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 2008-10-23

00:00:15 Quit tvelocity ("Αποχώρησε")
00:00:26ThedjatclubrockCan iTunes export as .m3u?
00:00:58scorcheThedjatclubrock: ask them...
00:01:28 Quit moos ("good night europe")
00:02:34 Quit mirak ("Ex-Chat")
00:04:08 Quit n1s ()
00:04:26fmlXavierGr: pong. I just added a task (FS #9504) that should help in FS #9455
00:04:29*bluebrother points to the IpodFAQ ...
00:05:13XavierGrok let me see
00:05:13 Join neddy [0] (n=john@nat/sun/x-7ce0d697352b667d)
00:06:37XavierGrfml: yes it would eliminate the last length parameter in my patch
00:06:50XavierGrfml: about strncat on lcd-x files
00:07:02fmlXavierGr: if FS #9504 is committed, you can make the definition of the max length of the padding string (in settings) as complicated as you want and will not to have to care about the matching definition in settings_list
00:07:17linuxstbfml: I see you changed a comment to say "filename (or another string)" - maybe the macro should be renamed, if it's not specific to a filename?
00:07:44fmllinuxstb: yes, it can handle any string, not only file names
00:08:03fmllinuxstb: but it would be another patch :-)
00:08:07ThedjatclubrockCan RB DB use iTunes Album Art?
00:08:19krazykitThedjatclubrock, no
00:08:20fmlXavierGr: what about strncat?
00:08:23XavierGrfml: for the last parameter of strncat, to make it safe, I have to subtract the size of the destination array with its current length
00:08:50XavierGrbut is the size of the destination array known?
00:08:54fmlXavierGr: correct. And then set the last char to 0.
00:09:01ThedjatclubrockThere should be a RB program, cross platform, Java or C with Qt
00:09:06ThedjatclubrockOr Swing
00:09:17linuxstbYou mean like Rockbox Utility?
00:09:23fmlXavierGr: yes, it's known. It's the size of the padding_string array in the firmware layer.
00:09:51XavierGrfml: sizeof(s->line) should return the size of the pointer instead right?
00:09:55fmlXavierGr: or just use snprintf. It does add the terminating 0
00:10:22XavierGrfml: I tried using snprintf but it failed for a reason, I didn't search thoroughly though
00:10:42krazykitThedjatclubrock, for converting album art? there probably is some software already written to do it, but feel free to write it yourself too.
00:11:02pixelmalinuxstb: I overlooked the "line" and thought you were talking about an overall size limit for .wps files... but with viewports it's not so clear anymore. Hmmm, I also wonder what lines are counted
00:11:43fmlXavierGr: where is that s-> line defined?
00:11:54fmlXavierGr: lcd.h?
00:12:08Thedjatclubrockkrazykit: No, like, iTunes for RockBox
00:12:21 Quit sarixe ("Ex-Chat")
00:12:31ThedjatclubrockOr a Banshee/RhythmBox plugin
00:12:35XavierGrfml: look at lcd-16bit.c where my patch does the strcat
00:13:28krazykitThedjatclubrock, er, itunes does work for rockbox, and since it's otherwise an MSC device, your media player may support syncing. i know amarok does, and i'd suspect others do too.
00:13:31XavierGrfml: hmmm, never mind scroll_engine.h which has the define for the size of that array is included
00:14:13fmlXavierGr: yes, but how is the "line" member defined? I see it's a part of the lcd_scroll_info.scroll struct. Where is it defined?
00:14:21Thedjatclubrock Amarok has native RB!
00:14:44XavierGrfml: why are we so fixed on making the last char 0? It seems to me that all methods append the null correctly. Or do we want better safe than sorry even if we sacrifice one command
00:14:44linuxstbThedjatclubrock: One of the design goals of Rockbox is that it's not dependant on any specific PC application - you simply need to drag and drop your files, and Rockbox will play them.
00:14:56fmlXavierGr: yes, the sizeof(s->line) should be fine
00:15:25XavierGrfml: I thought that this (sizeof(s->line)) will return the size of the pointer not of the array
00:15:42Thedjatclubrocklinuxstb: Yes, great. I'd also suggest that there be a plugin in iTunes, Banshee and RhtyhmBox to port album art and playlists and a DB.
00:15:47 Quit meven (Remote closed the connection)
00:16:04fmlXavierGr: because it can happen that the size of the dest array is less that the copied string. Then, when you use strnXXX, you won't overrun the array but the string in the dest array will not be zero terminated
00:16:10XavierGrfml: scroll_engine.h struct scrollinfo
00:16:25 Join stripwax [0] (
00:16:27linuxstbThedjatclubrock: Just use any PC application that uses the standard m3u playlists, and there is no need to port them. As for the DB, Rockbox builds it itself.
00:16:47XavierGrfml: I understand the use strncpy and strncat, I don't understand the extra command of zeroing out the last char.
00:17:00fmlXavierGr: yes, found that. The field "line" is a char array ( char line[xxx]) hence sizeof (line) will give the size of that array
00:17:12 Part L29Ah
00:17:18fmlXavierGr: see my explanation above
00:17:24scorcheThedjatclubrock: you are free to make one yourself...
00:17:33stripwaxstrncpy won't guarantee the string is zero terminated
00:17:40fmlXavierGr: this is to make sure that the dest string is always zero terminated
00:17:49ThedjatclubrockI would, but I am not great with my dev skills
00:18:18XavierGrfml: yeah, that's my point, it is always zero terminated by other functions, eg. settings loading and virtual keyboard in our case
00:18:27 Quit culture (Read error: 110 (Connection timed out))
00:18:56fmlThe buffer size in the firmware layer must not necessarily match the setting size in the app layer (since these two pieces of code are not so closely related).
00:18:58XavierGrI mean if the dest string is always null terminated and the appended string is null terminated how the new string can end without null?
00:19:38XavierGrbut okay I guess one more command shouldn't hurt
00:20:11fmlXavierGr: if the appended string is truncated (because the dest buffer is not large enough to hold the whole string), then the dest buffer will not be zero terminated after strncat
00:21:03fmlBut it will always be after snprintf. So either strnxxx + set the last char to 0 or just snprintf
00:21:10XavierGrfml: about the MAX macro. Currently it is not included on the files the padding is defined, I am not so sure it will be good to define them just for the player
00:21:24XavierGrfml: which is the only target with so little screen width
00:21:37XavierGrI think ifdefing for that case alone is enough
00:22:06XavierGrbecause the ifdefs are already there for e.g scroll step
00:22:21fmlXavierGr: I'm actually not sure we have to ensure that the setting is at least 10 chars. Of course we don't have to define a macro for that, this was just a scheme.
00:23:00 Quit mcuelenaere ("Zzzzz")
00:23:03fmlYou can write e.g. LCD_WIDTH/4 < 10 ? 10 : LCD_WIDTH/4
00:23:26XavierGrOndio seems to be next with 112 pixels LCD_WIDTH which is plenty after the division with 4 (to fill a screen)
00:23:43XavierGrfml: yeah that can be done too, but won't some consider it more confusing that the ifdef?
00:23:43fmlThis is a rather complex definition. Hence it's highly desireable that it doesn't have to be repeated anywhere else. Hence FS #9504.
00:24:34fmlXavierGr: we need that only if we want to ensure pad length >= 10. But do we want that? I'd say no for the first try.
00:24:45 Quit colesterol_dog (Read error: 104 (Connection reset by peer))
00:25:02XavierGrfml: LCD_WIDTH / 4 is used currently on 3 files
00:25:16XavierGrthe one you tried to avoid with your patch
00:25:24 Join m0f0x [0] (
00:25:42XavierGrsettings.h and scroll_engine.h will still need it
00:25:50fmlXavierGr: hence leave this part of the code as it is. With FS #9504 you won't have to repeat anything. Later you can make the size definition (in settings.h) more complicated −− and still not have to adjust anything else!
00:26:27XavierGrfml: what do you mean exactly, leave the code with the ifdef?
00:26:35XavierGr(for the player)
00:27:28fmlXavierGr: where? settings.h −−> ok. settings_list.c −−> eliminated with FS #9504; firmware code? With proper usage of strn-funcs, you can set the size of the string in the firmware layer to another value, everything will still work.
00:28:01fmlXavierGr: leave the code = no need to ensure that the length is >= 10
00:28:29 Quit feisar (Remote closed the connection)
00:28:35XavierGrfml: scroll_engine.h also uses LCD_WIDTH /4
00:29:09XavierGralso if I don't ifdef the player the setting is useless for that target 11/4 only 2 chars, lower than the default. (actually one char and the other the null)
00:30:07XavierGrso I will just ifdef player and later we'll see
00:30:15fmlXavierGr: yes, but IMHO it doesn't have to. I mean, the firmware value can be seen as a capability of the hardware. Of course it's good if we defined our settings' (app layer) size to exactly match that value. But we don't have to.
00:31:36fmlXavierGr: no,no I mean the defferentiating between char and dot displays is OK. My proposal was to ensure (only for dot displays!) that the max size is at least 10
00:31:51XavierGramiconn: is the M3s remote screen defined as a main LCD or REMOTE_LCD?
00:31:59fmlBut, as I said, I doubt we need that.
00:32:24XavierGrfml: ah okay then
00:32:43pixelmaXavierGr: main LCD
00:32:46 Join feisar [0] (
00:33:04XavierGrpixelma: phew, thanks, that saves me from something :)
00:33:13pixelmaon the M3 but it's the same remote you can use with an M5 or X5
00:33:13fmlSo leave this as it is. If some users of dot displays with WIDTH<40 will complain, we will add this "10". Nothing else will need to be adjusted.
00:33:53XavierGrfml: agreed. So nice we can finally talk about it in IRC, the tracker was tiresome for little details
00:34:13XavierGrfml: anything else?
00:34:31fml:-) I hope all points are clear now.
00:35:09XavierGrok most changes were done yesterday, I will add some more and upload a new patch.
00:35:12fmlXavierGr: but let me look at the next patch (tomorrow, I'll leave now) :-)
00:35:34fmlXavierGr: bye
00:35:39 Quit fml ("CGI:IRC 0.5.9 (2006/06/06)")
00:47:06 Quit bluebrother ("leaving")
01:00:01 Quit faemir (Remote closed the connection)
01:03:39***Saving seen data "./dancer.seen"
01:09:41 Join Nibbl [0] (
01:12:46 Quit DerDome ("Leaving.")
01:20:21 Quit Thundercloud (Read error: 104 (Connection reset by peer))
01:21:35 Quit AJCantos (Read error: 110 (Connection timed out))
01:29:46 Nick fxb is now known as fxb__ (
01:36:52 Quit krazykit ("dinnertime")
01:42:23 Join AhtiK [0] (
01:44:53 Quit n17ikh|Lappy ()
01:50:43 Join lordpil [0] (
01:55:29 Join fdinel [0] (
01:55:50 Quit fyrestorm (Read error: 113 (No route to host))
01:59:12 Part pixelma
01:59:34 Join pixelma2 [0] (n=marianne@rockbox/staff/pixelma)
02:04:34soaphmm, saratoga, then why did Bagder tag that task (FS #5796) for "later"? Or did "later" come and the task never got updated?
02:13:48 Join saratoga [0] (n=41becb3b@gateway/web/cgi-irc/
02:14:03 Nick JdGordon|zzz is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
02:14:07saratogasoap: I don't know much of anything about vorbis
02:14:08JdGordonlinuxstb: whats free?
02:14:24saratogabut i guess floor0 worked at one point, or it may still work now but for only certain files
02:17:54JdGordonwhitespace is for whimps!
02:18:38 Join n17ikh|Lappy [0] (
02:20:21 Quit Hillshum ("ChatZilla 0.9.83 [Firefox 3.0.3/2008092417]")
02:21:35JdGordonLambdaCalculus37: do you know how to make a e200v2 bootloader that is usable?
02:22:11LambdaCalculus37JdGordon: I don't even have an e200v2.
02:22:31JdGordonits not the same as the other v2s?
02:23:34LambdaCalculus37It is, but I don't have one, nor do I know how to make a v2 bootloader.
02:23:57LambdaCalculus37By the way, can you give the Sansapatcher at FS #9639 a kick around?
02:24:14JdGordonyep, ill do that today
02:24:27JdGordonstill havnt got very far out of bed this morning :p
02:24:34 Join cool_walking_ [0] (i=cb3b81c3@gateway/web/ajax/
02:41:28 Quit saratoga ("CGI:IRC")
02:49:55 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
03:03:40***Saving seen data "./dancer.seen"
03:09:11 Join HBK [0] (
03:13:31 Quit gromit`` ("Coyote finally caught me")
03:13:50 Join gromit` [0] (
03:15:05 Quit PaulJam (".")
03:15:38 Quit neddy ("Leaving.")
03:18:22 Quit HBK (Read error: 104 (Connection reset by peer))
03:18:27 Join HBK- [0] (
03:30:32 Quit stripwax (Read error: 104 (Connection reset by peer))
03:37:40 Join xFinch713x` [0] (
03:40:28 Join ajonat [0] (n=ajonat@
03:52:50 Quit m0f0x ()
03:52:57 Join axionix [0] (
03:54:53 Join neddy [0] (n=john@nat/sun/x-f2770b0f58577360)
03:56:13 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
04:01:35 Quit AndyI ()
04:02:24 Quit LambdaCalculus37 ("Ka-chunka")
04:08:18 Join AndyI [0] (i=AndyI@
04:17:11 Quit XavierGr ()
04:21:56 Join blkhawk- [0] (
04:23:49 Quit BHSPitMonkey (
04:23:49 Quit Zarggg (
04:23:49 Quit blkhawk (
04:23:49 Quit Tuplanolla (
04:23:49 Quit blahrus (
04:23:49 Quit blithe (
04:23:49 Quit DaCapn (
04:23:49 Quit Bjoern-Erik (
04:24:37NJoinBHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
04:24:37NJoinblkhawk [0] (
04:24:37NJoinTuplanolla [0] (n=jani@
04:24:37NJoinblahrus [0] (n=blahrus@
04:24:37NJoinBjoern-Erik [0] (
04:24:37NJoinDaCapn [0] (
04:24:37NJoinblithe [0] (
04:24:56NJoinZarggg [0] (
04:26:22 Join Tuplanol1a [0] (n=jani@
04:28:38 Quit blithe (
04:28:38 Quit DaCapn (
04:28:38 Quit Bjoern-Erik (
04:28:38 Quit blahrus (
04:28:38 Quit Tuplanolla (
04:28:38 Quit BHSPitMonkey (
04:28:38 Quit blkhawk (
04:29:26NJoinBHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
04:29:26NJoinblkhawk [0] (
04:29:26NJoinblahrus [0] (n=blahrus@
04:29:26NJoinBjoern-Erik [0] (
04:29:26NJoinDaCapn [0] (
04:29:26NJoinblithe [0] (
04:33:04 Quit Horschti ("Snak 5.3.3 Unregistered copy. Evaluation period is over. Program will now quit. Thanks for using Snak.")
04:33:13 Part xFinch713x`
04:39:24 Quit blkhawk (No route to host)
04:40:21 Quit einhirn (Read error: 104 (Connection reset by peer))
04:55:24 Join blkhawk [0] (
04:56:42 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
05:01:25 Quit ajonat ()
05:03:41***Saving seen data "./dancer.seen"
05:12:53 Quit blkhawk- (Read error: 110 (Connection timed out))
05:16:45 Quit BHSPitMonkey (Remote closed the connection)
05:18:00 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
05:27:00 Quit Horscht ("electromagnetic radiation from satellite debris")
06:03:49 Join goffa_ [0] (n=goffa@
06:12:58 Quit Zarggg ()
06:15:49 Nick HBK- is now known as HBK (
06:17:03 Quit goffa (Read error: 110 (Connection timed out))
06:21:49 Join nuonguy [0] (
07:03:45***Saving seen data "./dancer.seen"
07:26:58 Quit JdGordon ("Konversation terminated!")
07:29:01 Join JdGordon [0] (
07:31:33 Join AndyIL [0] (i=AndyI@
07:35:26 Nick fxb__ is now known as fxb (
07:43:16 Nick fxb is now known as fxb__ (
07:44:13 Quit AndyI (Read error: 110 (Connection timed out))
07:50:36 Nick MU{lappy} is now known as Unhelpful (
07:53:08 Quit kachna (Read error: 60 (Operation timed out))
08:00:04 Quit denes_ (Read error: 110 (Connection timed out))
08:00:16 Join denes_ [0] (
08:03:41 Quit JdGordon (Read error: 104 (Connection reset by peer))
08:09:32 Join fyrestorm [0] (
08:13:17 Join JdGordon [0] (
08:14:37 Join goffa [0] (n=goffa@
08:15:39 Join reacocard [0] (n=reacocar@WL-112.CINE.HMC.Edu)
08:19:07 Quit DataGhost (Nick collision from services.)
08:19:15 Join DataGhost [0] (i=dataghos@unaffiliated/dataghost)
08:19:15 Quit BHSPitMonkey (Remote closed the connection)
08:26:16 Quit goffa_ (Read error: 110 (Connection timed out))
08:29:50 Join goffa_ [0] (n=goffa@
08:30:44 Quit perrikwp (" ajax IRC Client")
08:31:28 Join GodEater [0] (i=c2cbc962@gateway/web/ajax/
08:32:08 Join perrikwp [0] (i=d1a8d351@gateway/web/ajax/
08:34:38 Quit BigBambi (Read error: 113 (No route to host))
08:36:40 Quit goffa (Read error: 110 (Connection timed out))
08:37:08JdGordonwhat size buffers does the hardware usually want from the pcm buffers?
08:38:09 Join ender` [0] (
08:44:47 Join bertrik [0] (
08:48:58 Quit Slack ("Ex-Chat")
08:54:19 Join Bagderr [241] (n=daniel@rockbox/developer/bagder)
08:54:44 Nick Bagderr is now known as B4gder (n=daniel@rockbox/developer/bagder)
08:55:26JdGordonam I right that the only major bits of pcmbuf.c are pcmbuf_request_buffer, pcmbuf_write_complete and pcmbuf_callback ? everything else just deals with stuff like crossfade and managing the buffer?
08:56:09JdGordonthose 3 funcs are called to 1) get a block to write some pcm to, 2) say we have fininshed writing to a block and 3) dump a block to the DAC?
08:57:59 Join Rob2222 [0] (
09:03:46***Saving seen data "./dancer.seen"
09:07:57 Join petur [50] (n=petur@rockbox/developer/petur)
09:16:42 Quit Rob2223 (Read error: 110 (Connection timed out))
09:18:46 Quit Nibbl ("Ex-Chat")
09:22:00 Join BigBambi [0] (i=86ceaf40@rockbox/staff/BigBambi)
09:23:15 Join kachna [0] (
09:24:05 Quit bertrik ("Leaving")
09:26:50*JdGordon needs NicoP :(
09:29:36 Join Zagor [0] (n=bjorn@rockbox/developer/Zagor)
09:36:07linuxstbJdGordon: PCM data is normally sent to the DAC in 32KB chunks - using DMA on Coldfire, and a fast interrupt (FIQ) feeding a FIFO on PP (and I think other ARMs).
09:37:07linuxstbI don't know about the other parts of pcmbuf.[ch], but pcmbuf_callback is the "get_more()" function which is called by either the DMA interrupt handler or the FIQ to get a pointer to next 32KB of data to transfer - that code runs in an ISR.
09:37:19JdGordonand the codecs look like they all request <1KB buffers to fill at a time?
09:38:20 Quit olegfink (Read error: 110 (Connection timed out))
09:38:33 Quit pixelma2 ("-")
09:38:34linuxstbThe codecs will generally fill one audio "frame" at a time - that framesize depends on the codec (and audio format).
09:38:44 Join pixelma [50] (i=pixelma@rockbox/staff/pixelma)
09:39:27linuxstbBut then it will go via the DSP layer...
09:39:45JdGordonsim seems to go into a infinite loop after starting playback and im not sure why :(
09:40:04linuxstbHave you done a "make clean" recently?
09:40:23JdGordonyep, and a mutilated code... its new code which is brocken
09:40:32 Join PaulJam [0] (
09:41:14linuxstbWhat are you planning?
09:42:43JdGordonhaving a go at a redone pcm buffer
09:44:45 Join olegfink [0] (n=olegfink@
09:44:54linuxstbRedone how?
09:45:25JdGordonmultiple channels, so we can have voice while paused
09:45:54 Join Thundercloud [0] (
09:50:43 Quit jhulst (Remote closed the connection)
10:01:03 Join mcuelenaere [0] (n=mcuelena@rockbox/developer/mcuelenaere)
10:01:07 Join olegfink^ [0] (n=olegfink@
10:02:10 Quit olegfink (Read error: 104 (Connection reset by peer))
10:02:21 Nick olegfink^ is now known as olegfink (n=olegfink@
10:03:05 Quit Seed ("cu, Andre")
10:17:09 Quit kachna ("Konversation terminated!")
10:21:16linuxstbGodEater: Talking about the request here - - I think it's feasible (but slow without dircache enabled) to build playlists based on more than just the current directory when a file is played. But I'm not sure if it's something I would ever want.
10:23:34 Join Bigbambi_L [0] (i=86ceaf34@gateway/web/ajax/
10:28:24 Quit Thundercloud (Remote closed the connection)
10:29:24GodEaterlinuxstb: it seems a very "niche" request to me
10:29:44GodEaterbut given the number of features of Rockbox I don't use, I may not be a typical user either ;)
10:33:02*JdGordon has half a sec of garbled sound :D
10:33:18ZagorJdGordon: yay!
10:36:47 Quit evilnick (" ajax IRC Client")
10:38:48 Quit cool_walking_ (" ajax IRC Client")
10:38:48 Quit perrikwp (" ajax IRC Client")
10:42:48 Quit BigBambi (" ajax IRC Client")
10:43:55 Quit Bigbambi_L (Read error: 104 (Connection reset by peer))
10:44:26JdGordonthis is sooo fucked... but i can definatly make out the music!
10:45:06 Join BigBambi_L [0] (i=86ceaf34@gateway/web/ajax/
10:45:34*linuxstb expects a "Gentleman (and Ladies)..." email soon
10:45:54JdGordoni dont think so...
10:46:36Llorean"We have sound, again, but this time..."?
10:46:58JdGordonyeah, something like that
10:47:59JdGordoncan anyone tell me whats probably wrong in this"> ?
10:48:43JdGordonassuming I should be able to take the pcm directly from the codecs and put it on the dac?
10:50:55linuxstbJdGordon: The codecs generally don't output 16-bit samples.
10:54:38JdGordonhmm... no, I mean once its finished going through the ci->callback (so after dsp_process()..)
10:59:48 Quit mcuelenaere ()
10:59:56JdGordonand I guess the speed is because I'm not giving the pcm callback 32K worth of data?
11:03:40linuxstbHow are you planning on mixing the channels?
11:03:51***Saving seen data "./dancer.seen"
11:06:33JdGordonim not sure
11:06:44JdGordonI was hoping to deal with chunks < 32K
11:07:17JdGordonatm I've got playback putting 512bytes of data in chunks and then passing them straight in
11:09:57linuxstbWhat's the reason for reducing the chunk size?
11:10:43JdGordonthe 512?
11:11:34 Join einhirn [0] (
11:14:32JdGordoneasier to deal with small chunks.. thats about what the codecs are asking for... I'm fiddling now to get that up to 32K
11:33:09*JdGordon has hit a buffer overrun and cant figure it out :'(
11:49:28 Join faemir [0] (
11:51:14 Join moos [0] (
12:00:07JdGordonZagor: seems the paypal link has been broken on the wiki pages?
12:07:55 Quit BigBambi_L (" ajax IRC Client")
12:11:09 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
12:18:24 Join BigBambi [0] (i=86ceaf40@rockbox/staff/BigBambi)
12:28:17scorcheerm...trying to install rockbox on my new H320...all of the links on seem to be dead?
12:28:48scorcheif they are all dead and rasher has them mirrored, why do we have those links still?
12:29:25JdGordon still works
12:29:44rasherDo I in fact have them all mirrored? I stopped paying attention to them a while ago..
12:30:24scorcheJdGordon: yes, but according to the manual that version is not supported and i get "could not detect firmware type" from rbutil using that
12:31:22scorcherasher: well, i just downloaded one from you, so i hope so ;)
12:32:47rasherI'm just wondering if I have the "full set"
12:33:01 Nick fxb__ is now known as fxb (
12:33:10 Quit nplus (Read error: 101 (Network is unreachable))
12:33:10rasherI don't mind hosting them, but maybe someone should keep a copy, just in case.
12:33:12scorcherasher: it sint the "full set", but it should be enough to install rockbox
12:34:40rasherIt'd be a pretty silly way to lose the ability to Rockbox irivers if I somehow lost them and no one has a copy
12:36:25scorcheit certainly would..
12:41:06linuxstbDo we even need to patch the OF any more? Couldn't "someone" build a .hex file just containing Rockbox code?
12:42:12*linuxstb forgot the h300 isn't flashed yet...
12:43:31scorcheeither way, it isnt very nice for new installers to come up against a wall of korean 404 messages and then perhaps mission rasher's small mirror link
12:43:54scorcheshould we just toss those links? (they should still be in the wiki page history...)
12:44:34rasherlinuxstb: I guess some people still want dual-boot
12:44:51rasherscorche: I don't see why not, really
12:45:04linuxstbrasher: I know... I guess I'm not suggesting it as the only install method, but as an option.
12:45:40 Quit nuonguy ("This computer has gone to sleep")
12:45:44linuxstbMaybe we could even rolo .hex files - I don't know how the irivers boot though, or the format of those files.
12:48:25ZagorJdGordon: fixed now. thanks.
12:53:38linuxstbrasher: I've just made some edits to the UISimulator wiki page - are you (and others...) happy for me to call your sim downloads "unofficial (but trusted)" ?
12:54:53B4gdersounds fine to me
12:55:58 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
12:57:01linuxstbAre we still against the idea of official sim downloads?
12:57:10rasherlinuxstb: I don't mind, if everyone else is fine with it
12:58:45rasherPersonally I think official (daily) sims would make sense
12:59:32linuxstbThe only downside I can think of is that we would get more support requests about them. But on the other hand, maybe getting more sim bug reports would be a good thing.
12:59:49 Join casainho [0] (
13:00:16casainhohello Rockbox people :-)
13:00:48rasherlinuxstb: The lack of sims already generates support requests, doesn't it?
13:01:16linuxstbrasher: I haven't really noticed any.
13:01:30rasherMaybe I'm just imagining it
13:01:36casainhoĨ would like to ask some questions about diffs: how should I organize files on my PC to work on RB code and make diffs? having two working directories, almost equals and one I should checkout always before making a diff?
13:01:42B4gderwell, assuming the sims are popular downloads we should probably make them official
13:02:01linuxstbcasainho: The easiest way is to work with only one directory - a checkout of the Rockbox SVN.
13:02:16B4gderor multiple checkouts...
13:02:19BigBambiI haven't noticed many requests about the lack of sims - it is just that from time to time people ask if they can preview WPSs, so (I personally) point them at rasher's sims
13:02:43BigBambiI think it is just that most people don't really know they exist
13:02:46 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
13:02:48linuxstbrasher: Do you have any download numbers for your sims?
13:03:31casainholinuxstb: and then I checkout RB trunk, and, how do I do diff? If diff needs to diferent files?
13:03:53***Saving seen data "./dancer.seen"
13:03:58scorchecasainho: have you read the WorkingWithPatches and SimpleGuideToDeveloping wiki pages?...they explin this
13:04:00rasherB4gder, linuxstb: I'll check my download numbers
13:04:37scorcheand SimpleGuideToCompiling..
13:04:38casainhoBigBambi: I already read that Working... I am reading it now
13:04:55BigBambiWell that explains how to do patching
13:06:20rasherB4gder, linuxstb: 2110 Ipod Video sims downloaded last month, 1372 e200, 1200 nanos.. and so on
13:06:20casainholooks very simple to make diffs, however now I have 2 different directories, source trees on my PC, but one of them is from an old version of RB, can it be a problem?
13:06:48B4gdercasainho: you should 'svn update' first to sync with Rockbox reality
13:06:59B4gdercasainho: then 'svn diff'
13:07:12B4gdereasy peacy
13:07:22casainhoor I should have only on one directory the only files I need to do a new port, and diff against the RB checkeout source tree?
13:08:19casainhoB4gder: hmmm... so I need to use svn diff, " diff -u -r olddir newdir > patch"??
13:08:23B4gdercasainho: 1) checkout rockbox 2) edit everything you need 3) svn diff
13:08:32B4gdersvn diff > patch
13:08:53 Quit feisar (Read error: 104 (Connection reset by peer))
13:09:47casainhoB4gder: since I am trying to do a new port, everytime I need to make a patch, does I need to edit the all port files?
13:09:59B4gderthe changes remain
13:10:00rasherI'm surprised that c200 is #4 sim download
13:10:02casainhoseems not practical
13:10:10B4gdercasainho: svn diff doesn't change anything in your files
13:10:14B4gderyou continue editing
13:10:21B4gderand svn diff every time you want to make a patch
13:10:28B4gderand svn update to keey synced
13:10:57 Quit faemir (Read error: 110 (Connection timed out))
13:11:13B4gderand as soon as you do things "right", we can commit the starts of the port to svn to allow more people to help out and see things
13:12:31casainhoB4gder: and files that are not new, files that I need to edit from actual RB source, what will happen to them? everytime I do svn update, will my chnges be lost to that files?
13:12:56B4gderdid you really and truly read those docs?
13:13:12BigBambi13:08B4gdercasainho: svn diff doesn't change anything in your files
13:13:15B4gdersvn update MERGES changes
13:13:29*BigBambi misread
13:13:34B4gderso your changes remain while you get the ones others have made
13:13:41*BigBambi apologises and decides to keep out of it
13:14:10B4gdercasainho: if this was a stupid way of working, why would hundreds of rockbox hackers do it like this?
13:14:27 Join feisar [0] (
13:14:50B4gder... all this is source code version control basics really if you ask me
13:15:02casainhoB4gder: I am not putting that on question. I was just afraid of losing my changes
13:15:20B4gdercasainho: yes, and that is explained in the docs we've referred you to
13:16:27casainhookok - so, how would I do to update all source , without merging changes? (I was afraid that svn update did that)
13:16:45B4gderwithout merging?
13:16:51B4gderthat's not updating at all
13:17:07B4gderso wipe everything and checkout fresh
13:17:36B4gderor 'svn -R revert .' first (kills all local changes) and then 'svn update'
13:18:08casainhookok - my question is not to never kill my local changes :-)
13:18:39B4gdersorry, I don't understand
13:18:48scorchehuh? are afraid that your local changes are going to get wiped upon update?
13:18:59casainhothank you −− I will work on in and hope to soon commit the pacthes - thanks.
13:19:15B4gdercasainho: you can't commit, but you can post your patch in the patch tracker
13:19:53casainho"commit" to the patch tracker - okok - and yes, I was afraid of get wiped my local changes...
13:20:07casainhobye bye :-)
13:20:11 Quit casainho ("ChatZilla 0.9.83 [Firefox 3.0.3/2008092510]")
13:26:26linuxstbB4gder, rasher: I guess the issue of providing sims is complicated by the question of which host OSes do we want to support? Would we just want to provide Windows sims?
13:27:23B4gderwell, we could provide what we manage
13:27:39B4gderI bet windows users are the biggest audience
13:30:41JdGordonif its just for sims, wouldnt it be better to get more effort put into the wps editor?
13:30:55JdGordons/sims/wps checking
13:31:12B4gderah yes
13:31:18B4gderbut we can do both
13:34:53linuxstbJdGordon: How do we get more effort put into the wps editor?
13:35:14JdGordonwait for that elusive Mr Someone :(
13:35:49linuxstbIt was a sad SoC this year...
13:35:56 Join nplus [0] (n=nplus@141.25.Globcom.Net)
13:36:13JdGordoncan we somehow make more prominant on the front page? It was made pretty obvious from the floss interview that something like this is more needed as a super quick start install guide for new comers
13:36:46*linuxstb wonders where that page came from...
13:37:09 Quit feisar ("leaving")
13:37:25JdGordonits an old one
13:37:38*linuxstb likes the "Quick links" on the IPL wiki homepage -
13:37:50 Join f__ [0] (
13:37:58JdGordonr1 - 07 Jul 2008 - 22:08:53 - DominikRiebeling
13:43:10linuxstbI think some of the screenshots on that page are out of date though - e.g. I thought the "configuration error" message had been changed?
13:43:18 Join casainho [0] (
13:43:39JdGordonyeah, I tihnk so
13:43:50XavierGrwhat happened to the SoC student responsible for wps editor?
13:44:33linuxstbI haven't noticed him here since the end of the summer.
13:45:00XavierGrpff he took the money and left.... if only the project was done or at least in a decent state.
13:45:59XavierGrIMHO evaluation should be stricter
13:46:59XavierGrI remember seeing him the last 2 days before the deadline and as soon as he got the 'ok' he never appeared again
13:47:25linuxstbStaying after the deadline isn't a requirement though... He was passed based on what he had achieved.
13:49:29XavierGrlinuxstb: true, though if I am not mistaken he was unofficially prompted to continue his work because it was half baked
13:49:29 Join MajorC [0] (
13:49:41pixelmaI am really disappointed about this year's communication. It was even worse than last year and we wanted it to get better
13:50:02pixelmawith GSoC students I mean, of course
13:50:07 Join mc2739 [0] (
13:50:08XavierGrin the end he got 4500USD for a weeks work
13:50:18*linuxstb thinks we need to do two things next year - 1) work harder on thinking of better projects and advertising ourselves to try and recruit better students; 2) be stricter about enforcing constant community involvement - e.g. appearing daily in IRC
13:50:25JdGordoncan anyone see the potential stack overflow in ? if PCM_BLOCK_SIZE is 512 is segfaults, 2048 it doesnt...
13:51:13XavierGrlinuxstb: let's hope next year it goes better.
13:52:28JdGordonthat segfault is actually coming out as stack corruption...
13:52:34 Join robin0800 [0] (
13:53:22linuxstbXavierGr: I don't think it will be any better unless we put more effort into it as a project.
13:54:40linuxstbXavierGr: It all starts with the quality of the applications - for some reason, we don't seem to attact the best students.
13:54:42 Quit GodEater (" ajax IRC Client")
13:54:47pixelmayeah, I think so too
13:55:21linuxstbMaybe people aren't interested because we're not Linux...
13:55:46gevaertsHow do they know?
13:56:50linuxstbscorche, markun: Maybe this is an issue for the mentors conference - what can projects do to advertise SoC and try to attract good quality students? Or maybe it's simply that google don't pay enough...
13:57:44B4gderthis year was not a good payment year for Euro people at least
13:58:04JdGordoneven in USD its harldy considered enough for the amount of work thats expected
13:58:29B4gderno, but euro-wise it was nearly 20% down compared to last year
13:58:41casainhoworking on a new RB port, I am getting this problem: "make[1]: *** No rule to make target `button-target.h', needed by `/home/cas/Documentos/rockbox/build/firmware/backlight.o'"
13:58:59JdGordonit means that file is missing
13:59:37casainhook, I don't hav e button-target.h because I don't make it yet, however also i don't have backlight also, I can I skio the compiling of backlight.c?
14:00:12casainhoI am just starting the bootloader, I don't have yet buttons, nor backlight
14:01:06 Quit mc2739 ()
14:02:05JdGordonNico_P: you round?
14:02:21*linuxstb wonders if half the number of students, getting paid twice the amount of money each, would be more productive...
14:02:24B4gdercasainho: our bootloaders generally have button support
14:02:35B4gdercasainho: but you can disable or edit out anything you like of course
14:02:38Nico_PJdGordon: yes
14:03:07linuxstbcasainho: You may want to look at the files for the Sansa E200v2 port - that is also in an extremely early stage, and many files just contain empty functions.
14:03:39casainhoB4gder: later my port will also use button for shure, but now, I can I disable the build of that files associated to buttons and LCD?
14:04:01B4gderyou can disably exactly everything you want
14:04:03linuxstbcasainho: I would recommend creating the required files, but with empty functions.
14:04:24linuxstbThat way you a) know what needs to be done; and b) don't need to re-enable things that you'd disabled.
14:05:09Nico_PJdGordon: a bit...
14:06:36JdGordonNico_P: do you know codec_pcmbuf_insert_callback()? at all? I have a feeling that somewhere in there is killing my buffers and dunno why (my changes to pcmbuf.c to reimplement the absolutly required calls are at
14:06:43JdGordonif you have a few min....
14:06:45B4gderyeah, creating the correct file set with empty stubs is surely the best way to start a port
14:06:59 Quit MajorC ("Leaving")
14:07:26casainhookok - I will do that :-)
14:07:52Nico_PJdGordon: no, I don't know that code
14:24:12 Quit amiconn (Nick collision from services.)
14:24:18 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
14:26:05 Join n1s [0] (n=nils@rockbox/developer/n1s)
14:33:42n1sgevaerts: the broken usb detection on c200 with the new bootloader that you mentioned in FS9369, is it something that used to work? if so could you perhaps file a detailed bugreport in the tracker so that "someone" can check it out?
14:34:06gevaertsn1s: I very strongly suspect a bad cable
14:35:49n1saha, anyone else with a c200 around that could test the bootloader in FS #9369 especially usb detection?
14:36:40 Join Rob2223 [0] (
14:37:32 Quit Rob2222 (Connection reset by peer)
14:37:50 Join GodEater [0] (i=c2cbc962@gateway/web/ajax/
14:39:27 Part B4gder
14:39:52 Join LambdaCalculus37 [0] (i=44a04303@gateway/web/ajax/
14:43:33 Join robin0800_ [0] (
14:49:02*JdGordon is getting closer to having actual music playing!
14:49:57LambdaCalculus37JdGordon: Sweet! Which device?
14:50:17JdGordonfiddling with the pcmbuf
14:54:42LambdaCalculus37JdGordon: You mentioned that the OS X Sansapatcher I rolled and posted at FS #9369 worked for you, correct?
14:55:31LambdaCalculus37Cool. Now we just need a Windows version and we're done. :)
14:56:19JdGordonlinuxstb: you busy?
14:56:27linuxstbI should be...
14:56:34 Quit robin0800 (Read error: 110 (Connection timed out))
14:56:43*JdGordon takes that as a no?
14:57:42JdGordoncodec_pcmbuf_insert_callback() should end up with usable pcm? ive finnaly got it working with full 32K buffers but the audoi is still playing way too fast and sounds wierd
14:58:30linuxstbIs that the function the codecs themselves call via the codec API?
14:59:01linuxstbThen that's the function where you can get PCM in lots of different formats.
14:59:03JdGordonits got a nice dsp_proccess() call in which which i assume makes it usable
14:59:29n1sLambdaCalculus37: does usb detection work on your c200 with the new bootloader, when connecting while it's off?
14:59:53Nico_PJdGordon: whilst you're fiddling with PCM code, do you think there could be a way to get info from the PCM data in order to show something like a spectrogram?
15:00:15LambdaCalculus37n1s: My c200 boots into the OF when connected via USB while powered off, but it doesn't establish a USB connection. It just charges.
15:00:18Nico_Pof course that means we need to look at the data going into the DAC, and not what the codec outputs
15:00:35n1sLambdaCalculus37: is that how it's supposed to work?
15:01:06casainhoI am getting this error: make[1]: *** No rule to make target `button-target.h', needed by `/home/cas/Documentos/rockbox/build/firmware/backlight.o'
15:01:13linuxstbNico_P: That sounds similar to the peakmeter. You should look to see how that's handled
15:01:19casainhohowever I have that button-target.h file
15:01:23n1sJdGordon: are you adding support for playback of more sample rates?
15:01:25 Quit PaulJam (".")
15:01:30casainhowhere can be the error?
15:01:40Nico_Plinuxstb: I have, and it seems to only get the peak values from a dedicated func
15:01:50linuxstbcasainho: Then you have it in the wrong place. Can you first post your changes to tools/configure somewhere?
15:01:55Nico_Psame for the vumeter and oscilloscope plugins
15:02:17JdGordonn1s: no, I'm trying to make a pcm buf that could be used to have voice playing while paused... atm just trying to get corect sound out of the codecs without anything fancy like fading...
15:02:27JdGordonI have sound, but its fast and a but garbeled
15:02:51n1sah, yes that would be a great improvement of the voice ui indeed :)
15:02:58JdGordongarbled is probably the wrong word
15:03:02LambdaCalculus37n1s: I don't think it's supposed to act like that. The connection works if I disconnect the Sansa and reconnect it. Maybe other c200 owners can give insight as well?
15:03:10Nico_PJdGordon: what kind of design are you aming for? a mixer with several entries?
15:03:13JdGordonn1s: do you know pcmbuf.c at all?
15:03:18JdGordonNico_P: yeah
15:03:25n1sJdGordon: not at all i'm afraid
15:03:32linuxstbNico_P: IIRC, the peakmeter values are calculated somewhere in the audio drivers in firmware/. If you don't do it there, you have to take account of the latency...
15:03:41casainholinuxstb: here:
15:03:54***Saving seen data "./dancer.seen"
15:04:06linuxstbcasainho: And where did you put your button-target.h file?
15:04:17*JdGordon cant see anywhere in pcmbuf.c where the pcm data is mangled for output
15:04:23Nico_Plinuxstb: maybe the drivers could be modified to export a few more values...
15:04:40jhMikeSis it gonna be too annoying if I have to force incompatibility of codecs and plugins to get rid of kernel cruft? :)
15:04:48casainholinuxstb: here: firmware/target/arm/at91sam/rockboxplayerlittle
15:05:10n1sjhMikeS: I don't think that is a problem
15:05:17linuxstbjhMikeS: No, I don't think it's ever an issue to break codec/plugin API compatibility... The world isn't full of third-party plugins...
15:05:31JdGordonjhMikeS: you know the pcm code dont you/
15:05:43n1sJdGordon: in dsp.c maybe?
15:05:48jhMikeSlinuxstb: I just try to ask before making a nuisance of myself :)
15:05:59jhMikeSJdGordon: somewhat
15:06:06linuxstbcasainho: That's the problem - the configure file says "rockboxplayer-little", but you've called the directory "rockboxplayerlittle" (t_model)
15:06:11 Nick f__ is now known as feisar (
15:06:28jhMikeSJdGordon: I don't think anyone does in full if you want my honest opinion :)
15:06:33 Quit feisar ("leaving")
15:06:35casainholinuxstb: thanks!! :-) I will rectify then :-)
15:07:40jhMikeSJdGordon: but....what particular bit were you wondering about? perhaps I can help.
15:08:12JdGordonjhMikeS: I'm trying to reimplement the pcmbuf so it doesnt do anything fancy (i.e fading), just to be able to get pcm from the codec and put it on the output... is the current patch
15:08:28JdGordonit works, but the output is really fast, and sounds like shit... and i dont know why :(
15:09:05JdGordonalso, line 106 of that patch, if that +4096 isnt there I get a stack smashing and I dunno why, but thats another issue
15:09:40casainholinuxstb: unfortunately, I am getting the same error :-( −− I did changed it on configure file and did "../tools/configure" again...
15:09:52jhMikeSJdGordon: ok, one moment I'll look
15:09:57linuxstbcasainho: Try a "make clean"
15:09:57JdGordonthanks :)
15:12:09casainholinuxstb: even after make clean, the error stills there :-(
15:12:41JdGordonjhMikeS: the patch is supposed to keep giving the codec a 512-1k buffer (whatever it asks for) and keep builing up groups of 32K which are then passed as is to the DAC
15:13:53jhMikeSwhy so large a chunk?
15:14:15JdGordonapparenrlty thats the size the hardware likes... but that can be shrunk
15:14:28*JdGordon would rather get it working at all before fiddling with sizes
15:14:50jhMikeSThe hardware shouldn't care what size it is other than aligned and a full pair of samples
15:14:52casainholinuxstb: here is the diff:
15:15:36 Join stoffel_ [0] (
15:16:15linuxstbcasainho: That diff doesn't include a "button-target.h"
15:16:42casainholinuxstb: eheh - forgot to do svn add :-)
15:17:01jhMikeSJdGordon: doesn't the buffer do 8192-byte chunks?
15:17:17*LambdaCalculus37 wonders if Mr Someone can make a Windows Sansapatcher so that all versions are up-to-date
15:17:55JdGordonjhMikeS: I'm not sure... the numbers can be fiddles with later
15:18:03casainholinuxstb: here:
15:18:12JdGordoni'm completly bypassing all the exsitsng pcm code
15:18:55Nico_PjhMikeS: nice to see you around :)
15:19:13JdGordon8k chunks seems to work just as well/badly as 32 :p
15:20:44Nico_PJdGordon: one nice application I see with the mixing is having plugins that can output audio while music is playing
15:21:04jhMikeSNico_P: hello there :) things move in cycles and so this has been the off cycle lately.
15:21:05 Join Schmogel [0] (
15:21:22Nico_Pit's quite the same here too
15:23:05jhMikeSI defintely don't see myself exiting permanently though but just compelled to visit other planes of being :)
15:23:58linuxstbcasainho: You patch is also missing config-rockboxplayerlittle.h
15:24:55linuxstbcasainho: If you type "svn status" from within your firmware/ directory, it will list any files you haven't done "svn add" on.
15:26:53casainholinuxstb: okok - looks like I didn't missing nothing now:
15:27:00JdGordonjhMikeS: any ideas?
15:27:28moosjhMikeS: wow, hi, indeed nice to see you around again :)
15:29:33linuxstbcasainho: That patch fails with an error about "backlight-target.h"
15:30:12jhMikeSmoos: hello. good to see you here too. you weren't here when I checked-in last.
15:30:23casainholinuxstb: patch failing? why? don't understand...
15:30:39jhMikeSJdGordon: not yet. I think I have to apply it and have a look.
15:31:11linuxstbcasainho: I mean when I try to compile Rockbox (rockboxplayerlittle, bootloader build) with your patch, I get an error saying backlight-target.h doesn't exist.
15:31:49moosjhMikeS: really busy here too with the real life...but always checking here and here about rockbox when I can
15:32:04linuxstbcasainho: You also need to define AT91SAM9260 in config.h - e.g. "#define AT91SAM9260 9260" (similar to all the other CPU defines).
15:32:39casainholinuxstb: oh, thats true :-) - sorry.
15:35:28casainholinuxstb: okok, I don't have now that warnings about missing the "AT91SAM9260" is not defined −− I will take care of backlight-target.h :-)
15:38:29jhMikeSJdGordon: I'll give my op. as soon as I figure out what is supposed to be happening
15:39:21jhMikeSJdGordon: that block_count variable isn't safe like that (inc/dec between thread and int context)
15:39:36JdGordonyeah, I know
15:40:14LambdaCalculus37n1s: USB detection of my c200 works in Windows XP.
15:40:31JdGordonand write_complete should actually be doing something useful, but it breaks if I try to change that
15:40:53*linuxstb hopes any pcmbuf rework will include support for using different hardware samplerates...
15:41:00*LambdaCalculus37 summons the mighty gevaerts
15:41:36jhMikeSJdGordon: that function should make the new data available to the callback
15:42:05JdGordonyeah, its safe how it is now, but will need fixing later
15:42:20JdGordonbut yeah, I just want to get sound working properly first :p
15:42:37*gevaerts feels summoned
15:43:14casainholinuxstb: now I need to have the adc-target.h, however I think I will not use any ADC on my player... do I really need to have that file?
15:43:23LambdaCalculus37gevaerts: Want to roll a Windows build of Sansapatcher using the new bootloaders from FS #9369?
15:43:34LambdaCalculus37That way we can have all versions up to date and they can go onto the build server.
15:43:35 Quit nplus (Read error: 104 (Connection reset by peer))
15:43:58linuxstbcasainho: adc is usually used for things like buttons and battery status. It's not audio ADC.
15:44:45casainholinuxstb: ok, but even like that, I will not use ADC for buttons, will be direct I/Os
15:44:56 Join J-23 [0] (
15:45:00casainholinuxstb: is there a way to skip that file?
15:46:27linuxstbcasainho: I think all existing targets have an adc - so if I was you I would just create a dummy file for now.
15:46:29 Join nplus [0] (n=nplus@141.25.Globcom.Net)
15:46:38gevaertsLambdaCalculus37: do I need to edit something to get the right version numbers?
15:46:43casainholinuxstb: ok :-)
15:49:52 Join dabujo [0] (
15:50:23LambdaCalculus37gevaerts: Sansapatcher should be version 0.7 with 5.0 bootloaders, so you may have to edit the version string.
15:50:41LambdaCalculus37n1s: Does gevaerts have to edit the version string for the sansapatcher source, or is that already done in SVN?
15:50:52 Part LinusN
15:51:29*LambdaCalculus37 checks
15:51:55LambdaCalculus37gevaerts: Edit rbutil/sansapatcher/main.c to add the appropriate version string.
15:53:12*LambdaCalculus37 will try to get a wireless connection later to bump the version string for Sansapatcher in SVN
15:53:30*jhMikeS is really confused by the new code (and perhaps out of practice) :\
15:54:57gevaertsLambdaCalculus37: done
15:55:46LambdaCalculus37gevaerts: Cool.
15:58:02 Quit Bagder ("*plopp*")
16:02:27 Join pvbcharon [0] (n=charon@
16:06:08*jhMikeS will just listen to the output which might say more about what's happening
16:08:23jhMikeSJdGordon: will it blow up if voice happens to be on?
16:08:48JdGordondunno.. probably completly ignored
16:08:58JdGordonplay it in the sim
16:10:22jhMikeSyeeha, data abort
16:12:02JdGordonon target?
16:12:11jhMikeSyes, but voice was on
16:12:20JdGordonyeah, just use the sim...
16:14:03jhMikeShrm, no sound but the seconds go by pretty fast
16:15:16casainhocan someone please help me on how should I include this file so I don't get this error? : CC target/arm/at91sam/rockboxplayerlittle/crt0.S
16:15:17casainhotarget/arm/at91sam/rockboxplayerlittle/crt0.S:40:73: error: firmware/target/arm/at91sam/rockboxplayerlittle/at91sam9260.h: No such file or directory
16:16:37casainhoI need to include file "at91sam9260.h" and I did put it on dir "target/arm/at91sam/rockboxplayerlittle/" - should it stay on another dir?
16:19:38jhMikeSJdGordon: one thing is that the count parameter is the number of samples (mono, or stereo pairs) and not bytes
16:20:11JdGordonoh... OHHH>>....
16:20:25jhMikeSthe buffer data should be in int16_t
16:22:18jhMikeSactually for the pcm buffer it's always 16-16 packed pairs so SIZE/4
16:23:35JdGordonAH.. thats why its doing *count<<2
16:23:49preglowJdGordon: oy, could you zip up your os x dev tree for me? absolutely no luck building compilers here
16:24:59n1spreglow: do you have any other arm-elf assemblers installed? iirc i saw that problem somewhere and it was because the wrong assembler was used
16:25:08JdGordonpreglow: umm, ok, lemme have a look
16:25:19preglown1s: nothing of the sort, but that does sound very much like the error i'm having
16:25:22linuxstbcasainho: It should probably go in firmware/export/ and you include it with "#include <at91sam9260.h>"
16:26:17preglown1s: can't ever recall having installed anything of the kind, and nothing prefixed "arm" exists here, but yeah, it does seem like the wrong assembler gets used
16:26:36preglown1s: where did you see this problem mentioned?
16:26:48casainholinuxstb: okok, I already did that, and it is working, don't know if is a good organization on RB source... −− almost know I am the bootloader building :-)
16:27:36n1spreglow: don't remember exactly but googling for "crtstuff" seems to bring up some hits that could be it
16:27:37jhMikeSJdGordon: yeah. counts rather than bytes made the DSP much simpler.
16:28:18rasherlinuxstb: I don't think providing only windows sims would be terrible
16:28:50rasherlinuxstb: I haven't had any requests for Linux sims while hosting the windows sims (but that's of course not representative of what Rockbox might get)
16:29:15JdGordonpreglow: ok.. I *thought* i had the CC's built... cant find them so im trying to build arm now to see if it works
16:29:21LambdaCalculus37n1s: I'm going to bump the version number for rbutil/sansapatcher/main.c in SVN.
16:29:56n1sLambdaCalculus37: ok
16:30:44 Join obo [0] (n=obo@rockbox/developer/obo)
16:30:51n1spreglow: here it is however it doesn't seem a solution was found there...
16:30:58jhMikeSJdGordon: HINT: The variable would be "size" rather than "count" if it were bytes, no? (/me renamed it for that reason anyway) :)
16:31:34JdGordoncount/size/blaa.. it all sounds the same to me :p
16:32:39*jhMikeS says to call it "blaa" then :)
16:33:44 Quit pvbcharon ()
16:37:11 Join kachna [0] (
16:37:12casainholinuxstb: I have now an error on the LDS script, can you please give a look? here is the patch file:
16:38:43 Join tvelocity [0] (
16:39:07JdGordonpreglow: ok, wierd... died... guess i cant build it :p
16:39:54JdGordonjhMikeS: ok, but size in the pcmbuf_callback() is in bytes?
16:40:36jhMikeSJdGordon: yes, the hardware takes bytes
16:44:44 Join PaulJam [0] (
16:48:50JdGordonjhMikeS: is closer to what it should look like?
16:54:53 Join dghohens [0] (n=dghohens@
16:55:07JdGordonoohhh... its almost correct :)
16:55:29jhMikeSJdGordon: Looks closer
16:56:03JdGordonstill have to add 4K onto the end of the buf though or it overflows somewhere
16:56:44jhMikeSthat should be found out no doubt
16:57:03JdGordonwell yeah :p
16:57:51JdGordonsetting it to 32k samples comes out almost perfect
16:58:14jhMikeSprobably before going anywhere. the codec won't go over but it could be buffer fragmenting perhaps?
16:59:01dghohensHey guys, I'm having a problem with Rockbox. I just installed it on my 5.5 Gen 30 GB Video iPod, and I put songs on it. The database only shows songs from titles starting with S to titles starting with Z or #. Also, when I try to play files, they don't play. Earlier, when I tried to play files, I got a high pitched beep when the file was supposed to be playing. Do you guys have any idea what my problem could be?
16:59:02JdGordonshouldnt be.. that 32k is grouped, and It shouldnt be able to lose a block...
16:59:48n1sdghohens: what kind of files?
17:00:02jhMikeSso even if one sample goes into the 32k block, the whole 32k block is used?
17:00:35dghohensn1s: I tried m4a and mp3
17:00:49linuxstbdghohens: Do the files play if you access them via the "Files" menu?
17:00:50n1sdghohens: and none of them play?
17:01:01JdGordonjhMikeS: that block will be filled
17:03:21 Quit Zagor ("Client exiting")
17:03:25JdGordonthere is still some artifact somewhere though
17:03:31JdGordonI can hear a click sort of sound
17:03:32dghohensn1s: I haven't had time to try to play all 1000 something files, but every one I've tried did not work. I had it playing yesterday, it might have been that it just likes m4a files. I thought m4a were the problem, so I put all mp3's on there. I'll put m4a's back on and check
17:03:55***Saving seen data "./dancer.seen"
17:04:44n1sdghohens: mp3 files with exotic tags can cause problems, do you have large embedded album art, lyrics or something like that in your tags?
17:05:09dghohensn1s: also, when I plug it into my computer, Rhythmbox shows it as having files a-z on there. I'm not aware of album art or lyrics in my tags.
17:06:21n1sdghohens: could you link to or post an example of a file that doesn't work somewhere?
17:07:32JdGordonjhMikeS: thanks for your help, im going to bed now... if you come up with anything else lemme know.. ill read the logs tomorow
17:07:34 Nick JdGordon is now known as JdGordon|zzz (n=jonno@rockbox/developer/JdGordon)
17:07:35dghohensn1s: Do you want me to just transfer you a file?
17:07:56linuxstbdghohens: When you say that the files "don't play" - what happens?
17:08:00n1sdghohens: you could try, not sure if it will work
17:08:50dghohensn1s: It stays at 0:00 and sometimes shows the file name. Under the file name, it shows (root) then under that, (root) again.
17:10:09linuxstbHave you checked that your filesystem isn't corrupt? e.g. with fsck.vfat ?
17:10:22dghohensn1s: I'm transferring you one of the songs that it shows the file name, then (root) (root)
17:10:39dghohenslinuxstb: I haven't... Would I just plug it in to the computer and check that with terminal?
17:11:26linuxstbYes, I forget the exact command though...
17:11:52n1sfsck.vfat ?
17:12:20linuxstbYes - what options (if any) do you need?
17:14:41 Join Slack [0] (
17:15:47dghohensfsck.vfat -a -d /media/DGHOHENS\'S/
17:16:00dghohensI tried that, it didn't change the problem.
17:17:07 Quit Slack (Client Quit)
17:17:09linuxstbI thought you had to specify the disk device - e.g. /dev/sdX2 ?
17:17:21 Join Slack [0] (
17:17:29dghohensOh... whoops. I'll try it that way.
17:17:56linuxstbYou should probably unmount first as well.
17:18:51 Join pierre- [0] (n=pierre@
17:21:07dghohenslinuxstb: How am I supposed to find out which letter to put in x's place?
17:21:44 Join bughunter2 [0] (n=Jelle@
17:22:31n1smount | grep "/media/DGHOHENS"
17:24:34dghohensand then what do I use for the fsck.vfat command?
17:25:38n1syou should get a line like "/dev/sda1 on /media/DGHOHEN" then you use /dev/sda1
17:25:47dghohensyeah, but what options do I use?
17:26:04dghohensfsck.vfat -a -d /def/sdg2/
17:26:08dghohenswould that work?
17:26:12linuxstbNo "-d"
17:26:20linuxstbJust fsck.vfat -a /dev/sdg2
17:26:36linuxstbAnd type "umount /dev/sdg2" first
17:26:57GodEateryes, scanning a mounted filesystem is a bad idea
17:27:22dghohens sudo fsck.vfat -a /dev/sdg2
17:27:22dghohensdosfsck 2.11, 12 Mar 2005, FAT32, LFN
17:27:22dghohens/dev/sdg2: 2069 files, 485509/1824490 clusters
17:27:26dghohensthat's my output
17:27:33dghohenswell, input, then output
17:28:15LambdaCalculus37dghohens: For future reference, can you use pastebin if you have to show us large amounts of data?
17:28:17linuxstbThat looks to me as if everything is fine.
17:29:15dghohensLambdaCalculus37: How do you use pastebin? Is that paste as plain text?
17:29:15*GodEater agrees with linuxstb
17:29:37GodEateryou paste you text there, and then paste the URL it gives you here
17:29:43LambdaCalculus37dghohens: That's pastebin.
17:29:58LambdaCalculus37It's also excellent for when you have to show stuff like code to people on IRC.
17:30:03GodEaterso we get one line from you in the channel, and can go off and look at your multi-line output in comfort on a website
17:30:38dghohensAhh... I see. Sure, I'll be sure to use that in the future.
17:31:18LambdaCalculus37dghohens: Actually, bookmark it. It's a great tool to have around.
17:32:05*linuxstb spots a 3 line paste, followed by 9 lines discussing pastebin... ;)
17:32:11*linuxstb made it 10
17:32:36dghohensOkay. So, any ideas on my problem? Would it fix it if I just used itunes to restore it and then reinstall rockbox?
17:34:24linuxstbdghohens: I would try 1) Resetting my settings (it's a menu option somewhere...); 2) Ignoring the database and just trying to play files from the Files menu.
17:34:48 Join Nibbl [0] (
17:35:00LambdaCalculus37dghohens: Restoring the iPod with iTunes is what we call a "last ditch resort". That's only to be done if all else fails.
17:37:50dghohenslinuxstb: I'm not able to access any songs from the files menu. I only have 3 folders available there, Calendars, Contacts, and Notes. If I put the files on the ipod by clicking and dragging, instead of using Rhythmbox, would they show up under files?
17:38:23n1sdghohens: itunes puts files in a hidden folder
17:38:39linuxstbdghohens: Yes. They are currently in a hidden folder called "iPod_Control". You need to change the "view files" setting in Rockbox to "all" in order to see them.
17:39:38dghohensAwesome... that's doing the job nicely
17:41:28dghohensThank you guys so much for your help, it works perfectly now.
17:44:13 Part dghohens
17:45:21 Quit casainho (Remote closed the connection)
17:48:16 Quit SmallR2002 (Read error: 110 (Connection timed out))
17:48:48 Join SmallR2002 [0] (
17:53:08 Join culture [0] (
17:57:28 Quit petur (
17:57:28 Quit blkhawk (
17:57:28 Quit axionix (
17:57:28 Quit lordpil (
17:57:28 Quit advcomp2019 (
17:57:28 Quit ameyer (
17:57:28 Quit gkffjcs (
17:57:28 Quit Thedjatclubrock (
17:57:28 Quit Llorean (
17:57:28 Quit jhMikeS (
17:57:28 Quit basti (
18:00:57NJoinpetur [50] (n=petur@rockbox/developer/petur)
18:00:57NJoinblkhawk [0] (
18:00:57NJoinaxionix [0] (
18:00:57NJoinlordpil [0] (
18:00:57NJoinThedjatclubrock [0] (
18:00:57NJoinadvcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
18:00:57NJoinameyer [0] (
18:00:57NJoingkffjcs [0] (n=john-cha@
18:00:57 Join Llorean [0] (n=DarkkOne@rockbox/administrator/Llorean)
18:00:57NJoinjhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
18:00:57NJoinbasti [0] (n=basti@
18:01:01 Quit pixelma (Read error: 110 (Connection timed out))
18:04:34 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
18:04:41 Quit linuxstb (Nick collision from services.)
18:04:43 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
18:05:33 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:05:59 Quit petur ("work->home")
18:12:52 Quit BigBambi (" ajax IRC Client")
18:15:14 Join faemir [0] (
18:24:07 Quit robin0800_ (Remote closed the connection)
18:26:40 Join karashata [0] (n=kimi@
18:33:31 Join perrikwp [0] (i=98213619@gateway/web/ajax/
18:34:20 Quit pierre- (Connection timed out)
18:36:45 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
18:37:20 Quit AhtiK (Connection timed out)
18:37:44 Nick fxb is now known as fxb__ (
18:39:40 Join Bagder [241] (n=daniel@rockbox/developer/bagder)
18:43:06 Join wpyh [0] (n=william@
18:45:06 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
18:48:15 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
18:48:20 Join nuonguy [0] (
18:48:32 Quit linuxstb (Nick collision from services.)
18:48:34 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
18:54:25 Quit Nibbl (Read error: 113 (No route to host))
18:54:42 Join Strife89 [0] (n=michael@
18:56:27 Join Nibbl [0] (
18:58:50 Quit tvelocity ("Αποχώρησε")
18:59:17 Join Thundercloud [0] (
19:01:55 Join bertrik [0] (
19:03:56***Saving seen data "./dancer.seen"
19:04:21 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
19:06:09 Quit moos ("Rockbox rules the DAP world")
19:06:39 Join miepchen^schlaf [0] (
19:06:40 Quit GodEater_ (Client Quit)
19:11:17 Quit HellDragon (Client Quit)
19:12:34 Join GodEater_ [0] (n=ge@rockbox/staff/GodEater)
19:19:27 Join karashata_ [0] (n=kimi@
19:19:27 Quit karashata (Read error: 104 (Connection reset by peer))
19:20:26 Nick karashata_ is now known as karashata (n=kimi@
19:22:05 Quit Nibbl (Remote closed the connection)
19:24:26 Quit karashata (Read error: 113 (No route to host))
19:24:31 Join karashata [0] (n=kimi@
19:30:53 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
19:36:25 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
19:39:08 Join Nibbl [0] (
19:44:13 Join Horschti [0] (
19:44:57 Quit Horscht (Nick collision from services.)
19:50:08 Quit miepchen^schlaf ()
19:50:32 Join kb__ [0] (
19:53:08kb__Hi. I have a 'semi-bricked' sansa e280. It boots into the original firmware and rockbox without problems, but I think the filesystem became corrupt - if I tried to browse files under rockbox, it would freeze, and under the original firmware, it said "bad track" if I tried to play anything. I was going to format it, but it first started showing up as two 0-byte drives under Windows, and then when I next plugged it in under Linux, it
19:53:08kb__wasn't even getting a device node. Now I have it in recovery mode, and I have /dev/sdc, but fdisk -l thinks it doesn't contain a valid partition table. How can I restore it to functional?
19:53:17 Join miepchen^schlaf [0] (
19:53:23kb__(It was working correctly for months before this happened)
19:53:51 Join mf0102 [0] (
19:54:48domonokykb__: maybe take look at:
19:56:35linuxstbkb__: STOP ;) You shouldn't need to use recovery mode.
19:56:54kb__linuxstb: ok, so, what should I need to do?
19:57:14linuxstbIf the original firmware is working, I think you can simply use the "reformat" option that's available there. (I know the c2x0 has that option, not 100% sure about the e200 though).
19:57:18kb__domonoky: i've read it - I just want to do the least-drastic thing possible
19:57:21kb__the e280 doesn't have it
19:57:27kb__I spent last night checking
19:57:54kb__the usual way to format the e280, according to the manufactor's videos, is to mount it under windows and right-click, selecting 'format'
19:58:32linuxstbHave you tried running chkdsk on it? (or fsck.vfat in Linux) ?
19:58:47kb__I ran fsck.vfat yesterday
19:59:01kb__it was after that that the volumes started appearing as 0 bytes under windows
19:59:55linuxstbOK, then I would probably try the "sansa.fmt" instructions on the page domonoky linked you to.
20:04:47kb__Hm. 2 things
20:04:56kb__a) None of the mi4 files are for the e280
20:05:37kb__b) I do have a working firmware (2, actually) as I mentioned
20:06:22linuxstbI was referring to these instructions -
20:07:26kb__ok, it's formatting
20:09:01kb__ok, it's stopped saying 'formatting' and seems to have gone off; I'll try booting it now?
20:09:10linuxstbI guess so.
20:10:10kb__I still get this in dmesg
20:10:12kb__usb 1-3: config 128 interface 0 altsetting 0 endpoint 0x81 has an invalid bInterval 100, changing to 10
20:10:18kb__and no device nodes :(
20:12:45 Join Lear [0] (i=chatzill@rockbox/developer/lear)
20:14:00 Join meven [0] (
20:16:50 Quit perrikwp (" ajax IRC Client")
20:18:04kb__why on earth can it get a device node in recovery mode, but not normal mode? :/
20:20:01 Quit HBK (Read error: 104 (Connection reset by peer))
20:20:03 Join HBK- [0] (
20:20:04 Quit dabujo (Read error: 104 (Connection reset by peer))
20:20:32 Join dabujo [0] (
20:22:19 Join pierre- [0] (
20:26:07linuxstbkb__: Are you in the original firmware when inserting USB? Is it in MSC (UMS) mode (if the E200 has such a setting)?
20:26:27 Quit Nibbl ("Ex-Chat")
20:27:24kb__a) yes, b) yes (the e200 doesn't have such a setting, but the e280 does)
20:27:35kb__ok, it seems to work now (crosses fingers)
20:27:59kb__so, after I formatted, it was in mtp mode, and I was still getting the above message in dmesg
20:28:22kb__I tried plugging it into windows - and media player came up, etc, which made me go "doh, it must be in mtp mode", so I switched it
20:28:38kb__then, I plugged it into windows again, and got told it was an unknown peripheral
20:28:49kb__then you asked your questions and I answered them, while plugging it back into linux
20:29:04kb__and under linux, in msc mode, I'm now seeing the device nodes correctly again
20:29:37 Quit J-23 ("Flying cow pressed ^D on my keyboard.")
20:31:26kb__I copied over a file, unmounted, and unplugged it
20:31:48kb__and now the bootloader is failing - it can't find rockbox.mi4 or rockbox.e200
20:32:16linuxstbI assume you installed Rockbox again after reformatting?
20:32:23domonokyyou formated it, so you need to reextract the ofcourse
20:32:24kb__ah - ok, holding down the left arrow I can still boot into the old firmware - I fat-fingered that the first time I tried it
20:32:51kb__I did realize rockbox wasn't there anymore, but it wasn't falling back to the original, and I mis-fingered trying to boot the original - user error, sorry!
20:34:43 Join Humbucker [0] (n=pIRCuser@
20:34:55Humbuckerhi people!)
20:35:22kb__hi Humbucker
20:35:27Humbuckeri'm from Russia, sorry for my english
20:35:38Humbuckerhave a question
20:35:59Humbuckeri'm user motorola z6. MontaVista 2.6.10
20:36:36Humbuckerusers e2 run rockbox
20:36:50kb__wow, the new version of rockbox is shiny :D - thanks a million, linuxstb and domonoky :)
20:36:52Humbuckerbut on z6 this port not working
20:37:28Humbuckerbecause version of kernel different
20:37:31linuxstbWe know nothing about Rockbox on the motorola phones - it's done by someone who hasn't contributed his changes back to us.
20:37:36Humbuckeror something like that
20:37:44 Quit miepchen^schlaf (Connection timed out)
20:38:07Humbuckerit's bad....
20:38:19Humbuckeror not so good)
20:39:36 Quit nplus (Remote closed the connection)
20:39:58 Join miepchen^schlaf [0] (
20:40:25Humbuckermmm . launch rockbox on monta vista linux - is possible?
20:40:46 Quit Thundercloud (Read error: 104 (Connection reset by peer))
20:40:51 Quit nuonguy ("This computer has gone to sleep")
20:42:08 Quit kb__ ("Everything works. Weeeee.")
20:42:25linuxstbRockbox itself is a full standalone firmware (operating system and application), but the Rockbox simulator can be compiled to run on a Linux kernel. It currently needs SDL, but that could also be changed quite easily.
20:42:42linuxstb(the simulator needs SDL - "real" rockbox doesn't use/need it)
20:43:19 Quit BigBambi (Read error: 110 (Connection timed out))
20:44:38Humbuckerit mean, possible, i think... But how? were i can read something about it?
20:46:08 Join lasser [0] (
20:46:25linuxstbYou can start here -
20:47:03linuxstbAnd then you will probably want to look at the source code (assuming it's available) of Rockbox on the other motorola phones.
20:48:06Humbuckerthank's a lot
20:48:29Humbuckeri'm try)
20:48:46Humbuckerbye, need to go)
20:49:10Humbuckerthank's one more)
20:49:19Humbuckersorry for english)
20:49:25 Quit Humbucker (" ")
20:54:57 Quit bmbl (Read error: 104 (Connection reset by peer))
20:55:16 Join pixelma [0] (i=pixelma@
20:56:43 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
20:58:03 Quit Thedjatclubrock (Read error: 60 (Operation timed out))
20:59:04 Quit pierre- (Read error: 110 (Connection timed out))
21:02:01 Join bluebrother [0] (n=dom@rockbox/staff/bluebrother)
21:02:18 Join Thedjatclubrock [0] (
21:03:59***Saving seen data "./dancer.seen"
21:07:16 Join webguest87 [0] (n=cd98f70f@gateway/web/cgi-irc/
21:07:29 Quit webguest87 (Client Quit)
21:09:29 Join barrywardell [0] (n=barrywar@
21:13:00 Join tessarakt [0] (
21:29:23 Join {phoenix} [0] (
21:36:54*linuxstb hrmfs
21:38:49 Quit karashata ("I go, only to return again some time...")
21:43:15linuxstbIs anyone familiar with the bug that a playlist full of invalid files (i.e. audio files with valid extensions, but ones which are rejected by the metadata parser) causes Rockbox to freeze?
21:46:11LearNot really... Got a number?
21:47:05linuxstbI think it's the cause for this report -
21:47:09 Join Thundercloud [0] (
21:47:30linuxstbI'm pretty sure I've experienced this bug before, but searching flyspray, I can't find it.
21:48:23LearRepeat must be enabled for this to happen, I guess?
21:49:21linuxstbNo - that was my first thought, but when I disabled repeat, it still happens.
21:49:42linuxstbIt's easy to recreated, just copy two mp3 files into a dir, and rename them to ".ogg" (for example)
21:50:51linuxstbIn fact, if you put two files in a directory, rename the second one to a different audio extension, and play that second one, it freezes.
21:51:03linuxstb(this is with repeat disabled)
21:52:46LearDon't see how that last case could happen... :/
21:55:32linuxstbIn fact, Rockbox doesn't freeze at all - just the WPS is stuck at 0:00
21:57:55LearBut it responds to key presses?
21:57:58 Quit dabujo ("( :: NoNameScript 4.2 :: )")
21:58:19LearMaybe playback.c should check for strange values for last_peek_offset and return an error somewhere...
21:58:28linuxstbAlso, my last example was (I think) a different bug... If you take a .mp3 file with an id3v2 tag, rename it to .ogg, then badness happens.
21:59:09 Quit jhulst (Connection timed out)
21:59:36LearDoesn't medata.c select parser based on extension? If so, loading fails... Not sure if I'd call that a bug, as it is by design.
22:00:48linuxstbI take back what I just said - Rockbox seems to (correctly, as you say), reject that mp3 file renamed to .ogg in the same way as any other invalid file.
22:01:40linuxstbI should probably just stop confusing the issue.
22:02:07 Quit {phoenix} (Remote closed the connection)
22:02:18linuxstbIs a possible solution to remove invalid tracks from the playlist, once they've been detected as invalid?
22:02:23LearHm, no freeze in the sim with a buch of bad files, even with repeat all. Sits in wps at 0:00, which doesn't look very good, but...
22:02:57linuxstbLear: Yes, it isn't a freeze as such - just nothing happening...
22:03:06LearThey're already marked as invalid, because removing playlist entries is bad too (e.g., breaks bookmarking).
22:03:48LearSo it should splash "Empty playlist" or "No playable tracks" or something, and error out...
22:04:22linuxstbWhat about the case of one good and one bad file in the same directory?
22:04:26domonokybertrik: i can confirm that your i2c work for as3525 also works on a m200v2 :-)
22:04:44 Join mc2739 [0] (
22:05:13LearPlay the good one, if repeat is on, just exit wps otherwise?
22:07:19mc2739Hi - Is is possible to build an e200v2 dualboot firmware from svn? If so, what is the correct procedure?
22:08:42linuxstbLear: Yes.
22:10:13domonokymc2739: should be possible, but wont give you much, because rockbox isnt ready on this device.. but if you want to help with codeing feel free :-)
22:10:13bertrikmc2739, yes I think, but I haven't tried it myself (I don't have an e200v2)
22:10:38linuxstbIt doesn't do anything at all at the moment...
22:11:06 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
22:11:39bertrikyou do realise that this is all very early experimental stuff that may require you to open your player up and short some pins if something goes wrong?
22:12:00linuxstbLear: Are you investigating?
22:12:09mc2739I understand that. I have used files from the git for testing, but didn't know if svn was able to do the same yet
22:12:46amiconnlinuxstb: Rockbox marks invalid tracks in the playlist so they aren't tried again (at least it used to do that on hwcodec)
22:13:06LearA little at least. Making a build with playlist logf, but I don't have much time for it today...
22:13:37amiconnThis was indicated in the playlist viewer. Not sure whether it still works - I don't have invalid tracks, and I don't tend to try such things regularly...
22:14:17Learamiconn: Works for swcodec too - or at least last time I checked, which wasn't too long ago.
22:14:33 Join perrikwp [0] (i=d1a8d351@gateway/web/ajax/
22:14:47*linuxstb sees that playback.c doesn't check the return value from get_metadata()...
22:15:33Learlinuxstb: buffering.c does actual reading, playback.c checks for success in audio_finish_load_track.
22:15:35 Quit tessarakt ("Client exiting")
22:16:07LearHm, with repeat on and only bad files, playback.c detects end of playlist properly, but the wps still doesn't exit...
22:16:39linuxstbLear: Yes, I'm just looking at buffering.c now, and it simply memsets the mp3entry struct with zeros if get_metadata fails...
22:19:42bertrikmc2739, the e200v2 is the only player of the sansa v2 series than can be "unbricked", so maybe you should try it out
22:20:12Nico_Plinuxstb: that's checked in audio_finish_load_track
22:20:22Nico_Pif (track_id3->length == 0 && track_id3->filesize == 0)
22:20:32Nico_Pline 1753
22:20:37Learplayback.c only sets filling to indicate end of playlist; if nothing was actually playing, playing isn't set to false. So the wps doesn't exit.
22:20:39bertrikthe biggest problem AFAIK is the lack of display support for the e200v2. On the m200v2 and clip we can do some primitive debugging with printf's to the display
22:20:57linuxstbamiconn: My playlist with one good and one bad file shows (ERR) for the bad track in the playlist viewer - so yes, it seems "id3->skipped" is being set.
22:21:28mc2739bertrik: ok, I'll give it a try
22:21:31amiconnNow the question remains what happens if all tracks are bad and repeat is on...
22:21:47LearMaybe add a check for filling == STATE_END_OF_PLAYLIST and not pcm_is_playing somewhere, and stop playback if so...
22:23:03 Join HellDragon [0] (
22:23:07Learamiconn: playlist_peek doesn't find any track, so audio_load_track aborts. playing is true at this point and doesn't get cleared.
22:23:09bertrikmc2739, basic steps are probably the same as for git: configure for e200v2 bootloader, build it, build mkamsboot, run it on the OF image and the bootloader you built to create a new flash image and put that image on the player
22:23:48amiconnIn mpeg.c?
22:24:22Learplayback.c. I'm only looking at swcodec...
22:25:10mc2739bertrik: thanks, mkamsboot didn't look to be the same as on the git, I didn't know if it worked the same
22:26:14LearHm, works as expected (in the sim, with repeat on) with one good track and a bunch of bad ones.
22:26:47domonokybertrik: your i2c work only uses the internal i2c (i2cAudMas) ? would it be possible to use the same/similar code for the external i2c (i2cmsi) ?
22:27:21bertrikyes, the two i2c buses are very similar
22:27:59bertrikthe internal one is strictly master, the external one can also be slave as far as I understand
22:28:02domonokyi think the TEA5767 (fm radio) in the m200v2 is on this externa i2c (and maybe also on the i2s)
22:29:14bertrikI just read that the FM chip in the clip and the fuze (Si4702) can also be accessed through i2c
22:30:20bluebrotherwhy should the TEA5767 use I2S?
22:30:34domonokybluebrother: to route the oudio output ?
22:30:54linuxstbIsn't the audio output normally analogue?
22:31:08bertriklinuxstb, exactly what I thought
22:31:14linuxstbi.e. wired to the audio codec
22:31:31bluebrotherIIRC that chip outputs analog audio
22:31:37domonokyso i am probably worng.. .:-)
22:31:49bluebrotherat least back the time I looked at that datasheet for the h100 ...
22:32:02amiconnTEA5767 only outputs analog
22:34:07bertrikwell maybe I need to separate the audio/PMU layer from the i2c layer again then
22:34:27amiconnIt's controlled via i2c (or 3-wire bus, but I don't know any target that uses this mode)
22:37:47domonokybertrik: your i2c code looks so small, we can easily make another on for the other i2c bus...
22:39:06bertrikyeah, it's a bit disappointing in the end to see that something so small took so much time ... :P
22:39:16*domonoky notices that Clip and e200v2 have buttons on those pins of the external i2c, so maybe they use other pins and software i2c for their fm chips ?
22:39:51 Join allele [0] (
22:40:05 Join DigitalisAkujin [0] (
22:42:00bertrikok, hadn't noticed that. The Si4702 can also be configured for a much faster 3-wire mode.
22:43:16alleleI'm getting a data abort message and my plugin freezes when I try exiting it. the only change I made to it was by adding more cells to some of the arrays.
22:45:43 Join shotofadds [0] (n=rob@rockbox/developer/shotofadds)
22:46:02bluebrotherallele: did you also allocate the required space correctly?
22:46:14domonokylooks like these pins for the external i2c are also already used on the m200v (by the lcd).. so the fm is probably not connected to this..
22:46:30Learlinuxstb: I have a fix that seem to work (in the sim at least). Doesn't trigger an end of playlist splash, but ordinary (in-ram playlist) playback doesn't trigger it either, so... Probably best to add it to the patch for some better testing.
22:46:51allelebluebrother: you mean changing the size of the array? if so then yes
22:48:35*shotofadds thinks this m200v2 business is going to get very confusing very quickly. In my hand I have a Telechips-based M200 which has a PCB printed with "revision v2.0". Rockbox's "m200v2" target is *rather* misleading...
22:48:55shotofadds..especially since we will need a v3 target for the HARP-based ones too
22:49:14domonokymaybe we should rename it to m200v4 ?
22:49:34shotofaddsI think that would probably be best
22:49:39 Join drowner [0] (
22:49:42shotofadds"someone" should do it ;)
22:49:49drownerG'day all
22:49:50bluebrotherwell, you could post your changes somewhere, then others could take a look
22:50:10LambdaCalculus37shotofadds: Wasn't that brought up before?
22:50:14bluebrothershotofadds: how 'bout renaming all those AMS targets to something -ams? I.e. e200-ams?
22:50:25allelethe plugin runs fine until the return point is reached
22:50:30drownerI got a 'new' refurbished ipod 80gig video (not the classic, the one before, 5.5) and rocbox utility (under linux) tells me it doesn't support it. Does anyone knnow why?'
22:50:35shotofaddsLambdaCalculus37: yes, but it was seemingly ignored :/
22:50:42bluebrotherhaving m200v4 and e200v2 being rather similar sounds kinda confusing too ;-)
22:51:00BagderI think the ams suffix idea sounds fine!
22:51:22alleleI'm wondering if the plugin would be iterating through one of the arrays while the plugin is ended
22:51:25shotofaddswell that's a nice little job for someone then ;)
22:51:53*Bagder is going away for a couple of days to let "someone" get some room for this
22:52:00drownerdo I have to install manually?
22:52:10drowneror is my firmware the wrong variety?
22:52:14bertrikIf we do that, then ams probably decides that their next chip will be called ams v2
22:52:43Bagderdrowner: wrong filesystem?
22:52:52drownerbadger, no its fat32
22:53:02shotofaddsbertrik: but that's ok, m200-ams-v2 is unambiguous...
22:53:10Bagderthen I don't know why it says that...
22:53:16drownerweird, i know
22:53:31domonokyrbutil detects it by the usb ids so its probably really a classic..
22:54:21bertrikwow this si4702 FM chip looks so much easier to program than the one in my sansa c200v1
22:54:28drownerdomonoky: That's what I first thought, but classics don't sync with linux at all without a hack, and mine syncs fine. I intentionally avoided a classic for this reason.
22:55:16domonokydrowner: you could try the manual way with ipodpatcher.. maybe it tells you more..
22:55:36drowneri might do that
22:55:37bluebrotherdrowner: have you checked the USB ID (from the sysinfo dialog)?
22:56:49drowneryou mean in utility or in general?
22:57:29bluebrotherin Rockbox Utility. There's a Sysinfo menu entry (below "About")
22:58:29drownerI get: VID 05ac PID 1261
22:58:39 Quit LambdaCalculus37 (" ajax IRC Client")
22:59:12Learlinuxstb: Patch added to FS #9502.
22:59:12drownerdoes that mean anything to anyone?
22:59:38bluebrotherwell, from the list in the configuration this is the ID of an Ipod Classic
22:59:52domonoky <- its a classic
23:00:00drownerIts a classic?
23:00:06*drowner shakes his fist at apple
23:00:12domonokyat least the usb-id says this..
23:00:15bluebrotheryou might want to double-check against apples KB article explaining the difference (check the IpodFAQ for the link)
23:00:20drownerTotally weird.
23:00:39drownerand, even weirder, it works with linux.
23:00:50bluebrotherwhy is that weird?
23:01:06drownerBecause classics don't work with linux
23:01:57bluebrotherthat's not true. Apple added some checksum to the itunesdb, but libitunesdb (or is it libitunes?) rather quickly learned how to deal with that
23:02:07bluebrotherunless my memory serves me wrong
23:02:07drowneroh ok
23:02:24drownerI was under the impression it needed to be hacked manually?
23:03:18bluebrotherah, it's libgpod these days
23:03:44drowneroh well
23:03:57drownerhow is the development for rockbox on classic going ;)
23:04:02***Saving seen data "./dancer.seen"
23:04:13bluebrotherit's going very well, given that there is no development ;-)
23:05:42drownerI just can't believe the nerve of apple
23:05:54drownerat least i got a cheap classic, i suppose
23:06:28drowneranyhow dudes, i'm off to work.
23:06:33drownerI shall catch you all
23:06:39drownerthankyou for your assistance. Bye!
23:06:39ameyerbluebrother: they reverse-engineered it within about a week if I recall correctly
23:06:44bertrikthere seems to be quite a bit of code for the e200v2 display
23:06:44 Quit drowner ("Ex-Chat")
23:06:50 Join hannesd [0] (
23:07:32bluebrotherameyer: now someone reverse-engineer the 6G bootloader in that time ;)
23:07:33bertriknow that we know how to access the audio/PMU part of the as3525, we may be able to light up the backlight
23:10:08 Quit Strife89 ("Leaving")
23:10:34 Quit mc2739 ()
23:13:39 Quit allele ("Java user signed off")
23:16:03linuxstbLear: I've just quickly tested your patch on FS9502, and it seems to fix all my problem directories.
23:17:25 Quit domonoky (Read error: 104 (Connection reset by peer))
23:20:38 Quit bmbl ("Woah!")
23:20:51 Quit PaulJam (".")
23:23:31Nico_PLear: the patch looks good to me
23:24:11 Quit bertrik ("Leaving")
23:24:27 Quit hannesd ("Client suicide")
23:25:02 Quit stoffel_ ("leaving")
23:26:52ameyerbluebrother: I somehow doubt rockbox will ever work on the 2G nano, 6G iPod, etc.
23:27:08 Join hannesd [0] (
23:27:12ameyerif anything as a F you to Apple for the encryption
23:27:34LearNico_P: OK. I'm off now, so feel free to commit it.
23:27:42 Quit Lear ("ChatZilla 0.9.83 [Firefox 3.1b2pre/20081022033543]")
23:27:52ameyerwell, that and the non-trivial encryption
23:28:13BigBambi Yes, we all think that it is mightily unlikely
23:28:54Nico_Pameyer: IMO a F*** you would be to crack it :)
23:29:31Nico_Psomeone on the linux4nano ML mentioned an attempt to brute force it, but hasn't posted any updates since
23:29:52ameyerIMO: "F you, we're not gonna help you sell iPods anymore"
23:29:55 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
23:30:07Nico_PI don't think they see it like this
23:30:17linuxstbameyer: But they don't sell them any more (or at least, some of them)...
23:30:47ameyerwouldn't surprise me to see an iTouch port, though
23:30:48Nico_Pif they saw us as any help selling ipods, they wouldn't have locked them
23:31:11ameyerI'm not a developer, so I don't count.
23:32:08ameyereven if I wanted rockbox on the newer ipods, it's not like I could do anything, so my opinion on the matter doesn't count
23:33:10 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
23:33:30saratogaanyone know if the H10s can charge the battery in rockbox?
23:38:18 Join matsl [0] (
23:39:03 Join petur [50] (n=petur@rockbox/developer/petur)
23:48:42 Quit bughunter2 ("bye")
23:49:58 Quit jhulst (Remote closed the connection)
23:50:13 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
23:50:34 Quit culture (Connection timed out)
23:51:12 Quit saratoga ("CGI:IRC (EOF)")
23:58:10 Quit barrywardell (Read error: 104 (Connection reset by peer))

Previous day | Next day