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 2010-09-11

00:05:05 Quit Llorean (Read error: Connection reset by peer)
00:07:16 Join clone4crw [0] (
00:08:53alexbobPso if people made rockbox for free, and it's awesome to use, how come companies that sell products for money can't manage to make firmwares that are even okay to use?
00:09:05 Join mjbahniu [0] (
00:09:50 Quit domonoky1 (Read error: Connection reset by peer)
00:12:20mjbahniuI was trying to download the 3.6 build of rockbox for a Sansa ClipV1 and the link gives a 404 error, anyone aware of this?
00:14:00gevaerts"aware"? There is no 3.6 for the clip
00:14:07gevaertshm, wait
00:14:17gevaertsThere shouldn't be
00:14:29mjbahniuwhat is the latest build then?
00:14:57gevaerts3.6 is the latest *release*
00:15:10saratogawhere did you find that link?
00:15:22gevaertssaratoga: it's on the release page for some reason
00:15:32mjbahniuk, the installer links to that 3.6 link i gave and wont work
00:16:08saratogaoh odd
00:16:25saratogayou want to edit the HTML? i don't have svn handy
00:16:48mjbahniu thats where that link is aswell
00:17:03gevaertssaratoga: that's generated
00:18:54gevaertshm, we seem to have a problem there. The system doesn't handle stable non-released targets right
00:19:24 Quit MItaly (Ping timeout: 276 seconds)
00:20:05 Quit DerPapst (Ping timeout: 252 seconds)
00:21:23gevaertsmjbahniu: try installing a current build instead. That will work
00:23:04mjbahniugevaerts: by current you mean the build on this page ?
00:23:26gevaertsmjbahniu: either that or in the installer
00:23:56mjbahniuwell the installer links to that 3.6 link that 404's
00:24:18gevaertsNot if you tell it to install a current build
00:24:35mjbahniuk, will do. thanks for the help
00:27:45 Quit mjbahniu (Quit: CGI:IRC (EOF))
00:28:01 Quit Gatz85 (Quit: Gatz85)
00:28:24*gevaerts adds FS #11621
00:30:42 Join fdinel [0] (
00:51:12 Join CaptainKwel [0] (
00:58:23 Quit bluebrother (Disconnected by services)
00:58:26 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
00:59:43 Quit ender` (Quit: Where there's a will, there's an inheritance tax.)
00:59:59***Saving seen data "./dancer.seen"
01:06:36 Quit bertrik (Ping timeout: 255 seconds)
01:06:39 Join evilnick [0] (~Evilnick@rockbox/staff/evilnick)
01:09:58 Join s1gma_ [0] (~d.d.derp@
01:12:14 Quit Jerom (Quit: Leaving.)
01:12:39 Quit pamaury (Remote host closed the connection)
01:13:51 Quit _s1gma (Ping timeout: 276 seconds)
01:15:50 Quit petur (Quit: here today, gone tomorrow)
01:16:35 Join liar [0] (
01:19:41 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
02:02:05 Quit z35 (Quit: Leaving)
02:09:20 Quit fyrestorm (Quit: Ur skills' fireproof like a wooden panel -- U got feds talking leet on your IRC channel!)
02:10:09*shuffle2 pokes gevaerts
02:10:48shuffle2how feasible is it to make a "plugin" which uses the pp502x usb controller in device mode?
02:11:08shuffle2afaict all the usb stuff in the kernel is based around host usage?
02:11:21shuffle2and doesn't really offer a large api
02:11:27gevaertsuhm, no. Rockbox doesn't do host mode at all
02:11:53shuffle2oh hm yes, that was stupid to say
02:12:50 Join fyrestorm [0] (
02:14:50gevaertsAnyway, we're never going to have an api that provides much more functionality than we actually *need*
02:17:15 Quit efyx (Remote host closed the connection)
02:22:51 Join Judas_PhD [0] (
02:49:02 Join MethoS- [0] (~clemens@
02:56:17 Quit evilnick (Quit: TTFN)
03:00:02***Saving seen data "./dancer.seen"
03:45:32 Join Horscht [0] (~Horscht@xbmc/user/horscht)
03:59:06 Quit S_a_i_n_t (Ping timeout: 276 seconds)
04:02:10 Join S_a_i_n_t [0] (S_a_i_n_t@
04:08:25 Quit S_a_i_n_t (Ping timeout: 264 seconds)
04:09:11 Quit amiconn (Disconnected by services)
04:09:11 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:09:11 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:09:11 Quit pixelma (Disconnected by services)
04:09:24 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:09:27 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:12:40 Join S_a_i_n_t [0] (S_a_i_n_t@
04:27:22 Join Barahir_ [0] (
04:30:09 Quit TheSeven (Ping timeout: 265 seconds)
04:30:58 Quit fyrestorm (Quit: lamers envy me like they envy bill g -- main boot xp, just the way it should be!)
04:31:18 Quit Barahir (Ping timeout: 272 seconds)
04:31:52 Join Chronon [0] (
04:33:17 Quit bieber (Ping timeout: 240 seconds)
04:33:33 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:33:35 Join bieber [0] (
04:33:56 Quit s1gma_ (Quit: Leaving)
04:45:44 Join z35 [0] (
05:00:06***Saving seen data "./dancer.seen"
05:03:45 Quit MethoS- (Read error: Connection reset by peer)
05:19:30 Quit kadoban (Read error: Connection reset by peer)
05:20:06 Join kadoban [0] (
05:34:14 Quit ps-auxw (Ping timeout: 258 seconds)
05:38:19 Quit bieber (Ping timeout: 255 seconds)
05:38:33 Join bieber [0] (
05:45:29 Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier.
05:45:45 Join ps-auxw [0] (
06:01:02 Quit clone4crw (Ping timeout: 276 seconds)
07:00:07***Saving seen data "./dancer.seen"
07:13:28 Quit antil33t (Read error: Connection reset by peer)
07:13:34 Join antil33t [0] (
07:25:46 Quit Judas_PhD (Quit: This is a quitting message)
07:40:01 Quit Horscht (Ping timeout: 252 seconds)
07:40:46 Join Horscht [0] (~Horscht@xbmc/user/horscht)
07:45:55 Quit milz (Ping timeout: 240 seconds)
07:49:50 Quit Chronon (Ping timeout: 265 seconds)
07:58:55 Quit panni_ (Read error: Connection reset by peer)
08:02:35 Join stoffel [0] (
08:09:09 Quit solexx_ (Ping timeout: 240 seconds)
08:11:04 Join solexx [0] (
08:16:41 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
08:22:24 Join Keripo [0] (
08:22:45 Part Keripo
08:23:07 Join Keripo [0] (
08:29:07 Part Keripo
08:32:32 Quit krazykit (Quit: awe yeeeeeee)
08:34:12 Join togetic [0] (~togetic@unaffiliated/ibuffy)
08:34:27 Quit anewuser ()
08:35:44 Join n1s [0] (~n1s@rockbox/developer/n1s)
08:37:23 Join Highlander [0] (
08:39:36 Quit togetic (Quit: WeeChat 0.3.0)
08:40:07 Join togetic [0] (~togetic@unaffiliated/ibuffy)
08:44:44docgnomeis there a way to clear the database completely? i have, for some reason, multiple entries for some files
08:51:42n1syeah, delete the *.tcd files in .rockbox just performing "Initialize Now" should reset it completely though
08:55:05S_a_i_n_tquite often duplicate entries point to a corrupted .tcd file, which may or may not be due to filesystem errors.
08:55:16S_a_i_n_tIt might pay to do a sick-check also.
08:55:28S_a_i_n_thehehe....sick check.
08:55:35S_a_i_n_t*disk* check, rather
08:59:08 Quit n1s (Quit: Lämnar)
09:00:10***Saving seen data "./dancer.seen"
09:05:15S_a_i_n_tgevaerts: Do you need testing on FS #11619 - Restore pitch and speed settings on resume?
09:05:26S_a_i_n_tThis seems like something I could realluy make use of.
09:12:56 Join Rob2222 [0] (
09:13:55 Join Chronon [0] (
09:16:35 Quit Rob2223 (Ping timeout: 255 seconds)
09:21:40 Join kramer3d [0] (~kramer@unaffiliated/kramer3d)
09:27:04 Join Judas_PhD [0] (
09:36:26 Quit Chronon (Ping timeout: 272 seconds)
09:56:20 Quit yosafbridge (Ping timeout: 240 seconds)
09:59:03 Quit sasquatch (Quit: WeeChat 0.3.2)
09:59:26 Join sasquatch [0] (
10:01:30 Join Jaykay [0] (
10:03:07 Join yosafbridge [0] (
10:11:29 Quit bieber (Ping timeout: 255 seconds)
10:11:46 Join bieber [0] (
10:14:25 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
10:15:30 Quit ascheel (Ping timeout: 240 seconds)
10:29:52 Join funman [0] (~fun@rockbox/developer/funman)
10:30:42 Join kramer3d_ [0] (~kramer@unaffiliated/kramer3d)
10:34:15 Quit kramer3d (Ping timeout: 276 seconds)
10:41:18 Join Borogrove [0] (
10:48:35 Join Dreamxtreme_ [0] (Dreamxtrem@
10:52:15 Quit Dreamxtreme (Ping timeout: 276 seconds)
11:00:13***Saving seen data "./dancer.seen"
11:00:34 Join JdGordon [0] (7bf38c1f@gateway/web/freenode/ip.
11:02:21 Join bimbel [0] (~Miranda@unaffiliated/bmbl)
11:06:13 Quit bmbl (Ping timeout: 258 seconds)
11:10:27 Join ender` [0] (
11:23:00 Quit yosafbridge (Ping timeout: 276 seconds)
11:26:28 Quit funman (Quit: leaving)
11:31:17 Join yosafbridge [0] (
12:22:21 Join Dogg006 [0] (
12:29:27 Join Jerom [0] (~heidi@
12:35:08 Quit kramer3d_ (Ping timeout: 240 seconds)
12:38:35 Join [sko] [0] (~sko]
12:48:39 Quit bertrik (Read error: Connection timed out)
12:49:15 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
12:55:33S_a_i_n_tJdGordon: I've been meaning to ask you for a while now what FS #11614 is, the summary just says "skin manager", so I'm a little in the dark there ;)
13:00:16***Saving seen data "./dancer.seen"
13:01:02 Quit ender` (Quit: Do not believe any statistic you didn't falsify yourself.)
13:01:08JdGordonits the usual "us3ers housldnt see any differencre"
13:02:00gevaertsPresumably there *is* a (non-user-visible, possibly) difference though
13:09:13 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
13:09:45 Quit bertrik (Read error: Connection timed out)
13:10:11 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
13:11:39gevaertsS_a_i_n_t: I wouldn't mind testing on FS #11619 of course, but my main concern right now is getting a consensus on what the exact behaviour should be. I personally think that what the patch does is a step in the right direction, but not everyone will agree, and possibly a slightly larger overhaul might be needed
13:13:06gevaertsPersonally I don't really care about it, which is why I only compiled it and didn't test :)
13:13:09S_a_i_n_tI think (if I have it correct) that the behaviour of the patch seems perfectly fair. == timestretch options are consistent when resumed? correct?
13:13:44S_a_i_n_tIf I start a *new* playback session...are timestretch options held then?
13:13:52gevaertsYes, I think...
13:13:57gevaertswait, when?
13:14:12S_a_i_n_tAh...right, that's something I think I *may* find annoying.
13:14:37gevaertsIf you reboot the player, and your start screen is *not* the WPS, and you don't resume, and you start a new playlist, timestretch options will be 0
13:15:15 Quit Judas_PhD (Quit: This is a quitting message)
13:15:17gevaertsHowever, if you resume, remember that this isn't what you wanted to listen to, and then start a new playlist, I *think* they'll be kept
13:15:49S_a_i_n_tgevaerts: Like if I'm playing a file with timestretch options set and shutdown then resume playback...timestretch should still be set. But, if I resume, stop playback, then start a *new* playback timestratch still set?
13:16:38S_a_i_n_toh, I think your last post answers that...Hmmm.
13:16:44gevaertsI think it is. I didn't change behaviour for that case. Is it currently reset if you just stop and start a new playlist without a shutdown?
13:17:09gevaertsI also didn't actually check what the current behaviour is in that case, I just assume it's not reset...
13:18:03JdGordonS_a_i_n_t: well, it is a massive rework of how skins are stored/loaded... not a very interesting change for anyone really...
13:19:08 Quit bieber (Ping timeout: 240 seconds)
13:19:22 Join bieber [0] (
13:21:28 Join Dreamxtreme [0] (Dreamxtrem@
13:22:51S_a_i_n_tI *think* timestretch is reset when playback is stopped...but, I don't have a device on hand to check that right now and I'm not 100% about it.
13:24:35 Quit Dreamxtreme_ (Ping timeout: 272 seconds)
13:26:10gevaertsIf it is, it's not
13:26:15*gevaerts tested
13:26:56gevaertswait, I still had the start of a sentence there...
13:27:00gevaertsIt's not reset
13:28:27*gevaerts thinks it should be
13:29:51S_a_i_n_tI do also...but I take it that's the (or one of the) "but not everyone will agree"s
13:31:16gevaertsI think we have several options:
13:31:47gevaerts(1) do nothing, and let resume mean something else for (a) resume after shutdown, (b) resume after "normal" stop, (c) resume from bookmarks
13:32:12gevaerts(2) do what FS #11619 does right now, and have resume after shutdown mean the same as resume from bookmarks
13:33:40gevaerts(3) FS #11619 + reset timestretch on new playback, which makes the three kinds of resume mean exactly the same, and also makes starting a new playlist mean the same after shutdown and normal stop (and in e.g. car mode, the user may not actually know if the player just booted)
13:34:12gevaerts(4) save timestretch in the settings, i.e. *never* reset it
13:34:36gevaertsI think (3) and (4) are the only ones that provides consistency and predictability, and I really don't like (4)
13:35:00gevaertsOh, there's also (5), have a setting to decide whether or not timestretch values are a setting...
13:35:15JdGordonmaybe a rethink of how rockbox deals with resumed playlists would be better... sort of what I tried (probably 2 years ago) with bookmarks... timestretch/spped whatever should be linked to the playlist, so write them straight into the .playlist_control file
13:36:24S_a_i_n_tgevaerts: I think (personally) that "(3) FS #11619 + reset timestretch on new playback" is the serious contender there.
13:39:33gevaertsJdGordon: I briefly looked at the playlist control file, and if I understood it correctly it really stores adds and removes from the playlist (in order, which allows it to keep shuffle order the same, and which allows it to not rewrite the file all the time). I couldn't figure out how to add the timestretch values in there cleanly, which is why I eventually picked nvram
13:39:51gevaertsJdGordon: I think the resulting behaviour would be the same anyway
13:40:19gevaerts(except in case of a crash, in which case nvram wouldn't be saved, but .playlist_control would)
13:40:38gevaertsS_a_i_n_t: you can probably tell that that's also what I think :)
13:41:17JdGordonindeed, but if bookmarks were changed to load from .playlist_Control (or that file type) then all these non-persistant-but-playlist-relevant settings would still work
13:41:26S_a_i_n_tgevaerts: Actually, I had no idea ;)
13:41:35S_a_i_n_tBut I'm glad we're on the same page. :P
13:41:46bertrikwe have basically three places to store settings, right: playlist, nvram.bin, config.cfg ?
13:44:04JdGordonplayist is only filenames and shuffle seed
13:44:31JdGordonnvram.bin is only i think 10 settings which arnt really settings.. (dircache size, resume index)
13:50:40JdGordonFS #9407....
13:51:33 Join Jaykay_ [0] (
13:51:34 Join _s1gma [0] (~d.d.derp@
13:52:15JdGordonif that patch ever got finished we would be able to bookmark any type of playlist which would solve a few annoying bugs
13:52:34JdGordonand yeah, I think that is the better place for these settings
13:54:08bertriksounds sensible to me, but I'm no bookmark/playlist expert
13:54:15 Quit Jaykay (Ping timeout: 276 seconds)
13:54:29 Nick Jaykay_ is now known as Jaykay (
13:55:31JdGordonnoone is :(
13:56:20*gevaerts tries to convince JdGordon that skins are boring
13:56:41 Quit antil33t (Read error: Connection reset by peer)
13:56:42gevaertsPlaylists and playback are *much* more fun to work on!
13:57:09*JdGordon isnt touching anything that possibly might get close to tagcahce.c :)
13:57:17JdGordonand playback/buffering
13:57:37bertrikat least tagcache seems a relatively independent module
13:57:53 Join ste_iriveruser [0] (5d2ed442@gateway/web/freenode/ip.
13:58:11 Join antil33t [0] (
13:58:15*gevaerts thinks that tagcache/skin integration isn't as nice as it could be :)
13:58:37*JdGordon throws something at gevaerts
13:59:09JdGordonanyway, the .playlist_control file should be where these settings are saved, its pretty easy to add iirc
13:59:48S_a_i_n_tAnd, we have a volunteer!
14:00:20JdGordonalso, you dont really want me to touch playback... it would turn into a mess of pretend OO and events and a completly seperate module...
14:00:26JdGordonoh wait.. that sounds like a good idea! :)
14:00:31 Join ender` [0] (
14:00:52gevaertsI agree with the first part (that the .playlist_control file would be the best place), I'm just not sure if it's really that easy
14:02:54gevaertshm, maybe just a line with both numbers everytime it's changed...
14:04:21gevaertsOr every time the file is synced, although maybe that might miss some changes
14:05:09JdGordonjuist add a control line each time it is changed, then on restart the final value will be used which is what is wanted
14:09:08 Quit CGL (Remote host closed the connection)
14:12:36 Join madalu [0] (~user@unaffiliated/madalu)
14:15:21 Join MethoS- [0] (~clemens@
14:17:26 Join arcnaut [0] (~clementsp@
14:24:40 Quit madalu (Remote host closed the connection)
14:32:39 Quit stoffel (Ping timeout: 245 seconds)
14:38:40 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
14:41:51 Join krazykit [0] (
15:00:19***Saving seen data "./dancer.seen"
15:01:55 Join kugel [0] (~kugel@rockbox/developer/kugel)
15:12:06 Quit S_a_i_n_t (Ping timeout: 255 seconds)
15:15:18 Quit Jaykay (Ping timeout: 240 seconds)
15:18:46 Join Jaykay [0] (
15:19:01 Quit Horscht (Quit: Verlassend)
15:25:48 Join Horscht [0] (~Horscht@xbmc/user/horscht)
15:30:11 Join panni_ [0] (
15:35:12 Join anewuser [0] (anewuser@unaffiliated/anewuser)
15:42:37 Join hebz0rl [0] (
15:44:05 Join stoffel [0] (
15:45:05 Join esperegu [0] (~quassel@
15:45:18 Join fyrestorm [0] (
15:48:32 Quit Dogg006 (Quit: C YA!!!)
15:49:52 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
15:52:52 Join S_a_i_n_t [0] (S_a_i_n_t@
15:56:20 Quit edboyer93 ()
16:02:22*gevaerts doesn't really see what the proper place for resetting those numbers would be
16:03:31JdGordonstopping [playback
16:03:55gevaertsYes, but where is that? :)
16:03:58bertrikah, this would be a place where pausing and stopping would actually make a difference ! :P
16:04:49gevaertsbertrik: not really, you can resume after stop. Of course timestretch while stopped would be 0, while it might be something else while paused, but most people won't hear the difference
16:04:50JdGordonwhere in the code you mean?
16:04:57S_a_i_n_twho is "audio=i" on IRC?
16:05:08S_a_i_n_tbah, "audio-i" rather.
16:05:12JdGordoni dunno :) somehwere in playback.c
16:05:17JdGordonaudio_stop() maybe
16:06:37gevaertsJdGordon: actually, is playback stopped when one starts a new playlist while the old one is still running?
16:08:02gevaertsThis is tricky stuff :)
16:10:47JdGordonor do it in playback restart then?
16:11:00 Join fdinel [0] (
16:11:32JdGordoni guess it depends how much of playback should be stopped?
16:12:38 Join earcar [0] (
16:14:55 Quit [sko] (Quit: Leaving.)
16:22:17 Join Strife89DS [0] (~nds@
16:23:33 Quit earcar (Quit: bye)
16:25:25*S_a_i_n_t wonders how (without looking at code at all) FS #11622 - Improved Synthesis of talk clips for CamelCased files and directories deals with band/album names that have deliberate capital letters and/or neglected spaces. (small hometown example AtOmMoTher)
16:27:01JdGordoni would hope it tries to send the band back to school and get a real job!
16:27:06 Quit anewuser ()
16:31:02 Quit tchan (Quit: WeeChat 0.3.3-dev)
16:32:16 Quit domonoky (Read error: Connection reset by peer)
16:32:37 Join Dreamxtreme_ [0] (~Dreamxtre@
16:33:59 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
16:34:13 Quit Dreamxtreme (Ping timeout: 276 seconds)
16:34:35 Nick Dreamxtreme_ is now known as Dreamxtreme (~Dreamxtre@
16:34:49 Quit Jaykay (Ping timeout: 276 seconds)
16:37:38 Quit Borogrove ()
16:40:21 Join anewuser [0] (anewuser@unaffiliated/anewuser)
16:44:32 Join evilnick [0] (~Evilnick@rockbox/staff/evilnick)
17:00:23***Saving seen data "./dancer.seen"
17:01:30 Quit bimbel (Quit: Bye!)
17:02:16 Quit Strife89DS (Quit: ClIRC - IRC client for Nintendo DS)
17:03:50shuffle2request_handler_device_get_descriptor() is never called? :s
17:04:04shuffle2sigh, nvm
17:07:14 Quit JdGordon (Ping timeout: 252 seconds)
17:08:56 Quit evilnick (Quit: TTFN)
17:21:39 Quit pamaury (Ping timeout: 265 seconds)
17:26:23 Join pamaury [0] (~quassel@
17:26:23 Quit pamaury (Changing host)
17:26:23 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
17:37:05 Quit ender` (Quit: It's not English. It's legalese, a completely different language which looks confusingly similiar.)
17:40:13 Join ender` [0] (
17:44:52shuffle2any ideas why i can't call usb_core* and usb_drv* funcs from a plugin? error is "undefined reference"...
17:45:12shuffle2i seem to be doing it just like debug_menu.c
17:46:20 Quit panni_ (Read error: Connection reset by peer)
17:46:27 Quit ender` (Quit: There is no reason anyone would want a computer in their home. -- Ken Olson, president, chairman and founder of DEC)
17:47:29kugelshuffle2: you need to add them to the plugin api, and call them with rb->
17:48:18kugeldebug menu is no plugin
17:48:37*kugel wonders what happened about the usb drivers in plugins idea
17:49:37kugelgevaerts: ?
17:49:48shuffle2is there an api for apps as well?
17:49:55shuffle2i would rather not change the plugin api
17:50:01 Quit Jerom (Read error: Connection reset by peer)
17:50:29shuffle2or however you classify what the debug_menu is
17:50:59kugelapps/firmware/debug menu is all a single binary
17:51:24kugelplugins are seperate binaries, therefore you need to access core functions via an API
17:53:36 Join ender` [0] (
18:04:25pamauryusb drivers in plugins is far away, because it would require support for several plugins at the time, otherwise you can only load one plugin. One can consider this a minor limitation of course
18:04:46pamaury*load one driver only and no other plugin
18:08:46pamauryapart from this point, developping usb drivers in plugins should not be that complicated, the drivers api is already defined, and one only need to export usb functions to plugins. However, you need to think about how/when/why you load them and when you unload them, that's not exactly clear I think.
18:09:25shuffle2at this point i'm just hacking on the specific usb driver for my platform
18:09:32shuffle2not going to bother with plugin stuff
18:10:46pamaurywhat are you hacking with the usb stuff ?
18:11:47 Join Strife89DS [0] (~nds@
18:13:24shuffle2making this work on a pp502x:
18:13:57pamauryah ok
18:19:34shuffle2also: if i happen to write a firmware to the ipod with a bugged usb driver, i can save it by going into "disk mode", right?
18:20:01shuffle2just wondering if there's anything that will make you unable to use disk mode
18:20:28S_a_i_n_tWell, an iTunes Restore will fix anything short of smashing it with something larger than it is ;)
18:21:48kugelpamaury: right, we need some way to load multiple plugins (and hence not at a fixed address)
18:22:48pamaurythat would be great for sure, but much more complicated than the current loading system because you need relocations :( Or PIC code
18:23:25*pamaury leaves this part to the apps guy :)
18:23:33*S_a_i_n_t wonders if any of gevaerts' Ppebbles" code could be salvaged for this.
18:24:28kugelno, the outcome was that pebbles need relocations too :)
18:24:37pamaurypebbles didn't have any relocation right ? That's seems quite a limitation no ? Or can you have completely PIC code ?
18:26:20kugelit's still on gsoc ideas page, maybe I can be bothered next year :)
18:27:50 Join darkham [0] (
18:28:43pamaurywell, I'm not sure it's complicated enough for a gsoc. Basically, you just have to write an elf reader, or use a simpler format (I think gevaerts mentionned one at the time but it's wasn't perfect) and do relocations. Nothing impossible, it seems. But you need to get into it and write the code and DEBUG it.
18:29:40kugelthat's only one part
18:29:52kugelmaking use of it for plugins/codecs/usb drivers is a second part
18:29:57shuffle2yes...debugging this is a pita :P apparently the logging over serial usb doesn't work on windows? the wiki said this
18:30:30kugeland once we have it we can adopt other parts too, like metadata parsers
18:31:21pamauryshuffle2: no because it's just plain logging, windows wan't a full <insert fucking usb spec name here> compliant device
18:32:02shuffle2sure, but it seems like it could be easy to fake that, at least for this purpose?
18:32:05 Join Chronon [0] (
18:32:17kugelalso you shouldn't forget that this task can't be done in the sim, so writing/testing/debugging is much harder (and more time consuming)
18:32:22pamauryI once tried and gave up, it was just a pain
18:32:30pamaurywait, I can point you to the spec
18:34:59pamaurythat CDC devices (Communication Device Class) spec
18:36:20 Join robin0800 [0] (
18:36:34pamauryyou need to implement ACM (abstract control model)
18:36:43shuffle2also...i'm not familiar with atmel usb implementation (what original psgroove is designed for) or pp502x usb stuff (which i'm learning from a mx31 doc)...but it seems like atmel has implemented some things in hardware? for instance on atmel, you write which endpoint is "active" and such, and the hardware acts like a state machine for you. on the pp502x, it seems like the hardware acts on pre-defined structs you hand it in memory,
18:37:11kugelpamaury: I think he meant bflt. but I think we'd need our own binary format anyway due to iram stuff
18:37:23pamauryyes, that's it
18:37:34pamaurywhich driver is used for pp502x ?
18:37:35kugeliram means we need custom linker scripts, and if we have those we can just as well use our own format
18:38:00shuffle2just was a bit confused when reading psgroove source, and seeing that operations on mmio registers were a lot simpler than what pp502x seems to need
18:38:26pamauryhum, this one is nice but is quite high level
18:39:08pamaurythe controller expects in-memory structures, QHs and TDs, to describe the transfers and then do everything for you
18:40:00kugelthinking about it, part of the task would also be a task manager (so you can re-enter or quit one of the running plugins) so I think it would qualify for gsoc
18:40:31pamauryhum, true, there is the apps/user part
18:41:08pamaurythat would definitely require a student who knows rockbox already, otherwise that's impossible :)
18:42:24shuffle2pamaury so if rockbox was made cdc/acm compliant, would the rockbox project also require special software on the windows host? or is this built in to windows?
18:43:02pamauryIt would just require a software to read froma serial port I think, there are tons on the web
18:43:06shuffle2asking because i found third party software that does it:
18:43:08shuffle2ah ok
18:43:15pamauryif you are interested, there is a cdc_acm on linux gadget source code
18:43:57pamauryiirc, it was more complicated than actually needed but I might be wrong
18:44:12pamauryWait a minute
18:44:42 Part ste_iriveruser
18:45:44pamaurythat drivers/usb/gadget/f_acm.c in linux code, I'm reading through it
18:46:41pamaurysome comments suggets it might not work because windows has a buggy driver but linux sometimes has buggy comments too :)
18:48:40 Quit antil33t (Read error: Connection reset by peer)
18:48:47 Join antil33t [0] (
18:49:15pamaurythe code looks simple, augmenting the rockbox driver shouldn't be too hard, I can have a look at it if you are really interested.
18:49:44 Join clone4crw [0] (
18:58:24 Join n1s [0] (~n1s@rockbox/developer/n1s)
18:59:17shuffle2pamaury not really iterested, just curious :)
19:00:24***Saving seen data "./dancer.seen"
19:01:19pamaurynobody really use logf over usb ;)
19:02:36 Join milz [0] (
19:02:50pamaurythe nice point about cdc_acm is that for linux, the driver would be automatically loaded, whereas you currently need to manually load usbserial with the PID/VID
19:05:44kugelpamaury: because nobody knows how it works :)
19:06:21pamaurythe wiki explains it ! But you need to recompile rockbox, that's true
19:09:02 Quit pamaury (Remote host closed the connection)
19:12:08bertrikIs anyone currently looking at the fuze+ OF?
19:12:45S_a_i_n_tIIUC funman has looked at it, whether he's looking at it *currently* or not is another story ;)
19:13:04 Join pamaury [0] (
19:13:05 Quit pamaury (Changing host)
19:13:05 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
19:13:33 Join Kitr88 [0] (
19:14:48 Join Jaykay [0] (
19:16:50 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
19:16:58 Quit Kitar|st (Ping timeout: 240 seconds)
19:18:30 Quit antil33t (Read error: Connection reset by peer)
19:18:37 Join antil33t [0] (
19:28:08 Quit mc2739 (Ping timeout: 272 seconds)
19:29:50 Quit Strife89DS (Ping timeout: 240 seconds)
19:29:57 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
19:42:02 Quit stoffel (Remote host closed the connection)
20:04:36 Quit robin0800 (Remote host closed the connection)
20:05:42 Quit n1s (Quit: Lämnar)
20:10:32pamauryfucking usb controller, I don't understand it randomly fails on clip+ ! The most incredible thing is that it keeps failing until you disconnect it !
20:11:08*pamaury concludes it's reliably random
20:11:26kugelthat's sure better than randomly random
20:11:45bertrikpamaury, didn't you also see a lot of CRC errors?
20:12:28pamauryyes but why it keeps failing ? When it works the first time, it works until you unplug usually. When it doesn't work, the kernel keeps resetting the device and retrying but it fails each time.
20:12:34 Quit MethoS- (Remote host closed the connection)
20:12:57pamauryYes I see CRC errors, the problem is more the rate of crc errors when it fails
20:13:08kugeltiming or cache are usually the answer to random problems but I don't know
20:13:12bertrikmaybe there's something wrong with the setup of the USB PHY (or whatever the thing is called that does the wire coding)
20:15:11pamaurykugel: that seems unlikely it this context, more something like usb phy setup or syncing, I don't really know, but something must be wrong
20:15:45 Join petur [0] (~petur@rockbox/developer/petur)
20:16:13 Join Misanthropos_ [0] (
20:16:19pamauryperhaps when it fails to much, we should do a hard reset ?!
20:16:20bertrikIIRC, PP502x USB only started working reliably once we got the magic init sequence of the PHY right
20:16:55kugeldoes this also happen when the nano2g code is used?
20:16:56pamaurywe already extracted the magic sequence from of, but it's true that funman modified it several times
20:17:14kugeli.e. with FS #11607?
20:18:37pamauryno, in svn. the current one is not the original one I extract from the of.
20:18:54pamaurynotably is the clock setup
20:19:08kugelif it works better with FS #11607 it should give a hint
20:19:17pamauryno it doesn't
20:19:26pamaurythere is the same problem
20:19:38kugelyou answered no to my question so I thought it would work better
20:21:01pamauryapparently not, funman told me he experienced the same kind of ramdom failure, I believe him
20:22:13kugelit worked nice for me
20:22:23kugelbut I didn't try very hard
20:22:26 Join Strife89DS [0] (~nds@
20:23:18pamauryyou can plug it 20 times before experiencing a failure
20:23:51kugelisn't linux currently buggy anyway?
20:24:09bertriksomehow I get that with the OF too, but maybe that's just my laptop
20:24:17kugelwith all my daps I often get no usb connected (until I switch the usb port), with both rockbox usb and OF usb
20:24:34kugelI think funman also experiences this
20:28:02 Join funman [0] (4dc183bc@rockbox/developer/funman)
20:28:54S_a_i_n_tfunman uses his special "someones talking about me" connect script ;)
20:29:23funmankugel: no it works fine, i had problems with a specific computer where when rockbox failed (on AMSv2) i had to rmmod / modprobe ehci_hcd
20:29:32*funman is watching you
20:30:11gevaertsS_a_i_n_t: something that's not generally known is that funman is actually logbot, but with a tie
20:33:19*pamaury tries crazy thing to try to solve the problem
20:34:09*S_a_i_n_t tells pamaury that yelling at it simply won't work.
20:34:31 Join Highlander_ [0] (
20:35:32funmanbertrik: about fuze+ OF it is some Sigmatel chip
20:35:43pamaurywhich one ?
20:36:14funmanwe'll know when we have pictures
20:36:48funman"STMP36" iirc (there is a "STMP" string in the OF but don't remember if there's a number)
20:38:18 Quit Highlander (Ping timeout: 240 seconds)
20:40:12pamauryfunman: did you had a look at the S3C usb driver in linux ?
20:41:05funmanyes, i was looking at it when i suggested to write to Ben Dooks did you forget?
20:41:43pamauryyes, did you notice that on usbreset it does far less than us ?
20:41:47funmanI could solder JTAG to one of my clips but I don't have a // port on any of my computers
20:42:19funmanbut IIUC it could be due to DMA
20:44:01pamauryfor example, it does not reset devadre*
20:44:01 Quit Misanthropos_ (Ping timeout: 258 seconds)
20:56:47 Join Misanthropos_ [0] (
21:00:28***Saving seen data "./dancer.seen"
21:03:51*S_a_i_n_t looks at and wonders why this isn't done already.
21:04:22S_a_i_n_tI have already inserted quite a few spaces in my english.lang file to make a few words sound even remotely human.
21:07:00bertrikthe first 6.5 MB of the fuzeplus .sb file looks like random data (could be the executable code, scrambled) , after that I see some images
21:10:30 Quit darkham (Read error: Connection reset by peer)
21:11:11 Quit Horscht (Quit: Verlassend)
21:11:23funmanbertrik: the .sb file format supports encryption
21:11:47funmanthere is a proprietary elftosb provided by sigmatel/freescale
21:11:55bertrikpamaury, you said you have experience with the .sb file format, right?
21:13:51bertrikI so see a bit of a "chunk directory" at the start of the file
21:16:09 Quit funman (Quit: Page closed)
21:18:35 Quit Misanthropos_ (Quit: Ex-Chat)
21:23:14 Join r0b- [0] (
21:24:19pamauryyes I experienced with the .sb format, it's just a binary format to fill to memory
21:24:56pamauryit doesn't really have a structure, at least with the chip I worked with
21:33:23 Quit Chronon (Ping timeout: 276 seconds)
21:34:49 Quit Jaykay (Ping timeout: 272 seconds)
21:35:36bertrikpamaury, there is a "crsr" tag ("rsrc" => resource) in the start of the file, with a little-endian number behind that points to the address just after the random data (when multiplied by 16)
21:36:08bertriknear that address is alos a text "RSRC"
21:36:24 Join Horscht [0] (~Horscht@xbmc/user/horscht)
21:40:59bertrikthe first part of about 6.5 MB looks completely without structure when viewed as a grayscale image, so I fear it's encrypted (or at least scrambled) indeed
21:44:23 Quit Strife89DS (Quit: ClIRC - IRC client for Nintendo DS)
21:45:58pamauryhum, to me a .sb file doesn't have any structure, it just says: write [chunk] at [address]. For rsrc format, I don't know
21:46:19pamauryand the data sent to the chip is usually encrypted. But perhaps it's not the same format !
21:48:08bertrikpamaury, I meant that it looks like random noise pixels when I said "without structure"
21:50:09pamaurythen perhaps the chip data is between CRSR tag and beginning of rsrc data
21:55:50 Join user890104 [0] (
21:58:37 Quit Dreamxtreme (Ping timeout: 252 seconds)
22:02:21 Join Dreamxtreme [0] (Dreamxtrem@
22:19:16 Join Chronon [0] (
22:23:57 Join thals1992 [0] (466968f3@gateway/web/freenode/ip.
22:25:06 Quit S_a_i_n_t (Quit: Even if you're lying, please tell me everythings going to be fine.)
22:36:16thals1992Hi. I have issues with my iPod Radio on the Nano (G1) with all recent builds. When I go into the FM Tuner and choose a station, it is inaudible over the headphone socket in the iPod itself, but when I plug it in the Radio, it is noisy (noise comes over headphones - not powered speakers) and extra noise comes in when keys are pressed on the Radio itself, when the screen scrolls the RDS data, and when the iPod wheel is touched.
22:36:26 Quit kugel (Ping timeout: 276 seconds)
22:37:01thals1992The music itself was not audible, so I bumped the volume to db 0 and even at one point 6 with the headphones plugged into the accessory and then the music over the radio can be heard, but faint. It is just fine when listened to the recording after amplifing (using the gain setting) 20 db and the vu meter underneath looks like it is receiving sound at a decent volume.
22:40:39 Join [sko] [0] (~sko]
22:40:40 Join Jaykay [0] (
22:41:00thals1992I posted the bug to the tracker and noone has replied in a month (FS #11559) so I was hoping this wouldn't go dead, because it is important to have this fixed.
22:42:00gevaertsI suspect not many developers actually have the hardware to reproduce this
22:42:34thals1992I figured so...
22:45:20thals1992Even though this occurs how would a patch be created to add gain to listening to it so that it can be heard over the ipod radio, because it will not work over the headphone socket on the ipod?
22:47:13shuffle2"<pamaury> nobody really use logf over usb ;)" > was this a joke or is there a better method?
22:48:10 Join kugel [0] (~kugel@rockbox/developer/kugel)
22:50:01 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
22:54:26 Join kramer3d [0] (~kramer@unaffiliated/kramer3d)
22:56:21 Quit Jaykay (Ping timeout: 240 seconds)
22:58:07thals1992How would I create a patch to increase listening volume (add gain) to the iPod Radio, without changing it on the iPod's volume as the to act like separate volume devices.
23:00:30***Saving seen data "./dancer.seen"
23:05:24pamauryshuffle2: it's just a fact, the only time I used it was when debugging mtp code
23:05:42bertrikthals1992, is this a recent problem?
23:05:47pamaurythe point is that usually I do usb stuff so you can't easily debug usb code with usb :)
23:05:50bertrikI mean, did it work ok before?
23:06:00shuffle2ya true :p
23:06:13pamauryand other people don't really use it, but it works
23:08:13thals1992it work only in recording.
23:09:25pamauryshuffle2: and it particular, nearly no dev uses windows so usbserial+windows is kind of unusual combinaison
23:10:40 Join edboyer93 [0] (
23:10:51thals1992The radio is audile only in the accessory audio socket and when the volume it completely turned up (6 db)
23:11:57gevaertsthals1992: just to make sure, it works in the OF?
23:12:07 Quit guymann (Quit: brb~)
23:12:10thals1992Yes, perfectly.
23:12:38thals1992Works in the ipod socket and the radio's socket
23:12:49thals1992but only in OF
23:14:01thals1992it occurs in every build I've used, even before build 3.6
23:14:54thals1992I just tried the lastest bulid to confirm, and it still occurs.
23:16:54thals1992The reason I brought it up was that it is unsuitable for using often. Its fine in recording, but listening is really difficult.
23:16:55pamauryfunman (for the logs): I tried to tweak the usb driver to make it work but no success. In particular, I had no success with the cgu_usb thing but I might have did it wrong there :) I'll try another method which consists in calling usb_drv_exit and then usb_drv_init after, say 3 usb resets. This way, perhaps it will reset the controller to a know ststae but I'm not too confident. Strangely, it seems that the problems only occur all the time or
23:16:55pamaury never, but it doesn't change until I disconnect the device. So perhaps a solution would be after say, 3 resets, to do hard disconnection if possible, that is have a low VBUS so the host see a disconnection. We never now, perhaps an electrical reset can solve our problem :)
23:18:37bertrikmaybe the radio problem has something to do with input source muxing
23:19:05 Quit bmbl (Quit: Bye!)
23:19:55 Join guymann [0] (~charlie@
23:21:07thals1992that's the reason, I thought it only needed an extra 20 db, like mics require.
23:23:19 Join cmmc [0] (
23:23:29 Quit petur (Quit: here today, gone tomorrow)
23:23:53 Quit Chronon (Ping timeout: 276 seconds)
23:29:58 Quit thals1992 (Quit: ?)
23:41:05 Quit edboyer93 ()
23:55:33 Join Judas_PhD [0] (
23:58:38 Quit ender` (Quit: Religion is regarded by the common people as true, by the wise as false, and by the rulers as useful. -- Seneca)

Previous day | Next day