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

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

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

#rockbox log for 2009-12-18

00:00:17gevaertsbertrik: it's not actually my patch. It's an ugly workaround...
00:02:19 Quit petur ("Zzzzz")
00:02:43bluebrotherAlexP: what's that \\* in your last commit?
00:03:02AlexPbluebrother: Forces a blank line
00:03:14JdGordon|what do we do about 5877? thats actually been closed and reopened once (abotu a year ago).. I see no point keeping it open as a bug
00:03:16AlexPlike \\ but with a blank line, not just new line
00:03:21bluebrotherAlexP: but why the *?
00:03:32bluebrotherah. Is that a good idea?
00:03:40AlexPIt looks much nicer IMO
00:03:53qwertzHiilsum: The last version was installed in December 2007 refereing to the directrories
00:04:02AlexPWhen everything runs together it is difficult to e.g. separate the note from the text
00:04:03gevaertsAlexP: rockbox isn't about eye candy! ;)
00:04:25bluebrotherwell, in general the text should not care about the visuals, that should be done by the style. If everyone likes something different better we'll soon end up with something inconsistent.
00:04:52bluebrotherso if we want a \note that is separated from the text we should have a macro for that. IMO :)
00:04:59 Quit fyre^OS ("lamers envy me like they envy bill g -- main boot xp, just the way it should be!")
00:04:59AlexPbluebrother: New paragraphs aren't something we can do with that though?
00:05:42AlexPSometimes you just want a new line, sometimes (e.g. for a new paragraph) you need a blank line
00:05:55AlexPI agree that for e.g. notes though that it could be incorporated
00:05:59bluebrotherAlexP: well, there's a difference between a new paragraph (that \n\n or \par) and forcing newlines with \\ or \newline. The first starts a new paragraph while the second only breaks the line.
00:06:27AlexPbluebrother: ah, does \par give a blank line for the new paragraph?
00:06:30bluebrotherthere are a few exceptions where it does make sense to specify it manually, but I've got the impression that it's overused.
00:06:34***Saving seen data "./dancer.seen"
00:06:45bluebrother\par starts a new paragraph. It's the same as writing two empty lines
00:07:09*AlexP gets confused with what latex eats and what it doesn't
00:07:37AlexPSo blah blah \par rrrr gives me blah blah, then a blank line, then rrr ?
00:08:10bluebrothera new paragraph does not necessarily mean some visual space between paragraphs. That's part of the style
00:08:21AlexPIt does in my world :)
00:08:32AlexPI find it next to useless otherwise
00:08:48 Quit pamaury ("exit(*(int *)0 / 0);")
00:09:22bertrikI see we have a meizuiaudiom3.h config file now, AFAIK the meizu m3 and the iaudio m3 are completely different players
00:09:24JdGordon|AlexP: 7295 is another bookmarking limitation bug
00:09:29bluebrotherLaTeX defaults to not having additional space between paragraphs but indents the first line. You can change that of course, and technical documentation tends to not indent paragraphs but use a visual space between paragraphs instead
00:09:48 Join Jonathan__ [0] (
00:09:59 Join fyrestorm [0] (
00:10:05AlexPbluebrother: I'm indifferent to the indent, but the blank line is essential IMO
00:10:27AlexPOtherwise you just end up staring at a massive block of text
00:10:53bluebrotherwell, it's a matter of taste and style. As I said, technical documentation tends to use more space between paragraphs. Literature tends to use the indent but no additional space.
00:11:15HillshumAny way to see the timestamp of comments?
00:11:25HillshumIn FS?
00:11:38JdGordon|the timestamp is shown
00:11:42JdGordon|or click history
00:11:42AlexPYes, choice indeed
00:12:02bluebrotherbut as of now we're terribly inconsistent. Check the plugin chapter for example: there's a huuuuge space after the first paragraph, but none after the first paragraph of the blackjack section (looking at mini manual)
00:12:24bertrikWhich player is the "ipod 5g" again? (I can't find it in the config directory in the sources)
00:12:31HillshumJdGordon: Not for me
00:12:35AlexPbluebrother: My vote is for a space, and we change the style to do that then prune the manual ones
00:12:36bluebrotherif we want space after paragraphs we can have them. But by changing the style, not by adding them as hard line breaks.
00:12:39 Nick Ypsy is now known as YPSY (
00:13:43Jonathan__I am trying to install Rockbox to my sansa Fuze, but it's not showing up in the "select your audio player" or the autodetect. Vista isn't showing me the drive letter of my player. am I going to have to install it manually?
00:13:44gevaertsbertrik: ipod video
00:13:54bluebrotherurgh. There's quite a lot of those hard "spaces" :(
00:14:02bluebrotherthis looks *ugly*
00:14:05*bluebrother screams
00:14:21*Jonathan__ is afraid he has to install it manually
00:14:25AlexPMostly my fault I suspect
00:14:34HillshumCan I search for bugs on a particular target?
00:15:00gevaertsJonathan__: you have to set it to MSC
00:15:01bluebrotherJonathan__: don't know about the fuze, but you need it to show up as drive letter in explorer. Check if you can switch it to MSC mode, it's likely to be in MTP mode
00:15:20Jonathan__ok, thanks, trying that now
00:15:30Jonathan__that did it, thanks
00:15:46Jonathan__now I just give it that letter?
00:15:54*bluebrother reminds the manual people of the fact that LaTeX does mean separating style from content which means not putting style into content
00:15:57JdGordon|254 :)
00:16:00 Join Strife89 [0] (
00:16:10bluebrotherJonathan__: windows is responsible for handing out the drive letter ...
00:16:31bluebrotherbut yeah, if autodetection doesn't work enter the correct one in the configuration dialog
00:16:36Jonathan__yah, what I mean is I just give the configuration the drive letter that windows is giving me
00:16:45 Quit bertrik ("De groeten")
00:16:47bluebrotherah. Sorry, misunderstood you
00:16:55saratogastripwax: no please commit
00:17:07saratogathe idea behind the branch was to give us a place to experiment anyway
00:17:42saratogastripwax: if you're looking at the mdct.c unrolling, you may want to look at the current ffmpeg version
00:17:45 Join KBH [0] (
00:17:59saratogathey've refactored it a bit, and their changes may be worth having
00:18:31JdGordon|oh crud... we can't move a bug to a feature request :(
00:18:39AlexPbluebrother: Sorry about that
00:19:01bluebrotherAlexP: I don't have a strong favour for or against space after paragraph, but someone should clean it up.
00:19:13bluebrotherAlexP: no problem :)
00:19:23bluebrotherat least we _can_ fix it ;-)
00:19:38AlexPbluebrother: I do have a strong preference, and will try and fix/tidy it in the next few days
00:19:43AlexP+ :)
00:20:35 Quit PaulJam_ (Read error: 104 (Connection reset by peer))
00:20:55*stripwax crosses his fingers
00:21:05 Quit KBH (Client Quit)
00:21:38CIA-6New commit by stripwax (r24058): Unifying various tables and functions across several wma+mdct+fft-mpeg files in FasterMDCT branch. Apply the equivalent patch to r17783 into this ...
00:22:01stripwaxsaratoga - commits 'on the branch' just work, right ....
00:22:28 Quit YPSY ("Oh noes! My ZNC obviously just shut down :(")
00:22:36kugelstripwax, saratoga: remember to merge trunk regularly, that makes merging back later easier
00:22:51AlexPbluebrother: cheers
00:23:14bluebrotherhmm, already half a year ago ;-)
00:23:39 Quit toffe82 (Read error: 104 (Connection reset by peer))
00:24:16AlexPDo we want the major changes in the manual to reflect the wiki page? e.g. in 3.1, 3.2, 3.3 etc
00:24:46qwertzbluebrother linuxstb : Thank you for you r help! Gone sleeping! Cheers! bye
00:25:18JdGordon|why do we bother with major changes in the manual at all?
00:25:29JdGordon|its just one more dynamic page to keep updated
00:25:47AlexPI tend to agree
00:25:55AlexPBut it isn't too bad from time to time
00:26:01AlexPi.e. along with a release
00:26:11JdGordon|the release notes has it, and the wiki has it
00:26:12Jonathan__I'm trying to install RB to my sansa fuze V1. I'm getting an error original firmware unknown please try an other version. tested c200 versions are: 3.02.05. Could not load C:/Users/*myuseraccount*/Desktop/fuzeA.bin. what's going on?
00:26:19 Quit kugel (Read error: 60 (Operation timed out))
00:26:19JdGordon|the manual shouldnt imo
00:26:38AlexPI'm not too bothered - what do people think?
00:26:43 Quit captain_kewl ("Page closed")
00:26:45AlexPEither I remove it or update it
00:26:47saratogastripwax: yes just commit to your checked out branch and svn will do the right thing
00:27:02stripwaxsaratoga - good! :-) (already done ... )
00:27:04 Quit qwertz ("CGI:IRC")
00:27:04saratogakugel: not a seriouis issue, the codecs are essentially unchanged for months or even years
00:27:23stripwaxcodeclib could conceivably change though in unrelated ways ..
00:27:24 Join kugel [0] (n=kugel@rockbox/developer/kugel)
00:27:31stripwaxor codec makefiles etc
00:28:48saratogastripwax: what do you think about the init for the new mdct?
00:29:11Jonathan__I'm trying to install RB to my sansa fuze V1. I'm getting an error original firmware unknown please try an other version. tested c200 versions are: 3.02.05. Could not load C:/Users/*myuseraccount*/Desktop/fuzeA.bin. what's going on?
00:29:14saratogai was thinking of maybe just having it take a void * to a buffer of suitable size instead of exporting the mdct.h header
00:29:17stripwaxit will eventually no longer be needed, is my guess, when all tables are static/fixed ..
00:29:36saratogaJonathan__: its a bug in the current installer giving you a silly message
00:29:43stripwaxoh you mean the MDCTContext stuff? yeah, maybe
00:29:49saratogayou need to use an older fuze firmware (any will work but the latest)
00:29:56*JdGordon| is moving various bugs to the manual when its known limitations with nothing that can really be done about them
00:30:01AlexPbluebrother: What do you think? I agree with JdGordon| that the changelog doesn't really belong in the manual
00:30:07Jonathan__oh... hmmm... where do I get an older one?
00:30:25stripwaxI actually did start with that idea (you'll see public/private interface in one or two of the headers) but gave up because it wasn't the interesting part of what I was looking into :)
00:30:31gevaertsJdGordon|: as long as you don't confuse limitations and bugs too much, that's a good idea :)
00:30:39saratogayeah we can certainly worry about this later
00:30:48saratogafor now vorbis and wma work which is good enough for testing
00:31:04Dhraakellianjust got an SD bank 1 error : 0x10
00:31:15Jonathan__I have V01.01.11A do I need to update my fuze before I do this as well?
00:31:16DhraakellianSansa Fuze, r24036
00:31:16saratogaare you interested in merging the trig stuff you figured out today?
00:31:29Dhraakellian8GB internal, 8GB class 2 SanDisk µSDHC
00:31:34saratogaupdate your rockbox
00:31:44stripwaxone concern about the r17783 diff you posted earlier ( - seems that now ff_mdct_init will calculate the fsincos tabs (for now, until they're flattened+made constant)
00:31:48Dhraakelliandon't have time to do any troubleshooting right now. just saying
00:32:11saratogayeah that may not be correct
00:32:22stripwaxBut I think we call ff_mdct_init for each block. We should do something similar to what I added to mdct.c, to ensure we only initialise each table once. Otherwise the performance is going to suck real bad.
00:32:45 Join Ypsy [0] (
00:32:47 Nick Ypsy is now known as YPSY (
00:33:01stripwaxI'm not going to look into the trig stuff. I'm still sick unfortunately. Got almost enough brainpower to build rockbox
00:33:20saratogaok i'll probably look at it after christmas then
00:33:26saratogaunless mt beats me to it
00:33:26stripwaxor alternatively only call ff_mdct_init when blocksize changes / new stream / etc
00:33:34stripwaxsaratoga - ok! have a good holiday btw.
00:34:02saratogano each call to ff_mdct_init should init one trig table
00:34:16saratogawhich should happen whenever a codec calls its own init functions
00:34:20Jonathan__Where am I supposed to get a sansa firmware that's not the newest one?
00:34:21stripwaxsaratoga - yeah and we should only call ff_mdct_init once 'globally', not once for every block we decode
00:34:45stripwaxMaybe I'm wrong.
00:34:55saratogaisn't that what we do?
00:34:57stripwaxLooks like wma does indeed init once in its own init fn
00:34:59JdGordon|9712 seems to be a simple fix... amiconn, do you tihnk its just a messed up config #define?
00:35:00stripwaxMaybe I screwed up vorbis
00:35:03saratogaalso i guess mdct_init_global is useless now isn't it?
00:35:43stripwaxsaratoga- kinda.
00:36:03stripwaxsaratoga- I think I'm totally misremembering the fft init thing. vorbis only inits in decoder init too.
00:36:36stripwax(_vds_init in block.c). So at least I think I managed to code that ok, phew
00:36:39 Quit Lss (Read error: 104 (Connection reset by peer))
00:36:42AlexPAnyone with manual/changelog thoughts? I'd like to commit this and go to bed :)
00:36:52bluebrotherAlexP: I agree. It might be interesting to have it in the manual, but it's almost never updated and pointless as the changelog is in the wiki and release notes
00:37:00saratogaoh its not
00:37:10AlexPbluebrother: Good enough for me :)
00:37:11saratogawhy did you put the call to ff_fft_init in ff_imdct_calc?
00:37:12stripwaxultimately, I think, we won't need to init *anything*. it should just be an arg to the thing we're calling and there should be no dynamic tables
00:37:23 Join Lss [0] (
00:37:53stripwaxsaratoga - I didn't. At least I'm pretty sure I didn't.
00:38:07bluebrothermaybe we could consider appending the release notes to the *release* manuals. Though I'm not sure if that's too useful. But at least for svn manuals the changelog doesn't make too much sense.
00:38:13stripwaxbut that's ff_fft_init. we were talking about ff_mdct_init
00:38:18stripwax(or at least I was :)
00:38:22AlexPbluebrother: yes, I agree
00:39:28stripwaxsaratoga - you put it there, on initial checkin, as far as I can tell ...
00:40:08stripwaxshouldn't ff_fft_init be called from ff_mdct_init .. ??
00:40:12 Quit saratoga (Ping timeout: 180 seconds)
00:41:04CIA-6New commit by alex (r24059): Update the NODO in the manual from the wiki and remove the changelog from the manual. Closes FS #9693.
00:41:12stripwaxOh. right, so I changed the type to be void*. Like we were just talking about... but I didn't put it there..
00:45:01 Join kugel_ [0] (
00:45:13 Quit kugel (Nick collision from services.)
00:45:17 Nick kugel_ is now known as kugel (
00:45:19kugelsaratoga: svn doesn't handle merges very well, I imagine it can fail horribly if you merge back very old code (with non-codecs code being old)
00:45:32kugelbut well, if you don't want to avoid problems...
00:45:35kugelJdGordon|: why do you edit the very first comment instead of adding one?
00:45:43stripwaxkugel - could we just 'not' merge the things we haven't changed?
00:46:01JdGordon|kugel: because I'm lazy
00:46:10JdGordon|same answer for everything!
00:46:20kugelJdGordon|: adding a comment is less work
00:46:40JdGordon|not when I edit the task at thesame time
00:46:57 Join saratoga [0] (i=9803c6dd@gateway/web/freenode/x-wrmvokpnbleconnh)
00:47:06stripwaxsaratoga - (for the logs) - naively/dumbly putting the fft init into ff_mdct_init doesn't seem to work (nasty white noise interspersed into output on vorbis), guessing it keeps some state around between calls that it probably shouldn't
00:47:15saratogayeah i just tried
00:47:21saratoga(was disconnected without realizing it)
00:47:23kugelstripwax: I'm not sure if svn can do that
00:47:28stripwaxoh you're back! :)
00:47:39Jonathan__Original firmware unknow, please try an other viersion. Tested c200 versions are : 3.02.05 Could not load C:/Users/*myusername*/Desktop/fuzeA.bin why am I getting this error and what can I do about it?
00:47:39stripwaxkugel - oh. I was guessing you could put the filenames on the merge commandline
00:47:41kugelsaratoga: same thing just happened to me..
00:47:52Jonathan__I can't find any older versions of firmware anywhere else
00:48:16stripwaxsaratoga - so if ff_fft_init is intensive (and I didn't really look at it at all), fixing that could be a big win.
00:48:28saratogastripwax: its extremely slow
00:48:43saratogait computes all the trig tables ...
00:48:56 Quit bzed ("üüüüääääääääöööö!")
00:49:06stripwaxah right, but only once (now)
00:49:15stripwaxthe multiple calls should be fast/free/cheap
00:49:17saratogai'm confused why it matters how its called
00:49:20stripwax(that was the addition I made)
00:49:34saratogastripwax: not it will recompute them everytime
00:49:37stripwaxthe if(NULL==tab) stuff. to prevent it being slow
00:49:37saratogano rather
00:49:59stripwaxsaratoga - it will? why?
00:50:09saratogahmm looks like mt added that
00:50:09stripwax(did you see my changes?)
00:50:11 Join bzed [0] (
00:50:13saratogait wasn't there the last time i looked
00:51:01stripwaxIt will *definitely* recompute revtab though. but not the trig tables. because the very first call sets them all up now. after that, the tabstabstabs tables point to the initialised data that gets reused
00:51:13stripwaxtabstabstabs is *obviously* not a good variable name,... and I apologise ...
00:52:16CIA-6New commit by alex (r24060): Document that bookmarks don't work with dynamic playlists - closes FS #7295.
00:53:10stripwaxthere must be a better/faster way to do the split radix indexation rather than using an array of size(4096).
00:54:02stripwaxOh, that's why we need to call init each time. the revtab CHANGES for different sized blocks. blarrrgggghhhh. if anything I'm sure we could reuse one giant revtab for all block sizes.
00:55:20 Quit kugel (Nick collision from services.)
00:55:24stripwaxwould be great if you could explain (or comment :) exactly what split_radix_permutation is doing. I get the idea, but I *cannot8 read that code .. do you know?
00:55:26 Join kugel [0] (
00:55:28stripwax^ *cannot*
00:55:40saratogastripwax: its just a regular fft permutation
00:55:55saratogabetween different stages of the split radix decomposition, the samples have to be reordered
00:56:06 Join dmb [0] (n=Dmb@unaffiliated/dmb)
00:56:31Jonathan__I just finished installing the bootloader for my sansa fuze, do I unplug my device and let it reboot before plugging it in, or do I simply install rockbox?
00:56:33saratogaor maybe its before
00:56:35saratogai can't remember
00:56:39stripwaxyeah but *what* is that code doing? Initialising a 4096-element array, recursively? blargh.
00:56:55stripwaxeach time we decode a block, I might add.
00:57:02gevaertsJonathan__: just install right away
00:57:10Jonathan__ok, thanks gevaerts
00:58:00saratogathe fft is basically a recursive process, so the init function probably just mirrors the recursion used in the fft
00:58:25stripwaxno, I realise that. but this bit of code is the reason we need to re-call init for each block.
00:58:33stripwaxplus, it's icky.
00:58:36Jonathan__gevaerts: I finished the install, now do I just safely remove and let it update, then reboot it?
00:59:22 Quit kugel (Nick collision from services.)
00:59:28 Join kugel [0] (
01:00:11stripwax−− and we only use it once at the prerotation step? we might as well compute the prerotation recursively ...
01:00:16stripwaxand save 4096*4 bytes..
01:00:40saratogayeah actually ffmpeg does that
01:00:47stripwaxHah. Ok, time for me to turn into a pumpkin. good-night!
01:00:56saratogabut i wasn't sure about doing it in place
01:01:05saratogai didn't really look into it
01:01:07saratogaok good night
01:01:23 Quit stripwax (Read error: 104 (Connection reset by peer))
01:01:44 Quit Hillshum (Remote closed the connection)
01:04:23 Join Hillshum [0] (
01:09:37 Join Riku [0] (
01:16:44 Join Lss__ [0] (
01:17:23 Quit Lss (Read error: 60 (Operation timed out))
01:17:31 Join AsaelReiter [0] (
01:18:40linuxstbAlexP: "Fixed" ? (FS #7295 - Bookmark feature doesn't work on user created playlists)
01:19:09AlexPlinuxstb: It had been changed to a manual shortcoming
01:19:20AlexPI was just responding to that
01:19:50AlexPBy JdGordon| I think
01:20:11JdGordon|yeah, I moved it to manual
01:20:36*linuxstb doesn't agree with closing hard bugs
01:21:59 Quit HellDragon ("Ω")
01:24:49JdGordon|bugs are things we can do something about...
01:25:10*JdGordon| back in 20
01:25:14 Quit JdGordon| ("Miranda IM! Smaller, Faster, Easier.")
01:25:42 Quit Jonathan__ ("ChatZilla 0.9.86 [Firefox 3.5.6/20091201220228]")
01:26:07linuxstbJdGordon: That comment makes no sense at all...
01:29:19linuxstbAlexP: What about the sentence in the manual that says "Bookmarks are saved on a per directory basis (for dynamic playlists) or for individual (saved) playlists."
01:29:55 Join HellDragon [0] (
01:30:01AlexPI didn't see that - I guess that it needs changing then
01:31:05AlexPI'll try and do it tomorrow if I can - if not Sunday
01:31:49AlexPIf nobody has first
01:33:31 Part linuxstb ("Leaving")
01:33:38 Quit Riku (Read error: 110 (Connection timed out))
01:34:12 Quit Hillshum (Read error: 60 (Operation timed out))
01:36:02 Quit saratoga (Ping timeout: 180 seconds)
01:36:42 Join shamblesben [0] (
01:36:58shamblesbenhello, is there anyone here who could possibly help me?
01:37:21 Quit Zarggg ()
01:38:18 Join JdGordon1 [0] (
01:38:32gevaertsshamblesben: it's usually a good idea to tell us what exactly you need help with
01:40:11JdGordon1woohoo! the configs were all moved into a new folder :D
01:40:34gevaertsJdGordon1: that was done ages ago!
01:40:39gevaertsor at least days :)
01:40:55shamblesbenok, well listen, basically i manually installed Rockbox onto my iPod Nano 2G, but now i wish to remove it as i am giving it to someone else who just wants the normal software, but i cant remove it!!
01:42:09 Join MethoS- [0] (n=clemens@
01:42:45JdGordon1such is the beuty of using in dvelopment sofgtware
01:43:06gevaertsshamblesben: you should be able to use itunes to restore it
01:43:08 Quit AsaelReiter ("CGI:IRC")
01:43:26shamblesbennope, itunes is not recognising it!
01:43:40gevaertsboot the original firmware first then
01:43:58shamblesbenyes actually that is what i was going to ask, how do i do that?
01:44:40gevaertsI don't know exactly (I don't have a nano), but the manual has that
01:45:33JdGordon1its usually menu or turn it on with hold on
01:45:41shamblesbenok well i couldnt see it, if i remember the nano rockbox firmware wasnt 100% complete
01:46:06shamblesbenive tried to hard reset the ipod o see if it gave me the option but ii doesnt
01:47:28JdGordon1contrast of 0 is pretty much the same as a blank screen right?
01:48:52shamblesbenok, il try thhat
01:49:55 Join Zarggg [0] (
01:54:17shamblesbenit works now! cheers gavaerts!
01:55:09shamblesbenno one know anymore about when Rockbox will be available for the 160gb iPod Classic! Guess its probly been asked a zillion times
01:55:34JdGordon1whats been asked?
01:55:39JdGordon1you didnt ask a question
01:55:48JdGordon1and dont ask it!
01:56:12shamblesbenok...... ??
01:56:41 Join Hillshum [0] (
01:58:14shamblesbenmay i ask why JdGordon1
01:58:32JdGordon1you may
01:58:40 Join HBK [0] (
01:58:51shamblesbenwhy then!
02:02:15shamblesbengreat help u are!
02:05:07 Quit shamblesben ("CGI:IRC")
02:06:38***Saving seen data "./dancer.seen"
02:10:29 Join root [0] (n=chatzill@
02:10:56 Nick root is now known as Guest83074 (n=chatzill@
02:11:03 Join webguest91 [0] (
02:11:13Guest83074Hi all, I want to port Cowon iAudio 9, but I've never done this before, any one could help?
02:11:37Guest83074I've already post a thread here:;topicseen
02:12:09Guest83074and I was informed that I need to go IRC for permissions of the Wiki
02:12:34 Nick Guest83074 is now known as MickeyShine (n=chatzill@
02:12:36 Quit HBK ()
02:12:51froggymanGuest83074: well, for starters it will take a lot of work on your part, as ports are done on a volunteer bases
02:13:35*froggyman doesnt know to give someone wiki permissions
02:14:33MickeyShinefroggyman: Thank you. What're the basic steps to start a new port?
02:15:09froggymanMickeyShine: well, there is already a lot of documention on how to port
02:15:25MickeyShineI only have some high-level programming knowledge (C/C++ etc), afraid it could not be done by only myself
02:15:44MickeyShinefroggyman: could you please give some links/refrences? Thank you
02:16:07froggymanC is considered a low-level language
02:16:22froggymanread through these articles:
02:16:55froggymanthats basically about everything for help with the ports
02:17:31froggymanthe next place you would want to start looking at would be compiling, and just understanding the rockbox source code
02:18:43MickeyShinefoggyman: Thank you. I am getting down to it and I will be back if I have anything to ask
02:20:01 Quit GeekShadow ("The cake is a lie !")
02:23:45 Quit bluebrother (Nick collision from services.)
02:23:47 Join bluebroth3r [0] (n=dom@rockbox/developer/bluebrother)
02:36:02MickeyShineExcuse me I am really new here and I am goint to port Cowon iAudio 9. I am looking at "Build a cross compiler", but I don't know how to "Find your target CPU". All I've got is some PCB pictures of iAudio 9: , is that enough for someone else to find the CPU type? (Sorry I knew little about hardware)
02:37:11 Quit JdGordon1 (Read error: 110 (Connection timed out))
02:43:25 Quit kugel (Read error: 113 (No route to host))
02:52:41 Join Jonathan__ [0] (
02:54:34kkurbjunMickeyShine: You can't tell from those pictures with the stickers over the parts
02:54:57kkurbjunthe chip with the green sticker looks to be the SOC which would tell you the processor
02:56:13kkurbjunif you can get a picture of the chip with the sticker removed it would probably tell you what you have
03:01:25 Quit froggyman (Read error: 110 (Connection timed out))
03:02:40MickeyShinekkurbjun: thank you. I am trying to find some more pictures
03:03:05 Quit webguest91 ("CGI:IRC")
03:03:37 Join funman [0] (n=fun@rockbox/developer/funman)
03:05:17 Join mikroflops [0] (
03:13:51 Quit Jonathan__ ("ChatZilla 0.9.86 [Firefox 3.5.6/20091201220228]")
03:16:03 Join CaptainKewl [0] (
03:43:17 Quit funman ("free(random());")
03:47:24 Join HBK [0] (
03:59:07 Part Strife89 ("Leaving")
04:01:27 Quit DerPapst1 ("Leaving.")
04:06:39***Saving seen data "./dancer.seen"
04:11:13 Join fdinel [0] (
04:25:32 Quit goffa ("leaving")
04:33:43 Join goffa [0] (n=goffa@
04:34:35 Join nandan [0] (n=nandan@
04:37:31CIA-6New commit by kkurbjun (r24061): Accept FS #10568 by Asael Reiter: Brickmania Lasers not working properly.
04:40:35 Part nandan
04:41:56 Join darkham [0] (
04:43:06darkhamhi people, only a question: how works 16gb microsdhc in sansa fuze with rockbox?
04:45:10 Quit HBK ()
04:45:28 Quit bzed (
04:45:28 Quit ps-auxw (
04:45:28 Quit rasher (
04:47:36 Part sdubois
04:50:18NJoinbzed [0] (
04:50:18NJoinps-auxw [0] (
04:50:18NJoinrasher [50] (n=rasher@rockbox/developer/rasher)
04:50:21 Quit ps-auxw (Remote closed the connection)
04:50:29tmztdarkham: do you mean does it work? I can confirm 8gb but not 16 as I don't have one
04:50:42scorchein theory, it should work fine
04:51:25darkhami would like to buy one for christmas
04:52:25darkhambut, if some known hard issue, i'll buy 2 8gb :)
04:52:59tmztsd 2.0 is the same, there's no hardware reason I know of for it not to work
04:53:44 Quit bzed (
04:53:44 Quit rasher (
04:54:13 Join HBK [0] (
04:58:37 Join ps-auxw [0] (
05:02:56NJoinbzed [0] (
05:02:56NJoinrasher [50] (n=rasher@rockbox/developer/rasher)
05:12:07CIA-6New commit by kkurbjun (r24062): Accept a form of FS #10561 by Clément Pit-Claudel: Fullscreen Brickmania for portrait players.
05:12:55 Join steve2 [0] (
05:13:41steve2could someone help me with WPS code? I setup the nano to use 12 hour clock but regardless the play now screen will show 24 hour clock. SO i read the custom wps page −− still confused :|
05:18:38 Quit panni_ ("( :: NoNameScript 3.81 :: )")
05:19:22 Quit steve2 (Read error: 104 (Connection reset by peer))
05:25:57CIA-6New commit by kkurbjun (r24063): Accept FS #10271 by Christian Beier: Automatically disable voice upon rockdoom startup.
05:30:13 Quit HBK ()
05:40:46 Quit MethoS- (Remote closed the connection)
05:56:51 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
06:06:40***Saving seen data "./dancer.seen"
06:07:13 Quit martian67 (Nick collision from services.)
06:07:13 Join lol2 [0] (
06:07:24 Join martian67 [0] (n=arkfar@about/linux/regular/martian67)
06:24:52 Quit killan (Read error: 110 (Connection timed out))
06:38:07 Join killan [0] (
06:49:31pixelmahmmm... I thought we once had a full height brickmania on portrait screens and people found it to easy so it was changed
06:50:21kkurbjunpixelma, the game speed scales with the screen height
06:50:33kkurbjunwell, the vertical speed scales
06:50:46kkurbjunso it should be a comparable difficulty
06:51:04 Join HBK [0] (
06:51:10kkurbjuntry it out and see what you think, if you don't think it works well we can revert it
06:52:02pixelmamaybe I also remember wrong... should get some more coffee first ;)
06:53:31CIA-6New commit by jdgordon (r24064): Fix FS #10433 - "Play Next" should use the "warn on playlist change" setting like the rest of the options that erase the playlist
06:54:38pixelmaJdGordon: which comment bug did you mean?
06:55:05kkurbjunit didn't use to scale, so it's possible/likely that people did find it to be too easy before.
06:55:11JdGordonpixelma: the one with # comments in the skins
06:56:18pixelmaI only remember a discussion whether or not # should eat the linefeed or not
06:56:47JdGordonthats the one
06:57:31pixelmacan't remember the outcome though but I like the current behaviour (eating lf)
06:58:09JdGordonI remember the argument got pretty heated (wel duh :p )
07:00:16Lloreankkurbjun: With fullscreen on portrait targets, are the bricks redrawn to be considerably taller?
07:00:31LloreanThen it changes gameplay considerably
07:00:47LloreanSince its vertical speed relative to brick layout is now considarably different.
07:01:03JdGordonit would be nice if it was an option
07:01:05kkurbjunthe ball crosses the board at the same relative speed, if it takes 1 second for it to cross from top to bottom it still takes 1 second
07:01:21LloreanBut now the bricks make up 5% of the total height instead of 20% (or whatever the actual percentages are)
07:01:29LloreanWhich changes the interaction of the player with them drastically
07:01:48kkurbjunI'm not concerned about minor gameplay changes
07:01:53LloreanIf you want to scale the ball movement, you need to scale the whole field, not just the empty portions.
07:02:04kkurbjunI would not consider it "considerable"
07:02:22Lloreankkurbjun: The whole reason ball speed scaling was introduced was to avoid similar issues.
07:02:27LloreanAnd you've just reintroduced them arbitrarily.
07:02:37kkurbjunI introduced the ball speed scaling
07:02:39LloreanThe fact that you consider them minor doesn't change the fact that it's now a different game than it was before.
07:02:40kkurbjunI think I know why it's there
07:03:14LloreanYes. Ball speed was added in to fix a bug. Then you reintroduced the same bug in a different presentation - the ball speed is once again improper relative to the playing field.
07:03:29kkurbjunthe game was unplayable on large screens
07:03:37kkurbjunand this is hardly the same situation
07:03:42LloreanThe sides of bricks are equally valid targets as the top/bottom, except now they're *significantly* harder to hit
07:03:56LloreanYou've increased difficulty drastically.
07:04:19kkurbjunsignificant and drastic are relative, I would tend to disagree that either statements are true
07:04:25LloreanIf the vertical speed is to be increased, the height of bricks needs to be increased by the same ratio
07:04:30kkurbjunbut it's easy enought ot change the brick sizes
07:04:41kkurbjunand you don't have to code to do it :)
07:05:10LloreanOr I could revert your change until you make a version of the patch that doesn't break gameplay
07:05:28kkurbjunagain, break gameplay is arbitrary too
07:05:34LloreanNo, it really isn't.
07:05:47kkurbjunI thought you were just a "user" now that you quite rockbox
07:05:51LloreanBrickmania works in a field with certain set parameters. You've changed the ratio of one to another - it changes game difficulty
07:06:26Lloreankkurbjun: Seriously. Revert it, or make me do it, or fix it. You really should not change gameplay on plugins when it *creates* gameplay bugs rather than fixes them
07:06:41LloreanYou make it *look* better, but changed the gameplay in the process so that it's no longer the same game as 4:3 targets
07:07:00kkurbjunstop being such a dick, you wonder why the users got so upset at you, bring it up in teh ml if you want to revert it
07:07:12LloreanOr I could just go and revert it.
07:07:20kkurbjunit doesn't create any bugs
07:07:30LloreanIt does.
07:07:35kkurbjungo for it I'll commit it again until you bring it ot an open discussion
07:07:36LloreanYou've changed the gameplay
07:07:41LloreanThat's a bug since it's no longer the same game.
07:07:48kkurbjunit was never the same game
07:07:50Lloreankkurbjun: you didn't bring it to open discussion when you broke it.
07:07:58kkurbjunI'm doing the work on it
07:08:10Lloreankkurbjun: How is it not the same game when previously the fields were the same aspect ratio and the movement was the same relative to the field?
07:08:32pixelmakkurbjun: users can have set opinions too and your discussion style isn't helpfull too (Llorean had arguments, yours was more like "I don't think so" and calling people names isn't nice either)
07:09:06kkurbjunI made it the way that it was, it didn't work like that before at all, so a minor change in the vertical speed is not significant or worth anyone's time arguing it
07:09:27 Join Horschti [0] (n=Horscht2@xbmc/user/horscht)
07:09:37pixelmanot everyone has to agree with your changes just because you do the work
07:10:01kkurbjunit didn't work like that before I did anything to it though
07:10:17Lloreankkurbjun: Changing gameplay mechanics is not minor even if it's a small mechanic change.
07:10:24kkurbjunso I don't see how Llorean's arguement is valid, he's taking a small snapshot in time and saying this is how the game should be forever
07:10:25LloreanDidn't work like *what* before?
07:11:05pixelmaLlorean: introducing a new speed scaling etc. recently
07:11:12LloreanYes, if it's the same game, the gameplay *shouldn't* be changed. It *is* how it should work forever, or at least the original (or most true possible) gameplay mode should be preserved as an option.
07:11:13kkurbjunthe game played significantly different on all the targets before I introduced the proper scaling
07:11:36kkurbjunI changed the gameplay on all of the targets except the H300 why wern't you complaining then?
07:11:49LloreanBecause that made the gameplay more consistent between targets
07:11:55LloreanIt made the rules for each system "the same"
07:12:05JdGordonif anyone cares about my view.... the plugins are not a core part of rockbox and any attention they get is good... which means if someone wants to change something I dont care at all
07:12:47LloreanJdGordon: Would you care if someone took the blackjack plugin and suddenly inversed the card values so that it was really a different game now?
07:13:11JdGordonthats not blackjack then
07:13:15kkurbjunthis is not a different game
07:13:21Lloreankkurbjun: But it *is*
07:13:26Lloreanyou've changed the playing field
07:14:01kkurbjunI'm done discussing this right now, if you want to bring it up in the ml do so −− if you revert it before then I WILL commit it back
07:14:18LloreanI've already brought it up on the mailing list.
07:14:36pixelmacommit wars? Sounds very childish :(
07:14:47*JdGordon wanted Llorean to complain about my commit
07:14:51*JdGordon is feeling left out :'(
07:15:01LloreanJdGordon: Why? Your change makes sense.
07:15:20Lloreankkurbjun: Has removed an existing game and replaced it with a variant he prefers.
07:15:25JdGordonits the excact same change as the last one you whinged about!
07:15:27LloreanLeaving those of us who liked the original unable to play it.
07:15:46LloreanJdGordon: "Play Next" is entirely different from erasing bookmarks.
07:17:23 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
07:17:44pixelmakkurbjun: what a reaction on the ml... this is much more noise because it doesn't add anything to the discussion. If you are angry could you please relax for an hour or so and answer then? :\
07:19:19kkurbjunif paul actually contributed something to the group other than anger and frustration I would be interested in his comments. That's why I asked you about it pixelma. I could give a shit about what he feels given the demeanor he brings to the group.
07:19:48JdGordonwithout takeing sides.... I'm with kkurbjun :D
07:20:58pixelmakkurbjun: well yours mail is full of anger and frustration too
07:21:07pixelmayour too
07:21:53Lloreankkurbjun: I love that. "Because of your attitude, it's impossible for you to be right."
07:22:02LloreanWay to keep an open mind and rise above things.
07:22:32JdGordonhang on... did Llorean just use the c word as an argument?
07:22:59LloreanJdGordon: Games have established rules. So one would assume them to be as consistent as reasonably possible between targets.
07:23:14JdGordonnot at all...
07:23:28 Quit Kitar|st (Read error: 60 (Operation timed out))
07:23:38JdGordonas lopng as each follows the rules of the game, there is nothing saying each target must be consistant with eachother
07:23:38kkurbjunlisten, I don't appreciate the way that you approached the change, I feel like you are making a mountain out of a mole hill as is often the case. It was not something that I did on my own, someone else contributed it initially, and they felt enough about it to create a patch for it, which seems like a perfectly reasonable change
07:23:39LloreanYou wouldn't mind if there were no hearts or diamonds on the mono screens "because they don't look good in black"?
07:23:57LloreanJdGordon: This patch changes the rules of the game.
07:24:03JdGordonno it didnt
07:24:07LloreanJdGordon: Yes, it did.
07:24:07JdGordonit changed the physics a bit
07:24:11LloreanWhich are part of the rules.
07:24:47JdGordonthe rules, as written by breakout 20(?)+ years ago was ball moves between paddle and obstacles, and on impact obstacles change
07:24:58Lloreankkurbjun: I was rather polite about it initially. In fact, I never insulted you along the way. I approached the change by assuming you'd actually done all the work necessary to not change the gameplay, and simply asked for confirmation. When you said no, I told you why I thought it was a negative change.
07:25:26LloreanJdGordon: So why not just make the game play sideways on portrait targets?
07:25:30pixelmakkurbjun: basing your answer to arguments on your opinion about people is much worse in my eyes
07:25:41JdGordonthat [atch was around for 6 months already.. plenty of time to complain
07:25:52JdGordonLlorean: if it made it better on that target then sure
07:25:58LloreanJdGordon: I didn't think anyone would commit it before it was more cmplete.
07:26:13LloreanThis *doesn't* make the game better on target. Certainly not as much as simply rotating the game would.
07:26:13JdGordonyou didnt comment on it at all
07:26:32LloreanI didn't comment on it because i thought "it would change gameplay" was so glaringly obvious I'd get ragged on for being condescending.
07:27:03LloreanIt's literally impossible for me to say anything without getting a complain. Now it's apparently impossible for me *not* to say anything without getting a complaint too.
07:27:30 Quit Horscht (Read error: 110 (Connection timed out))
07:28:20kkurbjunLlorean: you didn't even try the patch before saying that it changed the game "considerably" which is why I have a hard time accepting those words
07:28:43kkurbjunwhich is what pissed me off
07:29:32Lloreankkurbjun: Are you claiming that the brick heights did change now?
07:29:58LloreanWhether it's "considerable" or not isn't the issue - it's a gameplay change you've made arbitrarily.
07:30:09kkurbjunthe game was arbitrary to begin with
07:30:32LloreanSo basically "I don't care about the gameplay, so I feel I'm free to change it however I see fit."
07:30:35kkurbjunpeople have added new powerups that changed the gameplay where were you then?
07:30:40LloreanWould you revert it if I decided the N powerup was no longer useful?
07:30:52kkurbjunshould we remove the small and big powerups too?
07:31:08LloreanIf they don't scale appropriately with screen size, they should be adjusted.
07:31:14kkurbjunthey don't
07:31:28 Quit fyrestorm ("lamers envy me like they envy bill g -- main boot xp, just the way it should be!")
07:31:35kkurbjunthe paddles are not appropriatly scaled either
07:31:41kkurbjunfor different targets
07:32:00LloreanSo what was the point in scaling the ball speed to make it more consistent between targets earlier?
07:32:02kkurbjunso the arguement about a minor gameplay change on portrait screens seems moot
07:32:17LloreanBecause you're taking it *away* from being the same game on all targets, rather than toward.
07:32:31kkurbjunI did it mainly to make the game playable on large/small screens, that was an added benefit, but not a hard "rule"
07:32:36LloreanYou've now added to the ways in which gameplay is different on different targets.
07:32:46LloreanThe game was playable before the speed change.
07:32:48kkurbjunit always has been
07:32:51LloreanIts difficulty was just different.
07:32:54kkurbjunit's playable now
07:33:05JdGordonLlorean: yes or no.. have you tried the new version?
07:33:06LloreanI didn't say it wasn't.
07:33:23Lloreankkurbjun: But you changed the ball speed to improve difficulty correlation between them. This height thing does the opposite.
07:33:26LloreanJdGordon: I have now.
07:33:33LloreanMy opinion is still the same.
07:34:03kkurbjunyeah, because that's where you started and you will carry this arguement till you die or someone gives up
07:34:16LloreanAh yes, because I've *never* changed my opinion before.
07:35:10 Join advcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
07:36:04kkurbjunthe height change is a minor difference, the ball speed is still the same across the board, the fact that the bricks are slightly smaller relative to the 4:3 speed is a small differential
07:36:54Lloreankkurbjun: If it's so minor, why are you so against reverting it?
07:37:04kkurbjunand when you consider all the other differences on other targets I don't see how this is an arguement till the game is all the same
07:37:19JdGordonboring... take it to pm
07:37:23JdGordonnoone else cares
07:37:35kkurbjunbecause you just started an argument on principal
07:37:52Lloreankkurbjun: It introduces bugs too
07:37:55LloreanMy ball is passing through bricks
07:37:56kkurbjunwhat bug?
07:38:02kkurbjunno, that was there before
07:38:07LloreanNever seen it before, at least.
07:38:14kkurbjunwell it was there
07:38:19kkurbjunyou just never noticed it
07:38:26LloreanOr it never happened to me.
07:38:33kkurbjunit's when it hits a side
07:38:36LloreanCould the increased vertical speed make it miss collisions better?
07:38:55kkurbjunno, it's because of the way the ball interacts with the side of the screen
07:39:20kkurbjunthe collision detection is (and always has been) messed up on the side collisions.
07:39:41kkurbjunit's a relatively minor fix, but it's just something that has not been done
07:40:13*Llorean finds the ball also seems to be leaving the paddle at what seem to be strange angles.
07:40:21LloreanI know it's because of the increased vertical speed, but that just _looks_ weird.
07:40:25LloreanOutside of the gameplay difference.
07:41:22kkurbjunthe ball bounces strange on some collisions too, and that is also an artifact of the way the game was written, fixing that would take a completely different collision detection system
07:41:37kkurbjunit usually happens on bounces that would happen on corner bounces
07:41:50kkurbjunand it's becuase of the order that the game checks the bricks
07:41:52LloreanNo, I was just commenting on how what used to be "low angle" bounces are now apparently much higher because of the increased speed
07:42:02LloreanIt's just a visual artifact, it doesn't affect gameplay because the field is also taller
07:42:12LloreanThe only thing that affects gameplay is that the bricks are a much smaller percentage of the field.
07:42:16kkurbjunyes, I understood, that I was just pointing out all the little gotchas
07:42:47LloreanAt the very least they should be lowered somewhat so that they're the same percentage of the field from the bottom as they were before, rather than still being aligned with the top
07:42:53pixelmathe first time I saw the ball going through the bricks was after the speed scaling change, it might have been a mess in the code before but I can't remember it being visual to the user
07:43:40Unhelpfulthat's a fairly easy fix, check for any bricks the ball *might* have intersected between frames..
07:44:12pixelmamaking the bricks taller would just need an edit of the related bitmaps
07:44:15kkurbjunthe ball does not really go through the bricks, it is checking line intersections, and if it checks one line intersection before the other it can cause a bit of a jump, usually it's fast enough or happens infrequently enought hat you don't notice
07:44:36LloreanUnhelpful: I think the paddle at least needs that. I've seen the ball at certain angles clip through the paddle, I think. But it's unrelated to this patch.
07:50:05 Quit advcomp2019_ (Read error: 113 (No route to host))
07:51:38CIA-6New commit by jdgordon (r24065): FS #10855 3 minor sbs bugs, spotted and fixed by Teruaki Kawashima
07:54:16pixelmawell... I think gigabeat screen shares the same bitmaps with H300 and e200 shares bitmaps with the Nano (maybe also the 160x128 screens), so it would need an edit - save under a different name and edit the bitmaps' SOURCES file
07:55:51kkurbjunyes, and I think the gigabeat bitmaps could be more appropriately scaled for the width of the screen at the same time
07:58:17JdGordonhow much do people care about the splash clipping in the database search status splashes?
08:03:54JdGordonis the database enabled on charcell?
08:04:04*JdGordon suspects no?
08:06:44***Saving seen data "./dancer.seen"
08:09:57amiconnJdGordon: You probably meant the 3G in your closing comment...
08:10:14JdGordonwhat did I put?
08:11:11amiconn5G... that definitely handles wheel repeat
08:11:43 Quit lol2 (Client Quit)
08:11:43 Quit martian67 ("xD")
08:11:57JdGordonyeah :/
08:13:01*JdGordon decides to not care about the splash clipping for now
08:13:10 Join martian67 [0] (n=arkfar@about/linux/regular/martian67)
08:16:51topiki was just about to search FS for first issue fixed in FS #10855. testing the fix now
08:18:22topikmhh, why in the world does my fuze start in fm radio now
08:19:28topikthanks for commiting r24065, JdGordon. and teru thanks for the patches
08:23:53JdGordoncan people please try the fix in FS #7423? its a bugfix for playlists
08:24:01JdGordonI need to know if it causes any wierdness
08:26:08*JdGordon quickly updates the patch
08:30:03JdGordon please test....
08:30:08JdGordonits a dangerous fix
08:30:34JdGordondangerous in that the patch is a year old and I dont remember why it worked!
08:52:39 Join hebz0rl [0] (
08:57:36 Quit CaptainKewl ("( :: NoNameScript 4.22 :: )")
09:00:18JdGordonbah, that patch has issues starting a new playlist while audio is running... any other issues would be helpful to know
09:10:18 Join Kitar|st [0] (
09:13:57 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
09:14:18 Join flydutch [0] (
09:19:00 Join Bagder [0] (
09:36:48 Join Zagor [242] (n=bjorn@rockbox/developer/Zagor)
09:41:41 Join petur [50] (n=petur@rockbox/developer/petur)
09:46:47 Join FOAD_ [0] (
09:54:26 Join Casainho [0] (
09:55:45 Join CQ [0] (
09:56:32CQmorning ... quick question: are the radio frequencies on devices limited in teh hardware, or is that a software limitation?
09:58:27kaniinii believe it's a hardware limit
09:59:06GodEateryep, hardware
09:59:36GodEaterhowever, most OFs also limit them further in software to comply with regional rules
10:00:17 Quit martian67 (Read error: 110 (Connection timed out))
10:01:32 Join DerPapst [0] (
10:01:52CIA-6New commit by alle (r24066): Fix some more places (sequel to r24053)
10:02:56 Quit FOAD (Read error: 110 (Connection timed out))
10:02:56 Nick FOAD_ is now known as FOAD (
10:06:47***Saving seen data "./dancer.seen"
10:10:43 Quit shai ("Leaving")
10:15:32 Quit CQ ("ChatZilla 0.9.86 [Firefox 3.5.6/20091201220228]")
10:18:50CIA-6New commit by zagor (r24067): Added date output. Fixed build names.
10:19:46CIA-6New commit by zagor (r24068): Fixes for the new buildnames
10:20:47CIA-6New commit by zagor (r24069): Fixes for the new buildnames
10:21:53 Quit Casainho ("ChatZilla 0.9.86 [Firefox 3.5.3/20091020122601]")
10:22:37Bagdercommit flood!
10:25:13 Join evilnick [0] (i=4d63ba66@rockbox/staff/evilnick)
10:25:17 Part evilnick
10:25:22GodEaterBagder: he should be kicked!
10:25:33 Join evilnick [0] (i=4d63ba66@rockbox/staff/evilnick)
10:30:37tmztCQ: the filters are limited
10:31:00GodEaterdoes anyone know if there's a filename extension convention for windows powershell scripts?
10:31:19tmztnothing in the i2c can really limit them further though and there should be no soft limit outside of OF when using i2c
10:31:35GodEateroops , wrong channel - sorry folks
10:36:17 Join webguest43 [0] (
10:37:55 Quit DerPapst ("Leaving.")
10:39:03 Join pamaury [0] (
10:44:00 Quit pamaury ("exit(*(int *)0 / 0);")
11:01:56 Quit webguest43 ("CGI:IRC (Ping timeout)")
11:05:21 Quit HellDragon (Read error: 110 (Connection timed out))
11:06:47 Join HellDragon [0] (
11:16:11 Join shai [0] (
11:16:38 Quit hebz0rl ("Ex-Chat")
11:24:19 Quit darkham (Connection reset by peer)
11:24:44 Join darkham [0] (
11:26:55 Quit flydutch (Remote closed the connection)
11:28:25 Quit MickeyShine (Read error: 110 (Connection timed out))
11:33:16 Join Omlet [0] (
11:35:19 Join flydutch [0] (
11:57:26 Quit Omlet (Read error: 54 (Connection reset by peer))
11:57:33 Join Omlet [0] (
11:57:56 Quit Omlet (Client Quit)
11:58:01 Join Omlet [0] (
12:06:27CIA-6New commit by mcuelenaere (r24070): Onda VX747/VX777: add ACTION_KBD_ABORT (fixes exiting boomshine)
12:06:49***Saving seen data "./dancer.seen"
12:12:55 Join watto [0] (n=watto@
12:28:35 Quit Sajber^1 (Read error: 104 (Connection reset by peer))
12:39:44 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
12:43:21 Quit Omlet (Read error: 104 (Connection reset by peer))
12:43:21 Join einhirn [0] (
12:47:37GodEaterpondlife: (for the logs) are you going to make it to DevConPoolPub tomorrow?
12:50:48 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
12:59:15 Join funman [0] (n=fun@rockbox/developer/funman)
13:03:37 Join stoffel [0] (
13:06:11 Join stripwax [0] (
13:06:54 Quit BHSPitMonkey (Read error: 113 (No route to host))
13:13:59 Quit BlakeJohnson86 (Read error: 104 (Connection reset by peer))
13:14:23 Join BlakeJohnson86 [0] (
13:14:32CIA-6New commit by alex (r24071): Remove reference to dynamic playlists from bookmarking section; these cannot currently be bookmarked.
13:14:34 Join teru [0] (
13:29:59CIA-6New commit by mcuelenaere (r24072): Onda VX747: add USBPOWER_BUTTON and USBPOWER_BTN_IGNORE
13:31:49CIA-6New commit by mcuelenaere (r24073): Forgot a file
13:41:01 Join Zaba_ [0] (n=zaba@about/goats/billygoat/zaba)
13:41:14CIA-6New commit by mcuelenaere (r24074): Onda VX747/VX777: fix FM radio
13:42:28 Join froggyman [0] (
13:51:17 Quit Zaba (Read error: 110 (Connection timed out))
13:52:05 Nick Zaba_ is now known as Zaba (n=zaba@about/goats/billygoat/zaba)
13:55:45 Join faemir [0] (n=faemir@
14:02:26 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
14:06:23CIA-6New commit by teru (r24075): jpeg,png: Separate code to load file from load_and_show(). ...
14:06:50***Saving seen data "./dancer.seen"
14:11:01 Join pamaury [0] (
14:15:40 Part Bagder
14:20:33 Join fml [0] (
14:20:43 Join wirr [0] (
14:22:34wirrhi all. i'm trying to find some info about power mgmt state on iaudio x5. can someone tell me what runtime can be expected compared to original firmware?
14:25:56evilnickwirr: I expect that would be on the wiki
14:26:54evilnickwirr: but pay close attention to when the tests were done
14:29:40 Quit Horschti ("Verlassend")
14:30:58AlexPGenerally as good as if not better IIRC
14:31:04AlexPnote the IIRC
14:32:50evilnickThe most recent was from August 2008 so there's every chance that your results will be better due to optimisations in the code since then
14:33:47 Quit bzed (
14:33:47 Quit rasher (
14:34:04wirrthat sounds great. so why does it say on wiki/IAudioPort that powerhandling is only 10% implemented?
14:34:15NJoinbzed [0] (
14:34:15NJoinrasher [50] (n=rasher@rockbox/developer/rasher)
14:35:17 Nick YPSY is now known as Ypsy (
14:35:33 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
14:36:22evilnickwirr: Simply because no-one has updated it, I'd guess :)
14:40:03wirrevilnick: so that's good news then
14:40:37wirrbtw, has anyone have a working recipe for an ssd replacement?
14:41:05 Nick fxb__ is now known as fxb (
14:42:50 Join DerPapst [0] (n=DerPapst@
14:45:10 Quit bzed (
14:45:10 Quit rasher (
14:46:27stripwaxsaratoga - Sorensen et al. apparently show how the data for split-radix dft can be rearranged to retain the same regular radix-2 bitreverse order. If that's true/easy then we can just get rid of the whole revtab stuff and use something simpler like bitrev16
14:46:40NJoinbzed [0] (
14:46:40NJoinrasher [50] (n=rasher@rockbox/developer/rasher)
14:46:41stripwaxI mean bitrev12
14:48:36 Quit bzed (
14:48:36 Quit rasher (
14:48:59NJoinbzed [0] (
14:48:59NJoinrasher [50] (n=rasher@rockbox/developer/rasher)
14:51:39 Quit bluebroth3r (Read error: 110 (Connection timed out))
15:01:49 Quit DerPapst ("Leaving.")
15:03:10 Part LinusN
15:05:12 Quit pamaury ("exit(*(int *)0 / 0);")
15:16:40 Quit stoffel (Read error: 113 (No route to host))
15:17:30CIA-6New commit by teru (r24076): make sure plugin reset backlight setting before exit. do code polish.
15:17:51 Join stripwax__ [0] (
15:19:34 Quit funman (Read error: 60 (Operation timed out))
15:22:41pixelmawirr, evilnick: the Iaudio ports lack charging in Rockbox hence support is incomplete
15:24:17pixelmaruntime is very fine though, usually better than with the original firmware
15:24:41wirrpixelma: i don't exactly understand. using rockbox the iaudio can only charge the batteries while powered off?
15:24:54CIA-6New commit by stripwax (r24077): Fix issue/bug where we were calling fft_init for every BLOCK in mdct, rather than once at init stage. Fixing the bug required reusing a single revtab ...
15:25:02 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
15:26:10pixelmawirr: you charge with the cowon loader (what you get when you plug USB or the charger from an off state)
15:26:47wirrpixelma: i see - so no charging whilst playing?
15:27:36evilnickpixelma: Thanks for the clarification
15:29:51pixelmawirr: no
15:29:57mc2739Zagor: it looks like there is still problems with the buildserver temporary block. My client was blocked for a missing log file and did not unblock until I restarted it approximately 15 hours later.
15:35:01 Quit stripwax (Read error: 110 (Connection timed out))
15:35:34 Quit fml ("CGI:IRC")
15:42:40 Quit Lss__ (Read error: 104 (Connection reset by peer))
15:43:46 Join Lss__ [0] (
15:56:38 Join n1s [0] (n=n1s@rockbox/developer/n1s)
15:58:54 Join DerPapst [0] (
16:03:58 Quit Utchybann ("I like core dumps")
16:06:53***Saving seen data "./dancer.seen"
16:10:54 Join Jaykay [0] (
16:13:45Jaykayanother bug which imo shouldnt be closed: FS #9386 it's a small bug, but i think it is one :)
16:13:52Jaykayjdgordon: ping
16:15:41 Join Lss [0] (
16:16:19 Join Riku [0] (
16:16:34 Quit Riku (Read error: 104 (Connection reset by peer))
16:19:57 Quit Lss__ (Connection timed out)
16:31:53 Quit wirr ("Leaving")
16:35:09 Quit Lss (Connection timed out)
16:35:55darkhamhi people, only a question: how works 16gb microsdhc in sansa fuze with rockbox?
16:36:15 Join martian67 [0] (n=arkfar@about/linux/regular/martian67)
16:37:38 Join panni_ [0] (
16:45:42FlynDicedarkham: I have not used one myself but there is no reason it should not. 8 GB sdhc works so 16 GB sdhc should work also. FS #10860 has a post from someone using a 16 GB uSD and his complaint is not about the card ... But once again, I myself have not tried it ;-)
16:47:23 Quit Zagor ("Don't panic")
16:53:35 Quit faemir (Read error: 104 (Connection reset by peer))
16:53:57 Join faemir [0] (n=faemir@
17:01:45DogBoydarkham, one thing to consider is the speed of the sd card
17:02:20darkhamfor class 4
17:02:31DogBoyit's just a concern for writing though
17:03:37DogBoyhad the same question back before I got an 8 gig card
17:08:56 Quit petur ("connection reset by beer (Leffe)")
17:10:13 Join toffe82 [0] (n=chatzill@
17:11:19darkhami've many of it+
17:11:36DogBoyyou what darkham
17:11:56darkhami've many 8gb cards
17:12:48darkhamno problems in read/write operations and in refresching content
17:13:24 Join captainkewler [0] (i=2669ecc2@gateway/web/freenode/x-umjhvnufsrmegzfh)
17:13:50 Quit bmbl ("Bye!")
17:37:07 Join kugel [0] (n=kugel@rockbox/developer/kugel)
17:41:02JdGordonJaykay: pong
17:42:31 Quit teru ("Quit")
17:47:30kugelJdGordon: do you close all bugs like that one?
17:49:20 Quit captainkewler ("Page closed")
17:55:20JdGordonsome context would be nice
17:55:22JdGordonwhat one?
17:55:44 Join Sajber^ [0] (
17:56:39kugelJdGordon: the one jaykay mentioned in the line before pinging you
17:57:48JdGordonwell... I've close probably 15 in the last few days.. if you really want an answer you could look at the close comments for each of them
17:58:06kugelI looked at that one
17:59:51 Quit Jaykay (Read error: 110 (Connection timed out))
18:03:27kugelJdGordon: and that one was pretty bad
18:04:47JdGordonso fix it
18:05:03kugelfix what?
18:05:28kugelthe close message?
18:06:55***Saving seen data "./dancer.seen"
18:09:06 Join LambdaCalculus37 [0] (n=LambdaCa@rockbox/staff/LambdaCalculus37)
18:14:37 Join AaronM [0] (
18:21:57*gevaerts agrees with kugel and Jaykay. FS #9386 should be reopened
18:34:54 Join mitk [0] (
18:35:24 Join JdGordon| [0] (n=Miranda@nat/microsoft/x-dswhpfihbrwygrzr)
18:35:50mitkAlso FS #10860 should be.
18:37:46 Quit mitk (Client Quit)
18:41:55 Join mitk [0] (
18:43:07mitkAnd btw Sandisk uSDHC card works for me from one year in Fuze with rockbox.
18:43:44mitkI mean 16gb uSDHC.
18:44:19Hillshummitk: 'from one year'?
18:44:36JdGordon|mitk: that bug has nothing to do with the inital report, or where its placed... if you want to reopen it file it again or properly edit the first post
18:46:04kugelmitk: I just spotted that one too
18:47:43kugelJdGordon|: yea, because it's easy for everyone to edit the title or first post..
18:48:06*kugel reopened that one
18:48:32mitkJdGordon| Bug is tracked down. Problem is between r23903 and r23904 which was "FS #10824 - viewport/statusbar API rework."
18:50:10kugelmitk: maybe leave a comment in a day or two whether you still experience it
18:51:23pixelmaI'm completely against editing the first post as the bug description there is important to understand if a bug is fixed or not. I remember an occasion where someone edited my bug report because he thought it was fixed but it wasn't (at least not completely) and I couldn't get the description back
18:51:34 Join stoffel [0] (
18:51:42 Join bertrik [0] (
18:52:21kugelpixelma: flyspray has a history
18:53:39pixelmawell, it is still not nice to edit someone else's bug report (except it is spam or worse)
18:54:18mitkkugel: I experienced "Data abort" while turning on dircache with all revisions beyond 23903 when sbs type theme was on. With non sbs them not.
18:54:25 Join kugel_ [0] (
18:54:42pixelmait's like editing the start of a discussion and then wondering what it is about
18:55:42mitkkugel: I tried all of this 5 hours ago
18:55:47 Join mikelietz [0] (
18:56:02mikelietzAny Sansa e270 experts around?
18:56:44FlynDicemikelietz: what's your question I'll try...
18:56:46*JdGordon| finds it slightly curious that noone else has mentioned anything about data aborts and dircache...
18:57:23mikelietzSometime recently I updated to the (then current) version. I don't know the version, nor the version I updated from.
18:57:34mikelietzI can't check, either, due to the problem I'm currently having.
18:57:37mitkJdGordon|: Anyone tried to do this
18:57:42mikelietzwhich is this: jewels is messed up.
18:57:59mikelietzit seems the textures are missing. So it loads the game screen, then row after row of grey balls fall
18:58:11mikelietzthe screen fills up with them
18:58:15mikelietzand then it starts over.
18:58:18mikelietzI can't break the loop.
18:58:27mikelietzPlugging it in doesn't work. Holding power doesn't work.
18:58:32mikelietzdo I need to open the battery cover?
18:58:49FlynDicemikelietz: v1 or v2?
18:58:55Hillshummikelietz: How long have you been holding power?
18:58:55 Join nossracer [0] (
18:59:03mikelietzI've tried about 5 seconds.
18:59:03gevaertshow long do you hold power? It *always* works unless the hardware is broken
18:59:05mikelietzI can try longer.
18:59:15gevaertsyou need 30 seconds
18:59:39mikelietzAh, that would do it. By my count it was around 20
18:59:41HillshumThat should do it
18:59:45mikelietzThanks :)
18:59:54kugel_by count it was 15 :)
18:59:57mikelietzNow that I've got your attention, what does it take to customize the menu order?
19:00:16mikelietzremove items, move things around, etc?
19:00:23 Quit kugel (Read error: 60 (Operation timed out))
19:00:23mikelietzIs there like an XML file I can edit?
19:00:34*kugel_ just got horrible fs corruption on the clip
19:01:17Hillshummikelietz: I think you need to hack the code
19:01:32mikelietzOn the Now Playing context menu, I'd really like two have two particular items - Add to playlist A and Add to playlist B, rather than going to Playlist Catalog -> Add to Playlist, then selecting the particular ones.
19:02:26mikelietzAny pointers to documentation for that sort of hacking?
19:05:03JdGordon|can you code in c?
19:05:23mikelietzI'm well out of practice, but I could probably re-learn it if I had to.
19:06:42JdGordon|get the source code... apps/onplay.c has that menu
19:10:42 Quit Xerion_ (" ")
19:10:56 Join YoYoFreakCJ [0] (
19:11:03YoYoFreakCJgood evening
19:11:12 Join j8048188 [0] (
19:11:50j8048188does the pitch_detector plugin work on the e200v2 yet?
19:13:00 Quit YoYoFreakCJ (Client Quit)
19:18:10j8048188i've played a 440 hz tone and i don't get any response whatsoever in the tuner screen
19:19:27bertrikThe ams sansas don't have support for recording yet, so they can't sample any input (there's a patch for it somewhere I think)
19:19:31 Quit mitk ("CGI:IRC (EOF)")
19:19:33mikelietzThanks all, and keep up the good work.
19:19:39 Part mikelietz ("Leaving")
19:19:46j8048188ok. Thanks!
19:20:25pixelmashouldn't the plugin be ifdefed out with HAVE_RECORDING then?
19:20:50 Quit Hillshum (Remote closed the connection)
19:20:56 Join Hillshum [0] (
19:21:05bertrikpixelma, yes that would make sense IMO
19:21:28n1sit is, isn't it?
19:22:02 Part toffe82
19:22:54 Nick Ypsy is now known as YPSY (
19:23:28n1sit is: #if (CONFIG_CODEC == SWCODEC) && defined(HAVE_RECORDING) && \
19:23:28n1s (defined(HAVE_LINE_IN) || defined(HAVE_MIC_IN))
19:25:09bertrikoh I was confused, e200v2 does have recording enabled, it was added in r23739 on nov 24
19:25:21j8048188that's what i thought
19:26:32FlynDiceI think fuze is enabled also
19:27:12FlynDicekugel_: corruption with latest rev or one that includes widebus?
19:27:27kugel_the latter i think
19:27:38FlynDicephew... ;-)
19:29:01kugel_format&resync would've probably been faster than checkdsk
19:29:56 Quit Unhelpful ("No Ping reply in 90 seconds.")
19:30:18 Join Unhelpful [0] (
19:30:51*gevaerts suspects that the issue with 10808 was the ambiguity in the phrasing of the close message, so he re-opened it to close it again with a clearer message
19:32:34 Quit amiconn (Nick collision from services.)
19:32:36 Join amiconn_ [0] (i=quassel@rockbox/developer/amiconn)
19:32:55 Join pixelma_ [0] (i=quassel@rockbox/staff/pixelma)
19:32:55 Quit pixelma (Nick collision from services.)
19:32:56 Nick amiconn_ is now known as amiconn (i=quassel@rockbox/developer/amiconn)
19:33:10 Nick pixelma_ is now known as pixelma (i=quassel@rockbox/staff/pixelma)
19:36:56j8048188when i go into the pitch_detector screen, i can hear an extremely rapid clicking on my earbuds, but nothing else happens.
19:38:26n1sj8048188: there might be an issue with recording and playing back at the same time on some players, pitch_detector does this
19:39:04n1swhich is what i think is causing it to crash on some players
19:39:20n1s(if playback is running when you start it)
19:40:42 Join arohtar [0] (n=faemir@
19:41:11 Quit faemir (Read error: 104 (Connection reset by peer))
19:41:20j8048188i found if i put my volume threshold down to 5 and turn the tone generator up (quite loud) i get a smattering of tone readings, but they are not consistent
19:43:49j8048188I'd be glad to do more testing later, but i have to go to work now :(
19:44:11j8048188i can be pm'd on the forums with this username.
19:44:24 Quit j8048188 ("CGI:IRC")
20:03:50 Quit DerPapst ("Leaving.")
20:03:58 Part watto
20:06:57***Saving seen data "./dancer.seen"
20:07:31 Quit bzed (
20:07:31 Quit rasher (
20:17:51NJoinbzed [0] (
20:17:51NJoinrasher [50] (n=rasher@rockbox/developer/rasher)
20:18:19 Quit bzed (
20:18:19 Quit rasher (
20:21:00NJoinbzed [0] (
20:21:00NJoinrasher [50] (n=rasher@rockbox/developer/rasher)
20:28:35 Join petur [50] (n=petur@rockbox/developer/petur)
20:41:19 Quit LambdaCalculus37 ("Leaving")
20:48:12 Quit Sajber^ (Read error: 104 (Connection reset by peer))
20:50:03 Quit kugel_ (Read error: 113 (No route to host))
21:01:42 Quit solexx (Read error: 113 (No route to host))
21:02:49 Join archivator [0] (n=archivat@
21:06:39 Quit gevaerts (Nick collision from services.)
21:06:48 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
21:06:57 Quit flydutch ("/* empty */")
21:15:03 Join Lss [0] (
21:22:55n1sis supposed to work?
21:24:58n1sit seems entirely unable to match addresses to symbols
21:31:05 Quit jfc^3 (Read error: 54 (Connection reset by peer))
21:31:27 Join jfc^3 [0] (
21:31:45 Join bluebrother [0] (n=dom@rockbox/developer/bluebrother)
21:32:45 Quit AaronM (Remote closed the connection)
21:35:08 Join AaronM [0] (
21:40:37 Nick fxb is now known as fxb__ (
21:42:31n1saddresses seem to be off...
21:44:27 Join angelwolf71885 [0] (
21:50:16 Quit angelwolf71885 ("ChatZilla 0.9.86 [Firefox 3.5.6/20091201220228]")
21:51:59n1sstripwax__: ping?
21:59:30 Join Grahack [0] (
22:06:03 Nick stripwax__ is now known as stripwax (
22:06:23n1sdid you get the profile_reader thingy to work on an arm target?
22:06:28 Join pamaury [0] (
22:07:00***Saving seen data "./dancer.seen"
22:08:11 Join stripwax__ [0] (
22:08:24stripwax__argh, who disconnected me?
22:09:00stripwax__../tools/profile_reader/ profile.out arm-elf-objdump vorbis.o 1
22:09:27stripwax__But recently I can't successfully generate profile.out on arm target for Codecs. I got instruction exceptions and all sorts
22:10:43stripwax__n1s - is it just failing to find addresses; or do the addresses it finds not make sense given the profile ticks/call counts?
22:11:52stripwax__Profiling anything that happens to have inline functions (by the way) will likely give garbage unless you add __attribute__((no_instrument_function)) to all those inline functions.
22:12:02n1sat least a few seem to diff just a little between the profile_out and the map
22:12:32stripwax__because gcc will instrument the "entry" and "exit" of the functions even though they're inlined, and they'll show up as extra calls to the parent function, which obviously they are not
22:12:38n1sdoesn't -finstrument-functions uninline ilnline functions?
22:12:46stripwax__n1s - hm, is the map stale?
22:12:51stripwax__n1s - I don't think so
22:13:14n1snah, the map is fresh
22:13:21stripwax__I guess if it did there'd be no need for the __attribute__ existing
22:13:37stripwax__I could be wrong. Maybe profile builds are missing some flags.
22:13:44n1sbut that shouldn't caus efunction address mismatches
22:13:45stripwax__n1s - hm, then I don't know I'm afraid
22:14:18n1slike for tremor, the function using the most ticks is mdct_v
22:14:28stripwax__mdct_v ?
22:14:34stripwax__Yep, that makes sense
22:14:57stripwax__you think it's wrong?
22:15:22n1sprofile.out lists the address 0x03d8e778 and the map has 0x03d8e768
22:15:30n1sso a 16Byte diff
22:15:56 Join Jaykay [0] (
22:16:03stripwax__are you using eabi gcc?
22:16:54n1sno, regular arm-elf-gcc, i can try eabi
22:19:09CIA-6New commit by stripwax (r24078): Removed now unused stuff from fft-ffmpeg.c/fft.h. Removed fixmul32 completely from lib (only fixmul32b now). First step in unifying tables between ...
22:19:20stripwax__Was just wondering, since I'm not using eabi, and last time I tried it did work for me
22:19:58stripwax__But I am **very** interested to know how you managed to profile vorbis/codec without getting bizarre instruction exceptions / prefetch abort stuff. Is this just a regular [P]rofile build or did you have to do some tweaks
22:20:17stripwax__ALL: my commit is not on trunk, it's on mdctexp branch
22:22:50 Quit shai ("Leaving")
22:22:53n1sstripwax: i did it on my beast, it doesn't use iram
22:24:56n1sstandard profile build, added $(PROFILE_OPTS) to some makefiles and calls to profile_thread and profstop, nothing special
22:26:31 Quit stripwax (Read error: 110 (Connection timed out))
22:28:27*n1s tries with the eabi toolchain
22:29:50 Quit Grahack ("Tu m'as vu ?")
22:33:48stripwax__interesting, yeah my aborts were in iram.
22:34:21stripwax__Anything in particular you're looking at profiling, for tremor? I posted some (recent-ish −− last ones that worked) for arm (ipod 5g)
22:34:33stripwax__I mean, a while ago, but probably still pretty valid
22:35:56n1snah, just interested, but also looking for what's burning cycles in the non transform part of the codec since it seems you guys are about to swap out the mdct anyway :)
22:36:15stripwax__mostly the bitstream stuff.
22:36:52n1si've managed some small speedups blindly but would like to get some numbers :)
22:37:05stripwax__decode_packed_block still slow , and any cases where decode_packed_block has to drop down to individual decode thingies are even slower. I had a look at trying out some of the Tremelo optimisations but (naively at least) they seemed slower
22:37:14stripwax__tiny? :)
22:37:54n1sthe ~0.5% kind of speedups, committed one but that was in the mdct though
22:39:03stripwax__oh, I missed that. which revision?
22:39:23stripwax__You can speed up the bitreverse part by only calling bitrev every other time
22:39:30n1sand i have one more, that gets 0.5% on coldfire at least but i need to verify that it's safe
22:39:36stripwax__(the other time is just a simple xor with 1<<n)
22:39:37 Quit Jaykay (Read error: 110 (Connection timed out))
22:40:17n1syeah, i though about that
22:40:40n1sdidn't test though, and coldfire seems to hurt a lot from branching
22:40:44 Quit stoffel (Remote closed the connection)
22:40:57n1sbut maybe i can thik of a clever way :)
22:43:03 Quit petur ("Zzzzz early night Zzzz")
22:48:56n1seabi toolchain gave no profile.out at all :/
22:49:00stripwax__if anything there'd be less branching
22:49:04stripwax__n1s - oh. urgh.
22:52:03 Join DerPapst [0] (
22:52:19 Quit Hillshum (Read error: 110 (Connection timed out))
22:53:29 Join Sajber^ [0] (
22:54:36n1sstripwax__: ah, yeah, it's called twice in every iteration, yeah
22:55:37 Join liar [0] (n=liar@
23:08:22stripwax__yep. replace one with a simple xor for an easy (but again tiny) improvement. in an mdct we're likely throwing away :)
23:22:15 Join Xerion [0] (
23:22:38 Join GeekShado_ [0] (
23:23:36 Quit GeekShadow (Read error: 104 (Connection reset by peer))
23:23:48 Quit AaronM ("Emo Time In My Corner... //_-")
23:30:28n1sstripwax__: yeah, it works, benching now, used a regular inclusive or, not that it matters
23:30:48 Join iamdooser [0] (i=48e00dc0@gateway/web/freenode/x-plryjeeqnjsmjunq)
23:31:36n1sanyway, this mdct replacement will likely not be done before the release so some people can enjoy tiny speedups for a while :)
23:31:42iamdooserHey everyone, can someone help me with an ipod 4th gen color?
23:32:00n1smight try to get that into tremor upstream too if i'm bored enough sometime
23:32:12n1siamdooser: if you tell us the problem we might
23:32:38iamdooserAfter uninstallation, it only boots if connected to firewire. I did a restore via iTunes to make it a mac pod, and connected to external power source. It does the procedure (see progress bar) then reboots, but stalls at the lit-up apple logo. It will go through this and let you select a language if you connect it to the computer with firewire. It also needs firewire to fully reboot using menu+select. It then works fine after
23:32:52 Join thegeek [0] (
23:32:56 Join knin [0] (
23:33:26iamdooserSo if I try to boot it any other way than with Firewire, it stalls at the apple logo.
23:33:26n1sso it only boots with firewire connected but works as usual if you disconnect after boot?
23:33:37iamdooserCorrect, n1s.
23:33:50n1ssounds like a battery that's died to me but i am no exper on ipods
23:34:13 Quit Tomis (Read error: 110 (Connection timed out))
23:34:14 Join blithe_ [0] (
23:34:33 Quit Galois (
23:34:33 Quit FlynDice (
23:34:33 Quit niekie (
23:34:33 Quit Dhraakellian (
23:34:33 Quit thegeek_ (
23:34:33 Quit zu (
23:34:33 Quit lostlogic (
23:34:33 Quit kaniini (
23:34:33 Quit sbhsu (
23:34:33 Quit shodanX (
23:34:33 Quit blithe (
23:34:33 Quit ThomasAH (
23:34:36NJoinlostlogic [50] (n=lostlogi@rockbox/developer/lostlogic)
23:34:42NJoinsbhsu [0] (
23:34:47NJoinFlynDice [0] (
23:34:50NJoinshodanX [0] (
23:35:00 Quit DerPapst ("Leaving.")
23:35:12NJoinzu [0] (
23:35:14iamdooseryou think so, even though it'll run for a while with the battery after it's properly booted?
23:35:34NJoinDhraakellian [0] (
23:35:36iamdooserAnd it seems odd it won't fully boot with the ac adapter.
23:36:20n1shmm, yeah, that seems odd
23:36:59n1sstripwax__: unless i bothced the benchmark it's actually a small slowdown on the beast, maybe because it needs another register
23:38:05 Join stickboy [0] (
23:39:01NJoinThomasAH [0] (
23:45:53 Quit JdGordon| ("Miranda IM! Smaller, Faster, Easier.")
23:57:55 Quit pamaury ("exit(*(int *)0 / 0);")

Previous day | Next day