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 2011-08-28

00:00:22bertrikOh, and diagnostic mode confirms that it has 8 MB of DRAM on a 16-bit bus just like the older AMSv2 clips
00:01:57 Quit ender` (Quit: The difference between something that can go wrong and something that can't go wrong is that, when the thing that can't go wrong goes wrong, it's usually impossible to get at or fix.)
00:04:32***Saving seen data "./dancer.seen"
00:04:43 Join Keripo [0] (
00:08:06 Join XavierGr [0] (~xavier@rockbox/staff/XavierGr)
00:16:30 Join ender` [0] (
00:26:54 Quit n1s (Remote host closed the connection)
00:33:16 Quit HaimN (Remote host closed the connection)
00:41:07 Quit Keripo (Ping timeout: 240 seconds)
00:41:20 Join powell14ski [0] (
00:43:18 Quit ender` (Quit: Knowledge is power. Power corrupts.)
00:51:13 Quit fs-bluebot (Quit: timeout?)
00:51:19 Join fs-bluebot [0] (
00:56:58 Join Keripo [0] (
00:57:24 Quit sideral (Remote host closed the connection)
00:57:51 Join sideral [0] (~sideral@rockbox/developer/sideral)
00:57:55 Quit matsl (Ping timeout: 272 seconds)
00:58:08 Quit sideral (Remote host closed the connection)
00:59:58 Join sideral [0] (~sideral@rockbox/developer/sideral)
01:10:54 Join matsl [0] (
01:17:56 Quit Judas_PhD (Quit: This is a quitting message)
01:18:11 Quit matsl (Ping timeout: 272 seconds)
01:25:52 Join Rob2222 [0] (
01:29:27 Quit Rob2223 (Ping timeout: 240 seconds)
01:31:30 Join matsl [0] (
01:32:58 Quit bertrik (Ping timeout: 245 seconds)
01:43:14 Quit matsl (Remote host closed the connection)
01:49:47 Join Rob2223 [0] (
01:53:17 Quit Rob2222 (Ping timeout: 264 seconds)
02:01:17 Quit sideral (Ping timeout: 245 seconds)
02:04:34***Saving seen data "./dancer.seen"
02:14:11 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
02:35:25 Quit Horscht (Ping timeout: 240 seconds)
02:38:07 Quit Rob2223 (Read error: Connection reset by peer)
02:38:22 Join Rob2222 [0] (
02:44:53 Join ChickeNES [0] (
02:55:25 Quit kadoban (Ping timeout: 260 seconds)
02:59:50 Join Horscht [0] (~Horscht@xbmc/user/horscht)
03:00:04 Quit Jerom1 (Quit: Leaving.)
03:23:59 Quit mgue (Ping timeout: 260 seconds)
03:25:15 Join mgue [0] (
03:34:28 Join mystica555 [0] (
03:55:20 Quit JesusFreak316 (Read error: Connection reset by peer)
04:04:38***Saving seen data "./dancer.seen"
04:06:26 Quit mikroflops (Read error: Operation timed out)
04:18:49 Quit Keripo (Read error: Connection reset by peer)
04:26:56 Quit nick-p (Quit: Leaving)
04:31:41 Quit pixelma (Disconnected by services)
04:31:43 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:31:45 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:31:53 Quit amiconn (Disconnected by services)
04:31:54 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:32:13 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:33:02 Quit antil33t (Ping timeout: 245 seconds)
04:34:12 Quit [Saint] (Ping timeout: 265 seconds)
04:39:56 Join [Saint] [0] (~st.lasciv@
04:40:11 Join antil33t [0] (
04:48:30 Quit TheSeven (Disconnected by services)
04:48:36 Quit antil33t (Ping timeout: 250 seconds)
04:48:36 Quit [Saint] (Ping timeout: 250 seconds)
04:48:45 Join [7] [0] (~TheSeven@rockbox/developer/TheSeven)
04:55:12 Quit XavierGr (Read error: Connection reset by peer)
04:55:44 Join [Saint] [0] (~st.lasciv@
04:55:55 Join antil33t [0] (
05:09:51 Quit antil33t (Read error: Network is unreachable)
05:10:05 Join antil33t [0] (
05:37:54 Join kadoban [0] (
05:45:42 Join Rob2223 [0] (
05:49:37 Quit Rob2222 (Ping timeout: 264 seconds)
05:57:54 Join Keripo [0] (
06:02:24 Join nick-p [0] (
06:04:39***Saving seen data "./dancer.seen"
06:13:12 Join robin0800 [0] (
06:20:28 Quit robin0800 (Ping timeout: 250 seconds)
06:21:32 Join robin0800 [0] (
06:36:04 Quit robin0800 (Quit: Leaving)
06:37:44 Join dunkaist [0] (~dunkaist@
06:55:22 Quit nick-p (Quit: Leaving)
08:04:41***Saving seen data "./dancer.seen"
08:09:58 Quit powell14ski (Quit: powell14ski)
08:15:53 Join n1s [0] (~quassel@rockbox/developer/n1s)
08:16:12 Quit amiconn (Remote host closed the connection)
08:16:13 Quit pixelma (Remote host closed the connection)
08:17:11 Join amiconn [0] (quassel@rockbox/developer/amiconn)
08:17:13 Join pixelma [0] (quassel@rockbox/staff/pixelma)
08:21:14 Join nick-p [0] (
08:38:43 Quit nick-p (Quit: Leaving)
09:13:10 Quit Unhelpful (Ping timeout: 240 seconds)
09:15:02 Join stoffel [0] (
09:18:22 Quit jhMikeS ()
09:23:50 Quit Galois (Remote host closed the connection)
09:24:23 Join Galois [0] (
09:39:28 Join nick-p [0] (
09:45:37CIA-14New commit by jethead71 (r30366): Commit work started in FS #12153 to put timing/position information in PCM ...
09:47:07 Join jhMikeS [0] (
09:47:07 Quit jhMikeS (Changing host)
09:47:07 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
09:48:20CIA-14r30366 build result: All green
09:50:54n1sjhMikeS: does the 0002 patch in fs#12240 look good to you?
09:50:55fs-bluebot rbcodec refactoring part 1 (patches, unconfirmed)
09:52:32jhMikeShmmm...haven't check it yet
09:52:55jhMikeSjust one sec
09:53:04n1sno problem
09:55:45 Join matsl [0] (
09:57:39jhMikeShmmm...he's trying to reverse things that were done on purpose, like limiting the dsp yields
09:59:21n1si assume it is so that the library doesn't need to see current_tick but yes i thought it looked like it was done for a reason
10:02:01jhMikeSwhy does the lib need yields?
10:02:59n1swhen used by rb on native targets, the plan is for rb to use the lib
10:04:39jhMikeSwell, sure, on native targets
10:04:44***Saving seen data "./dancer.seen"
10:05:49jhMikeSthe lib ought to include data types
10:10:54*jhMikeS lays some shit on wtachi
10:12:46 Join sideral [0] (~sideral@rockbox/developer/sideral)
10:13:19*[Saint] summons someone to dispatch of "aphoto" from the forums...
10:13:36 Quit matsl (Ping timeout: 252 seconds)
10:19:46jhMikeSa photo of something, or user name aphoto
10:25:31jhMikeSDoes "report to moderator" work cause I just said "spam, spam, spam, spam"? I cannot dispatch of anyone.
10:26:36 Join matsl [0] (
10:36:52 Join bertrik [0] (
10:36:52 Quit bertrik (Changing host)
10:36:52 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
10:39:14 Join y4n [0] (y4n@unaffiliated/y4ndexx)
10:41:08jhMikeSreally odd that the delta table shows r30366 but the build table is lacking it
10:41:46 Join Buschel [0] (
10:41:59jhMikeSmaybe it doesn't like long-winded commit messages :P
10:44:38[Saint]If that was true, sideral would have broken it long ago ;)
10:45:26jhMikeSBuschel: message? I thought I tried building that
10:45:44*Buschel just restart his VM
10:47:28Buschelpcmbuf.c: 1020: error: 'INT_MAX' undeclared
10:47:58jhMikeSwell isn't that nice :\
10:47:58jacekowskiBuschel: limits.h
10:48:26jacekowski#include <limits.h>
10:49:23jhMikeSI could just nix INT_MAX too
10:49:35jhMikeSnot really heavily important there
10:51:50jhMikeSeek, it's wrong anyway, left over from some previous incarnation
10:59:16CIA-14New commit by jethead71 (r30367): Remove INT_MAX from pcmbuf.c. Win32 sim compained about it and it isn't specifically important enough for another #include - it just needs a great ...
11:02:50CIA-14r30367 build result: All green
11:05:23Buschelcompiles fine now
11:05:31 Quit matsl (Remote host closed the connection)
11:07:48Buschelmpc resume takes ages since this change :/
11:10:15jhMikeSit ought not touch resume itself
11:12:46jhMikeSdid I goof something in the codec?
11:13:24BuschelI am not sure, I just rolled back your changes to mpc.c and resume still takes far too long
11:13:42Buschelif I roll back to r30365 everything works as intended
11:13:42jhMikeSI see I left an extra set_elapsed that should have been taken out :\
11:14:23Buschellet me check that change to mpc.c again
11:14:45 Join ender` [0] (
11:15:17BuscheljhMikeS: which one should have been removed?
11:15:48jhMikeSthe one inside the if (mpc_demux....)
11:15:56jhMikeSsince it always must set it now
11:16:09jhMikeSbut that never blocks anything
11:16:21Buschelcompiling now
11:18:02Buschelstill haevily lagged
11:19:40jhMikeSdoes it make any internal assumptions about ci.id3->offset ?
11:23:18Buschelbut it uses ci->read_filebuf, ci->seek_buffer, ci->curpos and ci->filesize
11:24:37jhMikeSno changes were made to handling that
11:28:01Buschelwell, there is a change that uses ci.curpos
11:30:00jhMikeSthat ends up setting id3->offset after a seek completes
11:38:05 Quit stoffel (Ping timeout: 250 seconds)
11:46:22jhMikeSis it laggy only on the sim?
11:46:33BuschelI noly tested on the sim for now
11:48:15Buschelany idea?
11:48:41 Join pondlife [0] (~Steve@rockbox/developer/pondlife)
11:48:54 Quit pondlife (Client Quit)
11:49:20jhMikeSBuschel: not a clue. it's not really making any sense.
11:52:31Buschelmpc_demux_seek_sample() takes way longer... but only when resuming, not when seeking.
11:53:08jhMikeSi did search beforhand for dependencies and id3->offset was used for resume in codecs, not internallyh
11:55:49jhMikeSwhat about some other format?
12:01:09bertrikoh cool, "visionox" is *not* the internal name of the clip zip hardware, it's an LCD manufacturer
12:02:57*jhMikeS keeps typing make in the wrong tree, wondering what's going on with win32 sim :\
12:03:25 Join stoffel [0] (
12:03:26 Join Benz1nko [0] (
12:04:47***Saving seen data "./dancer.seen"
12:05:04bertrikunfortunately I can't find a 96x96 colour display at their website
12:05:34Benz1nkoRe, I have a question about rockbox and cowon d2, is it without primary drive support ? sorry for my english.
12:06:28Benz1nkoprimary usb*
12:10:43BuscheljhMikeS: back again. I tried m4a and mp3, both work fine with your changes. what is special wiith mpc resume/seek? -> mpc parses the full bitstream. so, mpc will read lots of data and requires quite some CPU during resume/seek.
12:11:23BuscheljhMikeS: is it possible that we experience priority issues? or is something waiting actively for seek_complete?
12:11:51Buschelthe other codecs will most likekly just seek to a byte position instead of parsing the whole stuff...
12:15:45jhMikeSonly mpa and musepack buffer with an offest at this time
12:15:55jhMikeSerr, wavepack
12:15:56 Quit Benz1nko (Quit: CGI:IRC (EOF))
12:17:00jhMikeSaudio just takes seek_complete whenever it comes, it's definitely not expected for resume
12:18:12 Quit bug2000 (Ping timeout: 276 seconds)
12:35:52 Join fml [0] (
12:36:30fmlHe-he, I wonder whether the sansa clip zip port will be usable before the player is sold via the german amazon!
12:36:33 Quit fml (Client Quit)
12:37:29 Quit tjb0607 (*.net *.split)
12:37:29 Quit scorche|sh (*.net *.split)
12:37:29 Quit alexbobp (*.net *.split)
12:37:29 Quit Elfish (*.net *.split)
12:37:29 Quit Staphylo (*.net *.split)
12:37:29 Quit GodEater_ (*.net *.split)
12:37:29 Quit user829385 (*.net *.split)
12:37:29 Quit ender| (*.net *.split)
12:37:29 Quit simonlnu (*.net *.split)
12:37:29 Quit TBCOOL (*.net *.split)
12:37:29 Quit niekie (*.net *.split)
12:37:29 Quit Poodlemastah (*.net *.split)
12:37:29 Quit pjm0616 (*.net *.split)
12:37:29 Quit jordan` (*.net *.split)
12:37:29 Quit zu (*.net *.split)
12:37:29 Quit eintopf (*.net *.split)
12:37:29 Quit bylzz (*.net *.split)
12:37:29 Quit ved (*.net *.split)
12:37:57 Join alexbobp [0] (~alex@
12:37:57 Join Elfish [0] (amba@2a01:4f8:100:90a1:abc:abc:abc:abc)
12:37:57 Join Staphylo [0] (
12:37:57 Join GodEater_ [0] (~bibble@rockbox/staff/GodEater)
12:37:57 Join tjb0607 [0] (
12:37:57 Join user829385 [0] (~aoeu@
12:37:57 Join ender| [0] (
12:37:57 Join simonlnu [0] (QfBLewJ3rV@unaffiliated/simonrvn)
12:37:57 Join TBCOOL [0] (
12:37:57 Join niekie [0] (~niek@CAcert/Assurer/niekie)
12:37:57 Join Poodlemastah [0] (
12:37:57 Join pjm0616 [0] (~user@
12:37:57 Join jordan` [0] (~gromit@2001:660:3302:2826:225:90ff:fe20:d9a8)
12:37:57 Join zu [0] (
12:37:57 Join bylzz [0] (
12:37:57 Join eintopf [0] (
12:37:57 Join ved [0] (
12:37:57 Join scorche|sh [0] (~scorche@rockbox/administrator/scorche)
12:38:02 Quit Galois (*.net *.split)
12:38:02 Quit maraz (*.net *.split)
12:38:02 Quit ps-auxw (*.net *.split)
12:38:02 Quit factor (*.net *.split)
12:38:02 Quit ukleinek (*.net *.split)
12:38:02 Quit parafin (*.net *.split)
12:38:03 Quit Hadaka (*.net *.split)
12:38:35 Join Galois [0] (
12:38:36 Join maraz [0] (
12:38:36 Join ps-auxw [0] (~arneb@2001:470:c807:0:1532:4e5f:2ad3:4123)
12:38:36 Join factor [0] (~factor@
12:38:36 Join ukleinek [0] (
12:38:36 Join parafin [0] (
12:38:36 Join Hadaka [0] (
12:38:46 Quit n1s (*.net *.split)
12:38:46 Quit [7] (*.net *.split)
12:38:46 Quit mgue (*.net *.split)
12:38:46 Quit ChickeNES (*.net *.split)
12:38:47 Quit Xerion (*.net *.split)
12:39:04 Join n1s [0] (~quassel@rockbox/developer/n1s)
12:39:04 Join [7] [0] (~TheSeven@rockbox/developer/TheSeven)
12:39:04 Join mgue [0] (
12:39:04 Join ChickeNES [0] (
12:39:04 Join Xerion [0] (
12:39:17 Quit aevin (*.net *.split)
12:39:17 Quit balintx (*.net *.split)
12:39:17 Quit markun (*.net *.split)
12:39:17 Quit efyx (*.net *.split)
12:39:19 Quit bluefoxx (*.net *.split)
12:39:19 Quit desowin_ (*.net *.split)
12:39:19 Quit Slasheri (*.net *.split)
12:39:19 Quit [fred] (*.net *.split)
12:39:31BuscheljhMikeS: if I move the resuming into the playback loop and delay resuming by ~100 frames the resuming is done instantly (as before)
12:39:35 Join aevin [0] (
12:39:36 Join balintx [0] (
12:39:36 Join markun [0] (~markun@rockbox/developer/markun)
12:39:36 Join efyx [0] (
12:39:36 Join bluefoxx [0] (
12:39:36 Join desowin_ [0] (
12:39:36 Join Slasheri [0] (miipekk@rockbox/developer/Slasheri)
12:39:36 Join [fred] [0] (
12:39:43BuscheljhMikeS: there is something fishy
12:39:56 Quit nick-p (*.net *.split)
12:39:56 Quit kadoban (*.net *.split)
12:39:56 Quit [Saint] (*.net *.split)
12:39:57 Quit advcomp2019_ (*.net *.split)
12:39:57 Quit madskiny (*.net *.split)
12:39:57 Quit GeekShadow (*.net *.split)
12:39:57 Quit mystica555_ (*.net *.split)
12:39:57 Quit Utchybann (*.net *.split)
12:39:57 Quit jfc (*.net *.split)
12:40:02Buschelsounds like some statemachine issue
12:40:14jhMikeSI don't seem to be having an issue
12:40:16 Quit ender` (*.net *.split)
12:40:16 Quit dunkaist (*.net *.split)
12:40:16 Quit guymann (*.net *.split)
12:40:23jhMikeSnot with the demo file anyway
12:40:26 Join nick-p [0] (
12:40:26 Join kadoban [0] (
12:40:26 Join [Saint] [0] (~st.lasciv@
12:40:26 Join advcomp2019_ [0] (~advcomp20@unaffiliated/advcomp2019)
12:40:26 Join madskiny [0] (dre@gateway/shell/
12:40:26 Join GeekShadow [0] (
12:40:26 Join mystica555_ [0] (
12:40:26 Join Utchybann [0] (
12:40:26 Join jfc [0] (
12:40:33 Join ender` [0] (
12:40:33 Join dunkaist [0] (~dunkaist@
12:40:33 Join guymann [0] (
12:40:37 Quit y4n (*.net *.split)
12:40:37 Quit sideral (*.net *.split)
12:40:37 Quit ReimuHakurei (*.net *.split)
12:40:37 Quit Guinness` (*.net *.split)
12:40:58jhMikeSwait a minute
12:41:03 Quit antil33t (*.net *.split)
12:41:03 Quit user890104 (*.net *.split)
12:41:03 Quit ruskie (*.net *.split)
12:41:03 Quit Farthen (*.net *.split)
12:41:03 Quit rudi_s (*.net *.split)
12:41:10 Quit crwl (*.net *.split)
12:41:10 Quit jae (*.net *.split)
12:41:10 Quit preglow (*.net *.split)
12:41:10 Quit JdGordon (*.net *.split)
12:41:10 Quit feisar-_ (*.net *.split)
12:41:57 Join antil33t [0] (
12:41:57 Join user890104 [0] (
12:41:57 Join ruskie [0] (ruskie@sourcemage/mage/ruskie)
12:41:57 Join Farthen [0] (
12:41:57 Join rudi_s [0] (
12:41:57 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
12:41:57 Join crwl [0] (
12:41:57 Join jae [0] (
12:41:57 Join preglow [0] (
12:41:57 Join feisar-_ [0] (
12:42:16 Join y4n [0] (y4n@unaffiliated/y4ndexx)
12:42:16 Join sideral [0] (~sideral@rockbox/developer/sideral)
12:42:16 Join ReimuHakurei [0] (
12:42:16 Join Guinness` [0] (
12:45:17 Quit Rob2223 (*.net *.split)
12:45:17 Quit Llorean (*.net *.split)
12:45:17 Quit kugel (*.net *.split)
12:45:34 Join Rob2223 [0] (
12:45:34 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
12:45:34 Join kugel [0] (~kugel@rockbox/developer/kugel)
12:52:03BuscheljhMikeS: look at this ->
12:52:23Buschelresume works as fast as before with this hack
12:52:43Buschelwith this hack the decoder plays a few frames and then resumes
12:53:09Buschelwhy should there be a difference (from the decoder point of view)?
12:57:03jhMikeSnot sure what you mean "plays a few frame then resumes"
12:57:11 Quit nick-p (Quit: Leaving)
13:00:19jhMikeSsamplesdone looks like it isn't initialized
13:01:40jhMikeSwait, oops
13:02:07Buschelthe patch moves the resume code. it is now *inside* mpc.c's do-while-loop. this way the decoder decodes/plays 40 frames (~1 second) and then performs the resume.
13:03:56Buschelbut even this patch will not always resume fast. seems like there is some timing dependency
13:05:17jhMikeSit could be thrashing the buffer around, but then, it would have been doing that for a long time now anyway
13:07:54Buschelwhen decoding 100 frames (~2.5s) before resuming it works more reliable −− there *is* some timing dependenc
13:09:51jhMikeSwell, that would push buffering ahread without needing a rebuffer
13:10:48jhMikeSbut then, nothing should be different in that respect from before
13:13:00Buschel -> line 155: "loop_idx == 0" -> resume takes ages, "loop_idx == 100" -> resume is done instantly
13:16:12jhMikeSis there anything with id3->elapsed that's important?
13:20:41n1scould there be some init that is missing when seeking straight away?
13:21:59Buschelat least not within the codec itself
13:23:08Buschelalso, why should the behaviour of my above patch change with the number of decoded frames before resuming?
13:23:15Buschelmaybe buffering?
13:28:11jhMikeSdumping a few frames in would also cause the thread priority to raise up initially
13:29:16jhMikeSis there album art involved or any other encumberance?
13:29:31Buschelwher is the codec priority set?
13:29:49jhMikeSin pcmbuf.c
13:32:52 Quit stoffel (Ping timeout: 260 seconds)
13:36:24Buscheleven setting PRIORITY_PLAYBACK to 5 does not help
13:39:57jhMikeScould try to see if it's getting loads of buffer callbacks at resume
13:40:45jhMikeSbut at this point, I still see no good reason a pcm buffer change ought to affect this, esp. given what was ruled out already :\
13:41:51Buschelwell, the playback engine was changed...
13:42:36Buscheldidn't you also change some the queue events?
13:44:32jhMikeSit wasn't changed in the resume area
13:45:38jhMikeSno queue events changed here, just how they're posted (via function in playback.c instead of directly in codec_thread.c)
13:46:40Buschelpossibly something moved? e.g. reading metadata? maybe the codec and something else have competing buffer/file access?
13:49:34jhMikeSI didn't move anything, not on-purpose
13:50:05 Join tguinot [0] (
13:51:21jhMikeSmaybe he changes to playing_id3_sync? it's really the only substantial change that is hit upon resume
13:52:31Buschelcan those just be rolled back to the former implementation?
13:54:15jhMikeSremove pcm_play_lock/unlock from that function and it would basically be what it was
13:55:49Buschelremoving those does not make a difference
13:56:34jhMikeSdidn't think so. the rest is transition and updating playing_id3's position info
13:57:58jhMikeSthose are integral to the patch, so can't be rolled by on their own
13:59:04*jhMikeS wonders what been done to buffering itself recently
14:04:49***Saving seen data "./dancer.seen"
14:13:09 Quit sideral (Ping timeout: 240 seconds)
14:15:56 Join stoffel [0] (
14:23:16BuscheljhMikeS: simple test ->
14:23:51Buschelthe patch adds some file-i/o to the aac codec. this behaves like mpc, it's heavily lagged since your changes
14:25:57jhMikeSbut not if before my changes?
14:26:43Buschelyes, if I roll back pcmbuf.c/.h, codec_thread.c and playback.c to r30365 it's working fine
14:26:54jhMikeSeven the aac patch?
14:28:07Buschelfwiw: my CPU-meter says the CPU's are idle during this several seconds of delay
14:30:00jhMikeSyou're saying it read 6553600 bytes quickly if aac patched and using 30365 <<== ??
14:33:16 Join Horschti [0] (~Horscht@xbmc/user/horscht)
14:34:07 Quit Horscht (Ping timeout: 258 seconds)
14:36:20jhMikeSI must admit I'm a bit incredulous
14:36:31Buscheljust test it :)
14:39:27jhMikeSin audio_on_buffering(), in the BUFFER_EVENT_REBUFFER case, check if it mysteriously jumps up between revisions
14:40:30jhMikeSor in buffer_event_buffer_low_callback <= better
14:40:58jhMikeS*buffer_event_rebuffer_callback (haven't slept enough)
14:41:56Buschelwhat exactly do you want me to check?
14:52:48jhMikeScompare the call count when resuming between revisions
14:56:50 Join Jerom1 [0] (~jerome@
14:57:05Buschelneither buffer_event_buffer_low_callback() nor buffer_event_rebuffer_callback() is called in both revisions
15:00:15jhMikeSthen it's not waiting for audio to clear anything and buffering is handling it internally (good)
15:00:50 Join Stummi [0] (~Stummi@rockbox/developer/Stummi)
15:08:27*jhMikeS wonders where the sim disk IO is for the app threads (not fiber threads in particular)
15:09:34jhMikeSactually *for* anything not SDL threads
15:20:52jhMikeSI'm also wondering if SDL threads do anything different
15:24:24 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
15:24:32Buschelwhat do you think of this -> in buffering.c prep_bufdata() the delay is caused by looping in the do-while loop
15:24:49Buschelthis is with your change. the old revision does not have this effect
15:30:54jhMikeSold revision doesn't do it at all, no matter how far one seeks into it?
15:31:01jhMikeSor resumes, rather
15:33:42Buschelusing the old revision it does it very few times when resuming 15-16 MB into a mpc song
15:34:35Buschelwith your changes it is more like several hundred times
15:49:28Buschelmy analysis was not fully correct. in case of delayed resume the "if (h->filerem > 0 && avail < realsize) {" path in prep_bufdata() is entered for *each* call of prep_bufdata(). with the old revision this happening only very few times.
15:52:45jhMikeShmmm...codec got way faster ? :)
15:53:40 Join nick-p [0] (
15:58:36Buschelas I said: timing dependeecies ;)
15:59:03*Buschel is in a period of typo's again
15:59:06jhMikeSit's not really a race per se
15:59:44jhMikeSI suspect you mean this occurs at the resume seek, before pcm is generated?
16:01:12jhMikeSa resume from stopped?
16:01:13 Quit antil33t (Read error: Connection reset by peer)
16:01:28Buschelyes, that is my usecase
16:01:32 Join antil33t [0] (
16:03:09*jhMikeS check if he messed up the pcmbuf_is_lowdata or something
16:04:50jhMikeSrevert the first "if (...) return false; " of pcmbuf_is_lowdata *only*
16:04:54***Saving seen data "./dancer.seen"
16:05:04Buschelgood point, this would also explain why resuming/seeking works after several frames have been decoded (= fill pcm buffer)
16:05:27jhMikeSI think I left something I didn't mean to
16:06:16jhMikeSbut actually, those should refer to the channel, not pcm, which also was something left that shouldn't have been
16:07:29Buschelno change
16:10:47Buschelif I always return false in pcmbuf_is_lowdata() , it works like charm
16:14:07Buschelit also works fine if I replace sleep(1) with yield() in buffering.c, line 728
16:17:23Buschelor if I just remove the sleep/yield section at those lines
16:17:38jhMikeS <= proper change
16:21:34jhMikeSthanks for sticking with that! :)
16:21:56Buschelyou're welcome :)
16:25:40CIA-14New commit by buschel (r30368): Remove obsolete 'ci->set_elapsed()' from mpc.
16:27:01CIA-14New commit by jethead71 (r30369): Restore functionality of pcmbuf_is_lowdata. It fell out of sync since the mixer code and then an incorrect change unintentionally remained in r30366.
16:28:22CIA-14r30368 build result: All green table still isn't keeping new revisions
16:30:46CIA-14r30369 build result: All green
16:57:05 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
16:59:45 Quit Buschel (Quit: ChatZilla 0.9.87 [Firefox 6.0/20110811165603])
17:06:08 Join MethoS- [0] (~clemens@
17:10:21 Quit antil33t (Read error: Connection reset by peer)
17:10:44 Join antil33t [0] (
17:19:47bertrikmeh, the build table is still showing the warnings after my commit.. :)
17:21:16 Quit nick-p (Quit: Leaving)
17:21:38bertrikthose are warnings from the host cc as far as I know, but our build scripts don't mention the host cc version
17:22:12bertrikShould we fix those warnings (or at least make them go away)?
17:38:26 Join Zarggg_ [0] (
17:40:54 Quit Zarggg (Ping timeout: 252 seconds)
17:46:16n1sbertrik: i think it's an ubuntu thing or something, anyway only happens when building tools
17:46:33 Nick Guinness` is now known as Guinness (
17:52:14 Join PurlingNayuki [0] (
17:55:19 Quit PurlingNayuki (Client Quit)
17:55:27 Join PurlingNayuki [0] (
17:55:36PurlingNayukiHi everyone.
17:56:00PurlingNayukiWhat's the use of 'Time scratch'?
17:58:23 Join bug2000 [0] (~bug@unaffiliated/bug2000)
18:04:58***Saving seen data "./dancer.seen"
18:06:58n1sPurlingNayuki: time *stretch*?
18:07:32PurlingNayukiOops, sorry. My typo.
18:08:43n1sit is for speeding up (or slowing down?) audio without changing pitch
18:09:19n1sthe intended use case is for audiobooks and podcasts and other spoken word
18:09:34PurlingNayukiThx, but does it really work?
18:09:57n1si think so but i haven't tried it
18:10:26PurlingNayukiWait a minute please and let me try iy.
18:10:40gevaertsWhy wouldn't it work?
18:10:43PurlingNayuki*it* lots of typo today:P
18:12:10PurlingNayukiWell it does work.
18:12:28PurlingNayukiI tried it before but it didn't work.
18:13:52PurlingNayukiAnd sorry, another question. I read the menu files and found a 'set_sleep_timer' function.
18:14:08PurlingNayukiHowever, I can't find where it's from.
18:14:39PurlingNayukiI searched 'time' for file but just got something useless.
18:20:26PurlingNayukiOK thanks. Finally I found it.
18:20:52PurlingNayukiIt's in firmware/powermmgmt.c
18:22:54 Quit PurlingNayuki (Quit: CGI:IRC (EOF))
18:26:18 Quit stoffel (Read error: Operation timed out)
18:35:46CIA-14New commit by nls (r30370): libtremor: comment out some unused functions and mark some file local functions static, saves a few hundred bytes and might give a tiny speedup.
18:36:43CIA-14New commit by nls (r30371): libtremor: remove some inline cf asm that is no longer needed with the new toolchain, no speed diff.
18:37:45n1sbenchmarking a 128kbps file said r30370 was ~0.1MHz faster but could easily be due to icache effects
18:38:12CIA-14r30370 build result: All green
18:40:35CIA-14r30371 build result: All green
18:52:31 Join ChickeNE_ [0] (
18:53:22 Quit ChickeNES (Ping timeout: 276 seconds)
19:04:10 Quit kadoban (Read error: Operation timed out)
19:08:42 Quit ChickeNE_ (Quit: Computer has gone to sleep.)
19:13:30 Quit Torne (Ping timeout: 260 seconds)
19:17:20 Join z180 [0] (
19:18:04 Join Torne [0] (~torne@rockbox/developer/Torne)
19:20:59 Join TheLemonMan [0] (
19:28:46CIA-14New commit by nls (r30372): libtremor: comment out some more unused functions, make a function param unsigned to simplify generated code, gives a small speedup on cf.
19:29:13 Quit TheLemonMan (Quit: leaving)
19:30:58CIA-14r30372 build result: All green
19:34:07 Quit Jerom1 (Quit: Leaving.)
19:43:00 Join bluebrother [0] (
19:43:00 Quit bluebrother (Changing host)
19:43:00 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
19:44:39 Quit fs-bluebot (Ping timeout: 245 seconds)
19:45:46 Join fs-bluebot [0] (
19:46:17 Join typn [0] (
19:46:19 Quit bluebroth3r (Ping timeout: 245 seconds)
19:46:45typnhi, I wonder what the anynomous SVN address is?
19:47:18bluebrotherthe same as the non-anymous
19:47:21typnI wanted to "borrow" some font-rendering code from rockbox for my project :-).
19:47:31typnoh thanks.
19:48:24bluebrothera bit more about it
19:48:57typngreat, got it, I appreciate it bluebrother
19:55:25 Join stoffel [0] (
19:56:46 Quit Keripo (Quit: Leaving.)
20:01:26 Quit stoffel (Remote host closed the connection)
20:04:59***Saving seen data "./dancer.seen"
20:10:04 Join TheLemonMan [0] (
20:15:45 Join wtachi [0] (
20:20:55 Join Keripo [0] (
20:25:00 Quit TheLemonMan (Quit: leaving)
20:25:22 Join TheLemonMan [0] (
20:26:28TheLemonManJ #xfce
20:43:02z180heytypn look at freetype it is used in ipod OF
20:44:54 Quit typn (Remote host closed the connection)
20:52:01 Join Topy44 [0] (
21:00:51 Join sideral [0] (~sideral@rockbox/developer/sideral)
21:04:46 Quit benedikt93 (Quit: Bye ;))
21:05:11 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
21:13:52gevaertsz180: what do you mean?
21:14:10 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
21:16:13 Join robin0800 [0] (~quassel@
21:22:28 Join pamaury [0] (
21:22:29 Quit pamaury (Changing host)
21:22:29 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
21:25:34z180not you ,sorry I meant that user who wanted the font code
21:26:37gevaertsah, right
21:28:19 Join kadoban [0] (
21:35:04 Quit benedikt93 (Quit: Bye ;))
21:41:50 Join ChickeNES [0] (~ChickeNES@
21:42:21 Quit ChickeNES (Read error: Connection reset by peer)
21:42:54 Join ChickeNES [0] (
21:48:29 Join froggyman [0] (~seth@
21:48:30 Quit froggyman (Changing host)
21:48:30 Join froggyman [0] (~seth@unaffiliated/froggyman)
21:48:43 Quit TheLemonMan (Quit: leaving)
21:49:01 Join TheLemonMan [0] (
22:05:01***Saving seen data "./dancer.seen"
22:09:29 Join liar [0] (
22:17:05 Quit robin0800 (Ping timeout: 240 seconds)
22:17:40 Quit z180 (Quit: ChatZilla 0.9.87 [Firefox 3.6.20/20110803131630])
22:19:22 Quit Stummi (Quit: Bye!)
22:21:53 Quit TheLemonMan (Quit: leaving)
22:22:12 Join TheLemonMan [0] (
22:36:50 Join robin0800 [0] (~quassel@
22:38:45 Quit wtachi (Quit: &)
22:48:48 Quit TheLemonMan (Quit: leaving)
22:49:05 Join ageis [0] (
22:54:42 Join TheLemonMan [0] (
22:55:10 Quit TheLemonMan (Client Quit)
22:56:34 Quit domonoky (Read error: Connection reset by peer)
22:57:12 Join ReimuHakurei_ [0] (
22:57:47 Quit ReimuHakurei (Ping timeout: 240 seconds)
22:59:39 Quit ageis ()
23:00:17GodEater_is there any way to make do it's activities seperately? Like - do all the downloads at once, and then do all the building?
23:00:27 Join TheLemonMan [0] (
23:01:14 Quit robin0800 (Ping timeout: 240 seconds)
23:03:39 Join robin0800 [0] (~quassel@
23:05:58 Quit Galois (Ping timeout: 264 seconds)
23:08:43saratogahuh we have no battery benchmarks for the ClipV2
23:13:16 Quit robin0800 (Read error: Connection reset by peer)
23:17:25 Quit n1s (Remote host closed the connection)
23:35:43 Join john [0] (
23:36:08johnHas anyone here ACTUALLY gotten rockbox to work on their ipod? (Cause I'm finding the task pretty impossible.)
23:36:18 Nick john is now known as evildaemon (
23:36:47bluebrotherevildaemon: a lot of people did
23:37:39bluebrotherinstallation is rather simple. At least for stable ports.
23:37:56bluebrotherbut since you didn't mention which Ipod you're talking about ...
23:38:16evildaemonAnyway, onto an actual issue, I've "installed" rockbox onto a 1st gen black ipod nano, and it won't recognize USB/Firewire input, any ideas?
23:39:35gevaertsUSB or firewire?
23:39:44evildaemonMy thought was that there might be a way to corrupt the firmware partition so I can go back to the (working) "restore your ipod" screen.
23:40:09gevaertsYou can *always* go back to the in-ROM recovery mode
23:40:24evildaemonWhere is this?
23:40:45 Join bertrik_ [0] (~bertrik@rockbox/developer/bertrik)
23:41:46gevaertsYou have to hold a button while booting. I can't remember which one right now. Let me search a bit...
23:42:24 Quit Torne (Ping timeout: 260 seconds)
23:44:45bluebrothershould be Select+Play
23:45:20bluebrotherdoes the Nano1G actually support Firewire? I thought Apple dropped that at some point
23:46:55bertrikI have the nano 1g and there's only a standard apple connector on it
23:47:11 Quit bertrik_ (Ping timeout: 258 seconds)
23:47:32evildaemonI thought there was a connector for USB, and another one for Firewire
23:47:35bertrikand inside it's basically just a portalplayer player as far as I know
23:47:44bluebrotherthat doesn't mean it does (not) have Firewire.
23:48:04 Join Torne [0] (~torne@rockbox/developer/Torne)
23:48:11evildaemonI'm starting to consider the possibility that the USB ports on my creeky machine are just breaking.
23:48:17bluebrotherthe Ipod has the dock connector. There's a cable for Firewire and one for USB, but IIRC newer Ipods don't support Firewire anymore
23:48:45bertrikdoesn't wikipedia know this?
23:49:02*evildaemon is too lazy to check.
23:49:12bluebrotherpossible :) The IPL wiki had a nice table back when it was around
23:49:39bluebrothereither way, how does this "not recognize USB" look like? Does Rockbox actually run? Does the Apple firmware run?
23:50:01evildaemonOne second, gonna try connecting to my laptop.
23:50:04bertrikwikipedia says the nano 1g has USB for data/charging and firewire only for charging
23:51:12evildaemonYour not going to believe the solution.
23:51:19evildaemonMy USB port was broken.
23:51:31*bluebrother isn't surprised
23:51:48evildaemonEither am I, this things gotta be from circa 2004.
23:51:58bertriksaratoga, I think I'll commit some lcd init code for the clip zip tomorrow
23:52:11bluebrotherflaky USB ports is an issue that occurs every now and then.
23:52:24gevaertsWe believe *everything* :)
23:52:28bertrikthe clip zip OF is already prepared to work with two different kinds of LCDs
23:52:41*bluebrother doesn't believe gevaerts :P
23:52:47evildaemonIt's funny, because it was working fine five minutes ago, lol.
23:54:02evildaemonI heard the "disk click of death" the other day, so I need new hardware regardless.
23:58:30 Join nick-p [0] (

Previous day | Next day