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 2010-07-28

00:02:07 Quit AlexP (Ping timeout: 276 seconds)
00:02:42 Join AlexP [0] (~alex@rockbox/staff/AlexP)
00:03:56 Quit Jerom (Read error: Connection reset by peer)
00:04:31 Quit krazykit (Ping timeout: 260 seconds)
00:04:50 Quit wodz (Ping timeout: 246 seconds)
00:05:38 Join wodz [0] (
00:06:29***Saving seen data "./dancer.seen"
00:07:22 Quit ender` (Quit: Never say "Oooops" - always say "Ahhh, interesting...")
00:12:57 Join wodz_ [0] (
00:13:15 Quit jgarvey (Quit: Leaving)
00:15:22kugelour thread engine works on android :
00:15:50 Quit wodz (Ping timeout: 258 seconds)
00:17:13kugeljhMikeS: ping
00:18:54 Join [Saint] [0] (S_a_i_n_t@
00:19:08 Join krazykit [0] (~kkit@
00:19:40 Quit S_a_i_n_t (Ping timeout: 258 seconds)
00:20:03 Quit AndyI (Ping timeout: 248 seconds)
00:23:03 Join nosa- [0] (
00:27:19 Quit bertrik (Quit: :tiuQ)
00:27:54 Join halmi_ [0] (
00:30:16 Quit halmi (Ping timeout: 240 seconds)
00:43:08 Quit wodz_ (Quit: Leaving)
00:46:18 Quit domonoky (Read error: Connection reset by peer)
00:46:59 Join Strife89 [0] (
00:50:19 Join peterwishere [0] (
00:51:25peterwishereis it appropriate to report a bug to the bug tracker even on an unstable port (sansa clip v2) ?
00:52:45kugelpeterwishere: I think we haven't decided that yet :)
00:54:41 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
00:56:47 Quit peterwishere (Quit: CGI:IRC (Ping timeout))
01:00:21dfkti wonder what his bug report would have been? ;)
01:01:16 Quit DerPapst (Quit: Leaving.)
01:15:50 Quit bmbl (Quit: Bye!)
01:16:42 Join Jennifur [0] (
01:17:27 Quit bunnyboi (Disconnected by services)
01:23:39 Join chronon [0] (
01:27:17 Quit JdGordon (Ping timeout: 276 seconds)
01:30:55 Nick 36DAAB493 is now known as advcomp2019 (
01:31:04 Quit advcomp2019 (Changing host)
01:31:04 Join advcomp2019 [0] (~advcomp20@unaffiliated/advcomp2019)
01:37:59 Quit kugel (Remote host closed the connection)
01:52:41 Join CaptainKwel [0] (
02:06:30***Saving seen data "./dancer.seen"
02:11:27 Quit chronon (Quit: chronon)
02:12:13 Quit Xerion (Read error: Connection reset by peer)
02:12:15 Quit halmi_ (Read error: Connection reset by peer)
02:12:49 Join halmi [0] (
02:12:56 Join Xerion [0] (~xerion@
02:15:50 Quit komputes (Excess Flood)
02:16:50 Join komputes [0] (~komputes@ubuntu/member/komputes)
02:17:01 Join JdGordon| [0] (
02:17:01 Quit JdGordon| (Changing host)
02:17:01 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
02:17:12 Quit halmi (Read error: Connection reset by peer)
02:22:02 Join kugel [0] (
02:22:58 Join czth [0] (
02:23:15 Join CIA-97 [0] (~CIA@
02:23:27 Quit CIA-71 (Ping timeout: 276 seconds)
02:26:17 Quit simonrvn (Quit: brb)
02:27:10czthSince the SansaAMS forum thread ( is locked, is there anywhere else to get current information about USB support? I can't find a current thread on it.
02:27:21 Quit stripwax (Read error: Connection reset by peer)
02:27:47krazykitczth, the wiki
02:29:27czthThat's where I started. :(
02:29:47czthSeems odd to close the forum thread with issues outstanding.
02:34:37mc2739czth: that is not odd at all. That thread was for the "New Port" and SansaAMS is no longer considered a new port.
02:37:22 Quit kugel (Quit: Bye)
02:37:30 Join kugel [0] (
02:38:40 Quit kugel (Client Quit)
02:40:16 Quit Battousai (Ping timeout: 265 seconds)
02:40:50 Join simonrvn [0] (
02:42:54czthmc2739: how is current work on USB tracked then? just SVN, or is it considered sufficient to use the OF for USB?
02:44:14 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
02:44:49mc2739czth: yes, svn and the the SansaAMS wiki page
02:46:47mc2739for the SansaAMS v2 devices, the OF is currently the only choice for USB connectivity
02:50:13 Join Battousai [0] (~bryan@gentoo/developer/battousai)
02:50:50 Quit MethoS- (Remote host closed the connection)
02:51:29czthfrustrating that SanDisk won't give out the specs
02:56:11 Join shai_ [0] (
03:00:02 Quit shai (Ping timeout: 276 seconds)
03:06:30 Quit komputes (Remote host closed the connection)
03:08:22 Join hebz0rl_ [0] (
03:11:23 Quit hebz0rl (Ping timeout: 245 seconds)
03:16:26 Quit hebz0rl_ (Quit: Ex-Chat)
03:19:27 Quit pixelma (Ping timeout: 246 seconds)
03:19:55 Quit amiconn (Ping timeout: 264 seconds)
03:33:48 Join pixelma [0] (quassel@rockbox/staff/pixelma)
03:34:21[Saint]why *would* Sandisk give out the specs...really?
03:34:23 Join amiconn [0] (quassel@rockbox/developer/amiconn)
03:35:03[Saint]Any dev here using a .kbd file for the virtual keyboard?
03:35:15*[Saint] wants to see if he can get someone to confirm a bug.
03:38:37[Saint]With current SVN the config.cfg file is definitely storing the info for the .kbd file, but the keyboard mask is not being applied at startup.
03:38:57[Saint]If I want to use a custom keyboard mask now I have to load it manually each time.
03:39:26JdGordon|keyboard mask?
03:39:43[Saint]didn't know what else to call it.
03:39:50[Saint]the .kbd ile
03:40:09JdGordon|I dont think the keyboard really gets much use (from devs anyway). If it did it would have been imporved by now
03:40:16[Saint]A lot of people don't use them, I know.
03:40:34[Saint]However, it *used* to apply at boot.
03:41:08 Join pokes_ [0] (
03:41:10[Saint]bah! s/ile/file/
03:41:16JdGordon|time to did into the code I think... if you really want to figure out what is happening anyway
03:41:35JdGordon|bassically check settings.c for how it uses the .kdb file setting
03:41:41JdGordon|make sure that is in settings_apply
03:42:07[Saint]heh, yeah...that'll get me *real* far ;)
03:42:15[Saint](though..I'll try)
03:42:54[Saint]Whatever is wrong, it used to only affect the sim.
03:43:10[Saint]The sim just *never* applied a .kbd file unless it was done manually.
03:43:14JdGordon|is there a bug report?
03:44:32[Saint]Nope, I'll have to make one. When I mentioned it about the SIM I was basicaly told "the sim is broken in lots of places..."
03:44:53 Quit pokes (Ping timeout: 276 seconds)
03:45:30JdGordon|who said that?
03:45:50JdGordon|things on the gui level should work 100% in the sim
03:46:49[Saint]I don't exactly was an age ago. I asked here if it was worthwhile making a bug report about it (re: sim) and got the impression I'd be wasting my time.
03:47:29[Saint]It *could* have just been the impression I got, but, it managed to convince me at the time.
03:47:42JdGordon|that last part is still probably true :) but if it isnt working in the sim there is probably a good reason
03:47:55JdGordon|and could explain why it isnt working anymore on target
03:48:34[Saint]AFAIk it never worked on target (well, in my experience). 'twas one of the first "bugs" I ever noticed.
03:48:48[Saint]Yet, it always (until now) worked on the sim.
03:49:09[Saint]aaaargh! s/target/sim/
03:49:34JdGordon|oh, /me is getting all confused
03:49:45JdGordon|target used to work, sim never did?
03:51:27JdGordon|ok, the keyboard is one part I have no clue about, if you are game open up apps/recorder/keyboard.c and try to see how it loads the setting
03:51:45 Join binaryhermit [0] (
03:51:47JdGordon|either it expects it to be loaded at boot, or it loads it the first time the keyboard is opened after boot
03:52:06JdGordon|global_settings.keyboard or .kbd_file iirc
03:52:30[Saint]Sweet, I'll have a nosey in there.
03:52:36[Saint]And, thanks.
03:52:47 Quit rasher (Quit: Lost terminal)
03:53:36 Join rasher [0] (
03:53:37 Quit rasher (Changing host)
03:53:37 Join rasher [0] (~rasher@rockbox/developer/rasher)
03:55:35 Join LambdaCalculus37 [0] (~rmenes@rockbox/staff/LambdaCalculus37)
04:06:32***Saving seen data "./dancer.seen"
04:07:58 Join Barahir_ [0] (
04:09:35 Quit Barahir (Read error: Operation timed out)
04:12:31 Quit TheSeven (Ping timeout: 248 seconds)
04:13:36 Quit amiconn (Disconnected by services)
04:13:38 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:13:45 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:13:45 Quit pixelma (Disconnected by services)
04:13:45 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:14:00 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:16:29 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:45:03 Join jHoNDoE [0] (
04:50:30 Quit guymann (Quit: emerge -uDNav world)
04:53:34 Quit Jennifur (Ping timeout: 240 seconds)
04:55:09 Part jHoNDoE
04:55:11 Join bunnyboi [0] (
04:57:19 Join Jennifur [0] (
05:00:46 Quit bunnyboi (Ping timeout: 276 seconds)
05:37:43 Join robin0800 [0] (
06:04:19 Quit robin0800 (Quit: Leaving)
06:05:58 Quit LambdaCalculus37 (Quit: bedtime)
06:06:36***Saving seen data "./dancer.seen"
06:14:13 Quit binaryhermit (Quit: Leaving)
06:20:46 Join Phelbore [0] (
06:21:29Phelboremaybe I missed it in the faq and manual, but is there a way to rebind a key? for instance, on a fuze, making the home button go to wps instead?
06:21:54[Saint]Not without editing, and recompiling, the
06:22:16Phelborei was hoping i wouldn't have to do that.
06:22:23 Quit panni_ (Quit: ( :: NoNameScript 3.81 :: ))
06:22:26[Saint]Um, another answer. "Not on the device, no"
06:23:10[Saint]Editing the keymaps is probably one of the easiest things a "non-coder" could do though.
06:23:22Phelborewell, i'm find with messing with the source, i do have a programming background, i was just hoping for a quick and easy way.
06:23:56 Quit anewuser (Quit: Votinz Remix Phase concluded. The votes! The mad votingzes!!)
06:25:57[Saint]Phelbore: Do you already have a development environment set up?
06:26:05[Saint]What OS are you using?
06:26:14[Saint](that would be the first step)
06:26:40[Saint](even if you edit the source code, it's not as simple as just dropping a new file on the device)
06:26:50Phelboreyeah i've got a development environment. i'm running xp now, but i've also got some linux-on-a-thumb-drive systems laying around.
06:27:11Phelborein xp, i've got codeblocks and mingw
06:27:25Phelboreand of course, msys
06:27:36[Saint]Go with Linux ;)
06:28:05[Saint]Or a *nix VM, I myself use CygWin but that seems to be a swear word in here :P
06:29:15Phelborei couldn't get cygwin working on my netbook, but msys seems to be working ok. i suppose... just to be weird, i could run a linux image thru qemu...
06:29:36[Saint]Uuuuurgh...I'd rather use CygWin ;)
06:29:42[Saint]It's really not hard to set up.
06:29:46[Saint]like, really.
06:30:07[Saint]ve a look here.
06:30:13Phelborelet me rephrase... cygwin set up and installed fine. i couldn't get it to play with wxwidgets.
06:31:28[Saint]There is a premade VM on the wiki, but, it is out of date.
06:31:43Phelborebut that takes most of the fun out of it!
06:32:14[Saint]Hahaha, there's nothing fun about compiling toolchains on CygWin ;)
06:32:32[Saint]<set it up>...<wait 12~13 hours>
06:32:38Phelborei wonder if i could make codeblocks compile it... it's a really nice ide.
06:33:03[Saint]You may be able to, sure,'d be on your own.
06:33:17[Saint]The methods on the wiki are there because they are known to work.
06:33:53Phelboreoh, i'm fine with going it alone. sounds like fun, even. i only needed to find out if there was a cheap easy way to do it in a config file
06:34:16[Saint]Well, there isn't ;)
06:34:30Phelborei was kinda hoping there wasn't.
06:34:38[Saint]remappable keys aren't going to happen any time soon I don't think.
06:34:51[Saint]Too much of a chance someone can get into a stupid situation.
06:35:30[Saint]"help, I remapped all my keys to X" for instance.
06:35:38Phelborebah, so just have them reinstall rockbox. not like it'll kill their library
06:35:53 Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier.
06:35:53Phelboreor delete the keymap file
06:36:34[Saint]I think the only way it would be considered is if there were an easy way to revert it from the device.
06:36:47[Saint](speculation, however)
06:36:51Phelborei will... keep that in mind.
06:54:09 Quit nosa- (Disconnected by services)
06:54:17 Quit TheSeven (Ping timeout: 246 seconds)
07:10:11 Quit Farthen (Ping timeout: 260 seconds)
07:18:11 Join Farthen [0] (
07:21:03 Quit CaptainKwel (Quit: Ex-Chat)
07:49:23 Part Phelbore
07:50:24 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
07:51:24 Join stoffel [0] (
08:06:37***Saving seen data "./dancer.seen"
08:10:45 Nick shai_ is now known as shai (
08:18:49 Join ender` [0] (
08:20:09 Quit ender` (Read error: Connection reset by peer)
08:20:18 Join ender` [0] (
08:21:07 Quit Utchybann (Ping timeout: 260 seconds)
08:24:17 Join Utchybann [0] (
08:31:14bieberMeh, qmake seems to be really, really peculiar about using literal $ characters in its function inputs
08:34:04bieberI don't suppose there's any way to get Subversion to use a different character for property substitution?
08:34:17 Quit Jennifur (Quit: Never look down on someone unless you're helping them up.)
08:42:38 Join funman [0] (~fun@rockbox/developer/funman)
08:44:24 Quit xavieran (Ping timeout: 240 seconds)
08:44:44funmanbieber: can tools/ help you?
08:45:44bieberIt looks like it's just a wrapper around svnversion, which I'm using now
08:45:56bieberI'd hoped to avoid it, but apparently SVN doesn't have a keyword for the global revision anyways
08:50:34funmanwhy avoid it? it works fine for rockbox
08:50:54CIA-97New commit by bieber (r27590): Theme Editor: Modified to call svnversion when qmake executes and include the revision in the version number, if building in debug ...
08:51:32bieberBecause it will fail if someone builds on a system without svn
08:52:12[Saint]git users will be happy ;)
08:52:23CIA-97r27590 build result: All green
08:54:14bieberI would assume that a git-svn checkout should still work with svnversion
08:54:17 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
08:57:42bieberNow I just need to find out if I can get at values from the .pro file from the application
09:00:30funmanbieber: how would someone build without svn, do you want to distribute the sources outside of svn repo?
09:01:00bieberI'd like to distribute a tarball, and have a source deb for Ubuntu PPA to build
09:03:59 Join ddalton [0] (
09:04:00funmanput a -DVERSION=xxx inside your debian/rules and use ifndef VERSION VERSION=../tools/ endif
09:04:13ddaltonhi guys, trying to install rb to my sansa e200 v2
09:04:52bieberfunman: You mean in source files to get the version?
09:04:54ddaltoni get "No e200s or C200s found" when running the sansapatcher under linux with sudo (root rights)
09:05:11funmanddalton: sansapatcher is for e200v1, check the manual for install procedure
09:05:23ddaltondo I need to unzip rockbox onto the player first, and I also see there is just sdc no sdc1 (so no partitions ont he device) is this usual?
09:05:39funmanhaving no partitions is quite usual yes
09:07:40ddaltonah right, thanks
09:07:49ddaltonah good:)
09:11:52ddaltonso as there is no partition, am i supposed to create a new one or do I just try and mount the device?
09:12:31ddaltonfunman: *thats for u:)
09:12:55 Quit crwl (Ping timeout: 260 seconds)
09:14:17 Join mt2 [0] (~chatzilla@
09:21:55 Join bimbel [0] (~Miranda@unaffiliated/bmbl)
09:24:35 Join xavieran [0] (
09:25:15 Quit mt2 (Ping timeout: 240 seconds)
09:25:29 Quit bmbl (Ping timeout: 258 seconds)
09:25:51 Join swilde [0] (
09:27:21 Join einhirn [0] (
09:27:58 Nick bimbel is now known as bmbl (~Miranda@unaffiliated/bmbl)
09:39:53 Quit JdGordon| (Quit: leaving)
09:41:35 Quit ddalton (Quit: leaving)
09:47:54 Join mitk [0] (~mitk@
09:54:12 Join Rob2223 [0] (
09:55:02 Join mt2 [0] (~chatzilla@
09:57:35 Quit Rob2222 (Ping timeout: 245 seconds)
10:06:40***Saving seen data "./dancer.seen"
10:21:24 Quit Strife89 (Ping timeout: 240 seconds)
10:22:13 Quit seani (Ping timeout: 258 seconds)
10:26:15 Quit mitk (Ping timeout: 240 seconds)
10:34:06 Join DerPapst [0] (~Alexander@
10:35:10 Join seani [0] (~seani@
10:39:46 Join mitk [0] (~mitk@
10:40:47 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
10:43:57 Join n1s [0] (~n1s@rockbox/developer/n1s)
10:47:39 Quit TheSeven (Ping timeout: 260 seconds)
10:50:35 Join kugel [0] (~kugel@rockbox/developer/kugel)
10:53:55 Quit mt2 (Ping timeout: 240 seconds)
10:56:25 Join crwl [0] (
11:11:09 Join mt2 [0] (~chatzilla@
11:23:25swildehi *, i just updated rockbox from svn (r27590) on my Sansa c200, now the theming of the WPS is broken (no matter what theme I select, its always a dull, text only, not even well fitting). Is this a known problem?
11:24:16funmanupdate your themes too
11:25:52swildefunman: aren't they in the zip generated by 'make zip'
11:26:44funmandefault themes, yes
11:29:04swildefunman: cabbiev2 is one of the default themes isn't it? (its in svn source) And the cabbiev2 wps doesn't work, too...
11:30:11 Join JdGordon1 [0] (~jonno@
11:31:16swildeAny ideas?
11:32:08 Quit mt2 (Read error: Connection reset by peer)
11:32:49 Join mt2 [0] (~chatzilla@
11:32:50 Quit JdGordon (Ping timeout: 265 seconds)
11:33:36swildemitk: Thanks. That explains part of the problem.
11:37:11n1sreproducible in the sim
11:37:24n1sERR: Failed parsing on line 39 : ERR: Invalid parameter list for token 81: "No token"
11:37:48mitkWhat screen size c200 uses?
11:40:09pixelmalooks like a missing comma
11:40:59swildeAh it actually explains everything, iCatcher and Rockboxed do work and all the other themes I have tried are outdated.
11:42:12pixelmaI'll try to fix it in the next hour
11:43:08mitkcompiling sim, can try it. You mean comma between 8 and - ?
11:47:30mitkpixelma: good shot. It works with line 40 changed to %V(86,69,16,8,-)
11:48:17pixelmaI'll take a look myself and then commit (busy with other things though but I guess I can spare some minutes)
11:51:17 Join MethoS- [0] (~clemens@
11:51:55 Quit mt2 (Ping timeout: 240 seconds)
11:52:21JdGordon1pixelma: did you try that hwcodec fix?
11:53:55 Nick JdGordon1 is now known as JdGordon (~jonno@
11:54:03swildemitk: I can confirm that this change fixes it on the real thing™, too. :)
11:54:59 Quit seani (Ping timeout: 258 seconds)
11:55:17 Quit kugel (Remote host closed the connection)
12:01:33pixelmaJdGordon: I just did. It still crashes
12:01:54JdGordonhmm, bugger
12:02:01JdGordonsame general place?
12:02:49pixelmahave to look it up in the map file
12:03:51pixelmait happens at the same point - current playlist finishes and it would have to advance to the next folder, create a new one
12:04:25 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
12:05:50 Join watto [0] (~watto@
12:06:41***Saving seen data "./dancer.seen"
12:07:14JdGordonaynone got a h10 hand?
12:07:15pixelmathe closest match I can find in the map is _parse_cuesheet and _cue_find_current_track, so I guess it's the same places
12:08:15 Join seani [0] (~seani@
12:11:49pixelmahmm, if I enable "Catch mem accesses > Zero area (all)" in the debug menu, I get "I0C: UserBrk at 090051B2" (OndioFM backlight build)
12:12:42JdGordonthat number is meaningless without the .map file
12:13:15JdGordonthis happens when cuesheet is DISABLED?
12:15:11pixelmacuesheet support is disabled
12:19:00JdGordonwell if you can do some builds I have a idea where to start, but otherwise aboslutly no clue what is going on
12:19:57pixelmahuh, that number was with the unpatched build... but the crash happens with the patched one all the same (addresses differs slightly). If I read the map correctly, the address is in firmware/libfirmware.a(strlcpy.o)
12:20:28funmancan you run gdb on hwcodec?
12:21:15pixelmaIIRC not on the Ondio and on the Archoses it needs some soldering (but I could totally confuse it with something else)
12:23:02*JdGordon wants a stack trace!
12:24:24JdGordonthis doesnt make any sense at all...
12:24:35JdGordonpixelma: what about with repeat all?
12:24:51pixelmaisn't it needed to find a new folder to play?
12:25:16JdGordonisnt what?
12:25:17pixelmadoes auto-change directory do something with repeat all? I can try though
12:26:37pixelmaI thought strlcpy is some string manipulation thing and *imagined* it is needed to find the path of the next folder
12:27:13JdGordonthe problem is it is somehow getting into the cuesheet code when it bloody shouldnt be!
12:27:29JdGordontry ending a playlist with repeat all enabled
12:28:50*[Saint] sees he missed a comma.
12:30:38pixelmaJdGordon: no crash with repeat all enabled
12:31:02pixelmaand the playlist starts again correctly
12:31:09JdGordoncan you do a new build?
12:32:09JdGordonmpeg.c line 2073, comment that whole if block out
12:32:14JdGordonof just make it "if (0)"
12:32:41pixelmaswilde: would you be opposed to me putting your full real name into the commit message (whois told me)?
12:33:21 Join dfkt [0] (dfkt@unaffiliated/dfkt)
12:35:14 Join robin0800 [0] (
12:35:32pixelmaJdGordon: till line 2084, I guess (just making sure as there also is a nested "if")
12:36:05 Join Kitar|st [0] (
12:39:22 Nick evilnick- is now known as evilnick (~evilnick@2001:0:4137:9e76:3cde:1636:b5a6:f655)
12:39:41 Quit Kitr88 (Ping timeout: 245 seconds)
12:40:17 Quit Kitar|st (Ping timeout: 240 seconds)
12:40:23 Join Kitar|st [0] (
12:44:23JdGordon[Saint]: how'd you go with the keyboard thing?
12:44:30 Join mt2 [0] (~chatzilla@
12:46:24pixelmaJdGordon: still crashes in the same place when the folder advance is due
12:46:40JdGordonok, good (!)
12:47:23JdGordonI need the exact function name... which of the two is it?
12:47:42pixelmathat was with repeat off, do you need with repeat all enabled again)?
12:49:28JdGordonno, thats fine
12:49:29JdGordonrep all was just a sanity check thing
12:49:56mt2n1s: ping
12:49:59pixelmawhat do you mean with the two functions?
12:50:14JdGordonthe patch I put on the tracker didnt work?
12:50:48[Saint]JdGordon: I've had a look in apps\recorder\keyboard.c and nothing jumps out at me, then again I could be missing something blindingly obvious.
12:51:30JdGordonpixelma: you're not sure which function the crash address is? I *really* need one funcion name
12:51:44JdGordonwho can tell I'm procrastinating from fixing more bugs in my patch :p
12:52:26pixelmaI guess I need objdump again (and forgot the syntax again :\ )
12:53:31JdGordonwhats the address? and the adrees of both those funcs?
12:53:39JdGordon[Saint]: which directory is your .kbd in?
12:53:59pixelmathe crash I now describe is with your patch applied (had to apply it manually from the .rej btw. as your patch didn't want to apply)
12:54:32[Saint]config.cfg never used to have to problem loading it from there.
12:54:39JdGordonwell there's the problem :)
12:54:45n1smt2: pong
12:54:49JdGordonthey should be in .rockbox/foo.kbd
12:55:18[Saint]it *used* to be loaded by my .cfg though...something changed?
12:55:38mt2n1s: Do you know if wma could do mdct, windowing and overlap in-place without needing a scratch buffer ?
12:55:58JdGordonI guess it could actually work if the keyboard/file bit is short enough
12:56:20[Saint]Ah, is it a path length thing?
12:56:30[Saint]the .kdb name recently got longer.
12:57:27JdGordon. /.rockbox/[folder/filename].kbd... the part is the []'s must be under abot 30 chars
12:57:44JdGordonshorter maybe
12:58:20*pixelma wonders since when there are audio dropouts after skipping tracks :\\
12:58:46n1smt2: i don't know that, asking saratoga is probably a good idea since he ported the wma codec
12:59:22[Saint]If my counting is correct, on device it will be 27 chars not including extension
12:59:38[Saint]I'll try a shorter name on the device when I get near one.
13:00:43mt2n1s: no problem he's away for the week, I thought I'd ask you since you did some work on wma. What was the memory problem for clip and wma (or am I thinking of something else :) )?
13:01:21pixelmaJdGordon: the crash is at 09011134, in the map I can find 0x0000000009011124 _cue_find_current_track
13:01:45JdGordonprobably the one then
13:02:04n1smt2: that was me, I merged some ffmpeg changes to the vlc decoding code which introduced a pretty large malloc which i replaced with a static buffe that made the wma codec too big for the clip codec buffer but i managed to shoehorn it back in
13:02:26 Quit T44 (Read error: Connection reset by peer)
13:03:23mt2n1s: Ah ok. I was looking into dropping the output buffer in codecs/wma.c like what was done in wma pro, which would save ~8K, but that's not needed now I guess ?
13:04:07n1smt2: it would be a good thing anyway i think
13:05:10JdGordonpixelma: add "id3->cuesheet = NULL" before the return in mpeg.c line 2084 (and you can re-enable that if although it shouldnt make a difference)
13:05:25n1smt2: ah, wma does interleave too
13:07:07n1sgetting rid of that would probably speed up decoding quite a bit (like for wma pro)
13:08:24mt2Well, I can get rid of interleaving - actually I've done it already - but will still need the scratch buffer.
13:08:34mt2so no memory saving
13:08:49mt2but maybe just some speedup
13:08:54pixelmaJdGordon: I guess that line needs a ; at the end?
13:09:08JdGordonwell yeah :)
13:15:29pixelmastill crashing, address is 09011190, _cue_find_current_track starts at 09011180
13:16:14pixelmaat least I found out that the audio dropouts after skipping tracks (which is not essential to the actual test) seem to be a case of bad MP3
13:16:29funmancan you disassemble _cue_find_current_track ?
13:16:42pixelmaif you tell me the syntax :)
13:17:13n1surg, wma is memmovig data too
13:19:04n1spixelma: sh-elf-objdump -d rockbox.elf > disasm.txt should do it
13:19:58JdGordonfunman: the problem is more how the bloody hell it is getting into that function
13:20:12JdGordonall calls to it make sure id3->cuesheet != NULL
13:20:24funmanpanic could print a bit of stack / link register
13:21:00funmandoes amiconn have a soldered archos for gdb?
13:22:08pixelma 9011190: 97 1c mov.w 90111cc <_cue_find_current_track+0x4c>,r7 ! 0x2d8
13:22:45pixelmafunman: no, and he's currently away
13:24:44 Join stooo [0] (
13:24:49 Part stooo
13:25:04mt2Could someone please measure the speedup for on a CF target ? (and hopefully an ARM target too since I don't have my e200 now)
13:25:34pixelmaJdGordon: is the completer _cue_find_current_track
13:25:44pixelmacomplete too
13:26:10 Join pyro_maniac [0] (
13:27:04 Join kugel [0] (~kugel@rockbox/developer/kugel)
13:27:16*pixelma needs to turn to RL things though
13:35:26 Quit mt2 (Ping timeout: 258 seconds)
13:41:29n1smt: sure, i'll measure on my h300
13:52:59 Join mt2 [0] (~chatzilla@
13:53:51n1smt: speedup of about 3.5MHz on the 128kbps wma sample :) buuut there's just garbage in the right channel :(
13:55:29mt2n1s: ! it worked fine for me here :(
13:55:40mt2I'll test on that sample
13:58:39n1smt2: it's the same in the sim
13:59:43mt2n1s: Ah yes I hear it now. apparently the sample I tested on had the same garbage but it was very quiet that I missed it. unline the 128k sample
14:04:15 Quit czth (Quit: CGI:IRC (Ping timeout))
14:05:25n1smt2: you moved the memove before the call to wma_decode_frame, maybe that matters?
14:06:45***Saving seen data "./dancer.seen"
14:07:12mt2n1s: the memove afaiu is for reusing half of the output for the next block.
14:07:12mt2so it's safe to be done before wma_decode_frame (like in wma pro)
14:08:59n1sit should in fact be a memcpy, since the halves of the buffer naturally doesn't overlap
14:09:33n1sthat probably doesn't matter much for speed tho
14:11:31 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
14:18:03 Quit bluebrother (Disconnected by services)
14:18:07 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
14:19:10amiconnfunman: Not readily available, but soldering the serial onto the Player isn't difficult
14:19:43amiconnThatM'll have to wait until I'm back home though
14:20:52JdGordonamiconn: 11355 almost certainly cant be fixed without some gdb happening... so when you get a chance
14:24:43 Part pokes_
14:37:35 Quit t0rc (Quit: Give someone code, help them with one project. Teach someone to code, help them rule the world.)
14:40:31mt2n1s: Fixed :
14:41:06n1smt2: nice, i'll test
14:41:07mitkmt2: clip+, wma 128 kbps sample: svn 33.61 MHz, with your patch - 31.42 MHz
14:41:13mt2Great, Thanks !
14:41:37mt2mitk: Thanks :)
14:41:43 Quit advcomp2019 (Read error: Connection reset by peer)
14:42:10 Join advcomp2019 [0] (~advcomp20@unaffiliated/advcomp2019)
14:42:28mt2n1s: I think you could skip redoing the speed test.
14:42:35mt2nothing different there
14:45:23CIA-97New commit by mt (r27591): Modify WMA to produce non-interleaved stereo output (FS #11503 by me). speeds up wma by 3.5MHz on CF and 2.2MHz on ARM.
14:46:59CIA-97r27591 build result: All green
14:48:33 Join kyle6513____ [0] (
14:48:39kyle6513____hey guys
14:48:50kyle6513____has anyone heard anything from the linux 4 nano team?
14:49:46 Quit robin0800 (Quit: Leaving)
14:50:13 Join robin0800 [0] (
14:51:08pixelmaswilde: did you see my question earlier?
14:51:28funmankyle6513____: no, and it's offtopic on #rockbox
14:54:13kyle6513____funman: appologies
14:55:20swildepixelma: oh, sorry, was away some time, using my real name is fine by me.
14:56:22kyle6513____funman: I just thought considering the fact that they are working towards getting rockbox on the Ipod Nanos, it would be appropriate to ask here considering their IRC channel is dead.
14:57:54funmanwhich ipod nano are you talking about?
14:58:17CIA-97New commit by pixelma (r27592): Fix broken cabbiev2 for the c200 screen (in r27574) as pointed out by Sascha Wilde in IRC.
14:58:43 Quit kugel (Remote host closed the connection)
14:59:20kyle6513____funman: well about a year ago the ipod nano second generation has rockbox ported (but unstable) as you may know and ibugger has been ported to the ipod nano 4th generation and I'm attempting to get in touch with the team again
14:59:55funmanafaik there's no team so feel free to start the work
15:00:32pixelmaswilde: if update your build, cabbiev2 should be working again. Thanks for reporting
15:00:44kyle6513____funman: heh, way out of my abilities there, besides since the firmware is encrypted I would need their exploit they have already managed to find.
15:01:03CIA-97r27592 build result: All green
15:01:08pixelmafunman: TheSeven started off as linuc4nano dev IIRC
15:01:13kyle6513____funman: but anyway thats far from the point, thanks.
15:01:31swildepixelma: I already patched my cabbiev2.wps in place (and reported here, that it helped)
15:01:33 Join Strife89 [0] (
15:01:47kyle6513____pixelma: any idea where he might be?
15:01:48 Quit antil33t (Read error: Connection reset by peer)
15:01:49pixelmaok :)
15:01:52swildepixelma: but I'll be happy when it's still working on the next update... ;-)
15:01:54 Join antil33t [0] (
15:02:43 Quit simonrvn (Read error: No route to host)
15:05:05 Quit Strife89 (Client Quit)
15:06:37 Quit MethoS- (Remote host closed the connection)
15:07:02JdGordonanyone got a h10?
15:07:35 Quit kyle6513____ (Quit: CGI:IRC (EOF))
15:08:51JdGordongevaerts: ping
15:09:03 Join MethoS- [0] (~clemens@
15:10:40 Nick mt2 is now known as gevearts (~chatzilla@
15:11:24 Join WalkGood [0] (~FU@unaffiliated/walkgood)
15:12:23 Nick gevearts is now known as mt2 (~chatzilla@
15:13:46 Join giovanni__ [0] (
15:14:15 Join simonrvn [0] (
15:16:54JdGordonis ROCKBOX not defined for the player build?
15:17:35JdGordonit is... so why the undefined referewnced?
15:19:15 Quit krazykit (Ping timeout: 276 seconds)
15:21:27 Join panni_ [0] (
15:21:56 Join detrickm [0] (
15:25:41 Join fdinel [0] (
15:25:54mt2JdGordon: h10 20gb compiles fine for me
15:26:11JdGordonyeah, 2 seperate issues :)
15:26:19mt2ah alright
15:30:05 Join jgarvey [0] (
15:32:28swildeYet another stupid question: is it intentional that not all fonts from svn/fonts are installed (using the zip file)?
15:32:43 Join krazykit [0] (~kkit@
15:32:47Torneyes, the fonts change rarely and are large. we normally distibute the font pack seperately
15:32:51Torneif you want them included, make fullzip
15:38:39 Quit simonrvn (Quit: .)
15:43:13 Quit storm` ()
15:44:46swildeTorne: Ah, thanks!
15:48:18 Quit mitk (Quit: Leaving)
15:50:32 Join simonrvn [0] (
15:51:40 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
15:53:46*amiconn wants 'make full7zip'
15:54:54JdGordonin about an hour the link to my builds will have all the updated builds... please test them
15:55:02 Quit JdGordon (Quit: Leaving.)
16:01:12CIA-97New commit by mt (r27593): Enable wmapro compilation for lowmem targets by excluding multichannel buffers for them.
16:02:54CIA-97r27593 build result: All green
16:06:47***Saving seen data "./dancer.seen"
16:10:16 Quit mt2 (Ping timeout: 240 seconds)
16:11:41 Join ajb [0] (
16:18:02 Quit cornflaku ()
16:18:53funmanmt: does it mean mono-only, or stereo+ will be decoded as mono?
16:21:11n1sfunman: no, it means that they will not support files with more than 2 channels, afaict
16:21:12mtfunman: means mono and stereo stream will work on those targets, but if at some point >2-channel streams are supported they'll not work on those targets.
16:22:15funmanstereo is multichannel, too :)
16:23:31mtYeah, but multichannel here was usually used to refer to >2 channel streams. But right, maybe I should have made the commit message clearer
16:24:01 Quit funman (Quit: free(random());)
16:26:38mtpixelma, AlexP : Do we change something in the manual for new codecs or just new file formats ?
16:27:33AlexPmt: Both :) e.g.
16:28:06AlexPmt: I assume r.e. wma pro?
16:29:41AlexPmt: also if %fc has changed
16:29:42mtAlexP: Yeah, but the RealAudio/Sony Audio entries there had me confused. I thought they would be Atrac3, cook, a52, ... etc. ?
16:30:40AlexPmt: I don't think there is a "rule" - it just depends who added it. If it were me I'd have wma and wma pro seperate I think
16:31:02AlexPDepends how different they are I suppose
16:31:06AlexPYou can chose :)
16:31:48 Quit giovanni__ (Quit: CGI:IRC)
16:32:22mtAlexP: Is that alright ?
16:33:53AlexPmt: A space between Standard and &, but yep :)
16:35:02mtAlexP: Ok, Thanks :)
16:40:50CIA-97New commit by mt (r27594): Add an entry for WMA Pro in the manual.
16:42:25 Join Topy44 [0] (
16:42:37CIA-97r27594 build result: All green
16:45:09 Quit Topy44 (Read error: Connection reset by peer)
16:45:25 Join auTONYmous [0] (
16:46:16auTONYmoushey, anybody here?
16:47:43mtauTONYmous: Just ask what you want and if someone is able to help they will.
16:48:30auTONYmousQuestion: I've used three different Rockbox builds (27591, one slightly older, I forget the build number, and 3.6 stable), and all have the same problem on a iPod Video 80G: track skip forward/reverse don't work in now playing
16:48:54auTONYmousholding for seek forward/reverse works fine, but skip doesn't
16:53:51 Join Topy44 [0] (
16:54:13 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
17:02:38 Join Strife89 [0] (
17:05:27pixelmaauTONYmous: maybe you enabled "prevent track skipping" accidentally (don't ask me where the setting is, I just know it exists)
17:05:44 Join AndyI [0] (~pasha_int@
17:09:52auTONYmouspixelma: You sir, are correct! Why did I think that setting was related to buffering somehow?
17:10:21pixelmaI'm not a sir though ;)
17:11:01auTONYmoussorry...the iNterWeb is so anonymous, it's hard to be polite
17:11:34pixelmano problem at all :)
17:13:27TorneauTONYmous: i suspect the answer to your semi-rhetorical question is "because you didn't read the manual" :)
17:14:02Tornerelevant followup: could we call it something else that would be clearer?
17:14:12auTONYmousyou're right. I didn't RTFM.
17:14:32Tornebecause it is a kinda tempting setting
17:14:35Torneis there a better name? :)
17:14:53auTONYmousI'll think on it
17:15:55auTONYmousoff the top: "Prevent MANUAL track skipping"
17:16:34 Quit robin0800 (Quit: Leaving)
17:27:47*[Saint] is of the opinion that a lot of people mistakenly believe it is the same as the similarly named "anit-skip" on CD's to hard to account for assumption, whatever its called someone will mess it up.
17:28:13[Saint]solution: Append RTFM to all menu entries ;)
17:28:32TorneI don't know that that's necessarily true
17:28:43pixelmathere's even a "Anti-Skip Buffer" setting on HD targets
17:28:52Torneit's a plausible confusion, so if there's a better name then that'd be good
17:30:11*[Saint] worked it out pretty easily by noting that the setting was present on his 'Flash' targets.
17:30:29[Saint]For HDD targets, OK, I guess I can see the confusion.
17:31:29 Quit BlakeJohnson861 (Ping timeout: 265 seconds)
17:35:54 Join Strife89DS [0] (
17:36:45evilnick_B"Prevent track changes"? It is pretty difficult to come up with a name that's good enough for this
17:37:47[Saint]too vague INO (changes)...and yes, it is indeed.
17:37:58Torneyes, or i would've changed it already :)
17:39:13[Saint]I think it's a completely valid name, there simply no accounting for people not R-ing the F-in' M.
17:39:55Torneno, that's just not true. there is some accounting for it.
17:40:08[Saint]its simply the mental association that seems to occur between the similarly named setting for portable CD players getting people into trouble.
17:40:10Tornewe don't name all the settings as random strings to force people to look them up in the manual :)
17:42:28 Join bimbel [0] (~Miranda@unaffiliated/bmbl)
17:43:00[Saint]walking into RB with any expectations on the way things should work, or what things do or don't do is asking for trouble from the start...some menu entries would be terribly, terribly long if the need to look certain things up from time to time were eliminated by more descriptive naming.
17:43:51 Quit bmbl (Ping timeout: 258 seconds)
17:43:56 Nick bimbel is now known as bmbl (~Miranda@unaffiliated/bmbl)
17:45:12[Saint]One of the only things I can see helping in this case would be descriptions of the menu entry currently selected available through the context menu.
17:45:36[Saint]But I'm sure there would be a pretty damn big binsize hit if something like that were to go in.
17:45:42 Quit auTONYmous (Ping timeout: 264 seconds)
17:46:12 Join BlakeJohnson86 [0] (
17:46:41 Join auTONYmous [0] (
17:49:12Torne[Saint]: i think you're being far too pessimistic; there are a few recurring cases where people make these mistakes but they are really infrequent; most menu entries really are obvious, or at least sufficiently baffling that people don't guess something *wrong*
17:49:36Tornewe've fixed them before with seeming good results: i changed the USB HID mode text and now people don't tell us they are in multimedia mode any more
17:50:05TorneI've not thought of a good idea for this particular one yet but it doesn't seem like it should be impossible ;)
17:50:32 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
17:50:40TorneIgnore track skip buttons
17:51:08Tornesince that's what it does, really. you can still change track any other way.
17:53:27 Quit BlakeJohnson86 (Ping timeout: 245 seconds)
17:54:38 Join kugel [0] (~kugel@rockbox/developer/kugel)
17:55:01pixelmaI just realised that the problem is not there in German as it has different words for the two meanings of "skip" - guess that applies to other languages as well
17:55:41 Quit bieber (Ping timeout: 260 seconds)
17:55:48Torneit does? what does the word for skipping audio derive from?
17:55:58 Join bieber [0] (
17:56:01[Saint]Context menu "tool-tips" have certainly been discussed before, but, I seem to remember the reaction to the idea being fairly mixed...its not a "NoDo", but, I'm certain it would have its opposition though. I'm sure it could be done in a sane way.
17:56:26Torne[Saint]: well, you could do it quite easily in the same way voicing works: if the file is htere allocate a buffer for it and load it at boot
17:56:33Tornethe actual code for the implementation would be pretty small
17:56:38Tornebut i dunno that it's really worth doing
17:56:54[Saint]I think it is, personally.
17:57:17Tornethe other issue, independant of implementation, is that this is then *another* kind of documentation to maintain
17:57:20 Join komputes [0] (~komputes@ubuntu/member/komputes)
17:57:28Torneunless you managed to generate the tooltips from the manual or vice versa
17:57:43 Join captainkewler [0] (2669ecc2@gateway/web/freenode/ip.
17:57:55[Saint]that would probably be quite possible.
17:58:06Tornepossible, yes
17:58:16Tornebut i'm not volunteering to write the implementatoin of that :)
17:58:16pixelmaTorne: I translated the "track skip" a bit more like "track change" and I'm not sure how the "anti-skip" setting is called exactly but it's probably more like "jumping"
17:58:55pixelmamaybe the "track change" can be an inspiration....
17:59:08Tornewell, in english track change sounds wrong since youc an still change tracks by letting them play out or just by picking them out of the playlist
17:59:12pixelmato find a new wording for the setting I mean
17:59:24Tornei think "ignore track skip buttons" is not too bad, tbh
17:59:30[Saint]out of curiousity...does anyone have an idea how big the manual ends up in .txt format (roughly)?
18:00:30pixelmano, haven't been able to build it - and you can probably forget it for cygwin
18:01:05kugelTorne: that doesn't work if skip length is set :)
18:01:23Tornekugel: ah, true
18:01:35Tornethough that is kinda weird
18:01:50kugelboth where the same feature initially
18:01:55Torne..why is prevent track skipping a seperate setting at all then?
18:01:57Tornethat seems redundant
18:02:05kugelI broke them into two because I find skip length unusable if it prevents track skipping
18:02:15pixelmaTorne: ok, now I'm not so sure about the translation anymore, got to look it up
18:02:58Tornekugel: i don't understand
18:03:02 Join BlakeJohnson86 [0] (
18:03:14Tornei just mean, prevent track skipping could be a skip length
18:03:26Torneskip by track, 30s, 10s, etc, disallow
18:03:26Torneor similar
18:03:36Tornesince then there's just one setting that says what pressing the skip button does
18:04:16kugelI like to be able to use skip length without loosing the ability to skip tracks
18:04:28pixelmadisallow sounds to me like it won't skip at all anymore... which it probably does...
18:04:30 Quit pamaury (Remote host closed the connection)
18:04:33kugelotherwise, if you set it to 30s, you'd need to play the last 30s out before getting to the next track
18:04:59Tornekugel: I can't see any reason why setting the skip length should *ever* have disallowed track skipping
18:05:09Torneif it worked like that before then i'd consider that just broken, rather than a bad UI
18:05:25kugelbut isn't that what you're proposing right now?
18:05:44Tornei'm proposing adding another option to the skip length list which is basically "0s"
18:05:48Tornei.e. don't skip at all
18:05:59Torneand removing the prevent track skip setting in favour of that
18:06:25kugelthen you lose the ability to skip 30s but with track skip prevention
18:06:37Torneyou do indeed.
18:06:40kugelwhich people like to jump around in big audio books
18:06:47*pixelma wonders how many people use "prevent track skip" at all or what the use case is
18:06:48***Saving seen data "./dancer.seen"
18:06:56Torneit seems like an extremely marginal use case
18:07:35Tornepreventing it in general is useful because it avoids losing your place in a long file
18:07:35kugelwell, the developer who made it initially is blind, and he said it greatly helps him enjoying his audio books
18:07:48[Saint]I have an audio-book .cfg I use for that exact use case...skipping around in 30s blocks, but not skipping the whol etrack.
18:07:55Tornebut if you have mapped the skip button to "skip 30s" and you accidentally skip past the end..
18:07:58Tornethen, er
18:08:02Torneyou know you were, previously, within 30s of the end
18:08:03[Saint]I'll admit it may be marginal, but nice to be able to do.
18:08:36kugelTorne: but you can't easily go to 30s before the end of the previous track
18:08:47Tornewhy not?
18:08:52kugelgoing back a track starts it from the beginning
18:08:58Tornewell that's dumb
18:09:15Torneif my ski plength was 30s I'd expect skipping backwards fro the start of one track to tak eme to end-30s on the previous one
18:09:23Torneall my audiobooks are cut into 5-10 minute chunks anyway
18:09:26kugelbut that's not the case
18:09:30Torneso it working the way you're saying would be unusable for me
18:09:53Tornesure. i regard that as a bug :)
18:10:03kugelthat's why there's that prevent track skipping setting
18:10:18Tornemy point is it seems *more useful* to just implement the logic the way i describe
18:10:30Tornesuch that if your skip length is 30s, it skips by 30s
18:10:35Torneeven if it happens to be within 30s of a track start/end
18:10:47Tornecurrently it must special-case being at the start/end of a track, no?
18:10:56kugelbut then you need to implement to start a track 30s before it's end
18:11:04Torneyes, we already can do this
18:11:08Tornebookmarks and resume playback do it
18:11:12kugelwhich is IIRC not doable with some of our codecs
18:11:34Tornewe don't have a way to say "30s from the end" at the moment
18:11:40Tornebut that's just arithmetic
18:11:56[Saint]there's a WPS tag that can do "X before the start/end"
18:11:59Tornethe playback engine already knows how to (try and) resume from arbitrary positions, within the limitations of the codec in question
18:12:10[Saint]so...I assume it is possible.
18:12:13Tornethe only extra logic is to calculate the position by subtracting some amount from the length
18:12:35Torneanyway, assume it's possible: is that better than having two seperate settings? :)
18:12:45kugelI think so
18:12:46Torneor is there an interesting use case it doesn't cover?
18:13:30Torneso instead of specialcasing being at the start/end of a track and turning that into a "normal" track skip event (or ignoring it entirely, if prevent thingy is on).. just do the math
18:13:47Tornethen it behaves consistently whether you have one big long file, or lots of shorter ones
18:14:05Torneit might not be exactly the time you specified when crossing a boundary because of codec limitations, but hey
18:14:10kugelI'll await your patch :)
18:14:21Tornewell, yaknow. maybe.
18:14:39Tornei was just curious why it was this way, once you brought it up :)
18:14:43Tornesince i've never actually used these options
18:16:39kugelI use skip length regularly
18:16:57Tornewell, all my audio books are cut into tracks, or possibly arbitrary chunks
18:17:09*pixelma remembers the days when it was called "study mode"
18:17:13Tornethe ones that aren't, rockbox refuses to play anyway because they are too big for the aac codec to load ;)
18:17:27Torneand i'm not a musician, so don't need to repeat bits of music :)
18:26:43 Quit evilnick (Quit: ZNC+Colloquy+iPhone4)
18:27:12 Join evilnick [0] (~evilnick@2001:0:4137:9e76:3cde:1636:b5a6:f655)
18:30:11 Quit pyro_maniac (Remote host closed the connection)
18:38:52 Quit Strife89 (Quit: Reboot.)
18:39:04 Quit auTONYmous (Quit: Leaving)
18:39:47 Join Lear [0] (chatzilla@rockbox/developer/lear)
18:45:09 Quit captainkewler (Quit: Page closed)
18:50:09 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
18:50:22 Quit swilde (Quit: ERC Version 5.3 (IRC client for Emacs))
18:55:54[Saint]What the hell could possibly be the explanation for a theme not working on a Nano2G sim, but working on the Nano1G sim?
18:56:45 Join hebz0rl [0] (
18:58:50 Join Jerom [0] (~heidi@
19:00:27detrickmare there any plans for an adjustable backlight level for ipods? it would be handy to dim the backlight to much lower levels at times, and I wonder if this would be possible since the backlight fade in/out can be adjusted already.
19:01:00[Saint]This is implemented...
19:01:10[Saint]what model do you mean specifically?
19:02:56[Saint]anyone have any insight as to what this rather interesting sim debug output is trying to tell me?
19:03:01[Saint]ERR: Failed parsing on line 125 : ERR: Invalid parameter list for token 196: "FIXME: TOKEN_MARKER_SOUND + 2
19:03:24[Saint]it's crying about this line: %t(5.0)%s%ac%?cs<iPod||%Sx(Recording)||%Sx(Playlist)>;%t(5.0)%ac%?cf<%cH:%cM|%cI:%cM%cp>
19:03:52[Saint]which incedently, the Nano1G sim has no problem parsing.
19:05:26pixelmadetrickm: which Ipod do you have?
19:05:42 Quit GodEater (Remote host closed the connection)
19:09:09[Saint]aha! the Nano2g needs to have the "Recording" string for this theme to be cross compatible with the Nano1 and 2G
19:09:22[Saint]...what a crock! :|
19:09:42pixelmahmm.. [Saint] already asked that. It's just that "this is implemented" is not completely correct, it misses a "where hardware allows". E.g. you can't adjust backlight brightness on the Minis. Doing the same as backlight fade in and out in software is very CPU intensive and hence eats battery - it's o.k. for a short period like the fading but not for a longer time
19:10:09 Join GodEater [0] (~bibble@rockbox/staff/GodEater)
19:10:51[Saint]pixelma: Yeah, as soon as I pressed enter I remembered that there was at least one iPod that couldn;t do it.
19:13:01[Saint]*duh*... [Saint] feels slightly silly.
19:13:40[Saint]I forgot I could wrap the "Recording" translation with a %?Rp
19:14:34 Nick [Saint] is now known as S_a_i_n_t (S_a_i_n_t@
19:15:23*S_a_i_n_t finds it interesting that the Nano2G even bothers parsing the .fms file
19:17:54bluebroth3rbieber: there's a reason for it also handles git and bzr repositories. svnversion doesn't work on git-svn checkouts (and I guess bzr either)
19:18:23bluebroth3rhowever, has the problem that it doesn't work on Windows. Unless you install sh explicitly of course.
19:19:36bluebroth3rI was thinking about adding a functionality to set a version number via once it's done (need to find out why some variables seem to be empty, but if that's done it should "just work")
19:20:32 Quit antil33t (Read error: Connection reset by peer)
19:20:38 Join antil33t [0] (
19:20:39bluebroth3ralso, version number locations are a bit scattered: Windows can have it in the resource of the file (not added yet, also not for Rockbox Utility), OS X has it in the .plist file
19:21:03bluebroth3rhow does the version set in the .pro file show up? I can't see that getting used anywhere
19:21:10detrickmsaint: I'm looking at it for an ipod mini specifically. couldn't find any documentation on it, nor a way to make it work, but I could be missing it.
19:21:36detrickmpixelma: ipod mini
19:21:56S_a_i_n_tNo, you're quite correct the time I said that I forgot about the mini, well, more the fact that the mini cannot do it.
19:22:05S_a_i_n_tI apologise.
19:22:25 Quit DerPapst (Quit: Leaving.)
19:22:26detrickminteresting. I thought it might be able to, since the fadein/fadeout timing can be adjusted.
19:23:16detrickmI see pixelma's comment too about cpu intensive
19:23:35S_a_i_n_tCorrect, I was just about to repost that ;)
19:23:42detrickmwhich makes sense ... I suspect doing it that way would be something like a loop of "fade in a bit, now fade otu a bit, now fade in, now out ...."
19:30:00mtkugel, : Just noticed that note about adding new "AFMT_*" entries to the end, (ehich I didn't do for wma pro back when I added it), should I change that ? Is there anything I should take care of regarding %fc ?
19:30:29kugelmt: I have no idea what you're talking about :)
19:31:20mtkugel: Alright, thought I'd ask you since I remembered you working with jdgordon on gui stuff, but nvm.
19:33:48*kugel has a go at building a android-raaa main binary
19:37:07 Join Strife89 [0] (
19:39:32preglowrb as an android app? :>
19:40:04preglowhow the hell'd you go about that, considering the weird limits on native code?
19:40:51 Join Stephen___ [0] (
19:41:09S_a_i_n_thave a look at the RaaA (Rockbox as an Application btw, the sights are set further than Android) wiki page.
19:41:14Stephen___Anyone ever taken apart an iaudio x5 ?
19:41:49pixelmadetrickm: it's switching the backlight on and off real quick
19:43:05pixelmaStephen___: early developers did (and there are pictures of it in the wiki) and I saw my M5 open when it got modded a bit. What do you want to know exactly?
19:43:26detrickmpixelma: pwm ... that much I understand well, I've done a bit of PIC assembly :)
19:43:34WalkGood <Stephen___> Anyone ever taken apart an iaudio x5 ? >>
19:43:35Stephen___i have two here wanting to swap the cases but i cant take the motherboard away fom it all the screws are out
19:43:39detrickmbut you're using software pwm, not hardware
19:43:43 Join Jaykay [0] (
19:43:55preglowdid read that, but didn't really make me any wiser. i was under the impression native code only could be run from java code, and native code not really having good access to for example ui functions
19:44:25Stephen___thanks WalkGood but i was wanting a bit deaper than that
19:44:52WalkGoodthen ask the source, he's here
19:44:56Stephen___the screen on one is broke and i want to replace wit with my already broken x5
19:45:35pixelmaI don't remember anything more than the 4 screws
19:46:16S_a_i_n_tAren't there 2 plastic tags/clips that hold the mb in the x5 as well as the screws?
19:46:32*S_a_i_n_t may well be mistaken.
19:46:49pixelmamaybe you have to be careful with the headphone jack on the side
19:47:21Stephen___yeah trying to pry it away
19:50:47Stephen___got the daughterbnoard away
19:54:22Stephen___damn side usb host port gah
19:58:58 Join Buschel [0] (
20:01:42Stephen___got it1
20:05:01Stephen___will have to resolder all teh cables.
20:06:52***Saving seen data "./dancer.seen"
20:11:53 Part henk
20:13:06kugelpreglow: I'm surprised it works well so far :)
20:13:13kugelnative code isn't limited at all
20:13:32kugelI even got our threading to work
20:15:55CIA-97New commit by Buschel (r27595): Submit next part of FS #11498. Unroll loop for minor speedup of libwmapro on ARM (1%).
20:17:42CIA-97r27595 build result: All green
20:17:57CIA-97New commit by Buschel (r27596): Submit FS #11502. Minor optimization to Coldfire assembler in libatrac's fixmul16(). Thanks to Nils Willménius.
20:19:35CIA-97r27596 build result: All green
20:19:42 Quit kugel (Ping timeout: 245 seconds)
20:23:14mtBuschel: Do you know if we could avoid the scratch buffer in wmadeci.c in the mdct, windowing and overlapping (wma_code_block()) ?
20:24:17Buschelmt: no, I do not know this code very well.
20:26:12 Quit stoffel (Remote host closed the connection)
20:26:18 Quit Strife89DS (Quit: Turning off DS.)
20:28:17 Join czth [0] (
20:28:51czth[Saint]: SanDisk might give out the specs because they're more interested in selling hardware?
20:30:07S_a_i_n_tbut, if they gave out the specs freely...then tons of cheap versions would turn up...and, they wouldn't sell as much as they do now.
20:30:17S_a_i_n_tamong other reasons.
20:32:04Buschelmt: your change 27592 (low mem build of wmapro) -> shouldn't WMAPRO_MAX_CHANNELS be set to 2 for low mem targets? and then use #if WMAPRO_MAX_CHANNELS > 2 ato encapsulate the g_out_multichannel buffer?
20:32:55Buschelmt: your change still allows decoding of >2 channels, but will lead to NULLPTR access on lowmem target when accessing these buffers.
20:33:34mtBuschel: I was going to take care of that when I start working on multichannel.
20:33:44mtBut yeah your suggestions is better
20:36:54Buschelmt: is it possible to decode mutlichannel and only use ch0 and ch1 sample output? in this case your solution should be refined via only allow access to first two out-buffers.
20:37:10CIA-97New commit by bluebrother (r27597): Fix a typo.
20:37:14CIA-97New commit by bluebrother (r27598): Restructure and rename ...
20:37:55mtBuschel: You mean rockbox's ch0 and 1 ? (the buffer pointers in codecs/wmapro.c) ?
20:38:19 Join captainkewll [0] (2669ecc2@gateway/web/freenode/ip.
20:38:55CIA-97r27597 build result: All green
20:39:12CIA-97New commit by mt (r27599): Correct the scaling of the coeffs to 17bits.
20:39:55 Join einhirn [0] (
20:40:30CIA-97r27598 build result: All green
20:41:06 Quit bieber (Ping timeout: 252 seconds)
20:41:23 Join bieber [0] (
20:41:30Buschelmt: I am talking of this ->
20:42:06CIA-97r27599 build result: All green
20:43:32mtAlright, to make sure I understand you correctly, are you asking if I'm going to allow for multichannel decoding on lowmem targets using 2 buffers only ? (if not, then I don't really get the question :) )
20:43:42mtBuschel: ^
20:43:57 Quit einhirn (Ping timeout: 240 seconds)
20:44:51Buschelmt: question is, if it is possible to allow support/decoding of multichannel files, but only outputting the first two channels. otherwise we would not support multichannel files at all (on lowmem targets)
20:45:39mtAh ok.. My intention was to disable multichannel support for low mem targets at first.
20:45:50 Quit bieber (Ping timeout: 260 seconds)
20:46:02mtBuschel : Is that why you said my solution would allow for NULLPTR accesses ?
20:46:12 Join bieber [0] (
20:46:23CIA-97New commit by bluebrother (r27600): Add deployment script for Theme Editor. ...
20:47:29Buschelmt: yes. the decoder might still bitstream decode multichannel and will try to requantize and synthesize to multichannel outputs. but onyl 2 output buffers exist. the other are NULLPTRs.
20:47:52CIA-97r27600 build result: All green
20:49:13czthSaint: perhaps not if they gave out only the specs to program them, not the internals.
20:49:14mtBuschel: What I intended was that I'd enable multichannel only for targets with MEMSIZE > 2 and skip those files with an error message for lowmem targets, so no multichannel decoding is going to be attempted in the first place.
20:49:29mti.e no NULLPTRs :)
20:49:33Buschelmt: ok.
20:50:15S_a_i_n_tczth: I sincerely doubt the care in the slightest about people developing third party software for the device if it isn't going to make them any money.
20:50:53Buschelmt: so, then it's just MAX_CHANNELS 2 for lowmem. there is an error message that implies the codec will then return with error. additional benefit is that several other arrays will be smaller as well
20:56:36 Quit Jaykay (Ping timeout: 260 seconds)
20:56:36mtBuschel: was *just* going to say this. :)
20:56:36mtI'm changing it now
20:56:36BuschelI already did
20:56:36Buscheljust testing
20:56:36 Quit Stephen___ (Quit: CGI:IRC (Ping timeout))
20:56:36 Quit detrickm (Ping timeout: 264 seconds)
20:56:36CIA-97New commit by learman (r27601): Update Swedish translation.
20:56:36mtBuschel: I was about to commit anyway, you want to do it ?
20:56:36 Quit czth (Ping timeout: 265 seconds)
20:57:01mtBuschel: Alright you commit then, seems you also took care of some tabs that crept in ?
20:57:16Buschelyep, I did. will commit now
20:57:36 Join insp [0] (
20:57:45insphi all
20:58:13CIA-97r27601 build result: All green
20:58:34 Join bunnyboi [0] (
20:58:40inspgosh, rockbox is great! i found it 2 hours ago, and installed it in anger that my ipod 5.5 lost its database!
20:58:42mtBuschel: alright, and thanks. :)
21:00:42inspand now i am listening to those songs that were in ipods database - really great! I never used itunes i hated that thing so much. I got through with winamp plugin. But now - no more!
21:01:30*S_a_i_n_t suggests you delete said database, and move to a more sensible system.
21:01:48CIA-97New commit by Buschel (r27602): Refinement of initial lowmem handling for libwmapro (r27593). Set maximum supported channels to 2 instead of 8 for low memory targets. This solution ...
21:01:54inspbut i havew a question, maybe somebody can show me the right direction - i changed my ipods battery and snapped the keylock wire.. is there some kind of software keylock in Rockbox?
21:01:54Buschelhere we go
21:01:57S_a_i_n_tyou'll find adding album art a real PITA with the sill truncated system that the iPods use.
21:02:28S_a_i_n_tinsp: Not for your player, no.
21:03:29CIA-97r27602 build result: All green
21:04:02S_a_i_n_tIf you have all your music on hand, you should really start again with a sanely tagged, Music/Artist/Album type directory structure.
21:04:06pixelmainsp: you could enable it for yours in the source - enabling is quite simple, finding a free button (combo) in the WPS maybe isn't
21:04:14S_a_i_n_tIt'll make life using Rockbox slightly easier.
21:05:25inspah my music is on my old pc that i havent turned on for about 2 years. :D but it read the tags really nicely and i can find it all in database by artist , genre , etc
21:06:19 Quit sasquatch (Ping timeout: 240 seconds)
21:06:38S_a_i_n_tyes, but the issue with the iPod directory structure really comes into play with using album art.
21:07:53S_a_i_n_tit's a lot easier to have "Folder.jpg" in the album directory than it is to have perfectly named aa based on the idV3 tags.
21:08:09 Part watto
21:09:54 Quit WalkGood ()
21:12:02 Quit insp (Quit: CGI:IRC)
21:12:05 Join insp [0] (
21:12:07 Quit insp (Client Quit)
21:12:09 Join insp [0] (
21:12:33 Quit insp (Client Quit)
21:12:47 Join insp [0] (
21:14:22 Quit insp (Client Quit)
21:14:43 Join insp [0] (
21:16:19 Quit insp (Client Quit)
21:17:18 Join insp_ [0] (
21:17:44insp_hi im back! installed ChatZilla
21:21:53 Join sasquatch [0] (
21:27:20 Join Venison [0] (~chatzilla@
21:29:26VenisonHow hard of a hit would it take to enable music playback (the players mp3s, oggs, and normal files) while playing gameboy titles or doom? How complicated would the development be to do that? Would a sansa player be able to cope with the hit and which targets would most likely not be able to?
21:31:02S_a_i_n_tI would say that doom+playback would be out of the question for anything other than the gigabeast.
21:31:11S_a_i_n_trockboy, I have no idea.
21:32:26S_a_i_n_tThese are often very low spec devices remember.
21:36:22CIA-97New commit by Buschel (r27603): Add ARM assembler to libwmapro vector_fixmul_scalar(). Speeds up decoding by 1% on PP5022.
21:37:42CIA-97r27603 build result: All green
21:42:56 Quit Lear (Read error: Connection reset by peer)
21:43:18 Join Lear [0] (chatzilla@rockbox/developer/lear)
21:44:16 Quit Venison (Quit: ChatZilla 0.9.86 [Firefox 3.6.6/20100628124739])
22:06:54***Saving seen data "./dancer.seen"
22:07:20 Join Jaykay [0] (
22:08:04 Join wodz [0] (
22:09:43wodzwhat's the performance of CF DMA?
22:12:44wodzI am wondering if it is worth to try using spare DMA to transfer data to lcd. Transfer size isn't big (128 bytes at max).
22:13:02 Quit bieber (Ping timeout: 245 seconds)
22:13:18 Join bieber [0] (
22:14:18amiconnwodz: Imo not worth it on monochrome/ greyscale
22:14:33 Quit Lear (Quit: ChatZilla 0.9.86 [Firefox 4.0b3pre/20100728031105])
22:14:42amiconnOn H300 we use it, on X5 we can'gt
22:14:46 Join komputes_ubuntu [0] (~komputes@ubuntu/member/komputes)
22:15:32 Quit komputes (Ping timeout: 245 seconds)
22:17:37 Quit bieber (Ping timeout: 260 seconds)
22:20:03 Join bieber [0] (
22:22:53 Quit bieber (Remote host closed the connection)
22:24:38 Join evilnick__ [0] (
22:29:33 Join bieber [0] (
22:32:33 Quit bmbl (Quit: Bye!)
22:33:56 Quit evilnick__ (Quit: evilnick__)
22:45:43 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
22:46:55CIA-97New commit by Buschel (r27604): Refactor asm macros in libwmapro's vector_fixmul_() functions. No change to output samples.
22:48:31CIA-97r27604 build result: All green
22:49:27*Buschel needs a reviewer for arm asm
22:49:42BuschelI cannot see the error in this asm ->
22:50:00Buschelthe output is different to the gcc compiled version
22:56:24CIA-97New commit by bluebrother (r27605): Update theme editor app name for OS X bundle. ...
22:58:04CIA-97r27605 build result: All green
23:05:20CIA-97New commit by bluebrother (r27606): Correct OS X post-build bundle copying. ...
23:06:25 Quit jgarvey (Quit: Leaving)
23:06:49CIA-97r27606 build result: All green
23:09:01 Quit Buschel (Quit: ChatZilla 0.9.86 [Firefox 3.6.7/20100713130626])
23:11:16 Quit wodz (Quit: Leaving)
23:15:22 Join anewuser [0] (anewuser@unaffiliated/anewuser)
23:21:20 Join halmi [0] (
23:22:33 Quit halmi (Client Quit)
23:22:51 Join halmi [0] (
23:25:14 Quit ender` (Quit: If at least two walls aren't shored up with balks of timber then it's not a true witch's cottage at all, but merely the home of some daft old bat who reads tea leaves and talks to her cat. -- Terry Pratchett: Lords and Ladies)
23:28:19Dhraakellianhmm... I *thought* I had a battery bench for my Clip+ with FM
23:28:32Dhraakellianbut I appear to have only recorded the first 2 hours or so
23:28:41 Quit Jaykay (Ping timeout: 246 seconds)
23:28:48Dhraakellianthis the danger of multi-part tests
23:35:57 Quit domonoky (Read error: Connection reset by peer)
23:37:58 Join chrisb [0] (
23:38:04 Quit captainkewll (Quit: Page closed)

Previous day | Next day