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 2010-08-26

00:03:34 Join Hillshum [0] (4ba5f419@gateway/web/freenode/ip.
00:08:00 Quit Devil_Weapon (Quit: KVIrc 4.0.2 Insomnia
00:08:34 Join Dreamxtreme [0] (~Dreamxtre@
00:08:39 Join GeekShad0w [0] (
00:11:10 Quit GeekShadow (Ping timeout: 240 seconds)
00:13:58 Nick eboyer93 is now known as edboyer93 (
00:16:41AlexPfiveofoh: He did, yes
00:17:31fiveofohOkay. It didn't sound like he did on this end. That's good then.
00:19:58 Quit Dreamxtreme (Read error: Connection reset by peer)
00:20:38 Quit stripwax (Quit:
00:22:04 Join madalu [0] (~matt@unaffiliated/madalu)
00:23:04 Quit _s1gma (Ping timeout: 272 seconds)
00:24:08 Join _s1gma [0] (~d.d.derp@
00:25:56 Quit Jerom (Read error: Connection reset by peer)
00:29:45froggymanis it possible to turn off the light on the scroll wheel on the fuzev2 in rockbox? or is it a HW problem?
00:31:14 Join fdinel [0] (
00:31:48 Join Mom [0] (
00:32:31MomDoes the manual contain information on which buttons have what features in various players using rockbox?
00:33:18saratogayes, the keymaps are in it
00:33:28MomFor example, the down button in my player doesn't bring up a menu, but it has a picture on the button saying that's what it does
00:33:41MomI'll check the manual then
00:38:37 Quit pamaury (Ping timeout: 265 seconds)
00:39:34MomDoes the "home" key on the sansa clip count as the "submenu" key? Or does the "down" key mean submenu?
00:39:57AlexPIt tells you at the beginning of the manual
00:39:59MomIn the original firmware, the down key when used at the Now Playing screen gives you a context menu
00:40:03AlexPThere is a diagram
00:40:08 Quit Strife89 (Ping timeout: 240 seconds)
00:40:17AlexPThe OF is irrelevant
00:40:36MomYou'd expect similar functionality though: the play key still plays
00:40:52AlexPThe functionality is completely different
00:40:58AlexPIt isn't possible to map directly
00:41:14AlexPAlso, Rockbox runs on lots of different players and we try to have some sort of consistency
00:41:46MomI understand :)
00:42:06MomI just want to make sure what I'm experiencing isn't supposed to happen
00:42:13Mombefore I see about reporting a bug or getting help
00:42:24MomIt's basically one button that doesn't do anything
00:42:24AlexPRead the manual first
00:42:30AlexPWe have spent lots of time on it
00:42:40Momyes I have it open now −− looking for that diagram
00:43:22MomIn the diagram the button is labeled as "down," however another page talks about "submenu" which isn't labeled to any key
00:43:51MomNevermind, down is submenu
00:44:44AlexPThat diagram is wrong though :)
00:44:46MomOK I have found a bug then: submenu doesn't work on the WPS
00:45:34Momit is a dead key on the WPS, but works as a down key in other places
00:45:43AlexPIdeally the diagram and description should match
00:45:46 Join Strife89 [0] (
00:45:56 Quit DerPapst (Quit: Leaving.)
00:46:07AlexPI wish I could find my damn clip
00:46:37Momhmmm actually...
00:46:50MomI found the problem
00:46:59Momyou have to Click and Hold it down for it to work
00:47:00AlexPThe manual is clearly wrong in the WPS keymap bit too
00:47:12AlexPIt has two functions on one action
00:47:34AlexPMom: Short and long click should do two different things
00:47:35Momif it is wrong, what would be the correct behavior?
00:47:49AlexPI don't know without looking at the source
00:47:51AlexPOne mo
00:48:08MomAlexP: ok in that case, "short" click doesn't do anything, but long click goes to the song list
00:48:24Momi'll wait :)
00:48:26 Quit liar (Ping timeout: 240 seconds)
00:49:04 Join liar [0] (
00:50:06AlexPhmmm, short down isn't defined for the wps in particular
00:50:29AlexPand repeat down is defined differently epending on whether the hotkey function is defined
00:50:39MomIt seems like the long click behavior should be mapped to short click instead
00:50:47 Quit bluebrother (Disconnected by services)
00:50:50 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
00:50:56 Quit evilnick (Quit: Page closed)
00:50:59AlexPquickscreen is on long home apparently
00:51:08AlexPMom: No, short should go to menu I think
00:51:20Momthe main menu?
00:51:29Momthere is already a menu button on this player
00:52:11AlexPah yes, home goes to it
00:52:20AlexPthat isn't in the manual either
00:52:21***Saving seen data "./dancer.seen"
00:52:28AlexPor rather it is, but is wrong
00:52:30Momso when you press the submenu, ideally it should give the list of songs
00:52:42AlexPOn what others do
00:52:57Momother buttons?
00:53:00 Quit Strife89 (Read error: Operation timed out)
00:53:25AlexPNo, other targets
00:53:39AlexPIf you hold down home you should get the quickscreen yes?
00:53:48AlexPaccording to the source
00:54:02MomI'm not sure what the quickscreen is, could you describe it? or i can look it up in the manual
00:54:38AlexPIt gives you four options
00:54:45AlexPfor left right up and down
00:54:52AlexPe.g. repeat mode is left IIRC
00:54:52Momno I don't see that
00:54:56MomI see a list of songs
00:55:04AlexPPressing long home?
00:55:12Momand pressing short home doesn't do anything
00:55:13AlexPsame as long down?
00:55:19Momoh sorry
00:55:33Momlong home is the quickscreen
00:56:17Momand short home takes you to the default starting point that you can set in settings
00:56:28AlexPyeah, the main menu
00:57:02MomI have to go now :( −− do you think this down key is a problem needing to be fixed?
00:57:02AlexPI can quickly fix the manual, but I should check what short down is meant to do before I fix it :)
00:57:08Momok :)
00:57:30AlexPI'll have a fiddle in the morning of nobody does first, it is late here :)
00:57:44AlexPThanks for reporting
00:57:48MomAlright :)
00:58:01MomI will come back in a day or so
00:58:07Momand ask what short down should do :)
00:58:10AlexPhehe :)
00:58:29Momthank you for your help :)
00:58:34Mombye bye
00:58:35AlexPno worries :)
00:58:42 Quit Mom (Quit: Leaving)
00:59:08 Join JdGordon [0] (~jonno@
00:59:12 Quit JdGordon (Changing host)
00:59:12 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
01:07:22 Join Strife89 [0] (
01:07:43 Quit Strife89 (Read error: Connection reset by peer)
01:07:44 Quit bmbl (Quit: Bye!)
01:08:02 Join Strife89 [0] (
01:10:54 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
01:13:07 Join Rob2223 [0] (
01:16:53 Quit Rob2222 (Ping timeout: 260 seconds)
01:18:30 Quit Brodel ()
01:30:12 Quit Rob2223 (Read error: Connection reset by peer)
01:33:07 Part toffe82
01:35:11 Quit GeekShad0w (Read error: Connection reset by peer)
01:42:25 Join Dreamxtreme_ [0] (~Dreamxtre@
01:42:25 Join Dreamxtreme [0] (~Dreamxtre@
01:42:36 Quit Dreamxtreme_ (Read error: Connection reset by peer)
01:58:53 Join JdGordon| [0] (~jonno@
01:58:53 Quit JdGordon| (Changing host)
01:58:53 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
02:05:12 Quit Judas_PhD (Quit: This is a quitting message)
02:10:57 Quit milz (Ping timeout: 255 seconds)
02:45:43 Join milz [0] (
02:49:02 Quit TheSphinX^ (Quit: XChat)
02:50:13 Join Xerion_ [0] (~xerion@
02:52:23***Saving seen data "./dancer.seen"
02:52:30 Quit Xerion (Ping timeout: 265 seconds)
02:52:30 Nick Xerion_ is now known as Xerion (~xerion@
02:56:23 Join antil33t [0] (
02:58:30 Join gourdcaptain [0] (
02:59:32gourdcaptainI'm having a problem with my Sanza Clip+ and USB access.
03:00:18gourdcaptainWhen I connect it with the original firmware under linux, it throws kernel errors about not being available after error recovery and doesn'
03:00:30gourdcaptaint mount. It does under Windows XP.
03:02:01JdGordon|we cant help if it is the OF's problem
03:03:09gourdcaptainIt did mount under Linux until it crashed partway through me loading music onto it.
03:11:08 Quit komputes (Remote host closed the connection)
03:15:21 Quit Xerion (Read error: Connection reset by peer)
03:17:01 Quit gourdcaptain (Quit: CGI:IRC (EOF))
03:24:10 Quit Strife89 (Quit: Reboot to install Ubuntu.)
03:32:46 Join Strife89DS [0] (
03:37:02 Quit Strife89DS (Read error: Connection reset by peer)
03:40:41 Join Strife89DS [0] (
03:43:41 Join bunnyboi [0] (
03:44:57 Quit Strife89DS (Read error: Connection reset by peer)
03:55:31 Join Strife89DS [0] (
03:57:33 Join Judas_PhD [0] (
04:04:58 Quit S_a_i_n_t (Ping timeout: 276 seconds)
04:10:55 Join steve|m1 [0] (
04:10:56 Quit steve|m (Disconnected by services)
04:11:11 Nick steve|m1 is now known as steve|m (
04:14:37 Quit MethoS- (Remote host closed the connection)
04:28:45 Quit amiconn (Disconnected by services)
04:28:47 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:28:47 Quit pixelma (Disconnected by services)
04:28:49 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:28:51 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:29:06 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:30:32 Quit TheSeven (Ping timeout: 245 seconds)
04:35:08 Join Strife89 [0] (
04:35:27 Quit Strife89DS (Quit: Shutting off DS.)
04:35:37 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:44:20 Quit madalu (Read error: Operation timed out)
04:49:04 Quit Hillshum (Quit: Page closed)
04:52:24***Saving seen data "./dancer.seen"
05:41:53 Quit panni_ (Read error: Connection reset by peer)
05:45:36 Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier.
06:00:11 Quit Horscht (Quit: Verlassend)
06:01:02 Quit mc2739 (Ping timeout: 240 seconds)
06:33:51 Join Jennifur [0] (
06:34:18 Join lestatar_ [0] (
06:34:49 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
06:37:35 Quit bunnyboi (Ping timeout: 276 seconds)
06:38:14 Quit lestatar (Ping timeout: 276 seconds)
06:38:15 Nick lestatar_ is now known as lestatar (
06:39:28 Quit Strife89 (Quit: Rebooting to finish off an update.)
06:44:14 Quit kadoban (Ping timeout: 276 seconds)
06:44:26 Join Strife89 [0] (
06:45:05 Quit _s1gma (Quit: irl ragequit)
06:48:54 Quit froggyman (Ping timeout: 265 seconds)
06:52:27***Saving seen data "./dancer.seen"
07:01:57 Quit Aslai (Ping timeout: 265 seconds)
07:09:56 Join Misanthropos_ [0] (
07:10:15 Join stoffel [0] (
07:16:03Misanthropos_heya. can somebody give me a link or some input on the usb-part of the clip+ (e.g. why is it working on the clip but not on the clip+?)
07:19:02Misanthropos_ah.. damn ... i knew i read something about it and just found it again
07:28:02 Join hebz0rl [0] (
07:28:44 Quit user890104 ()
07:38:38 Quit Misanthropos_ (Quit: Ex-Chat)
07:47:01 Quit hebz0rl (Quit: Ex-Chat)
07:52:12 Quit t0rc (Remote host closed the connection)
08:10:20 Join froggyman [0] (~me@unaffiliated/froggyman)
08:18:19 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
08:21:34 Quit anewuser (Quit: FAMICOMPO 7 - August 15 Through August 31 JST -
08:21:54 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
08:43:58 Quit AndyI (Ping timeout: 272 seconds)
08:49:51 Join einhirn [0] (
08:52:31***Saving seen data "./dancer.seen"
08:57:22 Join ender` [0] (
08:59:49 Join S_a_i_n_t [0] (S_a_i_n_t@
09:00:14 Quit bertrik (Quit: :tiuQ)
09:07:54 Quit milz (Ping timeout: 240 seconds)
09:10:36 Join milz [0] (
09:22:54 Join AndyI [0] (~pasha_int@
09:25:45 Quit antil33t (Read error: Connection reset by peer)
09:25:53 Join antil33t [0] (
09:29:38 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
09:29:54 Join LinusN [0] (~linus@rockbox/developer/LinusN)
09:31:38 Quit pamaury (Read error: No route to host)
09:31:39 Join pamaury_ [0] (
09:32:06 Quit AndyI (Ping timeout: 272 seconds)
09:32:26 Nick pamaury_ is now known as pamaury (
09:32:40 Quit pamaury (Changing host)
09:32:40 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
09:34:40 Join AndyI [0] (~pasha_int@
09:39:17 Quit JdGordon| (Quit: leaving)
09:39:43 Quit AndyI (Ping timeout: 272 seconds)
09:40:30 Join petur [0] (d408b802@rockbox/developer/petur)
09:46:19 Join AndyI [0] (~pasha_int@
09:50:22 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
09:51:07 Quit AndyI (Ping timeout: 272 seconds)
10:14:29 Join n1s [0] (~n1s@rockbox/developer/n1s)
10:25:45 Part LinusN
10:25:45 Join mischasworld [0] (
10:35:13 Join DerPapst [0] (~Alexander@
10:40:47 Join LinusN [0] (~linus@rockbox/developer/LinusN)
10:52:27*JdGordon managed to get rockbox into a wonderful state where it is impossible to restart playback without digging through menus
10:52:34***Saving seen data "./dancer.seen"
10:52:39JdGordoncrappy yes/no screen with no touch support
11:02:06 Join AndyI [0] (~pasha_int@
11:02:53 Join Rob2222 [0] (
11:03:21 Quit Rob2222 (Client Quit)
11:04:00 Join Rob2222 [0] (
11:06:32 Quit n1s (Quit: Lämnar)
11:07:07 Quit AndyI (Ping timeout: 272 seconds)
11:07:49 Quit jfc (Read error: No route to host)
11:08:16 Join jfc [0] (
11:09:16 Quit Judas_PhD (Quit: This is a quitting message)
11:09:19 Quit jfc (Read error: Connection reset by peer)
11:10:16 Join jfc [0] (
11:25:50 Join [Saint] [0] (S_a_i_n_t@
11:26:43 Quit S_a_i_n_t (Ping timeout: 265 seconds)
11:29:01 Join pyro_maniac_ [0] (
11:32:18 Quit pyro_maniac (Ping timeout: 272 seconds)
11:42:19 Quit [Saint] (Ping timeout: 240 seconds)
11:42:27 Join S_a_i_n_t [0] (S_a_i_n_t@
11:53:04 Quit stoffel (Ping timeout: 240 seconds)
11:54:08 Nick pyro_maniac_ is now known as pyro_maniac (
12:09:04JdGordonBagder: ping?
12:10:45 Quit sasquatch (Read error: Operation timed out)
12:14:43 Quit l403 (Ping timeout: 240 seconds)
12:15:07 Join l403 [0] (
12:20:19CIA-81New commit by jdgordon (r27890): Make the yes/no screen useable in the stylus touchscreen mode, add 2 soft-buttons. Also make bottom left yes in grid mode to line up with the new ...
12:21:58CIA-81r27890 build result: All green
12:27:02 Join sasquatch [0] (
12:27:07 Quit JdGordon (Ping timeout: 240 seconds)
12:31:23 Join stoffel [0] (
12:35:11 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
12:36:29 Join JdGordon1 [0] (~jonno@
12:37:18JdGordon1does anyone know why kugel chose the D2 pad for the sdl app instead of making a new keymap?
12:37:19JdGordon1its pretty close but still not ideal :(
12:38:04 Join AsusFreak [0] (
12:39:45 Join dfkt [0] (dfkt@unaffiliated/dfkt)
12:39:48 Quit JdGordon (Ping timeout: 265 seconds)
12:41:49 Join Xerion [0] (~xerion@
12:44:32 Quit AsusFreak (Quit: CGI:IRC)
12:52:38***Saving seen data "./dancer.seen"
13:09:25 Join kadoban [0] (
13:12:17 Join n1s [0] (~n1s@rockbox/developer/n1s)
13:18:47n1summ, did we decide on releasing?
13:20:24 Quit mischasworld (Ping timeout: 265 seconds)
13:43:54CIA-81New commit by jdgordon (r27891): Move to a proper sdl key config instead of using the d2 pad. make the mouse wheel work, middle click is "select" and right click is "back"
13:46:06CIA-81r27891 build result: All green
13:47:49 Nick JdGordon1 is now known as JdGordon (~jonno@
13:48:01 Quit JdGordon (Changing host)
13:48:01 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
13:52:47 Quit l403 (Ping timeout: 265 seconds)
13:56:56 Join maffe [0] (
13:58:34 Join Rob2223 [0] (
14:02:27 Quit Rob2222 (Ping timeout: 264 seconds)
14:07:22 Quit maffe (Quit: Miranda IM! Smaller, Faster, Easier.
14:08:45 Join maffe [0] (
14:09:37 Quit maffe (Client Quit)
14:14:31JdGordonanyone know where the fonts are supposed to live in the sdl app build?
14:23:53JdGordonn1s: nope. someone needs to take charge if we want to do a release
14:29:39 Join funman [0] (~fun@rockbox/developer/funman)
14:30:35 Quit n1s (Quit: Lämnar)
14:40:59 Join panni_ [0] (
14:43:54JdGordonkugel: incase I forget, should rockbx on android be using 16% cpu when in the filebrowser (and minimised) while playing music?
14:45:20funmanJdGordon: did you try to define CPU_ARM ?
14:45:50JdGordonno, is that not being done yet?
14:46:47funmandunno, iirc kugel wanted to postpone this in case there were problems but a very quick test worked fine for me
14:46:56funman(i didn't do performance test though)
14:47:36JdGordonok, ill give it a go..
14:48:25JdGordonwe shouldnt assume android is arm though... iirc it builds (in theory) for x86 and mips also, no?
14:49:45JdGordonjust #define CPU_ARM ?
14:49:51JdGordondoes it need the ARM_ARCH also?
14:50:42funmanJdGordon: C code is built for ARM though
14:51:12funmani had added the ANDROID case to CPU_ARM + ARM_ARCH 5, but ARCH 6 might work depending on your phone
14:52:42***Saving seen data "./dancer.seen"
14:55:10JdGordon/home/jonno/rockbox-git/apps/codecs/codec_crt0.c:60: error: 'struct codec_api' has no member named '__div0' :(
14:56:30funmanuse the same condition than in codecs.h
14:57:53 Join NTKM [0] (
14:58:00JdGordonCONFIG_PLATFORM & PLATFORM_NATIVE is wrong though?
14:58:35funmani suppose division by 0 exception is handled by the native libc
14:59:16JdGordonoh right... im half asleep
15:00:52 Quit antil33t (Read error: Connection reset by peer)
15:00:59 Join antil33t [0] (
15:02:06 Join robin0800 [0] (
15:03:08 Join robin0800_ [0] (
15:04:08 Quit robin0800 (Remote host closed the connection)
15:07:52 Quit dionoea (Ping timeout: 258 seconds)
15:08:27 Quit Utchybann (Ping timeout: 260 seconds)
15:08:28*JdGordon fails with
15:12:08 Join Utchybann [0] (
15:14:33JdGordoneclipse cant open the project!
15:14:41 Join evilnick [0] (0c140464@rockbox/staff/evilnick)
15:14:47 Join _s1gma [0] (~d.d.derp@ has lots of errors
15:15:09 Quit _s1gma (Max SendQ exceeded)
15:15:47JdGordonoh ruddy hell... it needs v8 api?
15:15:56 Join dionoea [0] (
15:16:46JdGordon[2010-08-26 23:13:23 - Rockbox] Unable to resolve target 'android-8'
15:16:55 Quit krazykit (Quit: bbl)
15:18:41 Quit NTKM (Read error: Connection reset by peer)
15:22:56JdGordonnuts, codec failure
15:29:42 Join kugel [0] (~kugel@rockbox/developer/kugel)
15:36:49 Join l403 [0] (
15:42:08 Quit JdGordon (Quit: I dont knwo why the fuck i bother)
15:45:16 Join Aslai [0] (
15:49:08 Quit antil33t (Read error: Connection reset by peer)
15:49:15 Join antil33t [0] (
15:55:31 Join jgarvey [0] (
16:05:09 Join Jaykay [0] (
16:05:56 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
16:07:36 Join krazykit [0] (~kkit@
16:14:59funmankugel: ping, did you see ?
16:15:52 Quit l403 (Ping timeout: 265 seconds)
16:16:53kugelfunman: I saw the first one yesterday while trying a different change with mingw32
16:17:10kugelmaking it _atexit and do #define atexit _atexit fixes it
16:17:59funmancan't we use the OS atexit() when available ? IIUC it should work when unloading .dll/.so
16:19:12 Quit parafin (Remote host closed the connection)
16:19:13kugelI don't know if that works
16:19:19 Join parafin [0] (
16:19:49kugelbut we only support 1 exit handler (for now) so they behave slightly different
16:20:22funmanit's close enough to the real function
16:21:29funmanit returns -1 if you try it a second time so it's perfectly valid
16:22:08kugelbut the host's atexit won't do that
16:22:26funmanif we only call atexit() one time it's ok?
16:22:36kugela plugin developed in the sim could then not work on target
16:23:22funmana plugin which calls atexit() more than 1 time could cause our wrath on its developer though :)
16:23:32 Join Misanthropos_ [0] (
16:24:55kugelI don't feel strong about it
16:25:24funmananyway it seems to work fine on linux: the rockbox atexit() is used and not the libc one
16:25:42kugelthat was the plan yes
16:26:10kugelit maybe has to do with some preprocessor magic, note that it complains about _atexit which we don't have
16:26:31funmandunno what's special for mingw
16:26:44kugeland maybe linking order
16:27:22kugelI can imagine it complains if it links "libc.dll" before plugin_crt0.o
16:27:33funmanatexit is in the crt.o/.a on debian's mingw
16:29:39 Quit t0rc (Ping timeout: 252 seconds)
16:35:04 Join l403 [0] (
16:38:07kugelfunman: atexit in .so's work with glibc, I don't know if it works on windows as well
16:38:18 Join freonx [0] (~freonx@
16:38:39funmani can try in wine
16:39:54 Quit freonx (Client Quit)
16:40:08 Join toffe82 [0] (~chatzilla@
16:40:26 Join Judas_PhD [0] (
16:42:08kugeldoesn't work for me
16:42:15kugeloh wait it does
16:42:40funmani can't use the '*' button for HOME on fuze sim
16:43:17funmanyes it works for me too
16:43:42kugelstill, I think it's a feature of the platforms C library and not something the C standard promises, right? if so I'd be recultant to use the host's atexit/exit
16:44:34funmanThe application should call sysconf() to obtain the value of {ATEXIT_MAX}, the number of functions that can be registered
16:46:19funmani dont' know if it's documented but according to google it's widely implemented
16:47:48funmanposix atexit/dlclose do not talk about each other function
16:47:56 Quit l403 (Ping timeout: 240 seconds)
16:49:07kugel"For GLibc, "atexit" is not in at all, but is instead part of a static library that is added to the link command of each generated shared library." (from, exactly what we do
16:50:00kugelbut it seems not implemented/broken on android (even though there's no plugins on android yet) so I think we should keep our atexit
16:50:01 Part LinusN
16:51:15 Join n1s [0] (~n1s@rockbox/developer/n1s)
16:52:45***Saving seen data "./dancer.seen"
16:53:12funmanmaybe rename to at_pluginexit ?
16:54:12kugelbut why?
16:56:31funmanbecause it doesn't conform to all existing atexit() implementations
16:56:44funmanalthough on the android bug they suggest to do like glibc
16:57:10kugelit doesn't?
16:57:33kugelit conforms to the standard which should be enough, doesn't it?
16:57:57kugel(we can bump the 1-exit-handler limitation if that's the problem)
16:58:01funmanyes but nothing in the standards says about dlopen/dlclose and IMO it's important
16:58:36kugelwhat has dlopen/dlclose to do with it?
16:59:14funmanif you register an exit handler in a shared library, and don't know if it will run at program exit or at said shared library dlclose()
16:59:37funmanjust like in the bug you linked
16:59:43TorneThe minimum acceptable value of ATEXIT_MAX is 32, so you can't claim ours is it. :)
16:59:51 Quit Jaykay (Remote host closed the connection)
16:59:56Tornerenaming it seems like a good idea
17:00:06Torneplugins that expect to use atexit() because they are ported from something else can just #define it
17:00:11kugelthe exit/atexit is in the plugin, shared library or not
17:00:15Tornelike I do for loads of stuff in frotz
17:00:18funmanVLC had a special member for its plugins named "linked_with_a_crap_library_which_uses_atexit", but now it's "cannot_unload_broken_library"
17:00:39funmansaid libraries think atexit handler will not run on dlclose(), or never thought about dlclose()
17:00:51kugelit's not like rockbox core provides atexit, it's linked into the plugin
17:01:44 Quit pyro_maniac (Remote host closed the connection)
17:02:28kugelTorne: so we can have 32 atexit handlers without problem, the array won't make plugins significantly larger
17:02:35Tornekugel: it's still not right
17:02:42 Join l403 [0] (
17:02:46 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
17:02:57Tornefunman is right: there's no reason why anyone should assume atexit() takes effect at dlclose() time
17:03:10funmanand there's no reason why anyone should assume it doesn't either?
17:03:11Torneit is specifically at process exit; plugins are not processes and don't act like them
17:03:16Tornefunman: yes they should
17:03:18Torneit says process exit
17:03:24kugelwe don't call it at dlclose?!
17:03:35Tornecalling it at dlclose() is *wrong*
17:03:37funmanNormal termination occurs either by a call to exit() or a return from main()
17:03:44kugeland we don't do it
17:04:01Tornefunman: right, but our plugins are not freestanding processes in the first place
17:04:07Torneso much of SUS can't be applied to them :)
17:04:12kugelwe call it exactly after the return from main() (plugin_start) or during exit()
17:04:33Tornekugel: But that's not main()
17:04:39funmanit could be renamed
17:04:45funmanbut we don't care
17:04:47Torneit doesn't matter what it's *called*
17:05:02*kugel fails to see that it doesn't conform
17:05:20Tornekugel: because we don't have processes? :)
17:05:25Torneso exit() is not conformant either
17:05:25 Quit l403 (Read error: Operation timed out)
17:05:25funmanit conforms to paper standards, but not to other broken implementations :)
17:05:32Tornefunman: no it doesn't
17:05:39Torneour exit() etc are all equally fake and nonconformant
17:05:52funmanthey are very good imitations though
17:05:55TorneNo, they're not
17:06:04Tornethey're minimally passable for the specific things that are ported using them :)
17:06:11Tornee.g. filehandles are not closed
17:06:14kugelwhere does it talk about processes? it's about the end of a *program*?
17:06:24funmanTorne: i'm pretty sure they are?
17:06:29Tornefunman: no they aren't
17:06:39Tornefrank looked at doing it, and it's too hard
17:06:42funmananyway what should we do for win32 builds?
17:06:48Tornewe should rename atexit to something else
17:07:10Torneand not care whether it's standards compliant or not because it isn't
17:07:14kugelit's not hard, the file handles opened by plugins are tracked and could be closed (we instead throw a panic now)
17:07:47Torneif plugins want to invoke atexit() then they can do so via a #define, but it's rare, no?
17:07:56CIA-81New commit by funman (r27892): remove executable bit on plugin_crt0.c
17:08:22Torne(frotz #defines all kinds of horrible hacks to pretend to be ANSI.. like fread() and friends ) :)
17:08:59Tornemy point is just that since it's neither actually a standard implementation, nor particularly important whether it's actually called atexit or not, renaming it is the easiest way to solve the build problem
17:09:19Torneif you wer ebeing picky about standards you should rename exit() as well :)
17:09:20kugelI'm still not convinced about the former
17:09:20funmanmakes sense
17:09:20Tornebut hey
17:09:29kugelbut renaming is easy enough
17:09:41Tornekugel: single pages of SUS are not meaningful on their own
17:09:49Torneeverything is interdependant
17:09:54funmanjust explicit that it behaves almost like atexit
17:09:55CIA-81r27892 build result: All green
17:09:57Tornesingle unix spec.
17:10:04Tornecurrent successor to posix
17:10:15 Join l403 [0] (
17:10:23 Join Jaykay [0] (
17:10:25Tornesame argument for ANSI C or whatever spec you want to look at, matters not
17:11:00Tornethese terms all have defined meanings and dynamically loading plugins and running them in an existing thread is not in any way equivalent to a process/program/whatever
17:11:08kugelwhat was the reason our atexit() doesn't conform to ansi c?
17:11:15Tornebecause a plugin isn't a program
17:11:20kugelit sure is
17:11:22TorneNo it isn't
17:12:42kugelonly because it doesn't have main()?
17:13:05Torneno, because it doesn't follow the lifecycle described in ansi c for a program in a host environment
17:13:18Tornebecause our "host environment" directly shares all its state with the plugin
17:13:21Tornebecause they are a single program
17:13:38Torneany implementation of exit() that doesn't exit rockbox entirely is not right :)
17:14:19Torneyou can't define the terms in the description of a function any way you like to say it conforms, you have to be using the same definitions of everyhting as the entire rest of the spec
17:14:43Torneit's handy that it behaves in a conceptually similar way, that's the point o fhaving it, to reduce porting effort
17:15:00Tornebut little tricks to reduce porting effort do not make an ANSI C environment :)
17:16:15kugelI can think of another reason, atexit isn't declared in stdlib.h but pluginlib_exit.h :)
17:16:22TorneThat too :)
17:16:42Tornebut srsly. doesn't matter. easy fix is to rename it. there's no actual downside to doing so, so yeah
17:16:53Torneat_plugin_exit or whatever
17:30:37 Quit robin0800_ (Quit: Leaving)
17:31:16 Quit funman (Quit: free(random());)
17:34:28 Join [sko] [0] (~sko]
17:42:51pixelmaS_a_i_n_t: I now go on with tracing the weird freeze on target with my WPS and it becomes even weirder. First I made a test.wps with only the one offending line and it still hangs, I then replaced the unusual tags with more common ones like %it etc. which I know work - no difference. Now I made the nested conditional one level less nested and now playback starts and it goes to the WPS (shows as expected) but then hangs when it finishes buffering...
17:44:35CIA-81New commit by learman (r27893): Fix building simulator plugins on Cygwin. A DLL stub containing atexit is linked to the plugins, so give our version a prefix.
17:44:41 Join Lear [0] (chatzilla@rockbox/developer/lear)
17:45:26kugelhm, Lear was faster than me
17:46:23CIA-81r27893 build result: 2627 errors, 0 warnings (learman committed)
17:46:52LearOops, forgot a file. :)
17:48:14CIA-81New commit by learman (r27894): Oops, forgot this file in r27893.
17:48:16kugel#include "lib/pluginlib_exit.h" in plugin_crt0.c might be a good idea
17:49:50CIA-81r27894 build result: All green
17:55:59pixelmaS_a_i_n_t: one more level less and it doesn't hang anymore
17:57:22*kugel finally understood why we're doing the temp_codec.dll thing
18:00:26 Join Kitr88 [0] (
18:03:23pixelmaand there is a difference between which "path" it has to take. As I said, I replaced one of the first with %?it - the freeze on finishing to buffer occurs on a file without title tag but not when the file has the title tag ?! :\
18:04:21 Quit Kitar|st (Ping timeout: 272 seconds)
18:04:35 Quit petur (Quit: Page closed)
18:04:50 Quit Kitr88 (Ping timeout: 240 seconds)
18:05:19pixelmaok, guess I need to find the commit that broke it
18:09:41 Join Kitar|st [0] (
18:18:07 Join Horscht [0] (~Horscht@xbmc/user/horscht)
18:21:34 Nick YPSY is now known as Ypsy (
18:26:31 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
18:32:05 Join Jerom [0] (~jerome@
18:35:43 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
18:38:30*kugel has a small code loading abstraction layer
18:44:24 Quit swilde (Quit: ERC Version 5.3 (IRC client for Emacs))
18:52:30 Join madalu [0] (~matt@unaffiliated/madalu)
18:52:47***Saving seen data "./dancer.seen"
19:03:52 Quit saratoga (Quit: Page closed)
19:04:35 Quit bmbl (Read error: Connection reset by peer)
19:06:51 Nick Ypsy is now known as YPSY (
19:20:29*kugel can hopefully get rid of io.c (and libuisimulator.a) entirely for the android port
19:26:09 Quit madalu (Remote host closed the connection)
19:27:51 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
19:29:18 Quit DerPapst (Quit: Leaving.)
19:31:40 Quit Lear (Quit: ChatZilla 0.9.86 [Firefox 4.0b5pre/20100824030940])
19:34:09 Join jey2 [0] (jey2@gateway/shell/
19:34:22jey2how easy is it to switch away from rockbox ?
19:34:27jey2tempeted to try it
19:34:43fiveofohHolding down the left-arrow key?
19:34:54fiveofohOr are you referring to something else?
19:34:59jey2(I meant switch back to ipod firmware)
19:35:10fiveofohHold down the left-arrow key while you're turning it on
19:35:20jey2oh, so it's dual boot?
19:35:28jey2but what if I want to actually remove it?
19:35:34jey2so I don't have to do that everytime
19:35:36pixelmafiveofoh: no, that's a different button sequence on the Ipods
19:35:59fiveofohWell hold down some device-specific keys while booting it then :P
19:36:21fiveofoh(my Sansa is showing ;)
19:36:30pixelmajey2: but yes, you can dual boot and in the case of the Ipods it will also stay in the original firmware until you reset it. You can also uninstall easily
19:37:21jey2okay, as long as I can uninstall easily :P
19:37:23fiveofohjey2: see
19:37:27jey2but first I need to get music on my ipod
19:37:42pixelmawhat kind of Ipod is it exactly?
19:38:05jey2i think it was one of the ones they called ipod classic
19:38:33jey2wait, is that covered under rockbox?
19:39:04pixelmaif it really is an Ipod *Classic* then Rockbox isn't an option anyway
19:39:15 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
19:39:54jey2well that's an easy option for me
19:40:00jey2aka, no rockbox
19:40:04 Join Lear [0] (chatzilla@rockbox/developer/lear)
19:40:10jey2by the way, does drspoon ever come on here?]
19:40:12pixelmalook at that page and tell us which model it is
19:41:46 Quit kugel (Remote host closed the connection)
19:46:10pixelmathen sorry, you're out of luck. I just asked again because sometimes people call the older Ipods "classic", it's a weird naming Apple got there
19:46:23jey2it's "funny"
19:46:31jey2my ipod says it has basically no free space on it
19:46:38jey2but sstill says "no music" :(
19:48:53AlexPpixelma: For the clip th eimage showing button names has a couple of the names different to what they are in the text - if I edit the svg, will the png and pdf files also be recreated, or do I need to generate them too?
19:49:23pixelmayou need to regenerate them too
19:49:36AlexPAlternative is that I change the names used in the text
19:49:42AlexPI don't know whhich is prefered
19:49:57AlexPThe image has e.g. "Down" and the text "Submenu"
19:50:06pixelmadepends on content I guess
19:50:35AlexPI don't actually care, but they should be the same :)
19:51:00pixelmahow do other Sansa manuals call the button? (Not that I'd be surprised if there aren't differences too)
19:51:08pixelmaerr... are
19:51:21AlexPdunno, one mo
19:52:05gevaertsThe text always calls it submenu I think
19:52:14AlexPclip+ is inconsistent too
19:52:21AlexPe200 uses e.g. submenu for both
19:52:28AlexPSO I'll change the image I think
19:52:41gevaertsFuze says Down in the image
19:52:57AlexPSo clip, clip+ and fuze need the images correcting?
19:53:18jey2thaqnks for the help guys
19:53:22 Part jey2
19:53:45*pixelma has the slight suspicion that the weird on the Ondio with that one line in the WPS got introduced between r27773 and r27860 and couldn't be tracked down further then :\\
19:53:55pixelmaweird crash even
19:54:16AlexPHow do I edit text in inkscape?
19:55:13 Join DerPapst [0] (
19:56:04fiveofohHmm, it may be vectorized
19:56:12fiveofohIn which case you'd just have to re-create it
19:56:49pixelmaAlexP: be careful that you are on the right layer and that it isn't grouped
19:57:10AlexPfiveofoh: I've got the svg, I just don't know what to do to it :)
19:57:16pixelmaI don't think that the text is vectorised in our SVGs
19:57:31AlexPI have no idea :)
19:57:55gevaertsWhich one are you looking at?
19:58:03fiveofohWell if it's not, I'm pretty sure you grab the text tool and click on the text
19:59:05AlexPgevaerts: sansaclip-front
19:59:06fiveofohOh it's probably in a couple groups
19:59:12AlexPoh ffs
19:59:17AlexPI'm an idiot
19:59:18fiveofohI just kept double-clicking on it until it came up with a text editing box
19:59:25AlexPI accidently opened the png :)
19:59:27 Quit evilnick (Quit: Page closed)
19:59:46fiveofohYeah, that would do it
20:02:01 Join Topy [0] (
20:04:17 Quit Strife89 (Quit: Packing up stuff.)
20:04:32AlexPpixelma: Are there any particular settings to generate the png and pdf?
20:05:56 Quit T44 (Ping timeout: 265 seconds)
20:06:06pixelma ;) ... although I'm not sure if there is a simpler way for the PDF - at least that is the safest method
20:17:49 Part fiveofoh
20:19:55 Join Topy44 [0] (
20:21:21CIA-81New commit by alex (r27895): Manual: Make the clip WPS buttons match what actually happens. Short submenu is not defined, is this intentional?
20:21:41CIA-81New commit by alex (r27896): Manual: Change the labels on the clip, clip+ and fuze images so that they actually match the text.
20:22:51 Quit Topy (Ping timeout: 265 seconds)
20:22:53AlexPpixelma: Do those look OK?
20:23:50CIA-81r27895 build result: All green
20:25:50CIA-81r27896 build result: All green
20:36:55 Part froggyman
20:36:55 Join [Saint] [0] (S_a_i_n_t@
20:38:45 Quit S_a_i_n_t (Ping timeout: 240 seconds)
20:44:46 Quit milz (Ping timeout: 245 seconds)
20:48:09pixelmaLear, amiconn: can you imagine something about r27773 that makes loading of music in my WPS on the Ondio break on a bit more complicated tag line? The reason I ask is because the WPS works fine in the sim etc. and I see weird behaviour while loading/buffering which I don't think are explainable with parser bugs alone.
20:48:09pixelmaWith the complicated, nested conditional line, the WPS loads but as soon as I want to start playback it hangs at the "loading" splash. If I take one level of nestedness I get normal playback or a crash right after the first buffering depending on which true/false path it has to go. Simplified one more level and all is well...
20:49:21 Join petur [0] (~petur@rockbox/developer/petur)
20:51:20 Nick [Saint] is now known as S_a_i_n_t (S_a_i_n_t@
20:51:27 Quit Topy44 (Quit: Leaving)
20:52:50***Saving seen data "./dancer.seen"
20:57:37 Join Topy44 [0] (
20:58:33 Quit t0rc (Quit: Leaving)
21:02:34 Quit stoffel (Remote host closed the connection)
21:04:21 Quit Jerom (Read error: Connection reset by peer)
21:04:34 Join Jerom [0] (~heidi@
21:05:10 Join Buschel [0] (
21:05:36BuschelLear: you there?
21:09:20LearBuschel: Yes.
21:09:21BuschelLear: regarding the aac issue with the sansa fuze v2 -> i can hear strange sounds in the sim as well. but i can only heat it on the sansa fuze v2 sim, and not on the ipod video sim.
21:09:40LearBuschel: Now, that's weird. :)
21:09:57BuschelLear: seems like somehow there are different build outputs (also the binsize of aac.codec is different)
21:11:27 Join m0ar [0] (
21:12:02m0arHow is the development with the clip+ going? Any fixes for the +1,5% playback speed issue in sight? :)
21:12:38bertrikm0ar, yes that was fixed in the past week or so
21:13:57m0arNice, then I've got that version already, hah.
21:14:23bertrikhaha, so you didn't even notice?
21:14:52m0arbertrik: Reformatted my amazingly corrupt clip+ last night, and updated to the newest RB-release in the same go :)
21:14:53BuschelLear: forget it... tested the wrong output...
21:15:27m0arWhat's left to fix before the clip+ is deemed to be stable?
21:15:38BuschelLear: at least I can reproduce it now :o)
21:15:53Torneusb that doesn't lock the player up for a week, presumably
21:16:02LearBuschel: So the ipod video sim has the same problem?
21:16:22Buschelyep, just didn't update the simdisk path...
21:16:31pixelmaTorne: I thought that affected only some FuzeV2s?
21:16:43m0arTorne: What? I've had no lockups when I've accidently USB'd the RB firmware?
21:16:50m0arTorne: It just keeps on rebooting
21:16:55Torneisn't it all the as3525v2 devices?
21:17:39m0arWhat's the problem there by the way, why is it so hard to fix? I'm not a developer myself, but I'd guess that the same kind of debugging has been done on most RB units?
21:17:42 Join funman [0] (~fun@rockbox/developer/funman)
21:17:48Learpixelma: Have you checked the "View OS stacks" in the debug menu, when using the working wps? The "main" stack is the most interesting.
21:18:26 Quit balintx (Remote host closed the connection)
21:18:32pixelmahmm, not yet. Will do
21:18:55Tornem0ar: we have no idea what the problem is
21:19:00Tornewhich is the usual reason for things being hard to fix
21:19:15m0arTorne: Ah, point taken
21:19:21 Join balintx [0] (
21:19:47m0arJust get it clear; when the clip+ is deemed stable it jumps on the full speed train, eg getting the new versions?
21:19:57Torneer, what?
21:20:04Torneno, when it's deemed stable it's deemed stable
21:20:08Tornenothing actually changes
21:20:09funmanall as3525v2, but now it's worked around
21:20:16m0arAh, well okay
21:20:25funman-> no reboot on usb plug, you need to reboot manually
21:20:31Tornewe already build every build for clip+, clipv2, etc
21:20:52m0arfunman: my clip+ reboot-loops when I plug it in without pressing left :)
21:21:06m0arfunman: boots RB, fails, reboots RB, fails, and, so, on
21:21:31funmaninstall a current build
21:21:31pixelmaLear: main is at 62% (same if I load my WPS - as I said I can load it but have trouble when starting playback)
21:22:14m0arTorne: I was in here a couple of days ago and someone told me that the clip+ isn't getting all the updates because it's not stable or something like that, this isn't true then?
21:22:35bertrikI've never had any problem with corruption because of USB plugin on my clip+ and I did plug in/out carelessly quite often
21:22:37Torneall players are built from exactly the same code
21:22:39Learpixelma: So no stack overflow then. Just eliminating one possibilities...
21:22:46LearWhat does the failing line look like?
21:22:52m0arTorne: Ah, thanks for clearing that out
21:23:04Tornesome of the code is only for particular devices..
21:23:15Tornebut it's the same for everything :)
21:23:39bertrikIMO, the clip+ is pretty close to stable, but I don't really care about the stable/unstable label anyway
21:24:02 Quit balintx (Remote host closed the connection)
21:24:25 Join balintx [0] (
21:24:30 Quit balintx (Read error: Connection reset by peer)
21:24:37*AlexP labels bertrik unstable :)
21:25:02BuschelLear: now I can approve that r27224 (same is used in 3.6) does not show the effect with aac playback
21:25:09m0arbertrik: Yeah, I feel the same. Haven't noticed ANY bugs in the clip, except the USB-RB-thing
21:25:21 Join balintx [0] (
21:26:08 Quit balintx (Read error: Connection reset by peer)
21:31:20pixelmaLear: moment, I'll pastebin the line together with some intermediates and comments
21:33:30 Join Hillshum [0] (4ba5f419@gateway/web/freenode/ip.
21:33:43pixelmaLear: (took a while, Firefox crashed :\ )
21:35:28funmanAlexP: r27895 : what is short submenu ?
21:35:42AlexPfunman: Nothing
21:35:52Learpixelma: Can you see if the problem started with r27815?
21:36:18AlexPfunman: as in, it is short press on submenu (down), which in the wps doesn't seem to do anything
21:36:24pixelmaLear: I most likely can't as that's between r27773 and r27860...
21:36:40LearAnd those are daily builds?
21:36:45pixelmaunless I backport the fix
21:37:10pixelmaLear: no, that's when playback on hwcodec was broken completely
21:37:21funmanAlexP: does it mean the clip has too much buttons? :)
21:37:24LearAh, ofcourse.
21:37:28AlexPfunman: hehe :)
21:37:34 Quit Judas_PhD (Quit: This is a quitting message)
21:37:56funmanit could alias the home button i think
21:38:16AlexPIf there is no missing function then yeah, seems sensible
21:38:58funmanlong home is pitchscreen
21:39:07funmanlong submenu is playlist
21:39:39funmanis there a pitchscreen wps key on other targets?
21:39:49pixelmaLear: ok, I'll try to backport the fix
21:40:03pixelmafunman: yes, there is
21:40:26BuschelLear: got it -> r27364 introduced the distortions. reviewing now
21:41:01pixelmaon the Ondio it's only available through the WPS context menu though (maybe Ipods too, don't know)
21:41:04funmana it's already: up+select (unfriendly IMO)
21:41:32S_a_i_n_tFor some reason the Nano2G won't "click" when keyclicks are set to 'On' unless there is currently audio being played...and udeas, or, idea where I should start looking?
21:42:08pixelmawhile music paused or stopped?
21:42:49bertrikWhat is the difference between music paused and music stopped again?
21:43:26funman <- short submenu is pitchscreen on clip
21:43:54pixelmafunman: is long submenu the quickscreen?
21:44:11 Join user890104 [0] (~Venci@
21:44:22funmanno it's playlist viewer (or hotkey if hotkey is defined)
21:45:18 Join h3po [0] (
21:45:48bertrikoh, "rewind before resume" is done active in stopped state, not in paused state
21:46:27 Part m0ar
21:46:53pixelmabertrik: something to do with DSP or whatever, it also causes voice to not work when paused
21:50:48funmanAlexP: how does that look to you ?
21:51:05funmannow i might accidentally find the pitchscreen and experiment with it :)
21:51:33funmanbutton combos with only the directional keypad are impossible (at least with my big fingers)
21:51:44AlexPfunman: I haven't bothered to check what other targets do, but it is better than unused :)
21:52:10AlexPand a single press is certainly better than a fiddly combo
21:52:24CIA-81New commit by funman (r27897): Clip keymap : move pitchscreen to short submenu in wps
21:52:30funmanbtw, for clipv1 to be stable, only thing missing is voice IMO
21:52:48funmanand for c200v2, the same thing + use the right button for dualboot in mkamsboot
21:53:03Hillshumfunman: I think the problem is that the pad with only push down in one area
21:53:36funmanHillshum: it seems possible to use 2 consecutive directions, or center + 1 direction (but not without trouble)à
21:53:58Hillshumfunman: Opposite or adjacent?
21:54:23CIA-81r27897 build result: All green
21:54:35funmanalso there are some combos forbidden by the hardware
21:54:58funmanleft+down i think
21:55:19Hillshumfunman: Have you tried opposite?
21:55:26*Hillshum lost his Clip months ago
21:56:12funmanwhen pressing up & bottom i hear too much clicks
21:57:16kaiscenefor the clip's keymap, how about moving the current Select action to short-Submenu (view database/file browser) and then allow WPS makers to use Select as a condition to change the WPS screen?
21:57:16Hillshumfunman: ? As in cracking or something?
21:57:22AlexPkaiscene: That would be a big change for all targets
21:57:35funmanHillshum: i suppose the center button is also being pressed
21:57:41AlexPkaiscene: We try to maintain some sort of consistency where possible, and select does the same on all
21:58:55HillshumThere isn't a debug screen showing what buttons are depressed is there?
21:59:01funmanno, it only shows the gpio status and not all clipv1 buttons will show a result in this screen
22:00:29pixelmakaiscene: and that change WPS state isn't implemented at all yet (*I* wouldn't want it anyway + there aren't enough buttons on all targets even for an already existing feature such as the "hotkey")
22:02:31kaiscenei see, but from a usability point of view, i think it's a mistake to avoid customizing the user input for each device. but from a developer's point of view, it makes sense to keep all targets as similiar as possible
22:03:38AlexPThe idea is you go to any Rockbox device and once you know what each button is, you know what functions are on them
22:03:51kaisceneyea that idea definately has merit
22:03:58AlexPAnd given it is largely arbitrary in the first place
22:04:25AlexPIt isn't always possible of course, but within reason
22:04:55 Quit MichaelAdams (Quit: CGI:IRC (EOF))
22:09:35pixelmaLear: ok, r27815 works with the fix backported, 27860 itself didn't. I'll go on with bin chopping then
22:10:06 Quit kramer3d (Ping timeout: 272 seconds)
22:10:23pixelmanothing to do with 27773 which I suspected first
22:12:18pixelmabut thanks for the input :)
22:13:11funmanwhat's the status with WPS syntax changes? is that over and tested?
22:14:09Learpixelma: Maybe r27846 is the problem...
22:15:53 Quit funman (Quit: free(random());)
22:17:01 Quit Hillshum (Quit: Page closed)
22:18:42pixelmawill test that next, I currently compile r27838 (just picked by rev number for a binary search)
22:19:25 Join user890104_ [0] (
22:19:53 Quit user890104 (Disconnected by services)
22:19:58 Nick user890104_ is now known as user890104 (
22:21:28pixelmasounds likely though - something to do with true and false branches in conditionals
22:21:48Learpixelma: Next skin-related change is in 27883.
22:22:40Learpixelma: But mainly for the new %if tag. It does increase stack usage though.
22:23:47pixelma27860 was already broken, now lets see what 27838 does
22:26:44 Join MethoS- [0] (~clemens@
22:26:58Learr27846 is pretty much the only related change left then.
22:28:24pixelmawhat I already tried btw. is enable "catch mem accesses" from the debug memory but it didn't trigger
22:35:17Learpixelma: You could see if this patch helps:
22:35:36LearNot very good if you use %?if, but just for testing...
22:36:14pixelmaI don't use %?if, I'm going to try this next
22:37:48CIA-81New commit by Buschel (r27898): Fix residual aac-he distortion reported via forum. Bug was introduded with r27364.
22:39:39CIA-81r27898 build result: All green
22:40:00 Join einhirn [0] (
22:41:20pixelmaLear: r27846 in itself is definitely broken
22:45:51BuschelLear: seems like r27898 also fixes FS #11540. now the distortion became obvious to my eyes. with r27898 it's gone :)
22:46:10Buschelahem, replace "eyes" with "ears"... :/
22:48:40pixelmaLear: your small test patch helps. :) So you have an idea what's going on?
22:49:00Learpixelma: Looks like stack overflow after all.
22:49:41pixelmaonly when playback starts then
22:50:36LearYes, the skin engine needs to start drawing the wps.
22:50:49pixelmaand I still wonder why it only happens with such a line in a WPS (there are other complex ones, maybe not so many nested conditionals though probably)
22:51:27LearThat nesting adds a few hundred bytes of stack usage. Could be enough to bring it over the edge.
22:52:41bertrikpreglow, do you know details about the crossfeed algorithm?
22:52:53***Saving seen data "./dancer.seen"
22:53:28pixelmaLear: could screen corruption which I experienced sometimes after loading themes also be caused by this (misplaced pixels)? Though they survived reboots and leaving/reentering didn't help, just loading a different theme... worst example I've seen so far:
22:54:31pixelmathat's the Archos cabbiev2 which didn't stall playback but often has a slightly misplaced progressbar
22:55:03pixelmaand a few extra pixels in the top left corner
22:56:10Learpixelma: Surviving reboots sounds odd, but screen corruption as such wouldn't be impossible. Depends on the memory layout.
22:56:14CIA-81New commit by alle (r27899): Describe the default value of the new optional tag parameter
22:57:30n1smemory corruption can cause all kinds of fun things :)
22:58:03CIA-81r27899 build result: All green
22:58:39 Quit n1s (Quit: Lämnar)
22:58:43pixelmaLear: ok, thanks for taking the time and the explanation. Do you already have an idea what or how to fix, or is it tracker timeß
22:59:37 Quit Jaykay (Ping timeout: 252 seconds)
22:59:40 Quit h3po (Quit: Leaving.)
23:00:54 Join Strife89 [0] (
23:02:22 Quit Buschel (Ping timeout: 265 seconds)
23:02:47 Quit petur (Quit: here today, gone tomorrow)
23:06:42Learpixelma: Making that temp_buf static might work... Letting JdGordon have a look at it would be good.
23:07:07 Join Hillshum [0] (4ba5f419@gateway/web/freenode/ip.
23:08:36 Quit Lear (Quit: ChatZilla 0.9.86 [Firefox 4.0b5pre/20100826030920])
23:12:31 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
23:16:15pixelmabertrik: not too long ago you said that the Clip's cabbiev2 looked weird, do you remember why you said that?
23:17:12bertrikpixelma, corruption that looked like what you just posted, and also the progressbar not showing properly (only the part that normally fills up was showing)
23:18:06pixelmaso I assume it went away after a while, or a reboot, or a theme reload
23:18:37bertrikI didn't really investigate it
23:18:47bertrikIt seems to work now
23:20:10pixelmaI believe amiconn said something similar too (screen corruption in the top left corner) about the Clip+, I wonder if it is related to one of the things mentioned here or not and how one can report it
23:24:23 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
23:32:45pixelmasomeone with a flashed H100 around or so?
23:34:04pixelma(I'm not sure which targets have "early USB" besides the Archoses and flashed H100, maybe there are others)
23:41:11 Join milz [0] (
23:42:49TorneAnyone got an ipod video, especially a 32MB one?
23:43:00TorneCan you test
23:43:08Torneit should do a splash on boot for 5 seconds telling you how much ram you have
23:43:09 Quit milz (Read error: Connection reset by peer)
23:43:21Tornei'm interested to know if it's actually right ;)
23:45:05 Quit Misanthropos_ (Ping timeout: 264 seconds)
23:46:00*Bagder adds committer #93
23:46:15AlexPWho isn't here to congratulate :)
23:46:26AlexPpixelma: Yes, somewhere :)
23:46:57BagderFred Bauer
23:47:27pixelmaAlexP: if you find it could you test early USB - plugging USB from an off-state, watch the status bar and what happens when you disconnect?
23:47:47AlexPWould you like a current build?
23:48:40pixelmayes, though the bug is already there for a while (since some SBS stuff)... oh and maybe a paper clip could be handy
23:48:41 Quit bmbl (Quit: Bye!)
23:49:09AlexPOK, give me a few minutes
23:52:24 Quit krazykit (Quit: Connection reset by beer)
23:52:25 Join froggyman [0] (~4845cf9e@unaffiliated/froggyman)
23:56:00AlexPpixelma: H100 has bootloader USB
23:56:40pixelmaoh, that has a different (or no) real USB screen then :\
23:56:58AlexPWhich hasn't been touched in years and works fine
23:58:22pixelmaok, thanks for testing anyway, not what I needed though then but I really don't know which other targets have that (maybe I remember wrongly that there are a few more)
23:58:29AlexPno worries
23:58:48 Join _s1gma [0] (~d.d.derp@

Previous day | Next day