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

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

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

#rockbox log for 2010-07-02

00:01:26Buschelfaad is faster with -O2 on PP5022 (31.6 MHz vs. 32.4 MHz). Is this also valid for other CPUs?
00:04:23bluebroth3rpixelma: sorry, got disconnected locally in between. My guess would be that the latter if possible at all would be anything than trivial, i.e. need someone with good LaTeX knowledge.
00:04:54bluebroth3rbut that's way beyond my knowledge, so I guess the only way would be to prepare things beforehand.
00:06:14 Quit bertrik (Quit: :tiuQ)
00:06:54 Join anewuser [0] (anewuser@unaffiliated/anewuser)
00:18:06n1sBuschel: is your change in libfaad/bits.h/get dword alignment safe? (the comment seems to suggest it is not)
00:18:30n1ss_/get dword_:get_dward_
00:21:32 Quit pamaury (Remote host closed the connection)
00:23:19 Quit Barahir_ (Ping timeout: 240 seconds)
00:24:25 Join Barahir [0] (
00:26:40 Quit Dhraakellian (Ping timeout: 248 seconds)
00:27:03Buscheln1s: I should have removed the comment. The "#ifdef ARM" path was not used as ARM was not defined. So, the special handling for ARM was not active −− there was and is only the difference between big and little endian.
00:28:23 Join Dhraakellian [0] (
00:28:52Buschelbtw, in the faad decoder there is some strange code and #ifdef'ing
00:29:23n1sit's pretty messy
00:29:56n1spretty nice speedup you got out of it though
00:30:16Buschelyep :)
00:32:49CIA-6New commit by Buschel (r27228): Remove false comment.
00:33:51funmanBuschel: i can test amsv2 if you want
00:34:23CIA-6r27228 build result: All green
00:34:41Buschelfunman: yes, would be fine. I will add a FS-entry to collect the results
00:39:40Buschelfunman: FS #11450
00:40:29n1sBuschel: i can try on h300 tomorrow
00:40:54 Part AzureWurk
00:41:27n1si have been thinking about making a larger scale compiler switch tuning, where all codecs are built with O[123] and tested in one go for my targets but haven't gotten around to it yet
00:42:07Buschelfunman/n1s: as I am on a 4-day trip until Monday evening, you will have plenty of time for testing ;)
00:42:38n1sok, i'll try to remember to do it tomorrow then :)
00:44:00Buschelhave a good night, I need to have some sleep now (hope to fall to sleep at all −− it is pretty warm here even at night...)
00:44:01***Saving seen data "./dancer.seen"
00:45:19 Quit Buschel ()
00:46:03funmanhm one of the test_file got fucked up on clip+ : 0.04s to decode
00:47:23 Quit panni_ (Read error: Connection reset by peer)
00:47:37 Join panni_ [0] (
00:48:06 Join krazykit [0] (~kkit@
00:49:19n1sfunman: from just switching compiler options?
00:49:30funmannope with svn
00:49:53funmanI'm pretty sure the iram is large enough on AMS but i must test the codec speed
00:51:17n1syes, the faad decoder can require a lot of memory, is the clip+ using the iram as codec buffer like the clip?
00:51:51funmanhm right :)
00:52:11funmanso it uses it already
00:52:21funmani must check on fuzev1 then
00:53:23funmanthe failing file is the he-aac one
00:58:04n1sthat probably isn't a very good testcase as it uses a lot of other features of the codec than regular aac AFAIU
00:58:16n1salso he-aac is dead slow currently
01:00:33funmani'm pretty sure it used to decode correctly
01:00:51funman-O2 gives ~ 5% speedup
01:01:34n1smaybe buschel's batch broke something in he decoding then
01:01:40saratogaAAC-HE worked on the clipv2 when i tried a while ago
01:02:00 Quit funman (Quit: leaving)
01:02:38 Quit ender` (Quit: I think I remember an episode of MacGyver where he overthrew a violent dictator with a rubber band, 2 bottle caps, and some navel lint.)
01:06:24saratogasomeone should update CodecPerformanceComparison now that AAC, MP3 have been improved
01:06:55*n1s points out that vorbis has gotten some small speeups too :)
01:08:05 Quit DerPapst (Quit: Leaving.)
01:10:17*n1s would also like to get his memory juggling patch fro tremor in but it still has a bug somewhere
01:14:37 Join bieber_ [0] (
01:15:14 Join binaryhermit [0] (
01:15:36 Quit n1s (Quit: Lämnar)
01:15:42bieber_JdGordon: The new newline handling is causing a problem: now anything that has Vf/Vb tags on the same line as a viewport declaration is inserting a blank line at the beginning of the viewport
01:30:08 Join evilnick_BS [0] (620ec24b@rockbox/staff/evilnick)
01:30:24 Nick fxb__ is now known as fxb (
01:37:03 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
01:40:38 Quit Ingrid_en_Henk (Ping timeout: 276 seconds)
01:41:38 Quit z35 (Read error: Connection reset by peer)
01:45:08 Quit Tarvis (Quit: /quit /quit /quit /quit /quit /quit /quit /quit /quit /quit /quit /quit /quit /quit /quit /quit /quit /quit /quit /quit /quit)
01:48:23 Join z35 [0] (
01:57:07 Quit binaryhermit (Read error: Connection reset by peer)
01:57:22 Quit z35 (Quit: Leaving)
01:57:25 Join binaryhermit [0] (
02:00:08 Quit bieber_ (Ping timeout: 276 seconds)
02:00:20 Quit kachna (Ping timeout: 265 seconds)
02:00:30 Quit Topy (*.net *.split)
02:00:31 Quit Strife89 (*.net *.split)
02:00:31 Quit BlakeJohnson86 (*.net *.split)
02:00:31 Quit delt (*.net *.split)
02:00:31 Quit aevin (*.net *.split)
02:00:31 Quit JdGordon (*.net *.split)
02:00:31 Quit leavittx (*.net *.split)
02:00:31 Quit Bowbles (*.net *.split)
02:00:31 Quit ThomasAH (*.net *.split)
02:00:32 Quit nk_ (*.net *.split)
02:00:32 Quit Hadaka (*.net *.split)
02:01:55 Join Strife89 [0] (
02:02:07 Join S_a_i_n_t [0] (S_a_i_n_t@
02:02:26 Join leavittx [0] (~leavittx@
02:03:54 Join Topy44 [0] (~topy@2a01:608:ffff:2619:2e0:4dff:fe83:a45e)
02:09:17 Join bieber_ [0] (
02:09:37 Quit evilnick_BS (Ping timeout: 252 seconds)
02:10:55 Join delt [0] (
02:11:31 Join aevin [0] (eivindsy@unaffiliated/aevin)
02:11:57 Join 13WAA3UUX [0] (
02:11:57 Join BlakeJohnson86 [0] (~bjohnson@2002:1876:a27b:0:227:13ff:fe65:1262)
02:11:57 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
02:11:57 Join Bowbles [0] (
02:11:57 Join ThomasAH [0] (
02:11:57 Join nk_ [0] (
02:11:57 Join Hadaka [0] (
02:12:01 Quit liar (Ping timeout: 240 seconds)
02:13:46 Join Tarvis [0] (
02:15:10TheSevenS_a_i_n_t: have you seen any "stkov nand" panics with recent nano2g builds?
02:15:48TheSevenrecent = since about a week
02:16:10S_a_i_n_tI haven't seen any stkov panics in the last two months or so, possibly longer.
02:16:29TheSeveneven with such recent versions? that's fine then
02:16:34S_a_i_n_tI have however been getting random data aborts that seem to relate to skipping tracks too quickly.
02:16:40 Quit MethoS- (Remote host closed the connection)
02:16:47TheSevenhmm that's a different issue
02:17:12*TheSeven was suspecting his storage idle callback commit to cause stkovs because someone reported one in the nand thread
02:17:47S_a_i_n_tBut, no, I update my builds almost daily and I have not seen a stkov panic on the Nano2G for quite literally months.
02:17:52 Nick fxb is now known as fxb__ (
02:19:29TheSevenand you aren't using my sector cache patch either, are you? (it might be hiding the stkov)
02:20:38CIA-6New commit by bieber (r27229): Theme Editor: Altered tag table entry for %pb to make image optional, implemented progress bar rendering in theme editor
02:20:46S_a_i_n_tOn one of my Nano2Gs I am using that patch.
02:20:53S_a_i_n_tOn the others, I'm not.
02:22:21CIA-6r27229 build result: All green
02:27:49CIA-6New commit by bieber (r27230): Theme Editor: Renderer will not insert a newline at the end of a line if it's the same line in the wps file as the viewport declaration, to avoid ...
02:29:36CIA-6r27230 build result: All green
02:30:43bieber_JdGordon: I've modified my renderer not to insert newlines when you have a %V(...)%Vf(...)%Vb(...) line, but we should fix the parser so it doesn't count the Vf and Vb as a separate line. My suggestion is to change the behavior of the \n flag in the parser so that it just eats a newline if one happens to occur immediately after the tag (or at least after some whitespace) and then put the \n flag on the end of both Vf and Vb's
02:30:43bieber_ argument lists in the tag table
02:32:43 Quit bieber_ (Remote host closed the connection)
02:32:53 Quit 13WAA3UUX (Ping timeout: 248 seconds)
02:42:16 Quit bieber (Ping timeout: 260 seconds)
02:42:32 Join bieber [0] (
02:44:02***Saving seen data "./dancer.seen"
02:44:41 Join leavittx_ [0] (~leavittx@
02:58:45JdGordonbieber: Vf, Vb should only not make a new line if it is attached to V() otherwise it should
03:11:10 Quit BlakeJohnson86 (*.net *.split)
03:11:10 Quit JdGordon (*.net *.split)
03:11:10 Quit Bowbles (*.net *.split)
03:11:10 Quit ThomasAH (*.net *.split)
03:11:10 Quit nk_ (*.net *.split)
03:11:10 Quit Hadaka (*.net *.split)
03:15:37 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
03:17:49 Quit bieber (Ping timeout: 240 seconds)
03:18:02 Join bieber [0] (
03:20:36 Join rvvs89 [0] (
03:22:54 Join BlakeJohnson86 [0] (~bjohnson@2002:1876:a27b:0:227:13ff:fe65:1262)
03:22:54 Join Bowbles [0] (
03:22:54 Join ThomasAH [0] (
03:22:54 Join nk_ [0] (
03:22:54 Join Hadaka [0] (
03:26:17 Join ruckus [0] (
03:26:38ruckushey people
03:27:07ruckusi reported the bug about the backlight with the metronome earlier today, and I see the change in the subversion
03:27:39ruckusif i donwnload the most current build, i shouldn't experience the bug anymore, right?
03:27:46JdGordonyeah sorry about that.. we usually dont fix bugs
03:28:30ruckustsk tsk, funny people
03:29:29ruckusseriously though, i downloaded the most current build about an hour ago, and it had the same problem
03:31:49 Quit panni_ (Read error: Connection reset by peer)
03:34:49S_a_i_n_tWhich specific commit are you talking about?
03:35:02S_a_i_n_truckus: ^
03:35:31S_a_i_n_tI don't personally remember seeing a commit specific to the text-viewer backlight myself
03:38:12JdGordonok, so comment in the bug pae that it isnt fixed
03:39:02 Quit Bowbles (Ping timeout: 248 seconds)
03:39:53 Join Bowbles [0] (
03:41:05S_a_i_n_tYes, you're quite correct. It is still occuring, however bugs will spring up from time to time, it's simply impossible to excpect everything to work as expected 100% of the time.
03:41:33S_a_i_n_truckus: Have you actually lodged a bug report of Flyspray, or simply commented about it here on IRC?
03:42:11ruckusi commented here
03:42:38ruckusand people who seemed like devs started talking about
03:43:28S_a_i_n_tHow do you determine "people who seem like devs"?
03:43:29ruckus(teru and pixelma)
03:43:40ruckusthey started talking about the code itself
03:43:47S_a_i_n_tMake a bug report on the bug tracker please, its easier for everyone.
03:48:22ruckusah crap. i am just about to leave work. :(
03:49:42 Quit BlakeJohnson86 (Ping timeout: 248 seconds)
03:50:15 Quit Tarvis (Read error: No route to host)
03:54:08 Quit ruckus (Quit: Leaving.)
04:08:50deltok, thanks for the help everyone!
04:08:51 Part delt
04:12:57 Join BlakeJohnson86 [0] (~bjohnson@2002:1876:a27b:0:227:13ff:fe65:1262)
04:16:44 Quit bieber (Ping timeout: 240 seconds)
04:19:18 Join bieber [0] (
04:23:25 Quit amiconn (Disconnected by services)
04:23:27 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:23:45 Quit pixelma (Disconnected by services)
04:23:47 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:23:47 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:24:02 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:28:00JdGordonbieber: quick request of the %Vp tag... can you make it so the 2nd half use the "info not found" code instead of the "found" code (i.e param 3 instead of 2?
04:31:43 Quit Barahir (Ping timeout: 260 seconds)
04:32:53 Join Barahir [0] (
04:36:51 Quit bieber (Ping timeout: 265 seconds)
04:37:09 Join bieber [0] (
04:43:14 Quit rado1 (Ping timeout: 240 seconds)
04:44:06***Saving seen data "./dancer.seen"
04:45:36 Quit TheSeven (Ping timeout: 264 seconds)
04:48:17 Join rado1 [0] (
04:49:01 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:53:37 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
05:08:34 Quit bieber (Ping timeout: 252 seconds)
05:08:49 Join bieber [0] (
05:16:27 Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier.
05:24:20 Join ThalionX [0] (
05:25:18 Quit Strife89 (Read error: Connection reset by peer)
05:30:55 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
05:34:32 Quit Horscht (Quit: Verlassend)
05:37:24 Join Strife89 [0] (
06:34:26bieberJdGordon: Sure thing
06:35:39CIA-6New commit by Buschel (r27231): r27225 broke AAC HE profile decoding due to missing check for buffersize. Introduce a check of needed buffersize and decide whether to use a static ...
06:37:28CIA-6r27231 build result: 18 errors, 0 warnings (Buschel committed)
06:40:26CIA-6New commit by Buschel (r27232): Fix red.
06:41:20 Quit anewuser (Quit: Sixth edition of 25 samples + 48 hours + detroit party (july 31) (maybe streamed online!))
06:42:02CIA-6r27232 build result: All green
06:44:08 Join CGL [0] (~CGL@
06:44:09***Saving seen data "./dancer.seen"
06:47:33CIA-6New commit by bieber (r27233): Theme Editor: Switched the order of the last two arguments to %Vp in rendering
06:49:07CIA-6r27233 build result: All green
06:49:16 Quit CGL (Remote host closed the connection)
06:50:27 Join bieber_ [0] (
06:50:30 Join wombat23 [0] (
06:52:37 Quit wombat23 (Remote host closed the connection)
06:59:34JdGordonbieber: that last commit? I meant if the list is going to show 6 items, then the first 3 should show the firt code param and the second 3 should show the 2nd code param
07:00:12JdGordonits always using the 2nd code param now?
07:00:19bieber_Ohh, you mean to make it look like the list is midway through loading?
07:01:24bieber_I thought you just meant to flip the order of the parameters
07:03:02bieber_Good idea :)
07:04:35JdGordonhmm, looks like you support fewer tags in the viewer than svn?
07:05:28JdGordonmaybe only 5 or 6 missing
07:05:48JdGordon*eventually* I want to make that support all tags
07:05:51bieber_I haven't done any of the RTL-related tags yet
07:06:17bieber_And I still need to handle all the special cases for the various conditional tags
07:07:28 Quit bieber (Ping timeout: 260 seconds)
07:07:30bieber_I _should_ have everything supported within the first couple days of next week
07:07:40 Join bieber [0] (
07:12:14 Join wombat23 [0] (
07:12:32 Quit binaryhermit (Ping timeout: 240 seconds)
07:16:28 Quit wombat23 (Ping timeout: 245 seconds)
07:18:00CIA-6New commit by bieber (r27234): Theme Editor: Switched the last two parameters of the Vp tag back to their original orientation, made the playlist display the first half with ID3 ...
07:19:29CIA-6r27234 build result: All green
07:20:50bieber_Aw shucks, you can't ammend a commit message, can you?
07:21:25 Quit t0rc (Quit: Leaving)
07:24:12JdGordonthis aint git!
07:24:58 Quit ThalionX (Quit: WrongPlanet forever!)
07:25:28JdGordonbieber: much better :) I see you havnt got it doing a full update after the reparse yet?
07:25:52 Join LinusN [0] (linus@rockbox/developer/LinusN)
07:26:02bieber_By all means, it should be. The image is being cleared and rerendered every time it gets reparsed, but for some reason the viewer isn't showing it
07:26:33bieber_I think the QGraphicsScene just isn't sending its sceneChanged signal every time it gets cleared, for some reason
07:28:42bieber_Oh wait, found it
07:29:06bieber_Apparently by default the viewer only changes its view if a certain amount of the scene changes
07:34:00bieber_Aanddd, it turns out that's not the real problem :/
07:34:00 Join ender` [0] (
07:39:21 Quit rvvs89 (Changing host)
07:39:21 Join rvvs89 [0] (ivo@pdpc/supporter/base/rvvs89)
07:49:52 Join kachna [0] (
07:58:08CIA-6New commit by bieber (r27235): Theme Editor: Fixed random outlining in default viewport, working on getting the viewer to refresh on re-render
07:59:49 Quit BHSPitMonkey (Remote host closed the connection)
07:59:50CIA-6r27235 build result: All green
08:00:02CIA-6New commit by jethead71 (r27236): AMS SoC's: Some register bit changes need interrupt protection: timer API and CGU_PERI.
08:01:42CIA-6r27236 build result: 0 errors, 32 warnings (jethead71 committed)
08:07:43CIA-6New commit by jethead71 (r27237): pl081: Fix r27236 warnings. Need system.h.
08:08:03DhraakellianDoes rockbox understand id3v2 replaygain tags as written by the mp3gain utility?
08:08:43Dhraakellianin other words, should I tell mp3gain to write id3v2 tags, or should I tell it to write APEv2 and run the files through
08:09:00 Quit Strife89 (Quit: Bed.)
08:09:13CIA-6r27237 build result: All green
08:10:03*Dhraakellian does both
08:11:05 Join esperegu [0] (~quassel@
08:15:34JdGordonrockbox doesnt read ape tags in mp3 files
08:15:46DhraakellianI know this
08:15:55Dhraakellianthus the comment about running it through
08:16:43*Dhraakellian ponders whether that's a well enough known thing for him to actually get snippy
08:17:09Dhraakellianit converts the APEv2 RG tags to more common id3v2 ones
08:18:01*Dhraakellian shrugs, as it's past his bedtime
08:27:26*JdGordon thinks it is never good to get snippy if you are asking questions
08:41:42bieber_Apparently the scene occasionally renders the last parse tree instead of the latest one, for no discernable reason :/
08:42:01bieber_After I spend forever trying to figure out why the viewer isn't refreshing, it turns out that it is
08:44:11***Saving seen data "./dancer.seen"
08:47:45JdGordonmessing up pointers somehow?
08:48:31bieber_I have no idea, I'm trying to track it down
08:48:51bieber_I have figured out that it's not random, which at least means it shouldn't be an uninitialized variable or anything like that
08:49:17bieber_If you open a new document, hit enter followed by "abcd", it fails to render the 'd' every time
08:49:42JdGordonthat is messed up :)
08:50:24JdGordonsomething seems to be eating the d
08:50:35bieber_But then if you continue with more stuff, the d shows up
08:50:51JdGordonnot if you keep hittting d
08:51:19JdGordond at the end of the line is killing it
08:51:41JdGordoni also
08:51:44JdGordonand I bet f
08:51:54JdGordonnot f
08:52:35bieber_Holy crap
08:52:41bieber_This is just bizarre
08:53:13JdGordonyour toString() must be eating it... those are the letters which do it
08:53:47bieber_Ohhh, it's not showing up in the parse tree either
08:53:55bieber_How did I not check that to begin with :/
08:54:21bieber_I wonder if the problem could be in the parser
08:54:57JdGordongood luck finding it :) I'm off
08:57:37 Quit bieber_ (Remote host closed the connection)
09:00:14 Join petur [0] (~petur@rockbox/developer/petur)
09:03:01 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
09:06:45bertrikjhMikeS, regarding r27236, wouldn't it have been easier to just move the CGU_PERI manipulation out of interrupt context?
09:07:21bertrikIt seems that ascodec-as3525 does the in-interrupt CGU_PERI manipulation (I think this file is way too complicated anyway, and it never really got clear to me why it had to be made so complicated)
09:12:13 Join flydutch [0] (
09:17:21 Join antil33t [0] (
09:17:44 Quit antil33t (Client Quit)
09:19:37 Quit bertrik (Quit: :tiuQ)
09:20:19 Join n1s [0] (~n1s@rockbox/developer/n1s)
09:21:49jhMikeSbertrik: perhaps, but then again, wouldn't it get left on if the transfer is completely async?
09:26:17 Join pondlife [0] (~Steve@rockbox/developer/pondlife)
09:28:18pondlifeSlightly OT question - is there any way I can change my e-mail address for the mailing lists without unsubscribing and resubscribing?
09:29:39pondlifenm, sussed it
09:30:15 Join Zagor [0] (bjst@rockbox/developer/Zagor)
09:30:39 Join einhirn [0] (
09:36:44 Part LinusN
09:36:47 Join LinusN [0] (linus@rockbox/developer/LinusN)
09:38:37 Join liar [0] (
09:43:19 Join Rob2223 [0] (
09:44:41 Part pondlife
09:46:26 Part Zagor
09:46:28 Join Zagor [0] (bjst@rockbox/developer/Zagor)
09:46:33 Quit Rob2222 (Ping timeout: 240 seconds)
10:02:38 Join lmjubuntu [0] (~lmjubuntu@
10:04:00 Join kugel [0] (~kugel@
10:04:02 Quit kugel (Changing host)
10:04:02 Join kugel [0] (~kugel@rockbox/developer/kugel)
10:05:05CIA-6New commit by bieber (r27238): Theme Editor: Removed some old hash code that caused viewer not to update due to hash collisions. Made Vf and Vb tags attached to viewport definition ...
10:06:45CIA-6r27238 build result: 0 errors, 75 warnings (bieber committed)
10:09:48*n1s starts massive test_codec run
10:13:33 Join lasser [0] (
10:19:44lasserHi there! I'm trying to compile the themeeditor from svn using Debian lenny after a successfull run of I'm using 'qmake-qt4' without errors, but with 'make' I get errors:
10:19:49lasserWhat is going wrong there?
10:24:41 Join Luca_S [0] (
10:26:30pixelmabieber: seen the warnings after your commit?
10:29:21 Join pondlife [0] (~Steve@rockbox/developer/pondlife)
10:33:48bieberpixelma: Those pop up from time to time, it shouldn't have anything to do with my commit, since none of the stuff I'm editing gets compiled by the build server (I freaked out a little last time it happened, but I've been told it's unrelated)
10:34:10bieberlasser: Compiling the theme editor shouldn't have anything to do with
10:34:59bieberWhat version of Qt are you using?
10:35:11lasserbieber: ok, it was only a possibility for me.
10:35:52 Join DerPapst [0] (~Alexander@
10:36:24lasserf$raven: qmake-qt4 −−version
10:36:26lasserQMake version 2.01a
10:36:27lasserUsing Qt version 4.4.3 in /usr/lib
10:36:56pixelmabieber: they don't *look* like a build system quirk
10:37:40bieberlasser: That's your problem. I believe some of the features I'm using came in as recently as Qt 4.5, maybe even 4.6, so you'll want to upgrade to the newest version
10:37:50bieberpixelma: Is there somewhere I can view the actual warnings in question?
10:38:34 Join Topy [0] (
10:38:35pixelmaif you click on a yellow cell in the build table, you get the log
10:38:47simonrvnlasser: check backports (
10:40:03bieberOh, the build server is building my parser
10:40:04 Quit Topy44 (Ping timeout: 276 seconds)
10:41:50bieberpixelma: The build table is showing one warning for a bunch of different builds, but the notification just said there were 75? How does that happen?
10:43:03CIA-6New commit by bieber (r27239): Theme Editor: Fixed a warning in the parser due to a stray asterisk
10:43:11lasserbieber: So I have to wait for squeeze for compiling. Are there compiled bins for download anywhere?
10:43:13lassersimonrvn: nothing in bpo, squeeze has 4.6
10:43:34kugelbieber: all warnings & errors are simply added
10:43:45bieberAh, okay
10:44:00bieberlasser: 32 bit or 64 bit?
10:44:12***Saving seen data "./dancer.seen"
10:44:33CIA-6r27239 build result: All green
10:46:29pixelmabieber: looks like one warning on 75 "targets" (devices and sims)
10:47:08 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
10:47:43lasserbieber: 32 bit
10:47:45bieberIs there a way to get GCC to ignore unused parameter warnings, by the way? The build server doesn't build it, but my theme editor has a lot of those that I can't get rid of, because the arguments are required for interfaces
10:48:31bieberlasser: I'll build it for you, just lemme fire up the VM
10:49:34kugelbieber: (void)param in the function or #__attribute__((unused)) int param' in the function declaration
10:49:47n1sbieber: it's kind of ugly but you can use some attribute to do that, or just the old (void)foo; trick
10:50:12bieberIs it generally considered good practice to suppress warnings like that, or just leave them there?
10:50:18*kugel disagrees that the attribute is ugly
10:50:45n1sbieber: we generally want 0 warnings, as they might hide new warnings that pop up
10:52:11 Quit mt (Ping timeout: 240 seconds)
10:52:46kugelbieber: perhaps you can put the attribute once in the interface declaration and all implementations will not show the warning anymore (I don't know c++/interfaces enough to be sure though)
10:53:08 Quit GeekShadow (Quit: The cake is a lie !)
10:53:25bieberIt might work in my own interfaces, but most of the abstract classes I'm implementing come from Qt, I'm afraid
10:55:35lasserbieber: woah, thank you!
10:55:47bieberlasser: Did you get it compiled?
10:56:00kugelvoid foo(__attribute__((unused)) int bar) {} or void foo(int bar) { (void)bar; } then. of course you can put the __attribute__ thing into a better looking #define
10:56:30lasserbieber: no. I mean thanks for building it!
10:56:41kugelthe compiler removes the (void)bar step when -O is given, so both are basically equivalent. in rockbox we generally use (void)bar
10:58:37bieberI don't know if it will run with the older version of Qt installed on your machine, but you can try it
11:02:42bieberOh wait, I uploaded the source package
11:03:29lasserbieber: yes, I just saw that, too.
11:04:57bieberlasser: Download the same link now, I replaced it with the right file
11:06:54simonrvnlasser: :/
11:08:11lasserbieber: f@raven:~$ ./themeeditor
11:08:12lasser./themeeditor: /usr/lib/ version `GLIBCXX_3.4.11' not found (required by ./themeeditor)
11:08:14lasserSeems I have to install squeeze in VirtualBox ;-) Thanks for your help!
11:08:37bieberGood luck
11:09:19bieberLet me know what you think of it when you get it running. Just ping me and I'll find your messages whenever I get back to my computer if I'm not around
11:12:05 Part pondlife
11:23:45 Quit Slasheri (Ping timeout: 240 seconds)
11:28:59 Join Slasheri [0] (
11:29:01 Quit Slasheri (Changing host)
11:29:01 Join Slasheri [0] (miipekk@rockbox/developer/Slasheri)
11:31:11 Quit kugel (Read error: Operation timed out)
11:34:10 Quit Rob2223 (Quit: Rob2223)
11:41:03 Join Rob2222 [0] (
11:43:21 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
11:45:01 Quit Rob2222 (Client Quit)
11:55:50S_a_i_n_tbieber: Do you remember the tag for "LANG == <LTR|RTL>"?
11:56:02*S_a_i_n_t can't seem to find it...
11:56:17S_a_i_n_t(I'm pretty sure one esists)
11:56:24S_a_i_n_t*exists even
11:56:48 Join Rob2222 [0] (
12:01:24 Join panni_ [0] (
12:03:24S_a_i_n_tDoes *anyone* remember what this tag is (%?'sometag'<lang_is_LTR|lang_is_RTL>)? I'm damn sure such a tag exists, but it's not in the manual, or the customWPS wiki page...
12:03:46S_a_i_n_t%Sr seems to sound familiar, but I'm not entirely sure about this.
12:05:37 Quit lmjubuntu (Read error: Operation timed out)
12:06:43 Quit petur (Quit: *plop*)
12:07:43*S_a_i_n_t gets annoyed by missing WPS tags in the manual...
12:08:46S_a_i_n_t"%ax : The next tag should follow the language direction (what that means is deifned by the individual tag)." Is missing from the manual
12:12:58S_a_i_n_tThe above tag/description is taken from the customWPS page (the description sucks btw), it really should say something along the lines of "%ax :This tag will mirror a viewport according to language direction" (or something to that effect).
12:13:44S_a_i_n_tAFAIK it is only support by %V, %Vl and %Cl
12:29:07n1sS_a_i_n_t: y'know, patches are welcome ;)
12:30:39S_a_i_n_tI'm not really so great with TeTex myself...I could put something up on flyspray with a more accurate description, but it'd most likely be plain text as opposed to an actual diff for the manual...
12:30:45S_a_i_n_tWould that be acceptable?
12:35:13n1syes, content is the hard part, tex is the easy part (usually, at least with simple additions like this)
12:36:42S_a_i_n_tSweet, well I'm whipping a theme into shape presently, but I shall put up a flyspray task regarding this particular tag in the very near future.
12:37:03S_a_i_n_tOne question though, what do I class it as? Is a missing tag in the manual a "bug"?
12:38:30 Join hebz0rl [0] (
12:39:32n1sS_a_i_n_t: yeah, bug is fine
12:40:15S_a_i_n_tSweet, thanks. I'll put it up soon(-ish)
12:41:30 Quit rvvs89 (Ping timeout: 276 seconds)
12:44:16***Saving seen data "./dancer.seen"
12:55:09 Quit lasser (Quit: ChatZilla 0.9.86 [Iceweasel 3.0.6/2010062510])
12:56:46 Join MethoS- [0] (~clemens@
13:04:37S_a_i_n_tHmmm...seems I need to add *two* tags to the manual, turns out I was right about the %?<LTR|RTL> language tag. It is indeed "%Sr" I had to find it by grepping the IRC logs lol ;)
13:04:56*S_a_i_n_t wonders how many other tags exist that are completely unknown...
13:13:40 Quit panni_ (Ping timeout: 265 seconds)
13:17:38 Join panni_ [0] (
13:18:58 Nick fxb__ is now known as fxb (
13:20:14 Join ucchan [0] (
13:26:09 Quit hebz0rl (Quit: Ex-Chat)
13:35:30CIA-6New commit by uchida (r27240): text viewer: fixed garbage data are not stored in the setting file.
13:37:06CIA-6r27240 build result: All green
13:39:14pixelmaS_a_i_n_t: ask JdGordon when he's around next time... ;)
13:49:58CIA-6New commit by wodz (r27241): HD200 - rework keymap
13:50:06CIA-6New commit by wodz (r27242): HD200 - improve keymap in simulator
13:50:58 Join wodz [0] (
13:51:34CIA-6r27241 build result: All green
13:53:19CIA-6r27242 build result: All green
13:55:42 Quit n1s (Quit: Lämnar)
13:56:10 Join n1s [0] (~n1s@rockbox/developer/n1s)
13:56:46 Quit pjm0616 (Ping timeout: 240 seconds)
13:57:53 Join pjm0616 [0] (~user@
13:58:19 Join stoffel [0] (
14:02:41 Quit pjm0616 (Ping timeout: 252 seconds)
14:03:09 Join pjm0616 [0] (~user@
14:07:49 Join anewuser [0] (anewuser@unaffiliated/anewuser)
14:13:44 Quit maraz (Quit: Reconnecting)
14:13:46 Join maraz [0] (
14:16:49CIA-6New commit by uchida (r27243): text viewer: for tv_action and tv_bookmark, the prototype of the initializer is the same arguments as other modules.
14:17:20S_a_i_n_tpixelma: Well, I've definitely found one tag I know of that isn't in the Manual's tag list or listed on the CustomWPS page on the wiki. The problem is, how do you know something is missing when you don't know if it exists or not? ;)
14:17:27 Part LinusN
14:18:12S_a_i_n_tThe only reason I knew of %Sr is that I was 99% confident I had used it before, and went looking for it in the Manual to make sure I was getting the syntax right.
14:18:21wodzyuppie, I managed to enable recording on MPIO :-)
14:18:25CIA-6r27243 build result: All green
14:18:42S_a_i_n_twodz: Awesome, congratulations \o/
14:19:43pixelmaS_a_i_n_t: that's why it would be nice if JdGordon would at least file a bug report (or a patch with a description at least as pure text) if he implement new tags
14:20:13pixelmawodz: congrats! :)
14:21:07S_a_i_n_tpixelma: I'm going to add a "bug report" for a propper description for %ax when I get the time later tonight, and also add %Sr now I found out that this is missing.
14:21:40S_a_i_n_tIt will just be plain text though, as I'm not great with TeTex, least it's something.
14:22:22pixelmato be fair, those were implemented by tomers
14:23:00S_a_i_n_tI'm trying to think of how to describe %ax without writing a novel...the description in the manual for it is *awfull*
14:23:22S_a_i_n_tI'll think of something when the time comes no doubt.
14:26:02 Part Zagor
14:26:41 Join Zagor [0] (bjst@rockbox/developer/Zagor)
14:31:24 Quit flydutch (Ping timeout: 260 seconds)
14:32:41 Join teru [0] (
14:36:00 Join Kitr88 [0] (
14:36:07 Quit literal (Read error: Connection reset by peer)
14:39:22 Join watto [0] (~watto@
14:39:34 Quit Kitar|st (Ping timeout: 260 seconds)
14:39:59 Quit ucchan (Quit: Leaving...)
14:40:19 Quit Kitr88 (Ping timeout: 248 seconds)
14:43:11 Quit wodz (Quit: Leaving)
14:44:17***Saving seen data "./dancer.seen"
14:44:29 Join thisisatest [0] (
14:45:38 Join Kitar|st [0] (
14:46:30 Quit pjm0616 (Ping timeout: 245 seconds)
14:48:06 Join literal [0] (
14:48:07 Quit thisisatest (Client Quit)
14:48:14 Join pjm0616 [0] (~user@
15:07:38*S_a_i_n_t is quite surprised that "Mute" and "Line Level" are not in engligh.lang.
15:08:02S_a_i_n_tWell, surprised and annoyed... :-/
15:08:02*Torne was more surprised that "OK" and "Cancel" weren't
15:08:06Torne(until recently)
15:09:57 Join mt [0] (~mtee@rockbox/developer/mt)
15:10:09 Quit kachna (Ping timeout: 265 seconds)
15:10:12S_a_i_n_t"Disc" is another one I'm surprised doesn't grace the lang files.
15:18:34 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
15:18:39pixelmathere's HD/microSD/MMC
15:20:35 Quit esperegu (Read error: Connection reset by peer)
15:21:54 Join esperegu [0] (~quassel@
15:23:02ZagorS_a_i_n_t: perhaps because we call it "disk" :)
15:23:43 Join komputes [0] (~komputes@ubuntu/member/komputes)
15:24:25S_a_i_n_tOk, well ""Disk" is another one I'm surprised doesn't grace the lang files." then ;)
15:25:42Zagorbut it does
15:25:53Zagorid: LANG_DISK_MENU
15:27:01S_a_i_n_tHmmmm...odd. I just see that now, I guess I kept missing it because I was grepping for "Disc"
15:27:09Zagorbut lang entries are full used phrases, not various useful words. so a lot of entries are not found. "song" for instance.
15:27:42Zagoror "track", which is the word used
15:28:05S_a_i_n_tThere's a patch on flyspray to add 'Disc" to the engligh lang file, I wonder if that guy didn't know that "Disk" was there also...
15:28:23S_a_i_n_tbah! "english"
15:28:34Zagorcan you find the patch?
15:28:57S_a_i_n_tI can sec.
15:29:24 Quit Rob2222 (Quit: Rob2222)
15:31:42S_a_i_n_t...err, sorry.
15:32:04pixelmaZagor: I believe "Title" is there or something like that, for the "Track info" viewer
15:32:48Zagorow, "this might be useful to someone". that is a terrible precedence.
15:33:05Zagorthere is no end to words that can be added with that reason
15:34:26Zagorbut we have a handful already
15:34:35S_a_i_n_tI'd personally like to have "Disc" in the lang files, but I understand that it shouldn;t be filled up with singular words that may, possibly* be used at some point etc. etc. I'd just like it for the exact reason that the poster of the task added it for, to me Dick == "Hard Disk", "Floppy Disk" etc. and Disc == "Compact Disc", Digital Video Disc" etc.
15:35:10S_a_i_n_toh, shit...BAD typo lol
15:35:22Zagoryes, I believe that is the generally accepted disc/disk meaning
15:36:25S_a_i_n_tIt's a pity none of the "CabbieV2" theme's use the word "Disc", then it would make sense to add it and then I could use it in my themes as well.
15:36:58Zagorwell. it is useful. it doesn't add bloat or problems. it doesn't introduce a problem that we don't already have. I say commit.
15:37:04S_a_i_n_tI'm not a fan of adding words to my themes that can't be translated (assuming the other langs have a translation for the string used also)
15:37:26S_a_i_n_t(re: Commit)
15:38:11 Join jgarvey [0] (
15:38:13gevaertswell, any new string does add some RAM use
15:38:37Zagoryes, but the small amount does not meet my definition of "bloat"
15:38:49S_a_i_n_tNot that I want to push my luck or anything...but as I said earlier I'm kinda surprised that "Mute" and "Line Level" aren;t in the lang file either.
15:39:09Zagor"line level"? when is that ever used?
15:39:42 Join funman [0] (~fun@rockbox/developer/funman)
15:40:08S_a_i_n_tI can quess that mute is probably almost universally known, but I'm almost positive that "line level" is something that would need translation for a non english speaker.
15:40:33ZagorS_a_i_n_t: yes, but I mean when do you use the phrase "line level" in a wps?
15:40:35S_a_i_n_tI like to use "line level" instead of 0dB in my themes.
15:41:14Zagor0db is much clear to me, but I guess that's a matter of taste
15:41:20Zagormuch more
15:41:35S_a_i_n_tI guess so.
15:42:22 Quit anewuser (Quit: Sixth edition of 25 samples + 48 hours + detroit party (july 31) (maybe streamed online!))
15:43:16S_a_i_n_tI actually like to have a numeric dB value and a text indicator for Mute, Line Level and Max Volume if space allows for it, however...non of the text options I use can be translated which I personally find to be a PITA
15:44:08pixelmathe problem with adding strings to be available to "some" WPS is the question where to stop. If it's used for the default - ok, the "Disk" one might justify itself if someone also implements the rest to make use of it in the "track info"
15:44:26Zagorperhaps we should add the possibility of translations to the wps format
15:44:30S_a_i_n_tJdGordon was saying at one point a few days ago that it wouldn;t be too hard to get the strings from a user created text file, and I guess that could be included in themes...which would be truly awesome.
15:45:22S_a_i_n_tI should bug him about that actually ;)
15:45:51ZagorI was rather thinking of a new tag
15:46:21Zagorbasically "for english, print this, for french, print this" etc.
15:46:43S_a_i_n_tI'm not sure exactly how he was planning on doing it, some form of user created file was the impression I got at the time but I may well be wrong.
15:47:16Zagorin any case I think that is a better idea than adding various useful strings to the core language files
15:47:27S_a_i_n_tOh, definitely.
15:47:52gevaertsThe advantage of adding them to the core language files is of course that the regular translators see them
15:48:47S_a_i_n_tyes, there is that..but as pixelma said, where to stop? And as you said, it adds to the RAM usage.
15:49:45S_a_i_n_tI like the idea of some user created file, but that could also lead to said user getting the translations horribly wrong, or deliberately getting them "wrong".
15:49:50 Quit literal (Remote host closed the connection)
15:50:47gevaertsThere might be the option of putting them in a skin-only pseudo-feature
15:51:58gevaertsThat would keep them out of the core, while allowing to export those translations to a separate file, to be included in the wps, in a way to be figured out by JdGordon :)
15:52:02 Join kugel [0] (~kugel@rockbox/developer/kugel)
15:52:25 Join rvvs89 [0] (
15:52:27 Join literal [0] (
15:52:28ZagorI think they should be tied to the specific wps. anything else just adds new exciting failure options.
15:53:00pixelmathen you won't see a lot translation I guess
15:53:33 Nick fxb is now known as fxb__ (
15:53:36*pixelma wonders how exactly that "user: " thing was meant in the langv2 format and if it can be of any use
15:56:15S_a_i_n_tI have often wondered about that myself, don't they *all* say "user: core"?
15:56:15Zagorpixelma: that is a possible downside. but any more general system has to be rather complex.
15:57:29pixelmaS_a_i_n_t: currently yes, but I remember talk about it when midgey worked on the localisation of plugins (which wasn't finished)
15:58:14S_a_i_n_tWhat was the intention behind it?
15:58:42 Join flydutch [0] (
16:11:01 Quit literal (Remote host closed the connection)
16:12:36 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
16:12:47 Join literal [0] (
16:13:14 Quit kugel (Ping timeout: 264 seconds)
16:14:07 Part Zagor
16:14:26 Quit flydutch (Ping timeout: 260 seconds)
16:14:31 Join wombat23 [0] (
16:18:10n1spixelma: iirc the user field was intended to specify the user of a string, like core or a plugin so the plugin only strings could go in plugin lang files and would not need to be loaded in the core
16:19:12 Quit wombat23 (Ping timeout: 240 seconds)
16:19:32Luca_Sthe patch in FS11297 is out of sync with svn :/
16:20:17Tornepatches in a lot of flyspray tasks are out of sync with svn
16:20:39Tornethis is why you should work on getting patches you are interested in committed :)
16:22:16 Join kugel [0] (
16:22:40pixelman1s: could this be used for WPS only lang files? (maybe one for all these strings)? Not sure how loading could be handled but afaik the WPS is only loaded once at boot (or after changing it) and then kept in RAM so I guess it could work (shot in the dark though as I talk about things I'm not sure of)
16:23:03Luca_STorne: i reported a few times that it worked fine, both on tracker and irc...
16:23:20Luca_Showever it does not work well on all AMS targets
16:23:39Luca_Sso it's there while being sorted out - but interest seem to have dropped
16:23:44S_a_i_n_tprobably why it hasn;t been commited then, no? ;)
16:24:07Torneif it breaks some targets, then no, it won't get committed..
16:24:13funmanprobably needs more testing on clipv2/clip+
16:24:30S_a_i_n_tLuca_S: It's only a very small patch, it would be fairly easy for you to patch it by hand.
16:24:32Luca_Ssure - i'm not complaining about committing, just saying that it can't get tested if can't be applied
16:25:05TorneLuca_S: so, resync it and upload it then
16:25:08funmanto make the wav encoder plugin able to use all encoders we need to pull recording code on all models or enable it only for models with recording
16:25:17Luca_SS_a_i_n_t: i resolved the wrong hunk, but it does not compile after that (missing method?)
16:25:48S_a_i_n_tpatch the hunks that fail by hand to a clean svn tree, svn diff > mypatchname.patch, boom...done, synced.
16:26:04TorneLuca_S: then investigate, and work out why, and fix it? :)
16:27:40 Join bluebrother [0] (
16:27:41 Quit bluebrother (Changing host)
16:27:41 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
16:29:12 Quit bluebroth3r (Read error: Operation timed out)
16:31:46 Quit liar (Ping timeout: 240 seconds)
16:34:07 Join teru0 [0] (
16:34:09 Quit teru0 (Client Quit)
16:34:10 Quit balintx (Remote host closed the connection)
16:34:42 Join liar [0] (
16:34:51 Quit kugel (Ping timeout: 245 seconds)
16:37:15 Quit teru (Ping timeout: 252 seconds)
16:44:19***Saving seen data "./dancer.seen"
16:45:18 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
16:46:10 Join lasser [0] (
16:46:38lasserbieber: ping
16:47:09 Join balintx [0] (
16:47:49 Join blu3t0oth [0] (d957399a@gateway/web/freenode/ip.
16:48:10blu3t0othI have a question
16:48:12blu3t0othJust wait a sec
16:48:14 Quit balintx (Read error: Connection reset by peer)
16:48:21S_a_i_n_tsweet, ask it.
16:48:41blu3t0othDoes Rockbox works well on this player?
16:48:56blu3t0othI want to buy a player with rockbox compatability
16:49:04lasserbieber: I successfully compiled themeeditor. And what shall I say. Nice :-) Since months my slefmade wps was broken, and now I could repair it in 10 minutes because of the error hints of your themeeditor. Thanks for the great work!
16:49:33lasserups... selfmade
16:49:33Luca_Suhm... FS11297 - the clock divider table was #defined for AS3525_PCLK_FREQ == 40000000, but in clock-target.h it gets defined to 41333334, that's why it des not compile...
16:49:40S_a_i_n_tblu3t0oth: Supported players on on the rockbox main page, if the player isn;t listed there rockbox will not run on that player.
16:50:14blu3t0othI found mine
16:50:26blu3t0othI don't know if the 2 gb version is working
16:50:43S_a_i_n_tWhat player is it? I can't be bothered looking at the link sorry...
16:51:10blu3t0othSandisk Sansa e250 2 GB
16:51:28S_a_i_n_tYes, rockbox runs on this device.
16:51:30funmanLuca_S: try the penultimate patch
16:51:41blu3t0othS_a_i_n_t: Thank you
16:51:51S_a_i_n_tIt doesn;t matter what size the internal storage is.
16:51:53Torneblu3t0oth: does the front page not make it clear?
16:51:57 Join balintx [0] (
16:52:55Torneit says e200 series (all models)
16:53:12blu3t0othAh ok, thanks
16:53:12Tornecan you suggest a way to make it clearer?
16:53:25TorneYou said you'd looked at that already
16:54:05blu3t0othYea, i was thinking that e250 is not e200
16:56:04 Quit funman (Quit: leaving)
16:56:06*Torne looks confused.. then why did you think it worked on the e250 at all?
16:56:15Tornewhat i mean is, what gave you the impression that the size would matter
16:56:43 Join lmjubuntu [0] (~lmjubuntu@
17:00:47 Quit blu3t0oth ()
17:01:32S_a_i_n_tYou scared him away Torne :P
17:08:54 Join |liar| [0] (~liar@
17:11:22 Quit liar (Ping timeout: 240 seconds)
17:19:51 Quit panni_ (Read error: Connection reset by peer)
17:23:52 Join flydutch [0] (
17:27:09Luca_Sfunman: the penultimate patch applies cleanly, compiles and runs :)
17:35:30 Join chaos [0] (~chaos@gentoo/user/ch4os)
17:38:08 Part chaos
17:45:58n1shmmm, this mdct compiler option test was rather interesting. O2 is faster than Os (which is used in svn) for all the codecs but O3 is significantly slower than O2 for all codecs *except* wma where it's actually quite a bit faster than O2...
17:47:27n1sO3 is 8% faster than O2 for the 128kbps wma test file
17:48:05n1sbut i suppose we don't want to build the lib with different options for different codecs, do we?
17:49:07n1spixelma: i suppose that could work
17:49:41n1swe would need different lang files for different wps's then or there wouldn't be much point
17:50:00n1swell, or cleverer loading
17:52:12 Quit bmbl (Read error: Connection reset by peer)
17:52:22Luca_Suhm... something strange happened! fuzev2, frequency switching patch, audio on pllb: panic stack underflow.. viewportmanagr
17:53:00 Quit |liar| (Ping timeout: 240 seconds)
17:53:12Luca_Sand before panicking, the fm radio started jumping across stations... from 92.6 to 92.3 then 108 and stopped tere
17:53:44Luca_Swhen i tried to change it, it went to 108 and a few seconds after, had the panic
17:53:49 Join panni_ [0] (
17:54:47Luca_Sstrange that it worked for about 45mins (the second half of brazil-nederland :D)
17:57:23 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
17:59:40saratogan1s: thats just changing the compiler flags for mdct.c?
18:00:26 Quit Luca_S (Quit: CGI:IRC (Ping timeout))
18:00:31 Quit panni_ (Read error: Connection reset by peer)
18:09:00n1ssaratoga: well, the whole codec lib but the other stuff shouldn't make this kind of difference
18:09:23 Join |liar| [0] (
18:09:26saratogan1s: this is for ARM?
18:09:32n1sno, coldfire
18:10:05n1sis wma using different sizes than the other codecs or something?
18:11:08n1sdoing the same tests for arm might be interesting though
18:12:04 Join swilde [0] (
18:12:07 Join panni_ [0] (
18:13:16saratogan1s: yes WMA/WMAPRO is weird
18:13:28saratogait uses 5 (or 6 for Pro) window sizes all at once
18:13:38saratogaall the other codecs use either 1 or 2 window sizes
18:13:59saratogathe difference for CF makes some sense, theres hardly any ASM for it
18:14:07saratogai would be surprised about aram
18:15:15n1si think i'll stick with O2 for cf and leave hacking up a solution for wma for someone else, it's still better than the Os used in svn
18:20:21 Quit swilde (Quit: ERC Version 5.3 (IRC client for Emacs))
18:30:16 Quit stoffel (Ping timeout: 265 seconds)
18:34:33 Join DerPapst1 [0] (
18:34:38 Join Jerom [0] (~jerome@
18:36:31 Quit DerPapst (Ping timeout: 245 seconds)
18:36:35 Join kachna [0] (
18:38:19 Quit jgarvey (Quit: Leaving)
18:42:43 Quit DerPapst1 (Ping timeout: 260 seconds)
18:43:39 Quit Jerom (Quit: Leaving.)
18:44:05 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
18:44:23***Saving seen data "./dancer.seen"
18:44:43 Join Jerom [0] (~jerome@
18:44:46 Quit simonrvn (Disconnected by services)
18:45:22 Join simonrvn [0] (
18:47:08 Quit storm` ()
18:52:37bertrikI can very reliably reproduce the extra button press after exiting a plugin problem now on my e200
18:53:09bertrikWhen starting the dict plugin and exiting with a brief press of the power button, the problem appears
18:53:48S_a_i_n_tAre you talking about exiting a Plugin eating the first button press?
18:55:44bertrikIt does not happen when I hold the power button a little longer, releasing it only when we have fully returned to the plugin menu (and the "Cancelled" popup has disappeared)
18:58:29bertrikIt happens both on my sansa e200 and on my sansa clip+, so it does not seem to have anything to do with having a scrollwheel or not
18:59:03S_a_i_n_tIt has been annoying me on my iPod Nano1/2Gs for as long as I remember...
19:00:35S_a_i_n_tI believe it may be linked to the scroll-wheel sometimes nut working until after a button-press also.
19:00:43S_a_i_n_t(again, on Nano1/2G)
19:02:36pixelmabertrik: is it specific to some plugins or does it happen for all (thinking about those using pluginlib actions and the others)?
19:06:35bertrikpixelma, I don't know yet, but it's very reproducible with the dict plugin, also with pictureflow if you press power quickly after entering it
19:07:12 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
19:07:21 Quit lmjubuntu (Ping timeout: 265 seconds)
19:11:18*amiconn grumbles
19:11:40amiconnThe old text viewer did chunked reading when the file didn't fit into the available buffer
19:11:42bertrikI think with dict, we don't really do much in the plugin, just waiting in rb->kbd_input
19:12:56amiconnThe new one apparently stops playback to use the audio buffer, which is (a) bad if you want to read something large-ish while listening to music, and (b) limiting the file size, especially on lowmem targets
19:13:41bertrikSounds like a bad idea indeed
19:14:24pixelmamaybe take it to the dev ml?
19:16:35 Join Strife89 [0] (
19:18:38 Join CGL [0] (~CGL@
19:24:22S_a_i_n_tThat does seem like a horid idea indeed, the few times I do actually use text_viewer I'm usually listening to music also. :/
19:27:59 Join DerPapst [0] (
19:32:51 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
19:33:02 Nick fxb__ is now known as fxb (
19:33:17bieberlasser: Thanks, glad to hear it worked well for you. Give it another week or two, and it should get really interesting :)
19:33:49 Join Rob2222 [0] (
19:34:29lasserbieber: Ok. What are you working on? Or is it a secret?
19:34:43bieberNo secrets, I'm working on the theme rendering
19:35:50 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
19:36:15bieberBy mid next week, I should have most everything implemented, and I'll have the device control panel fixed up so that there aren't any redundant settings. Then I'm going to add a timer panel to automatically advance time so you can watch things move
19:37:21bieberThen after that I'll be working on some drag and drop editing for WPS files
19:40:12 Quit flydutch (Quit: /* empty */)
19:40:48 Quit pamaury (Remote host closed the connection)
19:42:43 Join funman [0] (~fun@rockbox/developer/funman)
19:43:47funmani need some help for apps/ code : i would like to confirm if encoder selection is handled as global settings
19:44:43funmanI would like to reuse enc_config.c code in a .wav encoder plugin, but I am not sure if plugins are allowed to change global (recording) settings
19:46:04simonrvnrec format: wvpk ? on mine - it's in .rockbox/config.cfg
19:46:13funmanamiconn: the text viewer isn't stealing audio buffer.
19:46:53simonrvn(i inserted that '?' fyi)
19:48:32funmansimonrvn: well yes we can record in wavpack, mp3, wav or aiff
19:48:40lasserbieber: What about inserting code snippets? Like "insert %V tag" with a popup for all the options? Or will that be the work for next gsoc? ;-)
19:49:49 Nick funman is now known as LambadaCalculus3 (~fun@rockbox/developer/funman)
19:49:51simonrvnfunman: yeh, that setting as you know was set from the settings sub-menus. but you said plugins. thought it might be useful, sorry.
19:52:21LambadaCalculus3simonrvn: well that answers my first question so thank you
19:53:05 Join Strife1989 [0] (
19:54:05LambadaCalculus3the options in recording menu are just a shortcut to the settings
19:54:08 Nick LambadaCalculus3 is now known as funman (~fun@rockbox/developer/funman)
19:55:17 Quit Strife89 (Ping timeout: 240 seconds)
19:58:19bieberlasser: That will be part of the graphical editing
19:58:41bieberFor viewports, you'll be able to click and drag an area on the preview, and then set its options
19:59:15simonrvnmm i see
19:59:19lasserbieber: cool!
20:00:43funmanJdGordon: ping? check my question at 19:44 (logs time)
20:01:31 Quit bmbl (Quit: Bye!)
20:02:54n1ssaratoga: i misread the test results pretty badly, that wma discrepancy isn't there...
20:03:12n1sah! it is, it's my table that weird
20:05:56 Join ryran [0] (
20:08:49ryranQQ: hopped onto my build machine to get a new build for my ipod for the first time in ... a month and a half maybe.. and upon doing a make reconf was greeted with an error about arm-elf-eabi-gcc being missing. Is this something new or changed in the build process for arm?
20:09:09n1snew toolchain
20:09:35Torneyou need to build a new one. delete the old toolchain directory, run again and ask for an eabi toolchain when it prompts
20:09:52ryranah that's it? cool!
20:10:09Tornewell, it will take a while :)
20:10:40AlexPpixelma: I had a look at the onda 777 manual patch - the English is fine (if a little informal), but I think the instructions need a little more detail. e.g. "item Compile the bootloader (this will generate a \fname{ccpmp.bin} file)."
20:11:02ryranto be clear, since this is NOT what I do everyday, the toolchain dir would be the dir in /usr/local right? (As in, /usr/local/arm-elf)
20:11:07AlexPHowever, I guess that is temporary as when appropriate (not yet?) we will supply that
20:11:25n1sryran: yes
20:11:35ryrann1s: THANKS
20:11:38saratogaany reason FS #7831 - FLAC files with "unknown" total samplecount are rejected hasn't been committed?
20:13:50funmansaratoga: i assume linuxstb opened the bug and then didn't see the patch ?
20:22:45ryran@ ns1 or Torne or anyonen else: do we still need the old arm toolchain to build for ipods or only arm-eabi?
20:24:36funmanonly for ipod4g
20:24:40funmanand for testing
20:24:59ryranI seeeeee. thanks funman.
20:25:41 Quit CGL (Ping timeout: 245 seconds)
20:27:08amiconnfunman: Since r27213 it seems to
20:27:36amiconnErr, that's the editor
20:27:57*amiconn slightly confused
20:29:09 Join webguest04 [0] (
20:29:33 Quit evilnick_B (Quit: Page closed)
20:29:36 Quit webguest04 (Client Quit)
20:31:33 Join fdinel [0] (
20:32:33 Join fml [0] (
20:33:25fmlAlexP: Hello. What do you think about ? It makes the first column in the tags tables narrower.
20:35:46 Quit TheSeven (Ping timeout: 260 seconds)
20:36:06 Quit n1s (Quit: Lämnar)
20:39:14CIA-6New commit by saratoga (r27244): Commit FS #7831. Corrects our flac decoder to not reject files that have samplecount set to 0.
20:40:08 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
20:40:58CIA-6r27244 build result: All green
20:41:02 Quit ryran (Quit: CGI:IRC (EOF))
20:41:17 Join ryran [0] (
20:41:53CIA-6New commit by saratoga (r27245): Add Joshua Hulst to credits for FS #7831.
20:43:32CIA-6r27245 build result: All green
20:43:51ryranfunman: I just noticed you left a comment 7days ago on a task I opened a while back. ( you told me to ping Blue_Dude ... who's that by the way?
20:44:26***Saving seen data "./dancer.seen"
20:45:13 Join CGL [0] (~CGL@
20:46:56saratogaanyone know why bigbambi closed FS #7565 - support for the TPA tag?
20:47:09AlexPsaratoga: You could ask me :)
20:47:20saratogasomehow i didn't see your name
20:47:44saratogaryran: hes the person who wrote the hotkey code
20:47:59fmlAlexP: have you seen my message above?
20:48:28AlexPfml: Yep, and sounds fine - I'll just have a quick look how it looks
20:48:49AlexPsaratoga: I can't actually remember - it was part of the devcon clean up
20:49:02fmlAlexP: ok, I'll wait
20:49:18saratogaAlexP: this is why its good to put a message when you reject something :)
20:49:24AlexPIndeed yes
20:50:25saratogagoolging around suggests that TPA is ancient, so thats probably why it was rejected
20:50:46AlexPI suspect it was as it is an obsolete tag and nobody had taken an interest so it was just rotting
20:51:41AlexPBut I don't suspect there would be any real opposition if you wanted to add it
20:52:49AlexPfml: Looks good to me
20:53:21fmlAlexP: commit?
20:53:28AlexPsaratoga: Incidently, you didn't see my name there as it isn't as such - bigbambi is an oooooold nick :)
20:53:32AlexPfml: Sure
20:54:10AlexPfml: I think it better to have the odd tag wrapped on mutiple lines than loads of the descriptions
20:54:11CIA-6New commit by alle (r27246): Make the first column in the tag tables narrower
20:55:06fmlAlexP: what do you mean?
20:55:45CIA-6r27246 build result: All green
20:56:41AlexPfml: In general the description is longer than the tag, so with a wider tag column fewer tags took multiple lines, but loads of descriptions did. Now with a narrower tag column, a couple of tags wrap but the descriptions look better, i.e. a narrower tag column is better
20:57:58CIA-6New commit by alle (r27247): No need to capitalize every word
20:58:58 Quit ryran (Quit: CGI:IRC (EOF))
20:59:27CIA-6r27247 build result: All green
20:59:29fmlAlexP: ah, yes, that's what I think too
21:00:49fmlAlexP: and what about the fact that each parameter (in e.g. album art tag) is explained in its own table row? It'd do it differently: only one row, but with line breaks in the cell.
21:02:47AlexPfml: Yes, I think I agree
21:05:41 Quit CGL (Quit:
21:10:32 Quit t0rc (Quit: Leaving)
21:12:41 Quit Jerom (Quit: Leaving.)
21:13:33 Join Jerom [0] (~heidi@
21:17:39 Join CGL [0] (~CGL@
21:27:16 Join dfkt [0] (dfkt@unaffiliated/dfkt)
21:27:45 Part watto
21:29:04fmlAlexP: why do we use \textless for < etc? It's only necessary in math mode (?) which we don't use.
21:32:50CIA-6New commit by alle (r27248): Onle one table row per WPS tag
21:34:21CIA-6r27248 build result: All green
21:38:06bertrikDoes anyone have any idea why the mod codec metadata handler put the title of the mod in the artist field?
21:38:32bertrikThere's even a comment explicitly saying so, so it must be intentional
21:38:58bertrik(See also
21:38:59 Quit |liar| (Ping timeout: 240 seconds)
21:39:04 Quit MethoS- (Remote host closed the connection)
21:42:54pixelma_if I remember correctly_ mod doesn't have metadata in its usual sense, just something like an attached text where lines can be interpreted as something. I believe there is/was some common style about the order to use but also lots of lots mod authors put something else there. (citation needed)
21:43:19AlexPfml: No idea
21:44:26bertrikpixelma, I looked into the mod format, there is only a song title and basically nothing else (well there are instrument names). Besides that there's only the filename which is often the same as the song title.
21:45:08pixelmayeah, some authors used the instrument names for more info
21:45:44pixelmaor "misused"
21:46:14bertrikOne reason that I can think of that it is done the way it is now, is that it at least looks semi-decent in CabbbieV2, no "(root)" displayed
21:48:10bertrikI looked at a few .mod files and couldn't see an obvious pattern yet in the names of the instruments, so I think we can't really rely on that
21:48:44 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
21:49:19balintxvcodec=mpeg2video:vbitrate=<bitrate> -vf scale=<width>:<height>,harddup -ofps 25 -o <filename>.mpg
21:49:21balintxwhat to bitrate?
21:50:45 Quit lasser (Quit: ChatZilla 0.9.86 [Iceweasel 3.0.6/2010062510])
21:51:11 Join |liar| [0] (
21:51:18bertrikIs there some overview of default metadata values in case they're not set by the metadata parser?
21:55:49CIA-6New commit by alle (r27249): No need to use textless, textgreater and textbar where <, > and | can be used
21:57:37CIA-6r27249 build result: All green
21:57:44 Quit funman (Read error: No route to host)
21:59:26 Quit leavittx_ (Read error: Connection reset by peer)
21:59:26 Quit leavittx (Read error: Connection reset by peer)
22:01:26 Join leavittx [0] (~leavittx@
22:05:47 Quit |liar| (Ping timeout: 240 seconds)
22:12:06 Join |liar| [0] (
22:17:55 Join leavittx_ [0] (~leavittx@
22:25:20 Quit Galois (Read error: Operation timed out)
22:28:24 Join mischasworld [0] (
22:36:39 Join Galois [0] (
22:38:26 Join binaryhermit [0] (
22:41:10pixelmaAlexP: oh, thanks. So do you think it can be committed this way for now?
22:41:55AlexPpixelma: Yes, it is a good start and it is best to get it in IMO
22:42:11 Quit bertrik (Remote host closed the connection)
22:42:45pixelmaunfortunatelly giovanni didn't answer since although he asked me quite a bit about it last week
22:42:56 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
22:44:28***Saving seen data "./dancer.seen"
22:55:59 Quit mischasworld (Ping timeout: 265 seconds)
23:06:52 Join Mobius-1 [0] (
23:08:09 Quit Mobius-1 (Client Quit)
23:08:49 Join wodz [0] (
23:09:30CIA-6New commit by wodz (r27250): HD200 - Setup codec as I2S master and enable recording
23:11:09CIA-6r27250 build result: All green
23:14:54 Quit wodz (Quit: Leaving)
23:17:41 Quit fml (Ping timeout: 252 seconds)
23:21:18 Nick fxb is now known as fxb__ (
23:21:30 Join fml [0] (
23:34:09 Join wombat23 [0] (
23:39:06 Nick Strife1989 is now known as Strife89 (
23:39:51 Join GeekShad0w [0] (
23:41:06 Quit esperegu (Remote host closed the connection)
23:42:07dfktwhat's the key combo to charge the clip+ with rockbox instead of OF?
23:42:12 Quit GeekShadow (Ping timeout: 240 seconds)
23:43:46saratogadfkt: hold select I think
23:45:04dfktthanks saratoga, that's it - seems it only works when rockbox is running, not when the player is off
23:47:11saratogadfkt: correct, the prebootloader will always boot the OF if theres USB connected
23:47:21saratogaits a safety thing meant to make it harder to brick a plaer
23:47:23 Quit wombat23 (Read error: Operation timed out)
23:49:02dfktok, i was just trying every possible key combo to get it charging with rockbox (simply to get the screen to shut off, which the OF doesn't do for half an hour or so) :)
23:49:54saratogawe'll probably change that once we get rockbox USB working and fully debugged
23:50:53dfkti hope one day rockbox will take over 100%, that would be the least hassle
23:56:29 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
23:58:14 Join storm` [0] (
23:58:45 Quit GeekShad0w (Ping timeout: 248 seconds)

Previous day | Next day