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 2005-06-04

00:00:32preglowhmm, temperature actually isn't half bad outside now
00:01:12amiconnBug is still there :(
00:01:39[IDC]Dragonlet me say godnight then
00:01:53 Quit ghode|afk (Read error: 104 (Connection reset by peer))
00:04:05 Quit [IDC]Dragon ()
00:04:14 Quit bg_ ("Chatzilla [Firefox 1.0.4/20050511]")
00:04:43 Join ghode|afk [0] (
00:04:48 Join Shari [0] (
00:05:29 Quit Chamois (" HydraIRC -> <- 100,000+ downloads can't be wrong")
00:06:01 Quit Harpy (Operation timed out)
00:06:39 Join webguest70 [0] (
00:06:42*Bagder runs to bed
00:08:00*t0mas too
00:08:04 Quit t0mas ("good night")
00:08:26Shariu all are lame boy
00:08:54Sharilamer dan d Backstreet boys
00:09:19 Part Shari
00:09:56preglowyes, yes we are
00:11:45preglowjust some random cockgobblers musings
00:11:45 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
00:12:03MoosCamaroyes probably
00:17:06HClwhat is .tt anyways
00:17:18amiconnTrinidad & Tobago
00:17:23HClyea, i just looked it up :P
00:17:26HClprobably a bouncer
00:17:41amiconnThey sell their cctld, like .tv does too
00:18:29preglowso .tv is actually a national thing
00:18:37MoosCamaroTT are good islands is Tuvalu iirc
00:21:09 Join Chamois [0] (
00:21:57 Quit webguest70 ("CGI:IRC (EOF)")
00:23:46 Quit DomZ ("CGI:IRC")
00:24:05 Quit spiralout ("ChatZilla [Firefox 1.0.4/20050511]")
00:30:13HClbah, i want stuff in cvs :/
00:30:19*HCl guesses he could work on playtime
00:33:14preglowwhy does it have to be in cvs?
00:33:19preglowyou've got what will be put in cvs anyway
00:33:27HCli know
00:33:32HCli don't want the patch to change
00:33:36HClwhen i've built a patch on top of it
00:36:20niobosHCl: quick question: the file-offsets: am I correct that artist_start is always 68?
00:36:30HCl :)
00:36:32*niobos is off to bed
00:36:36*HCl is glad his wiki description is proven readable
00:36:50niobosbye all
00:37:11 Quit niobos ("hmm, some nice sleep...")
01:06:01MoosCamarogood night bob
01:06:11 Quit bobTHC ("smoke weed every day!")
01:06:15MoosCamarogo too
01:06:26MoosCamaroGn8 all
01:06:43 Part MoosCamaro
01:11:16 Quit preglow ("leaving")
01:13:56 Quit Shagnar (Read error: 60 (Operation timed out))
01:18:38***Saving seen data "./dancer.seen"
01:22:09 Quit mrlala (Read error: 110 (Connection timed out))
01:26:26Seedany native German around?
01:27:54 Quit webguest34 ("CGI:IRC")
01:47:40 Quit Chamois (" HydraIRC -> <- IRC for those that like to be different")
01:52:58 Quit Sucka ("a bird in the bush is worth two in your house")
01:58:51 Join muesli- [0] (
02:02:13 Join thegeek_ [0] (
02:02:13 Quit thegeek (Read error: 54 (Connection reset by peer))
02:07:53 Join ehntoo [0] (
02:13:57 Nick thegeek_ is now known as thegeek (
02:16:25 Quit muesli- (Read error: 60 (Operation timed out))
02:21:25 Join mrlala [0] (
02:22:46 Join StrathAFK [0] (
02:39:59 Quit Strath (Read error: 110 (Connection timed out))
02:42:57amiconnphun :/
02:53:10 Join Stryke` [0] (
02:58:25 Part mrlala
03:07:26 Join mrlala [0] (
03:18:42***Saving seen data "./dancer.seen"
03:25:55 Join BTKDaImMaikata [0] (~Bager@
03:25:57 Join tvelocity [0] (
03:26:34 Quit Bger (Read error: 101 (Network is unreachable))
03:29:03 Join amiconn_ [0] (
03:47:17 Quit amiconn (Read error: 110 (Connection timed out))
03:47:18 Nick amiconn_ is now known as amiconn (
03:55:50 Quit mrlala (Read error: 60 (Operation timed out))
04:05:56 Join QT_ [0] (as@area51.users.madwifi)
04:15:45 Quit QT (Read error: 110 (Connection timed out))
04:28:20 Join ashridah [0] (
04:43:07 Quit tvelocity ("Leaving")
04:43:14 Join bipak [0] (
04:50:42 Quit bipak_ (Read error: 145 (Connection timed out))
05:00:38 Quit TROGDORRR ("( :: NoNameScript 3.81 :: )")
05:11:00 Quit ehntoo ("Leaving")
05:18:45***Saving seen data "./dancer.seen"
05:29:09 Join mrlala [0] (
05:32:51 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
05:40:27 Join tvelocity [0] (
05:58:45 Quit mrlala (Read error: 110 (Connection timed out))
06:36:38 Join webguest71 [0] (
06:38:25 Quit webguest71 (Client Quit)
06:38:44 Join webguest68 [0] (
06:40:35 Quit webguest68 (Client Quit)
07:18:50***Saving seen data "./dancer.seen"
07:32:56 Join mrlala [0] (
07:32:57 Quit ghode|afk (Read error: 104 (Connection reset by peer))
07:46:09 Quit Bagder ("Off to search for that connect-resetting peer guy!")
07:57:01 Quit QT_ (
07:57:01 Quit CoCoLUS (
07:57:01 Quit odd (
07:57:01 Quit mbr (
07:57:01 Quit silencer_ (
07:57:01 Quit rob- (
07:57:01 Quit gromit` (
07:57:44NJoinQT_ [0] (as@area51.users.madwifi)
07:57:44NJoingromit` [0] (~gromit`
07:57:44NJoinCoCoLUS [0] (
07:57:44NJoinodd [0] (
07:57:44NJoinmbr [0] (
07:57:44NJoinsilencer_ [0] (
07:57:44NJoinrob- [0] (
08:04:09 Quit mrlala (Read error: 110 (Connection timed out))
08:28:18 Join mrlala [0] (
08:48:15 Quit tvelocity ("Leaving")
09:18:52***Saving seen data "./dancer.seen"
09:21:59 Join Harpy [0] (
09:41:00 Quit mrlala (Read error: 110 (Connection timed out))
09:42:44 Join tvelocity [0] (
09:58:00 Part Simplet_Onasme
10:11:47 Quit ashridah ("Leaving")
10:18:53 Join webguest29 [0] (
10:25:48 Join niobos [0] (
10:33:47 Join Chamois [0] (
10:35:37 Quit Chamois (Client Quit)
10:39:08 Join Chamois [0] (
10:39:50 Join ashridah [0] (
10:40:30 Quit Chamois (Client Quit)
10:41:06 Join Chamois [0] (
10:48:22 Quit webguest29 ("CGI:IRC")
10:48:36 Join webguest40 [0] (
10:57:32 Join mrlala [0] (
11:03:57 Join t0mas [0] (
11:04:13 Quit t0mas (Read error: 54 (Connection reset by peer))
11:04:18 Join t0mas [0] (
11:04:34t0masmorning :)
11:06:53Seedgood morning
11:12:40 Quit mrlala (Read error: 60 (Operation timed out))
11:18:55***Saving seen data "./dancer.seen"
11:26:52 Quit webguest40 ("CGI:IRC (EOF)")
11:44:00t0mashm... was it preglow who "fixed" a dead pixel on a tft?
11:44:40t0masI have 1 dead pixel... and the shop is willing to take it back (payed 15 euros for that)... but they don't have a replacement screen for me
11:45:22t0masso they offered 100 euro discount... (as they sell the replaced screens for 100 euro less)... and if I can fix the pixel... then it would be nice to have the discount :)
11:46:32 Join webguest85 [0] (
11:50:58 Join Musicmad [0] (
12:01:47 Quit Chamois (Read error: 110 (Connection timed out))
12:02:37 Join bipak_ [0] (
12:04:01 Join dapureplaya [0] (
12:04:58 Join webguest87 [0] (
12:04:59 Quit dapureplaya (Client Quit)
12:05:55 Part webguest87
12:14:22 Join Chamois [0] (
12:15:09Chamoishi slasheri
12:15:14Chamoisi found the problem
12:15:22Chamoiswith the gap during playback
12:15:25Slasherinow mp3 playback should be quite true gapless, i just made a simple dsp that removes silence from beginning of the song
12:15:44Chamoisi erase my configuration file
12:15:53Chamoisand hop playback is perfect
12:18:19 Nick QT_ is now known as QT (as@area51.users.madwifi)
12:18:28 Quit bipak (Read error: 110 (Connection timed out))
12:19:02*HCl has a cold :/
12:24:35 Join MoosCamaro [0] (
12:24:49MoosCamaroMorning all
12:25:02 Join preglow [0] (
12:25:27HClits preglow!
12:25:41*preglow looks threatening
12:26:00*HCl hands preglow a cookie
12:26:12*preglow grabs greedily
12:26:16preglowSlasheri: what silence do you remove? the amount described in the lame header encoder delay field?
12:26:21*HCl runs while preglow is distracted
12:28:10Slasheripreglow: I am not sure about those delays yet (because i haven't yet managed to get full gapless output using them), but now the codec removes all silence from the beginning of the song and that seems to do true gapless playback
12:29:49niobosHCl: the offsets in the DB: those are absolute offsets in the file? not index-numberd
12:30:21preglowSlasheri: you should mail someone who knows about these things to find the true behaviour you're after
12:30:23HClyea, absolute
12:30:39HClat first i wanted to make them index numbered, but there were people that were against that
12:30:49preglowSlasheri: the guy who made the lame info header spec page most certainly will now
12:30:55Slasheripreglow: Hmm, that's true. or just wait the patch gets to cvs and somebody will fix it
12:30:59niobosand it should be all sorted, I guess;
12:31:24niobosI asume binary sort will do? or should it be case-insensitive?
12:31:25amiconnSlasheri: An mp3 encoded with lame −−nogap should not contain any unwanted silence
12:32:16Slasherido you know if there are any specs on the web about that option?
12:32:28*niobos goes back to his books, and calculates the electrical field inside a coax-cable
12:32:41HClcase insensitive
12:32:58HClbut that doesn't mean you can't do binary search in it.
12:33:14nioboshmm... explain
12:33:20amiconnSlasheri: −−nogap does a rather simple thing
12:33:37amiconnIt can only work if you give lame more than one .wav at once
12:34:13amiconnlame then simply joins the .wav data for encoding, and splits at the nearest frame boundary after encoding
12:34:37SlasheriAh.. So there should be no silence between tracks..
12:34:44niobosjust compare each item then?
12:34:59HClniobos: they're still sorted, its just that lower and uppercase are equal
12:35:03niobosand don't asume entries[i] > search -> it isn't in
12:35:08amiconnThe result is actually one consecutive mp3 stream, just split in two (or more) files
12:35:21Slasherihowever, now the dsp works also with many songs that are not made right way with lame
12:35:24amiconnSlasheri: Do you reinitialise the codec on track change?
12:35:31Slasheriamiconn: no i don't
12:35:42preglowthere's no point in reinitialising an mp3 codec
12:35:51pregloweven if he did it wouldn't matter
12:35:51SlasheriBut i have tried also reinitialization and that has no difference..
12:36:08amiconnpreglow: What if the mpeg version/layer/sample frequency changes?
12:36:10nioboswell, sorting is not for today yet, so I'll think about it.
12:36:18amiconnSlasheri: You should not initialise
12:36:49*HCl yawns
12:37:03preglowamiconn: shouldn
12:37:14HClany objections if i put my java tool in cvs, and start upgrading the database to v3 ?
12:37:23HClnot really big changes, just adding some fields to the song structure
12:37:24preglowamiconn: shouldn't need to reinitialise then either, but you'd of course have to handle the sample rate change some other way
12:37:27HClplaytime, track number
12:38:52niobosHCl: please do update the wiki then... so I can do it for V3 without wasting time to V2
12:39:35HCli'll get working on it
12:39:50amiconnHCl: Yes I have objections if the v3 database can only be generated with the java tool
12:40:17HClamiconn: it won't. i'll adjust the songdb tool to fill in invalids for the extra fields
12:40:49niobosamiconn: and I'w writing songdb in C... but that could take a while...
12:41:36amiconnI know that you are doing that
12:41:44*HCl goes to adjust his java code structure to make it even better..
12:42:04amiconnI'd still prefer perl though (or any other widespread scripting language), since this allows to run it on almost any machine
12:42:16amiconn...without extensive setup procedures
12:44:25amiconnImho the only reason for having a .c version would be to run it on the box itself, but then it should be able to deal with low memory
12:44:31amiconn(a non-trivial task)
12:45:07niobosamiconn: is it worth for me to continue then
12:45:57t0masniobos & amiconn: wxwidgets?
12:46:08t0masthat's a toolkit for graphical C apps for linux and windows and mac
12:46:24t0masso maybe it's a good idea to chose for that for all end-user rockbox tools?
12:46:26preglowwhat do you want to use that for?
12:46:39t0masfor making a crossplatform end-user tool?
12:46:48t0masso windows users can use rockbox good too?
12:46:50amiconnt0mas: The problem ist that it is necessary to compile a .c version before it can be used
12:47:05t0masyes, but we could distribute an executable for that
12:47:16niobosexecutables are platform dependent
12:47:27amiconnFor windows, you could carry a precompiled .exe, since an .exe should run on any Win32 box, but that's not possible for Linux
12:47:38t0masyes, but on linux everybody can compile something
12:47:47preglownot true
12:47:48t0masand on windows we need an exe tool...
12:47:52preglowlots of distros don't bundle compilers
12:48:04t0masthen make some rpm package?
12:48:09t0masalmost all distros can use that?
12:48:11preglowamiconn: it is more or less possible for linux
12:48:16niobospreglow: lots of distro's dont install compilers... all of them supply one...
12:48:22amiconnEven if you can compile on a linux box, it's still a lot more work than simply running a perl script
12:48:29t0masbut an all-in-1 tool, for patching firmware... converting video... converting dictionary and makeing a songdb would be nice to have...
12:48:47preglowamiconn: a precompiled binary should work fine on more or less all boxes
12:48:59amiconnpreglow: Iirc different linux distros have different lib versions, residing in different dirs etc
12:49:06preglowamiconn: static linking
12:49:19preglowdirs doesn't matter, you don't have to worry about linking
12:49:19amiconnWhat about non-x86?
12:49:22preglowldd does that
12:49:28preglowamiconn: i did say 'most' boxes
12:49:41preglowamiconn: if you follow that reasoning, you also have to worry about windows 2000 on alpha computers
12:49:44t0masnon-x86 people are used to compiling
12:50:23amiconnIt's not that this would hit often, but we would give away some flexibility we currently have, and that's my point
12:50:55niobosI'm learning Perl too, so if you want I could try to make it in Perl...
12:51:09nioboshowever i'm an absolute perl noob...
12:51:27t0masstill... I think a windows tool would be nice...
12:51:37t0masbecause not every noob can run perl scripts
12:51:48preglowt0mas: then start working
12:51:58nioboswindows tool as in: a .exe that runs under windows
12:51:59t0maspreglow: vacation in 2 weeks :)
12:52:01preglowt0mas: i for one hate gui programming, so you wont see mee coding that
12:52:02niobosor as in: a GUI?
12:52:12t0maspreglow: I can code the gui...
12:52:15amiconnniobos: I'm not saying c code is useless. I would try to do this myself, since it would be cool to run the db generation on the box itself
12:52:21t0masand maybe you can embed the tools we have now?
12:52:37amiconnBut, as said, it should be able to deal with low memory, i.e. database larger than free ram
12:52:50niobosthat can be a problem...
12:53:01nioboscurrently all DB-manipulation is done in RAM...
12:53:08niobosbut I can adapt for that, I think
12:53:25t0maspreglow: maybe we van include the firmware patching stuff too? so all noobs can use it?
12:53:54amiconnniobos: I already though about how to solve that.
12:54:02niobosamiconn: should it ALWAYS use the least amout of RAM, or should it try to do it in RAM and, if failing, fall back to disk?
12:54:27amiconnThe biggest problem is sorting a huge table, but that can be solved by doing a disk-based merge sort
12:54:46amiconnniobos: I think you could use all available ram
12:55:09preglowdemanding to play music while a bloody db build is being done is a bit unreasonable
12:55:18niobosi agree
12:55:31preglowso you should have around two megs of memory to do it in
12:56:06nioboshmm... if I use a (few) temp-files, where should I place them? in the root? in the .rockbox/?
12:56:54amiconnImho it doesn't matter as long as you always do a proper cleanup
12:56:57HClyay for code restructurization
12:57:12amiconnI'd use the root dir, because of the shorter path
12:57:43 Join mrlala [0] (
12:58:03niobosusing the .rockbox/ would avoid file-name-collisions... since we "own" the dir
12:58:49amiconnWe do already generate unique filenames in a number of places
12:59:06amiconn...I want to convert that to one central function
12:59:48niobosk, I'll continue to write the C-thing and implement a HDD-fallback on mem-alloc failure
13:00:19preglowlike mkstemp?
13:00:29niobosI'll put a dummy mkstemp in to test it
13:00:44niobosor what will it be called?
13:01:09amiconnpreglow: Not exactly
13:01:39amiconnCurrently there are 3 places that generate unique file names using one of two methods
13:02:26 Quit edx (Read error: 145 (Connection timed out))
13:02:52amiconn(1) Using a number sequence: config file name, recording file name (on units without rtc), screenshot file name (on units without rtc)
13:03:12amiconn(2) Using a time string: recording file name (on units with rtc), screenshot file name (on units with rtc)
13:03:24 Quit webguest85 ("CGI:IRC (EOF)")
13:03:48amiconnBoth methods use a similar pattern: <prefix><uniqueness_part>.<ext>
13:03:59amiconnE.g. config-01.cfg
13:04:45HClyay for more code restructurization :3
13:05:02amiconnor rec_0002.mp3
13:05:21niobosamiconn: I guess I can find those functions in the CVS?
13:05:52amiconnThere is no central function yet, but I planned to do this
13:06:01amiconnCurrently, each place implements its own version
13:06:09niobosk, but IF you do it'll be in CVS?
13:06:28*HCl pokes java.
13:06:34niobosSo I might do a checkout and see how I should call it
13:06:36HClbah. i'm not allowed to do an abstract static :/
13:07:11niobosHCl: sure not... that's like asking for a variable constant...
13:07:23HClhow so?
13:07:49niobosstatic = this function is 1 function for ALL objects, and is once defined and can be called even without an object
13:07:54amiconnIt will be something like char *create_unique_name(char *buffer, const char *prefix, const char *suffix, int digits)
13:08:11niobosabstract = this is not realy a function, this is a template that must be overwritten by childs
13:08:22HCli want a global per-class function
13:08:23amiconnIf digits is 0, it would then create a timestring name on units with rtc
13:08:25HClthats abstract
13:08:26niobosyou can't have a function that must be overwritten and that is already defined...
13:08:40HClwhy not?
13:08:55HCli want an "int entrySize()"
13:08:59HClfor all my entry classes
13:09:07HClbut it should be static since it won't change per instance
13:09:19HClbut merf, i'll just make it non static
13:09:53niobosamiconn: k, I'll use that definition
13:10:21amiconnI may still change that, but certainly not much
13:10:40*niobos is back off to the study-books...
13:10:54HCli meant i'll make it non abstract o.o
13:14:13HClthat was too easy :)
13:14:25*HCl goes to update the wiki, rockbox code and perl script..
13:17:32HClniobos: updated the wiki..
13:18:58***Saving seen data "./dancer.seen"
13:27:17ashridahrockboy seems to hang with the version of rockbox i've got currently (an oldish version of slasheri's patches)
13:27:33HClyea, muesli reported that as well.
13:28:33HClback. dunno why it happens, haven't looked into it
13:31:09ashridahbehold! the power of entropy!
13:31:19HCli can't compile latest cvs anymore with my toolchain..
13:31:24HClloads of errors...
13:31:36HClCC plugin.c
13:31:36HClIn file included from /home/hcl/rockbox/cross/lib/gcc/m68k-elf/3.4.3/../../../../m68k-elf/include/unistd.h:4,
13:31:39HCl from plugin.c:25:
13:31:42HCl/home/hcl/rockbox/cross/lib/gcc/m68k-elf/3.4.3/../../../../m68k-elf/include/sys/unistd.h:23: error: parse error before "uid_t"
13:31:54HClmaybe thats me
13:31:57HClhold on o.o.
13:33:36HClthere o.o nevermind :p
13:33:39HClit was my fault
13:36:53niobosHCl: samplerate in Hz or in kHz?
13:37:50HClno idea :P
13:38:01HCli'll check in a bit.
13:38:07HCli'm assuming it will be in hz
13:38:18HClsince otherwise you'd need floating point
13:38:20BTKDaImMaikatais it int ?
13:38:25 Nick BTKDaImMaikata is now known as Bgr (~Bager@
13:38:34Bgrthen i'm almost sure it's in Hz
13:38:46HClwe'll see what my backend gives me in a bit
13:38:47niobosso it'll be limited to 65kHz (since it's 16bit)
13:38:49Bgror is fixed point
13:39:08niobosno support for (maybe) 96kHz?
13:39:15HCldo we want to support that?
13:39:23Bgrthere's no such hardware for now
13:39:37Bgrbtw, hi :)
13:39:57BgrHCl what caused the error ?
13:40:24HClwhat error?
13:40:33niobosHCl: you decide... We could extend it to 32bit... and leave a 16bit gap for future extpansion?
13:40:38Bgrwith compiling cvs
13:41:03HClniobos: meh, we'll just expand it when needed, its not a major surgery of code as long as its written maintainably
13:41:17niobosk 16bit in Hz then
13:41:20HClBgr: oh, my cruddy attempt at trying to get the searchengine to write to the playlist rather than /search.m3u
13:41:39HClbut i'll fix that later.
13:41:47*HCl goes to generate and try v3
13:41:52HCli'm hoping it will work
13:41:59Bgrv3 of db?
13:42:36HClminor changes
13:42:41HClplaytime, track number, samplingrate
13:42:51HClits mostly the playtime thats important
13:43:06HClthe rest just adds search options to the searchengine
13:43:28HClbut since we can't do versions 2.1 or so, it'll have to be 3
13:43:28HCl :P
13:45:15HClhm cool.
13:45:19HClbootloader usb mode
13:45:25HClnever had that before
13:45:56Bgrreally cool
13:46:11Bgrnot that i have tried it ...
13:46:13*niobos is off to lunch
13:46:23HClit sucks battery cause it keeps the hdd spinning, i think.
13:46:39HClbut its great for emergencys
13:47:11HClwoops, forgot to increase the db version in my rockbox code..
13:47:24preglowdoesn't exactly suck more battery than ordinary usb mode...
13:47:43Bgrwhat do you mean with spinning ?
13:47:57Bgri think so also
13:48:00Bgrthere's no reason
13:48:07HClit doesn't?
13:48:15Bgroriginal USB mode also spins the HDD all the time
13:48:18 Join webguest62 [0] (
13:48:34preglowordinary usb mode sure as hell never spins down the drive
13:48:49preglowrockbox can't control the disk in any way in usb mode
13:48:56BgrHCl the USB2.0 chip is MSD
13:49:04Bgrmass storage device class
13:49:30HClpreglow: are you sure? i thought i saw code in the bootloader to keep the disk spinning
13:49:51Bgrso the firmware (rockbox or iriver's) just flushes buffers and goes to sleep, waiting for interrupt (USB disconnected)
13:49:54HClunsupported database version 2 :)
13:50:07Bgrand also makes sure that the disk doesn't spin down
13:51:05*Bgr looks @ the wiki
13:51:53 Quit mrlala (Read error: 110 (Connection timed out))
13:54:28*HCl tries to generate a v3 database
13:54:53HCli still wonder what the delay at the end is :/
13:55:21 Join RED_M_CHIU [0] (
13:56:04HClbug :)
13:56:21ashridahanyone got a link to Slasheri's site?
13:56:26ashridahi manage to keep forgetting it
13:57:31thegeekone small thing
13:57:35thegeekin the translations
13:57:42thegeekI don't think things like "MMC"
13:57:45thegeekshould be translated
13:57:58thegeekthe .no translation says "MMK"
13:58:07thegeekwhich imho is just stupid
13:58:18thegeekI think it should still be MMC
13:58:24preglowyeah, i know
13:58:25thegeeksince that is the accept industry standard
13:58:27preglowi haven't botjered to fix it
13:58:27HClyea, i agree, but then, i don't care about translation
13:58:29thegeekregardless of language
13:58:33thegeekneither do I
13:58:40thegeekI just saw it;)
13:58:43HCl :)
13:58:50amiconnashridah: That's what the irc log is for...
13:58:56ashridahamiconn: mmhf.
13:58:58preglowthere's a lot about that translation i'm not satisfied with
13:59:08preglowbut i won't ever use it myself, so i don't want to use time on it
13:59:11thegeekI would not worry
13:59:20thegeekI doubt it's used much;)
13:59:27thegeekI would certainly never use it
13:59:34ashridahamiconn: maybe if they were searchable
13:59:57MoosCamaroashridah :
14:00:14thegeekyou're too late, muahaahah!
14:00:24RED_M_CHIUasfridah: ctrl+F to search logs?
14:01:11HClthat was *WAY* too easy :D
14:01:14HCli love maintainable code
14:01:24HClupgrading rockbox code to v3 was a matter of changing 1 simple define
14:01:24HCl :)
14:01:26thegeekyou lazy bastard;p
14:01:30ashridahRED_M_CHIU: there's one file per day
14:01:48preglowyou should consider writing maintainable code all the time, then
14:02:05thegeekgod damn
14:02:07HClpreglow: thats what my intention was with the java version of songdb :)
14:02:09thegeekI love the new coldplay album
14:02:21HClas for the changes in the java version, 3 lines. in one file :)
14:02:30HCloh wait
14:02:32HCl4 lines
14:02:36HCl3 in one file, 1 in another
14:02:36HCl :)
14:02:42HCl(rockbox database version number)
14:03:39thegeekColdplay - Talk
14:03:55thegeekin fact
14:05:34ashridahmusic doesn't stop when i plug in the usb jack ;)
14:05:49thegeekwhy complain?
14:05:57*ashridah suspects there's impending doom
14:06:10thegeekenjoy it while you can;p
14:06:13ashridahthegeek: because when the watermark hits, the code's probably goign to attempt to use the disk.
14:06:23thegeekI know I know
14:06:40thegeeksomething bad will happen
14:07:10thegeekright now I could kiss that coldplay dude
14:07:17HCli found a bug...
14:07:28ashridaha bug? NEVER!
14:07:56preglowi'm streaming some "lovely danseband-musikk from the tv
14:07:57thegeeksquash it for us HCl
14:08:29thegeekI'll refrain from saying what I thought
14:08:34thegeeksince it could be considered offensive
14:08:44thegeekand I don't want that... now do I?
14:08:57preglowit's more or less the only genre of music i don't like
14:09:10amiconnBugs tend to come in heaps. If you squash one, you'll uncover the next
14:09:21thegeekIf I ever begin listening to it
14:09:31thegeekremind me to kill myself
14:09:36thegeekI mean
14:09:39thegeekgod damn
14:09:41thegeekit's horrible
14:09:47thegeekit's the devil incarnate
14:09:57preglowyes, indeed
14:10:06amiconnThe overwritten recording was a really tricky one. Fortunately Bagder implemented logf()...
14:10:38HCli have a lot of badly tagged files o_o
14:10:53preglowHCl: freedb masstagger is your friend
14:10:56thegeekI don't
14:11:02HClpreglow: where do i get that?
14:11:05thegeekeverything I own is more or less nice now
14:11:22thegeekyou poor ignorant fool
14:11:33thegeekmasstagger forever.
14:11:38HCloh crud.
14:11:48thegeek(I'm really just trying to convey my fanaticism)
14:11:51HCli shouldn't have selected a song, should i
14:11:54MoosCamarofoobar god father, tag & rename...
14:12:28thegeekgodfather has a nice way to interface with scripts)
14:12:35thegeekbut foobar2000 is way better imho
14:12:42ashridahneat. rockbox HEAD and slasheri's patch have a progress counter
14:12:42thegeekI just love TAGZ
14:13:14HClthis can't be right.. :/
14:18:45HClbackend is failing on files that do have a tag...
14:18:45ashridahhm. okay. rockbox's progress screen has three numbers. the time into the song, the total time, and another time in brackets
14:19:03ashridahwhat the hell is the third time?
14:19:16Bgrremaining ?
14:19:18thegeek(insert bad joke)
14:19:32amiconnThis is just the default wps, and the numbers within the brackets are not a time
14:19:35ashridahBgr: it's bigger than the other two, and doesn't change
14:19:43HClmy backend does not read id3v1 but does read id3v2
14:19:45ashridahokay, what's a WPS?
14:19:46HClwth is up with that XD
14:19:48amiconn...but the current and total tracks in the playlist
14:20:04amiconnWPS = While Playing Screen
14:20:07thegeekspeaking of wps's
14:20:11thegeekhas anyone made one for the iriver?
14:20:18ashridahstrange then. it changed to a new song, but it's still on 4:14
14:20:21amiconnIt is user configurable; what you're seeing is just the default
14:20:25ashridahit was 4:14 last song too
14:20:51amiconnMaybe there's a bug in Slasheri's track change code
14:21:07ashridahentirely possible :)
14:22:21ashridahah, so the playback progress screen is configurable by a config file?
14:24:43preglowrockbox bundles no other wps screens than the default onm?
14:25:06amiconnNo, why should it?
14:25:48thegeekwhy not?
14:25:54thegeekit's not like they take up that much space
14:26:10thegeekand there are always noobs that won't understand it
14:26:34preglowsure looks like there'll be alot of space left on the h1x0 wps
14:26:37amiconnWe could offer the ones in the gallery as a 'wps download pack'
14:27:04amiconnpreglow: yup
14:27:27ashridahindeed, need to make sure a) we can use that space b) loading one for iriver doesn't break older players
14:27:44ashridahinfact, didn't someoen say the iriver remote is roughly the size of the lcd's on archos players?
14:29:03Bgrashridah as width - yes
14:29:14Bgrbut it's smaller in height
14:29:16ashridahah, so as height, no?
14:29:40Bgr as usual ;)
14:30:06Bgr Bitmap 112x64 Bitmap 112x64 Bitmap 112x64 Bitmap 112x64 Bitmap 112x64 Bitmap 112x64
14:30:29Bgrremote LCD is Bitmap 128x64
14:30:47Bgri was wrong
14:32:45preglownone of the archos players have hold switches???
14:33:31Bgrwhy for H3x0 is written that they don't have wake-up alarm ?
14:33:44amiconnpreglow: No hold switches
14:33:49preglowthat's bad
14:33:52amiconn...and I don't miss a hold switch
14:34:03preglowi sure like hold switched when i keep the player in my pocket
14:34:58amiconnI certainly won't put my player in my pocket, so maybe that's why... Apat from that, an archos jukebox would be a bit large for doing that
14:34:59preglownice when i've got it in a bag as well
14:35:04Bgrpreglow it can be done in software (except the case you don't want to turn it on occasionally)
14:35:18preglowBgr: yes, of course, but then you'd need fancy button combinations :P
14:35:24amiconnThere is software button lock in rockbox
14:35:31amiconnI almost never use it
14:35:35Bgrpreglow as in most GSMs ...
14:35:38preglowyes, indeed
14:35:52preglowi've had my cellphone switch off its hold when i carry it in my pocket
14:36:08pregloweven though that should be a bit hard
14:37:03Bgri've never had this luck ...
14:37:42Bgrregardless of the fact that i have some problem with my GSM, causing it to do autorestart sometimes...
14:38:05Bgr(and then, of course, its keypad is not locked)
14:41:09amiconnpreglow: At least on the older archos jukeboxes (player and recorder v1) it's pretty unlikely to accidentally press a button when they are in a pocket/bag, due to the bumpers and battery compartments
14:41:47preglowmight be, i've never actually seen one
14:41:49amiconnOf course this isn't true on the Ondio, and that is where I use button lock occasionally
14:42:31preglowon h1x0 you're bound to need it
14:42:44preglowthe joystick gets pressed all the frigging time when you've got it in a tight pocket
14:43:05HClstupid java backend does not read id3v1 tags
14:43:09HClwhat kind of crap is that.
14:43:27HClit reads id3v2 though
14:43:27preglowreading id3v1 is the very epitome of trivial programming
14:43:59HCli probably need a better reader backend, maybe one self-written
14:44:01preglowit's just a 128 byte struct
14:44:09HClfixed offset?
14:44:21amiconnlast 128 bytes of the file
14:44:38HCldoesn't sound hard to add to this. but first i want to see whats up with this..
14:44:52preglowyou can code a id3v1 reader in five minutes
14:45:42HClwell, there was one, but it doesn't support ogg :/
14:46:21HCldoes that support vorbis?
14:47:07preglowif it does, it sure as hell doesn't mention it
14:47:44HCl :/
14:53:35 Join edx [0] (
14:55:53 Quit webguest62 ("CGI:IRC (EOF)")
15:00:49 Join Sucka [0] (
15:10:58 Quit t0mas ("brb... time to test my new kernel...")
15:18:59***Saving seen data "./dancer.seen"
15:21:02preglowSlasheri: have you updated the patch in any way?
15:23:25Slasheripreglow: nope, i think i will wait it will be commited to cvs.
15:24:28MoosCamaroit's possibly than Linus don't appear before monday
15:24:45SlasheriMaybe i will do some bugfixes soon but not any new features before the commit
15:25:40MoosCamaroplease do ;)
15:26:59Musicmadanybody care to make a bin available somewhere? :)
15:28:08SlasheriHmm, perhaps i should make also flac and wav codecs available
15:28:29amiconngtg, cu
15:28:39 Join mrlala [0] (
15:28:53 Part amiconn
15:32:44 Join tucoz [0] (
15:33:43 Join leftright [0] (
15:34:00tucozMusicMad, I can put a zip up, but beware, it's highly unstable atm
15:34:21tucozdon't know if anyone has any objections to that
15:34:48Musicmadtucoz: thanks a lot. Decided to try to build my own though - let me get back to you if it fails :)
15:35:15leftrightdoes r-box for the h140 have a EQ now ?
15:36:04Musicmadwhich wiki page is the best to follow if I want to build on windows?
15:36:17Musicmadhad some problems with the one describing the use of Erics mirror.
15:37:40 Part leftright
15:37:44 Join t0mas [0] (
15:38:12tucozIsn't the build instructions for linux/cygwin ok?
15:38:17Musicmadand the alternative one seems to be dead.
15:38:51Musicmadtucoz: I think so. It's the instructions on how to be cygwin running with the correct stuff I need.
15:39:04Musicmador maybe don't need.
15:39:20tucozoh, I see.
15:45:19Musicmadwhere can I get this stuff: # Devel - m68k-elf-binutils (latest version is 2.15.96)
15:45:19Musicmad# Devel - m68k-elf-gcc (latest version is 3.3.5)
15:46:29tucozdon't know anything about cygwin. But it seems to be a hassle to get the toolchain working with that
15:46:41Musicmadok thanks
15:46:49MusicmadI'll give it a go.
15:47:07bipak_does m68k-elf-gcc comes with the gcc package (linux)?
15:47:25tucozbipak, no you have to build it yourself
15:47:33Musicmadbipak_: doesn't seem like it (judging from the description of setting up the stuff)
15:47:52tucozcheck out the link I just posted
15:48:02bipak_yeah, i mean if it comes with the gcc source, or is it a special package
15:48:20Musicmadtucoz: i will
15:49:00tucozMusicmad: ;) I answered bipak
15:49:13Musicmadok :) - shows how much I know.
15:49:55preglowSlasheri: it's more than possible linus wont be on before monday, so don't hold your breath
15:51:38Slasheripreglow: Ah, ok.. i will do then some general code improvement meanwhile :)
15:51:40preglowi guess i could commit it, but here probably wants a look at it first
15:51:46preglowhere = he
15:52:05HClwho was the person who had experience with reading ogg tags and mp3 tags
15:52:08preglowso anyway
15:52:17Slasheriyes, i think also he wants to see it
15:53:46 Quit TCK (Read error: 54 (Connection reset by peer))
15:55:11Slasheribtw, it would be great to get the current wps in some form to the remote control or remote control keys working at least
15:55:41MoosCamaroyou can improve your patch in waiting, with A bit of chance he will can appear this night :)
15:56:10tucozSlasheri: I think you deserve spending the weekend at a spa
15:56:27HCli need a proper tag reading library
15:57:18HCli think i might switch my design to c++
15:57:44*preglow cheers for hcl
15:57:49preglowatta boy!
15:57:59HClnot c, c++
15:58:02preglowi know
15:58:06preglowc++ > java
15:58:08HCland i said might
15:58:18HClwell, c++ is more powerful due to having the control of C
15:58:23tucozisn't c plus plus frowned upon from the rockbox crew?
15:58:25HClbut its more messy also, having no garbage collector
15:58:29preglowc++ is more powerful due to not begin retarded
15:58:36preglowHCl: never use ordinary pointers in c++
15:58:47HClthen what am i supposed to use?
15:58:56preglowHCl: use pointer classes that handle the deallocation for you
15:59:14preglowboost::shared_ptr, for example
15:59:24preglowor std::auto_ptr, but that's not as good
15:59:50HClwhat does that do / how does it work
16:00:35preglowlike, shared_array<int> my_int_array(new int[4096]);
16:00:42preglowfrom then on, you use it like normal
16:00:57preglowand just forget to deallocate it, it will deallocate the memory in its destructor
16:01:13HClgot any documents on this?
16:01:16HCli don't exactly know c++
16:02:29preglowi sorely wish boost was standard in c++
16:02:32preglowat least parts of it has an abundance of documentation
16:02:38preglowc++ _sorely_ needs a good library
16:02:51 Join asdsd____ [0] (
16:03:22 Part asdsd____
16:03:27tucozpreglow, have you used the boost lib and would recommend using it?
16:04:06preglowHCl: that should give you an idea of what they are about
16:04:14preglowtucoz: i would very much recommend using parts of it, at least
16:04:27preglowtucoz: the smart pointers are excellent, as is the portable file system library
16:04:28HClnow i'm starting to think i should just code id3v1 support ;/
16:04:45preglowHCl: if that's all that's stopping you, do that
16:04:46HClin fact
16:04:48preglowHCl: it's so incredibly simple
16:04:49HCli'm sure thats easier
16:04:49HCl ;P
16:04:49 Quit Chamois (Read error: 104 (Connection reset by peer))
16:05:17tucozpreglow: as in writing file system stuff cross-platform?
16:05:45preglowtucoz: yes
16:06:18 Join Chamois [0] (
16:06:24tucozpreglow: that sounds absolutely great. One of my concearns with the stuff I'm doing right now
16:06:41preglowi had a concern with that in the ansi viewer i am/was writing as well
16:07:18preglowit's also got great stuff like platform independent threading
16:07:32tucozI compiled boost a couple of months ago, but got scared and bailed out. Guess I'll have a look at it again :) thanks for the tip
16:08:11preglowit's not scary at all, granted that you know c++ properly
16:08:20tucozWas writing a parser, but ended up with using perl instead
16:08:34preglowboost actually has a parser generator
16:08:40preglowworks great, but is very, very slow to compile
16:08:50preglowthanks to template magic
16:08:56MusicmadShould I worry about these: make[1]: /cygdrive/c/Private/rockbox/tools/convbdf: Command not found
16:08:56Musicmadmake[1]: *** [/cygdrive/c/Private/rockbox/build/firmware/sysfont.o] Error 127
16:08:56Musicmadmake: *** [all] Error 2
16:09:02tucozyes, that's why I checked it out in the first place, and the regex tokenizer
16:09:10preglowMusicmad: 'make' in the tools dir
16:09:12ashridahdamnit. player just crashed while playing music. got a good 3/4ths of an hour out of it at least
16:10:16Musicmadthanks preglow
16:10:24tucozpreglow, can't say I know cpp properly. Still in the java-to-cpp phase
16:10:52tucozwhich is...well..interesting
16:11:35preglowthen you probably need to read up on templates, if you don't know them already
16:12:30tucozYep, templates are nice.
16:12:46HCljava got templates recently, i think..
16:12:47tucozBut, java 1.5 also has templates right?
16:12:50Musicmadpreglow: after I run make - is the iriver file created then?
16:13:03tucozMusicmad: make zip
16:13:27preglowMusicmad: should be
16:13:35tucozthen you get a to copy to the player and unzip
16:13:42preglowMusicmad: that is, after you go back to making in the build dir again
16:13:55Musicmadmake: *** No rule to make target `zip'. Stop.
16:14:06preglowyes, both java and c# have gotten template like abilities, after they realized they screwed up in not having them
16:14:48tucozyes it sure was casting-hell in pre 1.5 java
16:16:03ashridahhm. about the only thing missing from this is tone control
16:17:15preglowmissing from rockbox?
16:17:28tucozMusicmad: Do you succed?
16:17:40preglowwhat kind of tone control, you mean bass/treble?
16:18:07ashridahpreglow: working tone control i mean. at least, it doesn't seem to change mid-playback. it might kick in at the start, i haven't checked.
16:18:10tucozsucceed i.e.
16:18:16ashridahi know the options are there
16:18:55preglowashridah: if you mean bass/treble controls, i'm waiting for a patch from austriancoder
16:18:59preglowashridah: i've already fixed it
16:21:13Musicmadpreglow: any thought on the make error?
16:21:38preglowMusicmad: you are standing in the build dir again, yeah?
16:22:46 Join sox [0] (
16:23:07soxhey all
16:23:18soxis slasheris's sound api in cvs?
16:23:52 Join webguest92 [0] (
16:23:58ashridahsox: not yet
16:23:59soxi got the impression that some people were actually playing mp3s on their irivers, was i wrong?
16:24:01ashridahsox: requires vetting
16:24:18ashridahsox: there are patches floating around, they're a little out of date
16:24:34soxbut working?
16:24:56ashridahi'm onto my 4th song from decks, drums and rock and roll
16:25:46soxso where do i find these patches?
16:27:09Musicmadpreglow: no :) - i'll try there.
16:27:21ashridahkeep in mind that one has to reboot the player if one hits 'stop', and it occasionally cuts out, AND you need to adjust the volume once before it kicks in properly, plus a bunch of other issues:
16:27:29preglowashridah: that's been fixed
16:27:35preglowashridah: the stop issue, that is
16:28:13ashridahah. so it has
16:28:32tucozpreglow, that beeping I experience has also disappeared
16:28:32ashridahof course, when the player starts again, bits of the old buffer are still there
16:29:02ashridahand it still wigs out occasionally
16:30:29tucozhmm, weird. rockbox refuses to play certain mp3's
16:31:29preglowwhat kind of mp3s?
16:31:37pregloware you certain it's just not you pressing 'stop' enough time? :P
16:31:39HCli found an id3v1 reader for java..
16:31:50preglowHCl: by now i could have coded an id3v1 reader five times
16:31:59preglowHCl: i seriously recommend you just do it yourself
16:31:59HClyea but i wasn't looking
16:32:03tucozyes, It refuses to play the same mp3's after reboot
16:32:07HClnah, i got something.
16:32:38preglowtucoz: let slasheri have it, then :P
16:37:41 Quit mrlala (Read error: 110 (Connection timed out))
16:38:03HCli found something even better
16:38:14tucozSlasheri: are you there?
16:38:19HClan Iriver DB generator written in java
16:38:22HClsupporting mp3 and ogg
16:38:29HCli'm just gonna plug it into the rockbox db generator
16:39:33tucozpreglow, would you like to try and see if it's just me doing something wrong?
16:39:49HClthis is great.
16:40:29 Quit ashridah ("sleep")
16:40:51HClhaha. i love java :)
16:41:18HCllimited though :/ i'll need to add routines for playtime..
16:41:43soxwhere in the rockbox file structure do i need to place slasheris patches and what's the patch command? patch -p1 < snapshot.diff ?
16:42:23preglowcodec* -> apps/plugins, codeclib* -> apps/plugins/lib playback* -> apps/, patch -p0
16:43:06preglowtucoz: there's nothing fancy to it, if it wont play for you, it won't play for me either, the only trick i know of is pressing stop several times before playing
16:44:21tucozok, yes, but it seems you know a lot more of sound-files and mp3's than me. Might be something with libmad maybe?
16:44:53Slasheritucoz: yes, could you send the file to me?
16:45:08preglowyes it might, but slasheri is in a much better position than me to find out what's wrong
16:45:15tucozand 2.mp3
16:46:07Musicmadzip warning: name not matched: rockbox.iriver
16:46:13Musicmadsomething is not right :)
16:46:13Slasheritucoz: thanks, i will take a look at that soon
16:46:23tucozSlasheri: thank you
16:46:24HClcan someone tell me what the proper way would be to calculate a song's playtime in seconds?
16:46:28HClogg/mp3 ?
16:46:58SlasheriHCl: mp3s playtimes are calculated by firmware/id3.c at least
16:47:21SlasheriThere is length field or something that has the song length in ms
16:47:25SlasheriJust take a look at that
16:48:36HClwell, i need it in java :/
16:48:49Slasherii think you can quite easily port the code..
16:50:08HClfair enough
16:50:14HCli need ogg too though
16:50:59HClthat code is not portable >.<
16:52:21HClat this rate i'll end up using two backends
16:52:27preglowit's a quite simply formula
16:52:32HClone for tags, one for other stuff.
16:52:44HClwhats the formula?
16:53:05preglowthat's the song length in samples
16:53:30HCli want seconds
16:53:39preglowwell, divide it by sample rate, then
16:53:49HClis it just frameRate*(filelength/frameSize) ?
16:54:15HCli suck at math.
16:54:19preglowyou're aware you're bumping into fixed point here again, yes? :P
16:54:25HCli am not.
16:54:30HClthis is for the java db generator
16:54:34preglowyes you are, song length in seconds is rary an integer
16:54:40HCli don't care
16:54:43HClit doesn't have to be exact
16:54:53HCl+- 1 second doesn't matter
16:55:09preglowi think you should save the song length in samples, actually
16:55:21preglowand let the conversion to seconds be dealt with by whatever wants to display it
16:55:25HClfilelength/(frameRate*frameSize) ?
16:55:33preglowyou can't use file length
16:55:39HClyes i can o.o;
16:55:43preglowthere might be all sorts of other data in the file
16:55:48preglowlike a huge id3v2 blob
16:55:57HClwell, i don't know how to deal with that.
16:56:02HClthis isn't going anywhere..
16:56:05preglowyou need the number of mp3 frames in the file
16:56:11preglowthe mp3 frame size in samples is a constant
16:56:23HCli don't have the number of frames
16:56:28HClits a generic interface
16:56:30HClits not limited to mp3
16:56:59HCli'll just merge the two backends i have.
16:57:01preglowwell, how do you propose to do it then?
16:57:14preglowyou need codec info to figure it out
16:57:26preglowthe only other solution is using file size, and that quite simply doesn't cut it
16:57:34HClyes, i'll just merge my two backends.
16:57:40HClbut that will result in tags being read twice
16:57:42HClwhich is a waste of time.
16:58:13preglowi think mp3 frame size is 1152 samples
16:58:24preglowi think you're trapped in oo hell right now
16:58:30HClnot really o.o;;;;
16:58:36preglowthere's no reason tags should have to be read twice
16:58:43preglowyou're just trapped in a bad abstraction
16:58:48HCli know that. i have one backend that does not read id3v1
16:58:51HCli don't know why.
16:59:04HCli have another backend thats great to read tags and fast, supports ogg, mp3, asf
16:59:08HClbut it doesn't give extra info
16:59:11HCllike song duration
16:59:13HClwhich i do need
16:59:22preglowyou might have to query the codec for that info
16:59:29preglowi don't know how mp3 players normally deal with it
17:00:03Musicmadpreglow: any idea why make zip says this: zip warning: name not matched: rockbox.iriver
17:00:14MusicmadI suspect I haven't even build the .iriver file yet :)
17:00:23HCli'll just use both backends..
17:00:30preglowMusicmad: have you done 'make' ?
17:00:39Musicmadin the tools dir though
17:00:47Musicmadshould I do one in the build dir too?
17:00:50preglowforget tools
17:00:54preglowbuild is where you need to be
17:00:59preglowtools has been accounted for now
17:01:01*Musicmad forgets tools
17:01:14preglowsomeone needs to fix the bloody makefile
17:01:29Musicmadso I'm in the build dir now.
17:06:23 Quit webguest92 ("CGI:IRC (EOF)")
17:06:49*bipak_ just compiled rockbox :>
17:07:11 Join spiralout [0] (
17:08:01 Quit tvelocity ("Leaving")
17:09:00preglowwhat do you cygwin guys do to compile rockbox these days?
17:09:06preglowis that even accurate anymore?
17:09:17preglowgcc 3.3 will not work for current rockbox thanks to dumb
17:11:28 Quit sox ("Time to move on...")
17:17:48bipak_is there any step by step doc, für getting rockbox on the ihp?
17:19:03***Saving seen data "./dancer.seen"
17:19:48t0masno... not afaik
17:20:04Musicmadwell there are some claiming to be :)
17:20:10t0maswell.. it's easy
17:20:16t0masdownload the firmware from the iriver site
17:20:25t0masand the fwpatcher.exe from the wiki
17:20:32bipak_im on linux
17:20:33t0masand unzip a daily-build zip to the iriver
17:20:35preglowyou flash it with the bootloader and unzip a rockbox zip onto your player
17:20:36preglowthat's that
17:21:03t0masbipak_: ok, most people asking that question use windows :)
17:21:18t0masI have a firmware file for you
17:21:52bipak_well, i've compiled rockbox successfully, didn't know that there is allready a bin available *g*
17:21:56t0masand HCl had one on his webspace too...
17:22:09t0masoh ok.. then you can make your own bootloader
17:22:11t0masw/o trouble
17:22:48bipak_thanks :>
17:23:27 Quit spiralout ("ChatZilla [Firefox 1.0.4/20050511]")
17:23:36 Join spiralout [0] (
17:23:54bipak_"Be careful! Do NOT attempt to build your own bootloader from CVS unless you know for sure what you're doing."
17:24:22tucozbipak, just get the bootloader.bin from the page
17:24:56tucozand follow the descramble, mkboot ... steps
17:24:57 Quit spiralout (Client Quit)
17:25:04t0masyes... that should be enough
17:25:05 Join spiralout [0] (
17:25:09bipak_k :)
17:25:25*HCl goes to give it a spin
17:25:38HClit might be slow as heck, but at least it'll be correct
17:28:08HCloh, its not that much slower, actually o.o
17:28:30tucozbipak: remember to check the md5 value of your patched .hex file, that it's the same as that on the wiki
17:28:30 Join DomZ [0] (
17:28:39DomZplop everybody
17:28:52bipak_tucoz: ok
17:34:51 Join Stryke` [0] (
17:35:03bipak_md5sum's match :)
17:35:51Bgrbipak_ then just go on :)
17:36:20*bipak_ jitters
17:37:11HClwhat should i do with files that i did get a tag on, but failed to read bitrate/playtime?
17:37:28HClskip them or fill in invalids for playtime/bitrate?
17:37:40*Bgr suggest the last
17:37:47HClyea, i'm tempted to that too.
17:37:54Bgrmore logic :)
17:38:04Bgrlogical i mean
17:49:56*HCl bites windows
17:50:00 Quit t0mas ("brb... kernel replacement again :)")
17:50:00HClwhat kind of bullshit is that.
17:50:06HClC:\programming\songdbv3j\classes>move rockbox.id3db G:\.rockbox\
17:50:06HClThe file exists.
17:50:39preglowit's already there
17:50:43preglowmove won't overwrite
17:50:47HClthats crap
17:50:54preglowyes, but that's the way it is
17:51:15preglowwindows command line has degraded steadily since the days of dos 6.22
17:52:07HClgee. i suddenly have 4 more albums under air
17:52:08 Join t0mas [0] (
17:52:17bipak_it ruuuuns
17:52:18HCli think the new backend is loads better for reading tags ;p
17:52:44bipak_yeah smaller fonts, that good
17:53:03 Join bagawk [0] (~Lee@bagawk.user)
17:56:07HClthis new songdb works great
17:56:18HClwhat people can give it a spin for me?
17:56:26HCl(after i've recompiled with 1.4)
17:57:05t0masI can test now
17:57:10HClokay, sweet. hold on just a second
17:57:32HClwhat version of java do you have?
17:57:58t0masdon't know
17:58:06HClthats fine
17:58:07HClhold on
17:58:12DomZjava -version
17:58:19t0mastomas@athlon tomas $ java −−version
17:58:19t0masUnrecognized option: −−version
17:58:22t0masoh wait
17:58:39t0masI tried -v and −−version not -version :)
17:58:40t0mastomas@athlon tomas $ java -version
17:58:40t0masjava version "1.4.2-01"
17:58:40DBUGEnqueued KICK t0mas
17:58:40t0masJava(TM) 2 Runtime Environment, Standard Edition (build Blackdown-1.4.2-01)
17:58:40t0masJava HotSpot(TM) Client VM (build Blackdown-1.4.2-01, mixed mode)
17:59:05HClthats an old version :X
17:59:07bipak_so how can i start the games? :>
17:59:12preglowoh, how portable java is
17:59:22HCli'm not even gonna be able to compile for that, since sun removed the sdk for 1.4 :/
17:59:34CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
17:59:34*preglow sings the java portability song
17:59:40*t0mas hates java...
17:59:48t0masI just have it because some stupid websites want me too
17:59:51DomZlike DomZ :)
17:59:53HClt0mas: upgrade it? ;/
18:00:01HClat least to 1.4.2_08
18:00:08HClunfortunately, i can't compile for lower :/
18:00:25DomZinstall the last because java 1.4.X has a lot of security flaws
18:00:30DomZthe last is 1.5.0
18:00:43t0masgentoo comes wit 1.4
18:00:48*t0mas looks at the testing list
18:02:14HClyou'll have to get 1.5 :/
18:02:32tucozbipak: if you want to check out the plugins, press a/b
18:03:21bipak_ahhh :)
18:03:38*HCl hits himself
18:04:16HClwindows has crap filesystem handling.
18:04:20bipak_verry nice guys :)
18:04:37bipak_the display seems a bit to slow for pong :(
18:05:07CoCoLUSwhy shouldn't it work with 1.4 ?
18:05:13tucozbipak:Don't know if this is set as standard, but you should also enable view supported files, in general settings->file view->show files
18:05:58tucozto be able to play gameboy games, if your're interested in that.
18:06:50t0masHCl: downloading
18:07:02t0masbut sun restricted their download server to 100 KB/s I guess
18:07:35t0masbecause it's at perfectly 100.0 KB/s since it started
18:09:27HCli'm giving my last version of songdb one more testrun
18:10:13tucozHCL: It seems that sun has an archive of old java versions here:
18:10:23t0masroot@athlon bin # java -version
18:10:23t0masjava version "1.4.2_08"
18:10:23t0masJava(TM) 2 Runtime Environment, Standard Edition (build 1.4.2_08-b03)
18:10:23t0masJava HotSpot(TM) Client VM (build 1.4.2_08-b03, mixed mode)
18:10:36HCland i just compiled for 1.5.. but okay..
18:10:40t0masthat new enough?
18:10:41HCli'll recompile everything
18:10:45 Quit thegeek (Read error: 54 (Connection reset by peer))
18:10:49t0masI can download 1.5 too...
18:10:58t0masI'll do
18:12:10bipak_wow there are diffrent fonts...
18:13:23bipak_does the display controller offer these fonts?
18:13:29t0mas1.5.0.03 HCl?
18:13:35t0mas(that's the one downloading now)
18:13:38CoCoLUSif you compile for 1.4, it should work for 1.4 and 1.5, right?
18:13:56HClt0mas: sure
18:14:03CoCoLUSthen why... :)
18:14:18HClCoCoLUS: part of my backend already uses 1.5 only code :/
18:14:35HCli can adjust it, but its also (on windows) a pain to compile everything
18:14:43HCli should move my dev environment to linux soon
18:14:55t0masanybody knows something about the creative I-Trigue 3200? ( )
18:15:01CoCoLUScan't the 1.5 compiler spit out 1.4 compatible bytecode from 1.5 only source?
18:15:04t0masgood/bad ?
18:15:07HClCoCoLUS: dunno.
18:15:17HClit can
18:15:38HCli'll generate an 1.4 version soonish.
18:15:54HClfirst i'd like someone else to test it too
18:15:57HCllet me upload the last version
18:16:12t0masyeah... 1.5 is installing
18:16:13 Join mrlala [0] (
18:16:16t0masso I can test in a few minues
18:16:22HCl :)
18:16:39HClsize almost doubled with the new tagreader backend :/
18:17:04tucozHCl: what shall I do to try it?
18:17:49t0masroot@athlon tomas # java -version
18:17:49t0masjava version "1.5.0_03"
18:17:57t0masHCl: url?
18:18:04 Join thegeek [0] (
18:19:10bipak_tucoz: what means incompatible model?
18:19:21bipak_tucoz: doesnt work with ihp?
18:19:31tucozbipak: correct
18:19:34HClrun with java -jar SongDB.jar
18:20:08HClpreferably, run it on your iriver..
18:20:12tucozthat means that the plugin or feature is not yet supported on the iriver rockbox
18:20:18HCli'll go and update the perl script and commit db v3
18:20:24t0mastomas@athlon tomas $ java SongDB.jar
18:20:24t0masException in thread "main" java.lang.NoClassDefFoundError: SongDB/jar
18:20:29t0maswhat am I doing wrong? :)
18:20:52bipak_ive downloaded the zip package, doesnt the fw in it support sound yet? :)
18:21:09t0masjava -jar SongDB.jar /mnt/iriver
18:21:18*t0mas starts looking for his iriver...
18:22:23t0mastomas@athlon tomas $ java -jar SongDB.jar /mnt/iriver
18:22:23t0masSorting artists..
18:22:23t0masSorting albums..
18:22:23***Alert Mode level 1
18:22:23t0masSorting songs..
18:22:25t0masSorting files..
18:22:27t0masException in thread "main" java.lang.NullPointerException
18:22:29t0mas at TagDatabase.calcLimits(
18:22:31t0mas at TagDatabase.writeDatabase(
18:22:33t0mas at SongDB.main(
18:22:55tucozI get some Resetting to invalid mark
18:23:23t0masHCl: good luck fixing ;)
18:23:26t0masI'm off to eat
18:24:09tucozBut except from that it seems to work nice
18:25:39tucozwhere do I put rockbox.id3db?
18:26:01HClhehe, thanks.
18:26:11HCltucoz: .rockbox
18:26:26t0masHCl: any idea why mine crashes
18:26:32HClit doesn't find any songs.
18:26:42HClat all.
18:27:19HCli dunno, it doesn't find any songs o.o; did you mount it?
18:27:27t0masyes I checked
18:27:37 Nick Sucka is now known as Sucka`away (
18:27:54HClit might be java's mount handling thats not correct
18:28:01HCltry /mnt/iriver/. ?
18:28:14HClor running it from in there with .
18:29:18tucozHCl: So, this db-file is not functional on the target right now or?
18:29:23HCli need to go get food, bbiab
18:29:23t0masgot it
18:29:27HClt0mas: what was it?
18:29:35HCltucoz: yes, i *just* committed v3 into cvs
18:29:44t0masusers need +x
18:29:46HClt0mas: doh.
18:30:04t0masweird default...
18:30:14HClcan you test the resulting database, see if you find any flaws? if you check out cvs, it should be able to browse it
18:30:19t0mas" Resetting to invalid mark"
18:30:20HClsearchengine will probably mess up or crash
18:30:24HClyea, those are semi normal
18:30:29HCli'll probably get them out soon
18:30:38HClit means the first backend failed reading the tags
18:30:47t0masah ok
18:30:56HClwhich is mostly for duration, bitrate, etc..
18:31:15tucozHCl: sure, no toolchain here. Have to wait for the bleeding edge to finish
18:31:30HClthere's a compilation for v3 on my ftp
18:31:42HCluse that
18:31:45HCli'll be back in a while
18:31:48HClneed to buy dinner
18:31:50t0masArtist with longest name (44) :Cesoul Allstars ft. Carl Henry And Cecile
18:31:50t0masArtist with most albums (47) :<various artists>
18:31:50t0masAlbum with longest name (56) :Life Is What Happens While Your Busy Making Other Plans
18:31:50***Alert Mode level 2
18:31:50t0masAlbum with most songs (109) :The Da Vinci Code
18:31:51***Alert Mode level 3
18:31:57t0maslooks like it works
18:32:10HCl :) can you test it on your player?
18:32:14 Quit tucoz ("CGI:IRC")
18:32:15t0masyes, after dinner
18:32:17HClk :)
18:32:50 Join tucoz [0] (
18:36:21 Quit bagawk ("Leaving")
18:40:10tucozHCl: I don't think it sorts albums quite right. Like for an artist I get from top to bottom: Metallogy(Disc 3), Painkiller, Metallogy(Disc 4), Metallogy(Disc 3)...Angel of retribution, <All songs>
18:41:38 Part Musicmad
18:41:52***Alert Mode OFF
18:42:41tucozHCl: And also some artists show up with no songs, even when entering <all songs>
18:43:55tucozHCl: regarding those metalogy albums, it's because I have the same song on two different locations on the disc
18:44:05tucoz...with the same tag
18:48:12tucozHCl: I don't know what this is, but I have an artist called Hallucinogen and when I enter that artist I get two entries: Hallucinogen and <all songs>.
18:48:34tucozHallucinogen shows up with 3 songs, and <all songs> show 1
18:49:46tucozNote that all this is when I browse for Artists
18:50:09tucozHaven't tested Albums, Songs, or Search
18:50:53t0mastucoz: maybe all 3 songs have Hallucinogen as the album tag
18:51:00t0masand only 1 has the artist tag right?
18:51:53 Join XavierGr [0] (~XavierGr@
18:52:06XavierGrHi all!
18:52:09tucozMight be
18:52:51 Quit DomZ ("CGI:IRC (EOF)")
18:53:56tucozt0mas, but in that case, they shouldn't have been under Hallucinogen in the first place (if artist tag was wrong) or?
18:54:49tucozAs I am browsing for artists
18:58:42tucozBut all the files that are tagged correctly are working as they should. Even oggs. Very nice.
19:11:28 Part tucoz
19:11:58Slasherinext/prev track feature has been implemented soon :)
19:12:14Slasherinow away a moment ->
19:12:34MoosCamaroexcellent ;)
19:13:05MoosCamaroSlasheri: have you updated your patch?
19:13:48Slasherinope, you can check the file change times from the web page where the patch is
19:14:14MoosCamarook thanks
19:15:06preglowSlasheri: just try not to introduce any bugs
19:15:31t0maspreglow: ?
19:15:40preglowt0mas: we'd like this in cvs soon
19:16:01t0masoh ok
19:19:05***Saving seen data "./dancer.seen"
19:19:56 Quit Stryke` (Read error: 145 (Connection timed out))
19:31:56 Quit mrlala (Read error: 110 (Connection timed out))
19:36:21preglowSlasheri: reading is quite a bit slower than in iriver firmware, it seems
19:37:04Slasheripreglow: Please also note that it will load much more tracks at once than original firmware does
19:37:40Slasheripreglow: It will fill all available space in the buffer, where stock firmware loaded only three songs
19:38:06preglowSlasheri: only three? i thought it loaded until it couldn't fit a whole song until the buffer anymore
19:38:15preglowuntil = into
19:38:31Slasherinope, it will load only three at a time
19:38:41XavierGr3? omg! why do such thing?
19:38:58Slasherirockbox can load 10 songs
19:39:21Slasheriand that value can be also set to almost anything but i think 10 is quite reasonable amount
19:39:21XavierGrwhat about liitle tiny ones
19:39:31Slasheri10 is the current limit
19:39:31 Join tucoz [0] (
19:39:37preglowSlasheri: why not load as many as you can?
19:40:04Slasheripreglow: we have to pre-allocate space to store the track_info structure
19:40:15Slasheriand it's hard to place that structure into the file buffer
19:40:21preglow10 should be more than ample
19:40:57tucozSlasheri: did you try those mp3's that didn't work for me?
19:41:22preglowSlasheri: it's worth feeding them to foobar to see what it says, foobar logs errors to its own console
19:42:16HCltucoz: ah, albums within artists, yea, i need to sort those
19:42:16Slasheritucoz: i haven't had time yet. i will try those soon
19:42:33Slasheripreglow: that's a good idea
19:42:38tucozok, I can check later on when I get home what foobar says also
19:42:51HCli'll also fix the artist without songs. hold on.
19:43:45Slasheritucoz: but i have already a feeling where the problem might be. And that can be fixed
19:44:18tucozSlasheri: Great
19:45:22*tucoz away
19:45:25HClokay, both bugs fixed. anything else?
19:45:41*tucoz back
19:45:46 Join Musicmad [0] (
19:45:50HClgood :p
19:45:51tucozHCl, did you see that other bug
19:46:00HClwith artists without songs
19:46:04tucozWith that Hallucinogen stuff
19:46:29tucozI have no idea what that is
19:46:48HCli don't quite get how that can happen either.
19:47:04HCli'd say its just some incorrect tagged files though
19:47:09tucozNo, me neither
19:47:39tucozBut, isn't it strange that <all songs> reports only one, but the album shows three?
19:47:42Musicmadhi guys
19:48:02HClno, that should be supported for albums with multiple songs
19:48:09HClthe old database code is a bit dodgy
19:48:19tucozhmm, ok
19:48:32tucozAhh, I might understand why that happens
19:48:40XavierGrI want to make a dos bat file that deletes a folder but when i run it asks for 'Y' or 'N' can i make that automated?does anyone knows something like an echo command?
19:48:58HCljust. echo.
19:49:25tucozIf Artist-> Hallucinogen has made an album called Hallucinogen, and on that album there are three songs, but only one is made by Hallucinogen
19:49:38tucozThat has to be it
19:49:42XavierGrafter the rd command i have typed "echo y" but it doesnt work....i will try again
19:49:47MusicmadXavierGr: got my pm?
19:50:03HClif they made an album with multiple artists, it should've put it under <various artists> though
19:50:14 Quit Chamois ("Leaving")
19:50:17tucozHCl, ok. Let me check
19:50:24MusicmadXavierGr: trying to start at chat with you :)
19:50:47XavierGrwell I dont see a chat seesion here!
19:51:06Musicmaddamn :) - did you manage to build rockbox on windows recently?
19:51:24XavierGryes I did!
19:51:38preglowtucoz: what, he's never released an album with the name hallucinogen...
19:52:48tucozpreglow: I don't know who he is, something a friend had that I copied. Never listened to it
19:53:06 Join Stryke` [0] (
19:55:16tucozIt seems to be a Shpongle album called Hallucinogen
19:56:15preglowshpongle never released an album called hallucinogen :>
19:56:19tucozHCl: I can send that album to you, and see what's wrong with it.
19:56:24tucozA stress test
19:56:28preglowshongle is hallucinogen + raja ram
19:56:33XavierGrdoes anyone remembers what is the url for the Devkit?
19:56:52HClnah, bit busy at the moment...
19:57:08tucozpreglow, ok. The songs are called. Shamanix, L.S.D. magic ...
19:57:45HCli think these things can be sorted with the dirisalbum options and stuff like that.
19:57:58HCli'll add support for those..
19:58:01XavierGrok I found it
19:58:09tucozHCl: that's true, you have a bunch of options. :D
19:58:41preglowtucoz: that's from hallucinogen - twisted
19:59:58tucozpreglow: ok, any good?
20:00:42tucoznevermind, I'll listen to them.
20:00:52tucozgot to go, bye bye
20:00:56 Part tucoz
20:00:58preglowit's psychedelic/goa
20:01:57MoosCamaroL.S.D :) very hallucinogen :)
20:03:33preglowgood old classic goa trance
20:04:27MoosCamaro:) it need narcotics ;)
20:07:29preglowit probably does benefit from it, yes
20:22:21HCltoo bad my testcase left
20:22:24*HCl prods t0mas
20:23:57 Nick Sucka`away is now known as Sucka (
20:27:24*HCl prods people
20:30:47preglowno java, no can do :/
20:30:51 Quit Stryke` (Read error: 145 (Connection timed out))
20:31:52HClpreglow, can you test the perl version though?
20:32:00HClthe one in current cvs
20:34:25 Join coob [0] (
20:34:35preglowjust /mnt/myplayer ?
20:34:39coobhow many (bogo)mips are you guys getting on the iriver? −−path
20:34:49preglowcoob: we never did a bogomips check
20:34:54preglowcoob: since we can see no reason for it
20:35:30preglowcoob: but if you paste me some quick code, i might indulge you
20:35:32coobwell, do you have any rouch ideas?
20:35:54preglowno, not apart from motorolas market speak
20:36:15preglowall the performance numbers i've got are codec timing info
20:37:12coobit's not importaant, i was just curious
20:37:57preglowonly place where performance matters much for us, is in codec usage, and that utilises the cpu in other ways than a bogomips measurement, so it'd be pretty pointless
20:38:11preglowcoob: you banging against some performance walls?
20:38:35coobnah we finally figured out how to set clock speed on the PP5020 :D
20:38:45coobour mp3/aac stuff is working wonderfully now
20:38:47HClits about time
20:38:49HCl ;p
20:38:52preglowhow wonderfully?
20:38:58coobso now we need to code a proper player interface/buffering system :)
20:38:59preglowyou figured out how to use the second core yet?
20:39:19coobi've yet to do some proper test, just audial atm :)
20:39:26cooboh we know how to use it
20:39:35coobour video player uses it
20:39:57preglowhow well does the helix aac decoder run, btw?
20:40:01coobbut the guy working on sticking the audio stuff has diassapeared, last we heard he had our aac stuff running on it
20:40:14preglowlike how far above realtime?
20:40:29coob123% for 128 AAC-LC on one 75mhz core
20:40:41preglowpretty decent
20:40:48cooband thats using the slow memory
20:40:55coobthere's so many optimisations we've yet to make
20:40:56preglowdid you try out he-aac? i'm very curious as to how that'll perform
20:41:24coobafaik that's for low bitrate/speech type stuff isn't it?
20:41:31preglowno portables support it yet
20:41:35preglowwould be cool to be a first
20:41:52preglowHCl: now what?
20:41:56coobwell it only needs a few lines changing with the helix stuff
20:42:03HCloh wait.
20:42:05HClyou only had ogg
20:42:06HCl :X
20:42:09coobso i'm sure someone will send in a patch, it doesn't really bother me
20:42:12preglowHCl: no, i have a bunch of mp3s as well
20:42:17coobi'm working on ogg now
20:42:18HCltest it on your iriver
20:42:21coobfor the ipod :)
20:42:27preglowHCl: done, now what?
20:42:36coobwe have the newer gens (photo/4g) running faster than the older gens now! :D
20:42:37preglowcoob: i have no ipod, so really doesn't benefit me much, heh
20:42:40HCljust see if you can find any bugs
20:42:40 Join Stryke` [0] (
20:42:59HCli thought you didn't have the newer generation working yet?
20:43:05preglowcoob: anywho, i'll set up faad for h1x0 soon, so i'll fund out
20:43:09preglowHCl: they do now
20:43:17coobit's just labeled as unsupported to keep the noobish hordes away from bugging us
20:43:24preglowHCl: yeah, i've run songdb, how do i actually use the db on the player?
20:43:28preglowHCl: i've never done this before
20:43:29coobfaad is floating point i thought?
20:43:37HClgeneral options -> file view -> show files -> id3 database
20:43:40preglowfully fixed point, as far as i can see
20:43:40HCli just tested it myself
20:43:41HClworks fine
20:43:52HClaside from having all the drawbacks of the perl version
20:44:13HCli'll commit my java version to cvs
20:44:19preglowHCl: i don't have that entry
20:44:25Suckawhich project are you coob?
20:44:33HClyou need to put rockbox.id3db in .rockbox
20:44:35HCland reboot the player
20:45:12Suckaah right
20:45:40preglowcoob: i'm willing to bet your core performs better than our coldfire with no tweaks
20:46:05pregloweven mp3 was waaay below realtime for us without any opts
20:46:22coobaac is less cpu intensive than mp3
20:46:25coobat least for us
20:46:48HClhave we even tried aac?
20:46:52preglowthis sucks, i don't have time to put up faad now, but i really want to
20:47:07preglowHCl: bah, i forgot i have to recompile
20:47:13HClyou make time :)
20:48:23preglowloads of 'no album tag' entries
20:49:43preglowapart from that it looks fine
20:50:06HClthe is just. still the old
20:50:19preglowknow <-
20:50:22HCli'm happy that the work i've done last time to make the code more maintainable has paid off
20:50:35HClit only took 2 line changes to update rockbox to the new format
20:50:40HCland another 2 in
20:50:41HCl :)
20:52:15HClpreglow: what os do you run?
20:52:32HCl64bit linux? would an gcj 32bit elf binary work for you?
20:53:19preglowif it's statically linked: maybe
20:53:25preglowit might work anyway
20:53:28preglowworth a shot
20:54:09HCli'll try to generate one
20:55:04HClafter i update my server
20:58:49preglowand i'm giving faad a spin...
20:58:53 Join mrlala [0] (
21:00:50preglowseems like sbr really needs a large amount of code
21:02:17preglowno, i'm not doing this
21:02:19preglowgo away
21:02:49preglowHCl: you commited a _binary_ ?
21:04:47preglowyeah, i know, needs extra libs, etc, but i just don't feel it belongs in cvs
21:05:51 Quit XavierGr ()
21:11:33HClpreglow: no
21:11:46HClits both binary and source
21:12:11HClbut the file itself is binary, yes.
21:12:24preglowboth binary and source?
21:12:34pregloweh? you've stuffed the .java files into the jar?
21:12:49preglowwell, keeping it as a binary in cvs completely destroys the meaning of keeping it in cvs
21:13:06HCli can commit it as a whole directory structure
21:13:26preglowpersonally i don't want you to commit it at all, seeing as how it's temporary, and it's java :P
21:13:36HCli dunno, i like it
21:14:01preglowof course, i can understand that, but i've got a feeling the majority of the rockbox people are with me on this one
21:14:22preglowif not, it isn't the first time i'm wrong ;)
21:14:42HCl ;p
21:14:51HClstill, i want it somewhere safe
21:14:56HClin case i have major computer crashes
21:14:57HClor whatever
21:15:04HClmaybe i should add it as attachment on the wiki instead?
21:15:30rashersounds better
21:16:03rasherit's not like you'll be able to view diffs on it or anything anyway
21:16:18HClnot in its current form
21:19:08***Saving seen data "./dancer.seen"
21:20:25 Quit mrlala (Read error: 145 (Connection timed out))
21:28:04HClpreglow: gcj chokes on it, by the way :/ something with stringbuilder...
21:30:22HClwhich its simply missing from its api o.o.
21:30:27HClhow handy.
21:31:41 Quit spiralout ("ChatZilla [Firefox 1.0.4/20050511]")
21:34:49 Join asdsd____ [0] (
21:35:28 Part asdsd____
21:42:20 Join DomZ [0] (
21:46:38 Part Musicmad
21:49:50 Join tucoz [0] (
21:50:01 Join mrlala [0] (
21:51:08tucozHCl: Songdb seems to be working fine now
21:52:02HCl :)
21:52:07tucozI only experienced a problem when I tried to run it so that rockbox.id3db would end up on target
21:52:16HClnew version has commandline options support
21:52:17tucozIt just never stopped
21:52:20HCland i fixed some stuff
21:52:24HClit has a long delay at the end
21:52:28HCli don't know what causes this yet.
21:52:38tucozIt was fine when I run it at home
21:53:01tucozand copied the file from there to target
21:53:13HCleither way, you might want to grab the new version
21:53:15HClthe one on the wiki
21:53:19tucozI did
21:53:27HClk :)
21:53:29tucozabout 20 minutes ago
21:53:37HClgood to hear it works okay
21:53:45HCli changed the old as well
21:53:54HClit works, but still as crappily as it used to ;p
21:54:29tucozIt's really nice, though I seldom use db-browsing. Haven't really gotten used to it because of irivers fw. Might do now thoug
21:55:39tucozHCl: regarding that Hallucinogenic stuff. Files that are so crappily named and lacks tags, deserve to end up in weird places
21:55:57HCl :P
21:59:44 Join zezayer [0] (~chatzilla@
22:04:08Slasheripatch updated
22:04:23Slasherifew problems fixed and now next track button should work too :)
22:07:22*tucoz tries it immediatly
22:07:52bill2or3anyone have a H340?
22:08:18preglowSlasheri: forgot the url
22:08:46preglowwhat problems have been fixed?
22:09:15Slasheristart/stop better fixed (now previous sounds from buffer)
22:09:31Slasherino need to touch volume when playing first track
22:10:06preglowexcellent, excellent, i will try it right away
22:10:11Slasheriand some small changes i can't remember
22:10:17preglowis much work needed to make wps work for non-mp3?
22:11:00Slasherimaybe.. quick hack would be to put ogg information to id3 fields but that is not a good solution
22:11:09SlasheriI don't want to touch the wps code
22:11:58tucozdid I do something wrong? got this
22:12:07tucozHunk #1 FAILED at 458.
22:12:11tucoz1 out of 1 hunk FAILED −− saving rejects to file firmware/sound.c.rej
22:12:16tucozpatching file firmware/export/pcm_playback.h
22:12:22tucozfresh cvs co
22:12:43preglowactually, here as well
22:13:06pregloweverything bloody fails here
22:13:21tucozhmm, I can change that manually though
22:13:57preglowi wonder if i forgot to unapply the old patch
22:14:00preglowi always do that
22:15:50preglowi can't apply this patch at all
22:15:52pregloweverything fails
22:17:52 Join Mirfle [0] (
22:18:30tucozcannot build it :(
22:18:58tucozargh, it was make in tools again
22:19:14preglowbut why does my patch fail miserably?
22:19:24SlasheriHmm, just a moment
22:19:30tucozhmm, I made a fresh cvs co
22:20:14preglowlooks there are carriage returns here
22:20:22tucozin sound.c?
22:20:47tucozafter the patch? for me aswell, and also in the .diff
22:21:02preglowlooks there are crlf everywhere in the patch
22:21:07SlasheriI made a new patch against fresh cvs
22:21:09preglowno wonder it fails, then
22:21:11SlasheriPlease try patching again
22:21:35rasherthe one I got seems to be working
22:21:41preglowsounds.c hunk still fails
22:21:46 Part coob
22:21:46preglowbut everything else worked now
22:21:50rasheryes, applied that manually
22:25:06Slasheripreglow: really strange if the patch has crlf..
22:25:16SlasheriI have done everything on linux
22:25:16preglowlooked like it here
22:25:17preglowthis one doesn't
22:26:16preglowget a bit of clicking when i switch track
22:26:20preglowother than that it works great ;)
22:26:21tucozcool, nex track really works
22:26:47*tucoz gives Slasheri a beer
22:27:07 Quit zezayer ("Chatzilla [SUSE 1.0.4-1.1/20050511]")
22:27:39tucozSlasheri: It also plays those songs that didn't work before
22:27:39Slasheritucoz: hehe, thanks ;)
22:28:00Slasheritucoz: great :) i also checked that it should play those
22:28:05tucozbye bye iriver fw.
22:28:34preglowmy, how this rocks
22:28:40rasherHrm, when stopping, I get a fade, and then a split-second of full-volume
22:28:45rasheris this known and accepted?
22:28:58preglowthat's wps behaviour
22:28:59Slasherirasher: will be fixed later
22:29:02preglowi switched off fading
22:29:17rasherSlasheri: alright, just wanted to check
22:29:35preglowthere's quite a big of ui lag
22:29:59preglowhow much do you read from the disk every time?
22:30:38preglowi kept 'volume down' pushed during the last five seconds of disk reading
22:30:45SlasheriHmm, 256 KiB.. That might be too much
22:30:54SlasheriTry go to settings menu etc.
22:30:58preglowafter i released it, it continued to adjust the volume for ten seconds
22:31:01SlasheriIt has no lag
22:31:07SlasheriI am not sure what causes that
22:31:18 Join pfavr [0] (
22:31:18preglowwell, we'll find out
22:31:22Slasheriyeah :)
22:31:32 Join zezayer [0] (~chatzilla@
22:31:49preglowseems like there's a value lying in the dac when you stop a track
22:31:56tucozit's really great that my iriver is turning into a rockbox.
22:32:07tucozgtg, bye
22:32:13 Part tucoz ("Leaving")
22:32:39preglowi think vorbis plays pretty marginally
22:34:44 Quit mrlala ("-=SysReset 2.53=-")
22:36:02rasherHrm.. I think I'm going to miss the iRiver's distinction between "shuffle dir" and "shuffle all"
22:36:20rasherunless I'm missing something, rockbox doesn't implement this, does it?
22:37:22preglowi'm not going to miss it period
22:38:04zezayerbut u can shuffle a playlist, just mkae a pl of eash folder
22:38:24Stryke`or use the tag database
22:41:41rasherwell shuffle dir is covered
22:41:54Stryke`including or excluding sub-directories?
22:42:12rasherdon't know
22:42:23 Quit Sucka ("a bird in the bush is worth two in your house")
22:44:01 Quit t0mas ("good night")
22:51:53 Quit pfavr ("ChatZilla 0.9.61 [Mozilla rv:1.7.8/20050513]")
22:51:57 Part Mirfle
22:52:35DomZsomeone can share the new rockbox compilation ? :)">
22:54:05DomZthanks rasher
22:54:09HClhey rasher
22:54:18HClwhat do you have on file hashing?
22:54:20rasherhas center-scrolling and a silly icon-change
22:54:46rasherHCl: the same I've had all along..">
22:55:18HClokay :)
22:55:37rasherI think that's just about where my abilities end
22:55:58Stryke`rasher: thats the latest build with Slasheri's new patch?
22:56:53rasherStryke`: yes (and markun's center-scrolling patch)
22:57:01Stryke`awesome, thanks
22:57:38HClwe should get that one in cvs too
22:57:43HCli'll prod him about it on monday
22:58:36rasherI'm not sure if anything's holding it back
22:59:11 Join amiconn [0] (
23:00:44HClglitch on font change
23:00:47HClbut otherwise its fine
23:01:14rasheractually.. DomZ/Stryke` .. codecvorbis.rock and codecmpa.rock is missing from that zip
23:01:45DomZhumm yes i have no sound :)
23:01:50DomZwith my mp3 :p
23:02:50DomZis center-scrolling defined as an option ?
23:04:38rashernot sure - codec*.rock uploaded to
23:05:02DomZthanks i will try now
23:07:00DomZhumm always no sound
23:08:27rashertry changing the volume
23:08:54SlasheriDomZ: Try also setting on the fade out feature (that might have a bug)
23:09:46rasherDomZ: the codecs need to be in .rockbox/codecs/
23:10:50DomZarf i am stupid :)
23:11:08DomZi forgot to put the file in codecs directory
23:11:17DomZit's run fine now
23:12:27DomZjoystick left is for restart playing from the beginning ?
23:12:46DomZand joystick right for the next song ?
23:12:54Slasheriit doesn't work yet
23:13:03DomZthat's work for me
23:13:03HClthat reminds me...
23:13:06Slasheriyep, joystick right will change to next track
23:13:11 Quit zezayer (Read error: 60 (Operation timed out))
23:13:17HCli'm wanting to change the way rockbox navigates through the database
23:13:50DomZjoysrick left for on the first second on song
23:14:12DomZjoystick left works for the first second of playing
23:14:24Slasheriyes, it _should_ skip to the previous song
23:14:30DomZah ok
23:14:37SlasheriBut that's still highly untested and may crash rockbox
23:14:38Stryke`i love how fast the usb bootloader mode boots
23:15:30Stryke`and center scrolling is very nice
23:16:01DomZi have notice any difference :)
23:16:37*Stryke` is listening to one of my favorite albums gaplessly ;-)
23:19:09***Saving seen data "./dancer.seen"
23:21:27DomZthis is quite unbelivable the improvements that has been made in just a few weeks !
23:21:48Stryke`yeah, it really seems we're in the home stretch
23:21:51preglowyes, and slasheri has the credit pretty much alone
23:22:04HCl :)
23:22:18 Join ghode|afk [0] (
23:22:20 Join zezayer [0] (~chatzilla@
23:23:07Stryke`I04:IllInstr at 32F4019E
23:23:08DomZis it due to better coding experience from Slasheri ?
23:23:35SlasheriStryke`: Did you do something before that?
23:23:44Stryke`view current playlist
23:23:51SlasheriAh, it has a known bug
23:23:59SlasheriAnd it will cause that..
23:24:27Stryke`wish there was a better way than just letting the battery run out =\
23:24:42rasherStryke`: use the reset button?
23:24:50Slasheriyep, reset will work
23:25:09Stryke`didn't work the last time i had a crash
23:25:14Stryke`ill give it a shot
23:25:24Slasherii have needed to use reset hundreds of time while coding this :)
23:25:42DomZare the id tad always read or there are stored somewhere ?
23:25:48DomZid tag
23:25:53SlasheriStryke`: reset should always power cycle the device
23:26:14SlasheriMaybe you didn't push it from a right position
23:26:17Stryke`maybe i just didn't hit it
23:27:01SlasheriDomZ: it's stored on track info structure. Read while buffering the songs
23:28:01Stryke`hard reset worked like a charm
23:28:22Stryke`i got used to iRiver resume
23:28:46Slasheribetter resume will be implemented later
23:28:50Stryke`Doesn't play on boot, but if you press play, it will resume
23:28:58Stryke`oh i have no doubt
23:29:01Slasherinow going to sleep, night :)
23:29:02HClyea, i like that option..
23:29:10Stryke`thanks for your work
23:30:54Stryke`any other known crashes?
23:31:24Stryke`that i may avoid
23:32:31*amiconn starts working on the LED issue (MMC icon displayed on iriver)
23:32:56amiconnI need something to code while a huge test recording is running on archos...
23:37:38 Quit zezayer (Read error: 60 (Operation timed out))
23:41:19 Join zezayer [0] (~chatzilla@
23:46:53 Join zezayer_ [0] (~chatzilla@
23:49:43preglowamiconn: so, had any look at memcpy and pals? with all the copying thats going on in tremor, we really need a fast copy :P
23:50:17amiconnThe first asm thing I'll do for iriver will be memset()
23:50:28amiconnI expect a speed increase of 8x
23:50:44preglowoh yes
23:50:49amiconn(maybe even more)
23:50:52 Quit Harpy (Read error: 60 (Operation timed out))
23:51:13amiconnI still have my mem(set|cpy) test plugins from back when I did this on archos
23:51:56HCl :)
23:52:11rasherHrm, a chicago.bdf with other characters than a-z would be nice
23:52:13amiconnThey do test both correctness and speed
23:52:20HCli'll look at some minor rockboy optimizations after database stuff is fully operational
23:52:29amiconnrasher: xtal-14 is based on chicago
23:52:36amiconn...and extended by me :)
23:52:58preglowHCl: sounds like a good opportunity to brush up on the asm skills
23:53:03preglowHCl: 68k asm is quite comfortable
23:53:08amiconnpreglow: I just found a slight optimisation possibility for memset() on archos...
23:53:18rasheramiconn: ah, nice
23:53:28amiconnpreglow: From what I've seen, I'd prefer SH asm
23:53:41preglowamiconn: please spell out some annoying differences
23:53:47preglowfor some freak reason i've started to enjoy asm
23:53:48amiconnI'm missing some nice instructions in m68k
23:54:26amiconnswapping bytes, swapping into another register
23:54:51preglowyeah, exg was unfortunately removed from the set for coldfire
23:55:05preglowyou can still do it with eor, i think, but of course, that's three instructions instead of one
23:55:25 Quit zezayer (Read error: 60 (Operation timed out))
23:55:57rasherhow do i get from the wps/blank screen to the file browser without stopping playback?
23:56:05preglowi thinks
23:56:07 Join ehntoo [0] (
23:56:27rasherthat gives the menu
23:56:30preglowindeed :/
23:56:31amiconnOn archos, the wps <-> browser button is ON
23:56:43amiconnSo I'd guess it is play on iriver
23:57:17rasherIt is indeed, but now the pause icon is showing :-\
23:57:25rasheroh well
23:57:30amiconnI did not yet try Slasheri's playback patch for 2 reasons
23:57:39rasherGuess that should be remapped - possibly to Select
23:57:46amiconn(1) Not much time to do that, busy with other rockbox stuff
23:58:15amiconn(2) There's not that much need for me, since I already have multiple units with rockbox being capable of playback
23:58:21amiconnI really have the choice...

Previous day | Next day