Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2011-01-14

00:02:15 Quit evilnick_B (Quit: Page closed)
00:04:57 Join boxocide [0] (
00:05:11 Quit GeekShadow (Ping timeout: 260 seconds)
00:07:01 Quit liar (Ping timeout: 255 seconds)
00:07:58 Quit pamaury (Remote host closed the connection)
00:09:29 Join Pandaa [0] (
00:11:26 Join JdGordon1 [0] (
00:13:45 Quit boxocide (Quit: changing servers)
00:16:42 Quit efyx (Quit: Quitte)
00:26:36 Quit froggyman (Quit: Ex-Chat)
00:27:13 Quit mystica555_ (Read error: Connection reset by peer)
00:29:34***Saving seen data "./dancer.seen"
00:30:12 Quit petteyg (Quit: the mushroom quiche was served an eviction notice)
00:30:14 Join mystica555_ [0] (
00:31:46 Quit ender` (Quit: I invented the term 'Object-Oriented', and I can tell you I did not have C++ in mind. -- Alan Kay)
00:32:01 Join Mikeb0ok [0] (
00:32:19 Quit kevku (Quit: KVIrc 4.0.2 Insomnia
00:45:30 Join utanapischti [0] (
00:49:20 Quit sasquatch (Ping timeout: 264 seconds)
00:50:39 Quit kugel (Remote host closed the connection)
01:00:34sideralgaeverts, Torne, Saint: I just managed to reproduce the apparent FS corruption while logging open/close/fsync. Nothing suspicious was going on file-wise (except for font and currently played mp3 still being open). But the FAT appeared broken to fsck: Lots of circular cluster chains, wrong file sizes etc. After disconnecting and reconnecting, suddenly the FS was clean.
01:00:45sideralLooks like a FAT sector wasn't written out in time for the Linux host's read, but was later updated with the correct version.
01:01:39sideralBut I still can't seem to find any cache that I might have forgotten to flush.
01:02:34sideralI give up for today.
01:02:42JdGordon1never give up~!
01:02:46JdGordon1never surrendeer!
01:03:11 Nick sideral is now known as Picard (~sideral@unaffiliated/sideral)
01:03:23PicardWe will never surrender!
01:03:30 Nick Picard is now known as sideral (~sideral@unaffiliated/sideral)
01:03:34*JdGordon1 was tihnking churchill?
01:05:24sideralThe only history I know is that of the Federation
01:06:09sideralGood night!
01:06:26 Join krazykit [0] (
01:10:21 Join webguest93 [0] (
01:13:44 Quit webguest93 (Client Quit)
01:15:11 Join FBI_Guy [0] (
01:18:14 Quit komputes (Quit: I haven't slept for ten days, because that would be too long.)
01:20:49 Quit dfkt (Quit: -= SysReset 2.53=- Sic gorgiamus allos subjectatos nunc.)
01:24:25 Join webguest617 [0] (
01:24:30 Quit 20QAAE1S1 (Quit: Leaving.)
01:24:31 Quit 52AAACRYL (Quit: Leaving.)
01:25:23 Nick webguest617 is now known as froggyman (
01:25:45 Quit froggyman (Changing host)
01:25:45 Join froggyman [0] (~cdd586f9@unaffiliated/froggyman)
01:28:58 Join liar [0] (
01:30:13 Join JesusFreak316 [0] (
01:34:09 Quit sideral (Quit: Leaving.)
01:36:56 Quit froggyman (Quit: CGI:IRC (EOF))
01:38:49 Join froggyman [0] (~seth@unaffiliated/froggyman)
01:44:56 Quit jhMikeS ()
02:05:31 Quit mystica555_ (Ping timeout: 240 seconds)
02:06:29 Join mystica555_ [0] (
02:12:58 Quit leavittx (Ping timeout: 255 seconds)
02:15:20 Join newClipUser [0] (
02:26:10 Quit FBI_Guy ()
02:29:36***Saving seen data "./dancer.seen"
02:32:45 Join BNHabs [0] (
02:32:57BNHabsCouple of questions.
02:33:40BNHabsDoes Rockbox improve SQ of music automatically without fittiling around with EQ settings?
02:33:43 Join G4JC [0] (~Gaming4JC@unaffiliated/gaming4jc)
02:34:19G4JCWould anyone know of a bookreader for rockbox? :)
02:34:36G4JCsimple ebook reader, pdf, text, etc.
02:34:45G4JCpreferably searchable
02:35:01 Join milk [0] (
02:35:23G4JCprobably isn't impliminted but I'm not sure... definatly something for the wishlist. :)
02:36:36 Quit BNHabs (Client Quit)
02:37:39*G4JC googles it some more...
02:38:40 Quit factor (Read error: Connection reset by peer)
02:39:24tmztthere a text editor/viewer
02:40:20G4JCtmzt: Would you know if there's a way to search for a specific word in it?
02:40:34 Join factor [0] (~factor@
02:40:38tmztI don't
02:41:44 Quit Judas_PhD (Ping timeout: 240 seconds)
02:42:24G4JCand would anyone know if it's possible to run libc/gcc on the nano? :D
02:46:00 Join Keripo [0] (
02:49:26 Quit factor (Read error: Connection reset by peer)
02:49:46 Quit Keripo (Read error: Connection reset by peer)
02:49:47 Join factor [0] (~factor@
02:50:31 Join Keripo [0] (
02:53:29 Quit Topy44 (Read error: Connection reset by peer)
02:53:41 Join Topy44 [0] (
02:53:59 Quit factor (Read error: Connection reset by peer)
02:54:58 Join factor [0] (~factor@
02:55:50 Join Judas_PhD [0] (
03:02:18[Saint]JdGordon|: Am I correct in thinking there's a touch area definition for pictureflow hotkey?
03:02:28[Saint]If so, do you happen to know what it is?
03:03:48JdGordon1[Saint]: hotkey
03:04:06JdGordon1and make sure the wps hotkey is set to pf
03:04:26[Saint]ah, right...thanks.
03:04:43JdGordon1I wanted to add a bunch more hotkeys for touchscreen
03:04:47JdGordon1but havnt done it yet
03:04:59[Saint]I was thinking that tapping and holding the AA for pictureflow hotkey would be great.
03:15:05milki'm looking to try out Rockbox on my new Sandisk Clip+, but see that the 3.7.1 notes state "USB is still too unstable on the FuzeV2, ClipV2 and Clip+ and is disabled in this release". USB as in basic USB connectivity? would i have to dualboot into the original firmware to add new musics?
03:17:01milkta :)
03:19:37 Quit newClipUser ()
03:21:54milkis amsv2, like, the chipset or somesuch?
03:24:37 Quit Judas_PhD (Ping timeout: 246 seconds)
03:33:05 Quit G4JC (Quit: KVIrc 4.0.2 Insomnia
03:36:20tmztmilk: sort of, its more a generic way of saying the second generation of the ams players
03:37:47 Join Judas_PhD [0] (
03:39:50milk*dives through the wiki some more*
03:41:06milkis there still any collaboration with ams post the 2007 stuff?
03:42:12 Quit Judas_PhD (Ping timeout: 250 seconds)
03:42:13 Quit Keripo (Quit: Leaving.)
03:42:36 Quit froggyman (Ping timeout: 255 seconds)
03:42:41 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
03:45:14JdGordon1collaboration is a bit of a misleading term
03:49:50 Join yelped [0] (
03:51:49yelpedDoes anyone else get a "cannot open archive" error while trying to extract the new Debian image linked to on the VMware development page?
03:52:25 Join Keripo [0] (
03:53:55 Join Judas_PhD [0] (
04:03:07krazykityelped, you probably have a corrupted download
04:06:18yelpedOk, I'll try downloading again.
04:06:52yelpedHow does that happen?
04:10:06krazykitpackets get lost. internet happens.
04:10:52yelpedOk. Thanks.
04:20:23 Quit yelped (Quit: CGI:IRC (Ping timeout))
04:22:40 Join froggyman [0] (~seth@
04:22:40 Quit froggyman (Changing host)
04:22:40 Join froggyman [0] (~seth@unaffiliated/froggyman)
04:29:00 Join Barahir_ [0] (
04:29:38***Saving seen data "./dancer.seen"
04:32:05 Quit Barahir (Ping timeout: 246 seconds)
04:33:41JdGordon1[Saint]: did you see the guide in the forum about making a clock from bmps?
04:34:48[Saint]I think that guy might feel a bit burned now...
04:35:05[Saint]I mean, all points for trying, but it was a bit of a fail all round.
04:35:27JdGordon1well his way still has merits
04:35:39JdGordon1but I was wondering about adding tags to make it actually usable
04:35:41JdGordon1or simpler
04:36:16[Saint]What I think we need (If in fact it really *does* need it at all) is clearer documentation as opposed to these "style guides" that will grow stale and get buried.
04:36:23 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:36:23 Quit amiconn (Disconnected by services)
04:36:40 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:36:44[Saint]JdGordon1: Had you not thought about doing a clock like that before? A few themes already do that.
04:37:38 Quit pixelma (Disconnected by services)
04:37:40 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:37:43 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:39:11JdGordon1I added a tag ages ago (which was rejected) to draw any text from a bmp
04:39:17JdGordon1but no, i couldnt be bothered with clocks
04:49:52 Quit TheSeven (Ping timeout: 264 seconds)
05:04:39 Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier.
05:07:59 Quit JesusFreak316 (Ping timeout: 240 seconds)
05:10:31 Quit Rob2222 (Ping timeout: 250 seconds)
05:12:16 Quit Keripo (Quit: Leaving.)
05:13:24 Join froggyman_ [0] (~seth@
05:13:34 Quit froggyman_ (Read error: Connection reset by peer)
05:17:07 Quit froggyman (Read error: Connection reset by peer)
05:17:14 Join froggyman [0] (~seth@unaffiliated/froggyman)
05:25:45 Quit Horscht (Quit: Verlassend)
05:34:26 Quit Facefoxdotcom (Quit:
05:43:16 Join Keripo [0] (
05:50:13 Join Rob2222 [0] (
05:57:05[Saint]JdGordon|: Can I lay mute and unmute overtop of each other with touch areas and have it automagically either mute or unmute depending ow what's true/false?
05:57:50[Saint]or would O need to distinct touch areas?
05:58:58JdGordon1mute does both you dill :)
05:59:16JdGordon1first press mutes, 2nd press resumes previous volume
05:59:23JdGordon1[Saint]: ^
06:02:39[Saint]ah, there is no "unmute", just the one "mute".
06:06:07 Quit parafin (Ping timeout: 260 seconds)
06:06:40 Quit Judas_PhD (Ping timeout: 264 seconds)
06:09:01 Join parafin [0] (
06:20:14 Join Judas_PhD [0] (
06:21:28 Quit panni_ (Quit: ( :: NoNameScript 3.81 :: ))
06:25:06 Quit BHSPitMonkey (Quit: Ex-Chat)
06:29:40***Saving seen data "./dancer.seen"
06:39:06 Quit parafin (Ping timeout: 276 seconds)
06:42:41 Join parafin [0] (
07:19:49 Join sideral [0] (~sideral@unaffiliated/sideral)
07:22:12 Quit JdGordon1 (Quit: leaving)
07:32:38 Quit sideral (Quit: Leaving.)
07:34:52 Join einhirn [0] (
07:43:41 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
07:59:01 Join DerPapst [0] (
07:59:08 Join DerPapst1 [0] (
08:01:38 Quit Keripo (Quit: Leaving.)
08:05:54 Join esperegu [0] (~quassel@
08:09:23 Join sideral [0] (~sideral@unaffiliated/sideral)
08:09:24[Saint] <−−- 240x320 Touch-Screen CabbieV2 (TEST)
08:09:44[Saint]JdGordon| ; Others: ^
08:11:10 Join bertrik [0] (
08:11:10 Quit bertrik (Changing host)
08:11:10 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
08:14:48JdGordon|stop using 7z!
08:16:38[Saint]WHat's wrong with 7z?
08:17:18 Quit t0rc (Quit: Give someone code, help them with one project. Teach someone to code, help them rule the world.)
08:18:29 Join Bagder [0] (
08:19:40 Join Keripo [0] (
08:19:54 Quit Bagder (Changing host)
08:19:54 Join Bagder [0] (~danielx@rockbox/developer/bagder)
08:21:10JdGordon|[Saint]: no worky
08:21:12 Join kugel [0] (~kugel@rockbox/developer/kugel)
08:21:35[Saint]Are you using current SVN?
08:21:50[Saint]it has new tags in.
08:21:58[Saint]namely mute
08:22:57[Saint]gah! what is happening? is it falling back to rockbox_failsafe?
08:23:03kugel[Saint]: are you designing it after the other touchscreen cabbies?
08:23:16[Saint]kugel: vaguely
08:23:30kugelthat means?
08:23:38JdGordon|[Saint]: fallback wps, seems to be fallback sbs also with giant font and icons
08:24:08[Saint]fucking hell...the themeeditor displays it correctly.
08:24:50[Saint]Oh...hahaha...I think I know what happened.
08:24:57JdGordon|wrong screen size :p
08:25:00[Saint]there's no .rockbox dir
08:25:17[Saint]what's wrong screen size?
08:25:36[Saint]you need to set the screen size manually if you open it in the themeditor
08:25:41JdGordon|app defaults to 320x480
08:25:44JdGordon|your 240x320?
08:26:21[Saint]I think it's not working because if you just unzipped it to simdisk/whatever...there's no .rockbox dir
08:26:26*JdGordon| is *suuure* someone has done the work to get a universal app build....
08:26:28[Saint]if that's not it...shit knows.
08:26:54kugelthe app has no .rockbox
08:27:27kugelwell, it has a equivalent, but it's not named ".rockbox"
08:27:50[Saint]ah, right...the theme is probably lookinf for everything in the wrong place then.
08:27:59JdGordon|the path in the config should still be /.rockbox/<foo> though
08:28:08 Quit Judas_PhD (Ping timeout: 250 seconds)
08:28:13[Saint]ah, ok.
08:28:38 Join BRi7X [0] (
08:28:39[Saint]I can't think of why it's failing then...
08:28:49[Saint]the themeeditor shows it correctly :/
08:28:51kugel[Saint]: what do you mean with "vaguely"?
08:29:05[Saint]"It is vaguely based on cabbie"
08:29:08JdGordon|240 width?
08:29:37kugelcan you go a bit more into detail?
08:29:41***Saving seen data "./dancer.seen"
08:29:43 Join T44 [0] (
08:29:43 Quit Keripo (Read error: Connection reset by peer)
08:30:27BRi7Xhowdy.. i was a little confused by the installation for the Gigabeat S... will I be able to install the firmware on a brand new MK3008GAL 30GB hard disk?
08:30:56[Saint]How much more detail can I go into? I'm really not is based on cabbie, but the layout and most of the functionality is completely different.
08:31:27JdGordon|it isnt going into svn if the AA area goes to playlist!
08:31:51[Saint]feel free to make suggestions.
08:31:58JdGordon|that was sarcasm
08:32:20 Quit DerPapst (Quit: Leaving.)
08:32:21 Quit DerPapst1 (Quit: Leaving.)
08:32:48 Quit Topy44 (Ping timeout: 240 seconds)
08:32:50kugel"completely different" is something we don't want to hear when it comes to cabbie
08:33:05[Saint]there's one TODO still, and that's make the play state relevant stuff in the .sbs conditional.
08:33:05 Join Keripo [0] (
08:33:40[Saint]kugel: I said it'd be nice if it gets into svn...but at the end of the day this is pretty much a theme solely for me.
08:33:48[Saint]if anyone else likes it, great.
08:34:30JdGordon|I thought you hated cabbie? :)
08:34:55[Saint]I could easily change the code to match what SVN cabbie is doing...I just changed it to what *I* wanted to use.
08:35:21kugeloh I thought it's supposed to be a cabbiev2 port, but you named it differently
08:35:23[Saint]I also fixed up all the alignment, the current cabbie of this size is pretty out of shape.
08:36:12[Saint]I have the intention of porting pieces of this to the best cabbie port, solely for the fixed alignment of some things.
08:37:06[Saint]this is all something I knocked up in about 24 hours though, it's still far from finished.
08:37:49 Quit markun (Read error: Connection reset by peer)
08:38:00kugelalso doesn't work for me
08:38:02[Saint]heh, just noticed one bug already ;)
08:38:30[Saint]I have no idea what the fuck is going on then, it works in the themeeditor.
08:38:44kugelyou are free to build the sdl app
08:39:00kugelor does it not work on your system (cygwin, right?)
08:39:10[Saint]there's no point...there's no relevant debug output.
08:40:16JdGordon|sdl-app-checkwps doesnt build
08:41:30kugel−−debugwps doesn't work in the sdl app
08:41:51[Saint]gah! this is'll be something small...I know it will be.
08:42:04 Join Judas_PhD [0] (
08:42:18[Saint]sim debug output getting lost in transition kinda sucked.
08:42:45kugelyou could try loading it in the gigabeat f sim, IIUC it won't fail for the touchregions
08:42:58 Join markun [0] (
08:42:58 Quit markun (Changing host)
08:42:58 Join markun [0] (~markun@rockbox/developer/markun)
08:43:19[Saint]yes, I believe I will try this...but I still wwon't get any debug output if it does fail :/
08:43:38kugel−−debugwps should work in the sim
08:44:14[Saint]last time I tried it gives no output
08:44:28[Saint]it hasn't since the syntax change IIRC
08:46:11 Join wodz [0] (
08:49:12wodzI was thinking about how to add .fms support to the themes. The lecture of makes me wondering why we complicate the thing so much? Wouldn't it be easier, cleaner and so forth to treat themes installed by default just like all other themes? I mean create .zips with the cabbies and install only relevant one (instead of hacky generation of the theme structure by
08:53:55wodzJdGordon|: Nice work with peakmeater, really. Now for visual appealing effects we should have skinnable simple fft.
08:53:59[Saint]wodz: The idea was to move everything except cabbie out of the main binary
08:54:21wodzI think cabbie is not available on all targets
08:54:34[Saint]there are still a few themes left to be moved to the themesite...but the guy who knows which ones has gone awol
08:55:13 Quit wodz (Quit: Leaving)
08:55:17[Saint]the idea was to only ship cabbie, and move the others to the themesite...most of the shipped themes can already be downloaded from the themesite.
08:55:32[Saint]a lot of the shipped themes *could* just be removed.
08:58:28 Quit bertrik (Ping timeout: 276 seconds)
09:01:03sideralSaint, gevaerts: I have a new patch to try for the filesystem-corruption issue. I have opened FS #11877 to collect our experiences, and have posted the patch there
09:01:03 Quit bluebroth3r (Read error: Connection reset by peer)
09:01:21 Join bluebrother [0] (
09:01:21 Quit bluebrother (Changing host)
09:01:21 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
09:04:36 Join ender` [0] (
09:05:11 Quit ender` (Read error: Connection reset by peer)
09:05:41 Join ender` [0] (
09:08:43 Join Nikolasz [0] (
09:09:06 Join DerPapst [0] (
09:09:11 Join DerPapst1 [0] (
09:10:40 Join LinusN [0] (~linus@rockbox/developer/LinusN)
09:10:53 Quit Nikolasz (Client Quit)
09:11:49 Join Saij [0] (
09:18:39 Quit Judas_PhD (Ping timeout: 272 seconds)
09:23:48 Join ruckus [0] (
09:28:37 Join swilde [0] (
09:30:57 Join Judas_PhD [0] (
09:31:38 Join petur [0] (d408b802@rockbox/developer/petur)
09:33:48 Quit DerPapst (Quit: Leaving.)
09:33:49 Quit DerPapst1 (Quit: Leaving.)
09:33:50 Quit Judas_PhD (Client Quit)
09:35:08pixelmawodz (logs): which targets are missing a cabbie port (except the charcell Player and maybe new resolutions that come with new ports)?
09:35:40 Quit JdGordon| (Quit: Lost terminal)
09:37:42 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
09:44:24 Quit milk (Ping timeout: 240 seconds)
09:47:19 Join wodz|work [0] (
09:48:30wodz|workpixelma: you enumerated what I had in a mind
09:52:48pixelmanew resolution are often taken care off during the port (and fully switching to "default theme only" would require a cabbiev2 port or whatever theme it would be), for the Player you could almost take any "theme" and call it cabbiev2
09:53:58pixelmaAFAIK when cabbiev2 is missing currently the target will display the failsafe WPS anyway and I guess that would stay this way
09:55:02wodz|workyes, exactly. So ?
09:55:11 Quit rasher (Quit: leaving)
09:57:16pixelmathere's no reason that prevents switching to one included theme only, I think (you didn't say directly but your statement about cabbiev2 missing on some targets sounded this way to me)
09:58:17pixelmabtw. if I remember correctly was initially written by Cassandra who hasn't been around for quite some time
09:58:21 Part ruckus
09:58:42wodz|workI was wondering more if we could get rid of autogenrated cabbie (or other themes).
09:59:24wodz|workIt would be way simpler to have zips with cabbie as we have for other themes and simply unpack one with right resolution.
10:00:50pixelmahmm, not sure if I understand your suggestion completely
10:01:54wodz|workpixelma: currently we have .wps, .sbs, bunch of graphics and during make install we parse WPSLIST, .wps, .rwps to produce config and install theme.
10:03:53wodz|workwe should either: 1) refine to not be wps centric 2) get rid of this entirely and treat default theme as other themes
10:03:58 Join rasher [0] (
10:03:58 Quit rasher (Changing host)
10:03:58 Join rasher [0] (~rasher@rockbox/developer/rasher)
10:04:05 Quit rasher (Client Quit)
10:04:23 Join rasher [0] (
10:04:24 Quit rasher (Changing host)
10:04:24 Join rasher [0] (~rasher@rockbox/developer/rasher)
10:10:21pixelmawodz|work: I think we need to make sure that the default theme is always included though, not a seperate installation. If you want to get rid of the automatic packaging then maybe we could have premade zips in the code and a much simpler script that'll take care of chosing and unzipping the right zip (which is not only resolution dependent, there could also be different hardware versions RTC etc.)
10:10:45 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
10:11:14wodz|workpixelma: maybe my wording was not strict but this exactly what I propose
10:11:27pixelmaif treating it like all other themes, how would you make sure a sim contains the default theme, or when you don't install with the Utility?
10:12:00pixelmaok, then I understand your suggestion better now. Glad we worked that out :)
10:12:24wodz|workby treating like other themes I mean to have cabbie in .zip with standard structure and unpack the right one during make install
10:14:26pixelmaaha, well there was my confusion - other themes that live on the theme site can be chosen when installing with the Utility or you can put it on your player manually
10:18:25pixelmaone disadvantage of premade zips could be things that are used by more than one port getting "out of sync" when things changed. I don't think that affects bitmaps currently, I'm more thinking of things set in the .cfg like line selector colours
10:19:50 Quit Saij (Read error: Connection reset by peer)
10:23:01kugelwhat's wrong with generating cabbie? seems to be a mess (according to wodz|work)
10:26:07wodz|workkugel: the main problem is that wpsbuild is wps centric. It would be hard to cleanly extend it to handle .fms, or recscreen skins (it that happens)
10:26:52kugeli don't think so
10:26:59kugeladding .sbs wasn't that hard
10:28:52kugelit's not a mess, it's perl ;)
10:29:43***Saving seen data "./dancer.seen"
10:31:24wodz|workkugel: sbs is quite simple compared to .fms
10:32:12wodz|work.fms should be scanned if it uses additional graphic than .wps and moreover it should have .rfms companion
10:32:19gevaertssideral: one problem I see: how do you guarantee that the default_event_handler_ex handler runs last for system_flush() and first for system_restore()?
10:32:50AlexPMy .sbs is a lot more complicated than my .sbs
10:33:01AlexPAll my graphics pretty much are in .sbs for instance
10:33:46wodz|workAlexP: your sbs is not a problem - you don't use to install it
10:34:14AlexPI know, I'm just pointing out that fms is not necessarily more complicated than fms
10:34:20AlexP*than sbs
10:34:37AlexPWho is to say that cabbiev2 won't get reworked or replaced?
10:35:08AlexPAnd why can't there be a remote sbs for instance?
10:35:21kugelwodz|work: that's the same for sbs
10:35:34kugelit can also have graphics and an rsbs
10:36:00kugelsbs is just as complex as wps or fms
10:36:16pixelmaand wps/rwps which works in SVN
10:36:28wodz|workwe are mixing two things: 1) what can and what can't be in wps, sbs, fms 2) what is and what is not handled by
10:37:00kugelyes, but you were saying adding fms to was more complex than adding sbs was. that's not true
10:37:58kugelinfact, only the extension is different, the rest is completely the same from the point of view
10:39:33wodz|workkugel: wpsbuild scan only wps/rwps for graphics AFAIK
10:40:48kugelthat would be a bug
10:48:47 Quit esperegu (Remote host closed the connection)
10:50:01wodz|workkugel: that is the point when I say wpsbuild is wps centric (hacked a bit to work with sbs for cabbie)
10:50:51kugelI'm fairly sure I added the scans for the graphics because I had a (temporary) during development
10:55:50kugelyou could collect the graphics for wps,sbs,fms first, and then scan for the images. they end up in the same folder anyway
10:57:37wodz|worklooking at the code once more I am fairly sure only wps/rwps are parsed for needed graphics
10:58:30kugelperhaps it broke in the meantime since we never had an sbs with graphics
10:59:38 Join ruckus [0] (
10:59:57kugelanyway, if it's wps centric then it's a bug (which should be fixed) and not by design
11:00:13ruckushey folks. i am trying to format my ipod's (video) hdd, but it stops at 8%. anyone have any ideas what could I do? i am on win7
11:02:23sideralgevaerts: Good question. Right now I cannot guarantee this. I have to admit that I do not fully understand how the system is quiesced and restored around USB mode.
11:03:15sideralgevaerts: For example, who calls default_event_handler{,_ex} and how is it ensured that it is called only and exactly once?
11:03:33gevaertsI'm not sure that's ensured
11:04:09 Quit kugel (Ping timeout: 240 seconds)
11:04:21gevaertsActually, it is
11:04:30gevaertsI think
11:04:41sideralAh good. Because if not, then system_flush/restore might be called multiple times
11:04:59 Join PurlingNayuki [0] (~PurlingNa@
11:05:08gevaertsdefault_event_handler() is called by the GUI thread (in various places) when the code waits for an event to deal with "generic" events
11:05:09sideralOr not? Can event handlers consume events and prevent further propagation?
11:06:00gevaertsWell, events don't actually propagate. if an event is sent to more than one thread, it's sent multiple times
11:07:36gevaertsIn this case, basically usb.c sends SYS_USB_CONNECTED to all threads that want events, which are then all supposed to send back SYS_USB_CONNECTED_ACK. If all ACKs have been received, usb.c proceeds to the next stage
11:07:52sideralOK, so because the default handler is only ever called by the GUI thread, it runs only once?
11:07:58gevaertsThe order in which those SYS_USB_CONNECTED events are handled by various threads is not defined
11:11:34sideralRe the corruption problem: I had one theory where it is too late to flush data to the disk once all threads have been quiesced, leading to updates not making out to disk before USB slave mode is entered. Hence my idea to do the flushing in the GUI thread. But if I understand you correctly, the default handler runs when the SYS_USB_CONNECTED event has already been sent, so again some or all threads may already be sleeping
11:12:40gevaertsWell, a thread that's already sleeping should really have closed all its filehandles
11:13:51sideralAs I've seen last night, this isn't actually the case in at least two cases (font and currently played audio file are still open)
11:15:59 Quit GeekShadow (Read error: Connection reset by peer)
11:16:10 Join GeekShad0w [0] (
11:17:07sideralI still don't understand how slave mode can export disk that looks broken to the host when first connected, but consistent when connected a second time. It's as if the USB disk slave starts serving out stuff too early, or some other weird caching problem...
11:20:50sideralgevaerts: Does one of the file/fat/disk/storage layers require another thread to write out data, or do writes happen asynchronously in any other way? So far I've found no indication that this may be the case
11:21:48 Join n1s [0] (
11:21:48 Quit n1s (Changing host)
11:21:48 Join n1s [0] (~n1s@rockbox/developer/n1s)
11:28:49 Quit TheSeven (Disconnected by services)
11:28:50 Join [7] [0] (~TheSeven@rockbox/developer/TheSeven)
11:32:36 Part PurlingNayuki
11:51:29 Join kugel [0] (~kugel@rockbox/developer/kugel)
11:53:37gevaertssideral: I'm not actually an expert in this area :)
11:56:23sideralNeither am I −− yet ;) . I appreciate you chiming in anyway. Is anyone an expert in this area?
12:09:00 Join pamaury [0] (
12:09:00 Quit pamaury (Changing host)
12:09:00 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
12:09:07[7]hm, I've seen the ATA driver defer I/O to a thread, but it should use proper mutex locking for this
12:09:36[7]if this wouldn't be the case, you wouldn't only see corruption with UMS, but also with rockbox disk accesses themselves
12:13:03sideralTheSeven: Can you point me to a place in the code? I don't doubt that everything is synchronized correctly. But if one important thread is in USB freeze before we flush stuff to the disk, this would explain why the disk update doesn't make it out until after we've returned from USB mode
12:13:50 Join thomasjfox [0] (
12:15:17[7]this is the actual disk access. i doubt that anyone can read anything via UMS while a write access is still in progress
12:15:27[7]the file i'm talking about is firmware/drivers/ata.c
12:16:25[7]which target are you testing on currently?
12:16:58sideralClipV2 −− which has SD storage, not ATA
12:18:23sideralas far as I understand the code, the SD is written to synchronously, but I could be wrong
12:25:02gevaertssideral: it can't really be ATA or SD buffering. USB also does storage transfers (e.g. to read the partition table, or your fsck), and that would seriously clobber any buffer there, so rockbox wouldn't fix the filesystem after disconnect if things were wrong in that area
12:25:11gevaertsIt *has* to be higher level
12:29:47***Saving seen data "./dancer.seen"
12:32:16gevaertskugel: what's your plan with the AA patch?
12:32:40kugelhave a closer look as soon as I have some time
12:32:58kugelI think someone needs to check if it works on the mrobe500
12:33:16[7]AA meaning AAfonts?
12:38:21[7]i'm strongly in favor of that feature :)
12:38:59[7]hmm... what should we do with the ipod classic target in the long term?
12:39:40[7]i still have to figure out why dualboot is failing, but how would we want to handle rbutil integration?
12:40:02[7]apparently on some ipod classics the firmware partition isn't accessible using usb any more
12:42:40user890104[7]: is it possible to use the same commands that iTunes sends when updating the firmware, to flash an AUPD with an installer, and then let it run on the device?
12:43:09[7]possibly, but i have no idea how to do that from the driver side of things
12:43:22[7]itunes seems to do it using custom scsi commands
12:43:49[7]a different approach would be using DFU, but that requires pressing buttons on the ipod and needs a driver as well
12:44:10*[7] wonders how various iphone tools manage to hijack itunes' dfu driver
12:44:24wodz|work[7]: what's the problem with custom scsi commands? We can do that with help of libusb
12:44:54[7]hm, installing a libusb driver on windows for a mass storage device won't be any better than installing it for a dfu device
12:45:39gevaertsYou don't need libusb for custom SCSI commands I think
12:46:05gevaertsrbutil does or did something with them to find out ipod video RAM size
12:46:25[7]hm, that might be worth looking at
12:47:37[7]we might be able go for a hdd-only based installation like on the older ipods...
12:48:56*[7] also wonders if anyone wants to join the ipod nano 3g efforts, benedikt93 and asmmonkey seem to have picked up that target...
12:49:53 Quit kugel (Ping timeout: 276 seconds)
12:58:54wodz|work[7]: what's wrong with hdd-only installation?
13:05:53 Quit thomasjfox (Remote host closed the connection)
13:12:36 Join thomasjfox [0] (
13:15:27 Join kugel [0] (~kugel@rockbox/developer/kugel)
13:25:45 Join BaDeD [0] (
13:31:20BaDeDHow to install Spanish language voice clip +?
13:31:50TorneGenerate one with the rockbox utility, as we told you yesterday
13:32:36Torneif you're having a problem with that then you'll have to explain what you did and what didn't work
13:33:55StummiBaDeD, and, please, stop bugging me per pm. When I can help you, i will answer here
13:34:38TorneIt's better that you figure out how to generate your own, because it needs to be regenerated whenever you update Rockbox
13:34:47Torneso if you get one from someone else now, you will have to get another one every time you update
13:34:52Torneand we are not here to do this for you :)
13:37:07[7]wodz|work: that it doesn't work yet :)
13:38:36 Quit BaDeD (Quit: CGI:IRC (Ping timeout))
13:39:21 Join leavittx [0] (~lev@
13:39:59 Join BaDeD [0] (
13:40:18BaDeDi´m sorry Stummi
13:42:34amiconnn1s: I seem to be unable to build the new m68k toolchain on cygwin.
13:43:27amiconnIt worked fine on linux (both my VM and my hardware box), but oon cygwin the gcc build fails in libgcc with a rather weird error:
13:43:45amiconn/tmp/rbdev-build/build-gcc/./gcc/as: fork: Resource temporarily unavailable
13:45:45 Quit BaDeD (Quit: CGI:IRC (Ping timeout))
13:45:58wodz|workn1s: I build new toolchain yesterday with the same prefix as old one and m68k-elf-gcc is not overwriten. I have now m68k-elf-gcc-3.4.6 and m68k-elf-gcc-4.5.2 *AND* m68k-elf-gcc but this gives 3.4.6 as its version
13:46:59 Quit [7] (Ping timeout: 260 seconds)
13:48:53 Quit kugel (Ping timeout: 276 seconds)
13:52:39n1swodz|work: hmm i was certain it would overwrite it
13:54:36n1swodz|work: but you can just manually copy or symlink the one you want to use then
13:58:15n1samiconn: hmm, it did build fine for me on cygwin once i installed g++
13:58:54 Quit kkit|sh (Ping timeout: 264 seconds)
13:59:38 Quit Keripo (Quit: Leaving.)
14:00:06amiconnHmm. What error did you get before that?
14:00:27amiconng++ is installed btw
14:00:46n1ssomething about /lib/cpp failing sanity check
14:00:48amiconnHmm, but gcc and g++ versions don't match
14:00:56amiconnI wonder whether this may be important
14:01:37n1sthe error you got gets some hits on google but i haven't found anything interesting about it
14:02:05n1sare you building with -j ?
14:03:47n1s says it's an issue with a "Logitech process monitor"
14:04:41n1sor maybe thta's just related to that specific ssh error...
14:05:54 Quit liar (Read error: Connection reset by peer)
14:06:59 Join liar [0] (
14:07:56n1sthey have made a list of apps that they blame for cygwin troubles here
14:10:21n1sso it looks like anti virus scanners etc is a major cause for that
14:12:28 Join ZhangNing [0] (~ZhangNing@
14:16:38 Quit GeekShad0w (Read error: Connection reset by peer)
14:16:59 Join GeekShad0w [0] (
14:19:59 Join Kitar|st [0] (
14:21:07amiconnI've disabled the virus scanner during build
14:22:01gevaertsn1s: we probably need a mail to -dev about the new compilers too
14:22:16n1sgevaerts: ah, yes
14:23:38n1shmm, what should that say then, if people just rerun they seem to still have the old compiler as m68k-elf-gcc (wodz di this)
14:24:54Bagderwell, I did that too and I got the new one as defaut
14:25:07 Join dfkt [0] (dfkt@unaffiliated/dfkt)
14:25:17n1sBagder: that is what i expected to happen
14:29:48***Saving seen data "./dancer.seen"
14:29:55wodz|workBadger: did you install in default dir (/usr/local)? I passed RBDEV_PREFIX (IIRC) to install in my home
14:30:40BagderI used the default prefix, yes
14:31:00wodz|workhmm maybe that makes problem
14:31:22gevaertswodz|work: so you didn't overwrite it but did a separate install?
14:31:51gevaertsIf so, this will depend on your exact PATH
14:32:00wodz|workgevaerts: my old m68k-elf-gcc was installed with the same RBDEV_PREFIX
14:33:32n1swodz|work: so you have m68k-elf-gcc-3.4.6 and m68k-elf-gcc-4.5.2 and m68k-elf-gcc (reporting version 3.4.6) in the same dir?
14:33:51wodz|workthis is something I have to double check
14:34:55 Quit FOAD (Ping timeout: 265 seconds)
14:35:00 Join FOAD [0] (~dok@
14:43:46 Quit GeekShad0w (Quit: The cake is a lie !)
14:43:53 Quit factor (Read error: Connection reset by peer)
14:51:42 Quit slooopy (Ping timeout: 255 seconds)
14:53:06ZhangNingI try to build Rockbox,but it says: Normal build selected
14:53:08ZhangNing../tools/configure: 1071: Syntax error: "esac" unexpected
14:55:01gevaertsZhangNing: what's /bin/sh on your system?
14:55:25*gevaerts wonders if tools/configure doesn't assume bash
14:55:40Bagderdoes it? we have lots of ubuntu people etc
14:56:16ZhangNing /bin/sh?
14:56:59n1sZhangNing: ls -la /bin/sh
14:57:28gevaertsZhangNing: also, which revision of the source is that. Line 1071 doesn't have esac here
14:57:45ZhangNinglrwxrwxrwx 1 root root 4 2010-11-30 14:51 /bin/sh -> dash
14:58:44n1sdash works fine here so it's probably something else
14:59:59ZhangNingI get the source from a friend,she gives me the website:
15:00:02ZhangNinggit clone git://
15:00:20Bagderthen you should ask your friend
15:00:26Bagderour code is not hosted there
15:00:27n1sZhangNing: that's some fork we don't support
15:01:00ZhangNingI'm sorry……
15:01:14ZhangNingBut I want to know
15:01:19 Join factor [0] (~factor@
15:01:21BagderZhangNing: we recommend you getting and using the main branch
15:01:44BagderZhangNing: your problem is that your configure script has one or more bugs, making it not dash compliant
15:02:06ZhangNingif you have a source for Onda only
15:02:34Bagderwe have source for onda _also_
15:02:42ZhangNingI think the main branch is too big
15:02:51Bagderthen you're on your own
15:03:06ZhangNingHow to download?
15:03:07Bagderbut I really doubt that
15:03:20n1sZhangNing: see the UsingSVN page in our wiki
15:04:22 Join slooopy [0] (
15:05:54 Join madalu [0] (~user@unaffiliated/madalu)
15:08:10 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
15:10:34 Join TheLemonMan [0] (
15:12:19 Join komputes [0] (~komputes@ubuntu/member/komputes)
15:13:21 Join kkit|sh [0] (
15:16:37 Quit wodz|work (Quit: CGI:IRC)
15:23:45 Join casainho [0] (
15:25:23 Quit Dreamxtreme (Quit: Hi, I'm a quit message virus. Please replace your old line with this line and help me take over the world of IRC.)
15:31:41 Quit mystica555_ (Ping timeout: 255 seconds)
15:32:50 Quit ZhangNing (Remote host closed the connection)
15:35:37 Quit Mikeb0ok (Read error: Connection reset by peer)
15:37:12 Quit krazykit (Quit: :|)
15:39:24 Join bmbl [0] (
15:39:24 Quit bmbl (Changing host)
15:39:24 Join bmbl [0] (~bmbl@unaffiliated/bmbl)
15:42:16 Join robin0800 [0] (
15:44:25 Quit robin0800 (Remote host closed the connection)
15:44:51 Join robin0800 [0] (
15:45:39 Join mystica555 [0] (
15:45:40 Quit mystica555 (Excess Flood)
15:45:46 Join mystica555_ [0] (
15:46:09 Join mystica555 [0] (
15:46:52 Join Dreamxtreme [0] (~Dre@
15:51:07 Quit madalu (Remote host closed the connection)
15:52:37 Quit evilnick (Quit: Leaving)
15:53:45 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
15:57:41 Nick robin0800 is now known as robin0800_ (
16:00:56 Quit TheSeven (Ping timeout: 246 seconds)
16:03:35 Quit GodEater (Ping timeout: 260 seconds)
16:11:09 Quit slooopy (Remote host closed the connection)
16:12:39 Join slooopy [0] (
16:12:40 Join GodEater [0] (
16:12:40 Quit GodEater (Changing host)
16:12:40 Join GodEater [0] (~bibble@rockbox/staff/GodEater)
16:12:40 Part LinusN
16:16:45 Join TheSeven|Mobile [0] (~theseven@rockbox/developer/TheSeven)
16:19:55 Quit sideral (Ping timeout: 260 seconds)
16:21:34 Join sideral [0] (~sideral@unaffiliated/sideral)
16:24:03 Quit sideral (Remote host closed the connection)
16:24:39 Join sideral [0] (~sideral@unaffiliated/sideral)
16:26:26 Quit TheSeven|Mobile (Ping timeout: 240 seconds)
16:27:40 Join TheSeven|Mobile [0] (~theseven@rockbox/developer/TheSeven)
16:29:49***Saving seen data "./dancer.seen"
16:31:29soapok, Buschel, thomasjfox, I'm back in town for testing. Sorry for disappearing for the week.
16:31:30BRi7Xhowdy, so I just got a brand new 30GB hard drive for a Gigabeat S i purchased a while ago.. as per usual, when plugged in through USB (and simultaneously AC), the hard disk does not show up, but another device does.. a '1GB' device with an mp3-player icon (win7)... beastpatcher failed so I decided to use the original firmware from the rockbox instruction page, it appears to be stuck on
16:31:31BRi7Xtransferring firmware and the program looks like it's completely locked up... i can't even move the window holding it, which is strange for a hung application on win7... I also have the dreaded Error 5 message on the gigabeat's screen. any suggestions?
16:33:04BRi7X(let me know if that message got clipped in any way, it was kind of a paragraph..)
16:35:08 Join sideral1 [0] (~sideral@unaffiliated/sideral)
16:38:18 Quit sideral (Ping timeout: 255 seconds)
16:39:53 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
16:44:29pixelmajust letting you know that the message got through but unfortunately I don't know much about the Gigabeat S and can't help you. My only advice would be to stay around for a while, maybe there'll be someone who can help you later
16:46:19BRi7Xalrighty, i'll sit tight... at least until i have to go to class
16:51:24 Quit petur (Quit: Page closed)
16:53:09BRi7Xeh it doesn't even feel like the hard drive's spinning up
16:53:44gevaertsMaybe it's not connected properly?
16:54:57BRi7Xcontact side face down into the hard drive (the hard drive having toshiba markings and information face up.), and contacts facing toward me on the zif connector.. is that the way?
16:55:31*gevaerts never disassembled a gigabeat S
16:55:58thomasjfoxsoap: Nice. Hope it was a vacation ;)
16:56:06 Join LinusN [0] (~linus@rockbox/developer/LinusN)
16:56:30thomasjfoxsoap: Rockbox for maemo is now in extras-devel
16:56:37 Join panni_ [0] (
16:56:40BRi7Xprobably a good thing, the gigabeat S is an unruly beast.. (is that why it's called beastpatcher?)
16:57:46thomasjfoxsoap: One user reported a very strange crash which I am unable to reproduce: If he sets the scrollbar size to f.e. 26 on his n8xx, rockbox crashes when he enters the "show track information" screen upon scrolling
16:58:02gevaertsThat all started with a typo, with someone typing gigabeast instead of gigabeats, but yes
16:58:12gevaertsHere it's commonly called the beast
16:58:17 Part LinusN
16:59:41BRi7Xnow i can see why support for it was a bit hesitant. if only A. my gigabeat F had a working backlight and/or B. the gigabeat S supported the IDE interface. eek
16:59:49 Quit Bagder (Remote host closed the connection)
17:01:26 Quit ruckus (Quit: Leaving.)
17:02:16 Join kugel [0] (~kugel@rockbox/developer/kugel)
17:02:43thomasjfoxhey kugel. I will give the ucontext branch a test run right now
17:03:22kugelthomasjfox: hey! I've had a quick glance at your patch, and I have a few remarks
17:03:59kugelI don't know what to do about the "n900 german keyboard hack"; amiconn and gevaerts had a quick talk about it at the DevConPub but I can't remember the details
17:04:50kugelthen I feel a bit uncomfortable with the changes to system-sdl.c as it now has a lot of maemo specific stuff, and not only sdl specific stuff
17:05:41thomasjfoxhmm. I wanted to avoid code duplication, so I put the maemo stuff in system-sdl.c
17:05:48thomasjfoxOther ideas?
17:05:56kugeland I think maemo-thread.c could be split up a bit, it seems to mix different functionality (lcd suspend, power/battery management, telephony))
17:06:13kugelI've not looked deeply enough to make suggestions yet
17:07:32kugela question that I had was do you even need simcc() (in tools/configure)? apart from finding sdl-config everything you add there seems to be maemo specific and could be a seperate function
17:08:38kugeland did I understand that n8xx does *not* use gstreamer due to some bug? I assume it uses sdl audio instead?
17:09:47thomasjfoxwe could have a standalone maemocc(). Then we would f.e. need to duplicate the gcc option filtering like androidcc does
17:10:13thomasjfoxn8xx doesn't have the "appsrc" gstreamer module which we use to inject the pcm data
17:10:23thomasjfoxIt uses SDL audio
17:10:58kugelis there any audio api which would work on both?
17:11:33thomasjfoxTalking to pulseaudio directly I guess
17:11:41thomasjfoxLet me have another look how the libplayback stuff works exactly
17:11:54thomasjfoxIIRC it uses the pulseaudio "stream name".
17:12:16thomasjfoxlibplayback is the part to stop playback on incoming calls or play audio while the phone is in silent mode
17:13:15 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
17:13:20 Quit TheSeven (Read error: Connection reset by peer)
17:13:29 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
17:13:57 Quit kugel (Remote host closed the connection)
17:14:02 Join kugel [0] (
17:14:03 Quit kugel (Changing host)
17:14:03 Join kugel [0] (~kugel@rockbox/developer/kugel)
17:14:03 Quit TheSeven|Mobile (Quit: TheSeven|Mobile)
17:15:14 Quit sideral1 (Ping timeout: 260 seconds)
17:15:57 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
17:16:18thomasjfoxkugel: Switching to pulseaudio directly would work
17:16:52thomasjfoxThat would save us from SDL's slow software mixing
17:18:35kugelcool. seems to have some advantages then (i suspect it'd take a tad bit less CPU if we avoid the gstreamer layer)
17:19:13thomasjfoxcould be. On the other hand if have to learn yet another API ;)
17:19:19thomasjfoxif = I
17:21:20thomasjfoxkugel: btw: What do you think about FS #11859?
17:21:22kugelperhaps it also allows us to avoid that feed_data workaround? I don't know pulseaudio at all
17:22:33pamauryI once used the simple api of pulseaudio and it was quite easy to use
17:22:39thomasjfoxkugel: Yes, the feed_data stuff is a specific workaround for gstreamer.
17:23:26thomasjfoxpamaury: I can't imagine it being harder than: init(), set_volume(), push_pcm_data()
17:23:27kugelI find it strange to use other optimization levels at all
17:24:20pamaurythomasjfox: if you use the complete api, I guarantee you that your imagination will break apart
17:25:21thomasjfoxpamaury: Ups, the simple api doesn't have volume control
17:25:34thomasjfoxkugel: Sorry, I don't get you. What do you mean by that?
17:26:15kugelall builds should use the same level imo
17:26:36pamauryThere must be a way to set the volume using the simple api
17:26:52kugel(unless there's a good reason like a measurable performance difference)
17:27:42thomasjfoxsimcc() currently kills all optimizations by filtering -O
17:27:59thomasjfoxThis might be good for debugging
17:28:29thomasjfoxpamaury: No volume control:
17:29:18thomasjfoxpamaury: Though the async API is not a problem, we already have a glib main loop for maemo and there's a pulseaudio loop <-> glib loop wrapper
17:29:21pamaurythe simple api is simple in the sense that it's synchronous. It doesn't mean you don't have access to the volume
17:30:04thomasjfoxThe page explicitly mentions no volume control...
17:30:28 Join stsquad [0] (
17:30:48pamauryah, then you must be right
17:30:57pamauryit's really a trivial and useless api :)
17:37:23kugelsynchronous isn't necessarily bad
17:38:14pamauryno but if you can't change any setting, you are quite limited
17:38:36 Join wodz [0] (
17:39:38wodzn1s: m68k-elf-gcc-3.4.6 and m68k-elf-gcc-4.5.2 are in different dirs
17:39:51thomasjfoxkugel: The sigaltstack of the ucontext branch seems to run fine
17:40:13thomasjfoxkugel: ASSEMBLER_THREADS still die with a stack overflow on database initialization
17:40:50thomasjfoxkugel: Where is the stack size define located again?
17:40:57 Quit TheSeven (Read error: Connection reset by peer)
17:41:02 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
17:42:19 Join stoffel [0] (
17:42:33thomasjfoxkugel: Found it (by looking at your diffs)
17:43:36 Quit robin0800_ (Quit: Konversation terminated!)
17:43:47 Join robin0800_ [0] (
17:45:32n1swodz: ah, so they infact used different prefixes, which m68k-elf-gcc that will be used is determined by the PATH
17:46:08n1sso you have rockbox-dev/m68k-elf/bin before rockbox-dev/bin in your PATH
17:46:11wodzn1s: yes but the prefix passed to was the same
17:46:53n1saha, maybe the script has changed
17:47:09gevaertsIt has, a while ago.
17:47:23Torneyou'll notice that your m68k-elf has an m68k-elf inside it
17:47:26Tornethat would be the old one :)
17:47:36gevaertswodz: I suspect your "old" setup may have been before those changes
17:47:40Tornewe used to install to $(ARCH)/{bin,lib,$(ARCH),etc}
17:47:46n1sanyways thanks for checking as this means just running the sript again will work
17:47:50Tornewe now install to {bin,lib,$(ARCH),etc}
17:47:56n1sah yes
17:47:56Tornethis can easily be confusing :)
17:50:03thomasjfoxgevaerts: Do you remember the details about the n900 german keyboard hack discussion?
17:51:06gevaertsthomasjfox: "discussion" is a bit inaccurate. amiconn noticed that I have dedicated arrow keys and he didn't. That was basically it
17:51:20thomasjfoxah ok
17:52:09thomasjfoxThe german layout for the cursor keys just sucks. Also SDL doesn't map it properly and the pre-installed tetris game is not playable on the german N900, probably because of the same bug.
17:52:49Tornethis isn't rockbox-specific.. i would suggest people with sucky N900 keyboard layouts change their keyboard layout
17:52:54Tornewith xmodmap/etc
17:53:23 Quit kugel (Ping timeout: 240 seconds)
17:53:23thomasjfoxTorne: The idea was to make rockbox work without the user tweaking keybord layouts
17:54:16thomasjfoxTorne: Therefore the "n900 german keyboard hack" in the maemo patch :)
17:54:23Torneyes, i understand
17:54:31Tornei just don't see why we should care
17:55:08Torneif we're going to be an app then it's up to the OS to tell us what keys are being pressed, no?
17:55:19Torneif it's wrong blame the OS
17:55:53thomasjfoxYes. It looks like the mapping is not done properly by SDL
17:56:12Torneright, but we're not isntalling SDL, are we?
17:56:16Torneit's a preexisting package
17:57:08thomasjfoxYes and there will be no further updates from Nokia I guess. So we have to workaround the issue with a small if() ;)
17:58:16Tornenokia don't have to update it.
17:59:32Tornepreinstalled packages are not magic, someone can just put a new one in extras
17:59:50 Join kugel [0] (
17:59:51 Quit kugel (Changing host)
17:59:51 Join kugel [0] (~kugel@rockbox/developer/kugel)
17:59:53 Quit kugel (Remote host closed the connection)
18:00:06 Join kugel [0] (
18:00:07 Quit kugel (Changing host)
18:00:07 Join kugel [0] (~kugel@rockbox/developer/kugel)
18:01:32thomasjfoxTorne: True. The less modified packages needed, the better
18:04:06thomasjfoxkugel: ASSEMBLER_THREADS seems to die regardless of the stack size, I now have it at 0x100000 and it still dies after scanning ~27.000 files
18:05:32 Quit TheLemonMan (Quit: free(me))
18:05:58 Join TheLemonMan [0] (
18:11:58thomasjfoxkugel: I'm not really into the ARM thread code, but what abour r12 on ARMv7? It only saves up to r11 as far as I understand it
18:14:03Torner12 is ip, it's not required to be preserved by function calls
18:14:55Torne(it's for the benefit of linker-generated shims that might need a temporary register: not preserved by calls but also not used for args/returnvals)
18:16:04thomasjfoxah thanks
18:16:39Torneso you only need to save r4-r11,r13,r14 if you're saving state from a function :)
18:17:26thomasjfoxSo the crash must be somewhere else :)
18:17:37TheSevenTorne: r13?
18:17:50thomasjfoxr13 is the stack IIRC
18:18:23 Quit kugel (Ping timeout: 240 seconds)
18:18:44Torneyah, you need to be able to restore whatever r13 was on entry to the context saving function
18:19:15 Join Mustermax [0] (
18:19:48MustermaxAnyone at home?
18:20:05thomasjfoxSo how do we tell the compiler not to use r12 as a general purpose scratch register?
18:20:51TheSevenwhy should one do that?
18:21:39thomasjfoxJust something I read here:
18:21:53 Join kugel [0] (
18:21:53 Quit kugel (Changing host)
18:21:53 Join kugel [0] (~kugel@rockbox/developer/kugel)
18:21:58 Quit Mustermax (Client Quit)
18:22:03thomasjfoxThough I must admit I'm not really into the details
18:22:07TheSevenTorne: oh, right, cooperative multitasking context switch... i usually think from a preemptive multitasking point of view
18:23:55TheSeventhomasjfox: why do you need the compiler to leave this register alone?
18:25:28thomasjfoxTheSeven: My first thought was that it's not saved on context switches, so the compiler might store temp values in there. Though this is just a wild guess
18:26:13TheSevenno, it can't, because code that calls a function cannot safely assume that r12 survives this
18:26:23kugelfor cooperative threads you only need to save non-scratch regs
18:26:36TheSevenboth the compiler and linker will use it as a scratch register in the called function
18:26:47TheSevenso the calling function needs to treat it like r0-r3
18:26:54amiconnkugel: On arm r0..r3 and r12 are scratch
18:26:55thomasjfoxah ok, now I got it: because yield() is a function :)
18:27:09kugelamiconn: i know
18:27:32kugelthomasjfox: exactly
18:27:37TheSeventhomasjfox: for cooperative multitasking, yes. for preemptive multitasking, you need to save absolutely everything.
18:29:35thomasjfoxkugel: Hmm. I modified it to save r12 for fun and now it doesn't crash. I'll check if I didn't mess up something else
18:29:52***Saving seen data "./dancer.seen"
18:30:35kugelisn't r12 the frame pointer in some configuration?
18:30:43 Quit swilde (Remote host closed the connection)
18:31:03TheSevenkugel fp is usually r11 IIRC
18:32:04thomasjfoxStack size is down to 0x1000 and it still doesn't crash.
18:32:08kugelI've also problems with initializing the db on android, but that worked a while ago. perhaps the gcc upgrade?
18:32:35thomasjfoxI'll revert to the previous code which only saves up to r11
18:32:38kugelbut I have no idea why r12 would make a difference
18:34:08kugeldoes it possibly have something to do with neon?
18:35:07amiconnNeon is accessed as a coprocessor afaiu
18:36:17thomasjfoxI guess I found the issue: I need a "make reconf" after tweaking the stack size. r12 seems unrelated, I'll verify it once more
18:36:32thomasjfoxSorry 'bout that :)
18:37:14 Join Horscht [0] (~Horscht@xbmc/user/horscht)
18:38:37kugelthat shouldn't be needed
18:38:55kugelyou change DEFAULT_STACK_SIZE in thread.h don't you?
18:39:45 Quit robin0800_ (Remote host closed the connection)
18:41:26thomasjfoxkugel: I've now reverted all my changes expect "thread_support=ASSEMBLER_THREADS" and will try again
18:41:45 Join LambdaCalculus37 [0] (~3f74f70d@rockbox/staff/LambdaCalculus37)
18:41:47thomasjfoxkugel: I don't like problems that "disappear"...
18:42:08kugelthomasjfox: ah you need make dep if you switch between threading, that's perhaps the reason
18:43:02kugeldependencies are generated in a seperate step, so changes that affect dependencies (#include etc) might need make dep to be catched
18:43:13thomasjfoxI'll check
18:43:23 Join DX3 [0] (~Dre@
18:43:45kugelalthough it shouldn't affect thread.h
18:44:15 Quit Dreamxtreme (Ping timeout: 272 seconds)
18:45:30 Quit amee2k (Remote host closed the connection)
18:45:51 Join Buschel [0] (
18:48:47 Quit DX3 (Quit: Never look down on someone unless you're helping them up.)
18:48:51thomasjfoxkugel: If I switch it to HAVE_SDL_THREADS, kernel-sdl.c complains about a missing do_exit() function
18:49:05thomasjfoxkugel: Oh wait
18:49:20thomasjfoxkugel: That might be an issue of the maemo code
18:49:46thomasjfoxkugel: Remember the code we put in there to prevent the shutdown hang
18:52:05 Quit TheSeven (Ping timeout: 260 seconds)
18:53:02 Join {phoenix} [0] (
18:53:19kugeli remember that
18:54:09thomasjfoxtick_timer() also checks the do_exit -variable- which is only defined if HAVE_SDL_THREADS is not defined
18:54:44thomasjfoxFirst guess: That part of the code is currently not compatible with upstream code
18:54:49kugelthat'd be a bug in my code I guess
18:57:38 Quit kugel (Remote host closed the connection)
18:58:10 Join kugel [0] (~kugel@
18:58:11 Quit kugel (Changing host)
18:58:11 Join kugel [0] (~kugel@rockbox/developer/kugel)
18:58:15 Join yelped [0] (
18:59:04 Join Dreamxtreme [0] (~Dre@
18:59:55yelpedI have some warnings in settings-list.c when I try to compile, but my main problem is that the rockbox zip is only about 1.1 MB and it's missing many of the things that make it run.
19:01:03 Nick T44 is now known as Topy44 (
19:02:38yelpedDoes anyone know what the issue may be?
19:09:20LambdaCalculus37jhMikeS: (for the logs) Are you going to do any major changes to the SA9200 bootloader soon? I want to roll a release candidate bootloader for it soon.
19:09:50 Join Keripo [0] (
19:09:51LambdaCalculus37We need art for the SA9200 as well.
19:14:31thomasjfoxkugel: It started crashing again with ASSEMBLER_THREADS. This time I built a clean .deb package from scratch
19:14:44thomasjfoxkugel: I try to save r12 again...
19:15:31yelpedI have some warnings in settings-list.c when I try to compile, but my main problem is that the rockbox zip is only about 1.1 MB and it's missing many of the things that make it run.Does anyone know what the issue may be?
19:16:07thomasjfoxyelped: What target did you compile for?
19:16:21yelpedSansa Fuze v1
19:16:59bluebrothern1s: bisected the crash, turned out Buschel's hint was right −− r28608 is the culprit.
19:17:02bluebrotherif I revert that change it works fine again (currently r29039, since I don't have m68k-elf 4.5.2 on this machine)
19:17:12 Quit Buschel (Ping timeout: 265 seconds)
19:17:21thomasjfoxyelped: Using which toolchain?
19:18:24yelpedUsing the new Debian Image posted on the VMware development page, with Vmware player.
19:19:47thomasjfoxyelped: Ok. And you just checked out the source code from SVN or are you using f.e. 3.7.1?
19:20:30yelpedI did
19:21:30thomasjfoxOk. Maybe someone with the VMware image could try to reproduce it, I don't have it
19:22:05yelpedI selected Normal build. That's good right?
19:22:09 Quit kugel (Ping timeout: 240 seconds)
19:22:48 Quit Keripo (Quit: Leaving.)
19:23:01yelpedSo, no suggestions?
19:23:28thomasjfoxSo you created a build_sansa or whatever directory, entered it and ran ../tools/configure
19:23:46thomasjfoxthen 'make' and after that 'make zip'?
19:25:25n1sbluebrother: ah, i don't use database so that's probably why it works for me then, but nice to know it wasn't the compiler change that caused it
19:27:01thomasjfoxSounds right to me
19:27:14bluebrothern1s: not sure if it only occurs when you use the database. When upgrading I first replaced the complete .rockbox folder so the database wasn't present at all
19:27:22bluebrotherplus, I don't use the database as well :)
19:27:39bluebrotherthough it's possible I had the database built on the h120 someday in the past
19:27:40thomasjfoxyelped: Maybe capture the complete build output into a file and put it on
19:27:45n1sbluebrother: well, that's odd then
19:28:04yelpedYou mean
19:28:08n1syelped: does configure print any warnings?
19:28:37yelpedYes, settings-list.c
19:28:46bluebrothern1s: but I can retest that :)
19:28:54n1syelped: _configure_
19:29:22 Join aalonso [0] (~aalonso@
19:29:44 Quit LambdaCalculus37 (Quit: back to work)
19:30:02 Quit stoffel (Remote host closed the connection)
19:30:37n1syelped: the script ypou select the build in is called configure, does it print any warnings?
19:32:18bluebrothern1s: ok, installed 3.7.1 again. No database, haven't experienced the bug yet
19:32:19 Join kugel [0] (~kugel@rockbox/developer/kugel)
19:32:33bluebrotherrebuilding the database now, let's see what happens when that's done
19:33:06aalonsoHi, How ipod partition tables are generated? mbr-*.bin
19:33:11 Quit slooopy (Remote host closed the connection)
19:33:46n1saalonso: they were dumped from ipods
19:34:32 Join Horschti [0] (~Horscht@xbmc/user/horscht)
19:34:46aalonsousing dd: dd if=/dev/sdx of=mbr-*.bin ??
19:35:29n1syes, and count=1 i think
19:35:58 Join slooopy [0] (
19:36:08aalonson1s: thanks
19:37:48dionoeahello. Is a rockbox control toolbar widget being worked on for the android application? (if not I'll probably give it a try)
19:37:59 Quit Horscht (Ping timeout: 240 seconds)
19:40:10 Join bertrik [0] (
19:40:10 Quit bertrik (Changing host)
19:40:10 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
19:44:47yelpedHere's the build output:
19:48:03thomasjfoxyelped: The .zip file doesn't contain any binary files
19:48:42thomasjfoxyelped: Please post a log from the "configure" to the "build zip" steps
19:48:51 Join solexx_ [0] (
19:49:01thomasjfoxkugel: I'll run a sigaltstack build the next days on my n900 to test the stability.
19:49:17thomasjfoxkugel: So far CPU usage is 10% down compared to normal SDL threads
19:49:43thomasjfoxkugel: (similar to ASSEMBLER_THREADS, expect sigaltstack doesn't crash ;))
19:51:48 Quit solexx (Ping timeout: 255 seconds)
19:52:06thomasjfoxDo it inside a terminal emulator that can save the scroll buffer to disc. KDE's konsole can do this, dunno if it's part of the virtual machine (I doubt it).
19:52:22thomasjfox"screen" should be able to do it too
19:52:33thomasjfoxAnyway I've got to run. Good luck!
19:53:07 Quit thomasjfox (Remote host closed the connection)
19:58:30 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
19:59:29 Quit evilnick_B (Quit: Page closed)
20:00:37kugeldionoea: give it a try :)
20:01:43 Quit casainho (Remote host closed the connection)
20:01:54dionoeathat almost sounds like someone already tried it and found it too hard :)
20:03:11 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
20:14:54yelpedHereis the Build logs :
20:16:05 Join stoffel [0] (
20:18:23 Join mudd1 [0] (
20:18:34n1syelped: have you changed the source at all?
20:20:31yelpedYes. I deleted all the #Ifdefine CONFIG TUNER from root_menu.c and settings_list.c
20:20:42mudd1I asked this already a couple of days ago but I had to go to sleep before anyone could answer so let me try again:
20:20:57mudd1Why did I empty my "Recent Bookmarks" list by calling bo
20:20:58n1syelped: well, that is probably why it broke
20:21:04 Quit factor (Read error: Connection reset by peer)
20:21:21mudd1or to put it differently: how *do* I just bookmark the current state to the recent bookmarks?
20:21:38yelpedSo what can I do to remove the radio without breaking anything?
20:21:47n1sthe first error is that code in settings_list.c is trying to use the address of a function that doesn't exist for targets without radio
20:21:53mudd1it is very well possible that I made a stupid mistake because I misunderstood something about the bookmarking concept in Rockbox
20:22:00yelpedI only need to remove it from the radio screen.
20:22:13yelpedthe *menu screen
20:22:49n1syelped: well, why not just delete the menu entry then? or set CONFIG_TUNER to 0
20:24:21yelpedWhere, meaning which one? Also how do I put everything back in? Could I sync it with current SVN without downloading everything again?
20:24:39n1ssvn revert -R .
20:25:58n1syou either want to delete the actual menu entry or just disable all radio stuff by setting CONFIG_TUNER to 0 in your target's config
20:27:14yelpedIt says not enough arguments provided.
20:27:32 Join DerPapst [0] (
20:27:36Lloreanyelped: Did you include the period?
20:27:39gevaertsYou probably thought n1s wasn't serious about the . :)
20:27:39 Join DerPapst1 [0] (
20:28:05 Join factor [0] (~factor@
20:28:15 Quit bmbl (Quit: Verlassend)
20:29:23yelped"." Is not a working copy. ".svn/entries' : No such file or directory
20:29:56***Saving seen data "./dancer.seen"
20:30:19*gevaerts bets that yelped is trying this from a build directory
20:30:25n1syou need to be in the dir you checked out the source in for that command to work, or substitute . with the path to it
20:31:20yelpedOk, thanks it's working.
20:33:53yelpedNow where do I put the zero? On the same line as Config Tuner? Near all of them?
20:36:01yelpedSorry for asking so much questions? :)
20:36:44 Quit factor (Read error: Connection reset by peer)
20:37:00 Join factor [0] (~factor@
20:43:06 Join krazykit [0] (
20:47:14yelpedNow where do I put the zero? On the same line as Config Tuner? Near all of them?
20:51:14Lloreanyelped: Set its value to 0 in your target's config, like n1s said.
20:54:33yelpedI'm sorry, but I don't understand. Where is my targets config?
20:56:19 Join kevku [0] (~kevku@2001:7d0:0:f9af:215:c5ff:fe7e:794d)
21:00:00Lloreanyelped: It's the config file for whichever player you're building rockbox for.
21:00:11 Part Llorean
21:02:08pixelmayelped: in your checkout under firmware/export/config/yourplayer.h
21:04:05yelpedHow do I Set its value to 0?
21:13:06 Join sideral [0] (~sideral@unaffiliated/sideral)
21:17:35pixelmaopen yourplayer.h in an editor and in the "CONFIG_TUNER BLAH" line, replace BLAH with a 0, save, compile again. I think you said you had a Fuze, be careful to chose the correct file (there should be two, one for the v1 and the other for the v2)
21:17:54 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
21:20:10yelpedThanks. With a space or without a space?
21:26:47 Join user890104_ [0] (~Venci@2001:0:5ef5:79fd:28e4:2366:2b16:10ec)
21:27:40 Quit user890104 (Ping timeout: 272 seconds)
21:29:29 Quit yelped (Quit: CGI:IRC (Ping timeout))
21:34:46 Join baptiste [0] (
21:37:51 Quit stoffel (Remote host closed the connection)
21:43:38 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
21:47:28 Join FBI_Guy [0] (
21:51:32 Join Buschel [0] (
21:56:47 Join KiwiCam [0] (
22:14:23 Quit DerPapst (Quit: Leaving.)
22:14:24 Quit DerPapst1 (Quit: Leaving.)
22:21:00dionoeakugel: do you know if audio playback is supposed to work with rockbox in the android emaulator?
22:21:47kugelit worked some time ago, but I didn't try since then since the emulator is generally too slow to play music
22:21:52kugelon my machine at least
22:22:44dionoeadoesn't seem to work here. The progress bar doesn't even move. I guess that I'll just test the basics on the emulator and do the rest on the device.
22:22:51dionoeareal device that is
22:24:14 Join DerPapst [0] (
22:24:17 Join DerPapst1 [0] (
22:25:36 Quit slooopy (Ping timeout: 246 seconds)
22:28:43 Join Keripo [0] (
22:30:00***Saving seen data "./dancer.seen"
22:41:12CIA-7New commit by alle (r29052): pitch_detector: add the key transposition feature which makes the plugin better usable with/for transposing instruments (FS #11752 by Robert Horn)
22:46:07CIA-7r29052 build result: All green
22:48:01 Join user890104 [0] (~Venci@2001:0:4137:9e76:3044:23eb:2b16:10ec)
22:50:00 Quit user890104_ (Ping timeout: 272 seconds)
22:52:02 Quit benedikt93 (Quit: Bye ;))
22:55:01*amiconn tries another time to build m68k-elf-gcc on cygwin
22:55:23 Quit evilnick_B (Quit: Page closed)
23:00:49 Quit baptiste (Remote host closed the connection)
23:02:13 Quit TheSeven (Ping timeout: 276 seconds)
23:03:56n1samiconn: from that link i posted this moring it seems like things hooking into the OS when starting processes caused it and that those things usually are virus scanners etc, and apparently disabling them is not always enough...
23:05:40amiconnI've doubled the system and per-desktop heap sizes in the registry.
23:06:16amiconnIt's strange - I can build all other toolchains - sh, arm-eabi, mips, and old m68k. Just m68k 4.5.2 fails
23:06:57amiconnUnfortunately each try takes >1 hour before the error occurs
23:07:36n1si just read this thing there's a heading called "Random fork() failures."
23:08:12amiconnIn this case it's not random
23:08:27n1samiconn: always the same place?
23:09:12amiconnWell, if you accept two times as "always"
23:16:50 Quit user890104 ()
23:16:56 Join user890104 [0] (
23:18:08 Quit [Saint] (Disconnected by services)
23:18:10 Join S_a_i_n_t [0] (S_a_i_n_t@
23:20:01 Quit Keripo (Quit: Leaving.)
23:21:57 Join Keripo [0] (
23:22:27 Join dfkt_ [0] (dfkt@unaffiliated/dfkt)
23:22:40 Quit dfkt (Read error: Connection reset by peer)
23:23:36 Nick dfkt_ is now known as dfkt (dfkt@unaffiliated/dfkt)
23:24:07 Quit FBI_Guy ()
23:35:50 Quit Buschel (Ping timeout: 240 seconds)
23:35:51 Quit Keripo (Read error: Connection reset by peer)
23:37:34 Join Keripo [0] (
23:38:49 Quit TheLemonMan (Quit: free(me))
23:39:06 Quit komputes (Remote host closed the connection)
23:39:37 Join saratoga [0] (9803c22e@gateway/web/freenode/ip.
23:40:17saratogaanyone tested mpegplayer on the gigabeatS?
23:41:12 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
23:41:46 Join Horscht [0] (~Horscht@xbmc/user/horscht)
23:43:01sideralSaint, gevaerts: Re FS corruption issue: I just made a scary discovery
23:43:11bluebrothersaratoga: tried it a couple of days ago
23:43:51 Quit Keripo (Quit: Leaving.)
23:43:55gevaertssideral: discoveries are always good!
23:44:10sideralI diffed a "good" fat (dd'd from the OF) against a bad one (dd'd after fsck said the FAT was bad)
23:44:58 Quit Horschti (Ping timeout: 276 seconds)
23:45:00sideralturns out that in the bad one many 4K-aligned bytes are replaced with 0x7d
23:45:12sideralthere are many 0x7d's
23:45:20sideralbut all are 4k aligned
23:45:40sideralso it might be a cache-flush issue after all
23:46:14sideralmight this also explain the audio glitching on the ClipV2?
23:46:32saratogabluebrother: what was the FPS like?
23:46:43gevaertsBut it still gets "fixed" if you disconnect and reboot to the OF?
23:46:51gevaertsOr if you disconnect and reconnect?
23:47:01sideralyes, unless a modified FAT gets written out again
23:47:21bluebrothersaratoga: 24.9something
23:47:32gevaertsWhich presumably happens if you give rockbox time to start doing things again
23:47:46sideralif there's some write activity (DB update, for example)
23:47:50bluebrotherthough the fps display was 0.0 for the first second or two
23:48:02 Quit Llorean (Quit: Leaving.)
23:48:14saratogabluebrother: are you sure, the wiki gives 48fps for the Gigabeat F
23:48:20saratogai would expect close to 100 from the S
23:48:25sideralI'll try disabling the aligned-data zero-copy optimization in the SD driver
23:48:45gevaertsFeel free, but I *really* doubt if that helps
23:48:53sideralthe 4k alignment is suspicious
23:49:08sideralcould this also be a bad flash?
23:49:12saratogathe his problem on the clipv2 seems to be analog (doesn't show up in test_codec output) so its probably a problem with the DAC or VRM initialization
23:49:16sideralno, as the OF gets it right
23:49:20bluebrothersaratoga: that's with limit fps enabled :)
23:49:21gevaertsI seriously doubt that
23:49:28saratogaah well yeah :)
23:49:57gevaertsThe very fact that if you disconnect and then shutdown the filesystem is clean seems to indicate that this isn't a hardware or low-level driver issue
23:49:59sideralsaratoga: good point about test_codec
23:50:07bluebrotherit's around 97 without (but then the sound is kinda out of sync :)
23:50:21saratogai'm mostly just curious if 640*480 would work
23:50:44saratogaso if we used the hardware rescaler we might be able to play DVD rips on the beast
23:51:02bluebrotherdo you have a test video around for that?
23:51:08sideralgevaerts: any other theories on what might be causing this?
23:51:53bluebrotherI was thinking about a 640x480 :)
23:51:58gevaertssideral: not right now, but then I'm tired and I have a headache, so...
23:52:01saratogaah sorry no
23:52:06saratogai'm not even sure if it'll work
23:52:16saratogayou might overflow some buffers
23:52:21*gevaerts packs his c250 to see if he can reproduce it one of the following days
23:52:22amiconnIf you want realistic values for a dvd rip, the test video should use a higher bitrate
23:52:31*bluebrother will try elephantsdream once its downloaded
23:52:35amiconnI'm also not sure whether mpegplayer supports ac3 audio
23:52:54sideralSaint, what platform is your often-corrupted player based on?
23:52:55saratogagoogle found this:
23:53:01saratogaah fuck its h.264
23:53:17saratogaits mp2/3 only right now, but that could be fixed
23:53:34amiconnIf we find out how to use hw video acceleration, it could probably do yuv->rgb as well
23:53:35saratogaalthough maybe theres an mp2 dvd out there somewhere, IIRC its possible in theory
23:53:51saratogaah that would be really nice
23:54:21 Quit wodz (Quit: Leaving)
23:55:15amiconnmpeg2 audio (not sure about the allowed layers) is allowed on dvds, but afaik it's rarely used.
23:55:46amiconnAc3 is mandatory, and dts, pcm and mpeg2 are optional
23:58:32 Quit kevku (Quit: KVIrc 4.0.2 Insomnia

Previous day | Next day