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

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

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

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

#rockbox log for 2010-02-09

00:00:26flybackhmm nothing about power on so far
00:00:31flybackmabye my unit is just canucked
00:01:02gevaertsamiconn: I updated the wiki
00:01:25Unhelpfulpamaury: seems worthwhile to me in the long run, but yeah, it'll be work, and i know nothing about designing such things :/
00:01:45flybackyou guys ever probe your devices for serial consoles
00:01:51flybackmight help debugging easier
00:02:15flybackif some of the soc's have a ttl level uart and the mp3 player exposed the pads on the pcb for factory testing
00:02:36flybackall you need is a ttl level converter to rs232 or ttl to usb console cable such as ftdi or pl2303
00:02:38Unhelpfulflyback: we have debug output over usb serial ;)
00:02:44***Saving seen data "./dancer.seen"
00:03:10flybackseems really well done
00:03:20flybackrockbox overall :)
00:03:34pamauryUnhelpful: it's true but it needs to be discussed before starting it :)
00:03:37flybackI am glad I bought this used unit and tracked down a new hd for it and took it apart, cleaned all the parts and reassembled it
00:03:41flybackit was worth the effort
00:04:01Unhelpfuli seem to recall talking to somebody last week or so about the possibility of running a debugger over a custom usb protocol, i think it was TheSeven?
00:04:46stripwaxUnhelpful - sounds like something he's looking at doing for the nano2g
00:05:13Unhelpfulstripwax: it would be doable for many other soft-usb targets too, though...
00:05:14stripwaxActually doesn't the archos port have a serial debug console (or am I on crack)
00:05:35Unhelpfuli'd love to be able to step and trace my e200 or beast in gdb :P
00:05:41stripwaxUnhelpful - yeah - although probably not regular builds, more dev/debug builds..
00:05:44domonokythe old archos ons have serial-gdb stub yes.
00:06:01stripwaxI meant the gdb stub not console obviously!
00:06:28amiconngevaerts: thanks
00:06:52amiconnThe fused asm is a little bit more difficult than I thought.
00:06:54TheSevenUnhelpful: Yes, I have a trivial implementation of such a thing running on nano2g
00:07:38TheSeventhe stub could in theory already do fancy stuff like singlestepping/breakpoints, it's just that one would need to write a pc-side client that sends it the right commands :-)
00:08:48amiconnThe swar macros for arm <=v5 are somewhat nasty. Needing extra registers and such
00:09:20TheSevenactually i wrote that *before* porting rockbox, in order to be able to run testing code and debug things in a nice fashion
00:09:25 Quit domonoky (Read error: Connection reset by peer)
00:09:51 Join PaulJam_ [0] (
00:09:59 Quit GeekShadow (Quit: The cake is a lie !)
00:10:32TheSevenand another nice thing it supports is the debug console, where the application running inside can just send debug messages via SWIs, that will end up on a console on your PC, and that's fast!
00:10:52TheSeveni would guess that the console manages to do a throughput of about 100KB/s
00:11:03TheSevenyou won't get that using a serial port :-P
00:13:22 Join saratoga [0] (~9803c6dd@rockbox/developer/saratoga)
00:14:41 Join Strife89 [0] (
00:14:48saratogastripwax: did you see the problem with coldfire?
00:15:21 Quit pamaury (Quit: abort();)
00:17:37saratogastripwax: you don't have the inlined complex mul defined on CF
00:17:40saratogajust the original one
00:18:05saratogai wasn't sure why theres two versions of it
00:18:44stripwaxerm, not sure what you mean (inlined complex mul vs 'the original one'). you mean XPROD31_R ?
00:19:05saratogaits not defined for coldfire
00:19:35saratogai used the c version when benchmarking and got that the new mdct is 1.5mhz slower, so with the asm version its probably pretty comparable
00:22:12saratogai suppose we should try and get some benchmarks for other targets as well to see if the new lib is faster for non-PP too
00:22:25stripwaxfair enough. it's exactly the same as XPROD31 it just stores the result in a local variable rather than in a memory location
00:22:37 Join S_a_i_n_t [0] (S_a_i_n_t@
00:22:54stripwaxsaratoga - btw when were you benchmarking (did you see the new changes I just put in?)
00:22:56saratogai tried changing it to work on CF, however I wasn't able to get gcc to take it
00:23:05saratogastripwax: yesterday
00:23:09 Quit Adub- (Read error: Connection reset by peer)
00:23:18 Join Adub- [0] (
00:23:31saratogai got someone on IRC to do it since i don't have CF
00:24:02stripwaxi.e. XPROD31_R a bit like XPROD32 really. (Probably XPROD31 and XNPROD31 should also be rewritten as macros like XPROD32 is, and then there'd only be one version rather than my crazy/weird _R stuff)
00:24:21 Quit PaulJam_ (Disconnected by services)
00:24:26 Join PaulJam__ [0] (
00:24:51saratogadoes the new transform still need more IRAM then the old?
00:25:28stripwaxWell, it depends what you mean by 'need'. It doesn't currently *use* more IRAM than the old.
00:25:52stripwaxi.e. the remaining two tables (revtab in fft and then final fsin/fcos in mdct) are not in iram
00:26:37saratogaah ok
00:26:44saratogaimpressive that we get such good performance then
00:26:45*TheSeven just spotted a bug!
00:26:53 Quit Adub- (Read error: Connection reset by peer)
00:26:56TheSeven(in apple's norboot!)
00:27:04gevaertsTheSeven: report it!
00:27:04stripwaxThe mdct final (post)twiddle is the thing that could be unified with the FFT trig if you make a few approximations (take a look at the last step of the tremor mdct in mdct2.c i.e. interpolating between trig values, for large block sizes, rather than having trig that covers everything)
00:27:17stripwaxis it exploitable? :-p
00:27:44TheSeventheir LCD code might fail if some certain kind of garbage was on the stack before
00:27:53saratogastripwax: which block sizes would need interpolation?
00:27:54 Join komputes [0] (~komputes@ubuntu/member/komputes)
00:27:56TheSeventhey're using an uninitialized local variable
00:27:56 Quit togetic (Quit: WeeChat 0.3.0)
00:28:37 Join togetic [0] (~togetic@unaffiliated/ibuffy)
00:28:44stripwaxsaratoga - I make it that having the posttwiddle factors not in iram costs somewhere between 0.5 and 1MHz (but that was kinda rough+ready measurements)
00:28:57stripwaxsaratoga - 4096 and 8192
00:29:02stripwaxhence not really a problem
00:29:21stripwaxI.e. Tremor is (unsurprisingly) well optimised for 2048 and smaller blocksizes
00:29:30 Quit ender` (Quit: PHP is a minor evil perpetrated and created by incompetent amateurs, whereas Perl is a great and insidious evil, perpetrated by skilled but perverted professionals.)
00:30:07stripwaxBut a bit weird that Tremor would use interpolated trig, rather than more accurate trig, for the largest blocksizes. but hey.
00:30:23saratogaAFAIK no codec except tremor uses 4096 blocks
00:30:28 Quit parafin|away (Read error: Operation timed out)
00:30:41saratogaand while the encoder can do as it likes, I would be very suprised if they are used frequently at 44.1kHz
00:31:02 Join parafin [0] (
00:31:03saratogamost codecs reserve them for higher sampling rates, as the temporal resolution is unacceptably low at 44.1khz with a 4096 block
00:31:51stripwaxyeah, probably vorbis too. anyway, like I said, if we did that it would be no worse than Tremor...
00:31:59saratogaFWIW, when we accidentally merged the rotation and fft tables (by pretending they were the same!) the result was accurate to almost 12 bits
00:32:12saratogai'm sure linear interpolation here gives >> 16 bits resolution
00:32:57stripwaxAnother (suspected - not measured) cause of slowness is that the revtab is not in iram, and I would really love to resolve that by changing the butterflies so that the input permutation is *regular* bitrev and not a crazy splitradix base2/base4 digit rev mess
00:33:04stripwaxsaratoga - I'm sure you're right
00:33:10stripwaxIt is after all just that very last step.
00:33:15saratogais the bitrev table that large?
00:33:41stripwaxit's sized for 4096 FFT, but sneakily stored in uint16_t so only 2K
00:33:58stripwaxPutting in iram might help
00:34:04saratogauint16_t revtab[1<<12];
00:34:07stripwaxI'd rather get rid of it totally though, to be honest
00:34:19saratogaso 8KB
00:34:48stripwaxErrr, yeah, that's what I meant. Honest.
00:35:09stripwax8KB would be a big old waste of iram
00:35:58stripwaxIf the radix-4 part of the fft could be rewritten/transposed to use regular bitrev, then that would also enable (in future) doing a real half imdct in-place (something I'm 99% certain cannot be done right now with that horrible split radix indexing scheme)
00:36:39stripwaxwith regular bitrev, you can just swap element X and Bitrev[X].
00:37:02 Join wookey_ [0] (
00:37:21saratogamost codecs do 98% 512 point FFTs, so you could probably get away with just a 1KB bitrev table in IRAM and the other sizes in DRAM
00:37:29wookey_hi all, just upgraded to v3.5 on a snasa e200 and all is lovely
00:37:59saratogaWMA is the exception, and even then its still ike 50% 512 point, so you'd get most of the speed up there
00:38:06stripwaxsaratoga - I'd rather have no table at all, than a 1KB table in IRAM and an 8KB table in DRAM :)
00:38:10saratoga(assuming its too hard to fixe the problem)
00:38:18gevaertswookey_: great! Does that mean the database duplication issue is gone?
00:38:20wookey_I previously installed in march 2008. Has the e200 bootloader changed since then in significant ways?
00:38:37wookey_gevaerts: don't know - haven't done the wife's machine yet - which weas the poorly one
00:38:41gevaertsah, ok
00:38:45saratogayes, it no longer uses the OF for USB
00:38:51stripwaxGetting rid of the mdct twiddles and merging with the regular fft trig also means all that fsincos initialisation disappears (which is not exactly lightweight)
00:38:56wookey_But that seems to be working alkready
00:38:57saratogaupgrade if that matters to you (it will still work fine if you don't)
00:39:14gevaertsThe only major change in the bootloader since then is that the current bootloader now always boots rockbox
00:39:29wookey_OK. I'll update
00:39:30gevaertsThe old one booted the OF if you booted by pluging in
00:40:16wookey_OK. right. I see. yes it still does that.
00:40:24saratogastripwax: have you ever looked at the code gcc generates for the prerotation?
00:41:05wookey_gevaerts: I may not have looked in the right place but I couldn;t eaily find a list of what had changed since when in the bootloader
00:41:20wookey_so eventually I just came and asked :-)
00:41:32gevaertshm, yes. We don't really have a decent changelog for the bootloaders
00:41:50wookey_BTW when I run btutil it complains about network config. what am I doing wrong?
00:42:37stripwaxsaratoga - yeah. not the worst thing I've ever seen (only one local variable on stack). But that's probably one more than ideal.
00:42:48wookey_'Network error:request aborted. Please check you r network and proxy settings'
00:42:50saratogastripwax: i don't suppose you have it handy to pastebin?
00:43:04gevaertsIIRC I've seen that a while ago if you used a 32 bit build on a 64 bit setup
00:43:19wookey_hmm. I carefully downloaded 640bit build
00:43:45*gevaerts looks at bluebrother
00:43:54wookey_And version says 1.2.2, whereas other archs are 1.2.4
00:44:53wookey_file says: ELF 64-bit LSB executable, x86-64
00:44:53gevaertsok, it happens here as well
00:47:11 Quit stripwax (Ping timeout: 260 seconds)
00:47:46gevaertsthe old version is because nobody has built the latest version yet, which is because we want a static build, which is a pain with qt
00:48:18wookey_why the static build? I'ch much prefer to be able to apt-get install rbutil
00:48:24wookey_with normal dynamci libs
00:48:38wookey_then it could download and install 'current'. Is that difficult?
00:49:42wookey_and shouldn't the x86 and amd64 build processes be identical?
00:50:01 Quit orzo (Quit: Leaving.)
00:50:15gevaertsthey are, but apparently the people who usually have their build setup ready run 32 bit
00:50:28TheSevendamn, found the first rev. 7 / rev. 9 difference
00:50:32TheSevenit's an LCD init value
00:50:44gevaertsa normal dynamic build is easy. I'm not sure how compatible those are between distributions these days though
00:51:18wookey_gevaerts: probably not very. I can see that attraction of a static build for the website-download 'just works' situation
00:51:24gevaertsAnd having the distributions package it isn't really ideal yet, there are still changes pretty regularly that need a fast turnaround
00:51:42wookey_even in the installer?
00:51:46gevaertsMaybe we should just try dynamic builds and see what happens
00:52:39gevaertsyes. 3.5 need rbutil 1.2.4 due to some changes in how build information is retrieved from the server.
00:53:20wookey_OK. Well I guess I'll use old-fashion sansapatcher then, or try a local build from svn
00:53:29 Join stripwax [0] (
00:53:51gevaertsWe've basically had that sort of thing for all recent releases, although I think the issues get smaller every time, so we might get there in the end :)
00:54:10wookey_A distro-packaged version may not always work, but it will probably work for a fair while before you break it again?
00:55:00wookey_I suppose the 'stable' version could be almost permanently 'too old'
00:55:03 Quit petur (Quit: Zzzzzz)
00:55:04gevaertsWe tend to break it just before releases (because that's when we notice the issues)
00:55:08*stripwax wonders what all happened there
00:55:12stripwaxsaratoga - mdct.asm
00:56:05wookey_the debian-way fix for this is to upload a .deb to your own repo and point debian/ubuntu people at that
00:56:15 Join Rob2223 [0] (
00:56:35gevaertsyes, either that or possibly volatile I guess
00:57:30wookey_letting the nice OBS people built it for $most distros would save you the trouble of maintainng this for loads of distros
00:59:24gevaertshm, I wasn't really aware of them. That might be a good solution
00:59:39 Quit bertrik (Quit: De groeten)
00:59:56 Quit Rob2222 (Ping timeout: 265 seconds)
01:00:10 Quit DerPapst (Quit: Leaving.)
01:00:14 Quit mrkiko (Quit: Lost terminal)
01:00:42saratogastripwax: thanks
01:01:31saratogano use of load/store multiple even though it loads 8 bytes at a stride
01:02:48saratogacould probably get another 1 MHz back by optimizing it
01:03:38stripwaxhrm, it doesn't really load 8 bytes at a stride. It loads a word from *in1, and then increments pointer by two words..
01:03:53stripwax(and reads a word from *in2, and then decrements pointer by two words)
01:04:09saratogaT[1], T[0], &z[j].re, &z[j].im
01:04:13saratogathose are both sequential
01:04:16stripwaxpossibly the best you could do for a quick optimise is using ldm to read the twiddles and write the outputs
01:04:22stripwaxYeah, what you said :)
01:04:30saratogafrom the ASM: 98: e5972004 ldr r2, [r7, #4] 9c: e5973000 ldr r3, [r7]
01:04:34saratogagcc is fucking stupid
01:04:42stripwaxthat is awful gcc
01:05:06stripwaxIs there a way to force gcc to use ldm instructions more often?
01:05:35saratoganot that I know of
01:05:48saratogai'm actually a little confused why it so frequently fails to use it
01:05:56stripwaxAgree, any/all of those sections could be hand optimised. take a look at the reflection stuff in ff_imdct_calc - it's also not exactly perfect. but again not the worst I've ever seen.
01:06:19saratogait may be that it doesn't realize that arm7tdmi is so much slower for single loads
01:06:43UnhelpfulTheSeven: how hard would it be for gdb to do that?
01:07:42TheSevenhonestly i have no clue about gdb internals
01:08:24gevaertsbluebrother, domonoky: have a look at That should be helpful for rbutil
01:08:27TheSevenyou can basically read/write memory/registers and halt/resume program execution
01:08:39wookey_gaeverts: bootloader upload beautifully painless and now it all works rather more slickly - top job :-)
01:08:41TheSevenand there is a breakpoint instruction
01:08:48TheSeven(some SWI)
01:09:02saratogaon the upside it does at least look like it was smart enough to avoid incrementing in1,in2 twice per loop by using an offset when loading :)
01:09:06wookey_If only I could make thunar-volman actually mount the thing everything would be lovely
01:09:14TheSevenin theory that's all you need for a debugger
01:10:01saratogaso 28mhz wma right now, and we could probably save another 1-2MHz in the prerotation, and a few more in the main fft, so i bet we could do < 25MHz WMA on PP
01:10:07 Quit jd (Read error: Connection reset by peer)
01:10:31 Join jd [0] (
01:10:31 Quit jd (Changing host)
01:10:31 Join jd [0] (~jd@Wikipedia/HellDragon)
01:10:57wookey_I'll attack the wife's now :-) and see if I can get gpodder syning working right too
01:11:12gevaertsgood luck!
01:12:06TheSevengah, it's too late for ARM condition codes today...
01:12:28wookey_what is the 'committing database' message about on boot? Tess's sansa does that 9 times on startup
01:12:30TheSevenneed to do a trivial timeout loop, and just don't get it right
01:12:40wookey_(with old software)
01:13:05TheSevensomething like start=tick; while(tick<(start+timeout)); (wrap-safe)
01:13:39JdGordon_use TIME_BEFORE() or TIME_AFTER() which are wrap safe
01:13:51stripwaxsaratoga - which bitrate? (trunk is already 27.56MHz for 96kbps)
01:13:56TheSevenit isn't in rockbox, and it needs to be ASM
01:14:14TheSeven(it is in very early norloader init code)
01:15:04saratogaoh misremembered, your test was 28.52 for 192k
01:15:24stripwaxyep (versus 29.4 for trunk)
01:15:55TheSevenso i assume i basically do add <end>, <tick>, <timeout>; label: cmp <tick>, <end>; b<cc> label
01:15:59saratogai can squeeze even more out of the codec itself too
01:16:01TheSevenbut what the heck is <cc>?
01:16:01wookey_ah. don;t upgrade the bootloader before upgrading rockbox - they run round switching from rockbox to native firmwares offering the other to do USB-connecting :-)
01:16:08saratogai wonder how far i can take the codec on PP
01:16:23stripwaxThe big speedup in my vorbis was getting rid of the ugly memcpy (that I put in right at the start of my fastermdct playing since ffmpeg mdct didn't work in place and so vorbis was screwed without a temp buffer) - but that temp buffer is gone now
01:16:57stripwaxTheSeven - "mi" ? I think
01:18:02 Quit anewuser (Quit: Another edition of chiptune gig WinterChip5! :O =ooo)
01:19:57wookey_oh dear - I'm getting piles of FAT: Directory bread(block 15271) failed
01:20:02gevaertswookey_: from what I can see (but I'm definitely no database expert), the database commits happen on boot if they went wrong earlier for some reason. I'd check if it still happens
01:20:17gevaertshm, time for fsck then...
01:20:32stripwaxTheSeven - (could always check the asm for TIME_BEFORE to check..)
01:20:43wookey_yes, good job there isn't anything valuable and un-backed-up on here
01:20:44 Quit JdGordon_ (Quit: Page closed)
01:20:50TheSevenstripwax: probably rather PL, or the args of the cmp reversed
01:21:03stripwaxTheSeven - ah, probably
01:21:26TheSevenor not - looks like my mind is screwed today...
01:22:25 Quit stripwax (Quit:
01:23:37 Quit MethoS- (Remote host closed the connection)
01:23:45 Join MethoS- [0] (~clemens@
01:26:28 Quit fyrestorm (Quit: lamers envy me like they envy bill g -- main boot xp, just the way it should be!)
01:28:32 Join fyrestorm [0] (
01:31:08 Quit MethoS- (Remote host closed the connection)
01:31:20 Join MethoS- [0] (~clemens@
01:36:54UnhelpfulTheSeven: cc is carry clear... equivalent to lo, cs equivalent to hs
01:37:57Torneflyback: Actually viewing all the data in detail can go in a plugin, yes
01:38:12TheSevenUnhelpful: i meant <cc> as a placeholder for "condition code" ;-)
01:38:22Torneflyback: but checking the overall health status and the pre-fail/old-age attributes would be good on startup
01:38:37Torneflyback: it can just display a splash telling you there may be a problem, and you can run the plugin if you want lots of details
01:38:39TheSevenand no, cs isn't equivalent to lo/cs :-P
01:38:52TheSevendamn, cc of course
01:39:13Torneflyback: the actual data is not hard to parse so that wouldn't be much code. only a human-friendly UI for looking at settings would be big :)
01:39:18Torneer, attributes, not settings
01:39:44 Quit PaulJam__ (Ping timeout: 245 seconds)
01:40:46 Join JdGordon1 [0] (~jonno@
01:41:16 Quit froggyman (Ping timeout: 272 seconds)
01:44:21Unhelpfuloh... ok. :)
01:45:00UnhelpfulTorne: cc == lo, cs == hs, yes?
01:45:36Unhelpfuli don't really use lo/hs... i just remember how arm does subtraction and that makes it clear enough when carry is set :)
01:49:15 Join CaptainKewl [0] (
01:49:37flybackyeah I know if you can do it, awesome
01:49:48flybackthen I can focus on things I want like network doom play
01:49:54flybackand more mod formats playback
01:50:06 Join Strife89|PalmTX [0] (
01:50:17flybackoh nice
01:50:20flybackyou got linux on your palm?
01:50:30flybackI just got a broken T5 off ebay to combine with mine into a working unit
01:50:34Strife89|PalmTXflyback: Not yet.
01:50:50Strife89|PalmTXflyback: VERSION me for details.
01:50:55*flyback also has 11 m130's he got off ebat he has to test so he can order cables and batteries
01:50:58JdGordon1flyback: seriously? we have to ask you again to stay on topic in here?
01:51:09JdGordon1that *is* your last warning
01:51:27flybackthat's ok i'lll probably just forget in 5 mins anyways
01:54:06 Join vegtoruci [0] (~vegtoruci@
01:58:00flybackI can't even remeber to hold onto my fucking mp3 player
01:58:08flybacknot even 24 hrs with rockbox on it and I dropped it
02:00:40 Quit saratoga (Disconnected by services)
02:00:40 Quit sbhsu (Disconnected by services)
02:02:45***Saving seen data "./dancer.seen"
02:02:46 Join saratoga [0] (~9803c6dd@rockbox/developer/saratoga)
02:03:31saratogawhat just happened
02:04:34flybacki'd love to comment but it's offf topic
02:05:22 Quit Strife89|PalmTX (Quit: Strife89|PalmTX)
02:05:46saratogalet me see if I can do this
02:06:14Mode"#rockbox +o saratoga" by ChanServ (ChanServ@services.)
02:06:23 Quit JdGordon1 (Ping timeout: 264 seconds)
02:12:26 Quit saratoga (Quit: CGI:IRC (EOF))
02:13:49flybackwelcome to
02:14:23 Join saratoga [0] (~9803c6dd@rockbox/developer/saratoga)
02:16:19 Join webguest35 [0] (
02:17:59webguest35still no malloc/free in rockbox?
02:18:11saratogaonly in plugins/codecs
02:18:53webguest35really? There's malloc in plugins??
02:19:27saratogayou can use the pluginlib one, or provide your own
02:19:39saratogai think most plugins use their own
02:20:58 Quit webguest35 (Client Quit)
02:41:02 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
02:53:20 Quit komputes (Remote host closed the connection)
02:54:34 Quit martian67 (Read error: Operation timed out)
02:59:29 Join froggyman [0] (
03:09:39 Join martian67 [0] (~martian67@about/linux/regular/martian67)
03:39:41 Quit MethoS- (Remote host closed the connection)
03:40:11 Part froggyman
03:43:22 Join JdGordon1 [0] (
03:44:25 Quit JdGordon1 (Client Quit)
03:51:14 Join Rob2222 [0] (
03:54:59 Quit Rob2223 (Ping timeout: 256 seconds)
03:59:41 Join phanboy4 [0] (
04:02:48***Saving seen data "./dancer.seen"
04:22:59 Quit TheSeven (Disconnected by services)
04:23:11 Join The_Seven [0] (~theseven@rockbox/developer/TheSeven)
04:23:22 Nick The_Seven is now known as TheSeven (~theseven@rockbox/developer/TheSeven)
04:31:48TheSevenfor anyone keen on testing norboot: I have made a new version, that will now hopefully support all board and lcd types
04:35:28 Quit tmzt (Ping timeout: 272 seconds)
04:39:40TheSevenbtw, it's using dma-based lcd updating, and it's working just fine
04:42:28 Join kugel [0] (~kugel@rockbox/developer/kugel)
04:44:51 Join Barahir [0] (
04:48:19 Quit Barahir_ (Ping timeout: 258 seconds)
04:51:00 Quit Strife89 (Quit: Bed.)
04:56:30 Join tmzt [0] (
05:03:57 Quit kugel (Ping timeout: 272 seconds)
05:06:54 Quit Tomis (Read error: Operation timed out)
05:06:57 Join Tomis [0] (~Tomis@
05:34:47 Quit dmb (Quit: Leaving)
05:35:18 Join dmb [0] (~Dmb@unaffiliated/dmb)
05:37:57 Quit Horscht (Quit: Verlassend)
05:38:07 Quit FlynDice (Remote host closed the connection)
05:41:21 Join FlynDice [0] (
05:43:43 Join `Assassin [0] (
06:02:50***Saving seen data "./dancer.seen"
06:28:40 Quit panni_ (Read error: Connection reset by peer)
07:01:49 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
07:26:20 Join einhirn [0] (
07:33:11JdGordonare the -Fixed fonts broken?
07:34:27JdGordonand the sazanami ones
07:41:32JdGordonnope, I broke font.c
07:56:16 Join evilnick [0] (~evilnick@rockbox/staff/evilnick)
08:02:52***Saving seen data "./dancer.seen"
08:03:29 Quit vegtoruci (Ping timeout: 248 seconds)
08:09:47 Join kaniini [0] (
08:26:06 Join ender` [0] (
08:28:35 Quit CaptainKewl (Quit: ( :: NoNameScript 4.22 :: ))
08:28:44 Join Zagor [0] (
08:28:44 Quit Zagor (Changing host)
08:28:44 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
08:35:21 Join Adub- [0] (
08:40:32 Quit jerl92 ()
08:44:27*JdGordon preemptively runs from the pitchfork wielding mob and annoys everyone with FS #10984
08:47:46 Join flydutch [0] (
08:49:30 Join wind [0] (
09:12:38 Join petur [0] (~petur@rockbox/developer/petur)
09:20:50JdGordonmultifont with this approach seems to be dead simple, so if someone is planning on objecting can you please speak up soonish?
09:26:57 Join funman [0] (~fun@rockbox/developer/funman)
09:35:11 Join stripwax [0] (
09:35:37 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
09:37:23 Quit stripwax (Client Quit)
09:39:06B4gderZagor: ohloh says Rockbox is "only" 920078 lines
09:39:15B4gderah right, they exclude empty ones
09:39:22B4gder115929 blank lines
09:39:44B4gder134638 comment lines!
09:40:03JdGordonohloh stopped updating us
09:40:19B4gder"Updated 13 Jan 2010"
09:40:54B4gderhm yes it says "Failed 25 days ago"
09:43:56*JdGordon has apparently touched 8% of the code! that cant be good :p
09:48:43 Quit S_a_i_n_t (Ping timeout: 265 seconds)
09:50:36 Join GeekShado_ [0] (
09:52:07 Quit phanboy4 (Ping timeout: 272 seconds)
09:53:15 Quit GeekShadow (Ping timeout: 256 seconds)
09:55:41 Join kugel [0] (~kugel@rockbox/developer/kugel)
10:01:53 Quit kugel (Ping timeout: 248 seconds)
10:02:55***Saving seen data "./dancer.seen"
10:06:10 Join _zic [0] (
10:12:01 Quit AndyIL (Ping timeout: 248 seconds)
10:15:21_zicwhile completing russian translation for rockbox utility yesterday, i noticed the <numerusform> frame
10:16:41_zicdoes anybody know what it is intended to do/be/mean/whatever ?
10:17:09 Join LinusN [0] (~linus@rockbox/developer/LinusN)
10:18:42 Join AndyI [0] (~pasha_int@
10:34:24 Quit BHSPitMonkey (Remote host closed the connection)
10:42:43 Quit Adub- (Read error: Connection reset by peer)
10:42:51linuxstbgevaerts: I don't think is a corrupted installation - the checksum differs by 5, which is also the difference between the 4g greyscale and Color/Photo model number...
10:43:15 Join Adub- [0] (
10:46:02funman_zic: i think it's for languages which have different plural forms
10:46:50 Quit rhodan_ ()
10:47:20 Join rhodan [0] (
10:47:57funmanI can't find doc for the <numerus> tag in Qt4 doc, do you use a text editor or qt linguist ?
10:48:31 Quit `Assassin (Read error: Connection reset by peer)
10:48:41_zicfunman, yes i also think so, and russian is one of these languages, but there's no explicit indication in which one to put one particular form
10:48:46_zicno, i use gedit
10:49:07funman < I would think the 3 <numerus> are the 3 rules
10:51:41 Join Zarggg_ [0] (
10:51:43 Quit Zarggg (Read error: Connection reset by peer)
10:53:57gevaertslinuxstb: does that model get into the crc?
10:55:11_zicfunman, kind of confusing... but thanks :) i think i'm going to install qt linguist on my windows machine
10:57:00_zicqt linguist doesn't seem to run on linux :(
10:58:57gevaertsit does
10:59:03linuxstbgevaerts: Yes, the bootloader checksum is initialised with the model number
10:59:25gevaertsah, ok
10:59:47linuxstbSo a small difference normally means the wrong version installed.
11:00:06gevaertsgood to know
11:03:31 Quit petur (Ping timeout: 246 seconds)
11:03:54 Join kugel [0] (kugel@
11:03:56 Quit kugel (Changing host)
11:03:56 Join kugel [0] (kugel@rockbox/developer/kugel)
11:04:42 Join petur [0] (~petur@rockbox/developer/petur)
11:05:57 Join S_a_i_n_t [0] (S_a_i_n_t@
11:06:00 Quit S_a_i_n_t (Client Quit)
11:06:06funman_zic: on ubuntu it is called "linguist" or "linguist-qt4" in qt4-dev-tools package
11:06:17 Join S_a_i_n_t [0] (S_a_i_n_t@
11:06:28 Quit S_a_i_n_t (Client Quit)
11:07:46 Join S_a_i_n_t [0] (S_a_i_n_t@
11:07:58 Quit S_a_i_n_t (Client Quit)
11:08:10 Join S_a_i_n_t [0] (S_a_i_n_t@
11:08:16 Quit S_a_i_n_t (Client Quit)
11:09:31 Join S_a_i_n_t [0] (~st.lasciv@
11:12:48_zicfunman, gevaerts: thanks :) didn't know it, google code page doesn't provide the linux version
11:14:51 Quit funman (Quit: free(random());)
11:15:35 Quit S_a_i_n_t (Quit: [St.] has exited mIRC™)
11:15:58 Join S_a_i_n_t [0] (S_a_i_n_t@
11:20:31 Quit jfc (Ping timeout: 260 seconds)
11:30:20 Quit S_a_i_n_t (Quit: [St.] has exited mIRC™)
11:31:18 Join jfc [0] (
11:51:47 Join DerPapst [0] (
11:54:42AlexPJdGordon: Flyback has already had numerous last warnings - just ban him next time
11:54:51AlexPOr mute at least
12:02:57***Saving seen data "./dancer.seen"
12:08:52 Part LinusN
12:09:54 Join Rob2223 [0] (
12:10:24 Quit pixelma (Read error: Connection reset by peer)
12:10:26 Join pixelma [0] (quassel@rockbox/staff/pixelma)
12:12:16 Quit Rob2222 (Ping timeout: 260 seconds)
12:20:33 Quit AndyI (Ping timeout: 248 seconds)
12:22:27 Join AndyI [0] (~pasha_int@
12:26:49 Quit kugel (Ping timeout: 258 seconds)
12:38:53 Join bmbl [0] (
12:38:55 Quit bmbl (Changing host)
12:38:55 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
12:43:44 Join MethoS- [0] (~clemens@
12:48:49 Quit AndyI (Ping timeout: 248 seconds)
12:53:36 Join kugel [0] (kugel@rockbox/developer/kugel)
12:54:02 Join AndyI [0] (~pasha_int@
13:01:54 Join kugel_ [0] (kugel@
13:01:56topikgevaerts: i'm reading your and betriks meizu experiments on the forum from december. did you do anything special to compile meizu_dfu and/or the bootloader?
13:10:58 Join perfectdrug [0] (
13:14:05gevaertstopik: no
13:15:08 Join perfectdrug_ [0] (
13:18:49 Quit perfectdrug (Ping timeout: 256 seconds)
13:21:41topikneither compiles atm
13:21:43topikfrom svn
13:24:12topikfrom what i read, you made some nice progress and i was hoping to contribute by testing/debugging on my m6sl
13:34:57 Join Kitr88 [0] (
13:37:21 Quit Kitar|st (Ping timeout: 248 seconds)
13:39:41 Quit Kitr88 (Ping timeout: 252 seconds)
13:40:58 Quit Zarggg_ (Read error: Connection reset by peer)
13:41:04 Join Zarggg_ [0] (
13:43:21TheSeventopik: pastebinning a copy of the compile errors might be helpful
13:44:42 Join Kitar|st [0] (
13:47:51TheSevenouch, that was me who broke it :-)
13:48:19TheSeventhere should be a stub for that file for the meizus though
13:48:38topikyeah, i figured it was because of the nano2g ftl
13:48:58TheSeventhe path where it's looking for is weird though
13:49:02TheSeventhis looks like a makefile bug
13:49:28TheSevenwrong include path for dependency generation?
13:49:52TheSevenhow does your make.dep look likeß
13:50:44topiki think you're right
13:52:14topikall i did was ../tools/configure, 110, B with current svn
13:53:30 Join Grahack [0] (
13:56:04TheSevenhm, it's scattering them along 3 places: /opt/rockbox/<proper path>, /opt/rockbox/build-meizu/<proper path>, /opt/rockbox/build-meizu/<plain filename>
13:56:13TheSevenobviously the first one is right
13:57:09 Join Farthen [0] (
13:59:22bug2000Rockbox is amazing. I would never guess my music player is capable of all this. Thank you people for doing such an amazing work!
13:59:52 Join Lss [0] (
14:00:00 Join nuCCk [0] (
14:00:37nuCCkAny thing new about ipod classic 160gb and rockbox?
14:01:08 Quit Grahack (Ping timeout: 258 seconds)
14:01:46TheSevennuCCk: the old one or the new one?
14:02:58***Saving seen data "./dancer.seen"
14:07:59 Join evilnick_B [0] (~0c140464@rockbox/staff/evilnick)
14:08:08nuCCkthink it the new it is about 2 years olde
14:09:08nuCCkcan i check it some place
14:10:36TheSeventhen it's the old one
14:10:48TheSeventhe new one was released just some months ago
14:11:15TheSevenwell, we can execute code on that one, but not much has been done with it yet
14:11:53nuCCkok whit the olde?
14:12:18nuCCkis there any work planed
14:14:36TheSeveni was referring to the old one
14:14:49TheSevenand nobody seems to be working on it currently
14:15:29topikTheSeven: if i edit make.dep manually, will it work?
14:15:45TheSevenbut you should find out what's going wrong in the first place
14:18:38-->"help" received from forkup (
14:19:12topikfor my fuze and nano2g things get generated fine
14:21:08topikthe meizu ms6sl stub files are just not there
14:27:40 Join Stephen__ [0] (
14:29:29 Join PaulJam__ [0] (
14:30:40Stephen__JdGordon, do you use the rockbox_none base skin ? if uiviewport is used it's not recognised without a .sbs on normal themes.
14:35:57*_zic just uploaded the russian translation update for rbutil. please can someone go accept it ?
14:36:26windCould I use screendump in the system-debug to print screen?
14:37:46evilnick_Bwind: Yes
14:38:01windHow to?
14:38:12 Quit kugel_ (Ping timeout: 240 seconds)
14:38:19 Quit kugel (Ping timeout: 258 seconds)
14:38:54evilnick_Bwind: Check the manual for your player, but usually it disables USB connectivity so each time you plug it into your computer then it'll take a screendump instead of connecting
14:40:58 Join Sajber^ [0] (
14:43:51windevilnick_BL:Could the Rockbox add a new function to achieve it?
14:44:40windInstead of plug the player into computer.
14:45:20evilnick_BI'm sure that it'd be possible to do so, but I don't see why it shouldn't be done the current way
14:46:15evilnick_BYou'd enable screendump, then plug into the computer each time you want it to "print screen", to then access each of these files, you would then disable screendump and re-connect then copy the files over (they are stored in the root of the drive)
14:46:43windBecause sometimes I want to save my screen.
14:49:26evilnick_BIn that case, I believe that you could (NB - I haven't looked into the code) but you'd need to compile your own build to do this.
14:49:42evilnick_BIt'd be a lot simpler to just use screendump when you're near a computer
14:49:50gevaertsyou'd need to find an available button combination
14:51:29TheSevenanybody with a nano2g with lcd type 2 or 3 around? (lcd detect: 0:0, 1:1 / 0:0, 1:0 / 0:1, 1:1)
14:52:04 Quit nuCCk ()
14:52:08*TheSeven needs to test the new driver
14:52:25windI dont want to use computer to achieve this function.
14:52:34 Join nuCCk [0] (
14:52:44linuxstbwind: Well, that's the only way Rockbox can do screendumps, unless someone writes a new feature...
14:53:55windCould we make a button just like PrtScr in windows?
14:53:58topikwhere do you see those lcd detect values, TheSeven ?
14:54:19TheSevenother => status
14:54:29TheSevenpress select to show the lower page
14:54:39gevaertswind: sure. Find a button that's unused and use that
14:55:30topik0:1, 1:0
14:55:37TheSeventhat's a type 7
14:55:42TheSeven(the one i also have)
14:55:49topikthen i'm unable to help :(
14:55:59windHow to do it without computer thats what I am concerned.
14:56:33 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
14:57:19linuxstbwind: You can't. Unless you're a C programmer, and can modify how Rockbox works.
14:57:59 Join vegtoruci [0] (~vegtoruci@
14:58:14topikit's a shame my nano2g is only 2GB. that's why i hoped for meizu m6sl progress. same cpu but mine is 8GB
14:58:23topiknano 2g really has good runtime
14:58:30windlinuxstb:But in fact,I can't.
14:59:12gevaertswind: technically you don't need a computer, since rockbox only looks at the power pins for this
14:59:40 Join rhodan_ [0] (
15:00:08 Join Strife89 [0] (~upirc@
15:00:39 Quit rhodan (Read error: Operation timed out)
15:00:43 Join kugel [0] (kugel@rockbox/developer/kugel)
15:00:44 Join kugel_ [0] (~kugel@
15:00:54 Nick Strife89 is now known as Strife89|PalmTX (~upirc@
15:00:59 Quit nuCCk ()
15:02:01windonly the power pins?
15:06:12 Part Strife89|PalmTX
15:07:09windI still have a problem: how to exit the screendump?
15:08:28Tornego and select the same option again, it's a toggle
15:08:33 Quit AndyI (Ping timeout: 248 seconds)
15:10:26windbut if I have open the screendump,then I put the USB line into the player,the player cannot do anything again.
15:10:48windI have to reset it
15:11:35TorneThat's not supposed to happen
15:13:13windTorne:But it has happened
15:14:02 Join AndyI [0] (~pasha_int@
15:15:57TorneYes, obviously
15:16:04Tornebut my point is there is no "way to exit the screendump"
15:16:16Tornebecause it's not supposed to make the player hang
15:16:25Torneit should just take the screendump, show a splash, then you can carry on using it
15:16:30evilnick_BTorne: After you disconnect USB, then you should be able to disable the screendump option again and it'll work as it did before when you connect USB
15:16:43evilnick_BEr... s/Torne/wind/
15:16:56Torneevilnick_B: yes, but his point is once he plugs in the usb cable for the dump his player stops working entirely
15:19:33 Quit evilnick (Ping timeout: 272 seconds)
15:21:25 Quit bmbl (Ping timeout: 245 seconds)
15:22:28 Join robin0800 [0] (
15:27:46 Quit vegtoruci (Ping timeout: 258 seconds)
15:36:39 Quit Lss (Read error: Connection reset by peer)
15:37:37 Join Lss [0] (
15:41:45 Quit PaulJam__ (Ping timeout: 256 seconds)
15:46:51 Join jgarvey [0] (
15:48:07 Join Riku [0] (
15:48:35 Nick Riku is now known as Guest45915 (
15:50:10 Quit Lss (Ping timeout: 245 seconds)
15:53:59 Join petur2 [0] (~petur@rockbox/developer/petur)
15:54:13 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
15:54:25 Quit petur (Ping timeout: 248 seconds)
15:56:53 Quit wind (Quit: CGI:IRC (EOF))
15:57:08 Quit Zambezi_ (Changing host)
15:57:09 Join Zambezi_ [0] (Zulu@unaffiliated/zambezi)
15:57:17 Nick Zambezi_ is now known as Zambezi (Zulu@unaffiliated/zambezi)
15:58:12 Join petur [0] (~petur@rockbox/developer/petur)
15:58:17 Quit petur2 (Ping timeout: 240 seconds)
16:03:00***Saving seen data "./dancer.seen"
16:06:13 Quit kugel_ (Quit: exit(0);)
16:06:18 Quit Farthen (Disconnected by services)
16:06:40 Join Farthen_ [0] (
16:06:41 Nick Farthen_ is now known as Farthen (
16:09:11TheSevendamn, the new LCD driver breaks alpha blending :-/
16:10:57 Join darkham [0] (
16:13:28 Join AndyIL [0] (~pasha_int@
16:15:25 Join thegeek [0] (
16:15:32 Join thegeek_ [0] (
16:15:56 Join dfkt [0] (dfkt@unaffiliated/dfkt)
16:16:06 Quit AndyI (Read error: Connection reset by peer)
16:31:05 Join kachna [0] (
16:31:54TheSevenreally nobody with another nano2g here?
16:32:03*TheSeven needs testers
16:32:30 Quit kachna (Remote host closed the connection)
16:50:20 Join dmb_ [0] (~Dmb@unaffiliated/dmb)
16:55:38 Quit robin0800 (Remote host closed the connection)
16:55:59 Quit dmb (Ping timeout: 256 seconds)
16:56:28 Quit kugel (Ping timeout: 265 seconds)
17:06:50 Join panni_ [0] (
17:10:16 Quit Zagor (Quit: Clint excited)
17:19:39 Join Strife89 [0] (~michael@
17:22:22 Quit Strife89 (Client Quit)
17:26:17 Join vegtoruci [0] (~vegtoruci@
17:26:27 Quit Topy44 (Ping timeout: 256 seconds)
17:28:11 Join Topy44 [0] (
17:28:48 Join evilnick [0] (
17:33:58 Join piotrekm [0] (~mirggi@unaffiliated/piotrekm)
17:36:51 Join pamaury [0] (
17:49:01 Quit parafin (Quit: So long and thanks for all the fish)
17:49:09 Join parafin [0] (
17:49:25 Join phanboy4 [0] (
17:52:12 Quit killan (Quit: ( :: NoNameScript 4.22 :: ))
17:52:31 Quit piotrekm (Ping timeout: 256 seconds)
17:53:40 Join killan [0] (
17:56:00 Join captainkewllll [0] (~2669ecc2@gateway/web/freenode/x-qmqbxpydqygutxwx)
17:57:09 Join kugel [0] (~kugel@rockbox/developer/kugel)
17:58:55 Join komputes [0] (~komputes@ubuntu/member/komputes)
17:59:56 Quit _zic (Quit: Ex-Chat)
18:00:47 Quit thegeek (Quit: ( :: NoNameScript 4.2 :: ))
18:03:03***Saving seen data "./dancer.seen"
18:10:43 Join akur [0] (~akur@2001:690:2200:9aaa:224:8cff:fe7b:1275)
18:11:37akuri would like have some wiki priveleges, please?
18:11:51linuxstbWhat's your wiki name?
18:13:14linuxstbakur: That's done.
18:19:02kugel we should've patented folder skip :) cowon implented it in the new E2 firmware (and it works exactly how it does in rb)
18:20:55dionoeawhat does folder skip do ?
18:21:42 Quit GeekShado_ (Read error: Connection reset by peer)
18:22:14 Join GeekShado_ [0] (
18:23:54*linuxstb guesses it skips folders
18:24:15 Quit evilnick (Ping timeout: 256 seconds)
18:24:47kugeldionoea: you can skip to the next/previous folder directly from the wps with a key combo
18:25:07dionoeaoh, neat.
18:25:18linuxstbDidn't that idea come from the iriver firmware originally?
18:25:27linuxstb(h1x0 or h3x0)
18:25:37dionoealinuxstb: it could have been several things :) (like skip this subfolder when inserting in playlist for example)
18:26:20kugellinuxstb: that's possible, I don't know
18:28:51dionoeais that documented somewhere? (I can't find anything in the e200, ipod video or archos recorder docs, searching for "folder")
18:29:13JdGordone200 its power+next/prev
18:29:27JdGordonipod is select+next/prev i think
18:29:53dionoeabah, I should have looked for "directory" :)
18:30:43kugelshort next, release, then long next (or prev) works on all targets I think
18:31:02 Join piotrekm [0] (~pm@unaffiliated/piotrekm)
18:31:38 Quit petur (Quit: *plop*)
18:33:41piotrekmTheSeven: i have just read the log, still want to test it?
18:34:22TheSevenit should finally work on yours
18:34:47piotrekmand what's that fuss about double-buffering, any visible difference?
18:35:04TheSevenwe haven't tried that with rockbox yet
18:35:22TheSevenbut norloader, iloader and ibugger have lightning-fast lcd updates now, even with caches disabled!
18:35:38TheSevenno more visible tearing during the updates etc.
18:36:18piotrekmsounds cool. And no possible interference with nand reading/writing or anything like that?
18:36:31piotrekmi just guess it also uses dma
18:36:46TheSevenwe have a total of 9 (woah!) dma channels on that device
18:37:23piotrekmthis little ipod keeps amazing me. About a month ago i considered it the most limited device i possess
18:38:56 Quit Tuplis (Ping timeout: 265 seconds)
18:41:28kugelwell, with the apple os it surely is on of the most limited ones :)
18:46:17 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
18:47:18TheSevenkugel: not neccessarily. they have quite a lot of features nowadays
18:47:32TheSevenbut it definitely is the most limited os compared to the hardware it's running on :-)
18:49:37darkhami would like to know some rockbox next step...
18:49:58 Join froggyman [0] (
18:50:09*TheSeven doesn't get that darkham means
18:51:46darkhamTheSeven, i mean i would like to know what developers are almost ready to release, in rockbox features
18:52:36 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
18:52:55piotrekmisn't that what is in svn?
18:53:23 Join bertrik [0] (
18:53:41Tornedarkham: anything that is in a suitable state to be checked in, is checked in
18:53:49Torneanything else, is not kept track of particularly.
18:54:32TorneWell, ok, some things are sitting as patches in FS
18:54:41 Quit GeekShado_ (Ping timeout: 248 seconds)
18:55:23piotrekmlike the sorting tags support *_*
18:56:10 Join Tuplanolla [0] (
18:56:13 Join Atlanx [0] (
18:56:23 Nick Tuplanolla is now known as Tuplis (
18:56:28Atlanxgood day
18:56:51kugelpiotrekm: there's a patch about it, but it costs a lot of ram
18:57:35piotrekmkugel: i saw it, but didn't try. much meaning 1MB or more?
18:57:43AtlanxI get a new error from Rockbox Uitlity for updating to RB 3.5.
18:57:53kugeldepends on how many music you have
18:58:28kugelthe database can get quite big, and that patch doubles that
18:58:36AtlanxProblem found: - Abweichendes Gerät erkannt. Installiertes Gerät: e200, gewähltes Gerät: Sansa e200v1 series -
18:58:46piotrekmon a 2GB ipod it's not a problem i suppose
18:59:52AlexPStorage size is nothing to do with it
18:59:54AlexPRAM usage is
19:00:27 Join freddyb [0] (
19:00:34AlexPAtlanx: Any chance of a rough English translation?
19:00:44piotrekmi mean, with 2GB i can't get a large amount of music, so my library wont swell much
19:00:46AlexPAtlanx: Is it complaining that the target name changed?
19:00:54AlexPpiotrekm: Ah right
19:00:56AlexPSorry :)
19:01:03piotrekmno problem;)
19:01:24Atlanxyes: Problem : Diffrent Device found. Installed Device: e200, choosen Device: Sansa e200v1 series
19:01:35AlexPAtlanx: You can ignore that
19:01:56AlexPAtlanx: There was an internal name change, you should only get that message once
19:02:29Atlanxyes , but it is an irritating error message. should i report it? - ok will ignore it
19:02:42AlexPNo, it will only happen once and is intentional
19:02:51Atlanxok thank you
19:02:55AlexPno worries
19:03:36freddybTorne: you still playing with Frotz?
19:03:52Tornei'm not developing it at the moment
19:04:03Tornebut I will, yes
19:06:28freddybI resyned my virtual keyboard patches and added a new accerated variation to the scroll_1d patch. You might try it you didn't like Keep_rollin or regular Scroll_1d.
19:06:37freddybFS #10763
19:06:42Tornei didn't actually get around to trying either of those
19:06:59Tornemessing with DMA and USB charging atm
19:07:04kugelinteresting, wikipedia already claims we have a 4 month release cycle now
19:07:36 Quit bluebrother (Ping timeout: 260 seconds)
19:08:56 Nick froggyman is now known as frog (
19:08:57 Join bluebrother [0] (
19:08:57 Quit bluebrother (Changing host)
19:08:57 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
19:09:09 Nick frog is now known as froggyman (
19:09:21gevaertsdomonoky, bluebrother: did you see my note last night?
19:10:27domonokygevaerts: no, which note ?
19:14:43TheSevenbtw, the 4-month release cycle has been in there since r309625126, 17:03, 23 August 2009
19:15:08*domonoky doesnt really sees the benefit of building a rpm via some buildservice, if we can just build one static binarys.
19:15:11 Quit Tomis (Ping timeout: 252 seconds)
19:15:15piotrekmTheSeven: So shall i expect the norloader soon, or just wait?
19:15:28 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
19:15:58TheSevenpiotrekm: i updated the files on my server
19:16:02TheSevensame procedure as yesterday
19:16:27gevaertsdomonoky: sure, but currently we don't have an up to date 64 bit binary, and the one that's there doesn't actually work. As long as static binaries are a pain to build, we should keep an eye open for alternatives
19:18:50domonokytrue, but there are probably better solutions. (like using distributed buildsystem, and getting clients for all 4 rbutil variants.)
19:19:07piotrekmTheSeven: ok, i'll try
19:20:38 Quit freddyb (Remote host closed the connection)
19:26:58 Join JdGordon_ [0] (~836b006a@gateway/web/freenode/x-eimyrajnsldrjybi)
19:29:18piotrekmTheSeven: i have to say it even got a bit worse... the screen is white after running " norloader.dfu" and i can't reset the ipod;>
19:29:58 Part froggyman
19:30:27Stephen__JdGordon_, see my message earlier ?
19:31:42 Join Crackerizer [0] (
19:31:57bluebrothergevaerts: building rbutil with static Qt isn't complicated at all, you "just" need to build a static binary of Qt. Which takes quite some time. Btw, I've already thought about creating an rpm for fedora.
19:32:11CrackerizerI need some help today.
19:33:02bluebrotherhaving distribution packages shouldn't be a problem, if we can trigger updated. Which I assume would be the problem. Plus the volunteers creating those packages (I have some knowledge about rpm, but around zero for deb)
19:33:06CrackerizerI'm trying to copy my songs back from ipod. But it seems hell slow.
19:33:58CrackerizerAnybody here have the same problem please give me suggestion.
19:34:42 Quit JdGordon_ (Ping timeout: 248 seconds)
19:35:06 Join pabouk [0] (
19:35:32 Quit Guest45915 (Ping timeout: 252 seconds)
19:37:38evilnick_BCrackerizer: Are you using Rockbox's USB stack or the OF's?
19:38:32gevaertsbluebrother: I didn't say it's complicated! I maintain that it's a pain though :)
19:38:34piotrekmTheSeven: now i managed to reset it finally, and with hold on repeated the "nano2brdfu...", and the only difference from yesterday is that today the lcd turned white back again, no further response
19:38:46 Join pabouk_ [0] (
19:39:11linuxstbCrackerizer: Which version of Rockbox are you using? 3.5?
19:40:05TheSevenpiotrekm: damn, so type-2 lcd init still doesn't work
19:40:11 Quit pabouk (Ping timeout: 240 seconds)
19:40:13pabouk_Could you please raise severity and/or priority of ? The latest iriver H120 builds are unbootable.
19:40:14TheSeven(but doesn't hang any more at least)
19:40:15 Nick pabouk_ is now known as pabouk (
19:40:59TheSevenwhen you reboot it with the hold switch on (and the backlight turning on), can you see the ipod in lsusb? (as "Unified iBugger" or ffff:4248 or such)
19:41:35piotrekmi reverted to iloader, i will tell you in a moment
19:41:36 Join JdGordon_ [0] (~836b006a@gateway/web/freenode/x-rkdpwjsyuudginqd)
19:41:44JdGordon_kugel: ping?
19:41:49 Join Grahack [0] (
19:42:04JdGordon_pixelma: any chance you can test the radio patch on the ondiofm today/tomorow?
19:42:38JdGordon_kugel: would you like to fix the skin buffer allocation so it uses only as much as needed?
19:42:40darkhamexcuse me, with rockbox 3.5 is changed something about it and sansa fuze?
19:42:51piotrekmTheSeven: i can't actually reboot it to get the white screen back again, only after rebooting i can load the norloader.dfu again
19:43:01darkhamit's ever a dual boot?
19:43:16kugelI would like to have a look yes, but I'm still during exams so it has to wait a bit
19:43:33JdGordon_ah sure
19:43:36JdGordon_exams suck :)
19:44:05gevaertsdarkham: with 3.5 the fuze was promoted to "stable"
19:44:45piotrekmTheSeven: but when i load the norboot-orig.dfu back again, i get a strange two-color apple logo before the iloader shows, does that have anything to do with the new norboot?
19:45:12darkhamgreat, must i ever upgrade it reaplacing the /.rockbox in the player?
19:45:34TheSevenpiotrekm: that happens because norboot can't find its logo, because you overwrote it (with iloader and such)
19:45:40JdGordon_AlexP: 10853 TEST IT! :)
19:45:54AlexPJdGordon_: Calm down tiger :)
19:46:00AlexPJdGordon_: And OK :)
19:46:11bug2000AlexP: I fell in love with Rockbox.
19:46:22AlexPbug2000: Good :)
19:46:29JdGordon_I cant really finish multifont untill fm is in (I dont like having many patches going in parallel)
19:46:42bug2000Only tiny issue left to fix is a font issue [no Hebrew >.<] with the theme I use. Didn't check if it does show it in the default theme yet.
19:47:24bug2000And it doesn't >.<
19:48:50evilnick_Bbug2000: Have you seen:
19:48:52JdGordon_then use a font with the same size that does have hebrew
19:48:53AlexPYou just need to pick or make a theme that does
19:49:02gevaertspabouk: did you test any revision other than the ones explicitely listed?
19:49:39linuxstbpabouk: Also, have you tried the usual things, such as completely removing the old .rockbox folder, and checking your disk for errors?
19:50:21bug2000evilnick_B: Accutally I was about to look for it. I'm just noting to self stuff :)
19:50:48 Join pabouk_ [0] (
19:50:53AlexPJdGordon_: I was thinking about album art for radio stations - I reckon it'd be best to have art named by preset name in a specific folder
19:51:29JdGordon_yeah, thats probably how I'd do it too
19:51:36JdGordon_frequency doesnt really make sense
19:52:06AlexPAnd by preset name would help for stations that change fequency in different places
19:52:10*gevaerts wonders if pabouk_ saw the questions
19:52:16AlexPThe art would work for different preset files
19:52:21 Quit jordan` (Remote host closed the connection)
19:52:39 Quit pabouk (Ping timeout: 256 seconds)
19:52:43 Nick pabouk_ is now known as pabouk (
19:52:49TheSeven[19:48]<gevaerts>pabouk: did you test any revision other than the ones explicitely listed?
19:52:51TheSeven[19:48]<linuxstb>pabouk: Also, have you tried the usual things, such as completely removing the old .rockbox folder, and checking your disk for errors?
19:53:01linuxstbWhat about people that travel around? many countries have a "radio 1"
19:53:23AlexPI have e.g. BBC Radio 1
19:53:27AlexPSo hadn't thought about that
19:53:35gevaertslinuxstb: in that case, use a more elaborate naming scheme
19:54:23gevaertsor add RDS station ids to the presets file, and exclusively use those. People will probably complain though
19:54:52 Quit Atlanx (Quit: ChatZilla 0.9.86 [Firefox 3.5.7/20091221164558])
19:55:04AlexPgevaerts: I'll stop complaining as soon as someone does beast RDS :)
19:55:17paboukgevaerts: Unfortunately I do not have a build environment so I was able to try what was available on the web. I upgrade almost every day so it was quick to find the latest working daily build.
19:55:30gevaertspabouk: ok, just making sure
19:56:08gevaertsit's just a 10-revision gap, so it's not too bad :)
19:56:59pabouklinuxstb: I am sorry I forgot about this. Checking at the moment. I will report back.
19:58:14 Quit Crackerizer (Quit: CGI:IRC)
19:58:45 Join Lss [0] (
20:00:41 Join pabouk_ [0] (
20:00:42JdGordon_Stephen__: ui viewport setting is ignored if a .sbs is loaded
20:01:04Stephen__yeah but if no sbs is loaded uiviewport seems to be ignored also
20:01:23 Join evilnick [0] (
20:01:27JdGordon_that's a bug then
20:01:32JdGordon_file it
20:01:49gevaertsjhulst: ping
20:01:54Stephen__will do thanks. wasn't sure if you knew about it or not probaly needs to be deleted
20:02:17 Quit pabouk (Ping timeout: 256 seconds)
20:02:23 Nick pabouk_ is now known as pabouk (
20:02:42gevaertsjhulst: jhulstlab-jhulst seems to put wrong dates in its build results, like r24571-000208 (instead of r24571-100208)
20:03:04***Saving seen data "./dancer.seen"
20:03:27 Quit amiconn (Remote host closed the connection)
20:03:27 Quit pixelma (Read error: Connection reset by peer)
20:05:04darkhamuser can set the registration destination?
20:05:24jhulstgevaerts: hmm, I'll take a look
20:05:38gevaertsjhulst: that's generated by "date -u +%y%m%d" normally
20:05:42 Join pabouk_ [0] (
20:05:51JdGordon_Stephen__: if its getting ignored it means you still have statusbar set to custom
20:06:06jhulstgevaerts: ah my lab machine, was having problems with NTP. Should be resolved now
20:06:08JdGordon_if you set it to off it will work
20:06:14JdGordon_(or top or bottom)
20:06:30gevaertsthanks for checking
20:06:44jhulstno problem
20:08:31 Quit pabouk (Ping timeout: 256 seconds)
20:08:32pabouk_linuxstb: I removed .rockbox and .scrobbler-timeless.log and the result is the same as in the tracker: I04:IllInstr at 31057FE8 _immediately_ after bootloader finishes its job.
20:08:39 Nick pabouk_ is now known as pabouk (
20:08:55 Join jordan` [0] (~jordan@
20:08:56Stephen__JdGordon_, nope still same problem
20:09:28JdGordon_what does it show? garbage backdrop?
20:09:41Stephen__yeah i can take a screenshot on sim ?
20:10:09JdGordon_no, I've seen it
20:10:42JdGordon_file the bug, say exactly what you are doing (what theme you started ith, how yu changed, etc)
20:11:00Stephen__it's any theme that has uiviewport set and no .sbs.
20:11:10Stephen__have to remember my password and it's regged to my work email.
20:11:15Stephen__can i sign up a new one ?
20:11:21 Quit evilnick (Ping timeout: 256 seconds)
20:11:24gevaertspabouk: have you tried checking the filesystem? I don't see any suspicious commit in that range
20:11:45linuxstbpabouk: OK... And the latest current build gives the same problem?
20:12:25gevaertslinuxstb: r24571 is the latest, so yes
20:14:00 Join Tomis [0] (~Tomis@
20:14:41 Join pabouk_ [0] (
20:15:45pabouk_linuxstb: Yes, it was tested with the latest one which is still the same as reported in the tracker: r24571-000208.
20:15:46linuxstbpabouk_: Have you been trying the daily builds, or current builds (or both) ?
20:16:27 Quit pabouk (Ping timeout: 256 seconds)
20:16:38 Nick pabouk_ is now known as pabouk (
20:16:49 Join Omlet [0] (
20:18:14*gevaerts still suspects filesystem corruption, unless someone else with a h1x0 sees the same
20:18:39 Join pabouk_ [0] (
20:18:50pabouk_linuxstb: The latest daily build is the last known working one (r24556-100207). Firstly I discovered the problem on r24568-100208 then I tested it on the current build r24571-000208. No other build appeared since reporting the problem.
20:20:27linuxstbpabouk_: The latest daily is also r24751. But something seems odd, at least with the date of that build (000208)
20:20:59 Quit pabouk (Ping timeout: 256 seconds)
20:21:06 Nick pabouk_ is now known as pabouk (
20:21:13gevaertslinuxstb: that date isn't the issue. It was caused by a wrong clock on the build client, and r24568 was built on a different one
20:22:26linuxstbgevaerts: OK, I didn't say it was related, just that that build was odd...
20:22:29pabouklinuxstb: The latest build on is r24571-000208. I downloaded it a while ago. I hope it could not be cached somewhere...
20:22:34 Join amiconn [0] (quassel@rockbox/developer/amiconn)
20:22:34 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
20:22:52 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
20:23:31JdGordon_Stephen__: you are sure the statusbar option is not set to custom?
20:23:47Stephen__i changed to top bottom off and cstom all same effect
20:24:25JdGordon_can you attache or pastebin nyour config.cfg?
20:24:35Stephen__yep two tics
20:28:46 Join pabouk_ [0] (
20:30:09 Quit pabouk (Ping timeout: 248 seconds)
20:30:14 Nick pabouk_ is now known as pabouk (
20:32:21bug2000Ok. After reading it a couple of times.
20:32:33bug2000Is there a quick way to know whenever a font supports Hebrew or not without going back and forth?
20:32:44 Join pabouk_ [0] (
20:32:52bug2000Also the Unifont [as much as it is nice] is too big. Size of 16...
20:33:26bug2000gevaerts: Thanks :)
20:33:51bug2000gevaerts: What are the grey's and blacks?
20:34:06gevaertsincomplete coverage
20:34:35 Quit pabouk (Ping timeout: 256 seconds)
20:34:41 Nick pabouk_ is now known as pabouk (
20:35:40bug2000gevaerts: I could ask how can it be incomplete coverage if there are only 22+5 letters in the Hebrew Alphabet.
20:36:06gevaertsI don't know. Maybe archaic letters?
20:36:17 Quit Farthen (Ping timeout: 256 seconds)
20:36:20gevaertsHave a look at the verbose stats for details
20:36:56bug2000No such thing as archaic letters. All were used back then. All are used now.
20:37:24 Join Farthen [0] (
20:37:42gevaertshave a look at the linked unicode chart
20:38:29 Quit Tomis (Ping timeout: 252 seconds)
20:39:41 Quit phanboy4 (Ping timeout: 256 seconds)
20:39:42bug2000gevaerts: Oh. I see what it is.
20:40:04bug2000gevaerts: Support for `Nikod`. On how to pronounce the letters.
20:40:35bug2000And `Teamim`.
20:40:44 Join pabouk_ [0] (
20:40:46bug2000So funny to look at the Hebrew PDF.
20:40:55 Join Tomis [0] (~Tomis@
20:41:01 Quit Tomis (Remote host closed the connection)
20:41:05JdGordon_rockbox doesnt show the nikudim
20:41:14gevaertsIf you only need the letters, I guess any grey font will work
20:41:16JdGordon_nikudot even
20:41:43bug2000gevaerts: Seems like it does :) It's perfect.
20:41:52bug200008-Fixed looks fine :)
20:42:25 Quit pabouk (Ping timeout: 248 seconds)
20:42:26 Nick pabouk_ is now known as pabouk (
20:43:20paboukgevaerts: I overlooked the question about filesystem, sorry. Yes I checked it.
20:43:31linuxstbpabouk: Can you try"> ?
20:43:32bug2000JdGordon_: It is Nikudim, I guess. Though Nikud refers to both plural and singular.
20:44:03Unhelpful14-rockbox-mix is, presumably, glyphs from several source fonts?
20:46:11kugelUnhelpful: you stopped working on subpixel aa fonts, right?
20:46:41 Join pabouk_ [0] (
20:47:06Unhelpfulkugel: never really got started... there are some nasty troubles in the font cache if we want to support subpixel. i basically got hung up on things i couldn't decide about instead of implementing what i knew :)
20:47:24linuxstbpabouk_: Can you try"> ?
20:47:34kugelthe other implementation isn't out yet is it?
20:47:35pabouk_linuxstb: Your booted flawlessly. Thank you for your work!
20:48:37 Quit Omlet (Quit: ( :: NoNameScript 4.22 :: ))
20:48:48Unhelpfulkugel: "out"? "other"?
20:49:04linuxstbpabouk_: OK, can you now try"> ?
20:49:18 Join evilnick [0] (
20:49:21 Quit pabouk (Ping timeout: 248 seconds)
20:49:23 Nick pabouk_ is now known as pabouk (
20:49:27kugelI mean, are we considering to get the aa font patch on the tracker?
20:50:12 Nick rhodan_ is now known as rhodan (
20:50:22bug2000gevaerts: I just checked. Those incomplete for fixed are 27 letters -> In other words only the alphabet without the Nikud / Ta`amim. I'm fine with it. Really :0
20:50:33*kugel would like to see it in if it the drawback for non-aa fonts is next to 0
20:50:42Unhelpfuli don't know... i thought for a start i might implement mono font support in convttf and at least merge that (as it's useful *now)
20:50:57Unhelpfulkugel: it'll never be quite 0 as there's a binsize bump
20:51:20kugelyes, I mean except that of course
20:52:03bug2000kugel: Your nick is Yummy.
20:54:25 Quit evilnick (Ping timeout: 256 seconds)
20:54:32 Join pabouk_ [0] (
20:56:07linuxstbpabouk_: Did you see my second link?
20:56:31Unhelpfulkugel: also there will be a call-via-pointer, or else an extra branch, per character drawn
20:56:41 Quit pabouk (Ping timeout: 240 seconds)
20:56:44 Nick pabouk_ is now known as pabouk (
20:56:47pabouklinuxstb: Sorry, my IRC client disconnects all the time. failed: I04:IllInstr at 31057FE4
20:57:07linuxstbpabouk_: OK, can you now try"> ?
20:57:07 Join Bullet` [0] (
20:57:34Unhelpfuli forget which way the code actually went, it's been a while
20:57:46Unhelpfulalso composing characters were done all wrong, unless that's fixed?
20:58:01*kugel hopes that branch doesn't make it unacceptable, as it's rather neglible compared to the rest of drawing in his eyes
20:58:17linuxstbpabouk: Also, you can check the IRC logs here, to see if you missed anything (or if any of your messages didn't get through) -
20:59:27AlexPJdGordon_: The sbs didn't change from showing the "wps stuff" (selected via) %?mp) to showing the "radio stuff" until I paused the radio - after that it was fine
21:00:59AlexPIt didn't seem to notice that the playback state had changed on screen entry, I had to force it
21:02:02AlexPAnd the radio screen is doing the same thing as the WPS was with only updating the backdrop in viewports - I guess the same issue as "we" fixed the other night?
21:02:21 Quit pabouk (Ping timeout: 256 seconds)
21:02:31AlexPOther than that, it seems grand :)
21:02:54 Join pabouk [0] (
21:03:06pabouklinuxstb: works.
21:03:20 Join Strife89 [0] (~michael@
21:03:34linuxstbpabouk: OK, I'll build another - give me a minute...
21:05:04 Join Curulan [0] (
21:05:08linuxstbBut as 24567 is a change to a plugin, r24566 seems the culprit (I'm building that now).
21:05:12 Quit Zarggg_ (Read error: Connection reset by peer)
21:05:46 Join pabouk_ [0] (
21:06:18 Join phanboy4 [0] (
21:07:02kugelmaybe it has something to do with the remote
21:07:44Unhelpfulkugel: *i* don't think the binsize hit is a huge big deal. maybe when i'm satisfied that i'm done w/ this division mess i'll take a look again.
21:08:35 Quit pabouk (Ping timeout: 256 seconds)
21:08:37linuxstbkugel: I would guess so - the h120 has a main backdrop, but no remote backdrop.
21:08:43 Nick pabouk_ is now known as pabouk (
21:08:57kugelis the remote 1bbp?
21:09:14pixelmathe Iriver remote is
21:09:20bertrikjobec, would you mind experimenting a bit with sansa clip power settings, to try to increase runtime?
21:09:36linuxstbpabouk: Next one is">
21:10:57kugelbackdrop_show() should be a stub
21:11:36linuxstbBut what about sb_get_backdrop() ?
21:12:30 Join pabouk_ [0] (
21:12:39pabouk_linuxstb: The file seems to be missing on your server: Object not found! Error 404
21:12:42kugeljust "return sb_skin[screen].data->backdrop;
21:12:42kugel" which should cause harm as is
21:13:59kugelJdGordon: sb_create_from_settings() lacks an else, the main ui vp config string is always taken
21:14:06linuxstbpabouk_: Sorry, fixed now.
21:14:15 Quit pabouk (Ping timeout: 256 seconds)
21:14:28 Nick pabouk_ is now known as pabouk (
21:16:41 Join pabouk_ [0] (
21:18:08pabouk_linuxstb: works.
21:18:29JdGordon_bug2000: yeah, I thought it was nikudim, but then something from the back of my mind made me think nikudot was correct :p
21:18:30linuxstbpabouk_: OK, so that makes r24568 the culprit - I won't bother with 24567, as that was a change to a plugin.
21:18:52AlexPJdGordon_: Hmmm, my fms seems to work in the sim but not on target
21:19:10JdGordon_kugel: oh? can you fix it?
21:19:15kugelnot right now
21:19:21 Quit pabouk (Ping timeout: 256 seconds)
21:19:23 Nick pabouk_ is now known as pabouk (
21:19:33linuxstbkugel: Would that be causing the h120/h140 crash?
21:19:37kugelthe backdrop related functions should check for the screen depth I think
21:19:50kugellinuxstb: no, I don't know what's crashing it
21:20:26 Join GeekShado_ [0] (
21:20:31JdGordon_AlexP: ok, I thought i fixed the backdrop issue in the fms
21:20:47JdGordon_and not taking effect on start is hopefully just a restore=true missing or something
21:20:55JdGordon_how does it not work on target?
21:21:19AlexPI get the default fm screen, but with my sbs (which works just fine)
21:21:38JdGordon_lame :p
21:21:46JdGordon_pastebin the fms?
21:22:12AlexPActually, the not taking effect on start doesn't happen on target, that is fine :)
21:22:45JdGordon_ok cool
21:22:45 Quit GeekShadow (Ping timeout: 256 seconds)
21:23:08 Quit zumbi (Ping timeout: 258 seconds)
21:23:48JdGordon_there is no fm AA yet, so you might need to remove th %C's
21:24:01AlexPI wondered about that, but it worked on the sim :)
21:24:04AlexPI'll try without
21:24:58 Join stripwax [0] (
21:26:40AlexPJdGordon_: OK, it was that indeed - funny how the sim didn't mind :)
21:27:01AlexPJdGordon_: OK, seems great bar the backdrop :)
21:27:24JdGordon_can you remind me in the tracker to make %C not break on target?
21:29:05piotrekmshall i save the backdrop images as X1R5G5B5 or R5G6B5 16bit image?
21:29:39 Quit Curulan (Read error: Connection reset by peer)
21:30:16 Quit Strife89 (Quit: Reboot.)
21:31:15 Join Zarggg [0] (
21:31:17 Join Tomis [0] (~Tomis@
21:33:00 Join _zic [0] (
21:34:23 Join zumbi [0] (~zumbi@
21:34:41 Join pabouk_ [0] (
21:35:27JdGordon_AlexP: and the yucky question, whats the skin buffer size (rockbox info screen)
21:36:18AlexPJdGordon_: 508 KB / 758 KB
21:36:21 Quit pabouk (Ping timeout: 256 seconds)
21:36:23 Nick pabouk_ is now known as pabouk (
21:36:52jobecbertrik: sure
21:37:10bertrikjobec, are you able to build rockbox from source?
21:37:35JdGordon_yeah, that buffer really balloons on those big screen targets
21:38:05jobecnope. or at least i haven't set the build environment up
21:38:13AlexPJdGordon_: If you reuse backdrops are they loaded twice?
21:38:22JdGordon_not in svn
21:38:24AlexPAs I use e.g. the same backdrop for wps and fms
21:38:34JdGordon_thats something I need to fix
21:38:41AlexPah, OK :)
21:38:42kugelfms AA almost requires disk spinning up on entering the screen as far as I see it
21:38:45JdGordon_but its boring work so I hacvnt :p
21:38:58JdGordon_kugel: yes, so?
21:39:57*kugel remembers that was being 1 argument against the fms backdrop (back when we thought we need to merge it into the wps backdrop buffer)
21:41:06bertrikjobec, can you place this file in the .rockbox folder of your clip?
21:41:29bertrikit's a firmware file modified to show some extra power settings in the debug menu
21:41:38JdGordon_kugel: thats not an argument against fms AA
21:42:20kugelwe should be able to bufopen each AA still within the fms, the only difference to the wps is that ring buffer doesn't move
21:42:41 Join pabouk_ [0] (
21:43:51kugelhas anyone thought of how to obtain the filename?
21:44:17 Quit pabouk (Ping timeout: 256 seconds)
21:44:23 Nick pabouk_ is now known as pabouk (
21:45:08jobecbertrik: it's there. which settings?
21:45:52bertriksystem/debug/view hw info
21:46:25bertrikthe bottom line on the first screen of debug info shows SYSTEM: XX CVDD: YY, I want to know XX and YY
21:46:37JdGordon_kugel: no audio buffer need to be initialised on fms entry for that to work, which might not work
21:46:52JdGordon_as long as we dont try merging fm and rec screens thats should actually be ok
21:46:53kugelwhy should that not work?
21:47:00 Part akur
21:47:13kugelaudio and buffering is initialized in main
21:47:20jobecsystem: 29 cvdd: 04
21:47:31JdGordon_it might just be easier to manually load the images
21:47:59 Join pyro_maniac_ [0] (
21:48:07kugelbut that sucks when changing the channels
21:48:11JdGordon_and load up as many as will fit so you can channel surf without hitting the disk
21:48:28bertrikjobec, thanks, it's the same as on my clip, I was hoping for a set of different numbers as that might explain something
21:49:05 Quit _zic (Quit: Ex-Chat)
21:49:13 Part pabouk
21:50:15kugelbuffering has the advantage of handling the scaling for us, also the skin engine uses the playback_aa_slot mechanism which is the layer between album art and buffering
21:50:28 Quit Zarggg (Quit: Zarggg)
21:50:31 Quit Farthen (Ping timeout: 256 seconds)
21:50:45kugelI guess we want to use that to easily switch the sbs to show the fms AA too
21:51:34jobecbertrik: ok. at least not very big thing to check
21:53:53JdGordon_kugel: ah, yeah, but can it be asked to arbitrarily load AA images? even without audio playing? or without audio tracks int heh buffer?
21:54:30kugelI don't see why not
21:56:27JdGordon_ok cool then
21:56:28 Quit Tomis (Quit: Tomis)
21:56:54kugelbasically we would just call bufopen directly without going through playback.c, but playback.c has the aa_slot handling so it's not that easy
21:57:09kugelbut I'm not foreseeing a big deal
21:59:02 Quit vegtoruci (Ping timeout: 265 seconds)
22:01:17 Join robin0800 [0] (
22:03:05***Saving seen data "./dancer.seen"
22:03:22 Join Farthen [0] (
22:03:29 Quit komputes (Ping timeout: 252 seconds)
22:04:19 Quit bertrik (Read error: No route to host)
22:04:41 Join bertrik [0] (
22:04:45 Quit robin0800 (Client Quit)
22:05:12 Join robin0800 [0] (
22:05:33 Quit Horscht (Quit: Verlassend)
22:08:40 Quit robin0800 (Client Quit)
22:09:09 Join robin0800 [0] (
22:12:38 Quit robin0800 (Client Quit)
22:13:05 Join robin0800 [0] (
22:16:00JdGordon_have we got an easy way to see all the avilable strings int he english.lang file without all the extra info?
22:16:11JdGordon_for people looking to make their skins translatable
22:16:53 Quit darkham (Read error: Connection reset by peer)
22:18:34kugelgrep :)
22:20:44 Join komputes [0] (~komputes@ubuntu/member/komputes)
22:26:28 Quit rhodan (Remote host closed the connection)
22:27:15 Quit phanboy4 (Ping timeout: 245 seconds)
22:28:16 Quit domonoky (Read error: Connection reset by peer)
22:32:33 Join phanboy4 [0] (
22:34:08bertrik<build>/lang/lang_core.c ?
22:35:31JdGordon_maybe thats something we stick on
22:38:59 Quit Lss (Read error: Connection reset by peer)
22:47:21 Quit togetic (Ping timeout: 246 seconds)
22:49:32 Join togetic [0] (~togetic@unaffiliated/ibuffy)
22:59:10 Quit robin0800 (Ping timeout: 240 seconds)
23:00:36 Quit Farthen (Disconnected by services)
23:00:41 Join robin0800 [0] (
23:00:58 Join Farthen_ [0] (
23:01:00 Nick Farthen_ is now known as Farthen (
23:02:04kugelslow stream? :(
23:03:23kugelah, the one is better
23:10:16 Quit GeekShado_ (Quit: The cake is a lie !)
23:13:21rasherI don't know if that actually works. This is my first attempt at creating a torrent
23:15:03 Quit Grahack (Quit: Tu m'as vu ?)
23:16:45 Quit togetic (Ping timeout: 246 seconds)
23:18:47rasherSeems to be working.
23:19:16 Quit kugel (Quit: exit(0);)
23:21:51pamauryhehehe, I've completely rewritten the dircache generation. Result: -591 bytes, -30kib ram :)
23:22:19pamaurylooks promissing
23:24:21 Quit robin0800 (Remote host closed the connection)
23:25:11 Join krazykit` [0] (
23:26:08 Quit krazykit (Ping timeout: 276 seconds)
23:29:25CIA-88New commit by 03FlynDice (r24573): Sansa Clip+: Unset pin B0 correctly in dualboot.S
23:30:43 Join togetic [0] (~togetic@unaffiliated/ibuffy)
23:30:54pamauryI think that with some work, it is possible to get completely rid of MAX_PATH in dircache and diruncache as well as in file.c, that is nearly in the whole core (firmware/*). I have a prototype for this.
23:33:33 Join froggyman [0] (
23:34:02JdGordon_pamaury: nice work. removing MAX_PATH could cause problems in apps though, what happens when really long filenames are asked for but are truncated?
23:34:34 Quit evilnick_B (Quit: Page closed)
23:37:09 Quit dmb_ (Read error: Operation timed out)
23:38:28pamauryJdGordon: removing MAX_PATH from apps seems hard for now but I think that the core can be made mostly independent of MAX_PATH. That is, it can open file/dir with arbitrary long filenames.
23:39:54pamauryBut that's not the priority. My priority is to be sure my new dircache generation is correct and to commit it. I would like that Slasheri review it also but I've not seen him for several days now
23:43:30 Quit phanboy4 (Ping timeout: 245 seconds)
23:46:49pamauryCall to tester: could you try "FS #10991 - Rewrite dircache generation to free ram and stack " please ?
23:46:50 Quit dfkt (Read error: Connection reset by peer)
23:47:06 Join dfkt [0] (dfkt@unaffiliated/dfkt)
23:47:15 Quit piotrekm (Ping timeout: 245 seconds)
23:50:03 Quit Bullet` (Quit: Bye les gens =))
23:51:53 Part pyro_maniac_ (" - Chat comfortably. Anywhere.")
23:55:13 Quit flydutch (Quit: /* empty */)
23:58:40CIA-88New commit by 03stripwax (r24574): Use the same trig table as fft (which is already in iram where appropriate) for mdct post twiddles (was already doing for pre twiddles). Also fix an ...

Previous day | Next day