform is that it should be faster at the expense of quality
01:55:56saratogajust look at MPC
01:56:02preglowbut that still uses fewer mac operations than subband filtering
01:56:12preglownice point, sir
01:56:12saratogaMPC is pure subband
01:56:28saratogayeah this whole thing makes no sense to me
01:56:37saratogamaybe mad just sucks
01:56:38preglowcache problems?
01:56:45preglowmad doesn't really suck too much
01:56:48preglowis my impressions
01:56:54saratogamine too
01:57:34saratogai will try to profile it and see what the hell its doing
01:57:48saratogain the mean time if anyone wants to try my wma patch on the X5 . . .
01:58:15jhMikeSsaratoga: what's it do?
01:58:19preglowspeed up wma
01:58:48jhMikeSok, I'll test. comparison of benchmarks?
01:59:03saratogano just tell me if it works
01:59:08saratogaone sec, i'll upload it
01:59:15n1ssaratoga: yes, making the profile parser work would be nice indeed :)
02:00:42 Join Rick [0] (
02:01:47saratogajhMikeS: here you go:
02:02:17saratogai spent a good while on the profiling problem, but unfortunately I don't really understand gcc well enough to figure out why the symbol addresses didn't parse
02:02:42saratogathe perl code itself seems fine, it just doesn't get the info it needs from whatever gcc util it runs (objdump i think?)
02:04:18jhMikeSsaratoga: building...
02:05:01preglowthis particular command _is meant to have a bad crc_ :////
02:05:15preglowwhy can't i even make a sofisticated mistake?
02:05:17jhMikeSpreglow: isn't the early stage of a port just a joy :)
02:05:22preglowalways these retarded bugs
02:05:44preglowwell, ok, now i only need to make the bloody card report it's powered up
02:06:02jhMikeSso if the CRC matches, something is wrong?
02:06:22jhMikeSwith an "f"?
02:06:45*jhMikeS gets "plugin IRAM is full"
02:06:58saratogajhMikeS: ok thanks
02:07:08 Quit joshn_ (Remote closed the connection)
02:10:51 Join AceNik [0] (n=AceNik@
02:11:00webguest99Does anyone here know how to install Pacboy on a 30GB iPod video?
02:11:37preglowten seconds, and this damned bastard won't say it's powered up
02:11:44 Join simonrvn_ [0] (i=simon@unaffiliated/simonrvn)
02:11:45preglowtime to discover another retarded mistake, then
02:11:55 Quit simonrvn (Nick collision from services.)
02:11:55 Nick simonrvn_ is now known as simonrvn (i=simon@unaffiliated/simonrvn)
02:12:03 Join Ezra [0] (
02:12:26AceNikguys this svn revision r17478, has not fixed the same bug on the h10[20GB] atleast, my ears blast & ache eveytime i restart the player & resume playback
02:12:49preglowwhat bug?
02:13:11*jhMikeS has r17502 on his big H10
02:14:10jhMikeSAceNik: start screen is the WPS?
02:14:14AceNikwell when you resume playback on the h10 after restart the volume is very high & remains so, but if you try setting bass again like if its on 13 jus set it to 13 again & you will notice a volume difference with bass reapplied, same goes with the treble
02:14:21 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
02:14:51AceNikjhMikeS: no start screen is main screen, & i wait for the the player to be stable for a minute before i resume playback
02:15:00AceNikill just file a bug report hold on
02:15:26*jhMikeS is trying to reproduce
02:16:57*jhMikeS isn't having any luck
02:19:22n1sAceNik: please try with the latest svn revision, reset settings etc
02:19:23 Quit webguest99 ("CGI:IRC (EOF)")
02:19:34saratogacan someone take a look at a map file for me and tell me why theres a huge gap where IRAM gets used up but no variable is listed?
02:19:49EzraCould someone please help me?
02:20:02EzraWhere are there some instructions on how to install?
02:20:05jhMikeSsaratoga: sure
02:20:11n1ssaratoga: static functions (and maybe vars) do not show in the mpa
02:20:25EzraI have no clue what to do with the files!
02:20:30EzraHELP ME PLZ.
02:20:35jhMikeSit should show that the area is used though
02:20:37n1sEzra: in the manual
02:20:45Ezrawhere though?
02:20:50EzraI downloaded the manual
02:20:51scorche|shEzra: read it...
02:20:52saratogan1s: oh I didn't realize that
02:21:08AceNikwell im using r17518, now, but i belive the bug started at r17214: or r17217:
02:21:20 Quit OlivierBorowski (Remote closed the connection)
02:21:25Ezraread what? I cant find the read me!
02:21:28saratoga... and sure enough i have a static variable exactly that size
02:21:38*jhMikeS opins that they completely botched things re: not showing statics in the .map
02:22:08scorche|shEzra: i dont see what i am going to do in a PM that i wont do the manual
02:22:28AceNikreproduce the bug: 1.Hope You have equalizer on, bass & treble already set at a value, 2. restart the player, 3.resume playback 4.just reapply the bass value & notice the volume difference
02:22:49Ezrawhat manual? I downoad the "manual" but i cant find any readmes or anything!!
02:23:11n1sEzra: what did you download?
02:23:14 Join DaCapn [0] (
02:23:15scorche|shEzra: have you not seen a pdf before? is similar to a word document
02:23:41EzraCan you tell me where to get this pdf file?
02:23:44n1sEzra: we even have online manuals
02:23:50EzraI may of downloaded teh wrong one..
02:24:31AceNikn1s: i did use the latest svn & i did reset also, i tried the normal builds also, r17478 claims it fixed a bug on a particular target, not the h10 also
02:25:28scorche|shEzra: caps arent needed, and there is a link on the side of the site that says "manual"
02:26:00jhMikeSAceNik: Indeed...and it scales with the Bass level
02:26:06EzraThere are two ways of installing Rockbox: automated and manual.
02:26:17Ezrahow do I get the program that does it automaticaly?
02:26:25AceNikjhMikeS: i told you so, so confirms ill file a report
02:26:29scorche|shdid you read the "automated" section?
02:26:37Ezrano...I didn't see one.
02:26:48n1sAceNik: can you duild yourself?
02:26:53scorche|shseriously...please just read it
02:27:01AceNikn1s:yes i can build
02:27:11*jhMikeS guesses other targets with SW tone control would be affected too
02:27:17AceNikn1s: you got a ptach already ? or any suggestions
02:27:22EzraI have the SanDisk, does that matter?
02:27:46n1sAceNik: could you try the same fix as austriancoder did, simply add #define AUDIOHW_CAPS (BASS_CAP | TREBLE_CAP) in wm8731.h (in firmware/export)
02:27:51 Quit linuxstb (Read error: 110 (Connection timed out))
02:27:51AceNikjhmikeS: well yeah they will, jus someone's not noticed yet cause everyone would be happy with the treble they are listening to
02:28:03scorche|shEzra: just as long as you clicked on the manual that was for your device...there were pictures, i believe
02:28:17AceNikn1s: hold on ill try
02:31:39Ezra"your configuration is invalid. Please go to the configuration dialog and make sure the selected values are correc.
02:31:43Ezrahow do I do that???
02:31:51n1sAceNik: hmm, since the wm 8731 has no bass/treble (at least it doesn't in rockbox) it probably will not work...
02:32:23AceNikn1s: what do i do then ?
02:32:37n1sfile a bugreport :P
02:32:52EzraNo! please help me.
02:33:08n1sand whenever austriancoder drives by next hope someone tells him about it...
02:33:38AceNikwell ill log\on tomorrow, its 6am now
02:35:49*preglow kicks his sd card
02:36:19AceNikn1s,jhMikeS: what do i summarize the bug as ?
02:36:56 Join DarkJesus [0] (
02:36:58 Part DarkJesus
02:37:18n1sBass and Treble settings are not correct after reboot (or something) (well, if i understood your issue correctly)
02:40:56*jhMikeS is aware of two playback bugs 1) rapid skipping to the last song can cause metadata not to be displayed 2) skipping up to a track that requires buffering can cause the last buffered track to repeat
02:41:40jhMikeSactually, seems very related to one another
02:42:43EzraI have e250, so mine wont work?
02:42:55Ezracan I downgrade?
02:42:58jhMikeSheh, once you trigger it, the last song in the album (which can also be the first if only one song exists in a directory) won't ever display metadata anymore
02:43:00 Quit daskhweg (Read error: 110 (Connection timed out))
02:43:17scorche|shEzra: what firmware version do you have?
02:43:23 Join daskhweg [0] (
02:44:20Ezrais that what you mean?
02:45:19scorche|shno...go into the original firmware (the one you usually use), go into the settings and you should be able to find the firmware version under...system i think?
02:45:31AceNik FS #8992 - Bass & Treble Not Rendered By Player Every Restart
02:45:57AceNikn1s:, jhMikeS: thank you see you
02:46:25*n1s should sleep too :)
02:46:43 Part AceNik
02:51:26*n1s thinks he spotted a bug in the 17274 rev of sound.s, in sound_set_bass we now call dsp_callback(DSP_CALLBACK_SET_BASS, current_bass); before assigning a value to current bass...
02:52:29Ezraok, i will chech it now..
02:52:48n1sso current_bass will always be 0 on the first call
02:54:00scorche|shEzra: then you can install rockbox just fine
02:54:18n1sand the same goes for the treble function...
02:54:29*n1s comments on the tracker
02:55:28EzraNo, it wont detect it in the automated one!
02:56:24 Join kkurbjun [0] (
02:56:29scorche|shmake sure it is plugged in, make sure you get a drive letter, etc
02:56:37kkurbjunjhMikeS: Nice work on the LCD sleep cleanup
02:56:46kkurbjunit looks way nicer than I had it
03:00:06Ezraplease help me.
03:00:23scorche|sh< scorche|sh> make sure it is plugged in, make sure you get a drive letter, etc
03:02:10EzraI get it when I go into My Computer
03:02:16Ezrabut no in the application
03:02:25LloreanEzra: What drive letter is it?
03:02:44EzraIt just says:
03:02:46EzraSansa e250
03:03:22Ezrashould i change it?
03:03:40LloreanThat means it's not in MSC mode.
03:03:42 Quit jhMikeS (Nick collision from services.)
03:03:48 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
03:04:07EzraI have to put it in MSC mode?
03:04:13scorche|sh...which those instructions in the manual tell you to make sure it it is in
03:04:28Ezralet me read them again lol
03:06:38 Join toffe82 [0] (
03:07:02saratogahmm I can't even build with profiling enabled now
03:09:00 Join JdGordon|uni [0] (i=768aa35e@gateway/web/ajax/
03:10:22n1shmm, ther's a (what seems to me old leftover sleep(HZ/2) in the init in main.c candidate for removal?
03:10:51n1sline 301
03:11:25n1sit was put in in rev 900 something directly following a call to show_logo
03:12:12jhMikeSonly for the sim
03:12:24JdGordon|uniyep, prob removable
03:12:49n1sah, crap i thought it was for target, /me goes to sleep now...
03:13:05JdGordon|unistill... no reason to slow down the sim loading...
03:13:06jhMikeSrev 900? are you an archaeologist?
03:13:35*n1s checked svn blame :)
03:13:43saratogahmm amiconn seems to have broken profiling
03:14:04saratogajhMikeS: any idea what the value of the COP argument should be in this commit?
03:15:10jhMikeSsaratoga: whatever core you want the timer to run on
03:15:17saratogais 0 a good answer to that?
03:15:41saratogadoes it make any difference?
03:16:11jhMikeSyes, the timer interrupt runs on the specified core. it's run on COP for grey targets
03:16:25jhMikeSfor greylib of course
03:16:49 Join loan [0] (
03:16:55jhMikeSCPU is the safer if it's for some more general purpose
03:16:56 Nick loan is now known as loan93 (
03:17:16loan93i need some help
03:17:28EzraOK, Now this is serious!!!
03:17:38EzraWhen i start my mp3 I get an error!
03:18:25LloreanEzra: An "mp3" is a file. do you mean "when I play an mp3" or "when I start my mp3 player"
03:18:32loan93i have an ipod mini 2g & i can't disconnect it safety
03:18:34LloreanAlso, nobody can help you without knowing what error you're talking about.
03:18:49Ezrathe second one!
03:18:55Ezrawhen i start my mp3 player!
03:19:00loan93help me
03:19:05EzraIt works when its plugged in
03:19:13Lloreanloan93: Seriously, please read the channel guidelines.
03:19:19Ezrabut when I take it out, I get an error
03:19:30LloreanEzra: As I said, you need to tell us what the error is...
03:19:43 Join Makuseru [0] (
03:20:18Ezracan't load rockbox.mi4:fie not found
03:20:41loan93can you help me?
03:20:46 Quit kugel ("ChatZilla [Firefox 3.0b5/2008050509]")
03:20:47Ezracan't load rockbox.e200:file not found
03:20:54LloreanEzra: Then you didn't finish the install.
03:21:04Ezrawhat do you seugegst I do?
03:21:15LloreanEzra: You've only installed the bootloader, not the firmware. Go back to the manual, again, and read it more carefully.
03:22:08Lloreanloan93: You've got a problem with your operating system. It's not really something that belongs in this channel, since your OS is what refuses to disconnect safely, and Rockbox doesn't have a USB mode on the ipod yet.
03:22:37loan93i just installed rockbox
03:23:10LloreanAnd it doesn't matter that you're running Rockbox. USB is handled entirely by Apple software right now.
03:23:42*preglow starts suspecting the sd card isn't powered correctly
03:23:54saratogajhMikeS: in profile.c will it work to run the timer on the CPU for all targets? or is a #ifdef needed to figure out where to run it?
03:24:30saratogai'd like to fix profiling in SVN if I can understand what that COP parameter actually does to it
03:24:53*jhMikeS wonder's if the gigabeat FX is going to start charging the battery while running one all this power saving is done :p
03:24:59loan93i know but it says that it can't be disconnected
03:25:22davenloan93: that's not rockbox's problem.
03:25:28loan93and i don't have any app reading my iPod
03:25:34loan93ok thanks
03:25:34davenloan93: OS?
03:25:41loan93windows xp
03:26:07jhMikeSsaratoga: You should be able to force greylib to run on CPU for debugging or make profile.c run dual-core
03:26:11davenright, try ``remove device safely''.
03:26:23Lloreandaven: Really, this is off-topic here.
03:27:02LloreanHe can always just shutdown his computer, and unplug it while shutdown if he doesn't want to simply unplug it unsafely.
03:27:05loan93can't stop the device in this moment
03:29:01jhMikeSsaratoga: I'd just use IF_COP(, CPU) in profile.c
03:29:02loan93thanks anyway
03:29:10 Quit loan93 ("ChatZilla 0.9.81 [Firefox]")
03:29:58preglowjhMikeS: do you know if sd cards need full power on their power pins to work in init mode?
03:31:30jhMikeSsaratoga: the timer interrupt simply runs on the core specified...nothing more or less than that. CURRENT_CORE from within the handler will return the core specified in that call.
03:32:41jhMikeSpreglow: I really don't know that. Most of what I did didn't involve caring about how to handle the SD itself.
03:32:58preglowthat was the level i was hoping to remain on...
03:33:19preglowbut the fact that the ocd power bit never gets set does seem to imply it has troubles in that area
03:33:19 Quit XavierGr (Nick collision from services.)
03:33:29 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
03:33:35jhMikeSpreglow: use the sansa driver as a guide?
03:33:44preglowthat doesn't seem to touch power at all
03:34:09jhMikeSit has to power the cards and wait I'm pretty sure.
03:34:28saratogaany ideas why I'd get an undefined reference to profile_thread when I try to profile? I've configured a profile build
03:34:29preglowit probably does the powering stuff through the uncommented gpio manipulations
03:35:08preglowhrm, i wonder if i can fit some leads in through this sd slot...
03:36:18jhMikeSsaratoga: hmmm...I could have broken that along the way but I did see the functions in profile.c
03:36:43jhMikeSpreglow: that's just a mux to pick the card, it never does powering more than once
03:37:15jhMikeSor do you mean ata_init?
03:38:00preglowwas looking at sd_init_device, but yeah, ata_init also does some
03:38:25jhMikeSsd_init_device is only called for the initial plugin
03:38:33preglowstill, the powering won't really be applicable to d2, i don't even know how the power lines are connected to the sd slot
03:38:54 Quit Ezra ()
03:39:13jhMikeS"pro" time? :)
03:39:17preglowi suppose it would make sense to turn the power supply off/low when no card is detected, and right now i don't touch any power stuff
03:39:39preglowmight just be pro time, yes :/
03:40:13preglowperhaps i should just mess around a bit with the pcf
03:40:49preglowhopefully i'll fry the shit out of this thing
03:41:17jhMikeSall voltage to max and hardware charge regulation off! :)
03:41:24preglowonly way to debug
03:41:29***Saving seen data "./dancer.seen"
03:42:02preglowbah, sleep first
03:42:20 Join loan93 [0] (
03:42:56jhMikeSsaratoga: profile_thread is defined on thread.c line 1840
03:43:38jhMikeSmaybe the #ifdef RB_PROFILE is wrong?
03:45:03 Quit n1s ()
03:48:09saratogajhMikeS: I thought so, but commenting out that ifdef in thread.c and thread.h doesn't change anything
03:48:22saratogaeven manually including thread.h in mpa.c didn't help
03:48:55jhMikeSwhat uses profile_thread?
03:48:56 Quit loan93 (Read error: 104 (Connection reset by peer))
03:49:15 Join loan93 [0] (
03:49:33saratogajhMikeS: in SVN, nothing, but if you want to profile something you put those two functions in and it gives you profile data about what happened in betweeen them
03:49:38jhMikeSIf it's from mpa.c, then it's needs adding to the codec api too
03:49:50saratogaor at least it did as of this winter
03:49:54saratoganot sure when it broke
03:50:20jhMikeSmaybe it was yanked from the codec API somewhere during a cleanup?
03:51:04loan93its normal rockbox to take very much time to build the database in an ipod mini 2g?
03:52:09jhMikeSit will take longer if music is playing and/or if dircache isn't enabled
03:52:13saratogajhMikeS: theres been no changes to codeclib.h or c since I last profiled
03:52:36 Quit Makuseru (Read error: 104 (Connection reset by peer))
03:53:25jhMikeS...and profile_thread is in struct codec_api too :\
03:53:48saratogamaybe i broke something playing around with this
03:59:45saratogai think i'm going to give up for tonight
04:01:16 Quit XavierGr ()
04:02:06jhMikeSdid a 'make clean' after enabling it/
04:03:49saratogajhMikeS: yes and I reran configure to make sure
04:04:17saratogaalso, if you're interested I've still got the profiling patch someone posted on IRC last january that used to fix profile builds
04:06:09jhMikeSlink up...I can't see any obvious reason it shouldn't find it
04:07:22 Quit loan93 ("I'm not here right now.")
04:07:43saratogaplus I have to disable building doom and alac since they don't share IRAM well
04:07:52saratogabut thats easy enough
04:07:58saratogathis linking problem though has me stumped
04:08:21jhMikeSwait, this is a sim build?
04:08:32 Join loan93 [0] (
04:08:39saratogajhMIkeS: no I'm building for the Sansa
04:09:09 Quit loan93 (Client Quit)
04:09:22jhMikeSjust asking because there's no profile_thread in thread-sdl.c
04:12:41 Quit saratoga ("CGI:IRC")
04:16:23 Join Kornfan71 [0] (
04:19:01jhMikeSin alac.elf I get a "relocation truncated to fit..." but mpa.codec compiled ok
04:20:10Kornfan71Hi, I'm using Rockbox on a Sansa e200. I played all of my 800+ songs and turned on shuffle. It made it to about 20 songs, then stopped about 20 seconds into the next song, stopped and said "Data Abort at 00008000 (0)"
04:20:27Kornfan71What happened to make it do that?
04:22:59 Part Kornfan71
04:23:04 Join langhaarrocker [0] (
04:23:39langhaarrockerJdGordon - are you awake?
04:24:22langhaarrockerSup? What does that mean? I know soup ....
04:24:32langhaarrockerAh "What's up" I suppose
04:25:26langhaarrocker:D I just read about your trigger screen patch in the mailing list. I'd like to to test it, but currently have no build environment set up.
04:25:41JdGordon|uniyuo use the trigger screen often?
04:26:20langhaarrockerI MADE the trigger screen, because I use it regularily for band practice, twice a week.
04:26:38JdGordon|uniok, which target?
04:27:05langhaarrockerI have a IRiver h120
04:29:24JdGordon|unipm me an email address?
04:29:42JdGordon|uniI havnt got my webspace setup properly yet :p
04:30:45jhMikeSsaratoga: (for the logs) this compiles (IRAM full at the doom plugin though):">
04:31:09langhaarrocker*smile* Just try phil t caranggde
04:33:03JdGordon|unilanghaarrocker: sent
04:33:04 Quit cool_walking_ (Read error: 104 (Connection reset by peer))
04:33:16JdGordon|uniyou can email me comments directly to that address
04:33:27 Join cool_walking_ [0] (
04:33:52langhaarrockerJdGordon: received
04:33:57langhaarrockerI'll mail you :D
04:34:10langhaarrockerThank you
04:34:13*langhaarrocker bows
04:34:34 Quit Thundercloud (Remote closed the connection)
04:43:23langhaarrockerJdGordon|uni: For the h120 I'd suggest to swap the control keys: up/dn should move the cursor, left/right should change the values.
04:43:58JdGordon|uniI tried not to change the controls... arnt they the same as svn
04:44:20JdGordon|uniback in 5
04:46:06 Quit BHSPitLappy ("Ex-Chat")
04:47:38LloreanJdGordon|uni: Yeah, but the SVN controls kinda suck. :)
04:47:47*langhaarrocker is happy that JdGordon|uni is back
04:48:26 Join m0f0x [0] (
04:48:27langhaarrockerThe little marks in the peak meter that mark the trigger thresholds don't move when I adjust the thresholds.
04:49:14JdGordon|uniand its talking when it shuoldnt
04:49:30JdGordon|unidoes the threshold stufff work at all?
04:49:37JdGordon|unithats the thing I was worried about breaking
04:49:53langhaarrockerI wouldn't notice that. I'm deaf due to to many band practices ;)
04:50:45JdGordon|uniok, I have svn on my e200 and cant see which marks your talking abot
04:51:00JdGordon|unioh.. looking at the wrong bar
04:51:13JdGordon|uniwhat is the top bar supposed to be?
04:51:30langhaarrockerIt displays the state of the trigger.
04:52:36*langhaarrocker searches the manual in the wiki
04:53:16 Join miepchen^schlaf [0] (
04:53:48langhaarrockerThis describes what the top bar should do:
04:55:27*JdGordon|uni suspects the peak meter drawing thing is asking for the wrong action context
04:56:42langhaarrockerAdjusting the values for the threshold is a bit annoying since there's no repeat. You have to tick the button for every single db.
04:57:03JdGordon|uniyep, will fix that
04:58:03 Quit Evilnick (Read error: 110 (Connection timed out))
04:58:39langhaarrockerAt least in "trigger simulation" (= during playback) the trigger status display doesn't work. It always displays "Ready" state.
05:07:16 Join finnmacool [0] (n=d1ca06e9@gateway/web/cgi-irc/
05:08:06finnmacoolfolks, i have a question about my iaudio x5. anyone give me a hand with it pls?
05:08:14langhaarrockerJdGordon: Why did you arrange the setting names and values vertically? Space problems on some targets?
05:08:51 Quit markun (Read error: 104 (Connection reset by peer))
05:08:53 Quit miepchen^schlaf_ (Connection timed out)
05:08:59JdGordon|uniyes, and it now uses the standard list drawing code to make things smaller
05:09:21JdGordon|uniI have another patch which can put the values right aligned which I might put in with this
05:10:23JdGordon|unifinnmacool: only if you ask a question.... there a way i can set original volume control (i.e joystick up/down) with rockbox installed?
05:12:20JdGordon|uniwhat do you mean?
05:13:27finnmacoolwith the original iaudio firmware, the way i adjusted the volume was pushing up or down on the joystick. this no longer works using rockbox. i would like it to work that way
05:14:23JdGordon|uniI dont know the x5 controls but I would guess it shuold do that only in the while playing screen? Llorean?
05:14:49finnmacoolit 'should' do it but doesn't seem to
05:14:52Lloreanfinnmacool: What happens instead when you're in the WPS?
05:16:50finnmacoolit might just have been a glitch, seems to be working just fine. thx vm
05:17:48 Quit finnmacool ("CGI:IRC (EOF)")
05:20:14langhaarrockerJdGordon|uni: At least for the h120 target the trigger doesn't trigger at all :(
05:20:26 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
05:20:37*JdGordon|uni back to the drawing board then
05:21:43*langhaarrocker pats JdGordon for sympathy
05:22:29JdGordon|uniits alsmot certainly the threshold stuff :(
05:22:46JdGordon|unithats the only thing which was acually changed (other than the drawing code)
05:30:56 Join toffe82_ [0] (
05:38:04 Part langhaarrocker
05:38:17 Quit Horscht ("I am root. If you see me laughing, you better have a backup")
05:40:14 Quit cool_walking_ (Remote closed the connection)
05:41:30***Saving seen data "./dancer.seen"
05:45:56 Quit toffe82 (Read error: 104 (Connection reset by peer))
05:46:12JdGordon|uniLlorean: the zip in shold be removed seen as its got a microsoft exe in it yeah?
06:12:56 Join mackes [0] (
06:13:21 Quit JdGordon|uni (" ajax IRC Client")
06:14:58 Join nuonguy [0] (
06:36:22 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
06:36:38 Quit SirFunk (Read error: 104 (Connection reset by peer))
06:37:07 Join SirFunk [0] (
06:49:54 Join linuxstb__ [0] (
07:03:53 Quit linuxstb_ (Read error: 110 (Connection timed out))
07:05:08 Quit mackes ("Mackes is out")
07:06:13 Quit linuxstb__ (Read error: 104 (Connection reset by peer))
07:06:21 Join linuxstb__ [0] (
07:10:39 Quit krazykit (Remote closed the connection)
07:10:43 Join krazykit [0] (
07:22:59 Join lymeca_ [0] (i=kino@unaffiliated/lymeca)
07:23:09lymeca_Hi, I just got an 80GB iPod Classic
07:23:17lymeca_I want to know if this is supported.
07:23:38lymeca_Is this no longer considered '5.5 gen'?
07:23:42lymeca_Is this 6th gen?
07:23:43 Nick fxb__ is now known as fxb (
07:23:47 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
07:30:55 Quit m0f0x ()
07:31:22 Quit miepchen^schlaf ()
07:31:39 Join draeath [0] (i=draeath@unaffiliated/draeath)
07:32:17draeathHas anyone else tested FLAC on rockbox, at 24-bit 96000hz?
07:32:41scorcheit will just be downscaled...
07:33:44 Join Bagderr [0] (n=daniel@gateway/web/cgi-irc/
07:34:09 Nick Bagderr is now known as B4gder (n=daniel@gateway/web/cgi-irc/
07:34:11 Join homielowe [0] (
07:34:18draeathany ideas what that scales down to? (hard to believe it's getting scaled, the codec uses 15% of 6-10hz (that's a wow in and of itself)
07:34:42draeathand, somehow, i can hear a difference between that same file at that, and at 44100/16
07:35:10draeath(subtle, and i have crossfeed and stereo seperation on)
07:38:11 Join MethoS- [0] (
07:41:32***Saving seen data "./dancer.seen"
07:41:35 Join daurn [0] (i=daurn@unaffiliated/daurnimator)
07:42:15amiconndraeath: Rockbox currently resamples everything to 44100/16 on all swcodec targets (and it has a rather crappy resampler...)
07:43:12draeathI must be loosing my mind - how could it possibly sound <better> if it's using a resampler like that?
07:43:21scorcheif you hear a difference, it isnt a good are better off playing regular 44.1 files
07:43:40scorchethe mind is a powerful thing ;)
07:44:43draeathhmm.. i wonder if the hardware I have supports anything higher than that... wonder where to dig for that
07:44:59LinusNon most targets, we don't have much of a choice, since the DAC is 16-bit, and the sample clock is limited
07:45:23scorchewhich device do you have?
07:45:41draeathsansa e200r
07:46:09draeathThe dac - is that a standalone component or is it buried in a PIC or something?
07:46:33LinusNit's often inside the CODEC chip
07:46:36*draeath only halfway knows what he is speaking of
07:46:37B4gderon the sansa e200 it is buried deep within
07:46:45draeathshoot. If it was a nactual component I could play doctor
07:46:54draeather, "an actual"
07:46:59B4gderI doubt you could do that anyway
07:47:14scorcheit really isnt so simple =/
07:47:44draeathgranted it would be a complete pain, even if i didn't have to worry about clock and pinouts etc
07:48:05draeathwell... does anyone know of a player that could handle higher rates?
07:48:47amiconnLinusN: On all coldfire targets we could do 88200/24, but then we'd have to drop DMA. 88200/16 is possible with dma
07:48:56scorchethe DAC is only 18bit
07:49:03scorche(of the sansa)
07:49:21LinusNamiconn: how could we do 24 bits when the audio bus is only 20 bits?
07:50:16scorchei seem to remember a forum post about that..
07:50:19draeathwe could drop the LSBs, and it would likely sound better than 16 bits anyways
07:50:45amiconnLinusN: Well, maybe just 20 bit then.
07:51:38LinusNdraeath: that is in fact what we do
07:51:58draeathWell... in any case, I'm still impressed this thing could handle that. Even more so that it was resampling live.
07:52:25amiconnI think that it's not worth it, because one interrupt every 4 sample pairs (fifo depth iirc) versus one interrupt every several thousand samples would cause quite some cpu load
07:53:24*amiconn really wants to see a better resampler in rockbox
07:53:39scorchesadly, that student got nabbed :(
07:53:51draeathSSRC is <supposed> to be fast and good.
07:54:13draeathbut Im only an amatuer so... i don't think i could code my way into a box, let alone out of one.
07:55:03 Nick fxb is now known as fxb__ (
07:57:42draeathThanks for the heads-up on the resambling - i was wasting a lot of trouble for little reason...
08:00:00 Quit daurnimator (Read error: 110 (Connection timed out))
08:02:29 Join Buschel [0] (
08:02:35amiconnI wonder whether it would make sense to allow using software pwm for backlight *brightness* on targets where it doesn't require boosting
08:02:53 Join GodEater [0] (i=c2cbc962@gateway/web/ajax/
08:03:30amiconnThe reason being that the mini's backlight is a real battery sucker, and it's also very bright
08:03:58amiconnSo the extra cpu load might be insignificant versus the saved battery power
08:05:01BuscheljhMikeS: thanks for correcting the #ifdef-ing to CONFIG_CPU. Seems i've submitted the wrong local version of my patch :/
08:05:03amiconnI wouldn't do this on coldfire (read: H1x0) though, as software pwm requires boosting on it (for stability)
08:06:13JdGordonBuschel: you did rockbox sync on the wiki right?
08:06:44 Quit draeath ("leaving")
08:07:28Buschelpreglow/saratoga: your discussion about subband synthesis -> mpc does the same than mp1/2/3 when synthesizing. we should compare the performance of the filterbank implementation mad vs. mpc.
08:08:15*JdGordon comes to the horrible realisation that he has to modify 30 keymap files! :'(
08:08:34Buschelpreglow/saratoga: dct32 = calc_new_V, full_synth = dewindowing
08:08:51BuschelJdGordon: pardon?
08:09:26JdGordonbah, sorry.. bascule is who im after...
08:11:16amiconnBacklight pwm probably needs to be redesigned for this, as it requires the user timer. If something else grabs the user timer (with higher priority), eg. the greylib, backlight releases it and just switches to the proper state. But it doesn't re-grab the timer when it becomes available again
08:11:51 Join Rob2222 [0] (
08:11:52amiconnThat's not a problem at all if it's only used for fading, but it would be if it's used for brightness
08:14:59Buschelpreglow/saratoga: libmad's synth.c looks a bit less straight forward than mpc's synth_filter.c −− but maybe i am a bit biased :)
08:15:39amiconnBuschel: On coldfire, libmad is actually very efficient
08:17:22amiconn420% realtime using the same test track as mp3 (231kbps), versus the 610% of the mpc version (173kbps)
08:17:38amiconnAnd that already includes the transform for mp3
08:18:07Buschelwes hould really compare efficiency of the different synthesis implementations on arm/cf
08:18:09*amiconn should try that as mp2
08:18:16Buschelmp2 = good idea
08:18:25 Quit BlakeJohnson86 (Read error: 110 (Connection timed out))
08:18:42Buschelmpc should be slower than mp2 because of the huffman coding
08:18:54amiconnBtw, the mpc file sounds clearly worse than the mp3. It's muddy, even on my EUR 25 PC speakers (creative 2.1 system)
08:19:43amiconnI wonder whether this is an mpc problem, or there is a bug in the decoder
08:21:03Buschelis this a general issue or connected to special files?
08:21:44amiconnI only tried this one album as mpc so far
08:22:23amiconnI wanted to see whether I could save some space by encoding as mpc or ogg instead of mp3, while keeping the quality
08:23:11amiconnMy primary goal is transparency though, not saving space.
08:23:49 Join BlakeJohnson86 [0] (
08:23:53 Join spiorf [0] (
08:24:09Buschelthat was/is the goal of mpc
08:25:11Buschelcan you make available the first few 100KBs of such file?
08:26:03amiconnYes, and that's why I wonder why it sounds like it does. I used mppenc −−standard and oggenc -q5, comparing to lame −−preset standard
08:26:33Buschelthis should work fine −− i've encoded (most of) my files the same
08:26:33amiconnMaybe I should try mppenc −−extreme and oggenc -q6 for a better comparison
08:26:55 Join pondlife [50] (n=Steve@rockbox/developer/pondlife)
08:27:42Buschel"muddy" = high frequencies missing or too high noise floor?
08:28:04amiconnHigh frequencies sounding imprecise
08:28:57 Quit BigBambi (Read error: 113 (No route to host))
08:29:27 Quit Rob2223 (Read error: 110 (Connection timed out))
08:31:10 Part toffe82_
08:31:41 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
08:33:29JdGordondoes changing a parameter to const in the plugin api break compatability?
08:34:20 Join Mathiasdm [0] (
08:37:38JdGordonamiconn: ^ ?
08:39:47 Join markun [50] (n=markun@rockbox/developer/markun)
08:41:47 Quit Buschel ()
08:41:52pondlifeJdGordon: I believe so.
08:42:01pondlifeconst ptr != ptr
08:42:10JdGordonconflicting answers :'(
08:42:11 Join ender` [0] (
08:42:17pondlifeI bumped the min API, just in case.
08:42:28JdGordonguess I will too then
08:43:27*pondlife hopes the const police have a new recruit.
08:43:29JdGordon*groan* a dozen functions to move into the correct places
08:44:20JdGordonno such luck... just fixing it now instead of adding casts to my patch
08:44:53 Quit jhulst ("Konversation terminated!")
08:47:06 Nick linuxstb__ is now known as linuxstb (
08:52:22 Quit BigBambi (Read error: 113 (No route to host))
08:55:30 Quit MethoS- (Read error: 110 (Connection timed out))
08:55:43 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
08:55:45 Join MethoS- [0] (
08:59:00*JdGordon does his first const raid commit!
08:59:28*LinusN prepares to shout "RED! RED!"
09:00:17 Join davina [0] (
09:00:26 Join petur [50] (n=petur@rockbox/developer/petur)
09:01:21 Join n1s [0] (n=nils@rockbox/developer/n1s)
09:02:01n1sJdGordon: it seems you missed bumping the actual version in 17520...
09:02:29JdGordonbah woops
09:03:21JdGordonplugin.h was out of data and i missed that in the diff :p
09:03:45JdGordonLinusN: thanks for the vote of confidence :D
09:06:33JdGordonBagder: seems the target text isnt high enough anymore... (in the build table).. zvm 30gb - sim is cut off
09:07:25 Quit spiorf (Remote closed the connection)
09:10:50 Join pixelma [50] (i=pixelma@rockbox/staff/pixelma)
09:11:20amiconnJdGordon: yllow...
09:11:52JdGordonthats why i said arg
09:12:01n1sLinusN: do you know that gives back a blank page now (for about a week)?
09:12:12 Quit linuxstb (Read error: 110 (Connection timed out))
09:12:17LinusNyes i know
09:12:39 Join tvelocity [0] (
09:17:00B4gderzen vision m is now zvm in the build table
09:17:46B4gderthen the text fit in the image ;-)
09:20:45B4gderyellow alert
09:20:59amiconnJdGordon: The delta is somewhat strange...
09:21:13peturwhat's up with the two remaining Zen Vision targets then?
09:21:18JdGordonamiconn: no, its expected
09:21:33JdGordonquickscreen targets had some adjusting
09:21:42amiconnah, quickscreen
09:22:05amiconnB4gder: The target order seems to become more and more shuffled...
09:22:35B4gderit's still alpha order
09:23:12LloreanThe important question, then, is "which alphabet" I suppose. ;)
09:23:35B4gdercase sensitive on the full build name
09:24:13*petur would have guessed the oldest targets would have Archos in front of them...
09:24:27B4gderno they don't, and that's why they are most confusing
09:24:55B4gderand I haven't changed their names since then they go in a new column
09:25:13peturB4gder: what's with the two Zen Vision entries to the left of the ZVM ones? Are they other builds?
09:25:21*amiconn wonders whether we should fix that
09:25:41amiconnThe case sensitiveness is also a source of confusion
09:26:02JdGordonwe should just unmber the builds... 1->X
09:26:02B4gderpetur: one is the old name that's now gone
09:26:13B4gderthe other is zen vision, not zvm
09:26:19B4gdera different target
09:26:20peturah ok
09:27:00peturI haven't been following this whole boatload of new targets ;)
09:28:14B4gdera lot of action on new targets lately indeed
09:32:41*pondlife wonders if B4gder had time to mock up the previously proposed new table format...
09:33:13B4gderno... :-/
09:33:30pondlifeNo worries, but we now have 2 more columns to scroll... :)
09:33:42pondlifeI'd do it myself if I knew the first thing about HTML
09:33:56pondlifeIt's probably not a simple one to start on though.
09:34:35pondlifeI mean, and get a decent looking result
09:34:36scorchei can only tolerate 1 more column before i have to start scrolling! ;)
09:34:46JdGordonthe problem with more than one table though is its not as fast to check for non green
09:35:06pondlifeJdGordon: The plan was one table, but multiple lines for target/boot/sim
09:35:17pondlife(and delta, IIRC)
09:35:26JdGordonoh yes, rotating the table 90deg?
09:35:48pondlifeBut the main thing is that the current build would still be completely visible with no scrolling
09:35:55pondlifei.e. scrolling only required for previous builds.
09:36:35pondlife(Hence integrating the deltas would help too.)
09:38:04 Quit Mathiasdm ("Yuuw!")
09:40:17JdGordonI tihnk deltas should stay seperate.. red there isnt as bad (obivously) and would look wierd in a table of green..
09:40:57 Join dan_a [0] (n=dan-mirc@
09:41:34pondlifeGood point
09:41:35***Saving seen data "./dancer.seen"
09:42:00pondlifeIt would be nice if the columns lined up though
09:45:45 Quit MethoS- (No route to host)
09:45:50 Join MethoS- [0] (
09:50:19 Join linuxstb__ [0] (
09:50:46 Quit linuxstb_ (Read error: 104 (Connection reset by peer))
09:55:56 Join chevron [0] (n=7cbf6166@gateway/web/cgi-irc/
09:56:19chevronanyone here
09:57:05 Quit chevron (Client Quit)
09:57:54peturvery patient....
09:58:13 Join Mathiasdm [0] (
09:58:50markunpetur: 1 minute is plenty of time for us to react
10:06:24 Join nplus_ [0] (n=nplus@141.25.Globcom.Net)
10:06:40 Quit JdGordon ("Konversation terminated!")
10:08:29 Join JdGordon [0] (
10:10:49 Quit MethoS- (Read error: 110 (Connection timed out))
10:12:05 Quit JdGordon (Read error: 104 (Connection reset by peer))
10:14:36 Join JdGordon [0] (
10:14:59 Quit nplus_ (Read error: 104 (Connection reset by peer))
10:15:52 Join nplus_ [0] (n=nplus@141.25.Globcom.Net)
10:19:52 Quit anon1532 ("<Insert satire here>")
10:24:10n1sdoes the database use up the audio buffer when building on hwcodec or is this only true for the committing?
10:25:51 Quit dan_a (Read error: 110 (Connection timed out))
10:30:38 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
10:32:51 Join voltagex [0] (
10:40:31 Quit nplus_ (Read error: 104 (Connection reset by peer))
10:45:31 Join voltagex_ [0] (
10:55:17pixelman1s: seems like it doesn't. I can keep listening while the database is building
10:57:14n1spixelma: ah, thanks :)
10:57:30 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
10:58:32 Join dabujo [0] (
10:59:43 Join OlivierBorowski [0] (
11:01:35JdGordonamiconn: you couldnt get your microdrive to work outside the mini?
11:01:35 Quit voltagex (Read error: 110 (Connection timed out))
11:02:12 Quit nuonguy ("This computer has gone to sleep")
11:03:36JdGordonwierd... mine connected fine first go
11:04:42 Nick cnakihlcm is now known as Anakin (i=0@
11:05:51amiconnThe card reader lets it spin up, then stops it after a few seconds (hasty emergency park), several times before giving up. My digicam doesn't spin it up at all, even though it should support microdrives
11:06:38LinusNsounds like a current consumtion problem
11:06:59 Join DerPapst [0] (
11:07:10JdGordonI'm running mine in a powered usb enclosure, so maybe it is a power thing?
11:12:20JdGordonanyone not like the idea of using disktidy_custom.config and disktidy.config instead of it being a viewer?
11:17:16pondlifeJdGordon: I think that makes more sense.
11:18:12peturI think most agree that it behaving like a viewer is not a logical thing
11:21:14 Join desowin [0] (n=desowin@atheme/member/desowin)
11:34:17 Quit linuxstb__ (Read error: 110 (Connection timed out))
11:36:11 Join MethoS- [0] (
11:41:03 Quit DerPapst (Read error: 104 (Connection reset by peer))
11:41:36 Join DerPapst [0] (
11:41:39***Saving seen data "./dancer.seen"
11:44:02*pixelma read logs and wonders about what JdGordon and langhaarrocker said about the trigger screen controls
11:45:50JdGordonnothing much, actually nothing that isnt in the log
11:45:53pixelmain current svn they work as I'd expect them (up/down to go through the list, left/right to change the values) on all 3 targets I have, haven't tried the patch though yet but someone said that controls are like in svn (IIUC)
11:46:52JdGordonthat depends on the targets keymap.. obviously some are warped
11:47:36pixelmahmm... ok, maybe that's an H100/300 thing?
11:47:54 Quit rexbron ("My dreams used to be insulated by pink fiberglass...")
11:48:51 Nick fxb__ is now known as fxb (
11:49:08pixelmaJdGordon: and you also said something about talking which I don't understand... I thought that the recording screens currently can't use voice, or were you referring to something else?
11:49:29JdGordonthat trigger screen talked!
11:50:22pixelmaprobably can't record then? ...
11:50:43JdGordondunno.. I tinhk it can talk if its not recording
11:50:52JdGordonwhich is wasnt when i was testing
11:51:07 Quit tvelocity ("Αποχώρησε")
11:52:51markunjhMikeS: the meizu video player supports pre-rotated 240x320 videos which gives a longer runtime (compared to 320x240). Would that make sense for us as well?
11:54:23JdGordonnew disktidy patch online @ FS #8637
12:00:20 Join webguest09 [0] (n=58500503@gateway/web/cgi-irc/
12:00:29 Quit webguest09 (Client Quit)
12:03:44 Quit MethoS- (Read error: 110 (Connection timed out))
12:09:15peturtalking (even while in pre-rec) used to work on h300 so the hardware can do it
12:14:24 Join linuxstb__ [0] (
12:16:06 Join gregzx [0] (
12:24:55 Quit Shaid (Read error: 110 (Connection timed out))
12:27:43 Join staen [0] (
12:28:35staenHi 2 every 1!
12:30:01peturstaen: hi. And please use proper English in here
12:30:35staenSorry for that.
12:32:41 Quit homielowe ()
12:33:11staenI already posted in the forums that I have a problem with starting rockbox on a 5.5g Ipod. The bootloader and rockbox (when started with loader2) isn't able to recognize the file system (as it occurs to me)
12:33:16 Quit DerPapst (Read error: 113 (No route to host))
12:34:45 Join crash91 [0] (n=crash@
12:34:53staenDoes anyone know at what time it is the best to look around here?
12:35:14pondlifestaen: Have you tried without loader2?
12:35:27pondlifei.e. just the latest Rockbox bootloader
12:35:44pondlife(or is that what you meant?)
12:36:13staenYes. I first tried with the rockbox bootloader. As it doesn't work, I tried loader2
12:36:38crash91is " export PATH=$PATH:/include /usr/local/arm-elf/bin" the correct command?
12:36:41pondlifeOK. Did you also try an iTunes restore?
12:37:06*pondlife heads to the forum
12:37:53staenAs you are in the forums:
12:38:30pondlifeYep - did you try "Restore to factory settings" in iTunes?
12:38:56staenYes, I did try it. More than once.
12:39:02pondlifeOK, just checking
12:39:33pondlifeI don't actually have an iPod myself, but I'm pretty sure there are people running Apple's 1.3 firmware with Rockbox ok.
12:39:58pondlifeIs there anything else you might have changed?
12:40:08staenI tried it with 1.2.1 (copied from my wifes ipod).
12:40:16pondlifeYes, I saw
12:41:01pondlifeBTW the "IPOD version: 0xFFFFFFFF" is a known issue (fixed in SVN, but not in the released bootloader). It shouldn't affect anything apart from the info display.
12:41:24pondlife7130MB is abnormal though, I believe
12:41:29 Join JdGordon|uni [0] (i=d31c5d08@gateway/web/ajax/
12:41:36staenI first wanted to install ipodlinux and repartitioned the ipod (which I thought was a good idea as long as I had no Music and files on it). But I set it back to factory settings already
12:41:55*pondlife better stand aside and see if an iPod owner can help...
12:42:10 Quit crash91 (Remote closed the connection)
12:42:57staenthe weird thing is, my wife has exactly the same harddrive and it works. for some reason my harddrive is misinterpreted
12:43:35*staen will wait until someone comes around.
12:47:52gevaerts7130MB does sound like a 512/2048 byte problem
12:53:13 Join cool_walking_ [0] (
12:53:44cool_walking_ gives me "Notice: Undefined index: 0 in /usr/share/flyspray/htdocs/includes/class.flyspray.php on line 344" up the top.
12:54:05 Join crash91 [0] (n=crash@
12:54:23crash91i got this error from rockboxdev: checking whether the C compiler (gcc ) works... no configure: error: installation or configuration problem: C compiler cannot create executables
12:54:54gevaertsDo you have gcc installed ?
12:55:14crash91apt-get says its the newest version..
12:55:41gevaertsTo make sure, try installing build-essentials
12:56:23crash91cant find package
12:56:37crash91here is the whole error log btw,
12:56:53 Join Thundercloud [0] (
12:57:08gevaertssorry, build-essential
12:57:11staengevaerts: What keeps me wondering is that in works on the one ipod but not on the other (the hardware seems to be the same). But when the partition isn't recognized correctly then it is unlikely that the filesystem on it can be interpreted.
12:57:29crash91gevaerts: ah, i guess that was the problem
12:57:36crash91gevaerts: thanks :)
12:58:25staenI'd like to have a look into the code, but I need some advice were to look when it comes to reading from the drive
12:58:29 Quit Siku (Read error: 145 (Connection timed out))
13:00:13 Join |desowin| [0] (
13:01:47 Quit desowin (Read error: 113 (No route to host))
13:02:16 Nick |desowin| is now known as desowin (
13:02:53gevaertsI suspect that the code which prints out the 'Partition 1: 0x0B 7130MB' is wrong. It seems to assume 512-byte sectors. That's not the real issue though, as this is in print-out-only code
13:04:38pixelmastaen: are you sure you have a 5.5g (where the drive has 2048 byte sectors) and not a 5g with 512 byte sectors? Though I don't know enough about Ipods to be of much help...
13:05:19staenI suspected that the assumption of 512-byte sectors is causing a problem in adressing the driver while reading from the drive
13:06:21staenpixelma: Yes, it has 2048-byte sectors (as shown by fdisk and ipodpatcher) and even has the correct serial number according to apple
13:08:21gevaertsProbably a silly question, but is the ipodpatcher you are using relatively recent ?
13:08:32staengevaerts: but isn't the same code which gets the drive information for the printout responsible for getting the drive information when it comes to adressing something on the drive (just a guess)
13:09:11gevaertsNo. This code divides by 2048 to get megabytes _in_ the printf statement.
13:09:17*gevaerts has to go now
13:09:33staenipodpatcher v2.0
13:12:42crash91hmm how long does the cross compiler take to build?
13:13:47crash91I know it depends from computer to computer but generally 15mins? 30mins?
13:13:50GodEaterdepends on the speed of your PC, but generally it's not quick
13:14:46crash91Well i guess that means lunch
13:15:13pondlifeDefinitely don't wait for it or you may starve
13:16:10crash91ooh came back to check , its done
13:16:18crash91that was about 15mins
13:17:56pondlifeTook quite a bit longer for me
13:18:34crash91>< how do i export path again? Something like this right? export PATH:/usr/local/arm-elf/bin:$PATH
13:19:28markunfirst : should be = I think
13:19:49GodEatermarkun: yep
13:21:19markunwell, do you have arm-elf-gcc in /usr/local/arm-elf/bin ?
13:21:41*GodEater is amazed to discover that pastebin is blocked by the work proxy
13:21:59pondlifeGodEater: and .ca ?
13:22:06crash91markun: yes it is
13:22:14GodEateronly tried .com so far appears ok
13:22:37markuncrash91: then it's probably not in your path.
13:22:46GodEaterecho $PATH
13:23:34markunwhat happens if you now type "arm-elf-gcc"?
13:23:55crash91its says "no input files"
13:24:03markunthen try to compile again
13:24:03GodEaterthen it's definitely in the path
13:24:23crash91oh wait that was in the arm-elf-gcc dir
13:24:31crash91command not found otherwise
13:24:36crash91should i gedit .bashrc?
13:25:25GodEaterit shouldn't work in the arm-elf/bin directory if it doesn't work anywhere else
13:25:32GodEateryou don't have "." in your PATH (and nor should you)
13:25:38markunputting it in your bashrc is a good idea anyway
13:25:43markunGodEater: yes, pretty scary
13:25:59crash91crash@crash-laptop:/usr/local/arm-elf/bin$ arm-elf-gcc arm-elf-gcc: no input files
13:26:15crash91crash@crash-laptop:~/rockbox/build$ arm-elf-gcc bash: arm-elf-gcc: command not found
13:26:21markunand you try all this in the same window?
13:26:29markunsame terminal I mean
13:26:41crash91terminal with two different tabs
13:26:49markunthen it doesn't work of course
13:26:49GodEaterthat won't work
13:27:00markunif you set PATH in one it doesn't change in the other
13:27:02GodEaterthe export command only affects the tab you typed it in
13:27:10crash91ah >_<
13:27:18GodEaterwhich is why it should be in your .bashrc
13:27:23crash91thanks, its compiling now
13:28:55 Quit staen (Read error: 104 (Connection reset by peer))
13:28:58crash91yay, i think its finally using my dual core properly now : real1m17.214s user1m47.551s
13:29:24B4gderjust do make -j
13:29:39crash91B4gder: thats what i did
13:31:19 Join staen [0] (
13:33:09crash91thanks guys :) Now for lunch....finally
13:33:12 Quit crash91 ("Ex-Chat")
13:34:13 Join MethoS- [0] (
13:38:33pixelmastaen: you said your wife has exactly the same Ipod (30GB with 2048 byte sector disk) and Rockbox runs there?
13:40:01staenyepp. same model-nr, same harddrive, same sn-suffix ... the only difference is the apple firmware, but I don't know how this should affect rockbox
13:40:54staenI even tried to dd the whole drive of her ipod to mine, but even that didn't help
13:41:44***Saving seen data "./dancer.seen"
13:44:22*staen is getting the source code and tries to find some hints
13:44:25pondlifestaen: Have you tried a physical disk scan? CHKDSK /R or whatever...
13:44:34*GodEater wonders if it's worth downgrading the firmware
13:44:37pixelmaI don't know if this leads to anything but you could compare the output of the bootloader (reported disk size). If everything works well it is silent but you can force it to show you the details by holding the >> button (I think).
13:44:49 Quit JdGordon|uni (" ajax IRC Client")
13:46:32staenpixelma: on the working ipod it reports 28xxx MB as partition size
13:46:59 Quit MethoS- (Read error: 104 (Connection reset by peer))
13:54:29 Quit dabujo ("( :: NoNameScript 4.2 :: )")
14:00:55 Join crash91 [0] (n=crash@
14:03:17 Quit advcomp2019 ("Hi, I'm a quit message virus. Please replace your old line with this line and help me take over the world of IRC.")
14:07:28 Join advcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
14:08:05 Join webguest71 [0] (n=408e5916@gateway/web/cgi-irc/
14:08:52 Join tvelocity [0] (
14:13:25 Quit GodEater (" ajax IRC Client")
14:13:31 Quit webguest71 ("CGI:IRC (Ping timeout)")
14:17:20 Quit Zarggg (Read error: 104 (Connection reset by peer))
14:17:37 Join Zarggg [0] (
14:18:47 Join webguest71 [0] (n=57766566@gateway/web/cgi-irc/
14:20:50 Quit webguest71 (Client Quit)
14:21:51 Join Shaid [0] (
14:22:24markunis there anything legal on rapidshare?
14:23:29B4gderonly when people upload the wrong files by mistake
14:29:11 Join DerDome [0] (
14:34:18 Join DerDom1 [0] (
14:34:18 Quit DerDome (Nick collision from services.)
14:34:30 Nick DerDom1 is now known as DerDome (
14:37:36 Join DerDom1 [0] (
14:37:36 Quit DerDome (Nick collision from services.)
14:37:48 Nick DerDom1 is now known as DerDome (
14:45:42 Part B4gder
14:50:35 Quit crash91 ("Ex-Chat")
14:55:40 Join ctaylorr [0] (
14:56:25 Join LambdaCalculus37 [0] (n=44a0430f@gateway/web/cgi-irc/
14:57:07 Quit DerDome (Nick collision from services.)
14:57:08 Join DerDom1 [0] (
14:57:20 Nick DerDom1 is now known as DerDome (
15:02:15 Join crash91 [0] (n=crash@
15:03:19 Quit cool_walking_ (Read error: 110 (Connection timed out))
15:05:42*LambdaCalculus37 kicks off the return of his Gigabeast and audio playback with "Gianna Sisters" by Machinae Supremacy
15:07:05LinusNgood choice
15:07:43LambdaCalculus37LinusN: :)
15:07:59LambdaCalculus37LinusN: I thank you guys for introducing me to them.
15:08:33*LambdaCalculus37 pokes in the debug info on the beast
15:09:09LambdaCalculus37Heh... high quality FLACs and no CPU boosting!
15:09:51LambdaCalculus37Even when the buffer is filling, it's not boosting.
15:10:00LambdaCalculus37Now *this* is a target. :)
15:13:55preglowi don't think high quality flacs boost even on ipods...
15:14:16preglowunless you're talking high sample rate flacs, in which case you should stop listening to them with rockbox at once
15:14:27LambdaCalculus37preglow: Actually, IIRC they do. But I'll check again.
15:14:44pondlife96k, all the way, one day
15:14:54 Part LinusN
15:14:57LambdaCalculus37preglow: Does a ~1145kb/s FLAC count? :)
15:15:01*LambdaCalculus37 is playing one now on his beast
15:15:15preglowand still 16bit/44100?
15:15:25pondlifeDo we resample for SPDIF out on H100?
15:15:32preglowwe always resample
15:15:34LambdaCalculus37preglow: Yep.
15:15:46preglowLambdaCalculus37: flac didn't much like that one, then
15:16:31LambdaCalculus37preglow: It was from Machinae Supremacy's site.
15:29:32 Join GodEater [0] (i=c2cbc962@gateway/web/ajax/
15:32:11 Nick linuxstb__ is now known as linuxstb (
15:32:26 Quit tvelocity (Remote closed the connection)
15:33:42preglowlinuxstb: you know anything about how the cpu is powered in telechips setups?
15:35:02preglowi can't find a cpu voltage from the pcf
15:35:14preglowit's 1.2 volts, i think
15:35:26preglowi can find something that can be used as io voltage, but that's that
15:36:07 Join hopenglish [0] (n=4b204236@gateway/web/cgi-irc/
15:39:04preglowi'm trying to find out if the sd card voltage is provided from the pcf...
15:41:03 Join DerPapst [0] (
15:41:23 Quit crash91 ("Ex-Chat")
15:41:46***Saving seen data "./dancer.seen"
15:43:38 Join tvelocity [0] (
15:47:40 Join K4rP4D [0] (n=KrPD@unaffiliated/krpd)
15:48:13 Quit hopenglish ("CGI:IRC (EOF)")
15:50:29 Quit perrikwp (" ajax IRC Client")
15:55:13 Join mackes-Office [0] (n=root@
15:56:15 Join Zom_ [0] (
15:58:11JdGordonanyone tried out the latest disktidy patch?
15:59:10 Quit nplus (Remote closed the connection)
16:02:03 Join nplus [0] (n=nplus@141.25.Globcom.Net)
16:03:47JdGordoncan FS #6552 be closed?
16:04:18staenJust nearly found a solution for rockbox not loading/able to read the filesystem on some 5.5g 30GB ipods
16:04:40preglowshutting down three of the voltages results in non-booting d2
16:04:46preglowone makes it reboot after a while
16:07:43 Quit Zom (Read error: 110 (Connection timed out))
16:08:36GodEaterstaen: "nearly" ?
16:12:34staenGodEater: in drivers/fat.c fat_mount reads the bytes per sector out of the first sector of a partition. For some reason, it determins the wrong sectorsize on some 5.5g ipods. By changing the value in the sourcecode into 2048 rockbox is able to access the filesystem...
16:12:57 Quit Ragnor (Nick collision from services.)
16:13:07 Join Ragnor [0] (
16:13:33GodEaterstaen: interesting. I wasn't aware our fat code cared about the sector size. From what I understand, it's all fiddled with at the ATA level.
16:13:43GodEateramiconn is the man to ask
16:13:43 Nick Zom_ is now known as Zom (
16:14:01gevaertsfat has to care
16:14:02staenGodEater: with nearly I mean that I now have to look, why rockbox gets the wrong value. Changing the value manually can't be the right conclusion, as it would break the code for 5g ipods
16:14:28GodEatergevaerts: does it?
16:14:36amiconnIt does
16:14:48staenis there any sourcecode of ipodpatcher around, as this program reports the correct sectorsize
16:14:53*GodEater will cease spouting nonsense then
16:15:10GodEaterstaen: it's in the rockbox code tree
16:15:14 Join joshn_ [0] (
16:15:20amiconnIt has to. It's a bit confusing first, as one has to realise that there are 2 different sector sizes: physical sector size (ata level) and logical sector size (fat level)
16:16:05amiconnFor ipods, there's even a 3rd one: the sector size reported via usb
16:16:20staenGodEater: as I understand it rockbox works internally with 512B sized sectors and a multiplier?
16:16:26gevaertsstaen: I think that ipodpatcher uses the sector size reported over USB, so that won't help
16:16:51GodEaterstaen: pay attention to what amiconn's saying - he's the expert ;)
16:18:07gevaertsstaen: can you upload the first few kb of the data partition somewhere ?
16:18:22staen:-) Maybe I look into the loader2 code, as it is able to find and load rockbox. But I fear it uses a completly different approach...
16:18:33 Join BillDMoose [0] (n=BillDMoo@nat/ibm/x-b3ece0a24d43b000)
16:18:39staengevaerts: just a moment...
16:22:21staengevaerts: Hope the first 1024 will do...
16:23:29JdGordonamiconn: (anyone with a AJBR): is FS #7733 still a problem?
16:24:40 Quit BillDMoose (Connection reset by peer)
16:24:43JdGordonit looks like it will only work if you press ON first
16:24:52JdGordonwhich should be ok
16:26:22amiconnstaen: Rockbox just guesses the fat sector size, but first trying to mount a partition assuming 512 bytes, if that fails, it retries assuming 1024 bytes, and finally 2048 bytes
16:27:21amiconnfirmware/common/disk.c, especially #ifdef MAX_LOG_SECTOR_SIZE
16:27:58Nico_PLambdaCalculus37: IIRC the beast can't boost
16:28:27LambdaCalculus37Nico_P: Ahh.
16:28:42Nico_Pbut it probably wouldn't need to anyway
16:29:18amiconnJdGordon: Umm, no idea. Another feature I don't use at all...
16:29:35Nico_Pbuffering always boosts, so I guess if it doesn't on the beast, it's proof that the beast can't boost
16:30:12gevaertsActually, it guesses the sector size used for the partition table, and validates guesses by trying to mount. The fat code itself properly reads the sector size from the FAT bootsector
16:30:14staenamiconn: what for is the "fat_bpb->bpb_bytespersec = BYTES2INT16(buf,BPB_BYTESPERSEC);" line for in firmware/drivers/fat.c (line 314)
16:31:13amiconnIt stores the bytes per sector for later use. But it can't even get there before figuring out how large the logical sectors are
16:31:45*gevaerts blames the PC partition table format
16:31:51amiconnThis is because the bpb is located in the partition's boot sector - whose location depends on the (assumed) logical sector size...
16:32:18staenamiconn: is it stored there by rockbox itself?
16:32:46amiconnNo, it's read from there. But your problem is very likely happening in disk.c already
16:34:12staenbut by changing the value to 2048 in the sourcecode manually, rockbox loads successfully and tells the right partition size... and works on my ipod.
16:34:28*staen has to go to the door (it's ringing)
16:35:19JdGordondouble click is definatly something we dont want right?
16:37:34pondlifeJdGordon: I don't want it ;)
16:38:43JdGordongood enough for me :)
16:39:07 Join Genre9mp3 [0] (n=yngwiejo@rockbox/contributor/Genre9mp3)
16:40:14 Join toffe82 [0] (
16:41:14 Join BillDMoose [0] (n=BillDMoo@nat/ibm/x-3fe23081410fbcc7)
16:44:54JdGordonuh oh!
16:46:03peturbuild server hang :/
16:49:09gevaertsstaen: could you try zeroing the first partition entirely, and then restoring (from itunes or by following ?
16:51:00*gevaerts found that the apple OF is small enough that any bogus FAT bootsectors at offsets assuming 512-byte sectors might not be overwritten
16:51:22 Join anfortas [0] (
16:52:29LambdaCalculus37Yes, it is. :)
16:53:02JdGordonwell bugger... i cant see any typos so that commit should go through fine.. bed time
16:53:05 Nick JdGordon is now known as JdGordon|zzz (i=jonno@rockbox/developer/JdGordon)
16:54:24anfortasbut why
16:54:55 Join mf0102 [0] (n=michi@
16:55:07*scorche|sh coughs
16:55:19anfortaswhy do you cough
16:55:22 Part djh
16:55:26*LambdaCalculus37 doesn't like lag
16:55:32*scorche|sh looks at gevaerts
16:56:12*gevaerts decides that scorche|sh should have his own throat lozenges by now
16:56:19*scorche|sh doesnt
16:56:34LambdaCalculus37anfortas: Never mind. Inside jokes and whatnot.
16:56:42*petur hands scorche|sh his boots
16:56:46*gevaerts relents and hands scorche|sh a throat lozenge
16:56:50LambdaCalculus37Now check out the page linked to so you can see what Rockbox gives you.
16:57:28GodEaterstaen: I have an older version of the G5.5 firmware here - do you want to try downgrading your ipod to see if it helps ?
16:57:54scorche|shanfortas: do you have a valid support/development question, or are you here to just ask that?
16:58:35anfortasdo you like to cough, scorchesh?
16:58:55scorche|shstrike 2...
16:59:04LambdaCalculus37One more and you're going to fly.
16:59:34anfortasLambdaCalculus37 is a moron
16:59:41LambdaCalculus37Strike 3.
16:59:42Mode"#rockbox +o scorche|sh " by ChanServ (ChanServ@services.)
16:59:59*LambdaCalculus37 gives scorche|sh the cricket bat
17:00:06Mode"#rockbox +b %*!* " by scorche|sh (n=scorche@rockbox/administrator/scorche)
17:00:37scorche|shanfortas: you are muted for 10 minutes...during that time i suggest you read the channel guidelines linked in the topic
17:01:14fxbAnyone interested in helping with the Nano 2G? We have a flash dump now. But it's partly encrypted.
17:01:44LambdaCalculus37I have a 2G nano that's sitting around doing nothing. I'll gladly help test.
17:01:48 Quit BillDMoose ("Leaving")
17:02:10 Join keith [0] (n=chatzill@nat/ibm/x-17ed7da3c1b3fc5c)
17:02:27markunfxb: I briefly looked at it and only the interrupt vectors are not encrypted, right?
17:02:59markunare there any firmware updates for the nano 2g?
17:03:13LambdaCalculus37markun: I have one.
17:03:43fxbmarkun: yep.
17:04:17Nico_Pfxb: you're one of the guys from the linux4nano project?
17:04:36fxbbut the linux4nano guys dumped it
17:04:41fxbwell, someone dumped it for them
17:04:51markunfxb: I'm working on the rockbox port to the meizu m6 btw (which is also is s5l8700 based)
17:04:53 Join Red_Rabbit [0] (
17:05:16fxbmarkun: ah, Bagder told me so.
17:05:31markunhi Red_Rabbit
17:05:45 Join perrikwp [0] (i=98213a3c@gateway/web/ajax/
17:05:45Red_Rabbitso uhh.. i used to have rockbox on my 30g video ipod
17:05:53Red_Rabbitbut took it off, because it killed the battery
17:06:03scorche|shwhen was this?
17:06:12Red_Rabbiteh, couple months ago
17:06:16markunRed_Rabbit: suggestions about the battery draining?
17:06:28LambdaCalculus37Red_Rabbit: How old is your iPod?
17:06:32fxball we know is, that there are different parts in the flash (like in the nano 1g flash) but for the nano 2g, there are 512 byte headers for each part.
17:06:50fxbprobably containing the scrambled key
17:06:55markunRed_Rabbit: there have been some improvements
17:06:59LambdaCalculus37fxb: Can you post this onto the wiki?
17:07:00 Quit keith (Remote closed the connection)
17:07:11Red_Rabbiti thought there might be, im thinking about installing it again
17:07:32markunfxb: the s5l8700 has 50KB NOR memory which it boots from. The key could be in there.
17:07:37LambdaCalculus37Red_Rabbit: Go for it. Quite a improvements in power handling have been made.
17:07:38fxbLambdaCalculus37: sure. but first i need to create an account
17:07:55fxbmarkun: yea. we read it in the datasheet.
17:07:57LambdaCalculus37fxb: Sure thing.
17:08:25markunfxb: so without running some code in the first place I don't think we can get to the key..
17:08:46LambdaCalculus37fxb: We also have a forum thread if you'd like to let people see what's going on:
17:09:11fxbi think it's stored in the header, because otherwise there would be no purpose to have a header.
17:09:34 Join keith [0] (n=chatzill@nat/ibm/x-01beaab2d642ba56)
17:09:38markunmaybe. I don't know much about it.
17:09:44pixelmafxb: and GodEater wanted to talk to you ;)
17:09:58GodEaterpixelma: I already got him in #ipodlinux
17:10:01fxbbut yea... scrambled key. maybe custom decryption algorithm in that embedded 50KB NOR memory.
17:10:23pixelmaah, sorry :)
17:10:30GodEaters'okay :)
17:10:33markunluckily we don't have such problems on the meizu
17:10:45Mode"#rockbox -b %*!* " by scorche|sh (n=scorche@rockbox/administrator/scorche)
17:10:58anfortashere I am again
17:11:01anfortasyou missed me?
17:11:10scorche|shplanning on behaving?
17:11:57anfortasyes :-)
17:13:03fxbmarkun: currently i'm just sitting before the dump and staring at it :-/ but nothing happens.
17:13:06markunanfortas: why did you come here anyway? Do you have an mp3 player you want to install rockbox on?
17:13:08fxbsoo, if anyone wants to help :P
17:13:58anfortasout of curiosity
17:14:14 Nick keith is now known as BillDMoose (n=chatzill@nat/ibm/x-01beaab2d642ba56)
17:17:55GodEaterfxb: I think linuxstb has taken a look at the dump too, along with Bagder
17:18:32*GodEater 's own contribution was limited to looking at the pictures of the rig they soldered up on linux4nano in wonder
17:18:53 Part anfortas
17:19:44LambdaCalculus37fxb: I'll help prod at the dump too.
17:19:53LambdaCalculus37I guess the 2G nano will be Target #8. :)
17:20:31BillDMooseiPod 3G - r17364-080504 - works fine for a while, then acts like someone else has wildly taken over the controls. Where do I start? search terms?
17:20:44fxb | Program to extract flshdisk, flshdiag, flshlogo parts with or without headers from 1G/2G nano flashs
17:20:47BillDMoose(and howdy, everyone!)
17:21:00GodEaterBillDMoose: I'd start by upgrading the version of Rockbox you're using to one from today
17:21:04Mode"#rockbox -o scorche|sh " by ChanServ (ChanServ@services.)
17:21:31BillDMoosecan do, but it's only two weeks old - relevant?
17:21:48LambdaCalculus37fxb: So just extract to a folder and type make?
17:21:54BillDMooseokay - thanks!
17:21:55GodEaterBillDMoose: we never bother bug hunting on old code
17:22:03BillDMooseI can see that.
17:22:11fxbLambdaCalculus37: yes.
17:22:39LambdaCalculus37fxb: Thanks. :)
17:24:19 Quit Mathiasdm ("Yuuw!")
17:24:36fxboh, i dunno where you got your dump. but you need to swap bytes when you have the original one from the linux4nano guys.
17:25:07fxb16-bit swap
17:25:22LambdaCalculus37I had gotten the firmware from here:
17:25:58GodEateryou mean fxb's site :)
17:26:20*GodEater watches LambdaCalculus37 connect the dots...
17:26:34*petur hands more dots..................
17:27:13 Join Mathiasdm [0] (
17:27:17GodEaterhe's clearly a bit slow today
17:27:27*LambdaCalculus37 connects the dots... LA LA LA LAA... :)
17:27:54fxbLambdaCalculus37: well, the 2g nano firmware from there is completely encrypted
17:28:12 Quit staen (Read error: 113 (No route to host))
17:28:15fxbwe have 2 layers of encryption (or more)
17:28:15GodEateryeah, you'll want the dump the linux4nano guys got
17:28:34fxbLambdaCalculus37: just pm me :)
17:30:42*LambdaCalculus37 now has the right files
17:30:55LambdaCalculus37I'll take a look at them later on today.
17:33:08 Quit DerPapst (Read error: 110 (Connection timed out))
17:33:47 Quit joshn_ (Read error: 113 (No route to host))
17:34:38 Join Mathiasdm2 [0] (
17:40:13gevaertsWhat I think is going on with the 5.5G 30GB ipods is that for some unknown reason something (maybe itunes ?) writes a FAT header at the offset that would be expected if the partition table were 512-byte sector based
17:40:59gevaertsThis FAT header still falls outside of the apple firmware, so it is left undisturbed
17:41:48 Join joshn_ [0] (
17:41:50***Saving seen data "./dancer.seen"
17:42:07gevaertsThis _could_ be related (although it sounds unlikely) to what rexbron saw yesterday :
17:42:13gevaerts23:38 < rexbron> gevaerts: ha! Looks like it is mounting two partitions and I installed it to the wrong one (though it has never done that before)
17:42:34gevaerts(although that was on a 5G)
17:43:04LambdaCalculus37gevaerts: Hmmm... very interesting. Has he used iTunes with the iPod at all prior to that?
17:43:35 Quit Mathiasdm (Nick collision from services.)
17:43:40 Nick Mathiasdm2 is now known as Mathiasdm (
17:43:59*LambdaCalculus37 thinks iTunes may be the culprit as well
17:44:27gevaertsno idea. All others (on the forums, and staen here) have upgraded to 1.3 (presumably using itunes)
17:45:40LambdaCalculus37I haven't upgraded to 1.3 and don't plan to.
17:46:11gevaertsIf this is the case it might be possible to solve this by reversing the probe order, or by guessing the sector size from the ATA disk size and the number of sectors according to the MBR
17:46:36gevaertsAnother solution would be to make ipodpatcher clear those ghost FAT bootsectors
17:47:28LambdaCalculus37gevaerts: Which do you think is safer?
17:48:04gevaertsI would go for the ipodpatcher way I think
17:48:19gevaertsbut first we need to verify this hypothesis
17:49:14amiconnA dump of the first 1MB or so should tell
17:49:25LambdaCalculus37I think the ipodpatcher method sounds safer as well.
17:49:54gevaerts1MB is not enough. the FAT partition starts at an offset of 48195 sectors
17:50:23amiconnOf course it would be better if disk.c wouldn't need to guess, but afaik the MBR doesn't tell anythink about the virtual sector size
17:50:35 Quit voltagex_ (Read error: 110 (Connection timed out))
17:52:16gevaertsdisk.c could look at the partition table and choose the sector size that makes the partition layout best fit the disk size
17:53:32gevaertsOr at least use that to order the probes. Sometimes the "best" match isn't the correct one
17:54:14*gevaerts remembers that the beast's data partition extends a few sectors past the end of the disk
17:54:47gevaertsHas anyone reported this on a 80GB ?
17:59:46 Join herrwaldo [0] (
18:01:13gevaerts could be the same thing on 80GB
18:01:24 Quit Red_Rabbit ()
18:01:54 Join Siku [0] (
18:03:23 Quit Genre9mp3 ()
18:04:59 Quit gevaerts ("work->home")
18:05:57 Join nuonguy [0] (
18:09:30preglowanyone working on a new vmware image yet? annoying to not be able to download packages i need :>
18:09:35 Quit petur ("work->home")
18:10:23LambdaCalculus37preglow: IIRC scorche|sh mentioned he may make a new image.
18:14:02 Quit Shaid ("*toink*")
18:14:30scorche|shLambdaCalculus37: did you ever put a suggestions thing up on the wiki page?
18:14:40preglowcan't even bloody download libusb in this one
18:14:58LambdaCalculus37scorche|sh: Haven't had a chance to.
18:15:04*LambdaCalculus37 will do so during lunch
18:15:38scorche|shpreglow: well, if there are certain packages you need, put them in a suggestions for new image section on the vmware wikipage so i wont forget to include them
18:16:31*scorche|sh wonders if he should keep fluxbox or go with xfce for people who whine about fluxbox
18:18:14*LambdaCalculus37 likes Fluxbox
18:18:31*scorche|sh does too, but others seem not to
18:19:23scorche|shLambdaCalculus37: while you are editing, it, can you take the v5 suggestion in the top and throw it in the sugestions area?
18:19:27*scorche|sh adds a g
18:19:49LambdaCalculus37scorche|sh: Roger that!
18:20:35 Join staen [0] (
18:24:21preglowi don't bloody care what window manager is used as long as the consoles are nice
18:24:45 Join DerPapst [0] (
18:24:51preglowscorche|sh: you don't need to remember it as long as the repos are still working
18:25:11preglowi can't even install my own packages without compiling them from scratch or digging up some ancient sarge repo
18:25:16preglowin this image, that is
18:25:21LambdaCalculus37scorche|sh: Okay, I just added a suggestions section to the wiki.
18:26:10 Quit Mouser_X (Read error: 110 (Connection timed out))
18:27:03scorche|shpreglow: well, ideally, you shouldnt have to install any extra packages to work with rockbox in the image
18:32:22toffe82scorche|sh: would it be possible to have some srcipt that run the compil and start the sim at the end ? a script than compile all target in sequence ?
18:33:56preglowscorche|sh: then libusb needs to be in for tcctool, at least
18:34:01preglowat least libusb-dev
18:34:32LambdaCalculus37And we also need libmtp and libmtp-dev.
18:34:58linuxstbDoes the vmware image include mingw for cross-compiling Windows exes?
18:36:24scorche|shpreglow: planning on it
18:37:04scorche|shlinuxstb: well, the next image wont be additions and changes to the old one as the past have been...planning on starting anew
18:39:20 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
18:40:31gevaertsstaen: are you using linux ?
18:42:06 Join kugel [0] (n=chatzill@unaffiliated/kugel)
18:43:01 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
18:44:17jhMikeSmarkun: you want to not rotate in mpegplayer but encode videos rotated?
18:44:42gevaertsstaen: if so, could you upload the output of "dd if=/dev/sdb of=/tmp/512.bin bs=512 skip=48195 count=1", "dd if=/dev/sdb of=/tmp/1024.bin bs=1024 skip=48195 count=1" and "dd if=/dev/sdb of=/tmp/2048.bin bs=2048 skip=48195 count=1" somewhere ? (replace sdb by the correct device for you)
18:46:28 Join petur [50] (n=petur@rockbox/developer/petur)
18:49:17 Join Mouser_X [0] (
18:50:34markunjhMikeS: the meizu supports either
18:50:37 Join bertrik [0] (
18:50:40markuneh, both
18:50:57 Join miepchen^schlaf [0] (
18:51:04 Part pondlife
18:57:37jhMikeSmarkun: I suppose how it's done makes the difference though. Perhaps OF rotates an image into an intermediate buffer?
18:59:29staengevaerts: Yes, I am using linux. You find the output here:
19:00:50 Quit nplus (Remote closed the connection)
19:01:21*gevaerts doesn't understand, and goes to have dinner first
19:02:17 Quit LambdaCalculus37 ("CGI:IRC")
19:02:51 Join LambdaCalculus37 [0] (n=44a0430f@gateway/web/cgi-irc/
19:05:46 Join carini [0] (n=chatzill@pdpc/supporter/active/carini)
19:07:06 Quit linuxstb (Read error: 110 (Connection timed out))
19:10:20 Join lee-qid [0] (
19:10:33 Quit culture (Read error: 110 (Connection timed out))
19:21:20 Join tdtooke [0] (
19:21:51 Quit carini (Remote closed the connection)
19:22:12 Join carini [0] (n=chatzill@pdpc/supporter/active/carini)
19:23:59 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
19:27:16gevaertsSo my hypothesis on the 5.5 ipod is wrong...
19:28:08 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
19:28:16staengevaerts: the hypothesis about left over data in the partition? I was checking that just in this moment.
19:28:49gevaerts"left over", or accidentally written FAT boot records
19:29:00 Quit carini ("ChatZilla [XULRunner]")
19:29:21gevaertsMy hypothesis basically was that this 512.bin would also contain a FAT boot record
19:30:12staeni zeroed the partition and wrote the original bootpartion and the bootloader. which didn't start
19:30:46gevaertsDid you use ipodpatcher for the bootloader ?
19:31:32 Quit Galois (Remote closed the connection)
19:31:51 Join Galois [0] (
19:32:39 Join moos [0] (
19:34:18staenno, that didn't work, because it didn't recognize the bootpartition as such. I used dd and wrote the bootpartition.bin into it wich I stored right at the beginning of my experiments
19:35:15gevaertsI _think_ that ipodpatcher also fixes up the OF to actually jump to the bootloader
19:35:27*staen is thinking that that saved partition could also had the information in it
19:36:31gevaertsYou could zero the partition, and then restore by following That should leave the zeroes
19:40:53staenI will try it again
19:40:53 Join gevaerts_ [0] (
19:41:04 Quit amiconn (Nick collision from services.)
19:41:10 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
19:41:28 Quit gevaerts (Nick collision from services.)
19:41:35 Nick gevaerts_ is now known as gevaerts (
19:41:51***Saving seen data "./dancer.seen"
19:43:38 Join spiorf [0] (
19:47:52 Quit tvelocity ("Αποχώρησε")
19:48:39 Join lee321987 [0] (n=04a11fdd@gateway/web/cgi-irc/
19:49:07 Quit lee321987 (Client Quit)
19:49:48 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
19:52:14 Join Lear [0] (i=chatzill@rockbox/developer/lear)
19:55:27 Join lee321987 [0] (
19:58:14*gevaerts waves
19:58:22lee321987Someone said yesterday, "I'm still convinced the file system corruption on PP targets is a userbase problem." Could someone be so kind as to tell me what "userbase" means?
19:59:46BigBambiuser error
19:59:58BigBambiIs what that statement looks like to me
20:00:36domonokyit probably means, that the user had done something wrong, and its not a rockbox bug..
20:03:06amiconngevaerts: Extending your idea, it would possibly make sense to probe the varying logical sector sizes going from largest to smallest, leaving out those where the partitions wouldn't fit
20:03:22lee321987If I were to enable Rockbox USB support on my Sansa c250, and only use it to transfer audio files, would the only danger be corrupting those audio files? What about safe charging?
20:03:51BigBambicharging is independent of mass storage, and isn't properly enabled on the sansa
20:04:32 Join mcuelenaere [0] (n=mcuelena@rockbox/developer/mcuelenaere)
20:04:37BigBambigevaerts: Is there the possibility of corrupting the FAT, or is it just the files themselves?
20:04:59LambdaCalculus37USB support on the c200 is safe enough at USB 1.1 speeds; I enabled USB on my c240 and have transferred files with no problems.
20:05:16mcuelenaereBagder: I fixed a typo in tools/configure which requires (again :)) changin the build script; could you look into it when you have some time?
20:05:16BigBambiLambdaCalculus37: Not necessarily I believe
20:05:28 Quit mcuelenaere (Client Quit)
20:05:36BigBambiLambdaCalculus37: I use high speed with no problems, but that isn't conclusive
20:06:50gevaertsBigBambi: if you have data corruption, it can easily be anywhere. However I haven't seen it much on internal flash, and I haven't seen it at all with the first patch in FS #8663
20:07:04BigBambigevaerts: thanks
20:07:18BigBambilee321987: I direct you to the reply from gevaerts
20:07:19lee321987so if charging isn't properly enabled does that mean it could be bad for my battery to use Rockbox's USB? (since it will be charging while I transfer files)
20:07:28BigBambiIt won't be charging
20:09:00gevaertsamiconn: that will probably be the best solution. I'm still trying to work out why it doesn't work now though, as staen doesn't have FAT bootsectors on 512 and 1024 byte calculated offsets
20:09:32 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
20:10:00amiconnA dump that includes everything from sector 0 up to the fat boot sector would be interesting
20:10:56gevaertsyes, but it would also be big. With some luck it would compress well though
20:11:50amiconnIt will probably compress very well
20:12:45gevaertsstaen: could you provide such a dump ? (dd if=/dev/sdb of=/tmp/dump.bin bs=2048 count=48197 ; bzip2 /tmp//tmp/dump.bin)
20:12:53lee321987So I could (though it seems unlikely) brick my c250, even if not using Rockbox's USB only for audio files?
20:13:17lee321987i meant "even if using"
20:13:47 Quit nuonguy ("This computer has gone to sleep")
20:15:04domonokylee321987: no you wont "brick" it, but you could get errors on the drive, and have to reformat it..
20:15:14gevaertslee321987: Not "brick" in the sense of "not fixable", but it might require some work to recover if you're unlucky. Do you have access to a linux machine ?
20:15:37lee321987no. XP using Cygwin
20:16:32gevaertsI would recommend applying the first patch on anyway. It seriously reduces the risk
20:16:53lee321987would those potential errors only be on the partition I'm writing to?
20:18:05 Join obo [0] (n=obo@rockbox/developer/obo)
20:18:18LambdaCalculus37BigBambi: True.
20:18:35lee321987how would I slow my USB speed down to 1.1?
20:18:41 Quit Mouser_X (Read error: 110 (Connection timed out))
20:18:43*gevaerts really doesn't understand this ipod issue.
20:22:47 Quit bertrik (Read error: 104 (Connection reset by peer))
20:23:41 Join bertrik [0] (
20:24:08gevaertsstaen: how exactly did you change the sector size to 2048 to make things work ?
20:24:34gevaertslee321987: just don't enable high speed
20:26:41kugelgevaerts: why is that udelay needed on e200 but not on c200?
20:26:47lee321987USB 1.1 is full-speed, and 2.0 is high-speed. Correct?
20:27:03gevaertskugel: I think that if we knew that the real bug would already be solved
20:27:17amiconnUSB 1.1 and 2.0 are protocol versions
20:27:43kugelgevaerts: but it is confirmed that removing it causes issues on the e200 (unlike to c200)?
20:27:47amiconn1.1 can operate at lowspeed and fullspeed. 2.0 can operate at lowspeed, fullspeed and highspeed
20:28:02gevaertskugel: yes
20:29:03kugelgevaerts: just asked, since it doesn't seem to be mentioned on that tracker post
20:29:37kugeland I didn't really get what desowin said there
20:29:40gevaertsSome people tried it before I put that on the tracker
20:29:54kugelah ok
20:29:57desowinkugel: ?
20:30:07kugeldesowin: hehe :)
20:30:17 Quit tdtooke (Read error: 110 (Connection timed out))
20:30:29kugeldesowin: I was talking about FS #8663
20:31:26staengevaerts: changing the sectorsize: I did it the hard way and changed line 314 in /firmware/drivers/fat.c to fat_bpb->bpb_bytspersec = 2048;
20:32:23amiconnstaen: Hmm. How did you create the data partition?
20:33:00lee321987I'm looking at the Wiki on how to enable USB on PPs. If I just want to be able to transfer files I don't need to worry about the " logf() over USB Serial" right?
20:33:04 Quit simonrvn ("new kernel")
20:33:16 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
20:33:20gevaertslee321987: no. That's an entirely different thing
20:34:29amiconnIf you changed it in fat.c, it means that the disk.c code found the partition, and the bytespersec value in the fat bootsector is actually wrong
20:34:39staenI tried many different ways :-) I restored my Ipod using Itunes, I did it like in, I even copied one from a working ipod
20:34:50 Join tdtooke [0] (
20:35:00 Join merbanan [0] (n=banan@
20:35:17kugeltdtooke: ping
20:35:55gevaertsThe bytespersec value looks correct to me in
20:36:34amiconnstaen: You copied the whole disk content?
20:37:02staengevaerts: here is the dump you asked for a bit earlier:
20:37:25kugeltdtooke: You wanted to talk to me?
20:37:26staenamiconn: Yes. Took 12hours :-)
20:38:29gevaertsThanks. Now we can have some fun with it :)
20:40:38 Join nplus [0] (n=nplus@141.25.Globcom.Net)
20:41:43 Quit Siku ()
20:44:55 Quit crope` ("Leaving")
20:46:49gevaertsstaen: just to verify : does the OF boot with this ?
20:47:43 Join crope` [0] (
20:49:17staengevaerts: booting with what exactly? the copied filesystem? that bootet into the apple firmware but not into rockbox
20:49:38 Join simonrvn [0] (i=simon@unaffiliated/simonrvn)
20:51:00 Join CFP [0] (n=554431fa@gateway/web/cgi-irc/
20:51:27 Quit XavierGr (Nick collision from services.)
20:51:38 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
20:52:39gevaertsThere is a valid FAT bootsector at the start of the _first_ partition assuming 512-byte sectors. I have no idea where this comes from
20:53:26lee321987I'm trying to enable USB on my Sansa. Anyone know why the file "firmware/target/arm/sandisk/ata-c200_e200.c" is not on my system?
20:53:42CFPhello !
20:54:05 Join nedd1 [0] (n=john@nat/sun/x-32345dd31cdaaeb1)
20:54:27CFPAnyone knows how I can solve this error : error: ‘struct plugin_api’ has no member named ‘lcd_set_viewport’
20:54:31CFPwhen using make
20:54:45CFP(I'm developping my own plugin)
20:55:02gevaertslee321987: apparently that file got renamed in the meantime. Try editting firmware/target/arm/ata-sd-pp.c instead
20:55:27lee321987<trying...thank you.
20:55:38linuxstbCFP: That function isn't available in the plugin API
20:56:06linuxstbCFP: You can use the Rockbox "screens" API - see the test_viewport.c plugin
20:57:40staengevaerts: is there any chance to correct this?
20:58:20gevaertsstaen: the "quick" way would be to just use dd to zero out that bit, but I think I can commit a "real" fix soon
20:58:36CFPthanks linuxstb, but how can I make use of the "screens" API ?
20:58:57gevaertsAre there ipods (modded, with ipl, whatever) where the first partition is the data partition ?
20:59:07linuxstbCFP: See the test_viewport.c plugin... Plus almost all the core Rockbox code uses it.
20:59:20CFPand is there a way to know the heigth of a char/line of text ?
20:59:34CFPok, I'll have a look
20:59:49linuxstbThe code in apps/gui/ will probably be helpful
20:59:50CFPthank you very much
21:01:08staengevaerts: i think I can wait for the real fix. until then I can use my hacked version. thank you so far!
21:01:34 Join m0f0x [0] (
21:05:31*gevaerts suddenly realizes that this would need a new bootloader ...
21:06:15gevaertslinuxstb: is it doable to change ipodpatcher to zero out a sector on 2048-byte sector ipods when installing the bootloader ?
21:06:21amiconngevaerts: That sector is odd. It shouldn't exist at all...
21:06:57amiconnstaen: You could zero out that sector using dd. Then you won't need the hack anymore
21:07:02gevaertsIndeed not, but it does exist, and it fits the symptoms
21:07:23amiconngevaerts: I wonder what the proper fix would be. Accept only valid partition types?
21:07:47amiconnRight now rockbox tries to mount everything assuming fat except type 0x05
21:07:55amiconn(extended partition)
21:08:15gevaertsAccept only valid partition types, or the trick used for Gigabeat S (i.e. skip the first partition)
21:08:46staengevaerts: I wonder where the wrong information comes from, as my ipod seems not to be the only one having problems.
21:08:53amiconnThe drawback of an explicit check would be that there are several types, depending on a combination of factors
21:09:25gevaertsamiconn: this one is marked "empty" (type 0). Skipping those should be doable
21:09:32 Quit petur ("switching")
21:09:40 Join petur [50] (n=petur@rockbox/developer/petur)
21:09:40amiconnYes, empty partitions could be skipped
21:10:13gevaertsAnyway, any "real" fix requires a new bootloader, while just zeroing out this sector from ipodpatcher (and rbutil) would not. I would consider doing both
21:11:17linuxstbgevaerts: Yes, that would be relatively trivial.
21:11:20gevaertsstaen: could you try "dd if=/dev/zero of=/dev/sdb bs=512 count=1 seek=63" ?
21:12:38staenin a moment, just have to fetch the ipod from the other room
21:12:48amiconnA proper check would need to allow the following types: 0x0b, 0x0c, 0x1b and 0x1c. When compiling with fat16 support, the following additional types need to be considered valid: 0x04, 0x06, 0x0e, 0x14, 0x16, 0x1e
21:16:18gevaertsMy thinking with the ipodpatcher change is that if this is caused by itunes in some way, only itunes will put it back (with a restore command), which needs re-running ipodpatcher anyway
21:16:25 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
21:23:57 Join jaczehack [0] (n=d572f7f1@gateway/web/cgi-irc/
21:25:08staengevaerts: got my ipod and zeroed the recomended bytes. And suddenly it worked!
21:25:23gevaertsDoes the Apple firmware also still work ?
21:25:37*gevaerts thinks that we should be careful
21:26:35staenIt boots up and is working.
21:27:28gevaertsThanks. Now to fix it for real, so other people are helped as well :)
21:28:10 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
21:28:18 Quit moos (Read error: 110 (Connection timed out))
21:28:30 Quit Horscht (Remote closed the connection)
21:29:28gevaertsAs I see it, fixing this properly will require either a new bootloader or a new ipodpatcher/rbutil. Which is easiest ?
21:30:03linuxstb_A new bootloader requires a new ipodpatcher (the bootloaders are embedded in it)...
21:30:22*gevaerts decides to let other people decide
21:30:53linuxstb_IMO, it would seem better suited for ipodpatcher.
21:31:45 Join tvelocity [0] (
21:32:35 Join bertrik_ [0] (
21:32:42 Quit DaCapn ()
21:34:06 Join moos [0] (
21:34:53domonokyfixing it in ipodpatcher needs new releases of it and rbutil. Fixing it in the bootloader means a new bootloader and ipodpatcher is needed, but rbutil would work without new release .. :-)
21:35:12*gevaerts lets domonoky and linuxstb fight it out :)
21:35:45 Quit CFP ("CGI:IRC (EOF)")
21:35:47*LambdaCalculus37 gets the popcorn :)
21:37:02gevaertsBut fixing it in the bootloader would need a lot more testing IMHO
21:39:22lee321987Alright! I've USB on my c250. So - my battery's charge state does not matter at all? Can I leave it hooked up for long periods of time and feel safe?
21:39:47 Quit linuxstb (Read error: 110 (Connection timed out))
21:40:15amiconngevaerts: It might actually be a good idea to check partition types.
21:40:25amiconnIt means a bit more code though
21:40:59amiconnA while ago I was thinking different, but then I didn't think of such weird coincidences...
21:41:18amiconnIt should be done in a size efficient way of course
21:41:54gevaertsYes, no sense in throwing 2kb at it
21:41:55***Saving seen data "./dancer.seen"
21:43:26amiconnPerhaps checking against an array of allowed types
21:43:50amiconnBit 4 could be masked out, that would cut the allowed number of types in half
21:45:08 Join ch4os_ [0] (n=ch4os@unaffiliated/ch4os/x-059673)
21:45:56*gevaerts isn't as quick in looking at bits
21:46:26 Join joshn__ [0] (
21:46:54 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
21:46:54 Quit BillDMoose ("ChatZilla [Firefox]")
21:47:57lee321987I know it's not guaranteed.
21:49:13 Quit bertrik (Read error: 113 (No route to host))
21:50:15lee321987But it's considered safe right?
21:50:59gevaertsIt shouldn't cause accidents, but I have no idea on the effect on battery life (I know very little about batteries and charging)
21:51:40amiconnLooks like a job for memchr()...
21:52:28 Nick fxb is now known as fxb__ (
21:53:07 Join Hillshum [0] (
21:53:19lee321987Ok, I'll keep an eye on it. Thanks guys, you've been a lot of help. Have a good one.
21:53:40 Part lee321987
21:54:41gevaertsmemchr could indeed be a good way. One condition, one function call, and 4 or 10 bytes of const RAM
21:55:03 Join DrMoos [0] (
21:55:07 Quit XavierGr (Read error: 110 (Connection timed out))
21:55:34amiconnWould be the first use of memchr() in the core - it is linked anyway because it's exported to plugins and codecs
21:56:32 Quit moos (Read error: 104 (Connection reset by peer))
21:56:35 Nick DrMoos is now known as moos (
21:56:49gevaertsamiconn: are you working on it, or shall I do it ?
21:56:54amiconnI am
21:57:49 Quit lee-qid ("aufwiederbyebientotsayonara")
21:58:37 Quit LambdaCalculus37 ("CGI:IRC 0.5.9 (2006/06/06)")
21:59:52 Quit ch4os (Read error: 110 (Connection timed out))
22:00:05 Quit joshn_ (Read error: 110 (Connection timed out))
22:01:03 Quit jaczehack ("CGI:IRC (EOF)")
22:01:07 Quit Lear ("ChatZilla 0.9.82 [Firefox 3.0pre/2008051206]")
22:03:08 Nick ch4os_ is now known as ch4os (n=ch4os@unaffiliated/ch4os/x-059673)
22:04:43peturpixelma: ping
22:04:58 Quit merbanan (Remote closed the connection)
22:05:55peturI noticed the manual says 'Quick Menu' and the lang file 'Quick Screen'. Which one do we take?
22:06:54 Part staen ("left")
22:07:33pixelmaI'm not sure myself...
22:08:19peturquick menu seems better to me
22:09:22*amiconn will test on various targets now
22:09:22*n1s thinks the am/pm thing is pretty stupid...
22:09:29linuxstb_It's not really a menu though...
22:09:36linuxstb_(compared to the other Rockbox menus)
22:09:43 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
22:10:17gevaertsSpeaking of the manual, I noticed that the "continue playing while plugged into usb by holding key X" isn't documented
22:10:57pixelmabut you can adjust settings... though we also have other screens where you can adjust "something" (starting with the WPS, radio screen...)
22:11:32n1sgevaerts: if you aren't going to fix it please file a bug or put it in the wiki
22:12:05 Quit mf0102 ("Ex-Chat")
22:12:26*gevaerts will file a bug. He's willing to fix this, but it will require a lot of study on how the manual works
22:12:33pixelmagevaerts: not the only thing you'll find undocumented, for sure... :\
22:13:43n1sgevaerts: basically add a new button macro to the platform files and use that probably with some opting (I'm sure some player doesn't support this)
22:14:32 Join BHSPitLappy_ [0] (
22:14:49 Join tessarakt [0] (
22:15:21gevaertsn1s: that's what I thought, but as I've never actually done any manual work, it will take me some time. I'll file a bug anyway in case I don't complete this
22:17:05amiconngevaerts: USB power mode should be documented, as it's a quite old feature already
22:17:16*gevaerts checks again
22:17:23amiconnAnd no, not all targets support it, but only those which can use power from USB
22:17:34amiconn(for charging or otherwise)
22:18:36 Join Fromspain [0] (
22:20:16gevaertsI can't find it anyway...
22:20:24pixelmameh... found many references to "charging/charger" in the Ondio manuals :(
22:20:25 Quit HellDragon (No route to host)
22:21:29*petur thinks about changing 'Quick Menu' to 'Quick Screen' in the manual
22:25:54 Part Fromspain
22:28:30 Quit BHSPitLappy (Read error: 110 (Connection timed out))
22:29:20gevaertsHow does a HAVE_* define get to be an \opt ? Is there a magical way to do this, or do I need to define them myself somewhere ?
22:29:23n1sanyone who knows sound.c around?
22:29:59 Join roseen [0] (n=51e31bbd@gateway/web/cgi-irc/
22:30:56kugelpetur: I find quick menu more suitable
22:31:45 Quit tessarakt (Read error: 113 (No route to host))
22:31:49peturwell it certainly doesn't look like a menu
22:32:07*BigBambi has always known it as and used Quick Screen
22:32:12Bagderand it doesn't behave like a menu
22:32:15kugelpetur: but a "screen" means to me, that there's nothing to do other than loocking at it
22:32:46pixelmagevaerts: there are both - if a suitable language "feature" exists you can use that (compare with the features.text in "apps") because there's a features.tex with some options which is automatically generated. Otherwise you'd have to define so called UseOptions yourself in the manual platform files
22:32:53peturyou can press various buttons on the wps and recording screen...
22:32:59kugelBut actually I'm also more familar with quick screen
22:33:13pixelmagevaerts: automatically generated from the features.txt
22:33:15kugelI read quick menu once in the manual, but the source is full of "quick screen"
22:33:21*gevaerts looks
22:33:42*gevaerts is lucky :)
22:34:13kugelpetur: but you can't really change settings in the wps/rec screen, do you?
22:34:24BigBambirec screen, yes
22:34:33BigBambiwps volume
22:34:36pixelmakugel: can't you change the volume - isn't it a setting?
22:34:45kugelanyway, I just find menu more suitable. I didn't said that it is the best name ;)
22:34:52BigBambii.e. you can on the rec screen too
22:35:20gevaertsNow the main issue.. where in the manual should this be put ?
22:36:00kugelIn the quickscreen, there's nothing special to look at, besides the settings you can change(which applies for e.g. a settings menu too)
22:36:11kugelarg, i said quick screen :/
22:36:16BigBambinothing special to look at?
22:36:41 Join tessarakt [0] (
22:36:53pixelmayou know a WPS for an eye candy junky ;)
22:37:23*gevaerts has no idea where to put the descripton of a feature that works in _every_ screen
22:37:26*BigBambi wants flashing lights and animations in the quick screen :)
22:37:33amiconngevaerts: Mounting partitions works on my mini with that check. I also tried a fresh bootloader
22:37:38BigBambigevaerts: tricky one
22:37:45amiconnWill try other targets now
22:37:51*n1s screams at the 'cleaned up' sound.c
22:37:51scorche|shwe need a WQS
22:40:25kugelgevaerts: Put it everywhere :)
22:40:35 Join raph1 [0] (
22:41:03pixelman1s: why doesn't that surprise me? What's wrong?
22:41:17 Quit roseen ("CGI:IRC (EOF)")
22:42:13 Part raph1
22:42:15 Join raph1 [0] (
22:42:19 Part raph1
22:43:15 Join vector011 [0] (
22:43:27n1spixelma: the order of things were changed so on the first call to set_bass/treble the bass/treble is set to 0 on sw tone control targets, the fix for that is easy but the calculation and the call using the value are #ifdeffed on different things...
22:43:57 Join iRelinquish [0] (
22:45:25n1sand these values are checked in another place too, #ifdeffed on a third condition
22:45:58*petur discovers 'Quick Screen Menus' in the manual and runs off screaming
22:46:22 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
22:46:46pixelmagevaerts: maybe the "turning player on and off"section? But I'm not sure...
22:48:05n1sI wonder if I should commit the half fix and leave the ifdef hell to someone else...
22:48:14pixelmaI'd want it somewhere in the beginning though, in a more "general" part
22:49:21pixelman1s: is it because the cleanup isn't finished yet, or something introduced recently?
22:50:24n1sthe real bug was introduced in one of the clean up commits the ifdef hell was only made slightly harder to follow :/
22:51:00 Quit vector011 ("")
22:52:59 Quit joshn__ (Read error: 113 (No route to host))
22:54:21 Quit Hillshum ("ChatZilla 0.9.82 [Firefox]")
22:54:45*n1s gives up, I will just break something else if I touch this
22:54:46pixelmapetur: it's always like that - you look at something small which you think is fixed quickly and while at it find other parts which should then be fixed too... (e.g. the dice.tex - the button table is wrong because things changed when someone decided to use pluginlib actions there and introduce a menu to set number of dice etc.)
22:57:49 Join joshn__ [0] (
22:58:53 Quit DerPapst (Read error: 113 (No route to host))
23:02:04 Join Casainho [0] (
23:04:20CasainhoI have a question: I did install arm-elf-gcc using and now I am trying to build a program for one ARM processor, that program uses a file "stdlib.h" that looks like should be in the PATH, but I don't have and I would like know what did I miss in the installation...
23:05:14Bagderthat sounds like a header file a libc should provide
23:05:18Bagdernot the compiler
23:05:31Bagderwhat program are you building?
23:07:08Casainhoits an bootstrap, for the Rockbox Player prototype −− AT91Bootstarp:
23:08:01*petur executes rfi
23:08:21peturpixelma: I came accross \section{\label{ref:QuickScreenMenus}Quick Screen Menus} - do you know what it is used for?
23:08:22Casainhoso, should I install Libc? - but there is some Libc for ARM?
23:08:44BagderI don't see why a bootstrap code would need a libc/stdlib.h
23:08:56pixelmapetur: yes, it's used for cross-references
23:09:29pixelmaor do you mean something else besides the "ref" part?
23:09:38BagderCasainho: typically if you need a libc you want a different gcc build
23:10:09CasainhoBagder: it is used on file "dataflash.c", in the driver directory...
23:10:09peturpixelma: I'm also a bit confused about the button two and three on recorder ;)
23:10:59BagderCasainho: what code?
23:11:03pixelmathe Recorder has more than one quick screen - that's all I know about it (my Ondio doesn't even have one) ;)
23:11:07Bagderthe link you gave me was an index
23:11:24CasainhoBagder: so, Libc comes with gcc build? - I am confused... I use normally a GCC-AVR and that comes with Libc-AVR...
23:11:31*petur tries the recorder manual
23:11:33 Join einhirn_ [0] (
23:11:48BagderCasainho: no you're mixing things
23:11:56Bagdergcc comes with no libc
23:12:03CasainhoBagder: thet code is in the "at91bootstrap":
23:12:04Bagderbut you can build it to use one
23:12:20BagderCasainho: in that pdf?
23:12:38 Join HellDragon [0] (i=jd@unaffiliated/helldragon)
23:12:57pixelmapetur: I think it's explained quite differently there (even in another chapter) - if I'm not remembering wrong
23:13:28peturI think I found two different descriptions for the same quickscreens :/
23:13:32CasainhoBagder: wait, I will find the link for the code...
23:16:25CasainhoBagder: here the code and the PDF:
23:17:12 Part tdtooke ("Leaving")
23:17:17 Join DerPapst [0] (
23:17:26Bagderso the README even mentions the libc they've used
23:17:33Bagdernewlib 1.12.0
23:17:36 Quit OlivierBorowski ("Konversation terminated!")
23:18:32CasainhoBagder: eheh, I miss the REDME :-) :-)
23:19:09Casainhoso I must install at least newlib 1.12.0... - I will do that then...
23:19:14BagderCasainho: so if you want the easy route, you build/get yourself a gcc setup like that
23:20:09Bagderof course I still can't see why they write a bootstrap code to use a libc
23:20:19CasainhoBagder: okok, thank you. What do you think about the idea of using chumby as a development hardware platform to get the Rockbox Player?
23:20:40Bagderif that has similar hw, then sure
23:20:54Bagderfor sound, buttons/inputs and lcd
23:21:16Bagderwhat kind of inputs does the chumby has?
23:21:24Bagderhm "have"
23:22:17CasainhoI don't know, I just started to looking for chumby today...
23:22:18LloreanThe Chumby is more of a non-portable PDA. It has a touchscreen, and is meant to be plugged in as a sort of lightweight wifi internet device for use in the home.
23:22:29LloreanIt has wifi, but last I checked, no battery.
23:22:46Bagderthen it seems like a less suitable dev platform
23:23:00LloreanI think it's a point of interest because it's open source, and uses one of the CPUs they're interested in.
23:23:27Bagdersure, but if the cpu is the only common ground there's not a lot to gain
23:23:45 Join Falafel [0] (
23:23:47CasainhoLooks like to me a cheap way to get the hardware for developing, was It it was a development board....
23:24:23LloreanBagder: The chumby uses the i.mx21
23:24:25CasainhoBagder: It's the CPU, memories and TFT display...
23:24:25BagderCasainho: you can get arm boards cheap sure, they just won't be anywhere near your actual rockbox-capable hw ...
23:24:48Bagdermemories (plural) ?
23:25:09CasainhoSDRAM and flash - sorry for my english
23:25:13peturpixelma: would it be ok to also remove the quick menu part from the recorder manual as it has a different section for that and it only describes one of the two quick screens...
23:25:26BagderCasainho: well, SDRAM isn't much to "get used to"
23:25:55Casainhoyes, right... maybe the important thing is the MCU and the TFT display..
23:26:01Bagderand "flash", are you saying has the same SD(?) interface you want for rockbox?
23:27:24Casainhowell, turns out to be the MCU and the TFT display the only things that we will win... looks like TFT displays are very expensive
23:28:02CasainhoI don't really know :-) :-) −− I will take some days to read more about chumby
23:29:20BagderI wouldn't bother much
23:29:35Bagderif you need a practise rockbox target, I thought you already had a sansa?
23:30:05Casainhowell, but If chumby is a specie of hardware that have USB, flash memory, display, audio codec and buttons, so It's good to have all in the same hardware
23:30:29BigBambidoesn't a e200 have that?
23:30:36Casainhoyes, I have sansa but others developers don't have sansa... in general sansa V2 in not availaby or is dificult to buy
23:30:37Bagderassuming you'd use the same set in your rockbox device, then sure it would be neat
23:30:50 Quit sup ("leaving")
23:31:07Casainhoah, and sansa do not have the hardware that we will use...
23:31:19pixelmapetur: so no info would be lost?
23:31:33 Quit mackes-Office ("Ex-Chat")
23:31:33peturconfusion would be lost ;)
23:32:23pixelmaI'd say go for it then, if it makes the manuals more consistent
23:32:37Casainhookok - bye bye good night
23:32:44 Quit Casainho ("ChatZilla [Firefox]")
23:34:11joshn__is anyone here responsible for the master build server?
23:34:30*petur points to Bagder
23:34:34BagderI am
23:35:26joshn__I had to regenerate the ssh hostkey for my build server, because of the debian bug. Just FYI.
23:35:41Bagderyes but we don't
23:36:06Bagderour set is (way) older than this bug
23:36:22joshn__yes, just wanted to let you know, in case the master build server complains about a changed key
23:36:26joshn__when connecting to my computer
23:36:41Bagderwhat host name is that?
23:37:58amiconnBagder: Hmm, does that master complain if the ssh key changed?
23:38:02amiconnMine changed too...
23:38:33*petur wonders why some devs put stuff on the tracker in stead of committing directly (and I'm not talking about controversial or experimental things)
23:39:05LloreanSomething specific in mind?
23:39:23peturlike 8975
23:39:49 Quit Falafel (" HydraIRC -> <- \o/")
23:39:54preglowdoes he have commit rights?
23:40:21pixelmaI think petur means JdGordon's latest version
23:40:30*linuxstb points preglow to docs/COMMITTERS
23:41:15linuxstbHe doesn't appear to have commit rights...
23:41:36BigBambiThat's LambdaCalculus
23:41:42peturwoops... confused him with somebody else :/
23:41:57***Saving seen data "./dancer.seen"
23:43:21linuxstbAm I misunderstanding, or is that patch wrong? It passes "2" as the size parameter to strncmp, so will delete all files starting "De"...
23:44:40gevaertsIt looks like it
23:45:05*gevaerts concludes that this patch is controversial after all
23:46:04linuxstbDoesn't seem controversial to me...
23:46:46 Quit tvelocity (Remote closed the connection)
23:46:55gevaertsIf it deletes all files starting with De ?
23:47:21linuxstb"controversial" requires some people in favour of it...
23:47:31peturthat's just a bug
23:47:48preglowwhat a weird way to use strncmp...
23:48:21amiconnlinuxstb: Hmm, do we know whether HW_REV can be used to reliably distinguish G5 and G5.5?
23:48:54Bagderbring your chinese dictionary
23:49:22*gevaerts can read most of the screenshots
23:49:26linuxstbamiconn: I don't think so. IIUC, HW_REV refers to the main board, not the disk... Some refurbished ipods have been reported to have 2048-byte sector disks with a 5G board (IIRC).
23:49:26scorche|shhow dare they not show a screenshot of robot finds kitten
23:50:00amiconnlinuxstb: It's not the disk that causes the G5.5 to use 2048 byte *logical* sectors
23:50:14preglowcomplete with the "wrapped two upper lines" bug
23:50:46peturwas just going to say that
23:50:58preglowfancy bug, that
23:51:15preglowthe wrapping doesn't happen if you tcctool transfer the binary :>
23:52:24preglowjhMikeS: btw, my nano has started behaving substantially nicer lately, is there a reason for it?
23:52:34preglowjhMikeS: disk access sluggishnes-wise
23:55:04amiconnpreglow: I'd guess that bug is caused by the flash access problems. Some weird error in the loaded binary
23:56:45amiconnHmm, perhaps not
23:57:07amiconnIt could also be a hardware init difference
23:58:57preglowamiconn: flash access problems?

Previous day | Next day