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).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2010-03-22

00:01:21 Join webguest87roko [0] (
00:02:26eggblokeThe website says a 2nd gen nano will be buggy but it seems fine
00:03:18 Quit robin0800 (Remote host closed the connection)
00:03:58 Join robin0800 [0] (
00:06:00 Quit robin0800 (Remote host closed the connection)
00:06:25 Join robin0800 [0] (
00:06:49 Quit kugel (Disconnected by services)
00:06:55 Join kugel [0] (
00:07:07 Quit kugel (Changing host)
00:07:07 Join kugel [0] (~kugel@rockbox/developer/kugel)
00:10:50 Quit EpicCyndaquil (Read error: Connection reset by peer)
00:10:54 Join kugel_ [0] (
00:11:06 Quit kugel (Disconnected by services)
00:11:10 Nick kugel_ is now known as kugel (
00:11:14 Quit kugel (Changing host)
00:11:14 Join kugel [0] (~kugel@rockbox/developer/kugel)
00:13:44 Join EpicCyndaquil [0] (
00:17:06 Quit webguest87roko (Quit: CGI:IRC)
00:22:33 Quit ender` (Quit: The moment a person forms a theory, his imagination sees in every object only the traits which favor that theory.-- Thomas Jefferson)
00:24:08 Part domonoky
00:25:57 Quit robin0800 (Quit: No Ping reply in 180 seconds.)
00:26:18 Join robin0800 [0] (
00:29:21 Quit DataGhost (Ping timeout: 258 seconds)
00:29:46 Join hebz0rl [0] (
00:31:11 Quit petur (Remote host closed the connection)
00:36:26 Quit eggbloke (Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115144158])
00:39:44 Join JdGordon [0] (~7bf38c1f@gateway/web/freenode/x-rphqcbnumazxrpvl)
00:42:22 Quit EpicCyndaquil (Read error: Connection reset by peer)
00:43:57 Join EpicCyndaquil [0] (
00:44:22 Quit mt (Ping timeout: 252 seconds)
00:44:34 Quit kadoban (Read error: Connection reset by peer)
00:45:00 Join kadoban [0] (
00:46:39 Join mt [0] (~mtee@rockbox/developer/mt)
00:49:50 Quit kadoban (Read error: Connection reset by peer)
00:50:50 Join kadoban [0] (
00:51:54 Quit soap (Remote host closed the connection)
00:56:05 Join linuxstb [0] (~linuxstb@rockbox/developer/linuxstb)
01:00:19 Join soap [0] (~soap@rockbox/staff/soap)
01:02:45 Join robin0800_ [0] (
01:03:21 Quit robin0800 (Ping timeout: 268 seconds)
01:05:26 Join n17ikh [0] (
01:07:56S_a_i_n_tHmmmm...r25280 seems to bork a couple of splash screens, it either doesn't display the splash screen at all (Shutting Down... for instance), or displays nothing but the splash screen (Database Init for example). In the cases where the splash is shown in only draws the splash and the backdrop (the status bar disappears) and when it is shown its very quickly overwritten by a blank backdrop.
01:08:52 Quit perfectdrug (Quit: perfectdrug)
01:08:58 Nick fidencio is now known as fidencio[AWAY] (
01:09:36S_a_i_n_tTheSeven saw similar things in 25279, if I had to put the blame on something I'd look at 25271 (but thats just a bit of a guess)
01:14:06 Join akg518 [0] (
01:16:00S_a_i_n_tthere's plenty of people have a question?
01:16:10S_a_i_n_t^ akg518
01:16:30***Saving seen data "./dancer.seen"
01:16:44 Quit robin0800_ (Quit: No Ping reply in 180 seconds.)
01:17:10 Join robin0800 [0] (
01:17:50akg518Nope just saying helo
01:19:27JdGordonS_a_i_n_t: i very much doubt that revision did anything
01:21:38S_a_i_n_tHmmm, I can't bisect it right now as I'm not near my build environment, I know it was working at 25269
01:21:47S_a_i_n_tso that at least narrows it to 10
01:27:26S_a_i_n_tHmmm, questioning the revision number in my head made me have an idea. Would it be possible for System/Rockboxinfo to display (if it can pull it from the build environment somewhere/how) the *previous) revision before update?
01:27:44S_a_i_n_tThis would be useful for 'last good/first bad's etc.
01:31:33 Quit akg518 (Quit: CGI:IRC (EOF))
01:32:48 Quit Rob2222 (Remote host closed the connection)
01:32:58 Join Rob2222 [0] (
01:35:39 Quit robin0800 (Quit: No Ping reply in 180 seconds.)
01:36:05 Join robin0800 [0] (
01:39:21JdGordonthere is only one change between them qhich could have broken anything
01:40:04moosS_a_i_n_t: there is indeed a bug but don't from which revision. The screen seems to blank fully until key press
01:40:37JdGordonblank how? where?
01:40:51moosI mean not just splashes
01:41:32 Quit moos (Remote host closed the connection)
01:41:58 Quit Alexandru_Criste ()
01:42:09 Join moos [0] (moos@rockbox/staff/moos)
01:42:20moosJdGordon: blank screen, just backdrop left, but I didn't check yet if that's setting related maybe...
01:42:32JdGordonwhich screen?
01:42:35mooswhen? I have it since boot
01:42:49moosevery :)
01:43:00JdGordondo you have a .sbs?
01:43:22moosI use the default one
01:44:05*moos goes for searching his beast, if retails is needed
01:44:21*S_a_i_n_t is seeing the behaviour he described while using all setting at default.
01:44:36*JdGordon is installing the latest build on his mini
01:46:17moosrectification, bacdrop and sbs are left here
01:46:56moosand I have it just after boot.
01:48:46 Quit ObsidianX (Disconnected by services)
01:48:57 Quit moos (Read error: Connection reset by peer)
01:49:17 Join moos [0] (moos@rockbox/staff/moos)
01:49:36moosit seems that the screen isn't upadted correctly somewhere
01:50:25JdGordonno issues with default settings on mini2g
01:51:20moosthe keypress that wakes up the screen blank the screen it seems for a few ms, let's see with default setting
01:52:06S_a_i_n_tJdGordon: try and see if you have the "shutting down..." splash
01:52:51JdGordonok, that is curious
01:53:01 Quit hebz0rl (Quit: Ex-Chat)
01:53:09moosS_a_i_n_t: that *just* splashes on your hand?
01:53:36*S_a_i_n_t resists...
01:55:11mooshehe, nice to see the sbs here after loading my saved confing. There must have a culprit setting here
01:55:40 Join saratoga_lab [0] (~9803c20d@gateway/web/freenode/x-yvhrljsatwiwyffe)
01:55:58mooswow crazy screen now? I'll look on my .cfg and paste it somwhere
01:56:04 Nick fidencio[AWAY] is now known as fidencio (
01:57:37 Quit Casainho (Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115133306])
01:59:36 Quit moos (Remote host closed the connection)
02:00:15 Join moos [0] (moos@rockbox/staff/moos)
02:01:40moosJdGordon: my congig if that helps
02:03:05JdGordonsbs: /.rockbox/wps/ might be problematic
02:03:19JdGordonall splashes work fine for me except the shutdown one!
02:04:03moosJdGordon: the system have changed since ".cfg file created by rockbox r24312-100121"?
02:04:51moosI must have missed the commit message that advise people
02:05:51mooswith this config there is an update screen problem it seems, if that could help you to find the cause
02:06:31 Quit Strife89 (Read error: Connection reset by peer)
02:06:44 Join Strife89 [0] (
02:07:22S_a_i_n_tI get the statusbar disappearing when the database init stuff splashes
02:07:35JdGordonthat is expected
02:07:45JdGordonthe splashes disable the theme
02:08:13moosHow the sbs line on the config have to look like?
02:08:31JdGordonno, the actual sbs might have issues
02:08:35JdGordonthe line is correct
02:08:40moosah sorry
02:08:43 Quit EpicCyndaquil (Read error: Connection reset by peer)
02:10:31 Join EpicCyndaquil [0] (
02:24:16 Quit EpicCyndaquil (Remote host closed the connection)
02:34:50 Quit adnyxo (Remote host closed the connection)
02:37:35 Quit Rob2222 (Quit: Rob2222)
02:47:27 Join planetbeing_ [0] (
02:48:31S_a_i_n_tJdGordon I figure you're the one to ask about this...can I / how can I set the splashes to usewhite text on a black backgroung instead of black text on a grey background.
02:48:38S_a_i_n_tI find the latter hard to read.
02:56:08 Join FlynDice_ [0] (~FlynDice@
03:16:34***Saving seen data "./dancer.seen"
03:18:00 Quit MethoS- (Remote host closed the connection)
03:21:51 Quit Schmogel (Quit: Miranda IM! Smaller, Faster, Easier.
03:24:40 Join shaggy-h [0] (
03:29:42CIA-5New commit by 03FlynDice (r25281): sd-as3525v2.c Use SD_DESELECT_CARD instead of SD_SELECT_CARD to make the code easier to understand. There is no functional difference. ...
03:29:47 Quit planetbeing_ (Quit: planetbeing_)
03:29:55CIA-5New commit by 03FlynDice (r25282): sd-as3525v2 Remove post transfer wait for state call and move pre transfer wait for state call inside the transfer loop. ...
03:30:01CIA-5New commit by 03FlynDice (r25283): sd-as3525v2.c Add macros to distinguish between a CMD error and a DATA error.
03:31:10S_a_i_n_tquick question FS6697 defines the place to look for the dictionary .desc files by using "#define #define DICTS_DIR "/dicts" " which places a 'dicts' folder in the players root. If I wanted it to be in the dir "/.rockbox/dicts" I'd change the define to "DICTS_DIR ROCKBOX_DIR "/dicts" "
03:31:21 Quit Strife89 (Quit: Bed.)
03:32:19kugelhow about you simply try it?
03:33:14S_a_i_n_tFuck, *so* sorry for wanting to clarifty something...
03:37:50 Join Strife89|PalmTX [0] (
03:38:50 Join xiainx [0] (
03:41:59 Quit kugel (Remote host closed the connection)
03:42:04xiainxHi, I saw Rockbox was accepted into GSoC, I looked at your proposals and some of them interest me
03:44:21 Join bzed_ [0] (
03:44:22 Quit bzed (Read error: Connection reset by peer)
03:44:53 Nick bzed_ is now known as bzed (
03:52:21saratogaxiainx: i have a minute, which were you itnerested in
03:55:31xiainxsaratoga: I'm interested in a few actually
03:55:44 Quit robin0800 (Quit: No Ping reply in 600 seconds.)
03:55:54xiainxsaratoga: The Filesystem storage engine, MySQL to Drizzle syntax translator, Unit testing framework, and the documentation projects are all ones im interested in
03:56:09 Join robin0800 [0] (
03:56:24xiainxno, that's another project
03:57:15xiainxsaratoga: the rockbox as an application, TTS, and online services are all ones i'm interested in
03:59:36saratogaxiainx: i worked on rockbox as an app during a previous summer of code
04:00:01saratogait involves a lot of digging through our code and then understanding some embedded environment
04:00:15saratogalike android or whatever
04:00:46xiainxsaratoga: Really, so like, what kind of things does one need to know/be capable of before starting on a project?
04:01:44saratogathat project would involve basically being really good at greping through code and hopefully have some experience with linux development (or i suppose winmobile or macos)
04:01:56saratogathe first part is getting the sim building from our target tree
04:02:12saratogabasically right now we have a target tree for hardware devices
04:02:21saratogaso each CPU family has a folder, then devices go in there, etc
04:02:37xiainxright, that's pretty typical directory structure
04:02:41saratogathe idea would be to add the sim as a "target" and maybe have an "android" cpu type or something similar
04:02:59saratogathat way android devices could compile just like arm or coldfire devices
04:03:24saratogabecause currently the sim is compiled seperately from everything else, and about a million #ifdef commands handle it
04:03:37saratogathats obviously unworkable for a device port since you can't really change things easily
04:04:15saratogaso first that would be fixed ( i have a rough patch for doing so on the tracker somewhere but it needs a lot of work)
04:04:33saratogathen there would be an actual port to some device as sort of an example of how compiling rockbox as an application would work
04:04:44xiainxLike.... What kind of device?
04:05:00saratogaandroid phone, iphone, ipod touch, wince, etc
04:05:03saratogadoesn't really matter
04:05:26saratogaeven just targeting the android emulator would be fine actually
04:06:53saratogathe idea is to have a proof of concept port at very least so that other people can see how its done, then from there try and get a usable port if at all possible
04:08:00xiainxRight, makes sense
04:08:09 Quit Strife89|PalmTX (Quit: Bed, for real this time.)
04:08:09 Quit moos (Ping timeout: 264 seconds)
04:11:04 Quit Darkknight512 (Remote host closed the connection)
04:11:17 Quit robin0800 (Quit: No Ping reply in 600 seconds.)
04:11:44 Join robin0800 [0] (
04:21:36tmzt_if there's an sdl port why is a specifc rockbox port needed?
04:21:49saratogatmzt_: elaborate on that thought
04:22:30tmzt_well, the process to port rockbox to android, probably using the ndk and opengl for graphics has already been done for libsdl
04:22:46tmzt_the same for the exz phones (old port) and webos (working sdl)
04:24:19 Join Barahir [0] (
04:25:19saratogatmzt_: yes using sdl is a very good option but its only a very small part of actually doing the project
04:25:45saratogascroll up a couple lines and get an idea whats involved
04:25:50tmzt_I wasn't getting the target/cpu stuff
04:26:03tmzt_since it looks like a simulator build on linux abi
04:26:15 Quit TheSeven (Disconnected by services)
04:26:25 Join unic0rn [0] (unic0rn@unaffiliated/unic0rn)
04:26:28 Join The_Seven [0] (~theseven@rockbox/developer/TheSeven)
04:26:32tmzt_of course sim has internal codecs and pcm output
04:26:38 Nick The_Seven is now known as TheSeven (~theseven@rockbox/developer/TheSeven)
04:27:04tmzt_and on something like android radio firmware does most of the audio decoding for things like mp3, aac, wma
04:27:50 Quit Barahir_ (Ping timeout: 276 seconds)
04:29:52saratogatmzt_: what?
04:29:58 Part unic0rn
04:30:09 Join moos [0] (moos@rockbox/staff/moos)
04:30:26tmzt_on qualcomm msm chipset devices like g1, mytouch, hero, nexus
04:30:55tmzt_the radio firmware provides many system functions including media through it's qdsp framework
04:31:04tmzt_media decoding
04:31:16saratogayou're trying to say that theres a DSP for decoding?
04:31:33saratogai doubt they use something like that for audio
04:31:50saratogadoesn't make much sense given how fast codecs are on arm
04:32:08tmzt_well they do, whether hardware dsp is used or not all audio playback is through radio firmware
04:33:21saratogai don't doubt that the device decodes audio
04:33:35saratogai'm just saying its probably done on the CPU
04:33:56tmzt_there's two cpu cores
04:34:02tmzt_apps core and modem core
04:34:18tmzt_most hardware registers are only accessible from modem core
04:34:27tmzt_through a shared memory interface
04:35:32tmzt_if you were running rockbox directly on the g1 you couldn't avoid this
04:35:51tmzt_but there's enough abstraction, if you do decoding in userspace, to just output pcm
04:35:57tmzt_so it could be a non-issue
04:39:33 Quit CaptainKewl (Remote host closed the connection)
04:39:48 Join chrisb [0] (
04:44:03saratogahmm actually google says there is a QDSP5000 based WMA Pro decoder, so at least that much could be done on the dsp
04:44:13saratogai wonder why they bother though, that codec is so fast to being with
04:44:24saratogaunless they have systems that are qdsp only with no arm core
04:44:54 Quit panni_ (Read error: Connection reset by peer)
04:47:39S_a_i_n_tSuggestion for the manual: WPS Tags, Next Track Info - Display the tags as I'n', D'n', F'n' (instead of Ix,Dx, Fx) so as to be less misleading and also for greater consitency with the rest of the " 'n' value tags listed "
04:49:03S_a_i_n_t*edit: Also give clarification on 'n' value, currently x has no clarifiaction and may lead to new users believing the tags actually *are %Ix, %Dx, %Fx etc.
04:50:20 Quit anewuser ()
04:56:56 Join anewuser [0] (anewuser@unaffiliated/anewuser)
05:03:01 Join angelwolf71885 [0] (
05:06:32 Quit angelwolf71885 (Client Quit)
05:12:11 Quit CGL (Quit: Saliendo)
05:14:47 Nick fidencio is now known as fidencio[AWAY] (
05:16:38***Saving seen data "./dancer.seen"
05:17:47JdGordonS_a_i_n_t: patches are always welcome :)
05:18:14JdGordonand im not sure how hard (probably not very) it would be to change the splash colours
05:20:40S_a_i_n_tDo you know where they are defined (the splash colours)? And, Would submitting the appropriate new wordinf for the manual suffice? I have no idea how to patch/build the manual.
05:21:51JdGordonI think for the manual they dont mind just getting a blurb with proper wording. and apps/gui/splash.c
05:22:06 Quit rvvs89 (Ping timeout: 276 seconds)
05:22:47 Join rvvs89 [0] (
05:27:13 Quit elinenbe (Ping timeout: 248 seconds)
05:29:04 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
05:30:18 Quit Horscht (Client Quit)
05:37:34 Join n1s [0] (~n1s@rockbox/developer/n1s)
05:38:31 Join Xerion_ [0] (
05:40:06 Join switchgirl [0] (~sara@unaffiliated/switchgirl)
05:40:18 Quit alexbobp (Quit: reboot)
05:41:02 Quit Xerion (Ping timeout: 260 seconds)
05:41:02 Nick Xerion_ is now known as Xerion (
05:41:45switchgirlhi i gots a light blue ipod shuffle (the type with a clip and about the size of a stamp) wadda i need to do (i'm also running Linux Ubuntu) to get rockbox on there - or to get it to play ogg
05:42:01 Quit n1s (Ping timeout: 252 seconds)
05:42:23S_a_i_n_tWell, first of'd need to develop a port to that target ;)
05:43:23*switchgirl looks at S_a_i_n_t like they just cured cancer lists supported targets, and unfortunately for yourself, the Shuffle isn;t one of them.
05:43:42switchgirlin otherwords what do you mean?
05:44:32 Quit xiainx (Quit: Good Bye!)
05:44:41S_a_i_n_tSorry to be the bearer of bad news...Your're actually the second person to ask about the Shuffle today.
05:45:04S_a_i_n_tFirst time I've ever seen anyone ask about it myself, well times(s)
05:45:06 Quit bluebrother (Disconnected by services)
05:45:07 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
05:46:26 Quit anewuser (Quit: lolbama care)
05:46:52switchgirlS_a_i_n_t: thanks
05:46:59 Quit Bagder (Remote host closed the connection)
05:50:26 Join n1s [0] (~n1s@rockbox/developer/n1s)
05:51:19 Join alexbobp [0] (~alex@
05:53:52 Nick fidencio[AWAY] is now known as fidencio (
05:56:12 Quit Zarggg (Quit: Zarggg)
05:57:10 Join FOAD_ [0] (
06:00:03 Join Zarggg [0] (
06:01:22 Nick fidencio is now known as fidencio[AWAY] (
06:01:25 Quit FOAD (Ping timeout: 276 seconds)
06:01:26 Nick FOAD_ is now known as FOAD (
06:07:32 Quit kaniini (Read error: Connection reset by peer)
06:20:59 Quit robin0800 (Ping timeout: 240 seconds)
06:23:49 Quit n1s (Ping timeout: 252 seconds)
06:41:15 Quit saratoga_lab (Quit: Page closed)
06:45:48 Quit BHSPitMonkey (Remote host closed the connection)
06:52:15 Join shai [0] (
07:09:10CIA-5New commit by 03FlynDice (r25284): sd-as3525v2.c Use MCI_MASK instead of data_transfer flag to determine if data transfer is happening.
07:09:17CIA-5New commit by 03FlynDice (r25285): sd-as3525v2.c Enable disk access icon.
07:09:20CIA-5New commit by 03FlynDice (r25286): sd-as3525v2.c We don't need to disable/enable MCI_CLKENA in send_cmd().
07:10:15 Join Bagder [0] (~daniel@rockbox/developer/bagder)
07:16:41***Saving seen data "./dancer.seen"
07:25:24 Join DV__ [0] (~DV@
07:27:34 Quit DV__ (Read error: Connection reset by peer)
07:27:56 Quit DV (Ping timeout: 264 seconds)
07:28:49 Join DV [0] (~DV@
07:29:44 Quit Adubb (Ping timeout: 264 seconds)
07:33:20 Join Adubb [0] (~aldubuc@
07:33:56*pixelma doesn't understan S_a_i_n_t's next track info tags proposal, not even with the clarification
07:38:21 Join merbzt [0] (~benlar@
07:39:37 Quit FlynDice_ (Remote host closed the connection)
07:41:11 Quit merbzt1 (Ping timeout: 276 seconds)
07:43:05 Join Zagor [0] (
07:43:05 Quit Zagor (Changing host)
07:43:05 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
07:44:25 Quit DV (Read error: Connection reset by peer)
07:48:33 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
07:48:36 Quit bmbl (Remote host closed the connection)
07:49:28 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
07:50:37pixelmaah... in the manual, now I get it
07:52:05 Join DerPapst [0] (
08:02:10 Quit Zagor (Quit: Clint excited)
08:04:02 Quit bmbl (Ping timeout: 246 seconds)
08:06:14 Join fejfighter [0] (
08:12:58 Quit Xerion (Ping timeout: 265 seconds)
08:18:37pixelmaI don't see why t that would be better - for a few reasons. Gotta run though, I'll explain later
08:28:41 Quit DerPapst (Quit: Leaving.)
08:31:23 Join ender` [0] (
08:32:47 Quit TheSeven (Ping timeout: 265 seconds)
08:34:18 Join planetbeing_ [0] (~planetbei@
08:34:34*JdGordon pretty much gives up on fixing the draw order in the skins
08:38:05 Quit planetbeing_ (Client Quit)
08:39:59LloreanGodEater: For the "play all albums with the tracks in album order" problem, wouldn't the best solution be to have the database always "insert" in the same order it presents things to the user? So if you "insert" all tracks, you get them in alphabetical, but if you insert "all albums" you get the albums in alphabetical order, but the tracks within each album in track order, etc?
08:40:16LloreanBasically, wouldn't the ideal be to walk the database, rather than always returning the "all tracks" sub-list?
08:40:20 Join planetbeing_ [0] (
08:40:20GodEaterthat might be more intuitive yes
08:40:31GodEaterbut I've no idea how much work that would be
08:40:43LloreanI thought it used to do that already, at some point in the past, but I must be wrong.
08:40:57LloreanIt seems like it would be the expected behaviour by the user, though, in general
08:41:06GodEatercertainly trying to get the tagnavi syntax to get it to do that caused a data sbort
08:41:07bluebroth3rJdGordon: you have the cross compilers running on OS X? On 10.6?
08:41:13LloreanIf you insert what appears to be a folder, the playlist contains what you'd see if you view that folder
08:41:41JdGordonbluebroth3r: I thought i did, but not yet.. I cant get them going if needed
08:41:45GodEaterbut if you do view the folder you can see they're not in album order
08:41:51JdGordonI'm not sure why anyone would need to cross compile on that box though?
08:42:12GodEaterI'm not sure if the problem is in the database code, or our shipping default tagnavi file
08:42:31GodEaterI would have thought the latter, since that's what defines those folders
08:42:50GodEaterbut the "correct" syntax doesn't work at all
08:42:56bluebroth3rJdGordon: I wanted to set them up but building the compilers fail (in most cases binutils work, but gcc fails with a "trap abort" or similar).
08:43:07bluebroth3rwanted to set them up on my box that is
08:43:25JdGordonI'll start the script and see what happens
08:43:34bluebroth3rso if you had them running I'd been interested how you managed to build them :)
08:44:59 Join flydutch [0] (
08:45:15 Join petur [0] (~petur@rockbox/developer/petur)
08:48:25LloreanGodEater: If you view the "Albums" folder, you see a list of albums, which when viewed are then in track order
08:48:35GodEaterah right
08:48:39GodEaterI see what you mean no
08:48:41LloreanSo you'd expect that this hierarchy would pass on to the playlist if you inserted the level above
08:49:02JdGordondepends when that sort happens
08:49:06LloreanInstead what you get is whatever you'd see if you choose the "all tracks" entry (which I like being alphabetical)
08:49:08GodEaterI didn't even try inserting from there
08:49:28LloreanI think inserting "All Tracks" should get you alphabetical, but going up one and inserting the parent should get you a walked-through order
08:49:32GodEateryeah, and trying to sort the all tracks entry is what causes the data abort
08:49:53LloreanTo me at least, it makes the most sense / is the most expected result.
08:54:22 Quit rvvs89 (Ping timeout: 256 seconds)
08:55:46 Join rvvs89 [0] (
09:01:02pixelmaLlorean: if you insert from the level above (assumung the "album" entry in the database) wouldn't you get the content of <all tracks> and the individual albums?
09:01:05 Join Xerion [0] (
09:02:44LloreanI would expect it to currently either just give all tracks, or just walk folders, but not both. Even though explicitly it seems like it might, I think most people recognize "all tracks" as a more or less 'extra' or 'virtual' entry
09:03:16 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
09:03:29 Join giLL0r_ [0] (
09:05:19 Quit giLL0r (Read error: Operation timed out)
09:14:43 Quit rvvs89 (Ping timeout: 264 seconds)
09:14:51 Join rvvs89 [0] (
09:16:42***Saving seen data "./dancer.seen"
09:18:25 Join DerPapst [0] (
09:19:58 Quit rvvs89 (Ping timeout: 240 seconds)
09:20:42 Join LinusN [0] (~linus@rockbox/developer/LinusN)
09:32:06CIA-5New commit by 03funman (r25287): mkamsboot: add Clipv2 OF v2.01.35
09:32:19 Join funman [0] (~fun@rockbox/developer/funman)
09:32:33 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
09:32:36 Join funman_ [0] (~fun@rockbox/developer/funman)
09:32:54 Quit funman_ (Client Quit)
09:46:25 Quit JdGordon (Ping timeout: 252 seconds)
09:46:27pixelmaabout the next track tag description in the manual suggestion - I don't think %I'n' is not better than %Ix because (a) I'm used to "n" standing for an unknown number, not a character (b) if someone wants, he could read 'n' literally too (c) "x" is used for one of the allowed characters elsewhere too, e.g. in the general description about conditionals (%?xx<true|false>)
09:47:17pixelmamaybe * could work better - or setting "x" in italic?
09:48:35 Join PaulJam [0] (
09:48:47*GodEater strives to unravel pixelma's double negatives
09:49:04 Join JdGordon [0] (~7bf38c1f@gateway/web/freenode/x-lwnggqhgsrsgzqpn)
09:50:02pixelmaerr... I think the new suggestion is not better - or I don't think it's better :)
09:50:49JdGordonis sqlite really the smallest c database engine?
09:50:49CIA-5New commit by 03funman (r25288): Tag release v1.2 of mkamsboot
09:52:14funmanJdGordon: probably not. the smallest c sql database perhaps
09:52:14ThomasAHFlynDice, funman: One of the changes since enabling sound on the clip+ (don't know if yours or of another developer) causes some minor display problems: The "Shutting down" box no longer shows up (or so short that I can't see it) and when playing a song, then browsing through menus, then using "Now Playing" will only show a parial screen (status bar and seconds played/remaining and progress bar)
09:52:39funmanThomasAH: probably related to other commits
09:52:57JdGordonprobably my commit.. although I havnt seen this behaviour
09:53:15*ThomasAH checks if the simulator exposes the problem
09:55:07ThomasAHJdGordon: clip+ simulator can reproduce the problem
09:55:18ThomasAHJdGordon: (with rockbox_default playing screen)
09:56:35*funman sent mkamsboot 1.2 binaries to B4gder
09:56:51GodEaterI think a re-implementation of our tagcache system would make a nice GSOC project.
09:57:12JdGordonthe db part? or the frontend?
09:57:14JdGordonor both?
09:57:15GodEaterI'd like to see a proper design doc first - one which tries to capture the use cases first (i.e. what queries it would support)
09:57:26GodEaterthe db part
09:59:50ThomasAHJdGordon: reverting r25271 makes the problem vanish
10:00:40 Join Luca_S [0] (
10:00:46JdGordonclip+ sim doesnt have the mouse button locations :/
10:00:54Luca_Sthe main page
10:00:56Luca_Shas been defaced
10:01:03Luca_Sis there a site admin?
10:01:22GodEaterso it has :(
10:01:24ThomasAHJdGordon: right, and I don't know how to press the home button, too, but you don't need that to reproduce the problem
10:01:31JdGordonlucky us!
10:01:48Luca_Sinternal pages work correctly, it seems like it's just the home page
10:01:48GodEaterZagor / Bagder : you guys here?
10:01:55ZagorI am
10:02:30GodEaterdamn script kiddies
10:02:49Zagorwho, me? :)
10:02:58GodEaterno, the gits who defaced it
10:03:15Zagoroh crap, didn't see that
10:03:18 Join swilde [0] (
10:03:26GodEaterwonder how they got in this time
10:03:32pamauryGodEater: I'm not sure reimplementing the whole db makes sense. I've a better understanding of it now and I think we should simplify it. Perhaps drops some "features" that add code and seems not so useful
10:03:32GodEaterforensics hats on :)
10:03:34*Llorean wonders if that's his real email address.
10:03:46 Quit Luca_S (Client Quit)
10:03:47GodEaterpamaury: I respectfully disagree
10:04:02GodEaterit can't even managed basic tasks in it's current form - it needs work
10:04:08Galoisnot sure if it is a script kiddie, but if it is, there's no excuse for being vulnerable to canned attacks
10:04:22pamauryGodEater: which basic tasks ?
10:04:28GodEatersee earlier in the channel
10:04:31JdGordonwhy do we even allow building the db on the target?
10:04:43JdGordonand if we really want that, why not move it into a plugin?
10:04:48GodEaterJdGordon: to make rockbox platform independent
10:05:00JdGordonthe db format is target dependant IIRC
10:05:22*Llorean agrees with the "move it into a plugin" suggestion
10:05:31GodEateryes, but building it off target requires a computer based app
10:05:34LloreanI thought it used to at least be dependent on architecture endianness.
10:05:43JdGordonGodEater: rbutil
10:05:46GodEaterwhich means we have more code to support that needs to be supported across multiple targers
10:05:54GodEaterJdGordon: yes, which we already have trouble supporting on OSX
10:06:06GodEaterI disagree with moving db building into it
10:06:26pamauryrockbox is down ?
10:06:28JdGordonthats not technically true either... isnt the OSX issue a problem with <10.4 and a usb thing?
10:06:34GodEaterpamaury: keep up
10:06:40GodEateryou're not paying attention!
10:06:51pamauryah true :)
10:06:55JdGordonthe db code is entirely arch/target/everything independant wherever its coded
10:07:10JdGordonespecially if the metadata lib happens (how is that going by the way?)
10:07:29GodEaterJdGordon: I've no idea where the OSX problem is - the BIG problem with OSX is that we have barely any people with OSX in the first place, and even less with dev skills on it.
10:07:52GodEatermetadata lib is currently stalled whilst I have too much work on
10:08:12JdGordonyes, but thats a moot point here because the db building is just writing to files which Qt makes OS independant
10:08:31GodEateryes, IF you can get the util to build in the first palce
10:08:42GodEaterwhich seems a big if
10:08:53JdGordonbluebroth3r: yep, arm gcc crapped out
10:09:17 Part switchgirl
10:10:03pamauryThe db code is also a mess because, imo, it supports too many modes: there is normal mode, dircache mode, dircache+tcram mode and tcram mode iirc !
10:10:32JdGordonshouldnt there be atleast one abstraction layer in there somewhere? :)
10:10:40funmanGodEater: i have access to osx computers so i can perform some tests
10:10:41GodEaterpamaury: yes - so refactoring from the ground up would be a good idea imo
10:10:57Lloreanpamaury: And which of those modes should be removed, and why?
10:10:58GodEaterfunman: seriously, that's awesome - but you're one of so few
10:11:18pixelmaGodEater: which problem with RbUtil on MacOS are you referring too? Works nicely on 10.4 (both Intel and PPC)
10:11:31GodEaterpixelma: isn't the build for OSX out of date currently ?
10:11:45funmanGodEater: i'm using videolan servers, we could probably ask dionoea_ to create an account for rockboxers
10:12:09pixelmaalthough I haven't tried installing a bootloader, but installing builds/themes/creating voices works
10:12:42pixelmaGodEater: not more than the others and bluebroth3r builds for MacOS
10:13:16GodEaterpixelma: ok - I must be mistaken then, I thought it was lagging behind because of some build issue
10:13:45GodEateror some bug bluebroth3r couldn't work out with the latest version which only manifested on OSX
10:14:09pixelmaI believe the downloadable one from the wiki doesn't contain the latest voice fixes but that's because there was no new tagged release since
10:14:34GodEaterI realise he builds on OSX himself, my point was more that *if* he runs into issues with it, his audience of fellow OSX hackers to bounce ideas off is incredibly shallow compared to windows / *nix
10:15:05pixelmamaybe there is something about bootloader installation
10:15:25 Join perfectdrug [0] (
10:15:32JdGordonnot too many people owning macs care about rockbox anyway though so big deal
10:15:33pixelmayeah, at least I can help out testing currently
10:15:47pamauryLlorean: I don't know. I believe that with a proper implementation, only normal mode should be needed. The db is not accessed often enough to justify such modes imo
10:15:57pixelmathat's not a reason to have broken things
10:17:05Lloreanpamaury: Doesn't that make it significantly slower?
10:17:10pixelmaand by the way, I suggested a custom tagnavi creator (maybe standalone with possible integration into RbUtil) as gsoc project too last week
10:17:24LloreanMany users change or reorganize their playlists requeuing songs rather frequently
10:17:55*JdGordon just wants the db to be able to "resume" a query .. is that too much to ask?
10:18:17pamauryLlorean: yes but on many targets, the db is not loaded to ram by default so it's disk only. The truth is that the current implementation si crappy because the db is not always sorted and doesn't really take advantage of a sorted one iirc
10:18:17JdGordonsomething like /artist ~ A/
10:19:09pamaurywhat do you mean by "resume" a query ?
10:20:30JdGordonbe able to have some sort of path system like the file browser... /music/some_artist/ will always get you to the same place, why can't the db have something like that? "/artist starts_with A/Album/"
10:20:34*pamaury wonders what was the starting point of the whole discussion because he can't access the logs
10:20:46Lloreanpamaury: Why should it be loaded to ram by default? Shouldn't it be the users choice of performance vs memory wasted? And what do you mean by "not always sorted"?
10:22:20pamauryLlorean: I don't really understand everything :) See:
10:22:20pamaury#define TAGCACHE_SORTED_TAGS
10:22:53pamauryJdGordon: you would like that to replace the whole db requests ? Access the db like the filesystem ?
10:22:59GodEaterpamaury: the starting point of the discussion was Llorean and I talking about how it's not possible to play all albums from the db where each album is played in it's track order
10:23:06 Quit m3dlg (Quit: RAGE QUIT)
10:23:22GodEaterand I was saying how I'd looked into that a few months back, and then brought up the popular topic of re-implementing the db in sqlite
10:23:50GodEaterwhich, despite the topic's popularity, is not a popular solution because of the memory footprint of sqlite compared to tagcache
10:24:07GodEaterhence I'd like to see it re-implemented from the ground up so that it can do basic tasks like that
10:24:14JdGordonpamaury: yes and no. My beef is that a db playlist is kept as an APPEND for every track added (in .playlist_control) whereas a file playlist (dirplay) is a single line saying APPEND /some/directory/
10:24:45pamauryyes that's because the search needs a state
10:26:17pamauryJdGordon: and how would you fs idea work ? I mean, if you want to search for an artist you access /artist=.../ for example ?
10:26:18pixelmaGodEater: would it not be possibly to do what you want with a sort string that contains album and track name? Although I admit that this wouldn't make a well readable list
10:27:38JdGordonpamaury: I'm not necessairy suggesting a fs type thing, but being able to store/load a text represenatation of the query would be a HUGE benefit
10:28:40pamauryok, I see your point
10:29:15 Quit blairb (Quit: Leaving)
10:31:05pamauryThat's quite independent of the current tagcache implementation although. But then GodEater is right, if we were to reimplement tagcache, the first work is to list use cases, wanted features and then implement it.
10:31:50GodEaterpixelma: I tried that, it crashes Rockbox though, in a non-obvious way
10:31:57JdGordonI was going to say it sounds like GodEater's issue is with the frontend (tagtree.c) and yours is with the actual DB (tagcache.c) and both should be fixed and arnt necessarily related?
10:32:16GodEaterI tried to track down the data-abort, but it's in our string handling somewhere, so not obviously in tagcache itself
10:32:19pamaurydepends of whether we change the interface or not
10:32:50pamauryBut yes my issue is wit the actual DB because the code is impossible to maintain
10:33:04*GodEater summons Slasheri
10:33:46pamauryNot sure Slasheri would do it, it was written so long ago :)
10:33:47JdGordonsholdnt the db just have a few tables, one for each the strings, another for *all* the int/static sized items, and another to link them all together?
10:33:54JdGordonwhats so hard about that?
10:34:03pamaurybecause it's on the disk
10:34:13pamauryso there a zillions of errors to handle
10:34:21amiconnThere are some special details to handle as well
10:34:22GodEaterJdGordon: awesome - if it's so easy, do it yourself !
10:34:24pamauryactually half of code is error handling
10:34:37pamauryAnd there are dircache and tcram modes also
10:34:38JdGordondon't allow errors to get in and there is no problem :)
10:34:44GodEaterwth is tcram mode?
10:35:07amiconnE.g. some lists have to be sorted, but also need to be browsable in chunks, so they have to be pre-sorted on disk in a suitable way
10:35:43JdGordonthe actual data sholdnt be sorted..
10:36:22amiconnThe actual data needs to be sorted...
10:36:54amiconnHow else would you implement chunked browsing?
10:37:23JdGordonso an insert means moving potentially a massive amount of data?
10:37:30JdGordonsurely only a index is kept sorted?
10:38:24JdGordonwhat is chunked browsing?
10:39:19amiconnThe browser can only buffer max_files_in_dir entries at once for display, but some database lists are longer than that
10:39:30amiconn(namely the global "all tracks" list)
10:40:00amiconnSo scrolling through such lists swaps max_files_in_dir chunks as needed
10:41:51amiconnYou could probably do this with an index, but it would probably slow down chunk loading on hdd targets noticeably, because of seeking
10:41:53 Join mitk [0] (~mitk@
10:42:07 Quit mitk (Client Quit)
10:42:43 Join mitk [0] (~mitk@
10:44:17 Quit DerPapst (Ping timeout: 240 seconds)
10:45:00JdGordonmax_file_in_dir is a hell of a lot bigger than can be shown on screen quickly so as long as it doesnt block while its building that shouldnt be such a huge issue
10:45:35JdGordoni.e just get the first 2 screens worth while blocked then get the rest as needed in the background
10:46:12 Join DerPapst [0] (
10:46:38JdGordonbut yeah, I guess reading is a lot more common than updateing so the actual data being sorted would have benefits
10:46:49pixelmayou can see chunked browsing - there is a noticeable delay (I even noticed on my Ondio which is flash)
10:47:25pixelmajust scroll quickly through a huge list
10:48:01pixelma(dircache or not probably has an influence)
10:50:48Zagoraaaand we're back
10:51:22B4gderless than an hour since the hack
10:53:33JdGordonwere any other sites taken down?
10:53:59funmanother than www.rockbox ?
10:54:01B4gderwell, we run several sites and all of them were down at the same time
10:54:20JdGordonany others hacked I mean :)
10:54:28 Join DerPapst1 [0] (
10:54:32B4gderyes, the attacker put his files all over really
10:54:46B4gderhe found a way to upload a php script as the server user
10:54:51 Quit DerPapst (Ping timeout: 268 seconds)
10:54:54B4gderand then he uploaded more files with that
10:55:15B4gder41.232.54.102 is the guy
10:56:17 Join Luca_S [0] (
10:56:26pamauryI can't understand how someone could take time to do that, just destroying others work
10:56:50JdGordonthat was most probably a script kiddie.. not someonee with enough brains to know how to actually do it
10:56:55JdGordonso no time at all
10:57:17Zagorpamaury: my 2-year old likes to smash her older brother's lego cars
10:57:42Zagorsame thing here. same mental age too.
10:57:43pamaury when you're young that's different
10:58:21JdGordonZagor: hang on.. there is something wrong with breaking lego cars?
10:58:28JdGordonthats fun-ist!
10:58:39Zagorher brother doesn't always agree :)
11:00:16 Quit funman (Quit: free(random());)
11:00:22pamauryGodEater: tcram mode is the when db is loaded into ram
11:00:51JdGordonwhy does the db care about dircache?
11:01:07pamaurybecause there is little optimization for file names
11:01:21pamauryThat's really hacky
11:01:41JdGordondirectly talking to dircache instead of the proper file/folder handling?
11:01:57JdGordon.... :/
11:02:07CIA-5New commit by 03uchida (r25289): fix bug: WAV file playback does not resume (FS #11077) ...
11:02:27GodEaterI thought *everything* did that wrt dircache?
11:02:37GodEaterassuming it's enabled that is
11:02:51 Join JohannesSM64 [0] (
11:02:52JdGordonthrough an abstraction layer
11:03:22JdGordondir_open() (or whatever it is) does the magic to decide where to get filenames from (IIRC)
11:03:32JdGordonor I'd bloody assume that would be the case
11:04:16pamauryopendir uses opendir_cached when dircache is enabled and opendir_uncached if not or something like that
11:05:09pamauryI don't like the idea of tagcache messing up with dircache.
11:08:16 Quit flydutch (Quit: /* empty */)
11:08:39pamauryAnd anyway, the current code doesn't allow to add a single database entry easily, does it ?
11:09:20JdGordonsounds like you currently know the code better than anyone else.. so you tell us :p
11:09:45JdGordonare there tiny database engines aroudn that arnt sql based that would make sense thinking about using?
11:09:47pamauryI think not, but 4000 lines is too many for a single mind to remind of :)
11:11:53pamauryAnyone against committing my FAT patch ?
11:12:36gevaertspamaury: yes. It makes things smaller, so it can't be called a FAT patch!
11:12:51gevaertsApart from that, no
11:13:05 Join aholic [0] (
11:13:22*pamaury throws a FAT patch on gevaerts to crush him
11:15:20pamaurygevaerts: I'll commit it in a min, I changed some comments, I feel some were unclear
11:16:45***Saving seen data "./dancer.seen"
11:17:46pamaurypeople that know FAT spec: do you think the FAT code should check for checksums in LFN entries ?
11:22:54 Join rvvs89 [0] (~rvvs89@pdpc/supporter/base/rvvs89)
11:24:38 Join _silentAssassin [0] (
11:32:37ranmachanpamaury: I'd say it's unnecessary. I think the checksums protect agains vFAT unaware OSes (old DOS) messing up the directory.
11:32:52ranmachanAnd who is still using an OS that doesn't handle vFAT these days?
11:33:38 Join xsteadfastx [0] (~spectrum@
11:33:39ranmachanOr rather they are for detecting that someone not aware of vFAT has messed up the directory entries.
11:33:45xsteadfastxhey :)
11:33:55xsteadfastxhow can i say if my sansa fuze is v1 or v2?
11:35:17ranmachanOn my C200v2 is says v2 in very small letters on the back.
11:35:27CIA-5New commit by 03pamaury (r25290): FS #11118: reduce the number of cached sector in FAT code because some are useless after a rewrite of LFN entries handling. Also makes LFN handling ...
11:35:33 Join wodz [0] (
11:35:41JdGordonxsteadfastx: the OF version is the only way to know for sure
11:35:47Luca_Sxsteadfastx: the only way is to check the OF version
11:36:53xsteadfastxhow can i do that?
11:37:11JdGordongo into the sansa firmware and find the version in the info screen somewhere
11:37:27xsteadfastxoh :) is there something in the wiki about it?
11:37:28Luca_Sin the OF: Settings => System Settings => Information
11:37:43Luca_S(or something like that, I have it localized)
11:39:04xsteadfastxah thank you alot
11:39:50xsteadfastxcool..then i will try that when im at home
11:39:54xsteadfastxi even have a g5 ipod 30gb
11:40:05xsteadfastxi dont use it anymore...maybe perfect fpr rockbox
11:40:22xsteadfastxand fuze has a strange tag encoding sure it will get better with rockbox
11:41:33*pamaury : nice green delta !
11:41:59JdGordonnot too shabby
11:42:26JdGordonuchida strieks again...
11:43:53 Join Rob2222 [0] (
11:45:56pamauryJdGordon: what did uchida dod ?
11:47:11 Quit jordan` (Ping timeout: 256 seconds)
11:47:12CIA-5New commit by 03uchida (r25291): fix yellow
11:49:21*pixelma waits for the dust settling on pamaury's commit... being affected by the previous filesystem errors on directory delete bug ;)
11:50:53pixelmawhich he also fixed then though
11:51:36pamauryand that I also introduced first :)
11:52:51 Quit RadicalR (Quit: Nettalk6 -
11:55:03 Join funman [0] (~fun@rockbox/developer/funman)
11:56:02*pamaury just remembered that there is still a sync bug between dircache and file browser...
11:59:57wodzI have hit small but irritating bug. If I select System menu it automaticly enters to the first item in this menu (rockbox info).
12:00:11Luca_Sfunman: could your patch to enable sound on Clip+ work for Clipv2? I'm asking because on the TargetStatus wiki page it says 'only tested on clip+'...
12:00:15wodzdo You see something like this?
12:00:35funmanLuca_S: i have no clipv2 to test, but the code is enabled anyway, so you could tell us if clipv2 sounds works
12:01:14Luca_SI was just curious, unluckily I only own a FuzeV2
12:01:33gevaertswodz: I've never seen that. A bug in your button driver maybe?
12:02:38 Join einhirn [0] (
12:03:47wodzI get this on ipod mini 1G and in my experimental port to HD200. But interesting thing is that I also get this in sim
12:04:11linuxstbMaybe a problem with your action definitions?
12:05:23PaulJamhi, i was trying to download an older daily build, but after selecting the target the table at the top of the page ( ) is empty. is this just a temporary issue or have you removed the daily builds?
12:06:20 Nick fidencio[AWAY] is now known as fidencio (
12:08:06 Join kugel [0] (~kugel@rockbox/developer/kugel)
12:08:14 Quit hd (Ping timeout: 258 seconds)
12:08:18 Nick fxb__ is now known as fxb (
12:08:59 Join mischasworld [0] (~quassel@
12:09:28JdGordonPaulJam: was hacked before.. maybe the files got wiped
12:09:59gevaertsno, they've been gone for at least a few days
12:17:03funmanhm i can't reboot my clip+ to rockbox when it panics (black screen), but rebooting to OF works (when holding the dualboot button)
12:17:39 Quit kugel (Remote host closed the connection)
12:18:24ThomasAHfunman: I can't confirm this ... cube crashed once yesterday with black screen, booting directly into Rockbox worked after >10seconds-poweroff
12:18:54funmanThomasAH: when in the panic screen (*PANIC*), you can reboot immediately with any keypres
12:19:16ThomasAHfunman: it was completely black screen for me, no *PANIC*
12:19:17PaulJamJdGordon: btw, since your commit yesterday there seems to be a problem with the quickscreen. the screen flashes shortly and the the screen stays blank. after changing a setting it shows up again.
12:19:52JdGordonI've been trying to track these issues down... no idea how that change did it
12:20:02JdGordonI'll revert it in the next few days if I cant fix it
12:20:37JdGordonit apparently is a problem with a setting.... no issues on my mini2g
12:24:40Luca_Son the forums, pxby says sound works on the ClipV2 too.. dunno if it's worth a gentlemen mail, but congrats again funman :D
12:28:40JdGordonso here is what *must* be the problem.... before that change every line in the default viewport in the sbs was run through *every single update loop*, now it only happens if the line has "dynamic" content
12:29:46JdGordonwhich means the current behaviour is more correct, but a full update appears to be missing somewhere
12:29:49JdGordonI think
12:30:30JdGordonPaulJam: can you change skin_display.c line 1180 to make that "if (1)" and see if that fixes things?
12:37:01 Join kugel [0] (~kugel@rockbox/developer/kugel)
12:38:53 Join elinenbe [0] (
12:40:57PaulJamJdGordon: no, that didn't fix it.
12:41:07JdGordonok, good :)
12:41:24JdGordonalthough that means i have no idea how that commit broke anything
12:44:16 Join DV [0] (~DV@
12:44:46 Join mischa [0] (~quassel@
12:44:53PaulJambtw, the remote display (in the uisim) works fine, this only happens in the main screen.
12:45:39 Quit mischasworld (Ping timeout: 260 seconds)
12:47:50kugelnothing was updated in the sbs' default vp, only the get_line call happened to evaluate conditionals
12:49:20 Quit mischa (Ping timeout: 252 seconds)
12:50:22 Join mischasworld [0] (~quassel@
12:52:14kugelit was obvious (to me) that something like this happens which was the reason I did it how it was and declared drawing in the default vp/%Vi is unsupported
12:52:50funmanFlynDice: ping
13:00:29JdGordonkugel: well take this any way you want.. thats just stupid
13:00:41xsteadfastxis there a player i can buy now and where i can use rockbox on? i mean im sure i dont get something like the fuze v1 no more
13:00:53JdGordonand if that was the logic then it shold have been handled explicitly
13:01:18 Join froggyman [0] (
13:01:18 Quit antil33t (Read error: Connection reset by peer)
13:01:26 Join antil33t [0] (
13:02:25wodzdo the rockbox ata driver support flush_cache command? I cant figure that out
13:02:54Luca_Sxsteadfastx: - according to that table, nothing supported is in production anymore, so I suggest some of the most recent sansas (except fuzev2), work on them is progressing quite nicely
13:04:11xsteadfastxLuca_S, something like the clip?
13:05:00JdGordonwhat was actually happening was that the default viewport AND the ui viewports were having all their lines parsed every single update loop when they shouldnt be
13:05:12Luca_SIt really depends on your needs. Have a look at the wiki pages, all that I could tell you is there.
13:05:38 Join perfectdrug_ [0] (
13:06:15kugelJdGordon: I don't understand
13:06:21 Join mischa [0] (~quassel@
13:06:29 Quit mischasworld (Read error: Connection reset by peer)
13:06:37JdGordonin the displayer.. I cant tinhk of a better word
13:07:21kugelI changed nothing (because I don't know the side effects of get_line, I only figured it's needed for conditionals), I only added the exception for hidden viewports so that drawing doesn't happen in them
13:07:46JdGordonexcept you didnt
13:07:49kugelI don't think I added anything to it which didn't already happen before
13:08:11kugelplease explain that for me
13:08:11JdGordonyes and no
13:08:52 Quit perfectdrug (Read error: Operation timed out)
13:09:05JdGordonget_line() does some drawing without checking the update mode, so if those tags are in the sbs' default viewport then they would be drawn
13:09:38JdGordonmy commit didnt fix that but made it more obvious. All your excpetion really wanted was to make sure the default viewport isnt cleared every update
13:09:51JdGordonI dont understand how that change breaks other screens though
13:10:07JdGordonPaulJam: got a min for another quick maybe fix?
13:11:04JdGordonadd "!(hidden_flags&VP_NEVER_CLEAR) &&" to the end of line 1227 after the &&'s in skin_display.c please
13:11:07 Part froggyman
13:12:11CIA-5New commit by 03funman (r25292): sd-as3525v2: read cid & csd in the correct order, fix disk info
13:16:48***Saving seen data "./dancer.seen"
13:17:30funmanFlynDice: i got wait for state errors (waiting for TRAN, still in PRG - so i suppose after STOP_TRANSMISSION cmd - ), when trying to print status bits from the main loop while (retry || count), not even trying to write at this point (just inserting the µSD)
13:18:50 Quit mischa (Quit: No Ping reply in 180 seconds.)
13:19:05 Join mischasworld [0] (~quassel@
13:20:00PaulJamJdGordon: that didn't fix it either.
13:20:10JdGordonok thanks
13:23:04 Join adnyxo [0] (
13:24:14kugelJdGordon: the conditionals logic and the drawing should probably more separated to have more control on what's being drawn and what not
13:24:34JdGordonyes, any idea how to do that though?
13:25:03JdGordonwe can make it only do conditional tags, but nothing forces them to not draw in those
13:25:15kugelif get_line draws, it's bad IMO, it would be better if there was a drawing function (maybe per skin viewport) which draws depending on the get_line result
13:25:17JdGordonespecially if we move the bmp drawing directly into the tag handling (which i want to do)
13:25:25kugelno, I have no idea how to do it right now
13:25:37JdGordonI dont really like the big drawing at the end
13:26:37JdGordonanother way it could have been done was just get rid of the hidden_vp var and use the vp_redraw_mode to figure out if we can draw or not
13:26:46xsteadfastxoh the clip+ looks nice..cant wait for rockbox support
13:26:52kugelI think I tried that without much success
13:27:12 Join mischa [0] (~quassel@
13:27:21JdGordonI want to have a fiddle with how the hiding viewports work.. I dont know if you saw, I want to allow the ui viewport to be conditioanlly chosen
13:29:19*kugel doesn't see the point of that, but well
13:29:59 Quit mischasworld (Ping timeout: 245 seconds)
13:30:11 Join froggymana [0] (
13:33:16 Join anewuser [0] (anewuser@unaffiliated/anewuser)
13:35:20 Join elinenbe_ [0] (
13:35:20 Quit elinenbe (Read error: Connection reset by peer)
13:35:20 Nick elinenbe_ is now known as elinenbe (
13:36:48elinenbeFlynDice: no dice on writing to the storage on the Clip+ yet?
13:37:16funmanelinenbe: when it'll be done, status will be updated on the wiki
13:40:41 Quit JdGordon (Ping timeout: 252 seconds)
13:41:35 Quit shaggy-h (Ping timeout: 240 seconds)
13:42:35 Quit elinenbe (Read error: Connection reset by peer)
13:42:44 Join elinenbe [0] (
13:49:27wodzdoes sleep() work with disabled interrupts?
13:49:52funmansleep() rely on the tick interrupt
13:50:30funmanPP have a µseconds counter which can be used as a usleep() when interrupts are disabled
13:51:14wodzcoldfire has also second harware counter. But I can not figure out how to use it with current implementation
13:52:10wodzam I correct that sleep() does not guarantee precise delay?
13:53:17funmanafaik sleep() just gives control to the next thread until the delay has elapsed so it could be longer
13:53:40wodzok thats how I understood this
13:54:49 Quit funman (Quit: free(random());)
13:56:05 Quit JohannesSM64 (Ping timeout: 240 seconds)
14:00:07 Quit chrisb (Remote host closed the connection)
14:10:49 Quit mischa (Ping timeout: 245 seconds)
14:10:58 Join fyrestorm [0] (
14:11:39 Quit rvvs89 (Ping timeout: 245 seconds)
14:11:47 Quit DV (Ping timeout: 276 seconds)
14:13:03 Join hd [0] (
14:13:03 Quit hd (Changing host)
14:13:03 Join hd [0] (~jd@Wikipedia/HellDragon)
14:13:26 Join DV [0] (~DV@
14:13:51 Join rvvs89 [0] (
14:17:29pamaurygevaerts: I'm trying to implement the new usb api I mentionned to you and found something nasty. I'm currently implementing queueing because it's necessary for most endpoints. So I set up a queue for EP_CONTROL in both direction of size 1 and of course everything worked well for send because usb_drv_send is blocking (so the queue was never full on request) but on recv, I encountered a situation where usb_drv_recv was called whereras the previ
14:17:29pamauryous one had not finished ! My code refused the transfer because the queue was full but on the current code, there is a potential overwrite (even though is was ack so not problematic)...
14:19:49gevaertspamaury: is that one queue for both directions, or one queue per direction?
14:19:57pamauryone queue for each
14:20:10pamauryso two queue for EP_CONTROL, one send, one recv
14:20:18pamauryof size 1 currently
14:20:25gevaertshm, that shouldn't happen as far as I can see...
14:21:00 Quit Rob2222 (Remote host closed the connection)
14:21:08pamaurythat's sound strange for me also but I'm telling you what I see :)
14:21:15 Join Rob2222 [0] (
14:22:39gevaertsI'll have a look tonight unless you've found it before
14:22:46pamauryI think it's due to the implementation of usb_drv_int: it check for a control received before a transfer completion. so if the transfer completion arises and a setup, the setup is treated and the queue is not emptied in my code
14:23:11pamaurySo it might be specific to my code then
14:23:37 Quit elinenbe (Read error: No route to host)
14:23:57 Join elinenbe [0] (
14:26:17xsteadfastxthe sansa e250 should work cause its in the e200 series?
14:27:22Luca_Sxsteadfastx: yes
14:27:37xsteadfastxthats cool...cause you still can get them :)
14:28:12pamaurygevaerts: but to be honest, there is something I don't like in usb_drv_recv. Because usb_core assumes that on a control transfer, usb_drv_recv will fill the buffer immediately. So it's assumed that it's sort of blocking whereras it is not but it's working because the data is already there... I am wrong ?
14:30:40gevaertspamaury: I don't know
14:37:10 Quit GHF (Ping timeout: 260 seconds)
14:38:26 Quit Tomis (Quit: Tomis)
14:40:10 Join Tomis [0] (~Tomis@
14:41:54 Nick fidencio is now known as fidencio[AWAY] (
14:41:56 Quit Llorean (Ping timeout: 246 seconds)
14:49:30 Quit elinenbe (Read error: Connection reset by peer)
14:49:50 Join elinenbe [0] (
14:50:03 Join byondo [0] (
14:50:09 Join evilnick_B [0] (~0c140464@rockbox/staff/evilnick)
14:51:35byondohello everyone
14:52:51 Join Schmogel [0] (
14:53:23byondoanyone to help me w/ clipv2?
14:54:42 Quit froggymana (Quit: CGI:IRC)
14:56:33evilnick_Bbyondo: Ask the question you want an answer to and if anyone knows then they'll help
15:01:51 Quit Unhelpful (Quit: No Ping reply in 180 seconds.)
15:01:51 Quit byondo (Quit: CGI:IRC (EOF))
15:01:52 Join Unhelpful [0] (
15:01:52 Quit Unhelpful (Changing host)
15:01:52 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
15:02:03 Part xsteadfastx
15:02:19 Quit AlexP (Remote host closed the connection)
15:02:36 Join AlexP [0] (~ap@rockbox/staff/AlexP)
15:02:58 Join maxc1006 [0] (
15:03:11maxc1006here I am once again
15:03:28 Nick maxc1006 is now known as byondo (
15:03:56byondoswitched to mirc client, web client was sooo lagged
15:04:07 Quit linuxguy3 (Ping timeout: 276 seconds)
15:04:16 Join elinenbe_ [0] (
15:04:29byondoso I want to test rockbox on my clipv2
15:04:32 Join linuxguy3 [0] (
15:04:34 Quit elinenbe (Read error: No route to host)
15:04:34 Nick elinenbe_ is now known as elinenbe (
15:05:08 Join JohannesSM64 [0] (
15:06:49byondoI have OF, mkamsboot, daily clipv2 compiled RB, I need bootloader...?
15:07:55Zagorisn't both clip v2 and clip+ ready to be tagged "unstable"?
15:08:16kugelwithout write support?
15:08:37byondoyes, I know
15:09:21byondojust want to try, but if you think I could birck it, I'll wait :)
15:09:23Zagorkugel: I'd say it depends. if it works ok as a read-only player, I'd say it should be "unstable"
15:11:09kugelI wouldn't mind :)
15:11:51Zagorand again i think "Complete" and "Incomplete" are better names for the tier 1&2 ports
15:14:13kugelexcept complete doesn't apply to all stable targets
15:14:18S_a_i_n_tHmmmmm S_a_i_n_t does indeed dislike the wording used in FS #11101. Do you think "USB: Charge By Default" is any better than "USB Charging By Default?"
15:14:29S_a_i_n_tits really the "?" in the menu I dislike.
15:14:37Zagorkugel: such as?
15:14:57byondoI found compiled RB for clipv2 in the daily folder
15:15:15TorneS_a_i_n_t: It's nothing to do with charging, though, no?
15:15:21TorneThat sounds *really* confusing
15:15:25 Join Linus [0] (~linus@gateway/web/cgi-irc/
15:15:35TorneIt charges (or not) according ot the value of a completely unrelated setting.
15:15:45S_a_i_n_tWell, it *is* kinda to do with charging
15:16:04S_a_i_n_tAll it does though is stop you from having to press menu to charge with the player on
15:16:09S_a_i_n_t*using the RB FW
15:16:10TorneYes, but it *charges either way*
15:16:17Torneunless you have disabled usb charging with the setting for that ;0
15:16:25TorneIt's enabling/disabling USB *data* connection
15:16:30Tornethis is nothing to do with charging
15:16:51***Saving seen data "./dancer.seen"
15:17:26 Join hebz0rl [0] (
15:17:35 Join giLL0r [0] (
15:17:38S_a_i_n_tHmmm, yes. I see your point.But "Use this setting so you don;t have to long press a key to keep the player on" is a little long ;P
15:18:14 Quit einhirn (Read error: Connection reset by peer)
15:18:15S_a_i_n_tI *really* can't come up with a wording that a: looks good, and b: fits the bill.
15:18:18 Join flydutch [0] (
15:18:57S_a_i_n_t"looks good" isn;t nearly as important, but I seem to dislike the "?" in the Menu entry for some reason...and see it as a "no no"
15:18:59 Join krazykit` [0] (
15:20:30S_a_i_n_tThe Authors second suggestion(s) on his FS# are just as bad (IMO) and I'd hate to see bad wording keep this from getting comitted.
15:20:56S_a_i_n_tAs its *kinda* usefull, the GF had been bugging me about it for some time now.
15:21:08TorneDefault behaviour on USB connection: Data transfer | Charging only
15:21:41 Quit DerPapst1 (Read error: Connection reset by peer)
15:21:47TorneOr just "Default USB connection action"?
15:21:49TorneI dunno.
15:21:52S_a_i_n_tthat line doesn't fit by a LONG shot though..well of course it will scroll, but I like to try and keep menu entries as short as possible
15:21:58 Join TopyMobile [0] (
15:22:01S_a_i_n_tThat is the best one I've seen so far though
15:22:02 Quit moos (Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115144158])
15:22:13TorneDefault USB mode, even?
15:22:36Zagorthat sounds like I'd choose between UMS and MTP
15:22:41S_a_i_n_tYeah, that works...but needs a COMPLETE rewrite of the the options are yes/no
15:22:48S_a_i_n_tso, that would make it more confusing.
15:22:56TorneS_a_i_n_t: Boolean settings with custom on/off phrasing are not hard
15:22:59 Join DerPapst [0] (
15:23:00TorneThere's a macro for it even
15:23:14S_a_i_n_tnot hard for you perhaps :P
15:23:21TorneZagor: Maybe, but so? If we added MTP in the future then that'd be a perfectly good place to put it as a third choice
15:23:29TorneZagor: i think that fits fine
15:23:51Zagoryeah, that makes sense
15:23:57 Quit _silentAssassin (*.net *.split)
15:23:57 Quit giLL0r_ (*.net *.split)
15:23:58 Quit krazykit (*.net *.split)
15:23:58 Quit xavieran (*.net *.split)
15:23:58 Quit Topy44 (*.net *.split)
15:23:59 Quit LinusN (*.net *.split)
15:24:10S_a_i_n_tYes, I actually quite like that suggestion (Torne's)..but implementing it (without guidance) is a little beyond me.
15:24:27TorneZagor: maybe something other than "Data transfer", then?
15:24:40TorneDefault USB mode: Storage device, charging only
15:24:57S_a_i_n_tThat's the best yet. IMO
15:25:00TorneI would think "storage device" implies MSC to people who know what that is, and still means somethign to people who don't
15:25:07Zagoranything wrong with "USB Mass Storage"?
15:25:13TorneZagor: I guess not
15:25:23S_a_i_n_tits longer? :P
15:25:26TorneWell, just mass storage
15:25:28Tornethe option already says usb
15:26:11Zagorin that case I think I'd rather opt for "Storage device". "Mass storage" sounds a bit weird alone. :)
15:26:23ZagorI want my bike shed purlple!
15:26:46S_a_i_n_tyes, but keep in mind it'l be prefixed with USB*something*: probably
15:26:53TorneZagor: yeah we are bikeshedding a bit
15:27:09TorneS_a_i_n_t: anyway i would fix up the patch right now but i'm supposed to be working
15:27:20TorneS_a_i_n_t: so either someone else can do it or you can poke me in six hours
15:27:25 Join FlynDice_ [0] (~FlynDice@
15:27:38pixelmaZagor: about "complete" - e.g. charging on the Iaudios
15:27:50S_a_i_n_tIs there some example you know of I could glean the info I need from in the source?
15:28:51Tornelook for BOOL_SETTING
15:29:05S_a_i_n_tthanks rather
15:29:08Torneas opposed to ONOFF_SETTING
15:29:34TorneBOOL has free choice of lang string and string to use in the .cfg
15:29:50Zagorpixelma: is that the m3 or all the iaudios?
15:31:34 Join xavieran [0] (
15:31:49Zagorand that doesn't make me like the "unstable" name any more. it's simply misleading.
15:32:47 Join _silentAssassin [0] (
15:33:59*gevaerts isn't always very happy about the preexisting release ports automatically getting the "stable" label
15:35:48Zagorright, "stable" is also misleading when we screw things up.
15:36:05Zagorit might even have been me suggesting the names from debian, but I still don't like them now.
15:36:45Zagorour release process isn't anywhere near debian's
15:36:55 Quit DV (Read error: Connection reset by peer)
15:37:06Zagorthe only thing the tiers say is how much code has been written for each model.
15:37:07 Join DV [0] (~DV@
15:40:14FlynDice_funman: (logs) re: (waiting for TRAN, still in PRG) I haven't seen this yet but I think it's a good sign. If we know it's getting into PRG mode then I think that means the card is actually getting some data into it's buffers.
15:40:31 Quit DV (Read error: Connection reset by peer)
15:40:50byondohello FlynDice
15:41:18 Quit DerPapst (Quit: Leaving.)
15:41:28byondoare you a sansaAMS expert, right? :)
15:41:53FlynDice_expert is stretching it quite a bit but I know a little
15:42:11byondoI'm trying to install the daily RB to my clipv2
15:42:37 Join DV [0] (~DV@
15:42:47byondoI have to do it with win32...
15:43:28byondoand mkamsboot is not compiled with latest "original firmware adding"
15:43:47byondoisn't it?
15:44:16byondolatest OF: 2.01.35
15:45:25FlynDice_I think I heard some thing about that. If that firmware is a recent update I would just try using mkamsboot with an older firmware that mkamsboot recognizes.
15:46:10byondook, so I will have a clip with dual firmware, right?
15:46:35byondothe original one patched w/ mkamsboot and RB?
15:46:44Luca_Sbyondo: did you try to use the latest SVN mkamsboot? just today funman committed the checksum for version 2.01.35
15:47:19Luca_Syou can still see the commit message on the home page
15:47:25 Quit DV (Read error: Connection reset by peer)
15:47:51byondoLuca_S, I see it, but it seems there's no win32 compiled mkamsboot w/ that commit
15:48:17wodzI am preparing initial patch to add new target. How should I treat binary file in the patch? (UI graphic for simulator)
15:48:36mitkFlynDice_: About 4 hours ago funman sent compiled binaries of mkamsboot v1.2 to Zagor or Badger. Don't remember to which one..
15:48:41 Join DV [0] (~DV@
15:48:44wodzsvn diff ignores binary files
15:49:21S_a_i_n_twodz: can't you add them with svn add? Or does it just ignore them completely?
15:49:33TorneYou'll need to just post the binary file seperately along with the patch
15:50:25 Quit anewuser ()
15:51:01FlynDice_mitk: Which ata error did you see with you 8GB clip+?
15:52:52mitkFlynDice: It was just question on the forum, not bug report. I can tell you that in 3 hours if you like. No clipp around.
15:53:56wodzhow verbose should be comment to the FS task with initial port patch?
15:54:55 Quit elinenbe (Read error: No route to host)
15:55:08 Quit DV (Ping timeout: 276 seconds)
15:55:18byondoso is there any risk to brick clipv2 with this r25292?
15:55:28 Join elinenbe [0] (
15:55:31FlynDice_mitk: Post it on the forum or in the logs here, yes I would like to know so I can see where it's failing. That info is related to devolopement.
15:55:44B4gderbyondo: the sansas are never risk-free
15:55:59gevaertsthe AMS sansas anyway
15:56:03mitkI will then
15:56:33FlynDice_byondo: there is always a risk but there have been no bricks reported with the clipv2, clip+ had at least 2 so far early on
15:57:31ThomasAHFlynDice_: With 3-4 hours of 8GB clip+ with 8GB SDHC usage during the weekend I did not have a single ATA error. Most songs came from the internal memory, but I played some from the card, too.
15:57:59byondoFlynDice: ok, so I wait for the new compiled mkamsboot, or use an old OF firmware?
15:58:05FlynDice_ThomasAH: Great, thanks for the info!
15:58:40 Quit mitk (Quit: Leaving)
15:59:05ThomasAHFlynDice_: SD access icon works, too, since your changes this morning
15:59:31FlynDice_byondo: Yes, to me those seem like those are your choices
16:01:05FlynDice_ThomasAH: Your 8GB is playing fine today since my changes last night?
16:01:35byondoI'm impatient, I will try w/ old OF :)
16:01:47wodzplease look at FS #11137
16:02:23kugelZagor: USB
16:05:00ThomasAHFlynDice_: Only checked few minutes ... should I try longer?
16:05:48 Join elinenbe_ [0] (
16:05:53 Quit elinenbe (Read error: No route to host)
16:05:54 Nick elinenbe_ is now known as elinenbe (
16:06:04pixelmaZagor: all
16:06:10S_a_i_n_tCan somebody that knows their way around the code please look at FS5111 and tell me how I managed to bork the piezo for iPod nano1g (and probably all the other targets except the nano2g which TheSeven adden support for afterward)?
16:06:14kugeland charging too
16:06:26S_a_i_n_tI can't see what I've done wrong :(
16:06:37pixelmaZagor: no real charging under Rockbox, M3 is worst though
16:07:04FlynDice_ThomasAH: No, that's fine, mitk was getting an ATA error on startup and I was wondering if it was related to his player being 8GB. If yours starts up fine and plays it must be another reason.
16:07:47byondoFlynDice: ah,err... "Model name "clv2" not found in bootloader-clip.sansa" :(
16:08:55linuxstbbyondo: Did you compile a "clip" (v1) bootloader?
16:09:06FlynDice_byondo: Did you do a bootloader build?
16:09:28byondolinuxstb & FlynDice: nope :)
16:09:30ThomasAHFlynDice_: just tested, works without inserted SDHC card, too :)
16:09:41FlynDice_ThomasAH: Thanks
16:09:57byondojust downloaded from rockbox site
16:10:15byondoeffectively it's outdated...
16:10:28 Join Farthen [0] (
16:10:31linuxstbbyondo: You downloaded a clipv1 bootloader.
16:10:43byondoyes, for sure...
16:10:55byondothink I have to wait
16:11:16S_a_i_n_tpixelma: I see your point about my suggestion for the Next Track tags in the manual...I guess my problem with them is that there is nothing that suggests that x is a variable
16:11:45S_a_i_n_tYour suggestion of italics, and or enclosing it in ' 's is a lot better.
16:13:38 Join DV [0] (~DV@
16:14:04pixelmaI thought the ' ' was yours but maybe you just added them here to point things out
16:15:17byondostrange, mkamsboot has clipv2 compatibility but there's no bootloader for clipv2 to use with...?
16:16:12kugelwe don't have released bootloaders for the newer ams sansas
16:16:20kugelyou need to build from svn
16:16:39wodzRegarding new port, I added initial patch to FS. Should I do something else or just sit and wait for reviews?
16:17:12 Join DerPapst [0] (
16:18:26byondokugel: ah ok thanks, so I just have to wait
16:18:30linuxstbbyondo: clipv2 is still in development. Once it has all the things a user needs to install, it will be listed in "unstable" status.
16:18:51 Quit DV (Ping timeout: 252 seconds)
16:19:07B4gderwodz: you can also try asking for feedback here or on the dev mailing list
16:19:56wodzok I'll post to dev ml also than
16:21:40byondolinuxstb: yes, I was just too happy from reading latest posts on the forum, I forgot it ;)
16:23:13 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
16:24:30byondogoodbye all and thanks for the info's :)
16:24:49 Quit byondo ()
16:30:02 Quit elinenbe (Read error: Connection reset by peer)
16:30:03 Join elinenbe_ [0] (
16:31:30 Quit kugel (Ping timeout: 252 seconds)
16:31:35 Join Topy44 [0] (
16:34:01linuxstbwodz: Do you need your own bootloader file? i.e. could you share an existing Coldfire bootloader?
16:34:08 Quit Adubb (Read error: Connection reset by peer)
16:34:13linuxstbwodz: There's also an unrelated whitespace change in bootloader/common.c
16:34:15 Join Adubb [0] (~aldubuc@
16:35:49linuxstbwodz: (I guess I should be posting to the tracker....) Maybe "mpio" isn't the best choice for model string, assuming there are other "mpio" targets apart from HD200.
16:37:00wodzlinuxstb: it has many similarities with iriver h300 bootloader but I think it is cleaner to have separate bootloader than nightmare of ifdefs
16:37:53wodzlinuxstb: I am aware that "mpio" is not perfect but 4 chars do not allow me to put hd200 as a model string
16:39:27linuxstbMaybe just hd20?
16:41:01wodzIf You think it is ok to truncate the name I can use this
16:42:29LloreanYou could go the old school route, HDCC. Roman numerals can manage 200 in 2 digits. :-P
16:43:30wodzLlorean: cool but I am afraid this may be missleading to someone reading sources in the future
16:43:45pixelmahdcc ;)
16:43:56B4gderc8 in hex!
16:43:56*pixelma was too slow
16:44:28LloreanHD20 is probably 'good enough' anyway.
16:44:48 Join captainkewllllll [0] (~2669ecc2@gateway/web/freenode/x-xrndmlrlohjwwuon)
16:44:49wodzbtw: why there is 4 chars limit?
16:46:53 Quit wodz (Quit: Leaving)
16:48:12linuxstbwodz (for the logs): Simply because it's always been a four-char value, since it was invented to distinguish between the h100, h120 and h300...
16:48:12 Quit elinenbe_ (Read error: Connection reset by peer)
16:48:22 Join elinenbe [0] (
16:49:04 Join Blue_Dude [0] (
16:51:25Blue_DudeA working patch for an assignable hotkey is available at FS #11081. It's transparent if the target doesn't have HOTKEY defined in its config file. So far its functions are limited but the framework is in place and it can be easliy expanded.
16:53:13 Quit DerPapst (Quit: Leaving.)
16:56:41saratogawow a new coldfire target :)
16:56:47 Quit saratoga (Quit: Page closed)
16:57:45 Join m3dlg [0] (~m3dlg@
16:59:49 Quit petur (Read error: Connection reset by peer)
17:00:02 Join petur [0] (~petur@rockbox/developer/petur)
17:00:33 Quit petur (Read error: Connection reset by peer)
17:01:07 Join petur [0] (~petur@rockbox/developer/petur)
17:02:33 Quit m3dlg (Ping timeout: 240 seconds)
17:03:23 Join m3dlg [0] (~m3dlg@
17:04:29 Quit Blue_Dude (Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115144158])
17:04:39LloreanBlue_Dude: You are aware that your patch is a configurable button, which has traditionally been a no-do.
17:05:46 Quit rvvs89 (Ping timeout: 246 seconds)
17:07:59 Join rvvs89 [0] (
17:09:00 Join komputes [0] (~komputes@ubuntu/member/komputes)
17:10:33 Quit elinenbe (Read error: Connection reset by peer)
17:10:49 Join elinenbe [0] (
17:11:30 Quit m3dlg (Ping timeout: 276 seconds)
17:13:48 Quit FlynDice_ (Remote host closed the connection)
17:14:22 Join merbzt1 [0] (~benlar@
17:16:52***Saving seen data "./dancer.seen"
17:17:40 Quit merbzt (Ping timeout: 276 seconds)
17:18:15 Quit captainkewllllll (Quit: Page closed)
17:19:38 Quit elinenbe (Read error: Connection reset by peer)
17:19:39 Join elinenbe_ [0] (
17:23:43 Nick fidencio[AWAY] is now known as fidencio (
17:24:38 Join kaniini [0] (
17:24:51*linuxstb has no problem with customising the action of otherwise unused buttons
17:28:09 Quit elinenbe_ (Read error: No route to host)
17:28:12 Join elinenbe [0] (
17:28:39*S_a_i_n_t spots "CPU: arm ; Manufacturer: s5l8700" in rockbox-info for the Nano2g. Shouldn't it be "CPU: s5l8700 ; Manufacturer: arm"?
17:31:21CIA-5New commit by 03funman (r25293): Clip+: arbitrary combo (home+select) for button lock
17:34:57 Quit fejfighter (Ping timeout: 260 seconds)
17:35:58 Join kugel [0] (~kugel@rockbox/developer/kugel)
17:38:19pixelmaarm is definitely not the manufacturer, it's the architecture saying that though s5l8700 isn't the manufacturer either though...
17:41:56 Nick fxb is now known as fxb__ (
17:42:14 Quit B4gder (Quit: It is time to say moo)
17:42:15 Quit PaulJam (Ping timeout: 265 seconds)
17:42:34linuxstbS_a_i_n_t: That's simply what the levels in the target tree system are (inaccurately) called. That's probably nonsense for a lot of targets.
17:42:40kugelS_a_i_n_t: that's old, the new prefered interpretation is soc
17:43:13S_a_i_n_tIf its old...why did it *just* compile that way?
17:43:43kugelbecause nobody changed the code yet
17:44:35kugelsee point 3 under core
17:50:14 Quit petur (Quit: work->home)
17:52:47Lloreanlinuxstb: Wasn't there discussion of making it an insert or queue button for some time?
17:54:55linuxstbLlorean: I've no idea. That could of course be one of the customisation options...
17:56:07 Quit pamaury (Quit: Quitte)
17:58:34 Join liar [0] (
18:02:43Schmogeli'm following those instructions: and i get this warning at step 3: "The compiler you must use (arm-elf-gcc) is not in your path!" i'm trying to compile for the sansa clip+
18:02:57Schmogelwhat am i doing wrong?
18:03:28LloreanIs arm-elf-gcc installed and in your path?
18:04:24Schmogelno, i dont know where i can find it
18:04:50 Part Linus
18:05:18linuxstbSchmogel: Try one of the two links at the top of that page
18:05:30*Llorean is surprised the CrossCompiler link doesn't seem to be on that page
18:05:48linuxstbLlorean: It's a wiki ;)
18:06:53 Quit FOAD (Quit: I'll be back)
18:08:04 Join MethoS- [0] (~clemens@
18:08:37*Llorean adds a bit.
18:13:00 Quit antil33t (Read error: Connection reset by peer)
18:13:06 Join antil33t [0] (
18:13:08 Join Chrolo [0] (~51e05113@gateway/web/freenode/x-wugdbwbgoiyzocvk)
18:13:58 Join TheSeven [0] (~theseven@rockbox/developer/TheSeven)
18:15:59 Quit Luca_S (Quit: CGI:IRC)
18:16:21 Join CGL [0] (~CGL@
18:19:10 Join FOAD [0] (
18:19:41Battousaiplayback is supposed to work on clip+ now right?
18:26:41linuxstbI believe so.
18:27:35CIA-5New commit by 03alex (r25294): Fix c&p error for %pE in the manual. Thanks to Alexander Levin for pointing out.
18:27:42Battousaianybody know which clip+ funman has?
18:30:02 Join einhirn [0] (
18:31:06 Join DataGhost [0] (
18:31:06 Quit DataGhost (Changing host)
18:31:06 Join DataGhost [0] (~dataghost@unaffiliated/dataghost)
18:32:26 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
18:32:27 Join Luca_S [0] (
18:32:40Luca_SBattousai: clip+2gb, according to his forum signature
18:33:00Battousaihis sig just says clip+
18:33:43Luca_Soops. you're right, I was reading a post from FlynDic e
18:33:43Battousaii hope there's not some playback difference between capacities
18:34:50 Join jgarvey [0] (
18:34:51Luca_Sare you experiencing problems with playback?
18:35:01Battousaii'm experiencing silence ;)
18:35:50Battousaiclip+4gb, i get no audio output, regardless of codec, bitrate, channels, etc
18:35:54 Join PaulJam [0] (
18:36:17 Quit Schmogel (Quit: Miranda IM! Smaller, Faster, Easier.
18:37:20Luca_Sand no panic, just no sound?
18:37:34Battousairight, the files seem to "play" normally, but it's just silent
18:38:15 Join JB2002 [0] (
18:38:29 Join FOAD_ [0] (
18:38:45ThomasAHBattousai: which rockbox svn version?
18:38:53domonokyjust remember that clip+ support is very new in rockbox, and also sandisk is know to put new/different hardware in otherwise identical looking and named device.
18:40:14ThomasAHBattousai: hmm, that one should have sound ...
18:40:24ThomasAHBattousai: and it has on my 8GB clip+,too
18:40:51Luca_S25264 seems also related to sound, maybe it's worth trying an update to latest svn...
18:40:58 Quit JB2002 (Quit: Leaving.)
18:41:02 Join JB2002 [0] (
18:41:18 Quit S_a_i_n_t (Ping timeout: 252 seconds)
18:41:25 Join S_a_i_n_t [0] (S_a_i_n_t@
18:41:40 Quit togetic (Ping timeout: 264 seconds)
18:41:48ThomasAHLuca_S: I think that one is just to not break other targets ... at least I had sound before and after that revision
18:42:10Battousaino harm trying ;)
18:42:27ThomasAHBattousai: 25262 is confirmed from the rockbox menu or from "I am sure that I installed it"? :)
18:42:32BattousaiThomasAH: confirmed
18:42:51Battousaijust built the latest svn, gonna push it to the device now
18:43:23Battousaiafter OF database update :(
18:44:36ThomasAHBattousai: lucky are those people who can access usb-storage via rockbox instead of OF ... my co-worker's c200 for example
18:44:48*ThomasAH waves at swilde :)
18:44:57Battousaii have an e280 that works great with rockbox :)
18:45:03 Quit JB2002 (Ping timeout: 240 seconds)
18:45:53*S_a_i_n_t hugs his 4 RB'd Nanos :)
18:46:15*ThomasAH does not want a proprietary cable, so his choice is limited
18:46:40Battousaiok, 25294 works
18:46:51Luca_Sworks as in you have sound?
18:47:06Battousaiand button lock :)
18:47:37Luca_Snice. another difference in sansa players would be worrying
18:48:10Luca_SI had an e280v1 too, but someone broke into my car and stole it :/ sansa players are extremely difficult to find here and I could replace it only with a fuzev2 :(
18:49:42 Quit planetbeing_ (Remote host closed the connection)
18:50:23 Quit Luca_S (Quit: See you later)
18:51:15 Join DV__ [0] (~DV@
18:53:47ThomasAHcan I use the voice files for Clip on Clip+? Or is there a daily builds page for unusable targets?
18:54:26 Join Kitr88 [0] (~Kitr88@
18:55:46 Join togetic [0] (~togetic@unaffiliated/ibuffy)
18:57:17 Quit Kitar|st (Ping timeout: 260 seconds)
18:58:45 Quit Kitr88 (Ping timeout: 258 seconds)
18:58:50 Join Kitar|st [0] (
19:01:23 Quit CGL (Remote host closed the connection)
19:02:28domonokyThomasAH: as the clip+ is a new target, voicefiles from Clip will probably be rejected (the target is in the header of the voicefile).
19:03:07domonokyThomasAH: but you can generate a voicefile for clip+ with the buildsystem. (I think rbutil doesnt have support for clip+ yet)
19:04:27 Quit DV__ (Ping timeout: 268 seconds)
19:05:38 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
19:08:08bluebroth3rdomonoky: what do you think about saving the system trace to the cache folder in all cases? While this is a nice functionality it seems too may people still don't find it and send a trace from an empty run instead
19:08:08 Join m3dlg [0] (~m3dlg@
19:08:54domonokybluebroth3r: do you think those people will find the log in the cache folder ? :-)
19:09:56domonokybluebroth3r: and the one on the mailinglist is strange. it seems to be a log from a full run, but its cutoff at the end.
19:10:51bluebroth3rdomonoky: no, but we could save the log at quit, so we can "restore" the previous log from the system trace dialog
19:11:07domonokybluebroth3r: might be helpfull.
19:11:36ThomasAHdomonoky: how do I generate it voice files? "make voicetools" works, but "make voice" does not
19:11:40domonokybluebroth3r: but wont the log then get longer and longer, never resetting ?
19:11:43ThomasAHmake: *** No rule to make target `voice'. Stop.
19:12:12domonokyThomasAH: you need to tell configure to generate a voice, and then use "make voice"
19:12:37 Join DV__ [0] (~DV@
19:12:54bluebroth3rdomonoky: no, the idea was to only keep one log around, i.e. start a new one on each start and overwrite it when quitting rbutil
19:12:54ThomasAHdomonoky: "Build (N)ormal, (A)dvanced, (S)imulator, (B)ootloader, (C)heckWPS, (D)atabase tool, (M)anual: (Defaults to N)" ... is the option only offered if voicetools are installed?
19:13:16*domonoky thinks it might be under advanced
19:13:42ThomasAHdomonoky: ah, right ... "Voice build selected"
19:14:22pamaurygevaerts: what would you think of making usb_drv_recv blocking on control endpoint(s) ? This make more sense to me. Or alternatively, (preferably ?) have blocking and nonblocking version of recv ?
19:14:22domonokybluebroth3r: ah so a system trace would normally contain two traces ?
19:14:48bluebroth3rdomonoky: well, kinda. The trace would only contain the current trace, and the last trace would be available on disk.
19:15:58domonokybluebroth3r: no automatic loading of the old trace ? so just another button in the Tracedialog to also save the last trace ?
19:16:03gevaertspamaury: I'd go for blocking and nonblocking versions then
19:16:53bluebroth3rdomonoky: yes, something like this. Though I'm still wondering what's the best way to present this to the user. Maybe a toggle button "show last trace" - "show session trace"
19:16:56***Saving seen data "./dancer.seen"
19:17:54pamaurygevaerts: and do you see the problem I'm talking about (well I talked about a few hours ago) or is it completely fuzzy ?
19:18:03domonokybluebroth3r: hm, it needs to be very easy to just save those traces. The normal user normally wont be able todo anything with it. We just want them to send us the full logs on errors.
19:19:16gevaertspamaury: I see the asymmetry. I don't immediately see why that can cause two recv()s at the same time, but for general cleanness it should be fixed
19:19:42bluebroth3rdomonoky: yes, the presentation is the thing I'm still thinking about. Or just a "Save previous trace" button next to the "Save trace" one?
19:19:54*domonoky likes that.
19:20:27 Join {phoenix} [0] (
19:20:47domonokythe toogle button might nice too, but additional to the two save buttons.
19:21:14bluebroth3rI'll play around with it a bit.
19:21:46pamaurygevaerts: it's not really about the two recv at the same time because this problem is rather specific to my implementation of queueing (but I think any queueing implementation will suffer from that). The "problem" is also that usb_core assume a blocking usb_drv_recv (look at the code: it does usb_drv_recv on a buffer and then reads from it immediately after)
19:22:33pamauryMore generally, any control transfer handling assumes a blocking usb_drv_recv
19:23:35gevaertshm, I can't find such a place, although it must be there. Do you have a line number?
19:23:55pamauryany usb_drv_recv in usb_core.c
19:24:35gevaertsThe ones I see are all to NULL
19:24:57 Quit DV__ (Read error: Connection reset by peer)
19:25:44 Join DV__ [0] (~DV@
19:25:57pamauryah sorry, the buffer ones are/is in usb_hid.
19:26:28pamauryline 686 of usb_hid.c
19:26:58pamauryThe problem of "ack" is different, more driver specific I could say
19:27:36gevaertshm, isn't that one just wrong? I need to check...
19:28:31pamauryWell, it should work because a control transfer is triggered only when the data is already there but there is a potential timing problem imo. I really don't like this one, I already spot it when developing usb for simulator
19:28:33 Join Luca_S [0] (
19:29:25 Join TopyMobile_ [0] (
19:29:29 Join T44 [0] (
19:29:41gevaertsIt's definitely a non-blocking call right now, yes
19:30:12pamauryFor this one at least, the recv should be blocking. And ack should be blocking in my opinion because it's easier for drivers to deal with if queueing is implemented :)
19:30:54gevaertseither the ack recv should be blocking or the ack send should be nonblocking. I personally don't care much
19:31:58pamaurywhy can't both be blocking ?
19:32:42gevaertsRight now the recv is nonblocking and the send is blocking. That's what's wrong
19:33:24 Quit Topy44 (Ping timeout: 258 seconds)
19:33:30 Quit TopyMobile (Ping timeout: 276 seconds)
19:33:33pamaurysemantically you mean ?
19:34:00gevaertsyes. They should be the same
19:34:24 Quit DV__ (Read error: Connection reset by peer)
19:34:33 Quit m3dlg (Ping timeout: 258 seconds)
19:34:51pamauryI agree
19:35:00 Join GHF [0] (~meow@unaffiliated/ghf)
19:35:05pamauryBut acks could all be blocking no ?
19:35:43gevaertsI don't mind either way
19:35:44 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
19:36:21*pamaury secretly dreams that it could be the cause of some HID problems but that's unlikely
19:37:49pamauryWhat should be do about names ? I personally prefer usb_drv_{send,recv}_{blocking,non_blocking} because it's more symmetrical but it will break most of the code. On the other end, we can't miss one :)
19:38:36 Quit linuxguy3 (Ping timeout: 258 seconds)
19:38:41gevaertsI think I agree. It will avoid mistakes
19:39:35 Join stripwax [0] (
19:39:50 Quit stripwax (Client Quit)
19:40:41 Quit PaulJam (Ping timeout: 252 seconds)
19:41:13pamauryOk, I'll have a quick try on usb-arc because normally this is a trivial fix, I don't for the other drivers
19:49:25 Quit perfectdrug_ (Quit: perfectdrug_)
19:52:27 Join linuxguy3 [0] (
19:54:38 Join perfectdrug [0] (
19:56:02*pamaury is falling under compile error :/
19:59:06*S_a_i_n_t finds HW keyclicks on his Nano2Gs quite step, get the keyclick to sound more like the OF ans stop it from repeating even though keyclick repeats are set to off...
19:59:24S_a_i_n_tThen, I'll have very little to complain about :P
20:00:20 Join Buschel [0] (
20:02:09 Join DV__ [0] (~DV@
20:05:04*pamaury hates e200 OF, especially when it's refreshing db while I only want to use usb
20:06:38 Join leavittx_ [0] (~leavittx@
20:06:50pamauryhum, apparently, trivial change for usb blocking/nonblocking don't work :( I'll need to use my brain then
20:09:13 Quit perfectdrug (Quit: Leaving.)
20:09:39 Join perfectdrug [0] (
20:11:32 Join bieber [0] (~bieber@
20:12:10 Part doomcup ("Leaving")
20:12:32bieberI've been looking into the Theme Editor project for Summer of Code, and I'm wondering if there's any particular parsing libraries or parser generators that would be preferred?
20:12:44 Nick hd is now known as jd (~jd@Wikipedia/HellDragon)
20:13:20GodEaterbieber: since we have no editor at all at the moment, I don't think it matters
20:13:30bieberAlso, would starting to work on the basics of it now help my application out any?
20:13:30gevaertsWe currently don't use any, so I suspect there's no preference
20:13:40gevaertsoh, definitely
20:13:46GodEateryes indeed
20:13:50gevaertsNo need to keep yaccing on about it though :)
20:13:52*gevaerts hides
20:14:09bieberNice one :)
20:14:25GodEaternow stay in your corner gevaerts
20:14:37bieberHas anyone else shown interest in the project? And once I've got something to show, who should I show it to?
20:14:50GodEaterI've not seen anyone else ask yet
20:14:57GodEaterand here would be a good place to start :)
20:16:25 Quit n17ikh (Ping timeout: 265 seconds)
20:19:41GodEaterbieber: also, please excuse the awful puns. gevaerts is under the illusion they're funny ;)
20:20:46 Join n17ikh [0] (
20:22:54bieberEh, bad puns are alright by me
20:24:52 Quit DV__ (Read error: Connection reset by peer)
20:28:24 Join jordan` [0] (~jordan@
20:30:59*pamaury suddently wonders is a receive ack makes sense....
20:32:10pamauryparticulary of some data is send after...
20:32:31pamaurygevaerts: what do you think ?
20:32:40GodEaterdid the tagcache chat get anywhere this morning?
20:32:52gevaertsthe data is not sent after. It's *scheduled* after
20:33:04pamauryno, but I think we get closer to the complete rewrite each time
20:33:06 Quit bieber (Quit: Leaving)
20:33:09gevaertsThe host will send the IN-token first
20:34:05pamauryso the ack has to be nonblocking
20:34:12gevaertsThis ordering is to make sure that the transfer is ready as soon as the host wants it, which is apparently needed with some controllers (nano2g, although I'm not entirely convinced)
20:34:16gevaertsoh, right
20:34:53pamauryI don't get it, why a recv ack before a send changes ?
20:35:30 Quit kaniini (Ping timeout: 258 seconds)
20:36:03*pamaury should perhaps refer to usb spec...
20:36:11gevaertsthe host sends the IN token as soon as it has received data. Apparently the nano2g controller gets confused if it's not ready for that IN token yet
20:36:32gevaertsso on our side we do the ack first, but on the wire it will be last
20:36:56pamauryah,... fucking broken controllers
20:37:21gevaertsI'm not convinced actually if it's not just a bug in the driver, i.e. a wring setting somewhere
20:37:45pamauryThat seems strange to me
20:38:36pamauryok, let's break nano2g on my computer for the sake on testing ;)
20:40:20 Join Guest23293 [0] (
20:40:31 Quit n17ikh (Disconnected by services)
20:40:35 Quit Guest23293 (Client Quit)
20:40:54 Join n17ikh [0] (
20:42:45pamauryah, working, except storage which I forgot of course
20:43:30 Quit flydutch (Quit: /* empty */)
20:43:48 Join shaggy-h [0] (
20:53:05 Quit Luca_S (Quit: CGI:IRC (EOF))
20:53:18 Quit S_a_i_n_t ()
20:55:42 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
20:56:13 Quit {phoenix} (Remote host closed the connection)
20:57:17 Join robin0800 [0] (
20:58:37amiconnZagor: ping
20:58:45Zagoramiconn: yes?
20:59:23amiconnYesterday I wanted to subscribe to that rbclient list, but there is a problem
21:00:20amiconnWhen I replied to the confirmation mail, I got an ndr
21:00:45amiconnrockbox-rbclient-request AT doesn't exist
21:01:00amiconnnon delivery report
21:05:32Zagoramiconn: fixed now
21:05:48amiconnCan I just reply again?
21:07:23 Join Amorstus [0] (
21:09:07AmorstusCould someone help me with installing both rockbox and the stock firmware?
21:09:50amiconnZagor: Thanks, this time it worked :)
21:10:30 Join xiainx [0] (xiainx@wpa062081.Wireless.McGill.CA)
21:10:45gevaertsAmorstus: it might help if you said which device you have
21:11:07AmorstusI have an ipod video 5G (not 5.5G, I have checked)
21:12:41gevaertsWhat sort of problems are you having?
21:12:48 Join m3dlg [0] (~m3dlg@
21:13:51AmorstusWell, I want to get the iPL installed and then use rockbox and the stock os with it
21:14:20AmorstusBut terminal continues to not want to install the loader
21:14:37AmorstusI can get rockbox by itself to install fine
21:14:52Amorstushowever, I cannot seem to get both
21:15:19Torneyouu don't need ipl to use rockbox and the original firmware
21:15:30 Quit pixelma (Disconnected by services)
21:15:31 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
21:15:33Torneour bootloader dual boots fine on its ow
21:15:37 Quit amiconn (Disconnected by services)
21:15:39 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
21:15:50 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
21:16:03Amorstusok, then how would I use this?
21:16:06 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
21:16:26Tornethe normal install already dual boots
21:16:53AmorstusUse the "prev" button for apple os?
21:16:58***Saving seen data "./dancer.seen"
21:17:00Tornehold menu instead of select to power on, or power on then turn on the hold switch, and it will boot the original firmware
21:17:08Tornethis is in the manual.
21:17:18AmorstusMust be blind then
21:17:34Amorstusthanks :D
21:20:16 Join fml [0] (
21:21:27 Join webguest40 [0] (
21:21:36 Quit webguest40 (Client Quit)
21:24:01fmlAlexP: hello. Have you seen FS #11135 ? BTW: I find the wording in the description of %pE a bit weird in the sense that the tag _becomes_ true (if we go along the time axis). And while you're at the manual: could you please try the proposed fix for FS #11117 ?
21:28:00 Quit komputes (Ping timeout: 260 seconds)
21:28:00 Quit bmbl (Ping timeout: 258 seconds)
21:29:10fmlmc2739: Hello. FS #11117 has a proposal for fixing which I can't try out right now. Could you?
21:29:30 Quit Zagor (Remote host closed the connection)
21:29:51 Join toffe82 [0] (~chatzilla@
21:32:11 Quit m3dlg (Ping timeout: 240 seconds)
21:36:59AlexPfml: OK I'll change it, no I didn't I'll close it, and not right now I don't have the time
21:37:53AlexPShort and sweet :)
21:39:11fmlAlexP: what is short and sweet? No problem, we're not in hurry.
21:39:25AlexPmy answer
21:41:06fmlYes, condensed and compact!
21:41:16CIA-5New commit by 03alex (r25295): Improve %pE description.
21:42:27 Quit Amorstus (Remote host closed the connection)
21:43:07AlexPthat description is crap too
21:43:11AlexPTeach me to rush
21:43:15 Join Adubbb [0] (~aldubuc@
21:43:53AlexPno, actually it is fine
21:44:14AlexPI was reading r25294 not r25295 :)
21:44:35 Quit Adubb (Read error: Connection reset by peer)
21:46:40 Join komputes [0] (~komputes@ubuntu/member/komputes)
21:48:26 Join wodz [0] (
21:50:15 Quit JohannesSM64 (Quit: WeeChat 0.3.2-dev)
21:50:19 Join CGL [0] (~CGL@
21:55:21 Nick fidencio is now known as fidencio[AWAY] (
21:58:09wodzwhat exactly HAVE_POWEROFF_WHILE_CHARGING option do? Does it simply block poweroff event?
21:59:29GodEaterwodz: it means the device can be powered off while the battery is charging
21:59:34GodEatersome devices do not allow that
22:01:41 Quit grndslm (Remote host closed the connection)
22:02:22wodzGodEater: ok but how does it work from the codeflow point of view. Does it simply change keys interpretation?
22:03:09 Quit fml (Quit: CGI:IRC 0.5.9 (2006/06/06))
22:10:35 Join Darkknight512 [0] (
22:10:40*Bagder adds committer #89
22:10:49Bagderranmachan: welcome!
22:11:15 Quit komputes (Ping timeout: 246 seconds)
22:11:23*domonoky welcomes ranmachan in the rockbox crew !
22:11:43AlexPHi :)
22:14:09kugelawesome \☺/
22:14:34 Quit fyrestorm (Ping timeout: 248 seconds)
22:20:47mtranmachan: congrats and welcome ! :)
22:22:39 Quit topik (Ping timeout: 258 seconds)
22:27:52 Quit xiainx (Ping timeout: 265 seconds)
22:28:18bluebroth3rdomonoky: any comments?
22:29:01 Join blairb [0] (
22:30:14domonokybluebroth3r: looks fine.
22:31:49*bluebroth3r starts to get pissed by this guy at the forums
22:32:22 Join saratoga [0] (~9803c6dd@gateway/web/freenode/x-vyitymjtbtkghrww)
22:32:44saratogawhat do people think about making make zip automatically include a diff if the build is from a modified svn checkout?
22:32:56bluebroth3rsaratoga: thanks for locking that thread in the forums. I'm just getting pissed by this guy :(
22:33:00saratogawould make it a lot simpler for third party builders to comply with the GPL
22:33:10Tornesaratoga: er, no it doesn't
22:33:22saratogayes i couldn't understand what he was talking about and it seemed to either to lock it then to parse whatever was going on there
22:33:37Tornei'm not saying it's not a reasonable idea, but it's not relevant to gpl compliance
22:33:43bluebroth3rsaratoga: hmm, but that would also mean that we should handle all vcs systems supports
22:33:52Tornea diff si not sufficient
22:34:01saratogaTorne: can you elaborate
22:34:28Tornesaratoga: a diff against someon e else's source does not satisfy the requirement to convey source in the GPL
22:34:43Torneyou have to provide the *entire* source used to build the binaries (on request, at least)
22:34:49saratogaprovided they used our svn checkout, it does
22:34:54Torneno, it doesn 't.
22:34:57saratogasince it tells you where to get the code to diff against
22:35:04TorneNo. That is not how the GPL works.
22:35:34bluebroth3rthe guy providing the changed build needs to provide the full sources, not someone else (ie. the original project)
22:35:36TorneYou must either provide the *full source used to build your bianries*, or provide a written offer to do so valid for some very long time
22:35:52 Join DerPapst [0] (
22:35:58saratogai don't recall a provision in there saying you had to host it yourself
22:36:28TorneYou don't have to *host* it, b ut you have to be able to provide the code on request for a very long time
22:36:38Torneunless you accompany it with the code to start with.
22:36:46TorneSeriously, this is a known thing
22:36:52Tornei am not just misunderstanding here :)
22:37:00Tornethi shas been discussed in here before multiple times
22:37:15saratogaif you are not required to host it, why is linking to a problem?
22:37:32Tornebecause what's on is not the code used to build the binary
22:37:46saratogacombined with the patch it is
22:37:52TorneYes, that is irrelevant
22:38:01Torneand even if it were unmodified, is not guaranteed to always exist for the lifetime of the binary
22:38:07saratogacan you cite the part of the GPL saying so?
22:38:15TorneThe requirements are much mroe strict than most people think
22:38:59Tornesection 3, gives you three options for providing source
22:39:33Tornea) requires that you provide the complete corresponding source with the bianry
22:39:52Tornelinking to someone else's server is not sufficient because it might not be available while the bianry still is.
22:40:08Torneb) requires that you provide an offer to give the source on request which you hav to honour for three years, which is also a huge pain :)
22:40:18Torneand c) is only allowed if you did not modify the code
22:40:27saratogaits impossible to garentee anything for a period of 3 years
22:40:39TorneYes, which is why almost everyone does option a)
22:40:44saratogayou simply claim in good faith that you will try
22:40:47TorneBut you have to either include the source *with* the binary
22:40:58Torneor have it prominently available right next to the binary, and available at the same time
22:41:00saratogaheh so by this logic rockbox's download links violate the GPL
22:41:05Tornewhch you cannot guarantee if you are relying on third party hosting.
22:41:07saratogasince we don't know that will exist for 3 years
22:41:16TorneNo, three years is only relevant for b)
22:41:18Tornewe are using a)
22:41:27saratogawe do not
22:41:27gevaertssaratoga: the same page has a link to the source, on the same server. That's (a)
22:41:38TorneIf the source is always available right there at the time of the download and the user chooses not to use it that's their fault and this is allowed.
22:41:39bluebroth3raren't we c) for the binaries?
22:41:49gevaertsbluebroth3r: no, because we're upstream
22:41:51Tornebluebroth3r: no, it's our own code, we can't be redistributing it ;)
22:41:51saratogayeah we don't package the source with the binary
22:42:01bluebroth3rtrue :)
22:42:04Tornesaratoga: you don't have to, it's okay for it to be available at the same time by the same process
22:42:10Tornei.e. a download from the same site
22:42:14Tornewith the link right there.
22:42:38bluebroth3rif every binary would need to include all source ... well, I don't want to think about the size of my linux installaiton then
22:43:16Tornebluebroth3r: inded, but it means that, for example, a debian mirror *must* also mirror the source debs if they mirror the bianry debs.
22:43:39saratogaso its ok as long as is the same physical machine?
22:43:39 Join planetbeing_ [0] (~planetbei@
22:43:44saratogawhat happens if we get two PCs
22:43:46bluebroth3rEven for mirrors?
22:43:49Tornesaratoga: it doesn't have to be the same amchine..
22:43:59Tornesaratoga: we are one entity, from copyright pov, more or less :)
22:44:18TorneIt is reasonable to expect that if our binary download server still exists, so does our svn repo
22:44:22saratogawe are Ahead ? because last I checked they have the copyrights on the aac decoder we distribute
22:44:57saratogaand what about VLC, i don't believe they host our source code at all, but they do have our binraries
22:45:03saratogaare they violating our licenses?
22:45:09Tornesaratoga: distribution is about who is doing the distribution, not who has the individual copyrights
22:45:14saratogaare they violating ahead's licenses?
22:45:19gevaertssaratoga: do they host anything but stable builds?
22:45:35Torne"the rockbox project" is the entity doing the distribution of our official binaries
22:45:48saratogai know at one point we had daily builds at least on the VLC machines
22:46:07saratogahaven't checked recently
22:46:37saratogaso when VLC distributes libfaad and links to thats fine, but if someone else distributes libfaad and links to thats not ok?
22:47:16Tornesaratoga: this covers exactly what we're talking about here :)
22:47:28domonokysaratoga: does VLC distribute modified binarys ?
22:47:28 Join komputes [0] (~komputes@ubuntu/member/komputes)
22:47:29saratogathat clearly doesn't apply to SVN though
22:47:35saratogasince it has version control
22:47:38Tornesure it does
22:47:43saratogaand anyway, the gnu faqs are not definitive
22:47:45Tornewho's to say our svn server will exist in five years?
22:47:47saratogathey've often quite stupid
22:48:03saratogaif we think that then we're violating the GPL!
22:48:04Tornemaybe we've all gotten sick of getti gn no help from manufacturers and pulled the plug
22:48:22Torneas soon as we stop distributin g bianries we are also allowed to kill our svn server ;)
22:48:31Tornebut anyone whose modified build is still up somewhere must still make source avaialble
22:48:41saratogaand they can do that
22:48:50saratogabut that doesn't mean they're violating the GPL before we pull the SVN server
22:48:56TorneYes, it does.
22:49:13Tornethat would be a stupid license :)
22:49:16gevaerts*If* they include the three-year written offer, they're fine
22:49:29saratogai mean if I know we're taking down at the end of the month i have until the end of the month to pull my binary or upload the source
22:49:29domonokyyes, no, yes :-) probably only a curt can decide this ? :-)
22:49:31TorneIf they take 3b) isntead they don't have to provide *anything* up front, other than that offer
22:49:53saratogawho needs a court, we have the GNU FAQs!
22:50:04Tornebut then they would need to hang on to all the details of what their build included for three years after the last person downloaded it
22:50:06saratogathey make up for the GPL not saying all sorts of things they want it to say
22:50:20TorneI think 3a) is pretty clear, tb h
22:50:28gevaertssaratoga: uhm, no. If you distribute the binary but not the source, you *have* to go for (b)
22:51:08saratogai don't think i disputed that?
22:51:33gevaerts"i have until the end of the month to pull my binary or upload the source"
22:51:50gevaertsThat implies that you're not distributing the source
22:52:07saratogayes, because I'm linking to
22:52:20gevaertslinking to is *not* distributing
22:52:24saratogai mean i give people rockbox bootloaders all the time on my own hosting, and tell them the source is in
22:52:27Torneif what *someone else* does means you could suddenly be violating the license without knowing about it, then, er, that would be stupid :)
22:52:37 Quit DerPapst (Quit: Leaving.)
22:52:54saratogathe GPLv2 is kind of stupid about a lot of things
22:53:33TorneIt's not stupid about this :)
22:53:33saratogalook all i'm saying is that we regularly put up builds on other hosting (and sometimes even on the wiki) and then say the source is in
22:54:06saratogaand i do this without perfect knowledge of the future state of the rockbox project
22:54:20saratogai would even argue that none of us is perfectly aware of future events for the next 3 years
22:54:34Torne(3 years has nthing to do with this)
22:54:40Torne(we do not use section 3b) :)
22:54:56Torne3a is for as long as you distribute the binaries.
22:54:57 Quit Zarggg (Quit: Zarggg)
22:55:06saratogawe do 3b a lot
22:55:09saratogago look at the wiki
22:55:17TorneNo, we don't
22:55:33TorneWhere does it say, anywhere ont eh wiki, "You may ask $FOO for the soruce for this build for 3 years fromt he current date"
22:55:34saratoga3b unless I am mistaken
22:55:56domonokysaratoga: the wiki also belongs to rockbox.
22:55:56TorneYes, you are mistaken
22:56:01Tornethere is no written offer to provide source there
22:56:13Torneyou don't even mention source :)
22:56:15saratogaok i'll fix it
22:56:33TorneI wouldn't, because you don't *want* to be distributing under 3b
22:56:38saratogaits arguably more stupid
22:56:41saratogabut its not 3b
22:56:50saratoganow 3b
22:57:06saratogawere i guess before it was vague because i didn't say how someone could find the source
22:57:10 Quit jgarvey (Quit: Leaving)
22:57:12domonokysaratoga: we provide the sourcecode for those builds, its in svn, which belongs to the same entity. belongs to
22:57:38Torneyes, you should really say "the source is r12345 from svn" or similar ;)
22:57:47saratogathe binary has that information in it when you boot it
22:58:01TorneThat also doesn't help
22:58:08Tornethe user needs to know when they download it
22:58:13Torneso that if they want the souce, they can download that too
22:58:14saratogahonestly putting next to every download link is stupid
22:58:17saratogawe have it on the front page
22:59:24saratogai'm still not even convinced what i'm doing is allowed under this retarded reading of stuff thats not in the GPL
22:59:37saratogai mean what if goes down and I die in a car accident
22:59:40*domonoky thinks that in this case the hosting on is done for rockbox, like a service (normally payed, but free this time) so it belongs to rockbox, like rented webspace :-)
22:59:52Torneanyway. my point was simply that if you implement such a feature in make zip, you *should not* state anywhere that this is sufficient for someone's modified build's GPL obligations
22:59:56Torneb ecause it is not.
23:00:04saratogai don't agree with that
23:00:16Tornewe have, numerous times in the past, told people providing modified builds that it is not :)
23:00:19 Quit evilnick_B (Quit: Page closed)
23:00:21saratogaand you have not made a reasonable argument to the contrary
23:00:32saratogawhoever said that was wrong
23:00:45Tornewell, i can't say anything else on this tbh. ask the sounding board.
23:02:21domonokyanyone hosting binarys not on should provide source. Cases where the main link to this binarys is only on are debateable :-)
23:02:36*gevaerts agrees with domonoky
23:03:08gevaertsWell, if "provide source" means "either have the source next to the binary or provide a written offer" :)
23:06:36 Join Zarggg [0] (
23:08:26 Join yelped [0] (
23:08:55yelpedsomrthing weird going on.
23:09:05*bluebroth3r wonders how him providing svn binaries of rbutil looks in that light.
23:09:36bluebroth3rso ideally I should also provide sources. Though the binaries are from unmodified svn.
23:09:45gevaertsyelped: I think that's what's JdGordon's been chasing for a few days now
23:10:17 Join m3dlg [0] (
23:10:36yelpedOh, I had updated today from a build a week ago.
23:10:57 Quit Buschel (Ping timeout: 258 seconds)
23:12:12yelpedin the forums?
23:12:15saratogawhat if i make a written offer, and when some one requests, i email them
23:13:45 Part Chrolo
23:16:44 Quit m3dlg (Ping timeout: 246 seconds)
23:17:00***Saving seen data "./dancer.seen"
23:17:19 Quit wodz (Quit: Leaving)
23:22:30 Nick fidencio[AWAY] is now known as fidencio (
23:26:37 Join bieber [0] (
23:27:34 Quit pamaury (Quit: Quitte)
23:29:10saratogaheh LOTs of places on the forums where people are just linking patches rather then complete source code archives
23:29:43saratogawas it actually our policy that this isn't allowed?
23:29:55saratogai never even heard of it, and we certainly didn't enforce it on the forums
23:30:11saratogaand apparently almost no one was aware of it
23:30:38saratogaheh i even see people just giving FS numbers without even hosting their changes
23:31:52bieberQuick question on parsing WPS files: how is whitespace supposed to be treated in commands (or should I call them tags?) with multiple parts? i.e. if you had a named identifier, what would happen if you did something like %Vi| a |...| ?
23:32:38 Join anewuser [0] (anewuser@unaffiliated/anewuser)
23:32:52bieberErr, named viewport
23:40:36 Join Schmogel [0] (
23:41:30saratogaheh i'm a little concerned that the guy who told funman the current mkamsboot code worked fine on the clipv2 proceeded to ask in the forums if someone could explain to him how to use mkamsboot the next day
23:41:30 Quit robin0800 (Ping timeout: 248 seconds)
23:41:48saratogagoing to go out on a limb and suggest that he may not actually have tested it
23:42:14 Quit Adubbb (Read error: Connection reset by peer)
23:44:03 Join Adubb [0] (~aldubuc@
23:46:41 Join m3dlg [0] (
23:48:45 Nick fidencio is now known as fidencio[AWAY] (
23:49:36 Join Topy44 [0] (
23:50:04 Quit T44 (Read error: Connection reset by peer)
23:50:41 Join angelwolf71885 [0] (
23:51:04 Quit m3dlg (Read error: Connection reset by peer)
23:51:28 Join m3dlg [0] (
23:51:48 Join elinenbe_ [0] (
23:52:05 Quit elinenbe (Read error: No route to host)
23:52:05 Nick elinenbe_ is now known as elinenbe (
23:53:07angelwolf71885hello rockbox :)
23:54:28 Quit angelwolf71885 (Client Quit)
23:56:06 Quit einhirn (Read error: Connection reset by peer)
23:56:42 Quit bieber (Quit: Leaving)

Previous day | Next day