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

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

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

#rockbox log for 2009-05-16

00:01:35mtNo, didn't try it on an actual target. I just wanted to know the main issues to lookout for in my current code. I'll test it on a real target later.
00:01:43linuxstbmt: Finally, I'm not sure that code should be committed until the codec is working, so maybe delay committing it...
00:01:56mtAh, so that's why it doesn't show the details ! :)
00:02:10linuxstbmt: I'm guessing that's the reason...
00:02:40mtlinuxstb : I didn't intend for this patch to be committed. (I should have made this clear in the comment. )
00:02:52linuxstbOK then ;)
00:03:14pixelmaStrife89: did you actually test the c200 keymap patch (thought I remember you saying you wanted to)
00:04:04Strife89pixelma: I never did, but if you link it I'll grab and apply the patch real quick.
00:04:08mtlinuxstb : So what would you suggest for the #undef uint32_t thing ?
00:04:59pixelmathe last one in the comments
00:05:14 Quit bmbl ("Woah!")
00:07:56 Quit itcheg (" ajax IRC Client")
00:08:17 Join karim_ [0] (
00:08:26 Quit karim_ (Read error: 104 (Connection reset by peer))
00:15:29 Quit ender` (" Connection Reset by Gypsies with Wire Cutters")
00:18:00JdGordon|pixelma: havnt looked at the paste yet.. but change that WPS_REFRESH_DYNAMIC to WPS_REFRESH_STATIC... its not really liekely the rtc will magically appear mid song :)
00:18:36pixelmayeah, thought so too but didn't see the STATIC one for some reason
00:20:01 Quit linuxstb (Remote closed the connection)
00:21:51 Join saratoga2 [0] (n=9803c6dd@gateway/web/cgi-irc/
00:22:02saratoga2mt: what target do you have?
00:23:32mtsaratoga2 : why ? :)
00:24:14 Join itcheg [0] (i=62db4767@gateway/web/ajax/
00:24:47 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
00:25:11saratoga2mt: just curious
00:26:15linuxstbmt: You can't just define your variables as uint32_t, instead of using "unsigned int" ?
00:28:06mtlinuxstb : They're all uint32_t, read_uint32be in metadata_common.h expects an unsigned int, that's why.
00:28:46 Join petur [0] (
00:28:58mtI could convert them all to unsigned int.
00:29:35linuxstbmt: I think the read_uint32be function should be changed.
00:30:36mtWouldn't this raise warnings for other files ?
00:31:25linuxstbI'm just testing now... But if it does, they should probably be fixed.
00:34:21mtOK .. I have to go get some sleep now. :)
00:34:38mtGoodnight everyone !
00:41:13 Join midgey|work [0] (i=836b0069@gateway/web/ajax/
00:41:32 Quit bluebrother ("Leaving")
00:45:07CIA-38New commit by dave (r20950): Change the type of read_uint32be to uint32_t and ensure all variables passed to that function are also of that type. No functional changes, but makes ...
00:45:43 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
00:46:32 Join xnyhps [0] (n=xnyhps@2001:470:1f14:da:219:e3ff:fed7:c57c)
00:52:26 Quit petur (Remote closed the connection)
00:53:40 Quit JdGordon| (" ajax IRC Client")
00:55:11Unhelpfulpixelma, LambdaCalculus37: the AA search order is unchanged except for the addition of folder.jpg, which is checked right after the track-specific searches. for each path that we search for album art, try_exts tests for <base>.jpeg, then <base>.jpg, then <base>.bmp
00:56:06 Quit mirak (Remote closed the connection)
00:56:15evilnickUnhelpful: Have you seen the patch that disputes this?
00:56:37Unhelpfulwhich patch?
00:57:01pixelmaUnhelpful: so... the manual still needs a bit refinement
00:57:11 Quit gregzx ("ChatZilla 0.9.84 [Firefox 3.0.10/2009042316]")
00:58:45 Quit evilnick (" ajax IRC Client")
01:00:42 Quit robin0800 (Read error: 104 (Connection reset by peer))
01:00:55Unhelpfuli'm agreeable to skipping folder.jpg when searching with a specific size. WPS might *still* find different AA than pictureflow, though, because pictureflow doesn't, and shouldn't, use the WPS size.
01:02:06 Join JdGordonn [0] (n=irchon@
01:03:00JdGordonnpixelma: what was the link again?
01:04:25pixelma (already changed to STATIC now)
01:04:32 Quit JdGordonn (Remote closed the connection)
01:05:01 Join JdGordonn [0] (n=irchon@
01:05:32 Quit JdGordonn (Client Quit)
01:06:44pixelmais it not possible to build a simulator with some of the advanced build options (like a RTC modded H100 sim)?
01:07:01 Join JdGordonn [0] (n=irchon@
01:07:12 Quit itcheg (" ajax IRC Client")
01:10:09 Quit salty-horse ("Leaving")
01:11:54 Quit JdGordonn (Remote closed the connection)
01:12:39 Join JdGordon| [0] (i=2099ab5d@gateway/web/ajax/
01:16:42 Quit JdGordon| (Client Quit)
01:17:16***Saving seen data "./dancer.seen"
01:20:36 Join JdGordonn [0] (n=irchon@
01:22:19JdGordonnpixelme: your almost there.. you need to call radio_present() (or something like that)
01:22:36JdGordonninstead of just return 'c'
01:25:19JdGordonnpixelma even... stupid client doesn't have tab completion
01:26:42CIA-38New commit by unhelpful (r20951): FS #10201, skip folder.jpg in search_albumart_files if size_string is not empty.
01:27:01JdGordonnsomething like that
01:27:04pixelmawhy that?
01:28:15JdGordonnart ignore me
01:28:16JdGordonndr some rtarted reason I had radio on the rain
01:28:37JdGordonnpatch is fine... u can commit it...
01:29:03JdGordonnjust change dynamic to static
01:29:14pixelmaalready did here
01:31:01pixelmathe c is there similar to the h (for hold or something), to fill it with something if you use it unconditionally (which I think no-one does/will do)
01:31:10 Quit JdGordonn (Remote closed the connection)
01:32:46 Quit midgey|work (" ajax IRC Client")
01:39:40 Quit Thundercloud (Remote closed the connection)
01:44:30 Quit bertrik (Read error: 113 (No route to host))
01:51:38 Part toffe82
01:54:24 Quit matsl (Read error: 110 (Connection timed out))
01:57:16 Quit Strife89 ("Signing off.")
02:03:39 Quit saratoga2 ("CGI:IRC (EOF)")
02:09:35 Join pixelma_ [50] (n=pixelma@rockbox/staff/pixelma)
02:09:35 Quit pixelma (Nick collision from services.)
02:09:40 Quit amiconn (Nick collision from services.)
02:09:42 Join amiconn_ [50] (n=jens@rockbox/developer/amiconn)
02:09:55 Nick pixelma_ is now known as pixelma (n=pixelma@rockbox/staff/pixelma)
02:10:02 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
02:12:08 Join mirak [0] (
02:13:37pixelmaJdGordon: thanks for looking btw. I noticed that I need to prepare something for the manual too, will commit in the morning
02:15:30 Join Thundercloud [0] (
02:24:20 Join acerimmer667 [0] (n=acealex@
02:24:25 Quit mr_pink ()
02:26:05 Part acerimmer667
02:26:05 Quit vixus (Read error: 104 (Connection reset by peer))
02:26:55 Quit xnyhps ()
02:30:21 Quit SirFunk (Read error: 54 (Connection reset by peer))
02:31:34 Join SirFunk [0] (
02:33:58 Quit SirFunk (Read error: 104 (Connection reset by peer))
02:34:39 Join SirFunk [0] (
02:41:00 Join timc`` [0] (n=aoeu@
02:42:27 Join Hillshum [0] (n=chatzill@unaffiliated/hillshum)
02:42:48 Join MarcGuay_ [0] (
02:44:37HillshumWhat does FS #9971 need to get committed?
02:45:13CIA-38New commit by unhelpful (r20952): Extend buflib to support shifting space out off the buffer for other use, and returning space to the buffer.
02:45:17CIA-38New commit by unhelpful (r20953): Use new buflib extensions to avoid static allocation for track list, by shifting space out of the buffer and freeing slides as needed.
02:50:49 Quit timc (Read error: 110 (Connection timed out))
02:50:54UnhelpfulHillshum: per docs/CONTRIBUTING, a real full name is needed, both for the credits file and for legal reasons.
02:54:30HillshumI think he is already in /docs/CREDITS
02:54:39Hillshumsorry got to go now
02:54:56 Quit Hillshum ("ChatZilla 0.9.83 [Firefox 3.0.3/2008092417]")
02:55:14Unhelpfulif that's the same Teruaki, and the patch is good, i don't see any other reason to keep it out...
02:58:09 Quit MarcGuay (Read error: 110 (Connection timed out))
03:03:55pixelmayes, if you look around in the closed tasks in flyspray you'll find the task where he gives his full name
03:06:09Unhelpfuli see... i'll try to review this one, then :)
03:15:54 Quit Thundercloud (Remote closed the connection)
03:16:47 Quit mirak ("Ex-Chat")
03:17:18***Saving seen data "./dancer.seen"
03:21:11froggymananybody know why i cant get monopoly to even start on my 5.5G iPod (with rockboy)
03:28:15Unhelpfulis there any kind of *message* when it fails to work?
03:28:37Unhelpfuldoes it work on gnuboy?
03:29:22 Join Ursinha [0] (n=ursula@canonical/launchpad/ursinha)
03:29:26 Quit timc`` (Remote closed the connection)
03:29:49 Part Ursinha (":wq")
03:29:57froggymanjust says: Plugin Returned Error
03:30:44froggymani am going to get a different dump and then try
03:31:30Unhelpfulno, it doesn't work on gnuboy? than i'd guess it's going to easier to 1) get a different dump or 2) take up the problem with the gnuboy developers
03:31:35froggymanhmm.... now this new ROM appears to be working
03:31:57froggymanmusthave been a bad dump
03:33:56 Join timc [0] (n=aoeu@
03:46:09 Join ncfi1013_ [0] (
03:47:32 Quit dfkt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
03:50:50 Quit gevaerts (Nick collision from services.)
03:51:02 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
03:51:22 Quit froggyman ("CGI:IRC")
03:52:50 Nick JdGordon is now known as JdGordon|afk (n=jonno@rockbox/developer/JdGordon)
03:58:53 Nick fxb is now known as fxb__ (
04:06:25CIA-38New commit by unhelpful (r20954): Attempt to fix yellow on sims (can't reproduce on my system).
04:14:55 Join timc`` [0] (n=aoeu@
04:18:04 Quit lightbulbjim (Remote closed the connection)
04:18:16 Join lightbulbjim [0] (
04:30:49 Join robin0800 [0] (
04:31:29 Quit timc (Read error: 110 (Connection timed out))
04:35:27 Join krazykit [0] (
04:39:32 Join webguest28 [0] (n=58d92868@gateway/web/cgi-irc/
04:40:27webguest28i have a samsung yp-p3, there is no gapless playback. does rockbox support samsung yp-p3?
04:40:58scorcheis that device listed on the front page?
04:41:41webguest28no, thats the reason for my question.
04:41:46webguest28is there any l
04:41:55webguest28plan to make it compatible?
04:42:05scorchethen it looks like you have an answer...
04:42:14scorchewe dont make "plans" for devices, but a port may be in progress
04:42:42 Part wincent ("Kopete 0.12.7 :")
04:43:14scorchecheck the forums under the "new ports" section, and see if there is a thread about it there.....port progress will likely be there
04:44:31 Quit efyx_ (Remote closed the connection)
04:46:56webguest28how much time does it normally need for such a port to be finished?
04:47:08webguest28seems to be kinda sophisticated
04:48:41saratogadepends on the target, some take a few months if they're really close to an existing target, others take a year or two, while others are never finished
04:48:50 Join firedix [0] (n=firedix@
04:50:58webguest28ok, thanks for help!
04:51:39 Quit webguest28 ("CGI:IRC (EOF)")
04:52:44 Quit MarcGuay_ (
04:52:44 Quit bzed (
04:52:51NJoinMarcGuay_ [0] (
04:52:51NJoinbzed [0] (
04:53:18 Quit miepchen^schlaf (Read error: 101 (Network is unreachable))
05:17:19***Saving seen data "./dancer.seen"
05:18:20 Join robin0800_ [0] (
05:23:54 Join ball [0] (
05:24:27ballDoes the iPod have a special MP3 decoder circuit in it, or is that all done in software?
05:24:28 Quit FlynDice (Read error: 54 (Connection reset by peer))
05:27:18 Join FlynDice [0] (
05:28:18martian67ball, mp3 is done in software
05:28:36martian67on ipod videos, h.264 playback is done in hardware
05:30:20ballInteresting. I asked about MP3 because I've heard rumours that Ogg/Vorbis is harder on battery life than MP3 and I wondered if the CPU had to work harder because it got less support from the hardware. It's not a problem for me, but I'm a geek and I wonder about these things.
05:30:38ballI *love* that RockBox lets me play Ogg/Vorbis on my iPod Mini!
05:36:31ballsaratoga: thanks.
05:38:01Unhelpfuli'm looking for opinions on whitespace and line-length cleanups (i'm prepping FS #9971 for merge). should i clean up only inserted material, should i clean up other parts of the file while i'm in there, should i leave any cleanup of existing code for another revision, or should i just not touch existing code at all?
05:40:28saratogathough i guess more vorbis optimizations have gone in since the most recent tests on that page
05:45:25 Quit robin0800 (Read error: 113 (No route to host))
06:04:05 Part krazykit ("Connection reset by beer")
06:04:26 Join Tristan [0] (
06:22:14 Quit _Auron_ ("Infinity repeatedly denies rumours of plotting with zero to bring down the Universe.")
06:24:47 Join robin0800 [0] (
06:25:35 Join _Auron_ [0] (
06:43:30 Join jmillikin [0] (
06:43:57 Part ball ("beware of the owls")
06:45:48 Quit saratoga ("CGI:IRC (EOF)")
06:49:07 Quit robin0800_ (Read error: 113 (No route to host))
06:50:17 Join Digital_Root [0] (
06:52:15Digital_RootHi there. I was wondering if anyone has tried to completely format an iPod 3rd gen, and install Rockbox that way?
06:53:06scorchethe install process for rockbox includes making sure that the apple firmware is present...why do you wish to completely format your device?
06:53:11Unhelpfuldepends on what exactly you mean by "completely format". is there a reason you can't follow one of the install guides?
06:54:04 Quit robin0800 (Read error: 104 (Connection reset by peer))
06:54:10Digital_RootOk well I was asking because currently it isn't possible to install on the 3rd gen nano
06:54:29scorcheoh...3rd gen NANO....yeah...wont work
06:54:35scorcheit realyl really isnt that simple
06:54:59Digital_RootSo what if I were to download the firmware for my 1st gen nano from iTunes and install it over the 3rd gen's firmware
06:55:09scorcheit really really isnt that simple
06:55:23Unhelpfulit's different hardware, and it has checks against starting non-apple firmware.
06:55:24scorchethere is encryption to get past and the hardware is completely different
06:55:51 Join saratoga [0] (n=9803c264@gateway/web/cgi-irc/
06:56:37Digital_Rootok, well encryption isn't really my thing but I am more than willing to help getting rockbox onto the 3rd gen Nano.
06:56:56saratogathis is like asking if you can somehow reconfigure a cat so that it will run Apple II software
06:56:58scorchewell, the encryption is pretty much the first barrier to get past
06:57:54Digital_RootDon't even threaten me like that, I just throw out my II e two months ago, I couldn't believe I still had it.
06:58:06scorchestay on topic please...
06:58:21Digital_RootOk how would I go about "finding" the encrypted firmware
06:58:37saratogayou can download it from apple's website i believe
06:59:07Digital_RootOk, and then the problem is opening the firmware to edit it then?
06:59:13saratogabut since its encrypted looking at it isn't very useful
07:00:49Digital_Rootok so the over all goal at the moment would be to unencrypt the firmware so that it could be read. Am I correct in that statement?
07:01:16UnhelpfulDigital_Root: more so that we can encrypt our own such that it will be recognized, but yes, that's the major problem.
07:02:30Digital_RootOk well I guess about the only way I can really be helpful there is to try and decrypt it.(I don't know why I said 'unencrypt', I know its not a word)
07:02:49scorcheand how are you going to try that?
07:03:24Unhelpfulgood luck, then. you should know that apple could have easily used any of a number of well-documented, public cryptographic methods which can not be broken by any practical effort. :)
07:03:53Digital_Rootscorche: Well, I don't really know at the moment, but trying never hurt anyone.
07:03:58 Quit advcomp2019 (Read error: 104 (Connection reset by peer))
07:04:22 Join advcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
07:04:24Digital_Rootthank you all for the advice and tips
07:06:00scorchegood luck then =)
07:11:03 Quit saratoga ("CGI:IRC (EOF)")
07:11:12 Join petur [0] (
07:17:22***Saving seen data "./dancer.seen"
07:25:44 Quit martian67 (Read error: 113 (No route to host))
07:34:30 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
07:45:01 Join Horschti [0] (n=Horscht@xbmc/user/horscht)
08:03:19 Quit Horscht (Read error: 110 (Connection timed out))
08:27:04 Quit amiconn (Remote closed the connection)
08:27:05 Quit pixelma (Read error: 104 (Connection reset by peer))
08:41:24 Join bertrik [0] (
08:45:51 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
08:45:53 Join pixelma [50] (n=pixelma@rockbox/staff/pixelma)
08:48:42 Quit pixelma (Remote closed the connection)
08:48:42 Quit amiconn (Remote closed the connection)
08:48:45 Join Rob2222 [0] (
08:49:03 Join _fml [0] (n=4fd3f3de@gateway/web/cgi-irc/
08:50:18_fmlAre the changes in the manual regarding the order of searched album art files correct? Isn't <trackname>.jpg tried before a more generic cover.jpg?
08:50:19 Join pixelma [50] (n=pixelma@rockbox/staff/pixelma)
08:50:19 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
08:51:47 Join flydutch [0] (
08:57:14bertrikUnhelpful, shouldn't there be an #ifdef JPEG_FROM_MEM #include "jpeg_mem.h" #endif in apps/recorder/jpeg_load.c?
08:58:55Unhelpfulit's: <track_basename>.<ext>, <album_name>.<ext>, cover.<ext>, folder.jpg (only if not searching with a size tag), /.rockbox/albumart/<artist>-<album>.<ext>, ../<album>.<ext>, ../cover.<ext>
08:59:33CIA-38New commit by alle (r20955): Make variables static where possible
09:00:26UnhelpfulWPS tries the list first with .<width>x<height> from the WPS AA size inserted before .<ext>, then without, and for each item, <ext> is substituted with jpeg, jpg, and bmp and tested
09:01:05Unhelpfulthe only exception to any of that is folder.jpg, which is a specific windows-promoted convention, and always has the same extension, and doesn't use the size tag.
09:01:37Unhelpfulbertrik:not ignoring you, i'll check in a second :)
09:02:42Unhelpfuli don't see any reason it needs to, those are prototype declarations for the functions defined when bulding with JPEG_FROM_MEM, they're there for plugins that would use it.
09:05:02_fmlUnhelpful: so the manual in its current form isn't quite accurate?
09:05:04 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
09:05:36Unhelpfuli'm sorry to say that i don't even know what the manual says at present... let me look :)
09:06:13_fmlUnhelpful: I mean r20949
09:06:15 Quit Rob2223 (Read error: 110 (Connection timed out))
09:09:18*Unhelpful sees an S manual... can official support be far away?
09:11:11bertrikUnhelpful, jpeg_load.c also does the jpeg_mem stuff AFAIK, so it should know about the prototypes I think
09:11:53Unhelpfulshould it? it defines those functions, but never uses them...
09:12:03 Join schrottplatz [0] (
09:13:49 Join nibbler [0] (
09:13:50bertrikI think either we should #include the header file for the functions that are instantiate, or maybe not instantiate the functions at all (and save some space)
09:15:59Unhelpfulnow i'm not sure i follow... JPEG_FROM_MEM is only defined by a wrapper file for pluginlib, the core always uses the from-disk version of the decoder. the functions don't exist in core, and they're dropped at link for plugins that don't use them...
09:17:24***Saving seen data "./dancer.seen"
09:20:27bertrikUnhelpful, I try to compile with -Wmissing-prototypes once in a while. It warns when a functions is defined without being declared. I'm getting a "no previous prototype" warning now for get_jpeg_dim_mem.c and jpeg_load.c in apps/plugins/lib/../../recorder/jpeg_load.c
09:22:08 Quit moos (Read error: 110 (Connection timed out))
09:22:56Unhelpfulahhhh... well, if you want to reduce noise when looking for code that might be harmful, that's a reason, and i can't think of a reason not to. :)
09:23:17_fmlUnhelpful: what do you think about defining, say, USE_JPEG_COVER in albumart.c as "defined(HAVE_JPEG) || defined(PLUGIN)" and then using it everywhere instead of repeating the condition?
09:24:16Unhelpfuli think it's not a bad idea... i didn't really think of doing such a thing in that file... i did in a few others, but the conditions were trickier there (and not always the same)
09:25:16Unhelpfulperhaps that macro should be defined in albumart.h? although i can't see what use it would be outside of the .c file...
09:25:18_fmlUnhelpful: and: wouldn't it be easier to just define the extensions array differently and use try_exts everywhere? This would simplify the code.
09:26:30Unhelpfulit would, but it would increase the code size when HAVE_JPEG is undefined, since that case streamlines things by putting the .bmp right in the snprintf.
09:27:28_fmlUnhelpful: hmm, yes. A tiny bit, but still the delta would be positive.
09:29:55Unhelpfuli'm for a USE_JPEG_COVER define, but i just don't think that little bit of code cleanup is worth increasing code size, although the gain is fairly small for both cases.
09:32:25_fmlUnhelpful: the gain would be improved (IMHO) code readability and, in the former case, no code duplication. I'll commit the define bit, ok?
09:33:36Unhelpfuli get the reduced duplication / improved readability argument... let me see what the actual cost would be. :)
09:34:39CIA-38New commit by alle (r20956): Define a symbol and use it instead of repeating the condition
09:35:26tmzta define isn't a symbol though is it?
09:35:36tmztI thought it was a preproccessor thing
09:35:53Unhelpfulhrm, perhaps a macro that would wrap the two cases might be appropriate? a good name would make it obvious what it does, and it could still use the slightly-smaller version if HAVE_JPEG is undef
09:37:03_fmlUnhelpful: I did only one part as the other would probably increase the bin size
09:37:37Unhelpful_fml: it shouldn't increase binsize for our normal builds. there are no targets that define HAVE_ALBUMART and not HAVE_JPEG
09:38:50pixelmayes, the manual is still wrong with the album art priority list. I start thinking that it is maybe better to have the list generic without extension or so and then add a sentence about the order of checking filename with size specification, jpeg, bmp
09:39:10_fmltmzt: define itself is not a symbol, but it defines a symbol. And yes, it's all done during preprocessing.
09:40:50Unhelpful_fml: hrm... on ARM, albumart.c with a crippled try_exts is a "whopping" 120B larger than with the current solution. i think i'll take a crack at doing a macro that can do things either way, and see if it's wretched or not. :)
09:42:03_fmlUnhelpful: "crippled" = only one extension in the array and no special branching on HAVE_JPEG?
09:42:47Unhelpful_fml: i didn't change the HAVE_JPEG stuff, i just removed the other extensions and cut out the folder.jpg check
09:43:08Unhelpfuland then compared that to a build with HAVE_JPEG undef'd at the top of albumart.c
09:44:51linuxstbUnhelpful: Thanks for fixing my yellow.
09:45:26 Quit Horschti ("Verlassend")
09:45:42Unhelpfullinuxstb: no problem... very strange the way it popped up only on some build servers. gcc can be so flaky about int vs long :/
09:46:51_fmlUnhelpful: I don't quite follow you. Why should changing try_exts that way do anything? I think the change in bin size come from removing the "fixed" folder.jpg check
09:47:49Unhelpful_fml: with HAVE_JPEG defined, and the extensions array reduced, isn't that exactly what you'd have if we got rid of the special case for HAVE_JPEG undefined?
09:49:31_fmlUnhelpful: ahh, yes! Now I see. The delta probably comes because more parameters have to be passed every time (2 vs. 1)
09:49:31Unhelpfulbertrik: i added an include in the wrapper file. will that suffice, at least until we have some reason to think about using the from-memory decoder in core?
09:50:06_fmlUnhelpful: (try_exts vs. file_exists)
09:50:57 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
09:51:08Unhelpfulthat, and the body of try_exts itself... nm says it's 104B on ARM.
09:52:44linuxstbUnhelpful: I'm guessing it was 64-bit hosts causing the problem, which is why the yellow moved around.
09:52:45_fmlUnhelpful: in that case, we can define try_exts differently (just as file_exists) if JPEG is not defined
09:52:47bertrikUnhelpful, yes, that's probably the cleanest fix and works fine for me
09:52:58amiconnlinuxstb, Unhelpful: I guess it has to do with 32bit vs. 64bit build boxes in conjunction with gcc versions (newer versions tend to spit out more warnings, whether they make sense or not is another matter...)
09:53:03Unhelpful_fml: yes, that's what i was just thinking, myself :)
09:53:35Unhelpfulsince there are more people here now, are there any opinions on ?
09:54:28Unhelpfulmaybe i should just commit the patch for now with cleanups only for the inserted code, and then ask the policy question on the mail listen, with a goal of putting the consensus on it into CONTRIBUTING?
09:55:07_fmlUnhelpful: ok, the easy part is left as an exercise for the reader :-))) I have to leave, bye!
09:55:17 Quit _fml ("CGI:IRC 0.5.9 (2006/06/06)")
09:56:04 Quit bertrik ("Leaving")
09:56:18 Join bertrik [0] (
09:57:13 Quit perrikwp (Remote closed the connection)
10:00:21linuxstbUnhelpful: IMO, the problem with whitespace cleanups is that it breaks "svn blame" - so it makes it harder to see who's really responsible for things. So I would probably prefer to leave things as they are, and only fix whitespace when changing that code.
10:02:03Unhelpfulah... i see what you mean, if i clean up the inserts, it's not a big deal, but if i clean up other things, blame thinks i wrote half the file... and will think that no matter how the cleanup happens. :/
10:03:30 Quit Llorean (Read error: 104 (Connection reset by peer))
10:09:26 Join ender` [0] (
10:10:04 Quit mt (Read error: 110 (Connection timed out))
10:18:33 Quit nibbler ("Ex-Chat")
10:20:43CIA-38New commit by unhelpful (r20957): Add include to pluginlib_jpeg_mem.c to prevent warning noise with -Wmissing-prototypes.
10:20:49CIA-38New commit by unhelpful (r20958): Clean up albumart.c a bit, reducing use of USE_JPEG_COVER inside search_albumart_files to only around the folder.jpg test and a strlen call. 0 delta ...
10:20:57***Alert Mode level 1
10:20:57CIA-38New commit by unhelpful (r20959): FS #9971 superdom fixes from Teruaki Kawashima.
10:24:15 Join nibbler [0] (
10:30:52 Quit schrottplatz ("o.O")
10:30:58***Alert Mode OFF
10:36:21 Join perrikwp [0] (i=18ac0c41@gateway/web/ajax/
10:41:33 Join moos [0] (i=mustapha@rockbox/staff/moos)
10:42:55 Quit tvelocity (Read error: 110 (Connection timed out))
10:43:26 Join tvelocity [0] (
10:59:48 Quit martian67 (Remote closed the connection)
11:02:16 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
11:03:07bertrikis there anyone coming to devconeuro2009 with a c200v2?
11:12:03 Join miepchen^schlaf [0] (
11:17:27***Saving seen data "./dancer.seen"
11:24:16 Join stripwax [0] (
11:42:22 Join einhirn [0] (
11:44:35 Quit einhirn (Read error: 104 (Connection reset by peer))
12:09:32 Join renke [0] (
12:10:11pixelmacould a native speaker help me out a bit? I'd like to describe that "RTC present" tag in the manual - list it in the RTC section (and non-RTC targets will get that one tag too then), mentioning that this can be useful for cross-target WPSs for display sizes where targets with and without RTC exist - also explaining the tag's use briefly
12:10:41pixelmahave some trouble phrasing that nicely
12:14:50AlexPSomething like "Returns "c" if a RTC is present, and can also be used as a conditional. This can be very useful for designing a WPS that works on multiple targets, some with and some without a RTC. By using this tag as a conditional it is possible to display current date and time on those targets that support this, or alternate information on those that don't"
12:15:38linuxstbWill people know what an "RTC" is?
12:15:53pixelmathe headline above is Real Time Clock
12:15:54AlexPor real time clock - I jut couldn't be bothered to type it here :)
12:16:38pixelmaAlexP: THANKS, SOUNDS GOOD :)
12:16:45pixelmaoops, sorry
12:16:49AlexPpixelma: NO PROBLEM :)
12:17:24AlexPand then an example
12:18:09linuxstbBut I mean, even "Real time clock" is a bit technical.
12:19:12AlexPJust "Clock" then?
12:19:19*Unhelpful would assume that's the type of clock being talked about, if you say "clock" without other qualifiers, and would expect most USians to, to the extent that the know of any other kind of clock at all...
12:19:39AlexPUnhelpful: I think that applies universally
12:20:39AlexPor "Time and date"
12:20:54pixelma"Clock" also in the headline then?
12:21:37AlexPor "Time/Date"
12:21:41linuxstb"Time and date" (or Date and time) works for me.
12:22:39*pixelma discovers that the "hold switch state" tags only appear in the H100 and H300 manuals.... will fix that too then
12:23:38pixelmaand "clock" in the description then?
12:23:58pixelmaworks better to explain where the "c" comes from... ;)
12:24:27AlexPI think I prefer Time/Date, but I'm not too fussed
12:25:52pixelma"Returns "c" if Time/ Date is present..."
12:26:05pixelmanot sure that works
12:26:36AlexPcould c be something else?
12:27:04AlexPI wouldn't like "returns whatever if a clock is present" either
12:27:10AlexPwell, it is OK
12:27:29AlexPif maybe clock is OK there
12:27:39AlexPpixelma: I don't know any more :)
12:27:43Unhelpfulhow often is the bare clock tag used, other than as a conditional, anyway?
12:28:00AlexPUnhelpful: Never so far, it doesn't exist :)
12:28:10AlexPUnhelpful: But I imagine it will be never
12:28:30Unhelpfulperhaps it would make sense to only speak of it as a conditional?
12:29:57pixelmathere are some other tags (like the virtual LED one) which behave the same and I think they are still explained
12:31:13pixelmamaybe "if the hardware" is present?
12:31:25Unhelpfulperhaps something along the lines of "but its main use is as a conditional", along with the example?
12:31:44AlexPpixelma: "If the necessary hardware is present"?
12:32:04pixelmaI'll go for that one
12:32:28*pixelma starts thinking that this can all be improved later anyways :)
12:32:37AlexPindeed so :)
12:46:25 Join schrottplatz [0] (
12:51:22CIA-38New commit by bluebrother (r20960): Support building beastpatcher with MinGW.
13:17:29***Saving seen data "./dancer.seen"
13:22:45CIA-38New commit by pixelma (r20961): New 'clock present' WPS tag: The tag checks for the presence of the clock hardware and can be used unconditionally (returning a 'c') or in a ...
13:25:01 Nick fxb__ is now known as fxb (
13:30:12 Quit stripwax (Read error: 104 (Connection reset by peer))
13:33:46 Join stripwax [0] (
13:35:47 Quit stripwax (Read error: 104 (Connection reset by peer))
13:36:16 Quit perrikwp (" ajax IRC Client")
13:36:28 Join perrikwp [0] (i=18ac0c41@gateway/web/ajax/
13:38:29 Join stripwax [0] (
13:42:31 Join Thundercloud [0] (
13:44:33 Join efyx_ [0] (
13:45:48 Quit petur ("reboot")
13:50:40 Quit Thundercloud (Remote closed the connection)
13:52:55 Join petur [0] (
13:52:58Unhelpfulwould exporting resize_on_load in the plugin API be objectionable to anybody? i'm working on a benchmarking plugin for the scaler, before i get into some ideas for speeding it up a bit... but there's nothing else that will likely use the scaler soon, so i can just keep it local.
13:56:12 Quit stripwax (Read error: 104 (Connection reset by peer))
13:56:39pixelmablah, I thought the "Hold switch" tag section was an easy fix but I just realised that it needs a different headline and a slightly new phrasing - because the tag can also be used on the software hold targets (just displays button lock status)
13:56:57pixelmafor a while...
13:57:57*pixelma has no time to do the necessary changes now
14:00:03 Join stripwax [0] (
14:04:53 Quit flydutch ("/* empty */")
14:12:26 Join Thundercloud [0] (
14:12:43 Quit firedix ("Ex-Chat")
14:14:57 Join dekaden|Z [0] (
14:21:31 Join stripwax5443 [0] (
14:21:37 Quit stripwax5443 (Client Quit)
14:22:23dekaden|Zhi guys and gals, i'd like to add a few words to the wiki on how to compile and use FontForge and convbdf on Windows
14:23:08dekaden|Zbut it seems i need write privileges first
14:24:24dekaden|Zwould someone bring the kindness to grant them to me, please :-]
14:28:24pixelmayou would need to give us your WikiName
14:28:48dekaden|Zshure: DavidFoerster
14:30:15pixelmayou should be able now
14:30:21dekaden|Z(argh, should have used query for this. now it's in the logs...)
14:33:37 Join James [0] (n=5d81c3a2@gateway/web/cgi-irc/
14:34:05 Nick James is now known as Guest55352 (n=5d81c3a2@gateway/web/cgi-irc/
14:34:49Guest55352is there a way to use rockbox and my ipod video with the remote control?
14:37:18AlexPdekaden|Z: It'll be obvious when you edit the page anyway
14:37:41Guest55352do you mean me?
14:37:47AlexPGuest55352: Possibly, it depends on which remote crontrol
14:38:01AlexPGuest55352: No, or I would have used your nick not so;eone elses
14:38:02Guest55352i have the original apple remote
14:38:22Guest55352and the original apple docking
14:38:30Guest55352with the ir-receiver
14:38:36AlexPGuest55352: There is a wiki page that might say if someone else has tried it, otherwise I'd just try it
14:38:48AlexPThere is a reasonable chance it will
14:38:50dekaden|ZAlexP: yeah, i thought about that too. but the logs are indexed by google and now there's a direct connection between my name and my nick
14:39:24AlexPdekaden|Z: Well Rockbox has a real name policy, so trying to hide it won't work in the end :)
14:39:26Guest55352AlexP do you have an adress of this wiki-page?
14:39:36AlexPGuest55352: Not off the top of my head, no
14:39:47AlexPGuest55352: The wiki has a search function
14:40:01 Join SirFunk_ [0] (
14:40:09Guest55352do you mean the wiki?
14:40:13dekaden|ZAlexP: it's just about the required effort to establish the conenction
14:40:15AlexPof course
14:40:22AlexP@ Guest55352
14:40:32AlexPdekaden|Z: Yeah, I see what you are saying
14:45:00Guest55352AlexP could you do a fovor for me and send me a link? :)
14:45:15AlexPGuest55352: I don't know the link
14:45:26AlexPGuest55352: You are just as capable of searching as I am
14:45:45AlexPGuest55352: Click on wiki on the left of, then click on search at the top right
14:46:08AlexPGuest55352: I guess you would search for something like Apple Accessories or Ipod Accessories
14:47:40 Quit Thundercloud (Read error: 113 (No route to host))
14:48:10AlexPGuest55352: If you haven't yet found it, it is
14:48:34 Join Thundercloud [0] (
14:49:02Guest55352AlexP i already was at this page - i haven't found something about the ipod remote
14:49:12AlexPSo then nobody has yet tested it
14:49:18AlexPWhy not just try it?
14:49:28Guest55352i tried it
14:49:38Guest55352but i doesnt work :(
14:49:45AlexPDid you turn accessory power supply etc. on?
14:50:05AlexPtry that then
14:50:12AlexPCheck the manual
14:50:17Guest55352in my roboxmenu?
14:50:22AlexPof course
14:51:38AlexPGuest55352: There are a couple of Ipod accessory related options if I remember correctly - it is a good idea to check them out. I haven't got an ipod, so I can't give you exact names
14:52:10 Quit SirFunk (Read error: 110 (Connection timed out))
14:55:38 Join wincent [0] (
14:56:19 Quit evilnick_home (Read error: 104 (Connection reset by peer))
14:56:32AlexPGuest55352: Any luck?
14:57:12Guest55352AlexP not yet
14:57:30Guest55352i found the point in my ipod menu but it doesnt work
14:57:32AlexPHave you turned on accessory power?
14:57:45Guest55352there have to be a way! :D
14:57:56AlexPsure, you can code the missing support
14:58:14 Quit ncfi1013_ ("Leaving")
14:58:50AlexPRockbox only partially implements the Apple accessory protocol - if your device needs commands that aren't yet supported, then someone that owns that device needs to figure out what those commands are and then add them
14:59:23 Join evilnick_home [0] (
14:59:27Guest55352they say it should work but not how
14:59:54AlexPThey are some random site, nothing to do with us
15:01:10 Join nibbler_ [0] (
15:01:20Guest55352but how could it be that third company docks works and not the original apple? :)
15:02:00AlexP<AlexP> Rockbox only partially implements the Apple accessory protocol - if your device needs commands that aren't yet supported, then someone that owns that device needs to figure out what those commands are and then add them
15:02:13AlexPIt depends on what commands it needs
15:03:44HorschtI myself have 2 accesory devices
15:04:17Horschta third party alarm clock and a third party radio remote. Both work fine using rockbox (except for radio on the remote)
15:06:10Guest55352î only whant to get the original remote working :D
15:07:10 Join SirFunk [0] (
15:07:56 Quit Digital_Root (Remote closed the connection)
15:08:49Horschtand it works with the official firmware?
15:10:19 Join SirFunk__ [0] (
15:10:21Guest55352thanks to AlexP
15:10:24Guest55352for helping
15:10:36AlexPNo problem, it is a shame it doesn't work
15:10:37Guest55352i installed the newest firmware
15:10:47Guest55352and it works!!! :D:D:D:D
15:10:58Guest55352laudness, skip tracks
15:11:05 Join ItalianPianist [0] (n=dfsff@
15:11:08 Part ItalianPianist
15:11:27AlexPGuest55352: I had assumed that you had already tried updating Rockbox before asking for help....
15:11:29 Quit SirFunk_ (Read error: 110 (Connection timed out))
15:11:49Guest55352that so great
15:11:56AlexPPlease do next time. It is rather frustrating trying to fix a problem that no longer exists
15:11:59Guest55352thanks a lot and sorry for me englishc
15:12:22AlexPGuest55352: Your English is fine, but please do update in future before asking for help
15:14:32 Quit nibbler (Read error: 113 (No route to host))
15:14:33AlexPGuest55352: However, I'm glad it works :)
15:17:33***Saving seen data "./dancer.seen"
15:19:32dekaden|Zwhat's the better way to file a font with a few added glyphs −− send a patch to flyspray or just commit it to the repository?
15:20:20AlexPWhich repository?
15:20:57dekaden|Zthis one
15:21:04AlexPIf you mean Rockbox SVN, then you can't without commit access, so add it to flyspray and hopefully someone that can will look at it
15:21:29dekaden|Z"hopefully"... :D
15:21:42AlexPWell it is a volunteer project
15:22:02AlexPIt just requires someone with the knoz-how and time to review it then commit it
15:22:15dekaden|Zi know. you mentioning it just made me laugh
15:22:37AlexPI didn't want to seem to commit to anything :)
15:24:35AlexPWhat are you adding to which font?
15:26:13 Quit SirFunk (Connection timed out)
15:26:54 Join merbanan [0] (
15:32:27dekaden|Zi'm adding ellipsis and emspace to 12-nimbus
15:34:25 Quit SirFunk__ (Read error: 110 (Connection timed out))
15:38:41 Quit Guest55352 ("CGI:IRC (EOF)")
15:39:41 Join SirFunk [0] (
15:50:32 Join SirFunk_ [0] (
15:51:47 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
15:52:13 Join MarcGuay [0] (
15:54:44 Join MarcGuay__ [0] (
15:55:19 Quit MarcGuay_ (Read error: 60 (Operation timed out))
15:55:32 Quit timc`` ("Leaving")
15:55:57 Join timc [0] (n=aoeu@
15:59:54 Quit SirFunk (Read error: 110 (Connection timed out))
16:10:38 Join __lifeless [0] (n=lifeless@
16:10:40 Join matsl [0] (
16:12:29 Quit _lifeless (Read error: 60 (Operation timed out))
16:12:57 Quit SirFunk_ (Read error: 110 (Connection timed out))
16:13:33 Quit MarcGuay (Read error: 110 (Connection timed out))
16:22:04 Join jspeter [0] (
16:23:17jspeteronce I install rockbox, having put OF.mi4, pp5022.mi4, and .rockbox/ on the player, can I delete these files, or are they required for operation?
16:23:29jspeter(e280r, if it matters)
16:23:44 Part schrottplatz ("o.O")
16:27:01 Join stripwax5443 [0] (
16:28:06 Join codemonkeyandrew [0] (
16:28:36codemonkeyandrewdoes anybody here happen to know if the doom plugin saves settings to a file?
16:29:00dekaden|ZDuh, what is this supposed to mean: "List of does not include characters" (
16:33:03dekaden|Znever mind, i found it myself
16:35:21 Join wpyh [0] (n=william@
16:35:58 Quit antil33t (Read error: 54 (Connection reset by peer))
16:36:11 Join antil33t [0] (n=Mudkips@
16:38:11 Join xnyhps [0] (n=thijs@2001:470:1f14:da:219:e3ff:fed7:c57c)
16:49:26 Quit matsl (Read error: 110 (Connection timed out))
16:49:32 Join CaptainKewl [0] (
16:49:55 Quit stripwax5443 (Read error: 110 (Connection timed out))
16:50:06 Join tomers [0] (
16:52:04gevaertsit looks like we'll be able to drop this USB_CHARGING_ONLY driver soon (at least for devices using the arc usb controller). The net cost of using HID instead seems to be 208 bytes
16:53:03 Quit tessarakt (Read error: 110 (Connection timed out))
16:53:17 Join tessarakt [0] (
16:54:53codemonkeyandrewjust in case somebody who knows has come online since i last asked, does anybody know if the doom plugin saves settings to a file?
16:59:04codemonkeyandrewcould i, maybe, get a "no" just to show there is actually somebody else alive on this channel?
17:05:47 Quit codemonkeyandrew ("ChatZilla 0.9.84 [Firefox 3.5b4/20090423204732]")
17:17:36***Saving seen data "./dancer.seen"
17:23:47 Quit tessarakt (Read error: 110 (Connection timed out))
17:24:15 Join tessarakt [0] (
17:30:12CIA-38New commit by gevaerts (r20962): Add working USB HID driver, by Tomer Shalev (part of his GSoC work). ...
17:30:42CIA-38New commit by gevaerts (r20963): Add usb_hid_usage_tables.h for HID to actually compile
17:31:03 Join toffe82 [0] (
17:31:35tomersHi there, I am working on the USB HID class driver, i.e. when the DAP is connected to USB, you can use it as a multimedia keyboard - Volume Up/Down/Mute, Next/Prev Track. I've created key mappings for the platform I have, Sansa E200. How can we define other target's key mappings? Show I do it, or other developers do it?
17:31:59tomersBecause I can't test platforms I don't have...
17:33:25***Alert Mode level 1
17:33:25CIA-38New commit by gevaerts (r20964): fix DEBUGF statement to print variables that actually exist (FS #10212 by Tomer Shalev)
17:34:30xnyhpstomers: Ah, cool! I just tried your patch out yesterday and added #define's for the iPod's controls, if you're intrested in that I could share it.
17:36:59tomersxnyhps: Thanks! I'm glad it works. I got it working even better since. Keep the changes now, as I switched from defines into actions in app/keymaps. When it will be committed, or pushed into the git repository, send it to me, or commit yourself. In other thought, you can send it to me now, and I'll add it myself
17:39:29gevaertstomers: I didn't commit most of the changes to isp1583.c, because I don't have any way to test them. If you care about those changes, maybe post them to flyspray as a separate patch?
17:40:13 Join Dauron [0] (
17:40:21tomersdevaerts: OK. I'll do it (posting for flyspray). I don't have it either, so it's definitely not tested.
17:41:12 Quit _Auron_ (Read error: 104 (Connection reset by peer))
17:41:12 Nick Dauron is now known as _Auron_ (
17:42:05tomersWhere is the build table, with all the reds and yellows? I find it very hard to find (it's not bookmarked). It is neither in Main/DocsIndex#For_Developers nor Main/BuildServer
17:43:26***Alert Mode OFF
17:43:42xnyhpstomers: This is what I was using, I'm not sure if "stop" was working, I don't think my computer recognized it.
17:43:58xnyhpsAnd I had no button left for "mute" :)
17:45:09tomersxnyhps: Thanks! I'll review it and use it. Mute is not at all important TMHO.
17:45:34gevaertsthese things can always be improved later om
17:46:31tomersI've implemented support for seek forward/backward. Apparently there's HID codes for these, but Windows did not respond to these codes, so I dropped them. I tested with my reference multimedia keyboard, and it doesn't able to seek either. Too bad
17:47:44tomersWindows also supports the Mail/Calendar/Calculator/Home(web)/Files(explorer) codes, but those are irrelevant IMHO. There is an HID code for launching audio player, but Windows doesn't support it :-(
17:47:59gevaertsActually, maybe it would make sense to make these configurable? Some prople are going to want to use this for other things than controlling audio
17:48:27tomersgevaerts: It is in my todo list (but not in the wiki ;-) )
17:48:32xnyhpsYeah, I was thinking if it would be cool if you could switch between keymaps? Might be fun to use it for presentations or something.
17:48:37xnyhpsAh, okay.
17:50:32tomersxnyhps: I thought about it either, but I didn't think of any useful scenario for it. But your suggestion for PowerPoint pointer like, is really cool idea. It should be easy to implement context menu, with support for switching code pages
17:50:46gevaertstomers: another way would be to have the screens.c one to be fixed, but export the HID event sending functions in the plugin api. That way you keep basic HID during an MSC connection, but if you connect in charging-only mode you can have a plugin that does whatever you want.
17:51:09gevaertsOf course it probably makes sense to wait with that until you've looked into the usb-in-plugins bit a bit morew
17:51:44 Nick J-23 is now known as DurczokSpecter (
17:51:52 Nick DurczokSpecter is now known as J-23 (
17:52:50tomersgevaerts: I'll think about it
17:55:21 Quit xnyhps (Remote closed the connection)
17:55:45 Join xnyhps [0] (n=sphynx@2001:470:1f14:da:219:e3ff:fed7:c57c)
17:55:47 Join LambdaCalculus37 [0] (n=rmenes@rockbox/staff/LambdaCalculus37)
17:58:25 Quit xnyhps (Remote closed the connection)
18:00:40kkurbjungevaerts or someone else that's knowledgeable on USB, I've been working on a driver for the renesas 66591 and I was wondering if you could take a look at this logf output and tell me if it looks like it's going in the right direction or if I'm not even close to getting the mass storage driver working: The log is backwards (the beginning is at the end).
18:02:21 Join xnyhps [0] (n=xnyhps@2001:470:1f14:da:219:e3ff:fed7:c57c)
18:03:27AlexPtomers: Another option to create basic mappings is to look at the manuals or the sims - they show what buttons targets have.
18:04:12gevaertskkurbjun: from what I can see it enumerates correctly. What seems to be missing is endpoint initialisation.
18:04:19tomersThanks. I'll try to figure the keymaps myself, hoping people will send comments
18:05:46AlexPI assume this works on linux too?
18:06:08gevaertsat least xev tells me that things happen
18:06:33kkurbjungevaerts: at the moment I don't have any logf's going when an endpoint is requested, when I do a lsusb -v it shows the device with endpoint 1 and 2 setup as in or out, or is there something else that should be going on
18:06:40AlexPDoes it support more/different of these codes than Windows?
18:08:35gevaertskkurbjun: The bit that tells the host about your endpoints seems to work (that's part of enumeration). That should mean it gets to usb_drv_request_endpoint() properly.
18:09:15 Quit bertrik (Read error: 113 (No route to host))
18:09:28kkurbjunThe device state goes from powered, to default, to address, but it hasn't transitioned to configured yet, I wasn't sure if I should be expecting that transition to configured to happen after it receives the information requested in usb_storage_init_connection
18:09:29gevaertskkurbjun: as soon as usb_storage is initialised it starts an incoming transfer of up to 1024 bytes. That's the "mxx: ep: 1 in, len: 1024" from your logs.
18:10:06kkurbjunyeah, that's where the renesas is failing at the moment, it says the fifo isn't ready and it never seems to go ready
18:11:11gevaertsas soon as you get that working, mass storage should work
18:12:04kkurbjundoes that receive call in usb_storage have to be blocking? since this is only running at fullspeed at the moment my fifo's are only 64k in size so I think I need to run this part of the transfer with interrupts and fill it in in pieces
18:12:35 Quit tvelocity (Read error: 110 (Connection timed out))
18:13:00 Join tvelocity [0] (
18:14:35 Join nibbler__ [0] (
18:14:40gevaertsusb_drv_recv() should be nonblocking. as soon as you have the data, the driver should call usb_core_transfer_complete()
18:14:55kkurbjunok, that makes sense
18:15:52kkurbjunone other thing; is there a way I can see more detailed information on what is happening on the host (linux), right now lsusb is all that I'm using at the moment
18:17:20gevaertsyou can compile usb-storage with debug output, and you can look at usb traffic using usbmon (and possibly wireshark, depending on compilation options)
18:18:01kkurbjungreat, thanks!
18:25:17 Quit wpyh (SendQ exceeded)
18:28:59 Join bluebrother [0] (
18:30:21 Quit nibbler_ (Read error: 110 (Connection timed out))
18:32:36 Join kugel [0] (n=kugel@rockbox/developer/kugel)
18:32:59 Join flydutch [0] (
18:33:57kugelamiconn, pixelma: any idea why player and ondio don't have a way to restart rockblox? and if there would be a free and usefull combo?
18:36:38 Join evilnick_7 [0] (i=0c140464@gateway/web/ajax/
18:48:21CIA-38New commit by kugel (r20965): Define a restart button for Ondio and Player for Rockblox.
19:01:56 Join Jaykay [0] (
19:04:44Jaykayisn't FS #7593 out of date with r20959?
19:06:20 Quit bmbl ("Woah!")
19:08:09 Quit Thundercloud (Remote closed the connection)
19:12:10CIA-38New commit by kugel (r20966): Use a struct for the necessary status variables for rockblox. Will make dumping the status easier for the upcoming resume addition.
19:14:29 Join mt [0] (n=MTee@
19:15:53 Join saratogahome [0] (n=41becb3b@gateway/web/cgi-irc/
19:17:38***Saving seen data "./dancer.seen"
19:19:49kkurbjuntomers: does USB hid have support for tablets (touchscreens) and the like? I don't know if this is practical, but I thought it would be pretty cool to be able to use the touchscreen targets as a drawing pad
19:21:30tomerskkurbjun: Yes, of course. I will definitely would like to add support for touchscreens, making them act as a touch pad. It will not be part of my GSoC though, so maybe I won't spend time on it now. I'll try to work on it later.
19:22:56stripwaxis there any way for a codec to pass data/stats to debug_menu ?
19:24:13saratogahomestripwax: i think logf messages are viewable from debug right?
19:24:19kkurbjuncool, yeah, I can see that it wouldn't be part of the main project; I was more curious if it was possible. I think the HID support adds some pretty interesting possibilities for rockbox, I'm looking forward to trying it once I get this driver working :-D
19:24:37kkurbjunyep, logf messages are viewable from the debug screen
19:25:13 Join bubsy [0] (i=Bubsy@
19:25:27stripwaxah, cool. haven't used logf but that should do what I need - thanks!
19:25:44 Quit CaptainKewl ("( :: NoNameScript 4.02 :: )")
19:26:38saratogahomestripwax: what are you looking at?
19:27:55 Join Llorean [0] (
19:28:28stripwaxsaratogahome - nothing specific just yet, was wondering how to check/confirm on-target what the iram usage of tremor is at any point in time
19:29:25saratogahomestripwax: maybe its worth making the iram malloc stuff run on the sim, but with the buffer allocated in DRAM
19:29:51 Join DrMoos [0] (
19:29:52stripwaxsaratogahome - I think it does, currently
19:30:57 Join CaptainKwel [0] (
19:31:16 Quit moos (Read error: 104 (Connection reset by peer))
19:32:02 Nick DrMoos is now known as moos (
19:33:14bluebrotherlinuxstb: which file do I need for beastpatcher? bootloader.bin for beastpatcher and nk.bin for sendfirm?
19:33:38LambdaCalculus37bluebrother: Yes.
19:34:22bluebrotherhmm. Ok, but why does the beast reject the single boot nk.bin built from svn? I only managed to get it run using a patched OF
19:34:53bluebrotherapart from that my mtp changes on w32 seem to work fine.
19:35:32LambdaCalculus37bluebrother: I've gotten the same thing on my beast; it kept rejecting the single boot nk.bin. I gave up and just used the dual boot nk.bin for now.
19:36:03bluebrotherhmm. Not exactly the solution I wanted to hear :o
19:36:07*LambdaCalculus37 is trying to figure out how to get beastpatcher to compile in OS X
19:37:18*bluebrother wonders how the legal status of a beastpatcher.dll would be that rbutil loads dynamically from the web
19:37:48saratogahomeis it GPL?
19:39:14bluebrotheryes and no: the original code (the MTP_DLL in svn) is believed to be public domain, our code is BSD. We need to link it against MS' WM SDK though.
19:41:06 Quit tvelocity (Read error: 110 (Connection timed out))
19:41:32 Join tvelocity [0] (
19:43:19saratogahomebluebrother: i think the WM SDK is fine since thats an OS library and therefore GPL safe
19:43:32saratogahomebut do you have permission to GPL the original public domain code?
19:43:52 Quit flydutch (Read error: 104 (Connection reset by peer))
19:44:12bluebrotherwell, mcuelaere made it BSD. From the README in that source folder I think this is ok.
19:44:35*bluebrother hates it when nick completion doesn't work :/
19:45:18saratogahomeprobably ok then
19:45:32saratogahomei guess if anyone complains you could make it an exe and just call it seperately
19:46:50bluebrotherhaving an exe would be the other option. Linking in rbutil itself is a no-go as that would require VS for building (which currently doesn't work at all for rbutil), and having a dependency on the MS lib is bad as that would make rbutil required at least WMP10
19:47:10bluebrotherso it needs to be something external that we can load from somewhere
19:48:54 Join flydutch [0] (
19:49:39 Join _lifeless [0] (n=lifeless@
19:53:22Jaykaynext try - isn't out of date with r20959?
19:53:36Lloreangevaerts: What's blocking 9708? Just instability?
19:54:18*Llorean thinks in its current state of "your battery runs out while you're using it", USB on HD-based iPod at least isn't ready for 3.3
19:54:26gevaertsLlorean: I don't really know. It's not exactly an area I'm familiar with
19:54:41 Quit __lifeless (Read error: 60 (Operation timed out))
19:57:46AlexPJaykay: r.e. FS #9880 - I think I'm going to close it. I've committed those parts of the real changes that I think should be, and for the other part I'm not sure there is any point in chabge for change's sake
19:59:24JaykayAlexP: imo the code "looks better" now
19:59:40AlexPnow as in in SVN?
19:59:54JaykayAlexP: also i splitted some lines which contains more than 80 caracters
19:59:55jspeteronce I install rockbox, having put OF.mi4, pp5022.mi4, and .rockbox/ on the player, can I delete these files, or are they required for operation?
20:00:18JaykayAlexP: sorry, i meant it looks better with the patch
20:00:36AlexPI don't think it makes any difference
20:01:17Lloreanjspeter: They are required.
20:01:23jspeterk, thanks
20:01:28AlexPJaykay: For instance, with a paragraph after an \opt, shy do you further inline all lines following the first?
20:01:54JaykayAlexP: where?
20:02:21xnyhpsI could change the volume in Debian/KDE (not sure where it's handled) with my iPod, haven't tried any media player for Play/Pause.
20:02:57AlexPJaykay: e.g. changes after \opt(ipod){ at @@ -55,23 +49,21 @@
20:02:59 Quit dekaden|Z ("( :: NoNameScript 4.2 :: )")
20:05:46JaykayAlexP: i don't know..... wait a moment, i have to set up vmware
20:06:09AlexPJaykay: brb
20:07:41 Quit jspeter (Remote closed the connection)
20:08:46 Quit miepchen^schlaf ()
20:11:04 Join miepchen^schlaf [0] (
20:14:57 Quit stripwax (Read error: 104 (Connection reset by peer))
20:22:14 Quit Jaykay ("ChatZilla 0.9.84 [Firefox 3.0.10/2009042316]")
20:24:34 Quit saratogahome ("CGI:IRC (EOF)")
20:25:27 Join Jaykay [0] (
20:27:31AlexPJaykay: You had chance to look yet?
20:28:19JaykayAlexP: no. i need to get a newer vmware version...
20:33:38 Nick fxb is now known as fxb__ (
20:35:46 Quit Jaykay (Read error: 104 (Connection reset by peer))
20:36:48 Join Traveler5 [0] (
20:37:26Traveler5is it possible to have rockbox's bootloader load the .rockbox folder from a microsd card?
20:37:44LloreanNot without compiling your own version
20:38:57Traveler5but does the bootloader require drivers to read the microsd card then, or can I just change the directory to load from?
20:40:10bluebrotherit does need to know how to "talk" to the msd card, so it obviously needs drivers
20:40:24 Join Jaykay [0] (
20:40:27Traveler5alright, thanks
20:44:07Traveler5so if I were to make a program write to the microsd card by default, would the path be "/<microSD1>/"? (assuming the player has a microsd)
20:46:02LloreanDo you mean a Rockbox plugin (not a program) or an application on your computer?
20:52:04CIA-38New commit by kugel (r20967): Resume feature for Rockblox by dumping the state upon exiting and loading it when starting rockblox.
20:52:33 Join kugel_ [0] (
20:52:47 Quit kugel (Nick collision from services.)
20:52:53 Nick kugel_ is now known as kugel (
20:55:50Traveler5rockbox plugin
20:58:21 Quit perrikwp (" ajax IRC Client")
20:59:12LambdaCalculus37Traveler5: Then in that case, yes.
20:59:37tomersHow can I compile for all targets on my local machine automatically (i.e. personal build server) in order to test my work? Is there a script for that? (I have all the toolchains installed already)
21:00:24AlexPtomers: tools/release
21:00:35tomersAlexP: Thanks. I'll try
21:00:46AlexPIn that dir :)
21:01:39mtIf I want to commit certain changes only out of lots of others, should I just delete all the entries in the commit message except for the files I want to change ? (I hope my words make sense :) )
21:02:01bluebrotherdelete entries in the commit message?
21:02:02AlexPmt: You can list the files, including using wildcards
21:02:33AlexPmt: so svn commit /apps/whatever/x.c apps/something/else.c /apps/important/stuff.*
21:02:33bluebrotheryou can't edit the list of files to get committed from the commit message
21:02:51AlexPer, without the first / of course :)
21:03:01mtAlexP : thanks :)
21:03:23bluebrotherAlexP: he could use absolute paths :)
21:03:37AlexPbluebrother: he could, but best not to mix them :)
21:03:38mtbluebrother : I meant, when I do : svn commit, I get to write the comment and after that the list of modifications are listed, that's what I was talking about.
21:04:06kugelthe list of files is just for your info
21:04:19kugelafaik it even says that it is ignored
21:04:40*kugel never commits via command line
21:05:52CIA-38New commit by mt (r20968): Add read_uint8() and read_uint16be() to metadata_common.c/h
21:06:02bluebrothermt: as kugel said: there's a "all lines below this are ignored" marker above that. It's just for your information (and a thing I really like to see ;-)
21:06:33 Quit Traveler5 ("Java user signed off")
21:06:58*kugel uses an editor plugin where you can select and deselect files to commit nicely
21:07:01mtNow I see it's good that I asked first before committing. hehe
21:07:47gevaertsmt: not being careful there is a good way to get highscores :)
21:08:20 Quit evilnick_7 (" ajax IRC Client")
21:08:51mtgevaerts : Thanks for the tip ! :)
21:09:31 Join Thundercloud [0] (
21:17:40***Saving seen data "./dancer.seen"
21:34:19mtlinuxstb, saratoga : get_rm_metadata() works fine now on both a normal and a sim build.
21:40:57 Join Digital_Root [0] (
21:41:14Digital_RootSo a funny thing happened just 5 minutes ago
21:41:41Digital_RootI got fed up with trying to decrypt the firmware for my ipod nano 3g, I wasn't getting anywhere
21:43:39Horschtthen you accidentally all the firmware
21:43:55Digital_Rootand I started to attempt to install rockbox via the rockbox utility, obviously nothing worked, but when I tried to install it using the install setup for the Olympus M: robe100 it worked. Now obviously I can't do anything with it, I don't have the right key config, but I have rockbox on my 3rd gen nano that ownt work
21:44:23gevaertswell obviously
21:44:45 Quit ender` (" Little expense had been spared to create the impression that no expense had been spared.")
21:44:49gevaertsYou can also install the mr100 build on an apple II...
21:44:57bluebrotherit obviously will work. The m:robe requires putting a bootloader file on the disk and that's all
21:45:14bluebrotherso basically you can install the m:robe bootloader on anything that is writeable
21:45:32Digital_Rootwell see thats good to know, because now I know that I didn't just brick my iPod
21:45:40bluebrotherIn fact I did this various times when reworking the bootloader class. The m:robe simply was a path in the filesystem.
21:46:05bluebrotherbricking an Ipod is almost impossible.
21:46:15bluebrothertoo much safety nets in the flash bootloader.
21:46:32 Join firedix [0] (n=firedix@
21:46:41Digital_Rootmeh hit it with a hammer
21:47:05Digital_Rootok, so about the farthest I got is as far as everyone else, the dmg extracted
21:47:13gevaertsthat is indeed the recommended way to brick it
21:51:13Digital_RootSo, no one on here has an ingenious idea as how to open up the dmg file then?
21:52:17mtSince RM only has four strings for metadata, It's OK to use id3v1buf[0-3] to read them, right ?
21:52:22 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/
21:52:28AlexPDigital_Root: No, otherwise it would have been done.
21:53:13 Join ender` [0] (
21:53:23 Join _fml [0] (n=4fd3f3de@gateway/web/cgi-irc/
21:53:57Digital_Rootmk, well I might make another go at it later today, I love beating my head against a brick wall after all.
21:54:31_fmlkugel: hello. I have a couple of suggestions regarding the last commit to rockblox. May I shoot?
21:54:58 Join robin0800 [0] (
21:56:39 Join saratoga [0] (i=9803c6dd@gateway/web/ajax/
21:56:49_fmlkugel: ok. 1. IMHO it would be more readable to use just sizeof(rockblox_status) instead of sizeof(struct _rockblox_status) The effect is the same, just more redable (IMO) If you'd pass the address to put the data to as a parameter then your variant would have to be used
21:57:30Jaykayscorche, scorche|sh: ping
21:57:32kugelI can't really agree
21:57:37_fmlkugel: 2. In the dump function, the check after file open should be "<0", not "<=0"
21:58:37 Quit Digital_Root ("Leaving")
21:59:10 Part saratoga
21:59:27 Quit jordan` (Read error: 104 (Connection reset by peer))
22:00:02 Join saratoga [0] (i=9803c6dd@rockbox/developer/saratoga)
22:00:20 Quit robin0800 (Remote closed the connection)
22:00:29_fmlkugel: I'd also change the dump function a bit to have close(fd) just once, at the end. But it's just my personal programming style. I define a "ok" variable and do cleaning up at the end of the function. E.g. if (fd>=0) close(fd)
22:01:25_fmlkugel: what don't agreee with in 1.? Readability?
22:02:01kugelI'd always use sizeof(struct...)
22:02:07 Join robin0800 [0] (
22:02:25kugelI want to write the complete struct. that makes it more clear
22:03:31kugelalso if I say sizeof(rockbox_status) it depends on that it's not a pointer. you can't do anything wrong with sizeof(struct...)
22:03:37_fmlLambdaCalculus37: hey! Are you sure your last change to the album art manual is correct? Isn't the specific trackname tried first? (I judge by the code, not by the intented behaviour)
22:03:59 Join mt_ [0] (n=MTee@
22:04:22kugelI think the word struct in it makes it more readable
22:04:44 Join jordan` [0] (i=gromit@
22:04:46Unhelpfulthe first thing tried is replacing the extension of the track filename with {jpeg,jpg,bmp}
22:04:57 Nick mt is now known as Guest73676 (n=MTee@rockbox/developer/mt)
22:04:59_fmlkugel: with the latter I agree (see above) With the former... maybe too. It could be that rockblox_status has two struct fields in it that are read with two file ops
22:05:41 Nick mt_ is now known as mt (n=MTee@
22:06:03LambdaCalculus37_fml: That was pixelma that pointed that out to me, so ask her.
22:06:06_fmlUnhelpful: BTW: wouldn't it be good to also try folder.jpeg, not only .jpg?
22:06:37kugelwell, we've discussed it
22:06:45saratogafolder.jpg is the WMP standard, which is why it was included
22:06:47Unhelpful_fml: WMP doesn't create folder.jpeg
22:07:00kugelfolder.jpg is just for compability with wmp, apparently
22:07:17kugeland since wmp doesn't see folder.jpeg, we don't need to be compatible with that
22:07:32Horschtmost players on PC support the folder.jpg method
22:07:39kugel(that's not my opinion but the one of several others)
22:07:40saratogalots of windows software uses it as well, i ran some foobar plugin to get album art the other day and it created Folder.jpg files too
22:07:41_fmlOh, sorry, didn't know. I just thought if, for JPEG, we try jpeg and jpg then it'd be consistent to do the same for the folder. And some programs create JPEGs as .jpeg
22:07:49Horschtoh wait... jpeg
22:08:57kugel_fml: can you rephrase it please?
22:09:49_fmlkugel: track.jpeg, track.jpg, but only folder.jpg (no folder.jpeg)
22:10:09kugel_fml: I mean this "with the latter I agree (see above) With the former... maybe too. It could be that rockblox_status has two struct fields in it that are read with two file ops"
22:10:33 Join fdinel [0] (
22:12:10_fmlkugel: the former = the thing with the pointer
22:12:42CIA-38New commit by unhelpful (r20969): Add a simple benchmark plugin for the scaler, using the pluginlib import of resize_on_load.
22:14:29_fmlkugel: i.e. if the struct to read to is a pointer then you have to use sizeof(struct ...)
22:15:37kugelyes, that's what I was saying ;)
22:16:37_fmlkugel: that's what *I* was saying :-)
22:17:44JaykayAlexP: sorry, vmware didn't work up to now... and now i want to go to bed, but i'll revisit the patch tomorrow.... maybe i can reduce it to a few more useful things
22:18:00 Join kugel_ [0] (
22:18:06AlexPOK, no worries
22:18:08 Quit kugel (Nick collision from services.)
22:18:26 Nick kugel_ is now known as kugel (
22:19:02kugelI didn't understand that you agree with the latter (you initially said you find sizeof(rockblox_status) more readable. I also didn't understand what you mean with "two struct fields are read with two file ops"
22:20:27 Quit Guest73676 (Read error: 110 (Connection timed out))
22:22:38_fmlkugel: if you had two fields in rockblox_status, say, a and b, which are structs as well, then you could read them with two different file reads (if the two blocks were not one after the other in the file). You'd use sizeof(struct _a) and similar for b.
22:24:32Unhelpfulthey needn't be structs, you could do the same for an int field
22:26:53_fmlUnhelpful: yes, but then you'd use sizeof(int) (type), not sizeof(field_name) I wanted to make clear that the type would be used and not the field name
22:27:57Unhelpfulah! i didn't understand that. wouldn't sizeof(struct.field) work for int, also?
22:28:31 Quit Jaykay ("ChatZilla 0.9.84 [Firefox 3.0.10/2009042316]")
22:33:45 Quit mt (Read error: 110 (Connection timed out))
22:33:57_fmlUnhelpful: probably yes. Would it work for structs? I'm not sure, but maybe yes too. But with sizeof(struct ...) it surely would. The drawback of using sizeof(struct ...) is that you duplicate information, in this case the type of the field (once in the field declaration and once in the read op)
22:34:27 Join mt [0] (n=MTee@
22:34:39 Join {phoenix} [0] (
22:35:03Unhelpfuli think that sizeof(var.field) might be more readable and maintainable than sizeof(type), if that's what you're suggesting :)
22:35:17 Join robiin0800_ [0] (
22:35:37kugelnot in the sizeof(struct...) case imo
22:36:22 Quit robiin0800_ (Remote closed the connection)
22:36:32kugelexcept if the struct is a field of another struct maybe.
22:36:45 Join robin0800_ [0] (
22:38:19_fmlUnhelpful: yes, that's what I was suggesting. But it only works if the struct you read to is a struct itself, not a pointer to it
22:38:49Unhelpfulyou couldn't use sizeof(struct_ptr->field) ?
22:38:55_fmlNow we have a struct itself so my suggestion could be used
22:39:20*kugel hands Unhelpful TIME_AFTER()
22:39:30_fmlBut if the struct is passed as a param to the function (as a pointer to struct), you'd have to use sizeof(struct ...)
22:40:24_fmlUnhelpful: this would probably work (sizeof(str_ptr->field)), but we don't have a pointer to a struct with a sub-struct field now
22:40:33Unhelpfulkugel: i don't see how TIME_AFTER is needed if i'm just busy-waiting for the tick?
22:40:47_fmlkugel: will you correct the file handle check or should I do this?
22:41:14kugel_fml: I'll do that
22:41:40kugelUnhelpful: "needed"? It's not like TIME_AFTER is a complex construction
22:41:41linuxstb_fml: In the case where a pointer to a struct is passed to a function, you can just do sizeof(*p)
22:41:48_fmlkugel: ok
22:42:00Unhelpfulkugel: yes, but i just don't see any reason to use it?
22:42:04linuxstbUnhelpful: The pointof TIME_AFTER() is that it is wrap-safe.
22:42:10kugelUnhelpful: better readable
22:42:17_fmllinuxstb: yes, that's what I was saying. But as of now, we use the struct itself, not the pointer to it
22:42:37Unhelpfullinuxstb: i know, but in this specific case, it's a busy wait for the tick, so the test t1 != t2 is also wrap-safe
22:43:06kugelI was rather talking about "while (t2 - t1 < HZ || count < 10)"
22:43:22kugelI don't see any reason do not use TIME_AFTER
22:43:26_fmlUnhelpful: could you please look at ? It's my shot at the search order for the AA in the manual.
22:44:16Unhelpfulah, you'd rather see that as TIME_AFTER(t1 + HZ, t2)?
22:45:03CIA-38New commit by kugel (r20970): Correct a little typo.
22:45:26kugelUnhelpful: yes
22:46:39*kugel also thinks we should have some kind of tick_t typedef when messing with current tick, instead of long
22:46:39Unhelpful_fml: that looks like what actually happens in the source. you might want to mention size tags - WPS searches with them, and they're not used for folder.jpg, so a size-specific file for any of the other patterns will override folder.jpg (but not for PictureFlow)
22:46:55 Join ZincAlloy [0] (n=d9eecd5a@gateway/web/cgi-irc/
22:48:49dfktsaratoga, i guess you're the person to ask that - could this be applied to rockbox? -
22:49:17dfktis this "open source", since they provide all info to their "bbe-like recipe"?
22:49:18UnhelpfulCONTRIBUTING is anti-typedef, though we have a few, like fb_data
22:50:09_fmlUnhelpful: try to describe that in a short and understandable way :-) I think I'll omit the sizes in the first try. If someone insists then we'll try :-)
22:50:38Unhelpful_fml: it's not *short*, but you could present the list twice :)
22:51:37_fmlUnhelpful: :-) Is the desired (=preferred) size specified in the WPS tag? Or where does it come from?
22:51:43saratogadftk: that algorithm could be implemented, but i don't think they provide any code
22:52:21dfktsaratoga, the diagram looks rather specific
22:52:35Unhelpful_fml: from the WPS cover load tag. that's why PictureFlow doesn't use one - it uses a calculated (and not always square) size based on the display dimensions
22:52:48dfktnot sure how much "secret sauce" goes into it, considering it's a free plugin
22:52:53Unhelpfulhow does this look?
22:52:55 Quit Thundercloud (Remote closed the connection)
22:53:13 Quit robin0800 (Read error: 113 (No route to host))
22:53:44CIA-38New commit by alle (r20971): (Maybe the first) correction of the order the album art images are searched. Also correct some typos. Preferred sizes are not described yet.
22:54:09saratogadftk: yeah I don't know much about the idea, but if you just want to apply compression to certain frequency bands, thats not a hard thing to do, though it will be fairly slow
22:55:05dfktsaratoga, i see... the vst plugins by that company can eat up quite some cpu cycles - probably not a as optimized as BBE stuff
22:56:18dfktit sounds very good, however :)
22:56:24saratogai don't know enough about DSP to say how it would be implemented though in rockbox, maybe as an extension to the EQ or the (work in progress) dynamic compression patch
22:56:57Unhelpfulif you're looking to compress overall loudness, you could go a bit more quickly by periodically evaluating either RMS of the audio or equal loudness curve...
22:57:26saratogai think the idea was to compress just the bass frequencies so it has more "kick"
22:57:40Unhelpfulthe latter will need an FFT, though, and at that point you could just as easily do things per-band....
22:57:41dfktyeah, it's a psychoacoustical "exciter"
22:58:26saratogaprobably best as an IIR + compression, though i'm not sure how you implement a compression curve in rockbox without a lot of memory or being really slow
23:00:52Unhelpfuli guess that should be + HZ - 1 if i'm using TIME_AFTER, actually. not that it makes a big difference, i just want to run for at least either 1s or 10 iterations
23:01:09 Quit petur ("Zzzzz")
23:05:17 Quit firedix ("Ex-Chat")
23:10:14 Quit mt (Read error: 110 (Connection timed out))
23:14:10 Join Ubuntuxer [0] (
23:14:28_fmlHow can I set svn properties for a file? They are not set for album_art.tex
23:16:01kugel_fml: svn prop-get something, then prop-list
23:16:50kugel(propget to show the properties filled for the file, the proplist to see the values for each individual property)
23:16:56CIA-38New commit by alle (r20972): Set SVN properties to the standard Rockbox values
23:17:39LambdaCalculus37_fml: svn pset svn:keywords "Author Id Date Revision" <filename>
23:17:41LambdaCalculus37_fml: svn pset svn:keywords "Author Id Date Revision" <filename>
23:17:41DBUGEnqueued KICK LambdaCalculus37
23:17:41LambdaCalculus37_fml: svn pset svn:keywords "Author Id Date Revision" <filename>
23:17:41LambdaCalculus37_fml: svn pset svn:keywords "Author Id Date Revision" <filename>
23:17:41_fmlUnhelpful: thanks
23:17:42***Saving seen data "./dancer.seen"
23:18:14LambdaCalculus37_fml: Well, once was enough, but blame X-Chat for being stupid and pasting it four times. :P
23:18:20_fmlLambdaCalculus37: got it! :-)
23:27:23 Join Thundercloud [0] (
23:33:42 Quit bluebrother (Read error: 113 (No route to host))
23:36:47 Quit _fml ("CGI:IRC 0.5.9 (2006/06/06)")
23:44:51 Quit rasher (Read error: 104 (Connection reset by peer))
23:48:07 Join rasher [0] (
23:51:22UbuntuxerHi, I'm dividing the Patch FS #10193. Please have a look.
23:58:02 Part Ubuntuxer
23:59:00 Join skull0 [0] (n=skull0@

Previous day | Next day