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 2006-03-02

00:00:06cool2bdaveIn wps it seems to copy the current and next
00:00:40cool2bdaveis that in addition to the N loaded songs
00:00:44Bagderthe wps doesn't copy the id3 struct
00:01:06Bagderbut yes it uses pointers to it and the next
00:01:35 Join darkless [0] (
00:01:47cool2bdaveof coarse
00:02:53cool2bdaveSo would it be sensible to make my own structure in wps, decide which song is playing and extract its lyrics there
00:03:17linuxstbThe person working on the album art patch is attempting to store the image in the audio buffer - immediately before the track itself. You could do a similar thing with the lyrics data, but you would need one static buffer to copy the lyrics into when that track starts playing.
00:04:47cool2bdaveits a lot of memory to take up staticly, Is dynamic possible?
00:05:06amiconnThe metadata buffers (id3 structure are static. 30 on swcodec, 16 on archos
00:05:16 Join ts-x [0] (
00:05:42amiconnwe have no malloc() in the core, on purpose
00:06:36JdGordoncool2bdave: the playlist viewer uses the plugin buffer, you could use that..
00:07:02amiconnImho it doesn't make much sense to do this kind of work right now. At devcon we will discuss ideas for improving the whole audio buffer system including metadata etc
00:07:29amiconnJdGordon: That's not possible for buffering lyrics
00:07:40JdGordony not?
00:07:44 Quit x1jmp (Read error: 113 (No route to host))
00:07:49 Quit petur ("here today, gone tomorrow")
00:07:59cool2bdaveSo should I hold off till then
00:08:36amiconnYou want to have lyrics accessible during playback. What if a plugin is loaded, or the user uses the playlist viewer?
00:08:40BagderJdGordon: well, you should of course be able to run a plugin while playing songs
00:09:06JdGordonye, so rebuild it after the plugin exits... you cant view lyrics while in a plugin anyway
00:09:08 Join quobl [0] (n=quobl@tor/session/x-46e0baad0b2bd4fa)
00:09:19Bagderyou mean reload from N songs?
00:09:19JdGordoni understand its not the best way.. but it works..
00:09:28Bagderseems... odd
00:09:48JdGordonoh whaa??? all the lyrics are loaded at the start? not just the current song?
00:09:59amiconnThe plugin buffer can be used for data, but only if the usage is "atomic", i.e. not interruptable by another process that would also use the plugin buffer
00:10:29BagderJdGordon: we must not spin up the disk for each song, that's not clever
00:10:30amiconne.g. as a buffer for copying files
00:10:33JdGordonrighto... anyway, i have to go.. damn uni
00:13:22cool2bdaveOn the same lines then I shouldnt store the current song data in gwps.c as that would spin up the disk each time too
00:15:32amiconnIf we are going to support lyrics, this would need to be part of the metadata handling,
00:15:51amiconnso that lyrics are loaded when the song gets buffered
00:15:52markuncool2bdave: try linuxstb's suggestion
00:16:42 Join bernd [0] (
00:17:05 Nick bernd is now known as bernd_b (
00:17:07cool2bdaveHow do I get access to the audio buffer
00:17:37cool2bdavewhich files?
00:19:03 Join ashridah [0] (
00:20:02cool2bdaveonly found audio.h
00:30:37cool2bdavejust one f
00:36:00 Quit Steven_ (Read error: 110 (Connection timed out))
00:37:11 Join Steven__ [0] (
00:37:19linuxstbcool2bdave: On software-codec platforms, I think playback.c manages the audio buffer.
00:37:47linuxstbThis calls the get_metadata() function immediately before it starts to load each track into the audio buffer.
00:38:45linuxstbMy suggestion would be to change playback.c to pass a pointer to some free space in the audio buffer to get_metadata(), and that function could store extra information, such as lyrics or album art in that space.
00:39:45linuxstbBut when that track starts to get decoded, the data needs to be moved elsewhere - so that part of the audio buffer can be freed to start buffering future tracks.
00:44:30cool2bdaveSo the currently playing track should have its lyrics stored somewhere else, but all buffered tracks will have their lyrics stored until they are played
00:46:24cool2bdaveI could use gwps.c
00:58:21 Quit markun (Read error: 113 (No route to host))
01:02:16 Join markun [0] (
01:04:15Hideo_Kuzelol, ?
01:04:27Hideo_Kuzebastard students
01:08:12 Quit bernd_b (Remote closed the connection)
01:09:00 Quit Thus0 ("Leaving")
01:09:04Hideo_Kuzetest failed
01:10:39cool2bdavethanks all and linuxstb, I have much code reading to do. playback.c is a long one. I will try the suggestion
01:10:52cool2bdaveThanks again TTFN
01:11:32 Quit cool2bdave ("CGI:IRC (EOF)")
01:15:11 Quit skwad (Read error: 113 (No route to host))
01:20:05***Saving seen data "./dancer.seen"
01:21:16markunHideo_Kuze: yep, that's me :) unfortunately my PC is also a bastard from time to time..
01:23:06amiconnHmm, am I right in thinking that pointers are usually the same size as longs?
01:24:01amiconni.e. sizeof(* <any_type>) == sizeof(long) for most architectures?
01:24:34amiconnsizeof(* <any_type>) == sizeof(int) holds for 32bit processors, but neither for 16bit nor for 64bit
01:25:04Paul_The_NerdIt seems likely that it'll be the same as a long for most architectures, I think.
01:25:25*amiconn is tackling the sim issues on amd64
01:25:56 Quit ashridah ("Leaving")
01:26:30Paul_The_NerdMaking good progress at least?
01:33:32 Quit Moos (Read error: 110 (Connection timed out))
01:51:06muesli__amiconn your wife is a bug seeker? ;-)
01:51:15amiconnMy sister...
01:51:27muesli__ah ok ;)
01:53:03 Quit xmixahlx ("blah blah blah")
01:54:02CassandraWas she bitten by a moose once?
01:57:02Shadowarrior13No, really!
01:57:38Shadowarrior13Sweden is nice this time of year. See the lakes.
01:57:45Shadowarrior13The wonderful telephone system.
02:07:43 Join virtualball2 [0] (
02:11:24virtualball2linuxstb, what are the files i need to put in my linux partition?
02:11:40virtualball2i know the iPL userland but what was the bin file?
02:12:27Paul_The_NerdThe bin file is the ipodlinux patched uclinux kernel. It should already have a .bin extension, and you just rename it to linux.bin and put it in the Fat32 partition
02:12:50virtualball2ok and i put the userland in the linux partition?
02:13:12virtualball2then how would i run linux then? Would i still have rockbox installed?
02:13:48Paul_The_NerdWhen you hold down Play while booting the iPod, it tells it to load linux.bin instead of rockbox.ipod (assuming you still have the rockbox bootloader installed)
02:14:14virtualball2ok cool
02:14:34virtualball2Thanks Paul and do i need to reformat if i used newfs_msdos?
02:15:00Paul_The_NerdHonestly, dunno.
02:15:18Paul_The_NerdProbably not
02:15:28virtualball2ok thanks anyways, once again, you guys are really nice about questions
02:15:54Paul_The_NerdAssuming the first partition is the non-FSed one, the second is formatted Fat32, and the third is ext2/ext3, you should be fine
02:16:36 Nick vmx_ is now known as vmx (
02:16:39 Quit imphasing ("Have mercy on my soul..")
02:16:44virtualball2ya can i paste my partition map[?
02:17:22 Join midkay [0] (n=midkay@
02:17:28Paul_The_NerdDo you not know what they're formatted as?
02:17:31 Join egotrippen [0] (
02:17:46 Quit midkay (Read error: 104 (Connection reset by peer))
02:17:50Paul_The_NerdThen it should all be fine.
02:18:00virtualball2i just dont know about my first one, it is 27.9 GB and so is the second one :\
02:18:12Paul_The_NerdThe first one shouldn't be 27.9gb
02:18:43virtualball2...well thats weird lol
02:18:44Paul_The_NerdBasically, it should be ~40 or 80mb (depending on which iPod you have), then Lots and lots of fat 32, then some small amount of ext2 (depending on how much you want. I made mine 150mb)
02:19:01virtualball2but theres an asterik next to it
02:20:57virtualball2well w/e im gunna try it, and one last thing, i type things like this "cp bin /dev/disk2s3" and it is done that weird?
02:20:59Paul_The_NerdI don't know. I'm not that good with partition stuff. I had all kinds of problems trying to get my iPod properly partitioned to run Linux when I tried it.
02:21:22Paul_The_NerdAnd I don't know how long that should take at all
02:21:31virtualball2mkay then nvm
02:21:38virtualball2thanks for your help
02:21:56Paul_The_NerdWell, for one thing, if you try formatting just one partition on an ipod in windows using the built in windows format (at least with the Nano) it destroys the partition table and consolidate the whole 'disk' into one partition.
02:22:09Paul_The_NerdThat caused me a few issues when I was trying to format the Fat32 portion
02:22:21virtualball2basically the same with mac
02:22:30virtualball2except HFS+ :P
02:23:45virtualball2im pretty mad at iPL :P I used to think rockbox was the sucky one and iPL was good cuz it was sooo easy, but its like it has turned its back on hehe, so im pretty much rootin for RB now :D
02:25:16Paul_The_NerdThey're pretty different anyway.
02:26:19virtualball2ya i guess, but not so much, one is basically for music the other is for games
02:26:28virtualball2but both basically have the exact same :-P
02:26:50 Join perldiver [0] (
02:28:41 Join JdG|uni [0] (
02:29:00JdG|unihey all
02:29:09JdG|uniuni sux :'( im soo bored here
02:30:36JdG|unieveryone gone to bed already have they?
02:30:47JdG|unidamn eurpopeans and yanks
02:31:44egotrippenthe Yanks should be up, it's only 8:30
02:31:49egotrippenLost hasn't even aired yet :)
02:32:14Paul_The_Nerd7:30 here.
02:32:30egotrippeni'm on the east coast
02:32:33 Quit muesli__ (Read error: 110 (Connection timed out))
02:32:35Paul_The_NerdI figured
02:33:02egotrippen's what america gets, having three time zones in it
02:33:17Paul_The_NerdAnd Arizona
02:33:41Paul_The_NerdArizona and their "We're too GOOD for your 'daylight savings time'"
02:33:43egotrippeni'm forgetting hawaii, right? and i don't know about arizona
02:33:54egotrippenthey got the right idea, DST is a pain
02:34:05Paul_The_NerdPacific, Mountain, Central, and Atlantic.
02:34:23JdG|uniyup, we r in DST also, i hate it so much.. why the ell do we need to the sun to be still up at 8.30pm ?
02:34:56egotrippenyeah - i'd rather have to wake up at the same real time every day than suddenly jump to match the sun
02:35:10Paul_The_NerdMan, is it so hard to just say "Alright, we want to save on our power bills, so starting next week, you all come in to work one hour early and leave one hour early" and NOT change the clocks?
02:35:12egotrippenthey should make it gradual if they're gonna change it at all
02:35:21virtualball2how would i move files to a hidden partition? Do i need to boot linux (Ubuntu is basically hell on eart−−well Mac lol
02:36:08Paul_The_NerdWell, if MacOS doesn't recognize ext2 or ext3, yeah, you'd probably have to boot it.
02:36:36egotrippencan ubuntu use a bootdisk? that's the wrong term
02:36:50Paul_The_NerdI do not know.
02:36:58egotrippenwhere you literally run the distro off of the CD, without installing anything
02:37:18 Join tvelocity [0] (
02:37:31JdG|uniubutunu has a livecd
02:37:42egotrippenthat's a great idea
02:37:43virtualball2egotrippen, yes i use the live
02:39:02egotrippeni miss Linux... i messed around with it for a few months before i got a laptop
02:39:25egotrippeni couldn't figure out a bunch of the laptop-y things, though, like touchpad and tv-out recognition
02:39:43egotrippengot it going pretty well on the desktop, though
02:46:41 Quit quobl (Remote closed the connection)
02:46:57 Join quobl [0] (n=quobl@tor/session/x-b4828d9ded90f3bb)
02:48:47 Quit JdG|uni ("CGI:IRC (EOF)")
02:55:54 Quit egotrippen ("CGI:IRC")
03:03:21 Quit Steven__ (Read error: 110 (Connection timed out))
03:06:23 Quit tvelocity ("Ex-Chat")
03:09:00 Join imphasing [0] (
03:10:49 Quit actionshrimp (Read error: 110 (Connection timed out))
03:13:06 Join sw13 [0] (
03:13:40 Join Lohnny [0] (
03:14:30 Quit BHSPitLappy (Read error: 104 (Connection reset by peer))
03:16:38 Join BHSPitLappy [0] (
03:20:09***Saving seen data "./dancer.seen"
03:20:37 Quit virtualball2 ("virtualball2 has no reason")
03:24:15 Quit Shadowarrior13 (Read error: 110 (Connection timed out))
03:24:25 Nick sw13 is now known as Shadowarrior13 (
03:28:21 Join nave7693 [0] (
03:31:09 Quit ashay (Read error: 110 (Connection timed out))
03:33:16 Part nave7693
03:51:41 Quit ts-x ("CGI:IRC (EOF)")
04:03:02 Quit Daishi ("Client exiting...")
04:12:02 Quit imphasing (Read error: 104 (Connection reset by peer))
04:14:46 Quit Hideo_Kuze ("off to lala land")
04:27:23 Quit mikearthur (Read error: 104 (Connection reset by peer))
04:38:09 Join nave7693 [0] (
04:38:10 Quit Shadowarrior13 ()
04:39:35 Join midkay [0] (n=midkay@
04:52:34 Part picohertz
05:01:54 Join Rob2222_ [0] (
05:04:28 Nick Lohnny is now known as ashay (
05:14:01 Quit Rob2222_ (Read error: 104 (Connection reset by peer))
05:16:47 Join Deviant [0] (
05:20:10 Quit Rob2222 (Read error: 110 (Connection timed out))
05:20:12***Saving seen data "./dancer.seen"
05:25:34 Quit Deviant ()
05:40:06 Quit perldiver (Read error: 110 (Connection timed out))
05:45:15 Part nave7693
05:47:14 Quit golf7 (Read error: 104 (Connection reset by peer))
05:52:47 Join StrathAFK [0] (
05:56:02 Join perldiver [0] (
05:59:14 Nick warewolf_ is now known as warewolf (
06:08:54 Quit Strath (Read error: 110 (Connection timed out))
06:10:46 Join Jungti1234 [0] (n=jungti12@
06:22:44Jungti1234hi all
06:38:40 Join nave7693 [0] (
06:38:50 Part nave7693
06:39:27 Part segphault
06:42:52 Quit RotAtoR ("zzzzzzzz")
06:52:21 Join linuxstb_ [0] (
06:52:22 Quit linuxstb (Read error: 113 (No route to host))
06:59:59 Nick Aditya_ is now known as Aditya|Zzzz (
07:00:34 Quit SereR0kR ("XChat Aqua")
07:07:41 Join nave7693 [0] (
07:07:53 Part nave7693
07:11:25 Join B4gder [0] (
07:20:14***Saving seen data "./dancer.seen"
07:26:21 Quit TCK ("well, if you say so.")
07:27:26JdGordonwhy cant we edit our own comments on fyspray? is that a fs limitation or its just not setup?
07:29:38B4gderit's the way it works, we haven't told it anything special to do so
07:29:50 Join egotrippen [0] (
07:29:51JdGordonits annoying.. but ok
07:30:16JdGordonanyway, i updated the search in pl pacth again...
07:31:04B4gderyes, I noticed
07:39:15Jungti1234 - wow...
07:44:40 Join ender` [0] (i=ychat@
07:45:07JdGordonJungti1234: what the heck?
07:51:29Jungti1234hehe, iDJ.
07:52:29 Join LinusN [0] (
07:52:37 Part LinusN
07:53:06 Join LinusN [0] (
08:06:58 Join amiconn_ [0] (
08:13:26 Quit Jungti1234 ()
08:13:34 Quit quobl ("Leaving")
08:14:03 Quit thegeek ()
08:17:21 Join thegeek [0] (
08:17:49 Quit egotrippen ("CGI:IRC (EOF)")
08:22:53 Join Siku [0] (
08:24:29 Quit amiconn (Read error: 110 (Connection timed out))
08:24:29 Nick amiconn_ is now known as amiconn (
08:29:36 Join needleboy [0] (
08:39:41 Join DJ_Dooms_Day [0] (
08:41:31 Join muesli__ [0] (n=muesli_t@
08:43:54 Join JonasNZ [0] (n=jbergler@unaffiliated/jonasnz)
08:44:21B4gderJdGordon: in the search playlist patch: why do you use button_get_w_tmo() when you don't do anything fun when it times out?
08:44:52B4gderand a little suggestion, use sizeof(search_str) instead of hard-coding '32' for the size of the buffer
08:45:02 Quit perldiver (Read error: 110 (Connection timed out))
08:45:18JonasNZwow, i want to congratulate this project, ive been trying for ages with ipl to get flac going, i find this and it just works :D
08:45:37B4gderJonasNZ: yay ;-)
08:46:25JonasNZthere are a couple of questions i have, is there a way to browse the music in certain folders like the ipod does it normally, by artist, or album etc?
08:47:00LinusNJonasNZ: there is a database mode
08:47:10LinusNbut it is quite buggy as it is now
08:47:23JdGordonB4gder: do u want me to change those 2 things?
08:48:03 Join webguest48 [0] (
08:48:06JonasNZLinusN: hmm ill have a play
08:48:13 Join nave7693 [0] (
08:48:33B4gderJdGordon: It would make it more or less "final" so why not. I haven't tried it though. And I am not a dir cache user so I don't epect to either...
08:48:42B4gderit'd kill me
08:48:56B4gderI use a 4000+ playlist
08:50:50 Join ender1 [0] (i=ychat@
08:51:32 Quit Paul_The_Nerd ("Leaving.")
08:52:00 Quit webguest48 (Read error: 104 (Connection reset by peer))
09:02:21 Quit kkurbjun (Read error: 110 (Connection timed out))
09:06:56 Quit ender` (Read error: 110 (Connection timed out))
09:07:05amiconnB4gder, LinusN: Is it okay to rely on sizeof pointer == sizeof long for fixing the 64 bit problems,
09:07:20B4gderI think so
09:07:31B4gderI don't know of any arch where that isn't true
09:07:32amiconnor should our inttypes.h rather implement intptr_t and uintptr_t ?
09:08:21 Quit linuxstb_ (Read error: 110 (Connection timed out))
09:08:36 Join linuxstb_ [0] (
09:09:20amiconnB4gder: I can't either, perhaps 8 bit archs?
09:09:32B4gderpossibly yes
09:10:10amiconnA good thing is that this requirement (casting pointers to and from 'long' is also part of the older 'long policy'
09:10:32B4gderah right
09:10:37amiconnSo my fixes would also help the gmini port in case it will ever be picked up again...
09:10:44 Join petur [0] (
09:11:07amiconn..and it also explains why there aren't *that* many of these warnings
09:11:53amiconnI'm down to 23 warnings with my last commit
09:11:59 Quit JonasNZ (Read error: 113 (No route to host))
09:12:05amiconnMost of them are now in playback.c
09:13:31amiconnI also found at least one other place where our code doesn't work correctly even though it gives no warning - the functions to extract bits from / put bits into the binary settings
09:14:30amiconnThe effect is e.g. that I can start a sim without sector3D.bin and it works fine, but if I leave it and start it again, colours are black on black...
09:19:00B4gdermy favourite combo ;-)
09:19:24Bg3rB4gder better white on white
09:20:15***Saving seen data "./dancer.seen"
09:21:41 Join safetydan [0] (
09:21:53JdGordondo i want to slow down the search by adding a button_get_w_to so the search can be stopped?
09:22:14 Quit needleboy ()
09:22:19B4gderah, yes you do
09:22:34B4gdersince it'll take *ages* with dir cache disabled
09:23:11JdGordonwhat is the abolute fastest timeout that would still work?
09:23:25B4gderno timeout at all, just poll for a button
09:23:58B4gderbutton_get(false) even
09:24:00*JdGordon slaps self
09:24:36JdGordonis BUTTON_OFF always defined?
09:24:58Bg3rJdGordon look at firmware/export/button.h
09:25:01B4gdermost probably no key is really certain
09:25:01Bg3rbut i doubt
09:25:14JdGordonok, just using TREE_OFF then
09:25:29 Join yeahx [0] (
09:26:28Bg3ryes, better this way
09:26:29amiconnJdGordon: X5 and Player have no BUTTON_OFF
09:26:34linuxstb_amiconn: I was looking at your 64-bit fixes, and wondering if you should use one of the posix types, such as size_t or off_t.
09:26:49amiconnlinuxstb_: ??
09:27:05linuxstb_size_t seems to be the type used with pointers.
09:27:09linuxstb_e.g. in memcpy
09:27:49amiconnI found several references to special intptr_t and uitptr_t types,
09:28:12amiconn(ints that match the size of a pointer)
09:28:54amiconnsize_t would not always work, as it is unsigned
09:29:35amiconn..and ssize_t is something very different
09:29:49B4gderyes, a signed size_t
09:29:55B4gderalthough not a standard type
09:30:05amiconnMeh, I confused sth
09:30:05linuxstb_Is there a reason you rejected intptr_t ? That would seem perfect.
09:30:18amiconnlinuxstb_: It doesn't seem necessary
09:30:44amiconnsizeof pointer == sizeof long holds for all architectures I know of,
09:31:37amiconnpossibly except 8bit. I know that on Z80 addresses are 16bit, iI'm not sure whether long is 32bit there
09:32:11amiconn..but will we ever see rockbox running on an 8bit CPU?
09:33:24linuxstb_It just seems cleaner somehow to use a specific type when we need to treat pointers as integers, rather than making an assumption.
09:33:51amiconnMaybe, but it also implies other problems
09:34:40amiconnIn case the value is propagated to other variables, it would be necessary to change these as well
09:35:19amiconnlinuxstb_: I have to fix inttypes.h to work correctly on varying archs anyway, perhaps I should add intptr_t and uintptr_t
09:36:28amiconnB4gder: When preprocessing I can only figure out datatype sizes by checking UINT_MAX & friends, or is sizeof() available?
09:36:36 Join Zagor [0] (n=bjst@pdpc/supporter/sustaining/Zagor)
09:36:51B4gderamiconn: nope, no sizeof() in preprocessing
09:37:04B4gderbut you can write a test to run at configure-time
09:37:11linuxstb_The sim should just use the system inttypes.h - does cygwin have it?
09:37:22B4gderI did that for the endian check
09:37:29amiconnThe sim uses our own inttypes.h
09:37:35linuxstb_That's a bug then.
09:37:45amiconn..the same way it uses our own errno.h
09:38:48ZagorI didn't know we had udev and devfs support in rockbox... ;-)
09:38:57linuxstb_Was that one of your recent changes to make our errno.h include /usr/include/errno.h ?
09:39:15B4gderZagor: must be that new linux compatibility layer ;-)
09:40:15 Quit Higgy_ ("Trillian (")
09:40:25 Join Higgy_ [0] (
09:40:37 Quit muesli__ (Read error: 110 (Connection timed out))
09:40:53amiconnlinuxstb_: Btw, from looking at the system includes, inttypes.h isn't the header where the int*_t types are actually defined, the definitions reside in stdint.h
09:41:15amiconninttypes.h includes that and then defines a bunch of other macros
09:41:26markunB4gder: got a reply from him yet?
09:41:32amiconnlinuxstb_: Yes that was one of my recent changes.
09:41:47B4gderbut then I think he's ignoring me already
09:41:54B4gderdue to previous events
09:42:33amiconnIt allows building the sim on an nptl linux system. The #ifdef magic makes it use our own errno for the win32 sim
09:42:47markunB4gder: what that before or after he left rockbox?
09:42:48amiconn..and the sdl sim when building for windows
09:46:02 Part yeahx
09:47:31 Join needleboy [0] (
09:54:04 Quit needleboy ()
09:54:41 Join needleboy [0] (
09:54:43linuxstb_amiconn: I know inttypes.h works at least on Linux, Mac OS X and Cygwin/mingw32 (I've written an application which uses it, and that compiles cleanly on those three platforms) So I think it should be OK to do the same #ifdef magic as errno.h in inttypes.h
09:59:47 Quit nave7693 (Read error: 110 (Connection timed out))
10:10:08Zagorwiki:IpodStatus says rockbox has "clickwheel lag". wasn't this fixed in the button driver?
10:10:25Zagoror isn't the fix good enough?
10:14:09linuxstb_The clickwheel still needs more work, but yes, the lag has been removed. I'll fix the page.
10:27:38 Join muesli__ [0] (n=muesli_t@
10:28:41JdGordondoes any1 know how to add a drive to windows (like a fake drive from a folder on the hdd)? i know its a long shot...
10:29:35LinusNsubst x: <path>
10:29:59merbananJdGordon: or the long way, share the folder and then map it
10:30:22safetydanthe drive manager snap-in should be able to do it as well
10:30:50JdGordonthanx guys... i was expecting some really idiotic code... damn this is gonna be sweet
10:34:06 Join mirak [0] (
10:35:06 Nick ender1 is now known as ender` (i=ychat@
10:48:33 Join ashridah [0] (
10:51:23JdGordondoes the splash function support multiline strings?
10:51:38Bg3rJdGordon splash itself rearranges the lines
10:51:57JdGordonso u can have a multiline msg?
10:54:27safetydanI don't think splash() even looks for \n
10:54:34safetydanit will wrap the string if it thinks it needs to
10:55:21 Join bagawk_ [0] (
10:56:35JdGordonB4gder: i just tested it with dir cache off and its not much slower
10:56:47JdGordonmaybe an extra 2 or 3 sec
10:56:50B4gderI don't understand how though
10:56:59B4gderof course the disk will spin the whole time
11:01:31 Quit bagawk (Read error: 110 (Connection timed out))
11:02:43JdGordon.. and im done... again...
11:05:04 Part LinusN
11:09:13 Quit Bg3r (Read error: 113 (No route to host))
11:14:17safetydanSomeone tested my cut/copy/paste patch on Ondio with a MMC card and it seems to work copy between them. Cool.
11:14:23safetydanNow... should I add a progress bar?
11:15:00JdGordon.. probably...
11:15:31peturand a completely random estimated time, like windows does :)
11:18:56 Nick linuxstb_ is now known as linuxstb (
11:20:18***Saving seen data "./dancer.seen"
11:21:04 Join skwad [0] (
11:21:12safetydanI was thinking of making it themeable so you could have a progress WPS...
11:21:17safetydanWhile Progressing Screen :)
11:21:19amiconnsafetydan: For moving, do you try rename() first and if that fails, resort to copying?
11:21:19skwadHi all
11:21:32safetydanamiconn, no
11:21:56JdGordonsafetydan: its dead simple to use the currne twps.. as long as u fake the id3 info...
11:22:00amiconnThat will be necessary for multivolume support
11:22:28safetydanamiconn, I hadn't got to multivolume support yet
11:22:38safetydanIs there a way to detect if the file is on a different volume first?
11:23:01amiconnNot with the standard file api
11:23:20amiconnThe fat code does know about the volume
11:23:32JdGordonwould any1 like to look at my latest brain-fart.. uh, i mean, patch and commit if its ok?
11:24:05amiconnHmm, when thinking about it, is there a standard errno value (Esomething) to indicate move across filesystem boundaries?
11:24:28B4gderEXDEV oldpath and newpath are not on the same filesystem.
11:25:17safetydanbut rename just sets EIO if fat_rename fails
11:25:21amiconnOkay, so file.c: rename should be made to set that for move attempts across filesystems
11:25:57amiconnRockbox already has EXDEV defined...
11:26:33safetydanshould be easy enough, fat_rename returns -1 in the case of cross-volume renames
11:26:43safetydanother errors have different return codes
11:27:00 Quit gtkspert (Read error: 110 (Connection timed out))
11:27:17amiconnyup, just another if inside the if(rc <0) ...
11:27:44amiconn..put in an #ifdef HAVE_MULTIVOLUME
11:28:22safetydanSo who wants to do it? I can't test it without a multivolume target, but I can certainly write the code.
11:28:54JdGordonmultiple volume as in more than 1 partition on the hdd?
11:29:04amiconnHmm, I have a nice challenge for Slasheri: Make dircache work properly with multivolume and hot mounting/unmounting of volumes.
11:29:44safetydanJdGordon, pretty much
11:29:55peturwill certainly need all this nice work when I ever get usbotg to work...
11:30:21JdGordonill test it out
11:30:31JdGordonassumeing its safe to partition the disk in winxp?
11:31:12safetydanJdGordon, only some targets actually support multiple volumes. By some I think that means just Ondio at the moment
11:31:25JdGordonoh ok
11:31:47skwadwhat is multiple volumes ?
11:31:53amiconnJdGordon: You can use multivolume support for multiple partitions as well,
11:32:21amiconnbut it was mainly developed for Ondio, which has 128MB of builtin flash memory, and an MMC slot
11:32:31safetydanamiconn, does Rockbox currently work on multiple partitions on anything other than Ondio?
11:32:45 Join webguest76 [0] (
11:34:04peturwith usbotg, it could be that a multi card reader gets attached that has n drives of which only one is valid
11:34:21skwadjust a suggestion : wouldn't it be logical to group the wps gallery by lcd sizes and not players ?
11:34:35B4gderskwad: yes, lcd size + color depth
11:35:15skwadcause x5 and h100 have same resolution, and h300 and ipod 4g too
11:35:51B4gderbut x5 is color and h100 is not
11:36:08*JdGordon thinks thats his point
11:37:27skwador just groupping the players which have same lcd size
11:41:25skwadyep B4gder but the x5 users could take the h100 b/w wps ....
11:41:51skwadmaking like you said; lcd size + color depth
11:42:28linuxstbskwad: Yes, I've been thinking the same thing. Headings like "220x176 16-bit color (iriver H300, iPod Color/Photo)"
11:43:03skwadperhaps I coul do it ... it's not very difficult
11:45:06linuxstbAnother heading could perhaps be "Text-only", and the WPS would need to indicate how many lines of text are required - these are dependent on the combination of LCD size and the font the user chooses.
11:46:26safetydanamiconn, this seem okay?
11:46:43safetydanI'm cautious about comitting things in that area since... well it's in the scary firmware directory :)
11:47:23 Quit dpassen1 (Read error: 110 (Connection timed out))
11:48:42amiconnsafetydan: In fact I tested multivolume support on archos player, to see whether it has problems with mixed FAT32 and FAT32
11:49:04amiconnWorked rather well after fixing some bugs.
11:49:22linuxstbsafetydan: Looking at your latest patch (v4), I noticed that in your main copy loop, you are checking bytesread instead of byteswritten after a write.
11:50:50safetydanlinuxstb, good spot
11:51:10amiconnErm of course mixing FAT16 and FAT32
11:51:19amiconnFAT16 was another Ondio first...
11:52:08linuxstbAlso, if the write fails, maybe you should delete the created file.
11:53:47safetydanlinuxstb, good idea. I'll fix that too
11:54:16amiconnsafetydan: Your #fdef #else #endif sequence is wrong.
11:54:36amiconnThe code wouldn't report other errors than EXDEV on multivolume targets
11:56:16safetydanyes, you're right
11:58:22amiconnCheck my suggestion on pastebin
11:58:24safetydanI'll change that to just do both checks
12:00:49safetydanOkay, that's the same thing I did
12:00:54safetydanI'll commit that now then
12:03:40webguest76talking to partition things, anyone now how to fix the Rockbox Infos concerning the HD space left?
12:04:59safetydangah... misspelt "volumes"
12:05:26safetydanat least it's only a debug string
12:06:30amiconnI purposely used 'tried' not 'failed'
12:06:37amiconnBut it's not important either
12:07:04amiconnIt's just a debug string, and debug bui.ds aren't possible on Ondio
12:12:33safetydanwow the builds are fast now
12:12:47safetydanit's almost quicker to commit things to the build server than to compile them locally
12:13:23 Join webguest60 [0] (
12:13:58webguest60webguest76: you can do it from the debug menu.. which player do you have?
12:14:49safetydanwebguest76, what's wrong with the free space count?
12:15:19webguest76that said me wrong free space
12:15:56webguest76exemple: my iriver h1xx said me 2.34 GO, but I reall y have just around 600 mo free
12:16:05webguest60info > debug (keep out!) > view disk info.. now press up 3 times (or until you get a listing of free space).. Press select and it should say "Scanning disk"
12:16:23webguest60That'll recalculate the free space count
12:16:33webguest76uh? didn't know this, let me check
12:16:48webguest76what is the select button btw?
12:16:57webguest60The joystick
12:17:03webguest60Pressing it down.
12:17:10webguest76ok thanks
12:17:14webguest76I'm testing
12:17:51webguest60Maybe that should be possible to do outside the debug screen, as someone mentioned on the mailinglist?
12:18:29safetydanIIRC it's from the FAT structure which Windows (or Linux) is supposed to update when modifying the file system
12:18:35safetydanobviously it doesn't work right
12:18:47webguest76ok, that doesn' automaticly refresh
12:18:59webguest76we need to scan like this manually
12:19:18B4gderor you file a bug report to microsoft :-)
12:20:24webguest60Yes, windows seems utterly broken in this area (as well)
12:20:28webguest76time to go, thanks webguest60 for the info
12:20:38 Quit webguest76 ("CGI:IRC")
12:21:02 Quit webguest60 ("CGI:IRC")
12:28:10 Nick ashridah is now known as Lost-ash (
12:32:14 Quit damaki (Read error: 110 (Connection timed out))
12:43:56JdGordondoes any1 know if it is possible to load the windows icon for a certain file if u know the filetype?
12:44:34Aditya|Zzzzits a shell call
12:44:49JdGordondo u know the name?
12:45:16Aditya|Zzzzwhat language you using?
12:46:04 Join Rob2222 [0] (
12:47:11Aditya|Zzzzwell I dont exactly know how to make shell calls from C but thats your guy
12:47:41JdGordonill have a look
12:47:57Aditya|Zzzzhave you tried google groups?
12:48:17JdGordonnot yet
12:48:31petursince when is C++ or higher required to do win32 api calls?
12:49:03peturJdGordon: check that on msdn (
12:50:20Aditya|ZzzzI didn't say that..
12:50:28Aditya|ZzzzI just didn't know HOW to make shell calls from C
12:50:44Aditya|ZzzzI use C# and C++ a lot and I only have had to use ShGetFileInfo from C# for a file browser
12:50:47preglowmsdn is very helpful for stuff like this
12:50:56preglowuse it all the time myself
12:51:16Aditya|Zzzzoff to uni.
12:51:18Aditya|Zzzzlater taters
12:51:25 Nick Aditya|Zzzz is now known as Aditya|Uni (
12:51:51petur... talks a lot in his sleep ... :P
12:53:28 Join erus` [0] (
12:55:23peturwhoa... 8 seconds / build
12:55:44B4gderdarned neat
12:57:43preglownot bad
12:57:46preglownot bad at all
12:57:49linuxstbHas anyone tried distcc to build Rockbox? e.g. on a home network?
12:58:04preglowpretty good looking build table too
12:58:12B4gderor rather, I build it on the build server with distcc to my homebox
12:58:33preglowhow many build boxes have we got?
12:58:39B4gderwhich was too slow network for distcc to be fun
12:58:50B4gderpreglow: 5 the last time I checked
12:59:16preglowvery nice
12:59:21preglowsolved that problem
13:01:14B4gdera few more servers and we'll be faster than what most people build locally ;-)
13:01:36CassandraMine will be coming on line in the next week or two, hopefully.
13:01:48peturbah, it already is... A LOT :(
13:02:53peturJdGordon: you had colinux running, right?
13:03:09JdGordonye, havnt got it setup properly yet tho
13:03:34peturmine runs, but refuses to know the dns servers I give it :(
13:03:52JdGordonhave u got networking at all?
13:03:59B4gderyou cygwin guys tried using ccache?
13:04:13B4gderI see it is available as a package
13:04:33B4gderinstall it, run configure −−ccache
13:04:39B4gdercompare build speed
13:04:54 Quit XavierGr (Read error: 110 (Connection timed out))
13:04:57*JdGordon doing
13:04:59peturJdGordon: I just tried pining my windows box, maybe I should try an ip address on the net - good point
13:05:41petursomebody said ccache made it worse
13:06:57safetydanB4gder, I think amiconn tried it. His results were something like 6 minutes without ccache, 8 minutes the first time with ccache and 7 minutes the second time with cccahe
13:07:10B4gderhaha, how crappy
13:07:37CassandraWhat is ccache?
13:07:50B4gderit caches compiler results
13:07:59B4gderand gets them instead of rebuilding
13:08:04B4gderif all things remain the same
13:08:09CassandraIt that the same or different to pch?
13:08:36B4gderI don't know pch
13:08:43CassandraPre-compiled headers.
13:08:51B4gderthis caches object files
13:09:05CassandraNice feature in mingw at least that cuts down on compile time a lot.
13:09:16CassandraNot sure if core gcc supports it.
13:09:38B4gderso if you build the same unmodified source again, it just gets the cached object file instead of recompiling
13:10:07B4gderbut of course it needs to preprocess first
13:10:17B4gderwhich I guess is what takes lots of the time on cygwin
13:11:34linuxstbDo you know how much ccache speeds up the builds on the build server?
13:11:42 Part erus` ("Leaving")
13:11:59B4gdernot exactly, but I remember it cut down the build time by a signigicant amount when I added it
13:12:14B4gderlike 20-30% or so
13:13:11CassandraReally we ought to be using pre-compiled headers if bog standard gcc supports it. It should increase build speed quite dramatically.
13:16:15CassandraI notice that doesn't seem to indicate how to generate them.
13:16:34B4gdernot with details, it just says "To create a precompiled header file, simply compile it as you would any other file, if necessary using the -x option to make the driver treat it as a C or C++ header file."
13:17:49B4gderbut that doesn't make sense to me
13:18:07MikachuOnly one precompiled header can be used in a particular compilation.
13:18:42B4gderweird limitation
13:18:47 Quit JdGordon ("Miranda IM! Smaller, Faster, Easier.")
13:19:07MikachuSome other command-line options starting with -f, -p, or -O must be defined in the same way as when the precompiled header was generated. At present, it's not clear which options are safe to change and which are not
13:20:19amiconnB4gder: safetydan is right, I tried ccache on cygwin and it is indeed counter-productive
13:20:22***Saving seen data "./dancer.seen"
13:21:33 Join gtkspert [0] (
13:22:03amiconnThe cygwin slowness seems to be filesystem related
13:22:41 Quit hyarion ("leaving")
13:24:27amiconnBuilding on my linux VM is almost 3 times as fast as on the host w/ cygwin, but guess what happened when I tried building within the linux VM but directly on the hgfs shared folder?
13:25:31amiconnAlmost as slow as under cygwin...
13:27:38amiconnEither it's the virus scanner or ntfs itself
13:42:44preglowfile system doesn't matter
13:42:48preglowit's just as bad on fat32
13:44:02preglowi've never used an active virus scanner, and it was just as bad on my pc
13:46:24 Nick Lost-ash is now known as ashridah (
13:46:43preglow <- nice to see someone's talking about me
13:48:19preglowwhich reminds me, anyone know where the latest wma fixed point port is? i can only seem to find what was posted on our forum
13:50:45B4gderthere was a fixed point port?
13:51:47 Join webguest31 [0] (
13:55:38preglowwell, yeah
13:55:49preglowa half-finished one
14:00:46ashridahi seem to recall someone mentioning they were going to look into it shortly after the H3xx started playing audio or something, that's all i've seen
14:04:39preglowa lot of people have said they'll look at it
14:04:52preglowi've never heard from anyone of them again, even those i said i'd help
14:10:37preglowwhat, you have to _check_ for size_t?
14:10:49preglowi thought that was a built-in type
14:13:47markunpreglow: I'm not sure. Maybe my commit isn't the best solution
14:15:15markunWithout it I get: include/sys/types.h:46: error: conflicting types for 'size_t'
14:15:32markun/usr/include/stddef.h:55: error: previous declaration of 'size_t' was here
14:15:32B4gdersize_t isn't built-in
14:15:36 Quit DJ_Dooms_Day ("Trillian (")
14:15:40B4gderonly char, int, long, short etc are
14:18:30preglowit should be :>
14:21:02 Quit Paprica (Read error: 104 (Connection reset by peer))
14:24:00 Join DJ_Dooms_Day [0] (
14:26:53 Join linuxstb_ [0] (
14:28:19 Join saa[b_r]ider [0] (n=saab_rid@
14:30:19 Join webguest33 [0] (
14:30:41linuxstb_preglow: Are you still looking for the wma code?
14:31:32 Join SereR0kR [0] (
14:33:05 Quit webguest31 ("CGI:IRC (Ping timeout)")
14:33:43linuxstb_preglow: I can't find the forum thread, but I think this is the zip file the poster made available:
14:35:26linuxstb_But looking again at it now, it doesn't seem very useful - it still uses float, and it still links against libm.
14:36:59 Quit SereR0kR ("XChat Aqua")
14:38:39 Join SereR0kR [0] (
14:38:41linuxstb_I seem to be talking to myself, but I've also found the forum topic -
14:39:54peturoh no, I listened... not that it helped much :)
14:40:16markunlinuxstb_: I also listened and found the same thread
14:40:32 Join Jungti1234 [0] (n=jungti12@
14:41:15 Quit B4gder ("time to say moo")
14:41:44Jungti1234hello all
14:46:09linuxstb_markun: Thanks for that link.
14:46:28lostlogicman, these 5 minute builds are phenominally nice, keep catching me off guard with their speed.
14:47:51linuxstb_There is some weird code in there - e.g. the initialisation of lookup tables using 256 assignment statements (table[0]=2; table[1]=1; etc...), or 256 if statements instead of a lookup table.
14:48:30 Join tvelocity [0] (
14:48:32linuxstb_It's like a lesson in how to write inefficient code...
14:48:47peturno wonder it required 100MHz
14:49:11markunlinuxstb_: was it already there in the ffmpeg source?
14:49:14linuxstb_But it's definitely got potential - those things are easily fixed.
14:49:21linuxstb_markun: I highly doubt it.
14:50:15markunlinuxstb_: well, fix it! :)
14:50:49linuxstb_Of course - those bits are the least of our worries.
14:53:55markunIt would be so nice if the codecs could use the same fft and mdct code
14:55:48 Join XavierGr [0] (
14:56:01linuxstb_I'm very tempted to give this wma decoder a go - to at least get it working (at whatever speed it can manage) under Rockbox.
14:56:13 Quit Jungti1234 ()
14:56:54webguest33go go go!
14:57:28XavierGryeah linuxstb_!!
14:57:38XavierGrMake the big leap :D
14:58:14webguest33Once wma runs on ipod, expect a nastygram from Apple
14:58:58markunDoes MS have problems with a GPL WMA decoder?
14:59:27webguest33I don't think there's much they can do
14:59:38webguest33except make the next codec harder to RE
15:00:25linuxstb_markun: The WMA decoder in ffmpeg isn't new. But I guess it has a relatively low profile.
15:01:04XavierGris there any chance of realtime?
15:01:23linuxstb_Of course - the original iriver firmware plays wma in realtime.
15:01:44linuxstb_But how much work it's going to need is another question.
15:01:52 Join mikearthur [0] (
15:01:56XavierGryeah but the ffmpeg decoder may be not so optimal
15:04:03safetydanWasn't the WMV spec submitted to a standards board? Wouldn't that also include the audio component?
15:04:32amiconnmarkun: (size_t and stuff for bsd) Maybe your solution isn't the best, but if it works... It goes in the same line as my errno.h fixes for nptl
15:04:58linuxstb_safetydan: That all depends what audio format was proposed to go with it. But yes, Microsoft want broadcasters to use their codecs, so they've had to submit them for standardisation.
15:05:12merbanansafetydan: only video
15:06:03merbananlinuxstb: what is needed is a fixedpoint implementation of the requantization of the mdct coeffs
15:06:14linuxstb_webguest33 - I don't think Apple would care - we're never going to support DRM'd WMA.
15:06:49webguest33linuxstb_: I was a) more thinking of the principle of the thing: "their format on our player! An outrage!" b) kidding
15:07:22merbananwebguest33: their new codec is not hard at all to RE, linspire released decoders with debug info in them
15:08:18webguest33merbanan: then they have failed, I guess
15:08:48merbananwebguest33: ps file of the call tree for wmapro init
15:13:04 Quit DJ_Dooms_Day (Read error: 110 (Connection timed out))
15:14:11 Join Matze [0] (
15:17:07 Join Daishi [0] (
15:18:41preglowlinuxstb_: i think you should! :)
15:19:10linuxstb_I only have one wma file though....
15:19:12preglowthe integration job is what i'm not looking forward to
15:19:14 Quit SereR0kR ("XChat Aqua")
15:19:17preglowi only have one wma file too...
15:19:22 Join _FireFly_ [0] (
15:19:40webguest33I'm sure microsoft will happily provide a few to show off how superior the format is
15:20:26***Saving seen data "./dancer.seen"
15:20:58linuxstb_I don't mind the integration job - as long as I can find some documentation on the asf format (or whatever wma is contained in...)
15:22:40amiconnEncoding to wma is not a problem with a windows box
15:23:21linuxstb_I'm sure there will be volunteers to give preglow and me sample files...
15:24:07webguest33those are short samples of different bitrates I think
15:24:15webguest33probably nice for testing
15:24:30muesli__linuxstb need a wma?
15:25:42linuxstb_muesli__: Not yet :)
15:25:49muesli__oki ;-)
15:26:05muesli__just to give u a perfect jump start ;)
15:26:38preglowlinuxstb_: i think contacting the wma port guy might also be clever, i think he's done some more work on it
15:26:41merbananlinuxstb: here are some samples for you
15:26:52 Quit kernelsensei (Read error: 104 (Connection reset by peer))
15:27:01 Join kernelsensei [0] (n=boris@gentoo/developer/kernelsensei)
15:28:35 Quit tvelocity ("Ex-Chat")
15:32:51preglowlinuxstb_: i get the impression there's a lot of unneeded code here
15:32:57 Join Paprica [0] (
15:34:56preglowaaghahahaha, half of the wmadec.c code is bloody lookup tables in disguise
15:35:05linuxstb_I noticed that :)
15:35:15 Join dpassen1 [0] (n=dpassen1@resnet-236-163.resnet.UMBC.EDU)
15:35:27preglowmy god
15:35:55linuxstb_I especially love the internal_05_power_ten() function
15:36:29preglowyes, that in particular struck a chord with me
15:36:47preglowthe programmer has to be from the moon or something"> <- best song ever.
15:37:32preglowlinuxstb_: the readme.txt says the ffmpeg code is taken from a 2004 branch, do you think much has changed since then?
15:38:24merbananpreglow: the most important fix commited was a framing issue
15:38:33preglowyeah, it has been changed
15:38:37preglowcouple of bug fixes
15:38:47preglownothing big
15:39:18 Join ts-x [0] (
15:39:51preglowhuuge parts of dsputil.c are uneeded
15:39:58preglowcpp, at that
15:40:16linuxstb_Yes, most of the time I spent on porting the FLAC decoder was deleting unused code.
15:40:40preglowbut yeah, in the near future, when we change to svn, the libffmpegFLAC dir should probably be renamed
15:41:09*preglow pokes Bagder in the ribs
15:41:40linuxstb_I can see more and more of ffmpeg being useful - especially if someone ever implements video.
15:41:42merbananpreglow: dsputil is unneeded, the mdct code from ffmpeg isn't used
15:42:04preglowi couldn only see dct code in dsputil anyway
15:42:09preglowdidn't look too close
15:42:48merbananit also gives you mdct routines
15:43:13preglownow, what the hell does simple_idct do here...
15:43:31merbananI guess to make dsputil happy
15:44:03preglowand then we've got two different implementations of the mdct
15:44:06preglowone needing an fft lib
15:50:38*linuxstb_ spots an asf file format document on Microsoft's website:
15:51:51amiconnlostlogic: Why did you remove the global_settings.runtime = 0; lines from apps/gui/statusbar.c ?
15:52:24amiconnThe runtime should be forced to 0 whenever a charger is connected
15:53:19amiconnBtw, this didn't work 100% on H300. It did work on the archos jukeboxes
15:53:55amiconnOn H300 it was only forced to 0 while actually charging, not when charging was complete
15:56:41linuxstb_merbanan: Do you know if that is still accurate? i.e. is v1.0 undocumented but used, and v2.0 documented but unused?
15:58:47 Join Thus0 [0] (n=Thus0@
15:59:30merbananlinuxstb_: not sure but new files still demux correctly with ffmpeg and mplayer
15:59:59merbanantry to correlate any fatal changes
16:07:41 Quit ashridah ("Leaving")
16:07:52 Join amiconn_ [0] (
16:10:30 Quit webguest33 ("CGI:IRC (EOF)")
16:16:26crashdany devs aboot?
16:17:21crashdhey preglow, ltns :)
16:17:25 Quit amiconn (Nick collision from services.)
16:17:26 Nick amiconn_ is now known as amiconn (
16:17:30preglowindeed, hehe
16:17:43crashdi wanted to ask about the bootloader actuall
16:17:59preglowgo ahead
16:18:05crashdon the other architectures rockbox has been ported too, what happens if the bootloader doesn't quite work
16:18:11crashddoes it, effectively, just brick the player?
16:18:17crashdi know the ipod has a special emergency mode
16:19:27preglowyes, brick
16:19:31preglowat least for iriver players
16:19:40preglowin the case of iaudio, you have something much like on ipod
16:19:55preglowbut yes, for iriver players, a brick is very probably if you break the bootloader
16:19:57crashdas im still within my 1year statutory period here in the uk with this iriver
16:20:05crashdso if i brick it, i can just get a new one from the store
16:20:18peturno you can't
16:20:30crashdive done it once alreayd
16:20:32crashdso im guessing i can
16:20:32preglowweeell, i don't exactly know uk laws, but i think flashing it with unauthorised firmware kind of voids that...
16:20:39crashdyeah, that's the idea
16:20:41safetydanyou have to demonstrate it was unfit for sale usually
16:20:43crashdbut they wouldnt know, as it's just a store
16:20:53crashdwell, when ive replaced hardware that's broken before
16:20:56muesli__nobody bricked his iriver so would be bloody nr.1 ;-p
16:21:05crashdespecially in high street chains, like Dixons
16:21:08crashdthey just exchange the unit
16:21:27peturthere's already a story of a rockboxed iriver gone bad, and no more warranty
16:21:42peturit wasn't rockbox that bricked it
16:21:52peturbut they still refused
16:22:03crashdwhen i left my h10 in a red hot car over the summer for a few days, the battery broke, and they just replaced the whole unit in the store i bought it from
16:22:07muesli__petur yepp, but the bootlaoder itself is safe
16:23:13peturthe guy's hdd broke, he returned it (couldn't reflash) and they said he void his warranty
16:23:23crashdwhere was this?
16:23:41peturread in the forum (here or MR, can't remember)
16:23:47crashdahh, fair deal
16:24:27muesli__petur that sounds logical...he shouldve built in a replacement hff before returning
16:24:27 Quit DT291 (Read error: 104 (Connection reset by peer))
16:24:37muesli__foolish if you ask me
16:24:48safetydanI don't think it would boot so he couldn't reflash it
16:24:49 Join DreamTactix291 [0] (
16:25:20muesli__get another drive and reflash
16:25:25 Join SereR0kR [0] (
16:25:56safetydanerr... he wanted the drive replaced under warranty
16:26:53muesli__organizations are always looking to avoid their warranty service
16:28:38safetydanon another topic... web services are annoying :(
16:28:39peturand a new report just came in: (happened in the UK)
16:33:03preglowperhaps we should make the rockbox bootloader not spam the screen so much :-)
16:35:59petursupport usbotg and flashing from the bootloader, so we can attach a memorystick and reflash from it :D
16:40:50 Join Menollo [0] (
16:42:40 Join Farpenoodle [0] (n=solo84@
16:42:56 Quit Menollo (Client Quit)
16:45:50muesli__petur lol
16:46:15muesli__that would be awesome
16:46:37peturfor bootloader v39 probably :P
16:46:45muesli__connecting a 5 1/4 floppy ;)
16:47:21 Quit vmx (Read error: 110 (Connection timed out))
16:47:30 Quit markun ("leaving")
16:47:48 Join vmx [0] (
16:48:45*petur looks at the address of markun and notes how the university of Twente thinks about its students
16:50:22Xerionhow so? :P
16:51:19Xerionwe choose our own hostnames :)
16:53:39 Quit needleboy ()
16:55:36 Join TheOneWhoKnows [0] (
16:55:38 Join markun [0] (
16:56:16 Join imphasing [0] (
16:58:27markunXerion: quite a lot of people from Twente who come here :)
16:59:22 Quit TheOneWhoKnows (Client Quit)
17:00:45 Quit Thus0 (Read error: 104 (Connection reset by peer))
17:00:57markunXerion: you, me, HCl and Pieter (not here right now)
17:01:16 Join Thus0 [0] (n=Thus0@
17:08:17 Join DJ_Dooms_Day [0] (
17:16:39 Join Paul_The_Nerd [0] (
17:20:28***Saving seen data "./dancer.seen"
17:26:14 Join Mongey|study [0] (
17:27:26Mongey|studyanyone speak french here?
17:28:12 Join needleboy [0] (
17:33:29 Quit linuxstb (Read error: 110 (Connection timed out))
17:33:40 Join linuxstb [0] (
17:34:44peturMongey|study: un peu ;)
17:35:16 Quit Thus0 ("Leaving")
17:36:41 Join IcyStorM [0] (
17:36:54Mongey|studycan u help moi?
17:37:23peturwith what?
17:37:51Mongey|studynvm its ok now
17:37:55Mongey|studythanks anyway
17:39:11 Quit Mongey|study ()
17:48:26 Join Farpnut [0] (n=solo84@
17:48:45 Quit safetydan ("Leaving")
17:50:50 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
17:55:22 Quit ts-x ("CGI:IRC (Ping timeout)")
18:01:46 Quit mirak (Remote closed the connection)
18:03:58 Join mirak [0] (
18:06:00 Quit petur ("damn, another evening without rockbox coding :(")
18:09:14 Quit Farpenoodle (Read error: 110 (Connection timed out))
18:12:01 Join safetydan [0] (
18:12:26safetydanamiconn, I just booted in to windows to try and reproduce the SDL sim "window dragging while playing sound"
18:12:33safetydanworks fine for me (un)fortunately
18:12:43safetydanSound plays no matter what I do to the sim window
18:15:50Paul_The_Nerdsafetydan: Have you tried holding an extended drag, and like... shaking it all over the place, for lack of a better way to describe it?
18:16:10Paul_The_NerdI haven't done it recently, but when I first heard about it causing issues, I had to go that far cause problems
18:18:26 Join mozetti [0] (
18:30:06safetydanPaul_The_Nerd, that's what I spent about five minutes doing :)
18:30:16safetydanMy wife was looking at me funny the whole time
18:30:22safetydan"I'm just debugging honey"
18:31:22Paul_The_Nerdsafetydan: Well, I haven't tried that recently, so it's very possible the problem's gone anyway. :) Man, I'm glad I don't have people to look over my shoulder when I do things like that
18:46:00 Quit IcyStorM (Read error: 110 (Connection timed out))
18:55:05 Join damaki [0] (
18:59:19Papricahow can i jumo a line when i read from file?
18:59:33 Join Nico_P [0] (
19:01:46 Join bluey [0] (
19:04:52 Quit ghode|afk (Read error: 104 (Connection reset by peer))
19:05:30 Join ts-x [0] (
19:08:24linuxstb_Paprica: What do you mean by "jumo" ?
19:08:42Papricago to the next line
19:10:20linuxstb_You have to read the whole file in order to know where the lines are - so you can't jump lines.
19:11:12 Quit Paul_The_Nerd ("Leaving.")
19:11:17 Quit imphasing (Read error: 104 (Connection reset by peer))
19:12:08 Quit Xerion (" ")
19:14:47 Quit DrumRBoy320|Away (Read error: 110 (Connection timed out))
19:15:29 Quit linuxstb_ ("Client Exiting")
19:20:32***Saving seen data "./dancer.seen"
19:21:23 Join SuperSnout [0] (
19:22:50 Join Xerion [0] (
19:22:53 Join egotrippen [0] (
19:22:55SuperSnoutHey.. I'm running rockbox on my H300.. are there any plugins extra plugins that people could recommed? Also.. where would you get other rockdoom .wad files to play the other versions.. and if you add those files to the doom folder.. how do you select which file rockdoom runs from?
19:23:06SuperSnout* -plugins
19:23:41egotrippeni'm getting an error message re 'Playlist Control File' on the H300 CVS
19:23:58egotrippenmost of the times when i try to resume a playlist, and sometimes when i try to add something to a playlist, it won't let me
19:24:05SuperSnoutegotrippen: still not sorted it out? =(
19:24:36egotrippeni don't think i've reported this problem yet... it just began happening a couple weeks ago, i think
19:24:48egotrippenit's starting to really bother me, though
19:24:52egotrippeni like playlist resume
19:25:22egotrippeni don't think we can talk about where to get .wad files unless they're open source, but to play them
19:25:32egotrippenjust rename them and put them where the one you have is
19:25:41egotrippeneither to doom.wad or doom2.wad, i think
19:27:01BHSPitLappydoom2.wad is commercial, but a lot of the other wad's you can get are freeware
19:27:15egotrippenyeah. freeware ones we can point you to
19:27:26egotrippenbut not like quake or hexen
19:27:32BHSPitLappygo to the forums at and search for "wad"
19:27:37SuperSnoutegotrippen: must have been someone else last night who had the same problem
19:27:57egotrippenthey get any info on it? i haven't heard anyone else with this problem
19:28:22SuperSnoutegotrippen: can't really remember.. sorry..
19:28:33egotrippeneh, s'alright
19:28:53SuperSnoutBHSPitLappy: thanks..
19:29:15 Join |Beowulf| [0] (
19:30:24SuperSnoutBHSPitLappy: "No page with this exact title exists, trying full text search."
19:30:42BHSPitLappySuperSnout: I said the forums, not the wiki
19:31:05BHSPitLappy" go to the forums at"
19:31:08SuperSnoutoh.. seperate.. ok
19:31:20SuperSnoutsorry.. i thought it was just a straight link..
19:31:45 Quit mikearthur (Read error: 104 (Connection reset by peer))
19:34:54SuperSnoutare they actually shareware BHS?
19:35:16SuperSnoutthey have doom2.wad..
19:35:23 Part mozetti
19:35:49 Join mikearthur [0] (
19:36:04BHSPitLappySuperSnout: no, they don't
19:37:07SuperSnoutBHSPitLappy: *sigh* i'm awfully sorry.. i'm seem to be a bit blind today
19:44:35 Join TCK- [0] (
19:45:32 Join damaki_ [0] (
19:46:11 Join Lear [0] (
19:48:07SuperSnouti actually have the game doom.. bought it a long time ago.. do you just take the wad files of the cd?
19:50:36SuperSnouti have ultimate doom.. would it run?
19:52:23egotrippenit should
19:52:47egotrippenjust rename the wad to what the one you have already is called
19:55:05 Quit mikearthur (Read error: 104 (Connection reset by peer))
19:55:13 Join mikearthur [0] (
19:58:45 Join skwad_ [0] (
19:58:45 Quit skwad (Read error: 104 (Connection reset by peer))
20:00:15 Quit damaki (Read error: 110 (Connection timed out))
20:05:03 Quit TCK- (Read error: 110 (Connection timed out))
20:05:15amiconnsafetydan: The crash still happens here
20:06:30amiconnI start playing music (tested both with ogg and mp3), then start moving around the sim window.
20:06:43safetydanamiconn, back trace? Or is it still the "wacky negative values to pcm_buf_len" issue?
20:06:48amiconnWithin a few seconds, music stops playing
20:07:10amiconn..and then either the sim freezes or exits
20:07:31amiconnI was only able to reproduce the latter effect right now
20:07:42amiconn(tried 4 times, 2 each with ogg and mp3)
20:08:31egotrippenyou guys have any ideas what would cause problems with the 'playlist control file' when resuming playback?
20:08:36egotrippenhave you heard of this bug before?
20:09:18safetydanegotrippen, maybe trying deleting the .playlist_control file and trying again? Or possibly a scandisk.
20:09:26amiconnArgh! Under gdb control, the former effect happens
20:09:39amiconnThe crash freezes the whole system for several seconds
20:10:36 Join Thus0 [0] (
20:11:37SuperSnoutokay.. i'm having problem.. i copied the wad file from my copy of ultimate doom and named it doom1.wad.. but when i play the game the startup screen is correct.. but all the rest is the same as the shareware version and i can only play knee deep in the dead..
20:11:44SuperSnout*a problem
20:12:30amiconndoom1.wad is the wrong name
20:12:41safetydanamiconn, that's crashing inside the Rockbox pcmbuf code
20:12:43 Join RotAtoR [0] (
20:13:06safetydanAlso, your sdl_audio_callback seems to get a much longer "len" value than mine
20:13:51 Quit Lear ("Chatzilla 0.9.71 [Firefox]")
20:15:45safetydanYeah, my audio callback only seems to want 8192 bytes at a time
20:16:25amiconnpcmbuf_current is a NULL pointer
20:16:28SuperSnout"just rename the wad to what the one you have already is called".. what do i name it then?
20:16:40safetydanHow big is a typical PCM block sent to the DAC? 32k? the dereferencing caused SIGSEGV
20:17:02linuxstbsafetydan: Yes, I think the callback normally gives 32KB
20:17:35amiconnSuperSnout: The .wad files need to have their proper names, as the doom code 'magically' adjusts its behaviour according to the name
20:17:44safetydanso it's almost like the pcmbuf code isn't keeping up in the sim and running out of buffers maybe
20:18:22egotrippensnout: try doom2.wad
20:18:25amiconnSuperSnout: The ultimate doom wad has to be called doomu.wad
20:18:37safetydanamiconn, on line 279 of sound.c, try lowering the value from 2048 to something smaller, like 512
20:18:43egotrippeni dunno, if the H300 doom whatever can read that... try it though
20:18:46 Quit Thus0 ("Leaving")
20:18:48egotrippenif it doesn't work, try doom2
20:18:52SuperSnoutamiconn thanks!
20:20:07amiconnsafetydan: You mean line 276?
20:20:22safetydanit's 279 on my system
20:20:34safetydanfmt.sampes = 2408;
20:20:42safetydancan't type
20:20:43amiconnYes, 276 here
20:20:45safetydanbut you get the idea
20:20:52amiconn* $Id: sound.c,v 1.9 2006-02-23 21:42:53 dan Exp $
20:21:03safetydanah... yeah I just added some debug stuff
20:21:13safetydansorry about that
20:21:29safetydananyway, that should tell SDL to ask for a smaller audio buffer
20:21:35amiconnI will have to fiddle with sound.c anyway, on amd64 there's two places with incompatible pointer type warnings
20:23:10amiconnSame effect with fmt.sampes = 512;
20:23:42safetydanMy only other guess is that we need a better mapping of Rockbox threads on to SDL threads.
20:23:47safetydanParticularly yield()
20:25:12 Join imphasing [0] (
20:25:25egotrippensafetydan: looks like deleting the file fixed it
20:25:52SuperSnoutamiconn: worked =) thanks
20:25:58BHSPitLappyhas rockboy been worked on lately?
20:26:10safetydanegotrippen, possibly the file got corrupted somewhere along the line
20:26:28safetydanBHSPitLappy, if you haven't seen any CVS commit messages then probably not
20:26:36egotrippenmm. it's an easy enough fix. thanks for the tip
20:26:42egotrippenit'd been happening on and off for awhile
20:29:40 Quit mikearthur (Read error: 104 (Connection reset by peer))
20:29:52 Join mikearthur [0] (
20:33:48 Quit |Beowulf| ("KVIrc 3.2.0 'Realia'")
20:35:16amiconnsafetydan: Maybe the problem is that the sdl sim playback code can't deal with the pcm buffer running empty?
20:36:16amiconnHmm, the problem happens in pcmbuf_callback(), which is core code...
20:36:47safetydanI did write the sim sound code to try and deal with that problem.
20:36:58safetydanThough it could probably do things better
20:38:53 Join |Beowulf| [0] (
20:39:04 Join Thus0 [0] (
20:42:45 Join matsl [0] (
20:43:29*preglow sees further ways of optimising the fiq handler...
20:44:51 Quit bluey (Read error: 110 (Connection timed out))
20:49:52amiconnsafetydan: Could be interesting to test whether pcmbuf_callback() does the same NULL pointer access when running out of data,
20:50:02amiconni.e. when the codec can't keep up
20:50:18safetydanWouldn't someone with an iPod discovered that by now?
20:50:43amiconnRockbox has no memory protection
20:50:43safetydanI thought there were issues with some codecs keeping up with realtime
20:50:58 Join TCK [0] (
20:51:01safetydanBut has anyone reported freezing on high bitrate files?
20:51:14preglowit doesn't freeze
20:51:17preglownot on ipod, at least
20:52:04 Join Paul_The_Nerd [0] (
20:52:24amiconn..but on iriver (and archos) I've implemented a 'memory guard' feature
20:52:48amiconnThe page talks about archos only, I should update it...
20:53:15amiconnSo, if we had an .aac test file, the test could be done...
20:54:03 Join Steven___ [0] (
20:54:24amiconnAnother option would be to disable cpu boost
20:54:39preglowamiconn: did you get cross compilers working on your amd64 box with no hassle?
20:54:56amiconnpreglow: sh and arm: yes, m68k: no go
20:55:05preglowamiconn: same problem i had, then...
20:55:18preglowm68k was the only thing i used at that time
20:55:24amiconnI tried 3.3.6, 3.4.5, 4.0.2 and 4.1.0
20:55:31preglowyeah, i tried bunches too
20:55:34preglowgoddamned annoying
20:55:36amiconnSame bug in all of them, at the same place
20:55:42preglowhave you reported it?
20:55:47amiconnEven rewriting that function didn't help
20:55:50amiconnNot yet
20:56:15amiconnYesterday I tried crosscompiling the crosscompiler
20:56:49amiconn..but couldn't do that because that requires to build an x86_64 crosscompiler on i686 first, which failed as well
20:57:33amiconnTotally different bug though
20:57:40HClwhy can't you just use an 32bit compiled crosscompiler?
20:57:42HClshouldn't it work fine?
20:58:55preglowbut ok, time to start the linux installer fun
20:59:38MikachuHCl: you need a complete 32bit library install then
20:59:48 Join leftright [0] (
20:59:55HClMikachu: ah...
21:00:07leftrightSlasheri you around ?
21:00:11HClstatic 32bit compiled crosscompiler? o-o
21:00:37 Join Farpenoodle [0] (n=solo84@
21:01:24 Join quobl [0] (i=tor@tor/session/x-0dd2143de59ee0ce)
21:02:46safetydanhrmm... my clipboard code adds 2k to onplay.o
21:03:40 Join linuxstb_ [0] (
21:04:45leftrightSlasheri; dir cache doesn't enable if its set to 'on' in a config file, one has to go to the dir cache menu and re-enable it again
21:04:55 Quit linuxstb (Read error: 110 (Connection timed out))
21:05:34 Join ghode|afk [0] (
21:05:41 Quit Nico_P ()
21:05:45leftrightSlasheri Lostlogic, I get the occasional pop when skipping tracks and when selecting play from pause.
21:08:29 Quit egotrippen ("CGI:IRC (EOF)")
21:08:43amiconnsafetydan: The .o size increase isn't important; what's important is the increase in final binary size
21:09:26 Quit ts-x ("CGI:IRC (EOF)")
21:10:04safetydanbinary increased about the same amount
21:10:07safetydanfor h120 at least
21:10:18amiconnoh :(
21:10:38safetydan1220 bytes precisely
21:10:59 Quit damaki_ (Read error: 104 (Connection reset by peer))
21:13:41 Part leftright
21:15:11 Join damaki [0] (
21:17:17 Nick linuxstb_ is now known as linuxstb (
21:18:27linuxstbpreglow: Did you see my changes to enable the C version of the FIQ handler for the ipod 3g? Audio on the 3g isn't working (there are no fiqs being fired), so if you spot anything odd, please let me know.
21:18:42 Quit damaki (Remote closed the connection)
21:19:14 Quit Farpnut (Read error: 110 (Connection timed out))
21:19:28 Join damaki [0] (
21:19:33 Join Farpnut [0] (n=solo84@
21:20:33SereR0kRif I got a fat32 formatted iPod, can i install RockBox from a mac without Virtual PC?
21:20:36***Saving seen data "./dancer.seen"
21:21:43 Quit Farpenoodle (Read error: 110 (Connection timed out))
21:25:50 Quit _FireFly_ ("Leaving")
21:26:15 Quit ghode|afk (Read error: 110 (Connection timed out))
21:29:40MikachuPaprica: #defint PREV_MONTH BUTTON_SCROLL_BACK
21:29:50Mikachujust noticed that line wasn't hilighted in vim
21:33:12Mikachudoes it work on nano?
21:33:24Papricanot yet
21:35:21 Join skwad__ [0] (
21:35:23 Quit skwad_ (Read error: 104 (Connection reset by peer))
21:36:28safetydanWell I got the cut/copy/paste code to be 16 bytes smaller. But I can't see anything else I can easily remove to make it smaller. If anything it will get larger if I add a progress indicator
21:37:07Paul_The_NerdMaybe it would be best if it was a "file manager"' kinda plugin then?
21:39:11safetydanIt depends on whether 1200 bytes is significant enough to warrant moving out of the core.
21:39:57safetydanBut yeah, norton commander style plugin would be cool
21:42:07amiconnIs this already including directory copy/move ?
21:42:36amiconn..and how much does it add for archos (where size matters most)
21:42:45safetydandon't have an sh1 tool chain locally
21:42:50safetydanI should probably do that
21:42:59*amiconn has all 3 toolchains
21:43:15Mikachui accidentally installed an sh toolchain when i didn't look at the instructions too carefully
21:43:39amiconnsafetydan: Is the patch on the tracker?
21:44:08safetydanamiconn, yes
21:44:16safetydanthe version 5 down the bottom
21:44:28safetydandoesn't include some minor size changes I've made recently though
21:44:49Papricablah, why cant i delete my files from the traker?
21:45:54safetydanIs gcc 3.4.5 okay for SH1? The crosscompiler page seems to say 3.3.x is better
21:47:40amiconnStick with 3.3.x
21:47:48amiconn3.4.x will generate larger binaries
21:48:30amiconn+1444 bytes for archos recorder
21:48:42amiconn(added by your patch)
21:49:28safetydanSo, is that too large?
21:50:12Mikachuisn't the splash image bigger?
21:51:27amiconnNot on archos
21:54:12amiconnclipboard_(clip|cut|copy) don't need a return value 'cause it's always true anyway
21:54:33amiconnclipboard_cut and clipboard_copy could be inline or macros
21:55:21 Quit imphasing (Read error: 104 (Connection reset by peer))
21:55:41safetydanclipboard_cut/copy are called by the menu system so I think they have to return bool and not be macros
21:56:14amiconnah ok, sorry
21:58:30amiconnHmm, I'm undecided on this
21:59:11amiconnOn one hand, I consider this a very nice feature, but otoh it adds quite some code
21:59:52amiconnI got an idea though how we could perhaps save some code, not only in your patch but everywhere were menus are built dynamically
21:59:54Paul_The_NerdI personally am of the opinion it should be in a plugin for a few more reasons now that I think about it.
22:00:11 Join ashridah [0] (
22:00:27Paul_The_NerdIt seems to me, just from my personal perspective, that through normal use a music player's software shouldn't modify any files that it doesn't create (database, etc.)
22:01:21Paul_The_NerdAnd that any functions with the potential to do that should require pretty explicit actions on the hand of the user (entering a plugin or changing a setting to enable it that doesn't save across reboots, or something similar)
22:01:28Paul_The_NerdI mean, that's just my personal view on the matter.
22:02:02amiconnThe file/folder delete function would then be another candidate
22:02:28amiconnOtoh plugins (still) have the disadvantage that they aren't localisable
22:02:43Paul_The_NerdThat is a problem then.
22:03:13safetydanI agree that this could work better as plugin. It's certainly not as "friendly" though.
22:03:27amiconnThe question is how it would work as a plugin
22:04:00 Join nobelium [0] (
22:04:25amiconnOh, and there are more file management functions which should be moved into a plugin if we decide to make file management a plugin
22:04:40amiconnThere's also 'rename' and 'create folder'
22:07:00safetydanThis is why I haven't comitted things yet :)
22:07:54 Join solexx_ [0] (
22:10:58amiconnsafetydan: You can save some code for most devices: The check for EXDEV and retry as copy is only necessary for #ifdef HAVE_MULTIVOLUME
22:11:44amiconnThe only devices where code size won't be reduced right now are the Ondios, but for these code size is less critical.
22:12:59safetydanthat's easy enough to do
22:13:26 Quit BHSPitLappy (Read error: 110 (Connection timed out))
22:15:42safetydanwell that saves another 100 bytes on h120
22:16:42 Quit Matze ("Miranda IM! Smaller, Faster, Easier.")
22:16:46 Join pre2k_ [0] (
22:16:52safetydanweird, gcc-3.3.6 won't compile
22:19:51 Quit solexx (Read error: 110 (Connection timed out))
22:20:03*Bagder downloads 78MB Cowon A2 firmware source code
22:20:17 Quit linuxstb (Read error: 110 (Connection timed out))
22:20:18 Quit nobelium- (Read error: 110 (Connection timed out))
22:20:31 Quit TCK (Read error: 110 (Connection timed out))
22:21:16 Join linuxstb [0] (
22:21:24 Quit SereR0kR ("XChat Aqua")
22:21:55Bagderat least it makes our source package seem small
22:22:55 Join TCK [0] (
22:25:50 Quit Steven___ (Read error: 110 (Connection timed out))
22:26:10safetydannot that it matters, but 1664 bytes on arm (iPod Nano)
22:26:41*Paul_The_Nerd does the "our code is the biggest" dance.
22:26:43Paul_The_NerdOr something.
22:26:54*Mikachu gives a 5.2
22:27:17*Bagder instructs Paul_The_Nerd on how to put his feet when doing that dance
22:28:26Bagder78MB is a lot of source
22:28:29Paul_The_NerdI'll remember that for next time.
22:28:39Paul_The_NerdIs that 78mb including images and other non-text files?
22:28:48BagderI don't know yet, their site is quite slow
22:28:56Bagderbut I guess it is
22:29:37Bagderin case anyone would be interested
22:30:05Paul_The_NerdI fear it.
22:30:38Paul_The_NerdI just know they've run some sort of quick script over it to strip out all comments and rename variables to random character strings, just to be spiteful.
22:30:42Bagderyou can tell what they based their hw on
22:31:18Bagderlinux 2.6.5 kernel
22:33:31Bagderplenty korean comments ;-)
22:33:42Bagder // ?õ ̵? ?? ?..
22:34:20Mikachuthat looks like it just might be in the wrong charset :)
22:34:31Bagderyes it is of course
22:34:41Bagderand I bet my irc client config screws it up even more
22:36:16Bagderand if you thought libFLAC uses long source lines, you should see these ;-)
22:36:22 Join moozooh [0] (n=moozooh@
22:36:29Bagder300 columns
22:36:52safetydanI don't even see lines that long in my Java day job
22:37:13Bagderbrickmania should be code policed
22:37:54 Quit quobl ("Leaving")
22:38:24 Quit nobelium (Read error: 110 (Connection timed out))
22:38:25BagderPaprica: 80 columns, not 180 ;-)
22:41:44 Join petur [0] (
22:43:41 Quit matsl (Read error: 104 (Connection reset by peer))
22:43:47BagderThe A2 has the same CPU as the new Neuros models are gonna use
22:44:23pre2k_cd ..
22:44:30Mikachuno such directory
22:44:34Bagderrm -rf /
22:45:12pre2k_uhmm, there's rockbox in there...
22:45:27safetydanmke2fs /dev/hda
22:46:17miraksafetydan: cat /dev/zero /dev/hda
22:46:28miraksafetydan: cat /dev/zero > /dev/hda
22:46:53mirakI am not sure this works
22:46:59miraksafetydan: cat /dev/random > /dev/hda
22:47:01mirakthis one works
22:48:20Mikachuyou probably want urandom
22:48:23 Quit darkless (Client Quit)
22:48:28 Join darkless [0] (
22:53:15 Join BHSPitLappy [0] (
22:54:50 Quit safetydan ("Leaving")
22:59:06preglowlinuxstb: so we can use the same fiq handler for both pp chips?
23:05:58 Join ghode|afk [0] (
23:06:53preglowgod, its been so long since i had trouble making my norwegian keyboard work properly in linux, id forgotten just how much fun it can be
23:09:03 Quit StrathAFK (Read error: 104 (Connection reset by peer))
23:13:15 Join Strath [0] (
23:13:48 Join LinusN [0] (
23:15:50linuxstbpreglow: With some minor changes, we should be able to, yes.
23:16:11linuxstbBut we need to get the C version working first...
23:16:27 Join Shadowarrior13 [0] (
23:18:27 Join JdGordon [0] (
23:19:08JdGordonmorning all
23:19:09preglowi can exploit the fact that all register addresses are a base address plus some offset to free a bunch of registers, plus init a couple of registers in fiq mode in crt0.S, so should be able to shave off more cycles at least
23:19:17preglownow if only i got my keyboard working....
23:19:59Mikachusetxkbmap no should be enough i think
23:20:28preglowid prefer to do it from gnome, so all parts of ubuntu remembers what i just did
23:20:37***Saving seen data "./dancer.seen"
23:20:40preglowits never bloody been a problem before, but now...
23:20:55preglowwhat the hell
23:20:58preglowthat command alone just worked
23:21:03preglowwhat happened just now :>
23:23:33preglowi didn't even give it a parameter, i just typed setxkbmap
23:25:44 Quit ghode|afk (Read error: 110 (Connection timed out))
23:32:31 Quit TCK (Read error: 110 (Connection timed out))
23:32:47 Quit Daishi (Remote closed the connection)
23:33:33 Join Daishi [0] (
23:34:02 Quit paugh ("Leaving")
23:37:50 Join stripwax [0] (
23:39:14stripwaxjust hanging out. still trying to get ddd to play nicely. is it just me or can I not set a breakpoint on an arbirary line of code?
23:39:32 Quit |Beowulf| (Read error: 110 (Connection timed out))
23:39:34stripwax(still part-time debugging this f**king auto-change-directory-not-working bug)
23:40:09 Join nobelium [0] (
23:44:15amiconnlinuxstb: Is there a reason why rockbox' inttypes.h uses #define and not typedef?
23:44:31stripwaxddd is a gui for gdb. only it's even slower and not much friendlier
23:44:50preglowyou can set a breakpoint on a line of code, yes
23:45:05preglowdon't remember how, though
23:45:17stripwaxyep, in gdb I can. how do I do that in ddd? maybe I'm just using it wrong
23:45:19preglowamiconn: because the rockbox rules say we don't use typedef? :>
23:47:31linuxstbamiconn: What preglow said. ;)
23:49:15Papricammm what do you think on open a sub forum for plugins?
23:49:35LinusNpreglow: that rule is there to protect against unnecessary typedefing of structs etc
23:49:53preglowLinusN: yes, i know
23:49:59LinusNinttypes is a valid use imho
23:50:07 Quit ender` (" If at first you don't succeed, destroy all evidence that you have tried.")
23:51:34JdGordon*couph* any1 want to look at the finishd search in play list patch?
23:51:43JdGordon.. if you have nothing else to do...
23:52:05preglowwill i be shot, killed or maimed if i start assuming variable order in assembler code?
23:52:30LinusNnah, we'd just drag you out in the forest and beat you up
23:53:04preglow'cuz i've got two variables that are declared in order here, and i want to free a register...
23:53:16ashridah* preglow moves to the desert
23:54:04 Quit mirak (Remote closed the connection)
23:54:06*preglow has embarked on the holy quest of the blazingly speedy fiq handler
23:56:36LinusNpreglow: put a comment where the variables are declared, explaining the situation
23:56:51Mikachupreferably not in korean
23:57:00preglowwould be nice, and i can't imagine why gcc would reorder them
23:57:05LinusNswedish or norwegian :-)
23:57:06preglowMikachu: but that's the only language i know...
23:57:13LinusNbork! bork!
23:58:18muesli__LinusN got assimilated? ;)

Previous day | Next day