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 2005-07-13

00:05:13 Part stripwax
00:05:18 Join stripwax [0] (
00:05:46 Quit Coldtoast (Read error: 148 (No route to host))
00:08:52Rorilol Elfman score on new wonka movie is fun
00:09:21RoriVery reminiscent of one of his first movie scores....tries to remember the name
00:10:13bill20r3forbidden zone?
00:10:42stripwaxbill20r3 - don't know that one
00:10:51RoriOingo Boingo r0x
00:11:11bill20r3some movie he was in when they were still the 'mystic knights of oingo boingo', he played the devil
00:11:26bill20r3also starring Herve Villachez (aka: tattoo from fantasy island)
00:12:01RoriI may have that if it's what I am thinking of
00:12:22Rorianother low budget oddity from what I again evades me
00:12:33Rorib&w I think
00:12:36 Part KasKaraK
00:12:50Rorinot name shot in black and white
00:18:12 Join Bazz [0] (
00:18:35Bazzare the any plans/projects underway to get rockbox working on iriver h3xx players?
00:21:04BBubit is planned, yes
00:22:24BazzBBub: cool, do you know what stage it's in at the moment?
00:22:41BBubafaik the team hasnt started yet
00:23:31Bazznod. do you know who is leading the team and if they what/need any help?
00:26:19RoriI'd like to see rockbox for h1x0 at least get to beta 1 first
00:26:33BBubBazz: they always need coders ;)
00:26:36HClthe stage for h3x0 is... well. we have a simulator, but aside from that, nothing.
00:26:44HCland nobody's working on it either
00:26:55HClbut it is planned. someday.
00:27:13Roriif h1x0 gets most of the functionality going including recording and remote then I would say h3x0 can go ahead ;)
00:27:43 Join ashridah [0] (
00:28:05Roribut everything is fluid
00:28:25Roripeople can work on what they like there is no obligation here
00:28:27BazzHCl: if i was interested in getting involved, how would you suggest i begin? i have a solid background in coding, but not much with embeded stuff.
00:30:18ashridahbazz: the simulator can let you poke at the general design of rockbox before worrying about the nitty-gritty details of the target.
00:30:33Bazzashridah: where does one get said simulator?
00:31:11ashridahit's part of the rockbox source, it follows the development as closely as possible.
00:31:45ashridahwhat were you looking to do development in? windows? linux?
00:31:50Bazzashridah: linux
00:32:09Bazzso if i just get the source tarball off the download page it'll be in there?
00:32:16ashridahthat's not a problem then. if you've already compiled other stuff on the system before, you'll more than likely have the gear for the simulator.
00:32:21ashridahwe tend to use CVS
00:32:56Bazzahh, k, i'll check it out then
00:33:18ashridahthere are daily drops, but they're precompiled. will cover what you want
00:33:30 Quit Aison ("( :: NoNameScript 3.72 :: )")
00:34:11ashridah covers building it
00:35:14ashridahnote that those two might reference 'archos', but if it's the iriver build you're after, that's available too (the differences should be relatively minor... in theory :) )
00:35:37ashridahwhich player are you interested in, btw?
00:37:20Bazzwhich i realize is probably the least worked on at the moment :)
00:37:54ashridahnot necessarily. bits and pieces are getting done, but since there's no working bootloader, and no LCD driver for it yet, it's kinda useless
00:38:01Bazzashridah: nod
00:38:23ashridahbut yeah, to build code to actually run on that (when that eventuality comes around), you'd need to build a cross-compiling version of binutils and gcc for yourself.
00:38:49Bazzashridah: yeah, i actually cross-compile for some other stuff so it's no problem
00:42:56Bazzhrmm, getting an error in the config script: sed: -e expression #1, char 7: Extra characters after command poking about in the script now, but i figured you might know right off what's wrong (it does say i have the fine version of sed right after that)
00:46:54RoriIs there a crossfeed plugin for Rockbox?
00:47:01 Join JAJDude [0] (
00:47:40RickProbably means crossfade
00:47:48Rorino crossfeed
00:47:55stripwaxRori and what is that?
00:48:03 Quit JAJDude (Client Quit)
00:48:20 Join JAJDude [0] (
00:48:30Roriit crossfeeds some of the left and right audio channels for a more natural 'speaker like' experience on headphones. There's more to it than that of course but that's the basics. Just learned of it and using it and it's good
00:48:42stripwaxRori - you mean like Steror S
00:48:44Roriplugin for winamp or foobar
00:48:49stripwax^Stereo Separation
00:49:10Roricrossfeed 'leaks' left into right and visa versa
00:49:21Roriplus some trickery with delay etc
00:49:38Rorigo checkout 'Headplug' for winamp or crossfeed plugin in foobar
00:49:40stripwaxRori - apart from delay trickery, that (should be) the same as Stereo Separation do?
00:49:52Rorigo check it out
00:50:02ashridahRori: no, there isn't a plugin for that yet, and i for one, would love to see it :)
00:50:46ashridahstripwax: lots of music is recorded with speakers in mind, thus, the left channel often doesn't play a time-shifted version of what the right channel has
00:51:00ashridahthis has the effect of making sound appear to be coming from directly inside your skull, when wearing headphones
00:51:12ashridahand is quite distracting with certain kinds of jazz, which tends to be VERY separated
00:51:14Roriwhen you listen on speakers you get left and right leaked into each ear anyhow ya see
00:51:17ashridahas well as lots of other music
00:51:44ashridahwhat rori wants is to have a timeshifted and muted version mixed into the other channel, and vice versa
00:52:02stripwaxoh right, so a kind of frequency-dependent delay?
00:52:12ashridahwhen done properly, this tends to move music that appears to be inside your head to infront of you
00:52:30Rorivery effective
00:52:54Rorilistening to a wonka soundtrack and it creates a big soundstack in front of me
00:52:56ashridahstripwax: you don't need to do frequency dependent work on it for simpler versions, just measure the average head and divide by the speed of sound
00:53:11stripwaxashridah - so that's just a delay then, right?
00:53:28stripwaxthat should be a piece of cake for a simple mp3 player to do
00:54:02ashridahi imagine it can be worked into the DSP stuff that currently covers resampling
00:54:27ashridahyou've got to be careful not to end up with a feedback loop tho :)
00:54:27stripwaxashridah - for a really flexible solution, the dsp code would be able to accept dsp plugins
00:55:31stripwaxor just preprocess all your mp3 files with foobar to apply the crossfeed before transferring to your player? ;-)
00:55:42ashridahno, thanks
00:55:55ashridahsince i occasionally listen to them with speakers
00:58:40ashridahif you want a better sounding version, you can model how the ear affects sound and start applying that too, but it's hard to get that right, since ears are dependent on each person's genetics :)
00:59:54RoriThe best recordings are those done with those mics in artificial heads
00:59:59Rorifor live stuff
01:00:15stripwaxRori - would those be best played back on headphones?
01:00:24Rorigood question
01:00:30RoriI would have thought so
01:00:41Roribinaural recordings are an odd thing
01:00:50 Join Coldtoast [0] (
01:02:18 Join BBub_ [0] (
01:02:18 Quit BBub (Read error: 104 (Connection reset by peer))
01:02:20 Nick BBub_ is now known as BBub (
01:02:27stripwaxgotta get some sleep - laters
01:02:31 Part stripwax
01:05:49Bazzsomeone know what this line is supposed to do in configure from cvs: lines=`echo "moo" | sed -e '/moo/cline1\nline2\nline3' | wc -l` (well, i realize it's supposed to generate more than one line to test your sed version, but it gives an error and i'm not sure what it's supposed to be)
01:07:13***Saving seen data "./dancer.seen"
01:10:20ashridahlines should get set to 3, by the looks of it
01:11:33ashridahwhich version of sed do you have?
01:11:52ashridahand what error are you getting out of that command?
01:13:07Bazzi have sed 3.02 and i get: sed: -e expression #1, char 7: Extra characters after command
01:13:41ashridahcan you upgrade your version of sed?
01:14:14ashridahinfact. tjat
01:14:16Coldtoastaaah hey
01:14:31Coldtoastjust found something somewhat confusing in teh wps
01:14:40ashridahthat's very interesting, since the test is actually looking for broken versions of sed 4.x, not 3.x
01:14:47Coldtoastusing %pp/%pe
01:15:15Bazzashridah: yeah, i realized that, i'm not sure if i can upgrade sed or not, i have to keep certain versions of things around for other stuff, but let me see
01:15:24Coldtoastif I select track 1 in a dir with 9 tracks, %pp/%pe displays "1/9"
01:15:36Coldtoastif I skip to the next track, it changes to "2/9"
01:15:37Coldtoastand so on
01:15:41*amiconn now has a solid greyscale rotating cube on his iriver :)
01:16:21Coldtoastbut if I go to the file tree and select a track in the dir, rockboc generates a new playlist and the track I selected becomes track 1 in the playlist
01:16:57Coldtoastit seems sort of odd
01:17:03Coldtoasteven tho I know how it's working
01:17:35ashridahit comes up as 5/14 here
01:17:45Coldtoastif you select it in the file tree?
01:18:04Coldtoastok. mine shows as 1/9
01:18:25Coldtoastah. hold on
01:19:07ashridahi select the first one, change tracks, goes to 2. press the navi button in, select '5', it builds a new playlist, and sets me up running track '4'
01:19:10ashridah'5' even
01:19:50ashridahBazz: well, you can build a version of sed and install it in a local directory in your homedirectory, then add it to the path when doing rockbox development
01:19:54Coldtoastah hah
01:20:00Coldtoastheh. ignore what I just said...
01:20:05 Quit Harpy (Read error: 110 (Connection timed out))
01:20:07Coldtoastjust realised why it was doing it......
01:20:14ashridahit's what you'd probably want to do with your cross compiler for m68k anyway, later on
01:20:24ColdtoastI had Shuffle enabled. heh
01:21:34Bazzashridah: that just what i'm doing :)
01:23:45 Quit BBub (Read error: 110 (Connection timed out))
01:28:30Bazzashridah: that fixed configure problems, now on to compile problems :)
01:28:30 Quit Primedeath|Away (Read error: 131 (Connection reset by peer))
01:28:42 Join Primedeath|Away [0] (
01:29:17*ashridah idly wonders how on earth bazz has such an old sed version
01:29:44Bazzashridah: try having to support old installs for your customers ;)
01:29:49 Quit Coldtoast (Read error: 148 (No route to host))
01:30:06Bazzwhat gcc version is recommended?
01:30:31ashridahwell, when you're building the cross compiler for m68k, you *must* use gcc 3.4.4
01:31:00ashridahi'm not sure about the simulator itself, however
01:35:25ashridahheh. as for supporting old installs, vmware, chroot, and user-mode-linux come in handy :)
01:45:24 Quit ashridah ("Leaving")
01:58:55 Quit pike ()
01:59:11 Join pike [0] (
02:00:38Bazzyeah, that's true
02:00:47Bazzwell, gotta run, thanks for the help, i'm sure i'll be back :)
02:00:49 Quit Bazz ("Client exiting")
02:09:58 Quit JAJDude ("CGI:IRC (EOF)")
02:19:26 Quit Sucka ("a bird in the bush is worth two in your house")
02:24:48 Quit hicks (Remote closed the connection)
02:31:34 Quit cYmen ("zZz")
02:52:44 Quit bipak (Read error: 60 (Operation timed out))
03:07:16***Saving seen data "./dancer.seen"
03:07:47 Join Maxime`Mrn [0] (
03:07:47 Quit Maxime (Read error: 104 (Connection reset by peer))
03:32:31 Join TCK [0] (
03:53:45 Quit Moos (Read error: 110 (Connection timed out))
04:05:14 Join QT_ [0] (as@area51.users.madwifi)
04:05:41 Join jamesshuang [0] (~james@
04:06:19jamesshuanghey everyone, I'm trying to get into the developer side of rockbox right now
04:06:36jamesshuangI'm having some trouble building the simulator though
04:07:25jamesshuangI'm on a standard gentoo system, and I'm running configure and everything, but for some odd reason, I'm getting this linker error:
04:07:43jamesshuang/usr/lib/gcc/i686-pc-linux-gnu/3.4.3-20050110/../../../../i686-pc-linux-gnu/bin/ld: errno: TLS definition in /lib/ section .tbss mismatches non-TLS reference in /home/james/rockbox-checkout/rockbox-bleeding/build-simulator/librockbox.a(errno.o)
04:09:06jamesshuanganyone have any idea how I can solve this problem? I'd rather not put a new copy of everything on my iriver each time :-p
04:17:22 Quit QT (No route to host)
04:18:23alxcmodd issue
04:18:27alxcmi'm on gentoo, i should try it
04:18:39alxcmis your glibc up to date?
04:18:50jamesshuangwell, whatever it was in portage about a month ago
04:19:30jamesshuangwhich happens to be version 2.3.5
04:21:09jamesshuangactually, I'm working on getting greyscale bmp's to load... I dunno what amiconn is doing with that but it looks like the bmp2rb program on the computer shares some code with bmp.c
04:21:19jamesshuangor at least some similar looking code
04:21:21alxcmhmm, you have a standard toolchain? no funky ld configs?
04:21:33jamesshuangyeah, standard toolchain, stage 1+3 compiled
04:21:50alxcmi'd have to give it a try then
04:22:21jamesshuangI'm just trying to artificially splice the bmp2rb code from the computer program into a plugin so I can load greyscale bmp's on rockbox
04:23:18alxcmoh, cool
04:23:31alxcmi hate linker errors
04:23:37alxcmso so hard to trace
04:23:50jamesshuangyeah, compiler errors I know how to solve... linker errors? I dunno what I'm doing
04:24:25alxcmdo you have nptl enabled by any chance?
04:24:29jamesshuangyes, I do
04:24:40alxcmthat's probably the issue
04:25:14jamesshuangyeah, I definitely do... it's in my USE vars
04:25:17 Quit Primedeath|Away (Read error: 110 (Connection timed out))
04:25:38jamesshuangthat's annoying... does that mean I have to reinstall gentoo without nptl? I thought glibc build with the native linuxheaders as well?
04:27:34alxcmit should
04:27:44alxcmactually it would have to
04:27:58alxcmunless you use nptlonly in the USE flags i think
04:28:08alxcmbut that would break a whole lot
04:28:15jamesshuangnope, definitely did not do that... heard quite a few horror stories
04:29:48alxcmnever tried it myself
04:29:51 Join DarkkOne [0] (
04:30:25alxcmit could also be glibc 2.3.5
04:30:37alxcmi'm still seeing only 2.3.4 in portage
04:30:41alxcmand i JUST emerge sync'd
04:31:10jamesshuangyeah, I decided to use ~86 because all the package hiding was pissing me off
04:31:29alxcmthat's probably why then
04:31:47alxcmyup, ACCEPT_KEYWORDS shows 2.3.5
04:43:47 Join JAJDude [0] (
04:45:12 Quit DarkkOne ("Chatzilla [Firefox 1.0.4/20050511]")
04:45:12 Quit JAJDude (Client Quit)
04:59:55 Join bipak_ [0] (
05:01:46 Quit bipak_ (Remote closed the connection)
05:03:26 Join amiconn_ [0] (
05:07:20***Saving seen data "./dancer.seen"
05:21:05 Quit amiconn (Read error: 110 (Connection timed out))
05:21:06 Nick amiconn_ is now known as amiconn (
06:36:50 Join LinusN [0] (
06:45:10 Quit pill (Read error: 145 (Connection timed out))
07:07:23***Saving seen data "./dancer.seen"
07:15:49 Join webguest11 [0] (
07:21:14 Join pill [0] (
07:56:36 Join B4gder [0] (
08:00:57B4gdergooood morning
08:13:20 Join oxygen77 [0] (
08:17:38 Quit alxcm (Read error: 110 (Connection timed out))
08:17:56LinusNamiconn: nice cube :-)
08:18:07LinusNreally shows off the sloooooow lcd :-)
08:18:21B4gderamiconn: you are now officially insane
08:18:54amiconnLinusN: tnx
08:19:03amiconnB4gder: why?
08:19:38B4gdersuch an effort for such a... usless thing
08:20:14B4gderbut then I haven't seen it in action yet
08:20:20B4gderno cable here
08:20:29amiconnThe effort wasn't *that* big
08:20:34LinusNfun is never useless, you should know that, after all the demos you've coded
08:20:46B4gderI know fun
08:20:52amiconnI could reuse some of the work of the ipodlinux guys
08:20:59amiconn(specifically coob)
08:21:09B4gderaha, nice
08:21:26amiconnB4gder: You can also run it in the sim, but do that with a windows sim
08:21:26B4gderthey ported our and you could semi-port it back! :-)
08:21:56B4gderyeah, need to fix the x11 sim
08:22:00amiconn(it will most probably either look dull or strange in the x11 sim)
08:22:04LinusNguys, how do we solve the memcpy problem with the plugins?
08:22:07B4gderI don't have any windows machine to try on
08:22:19LinusNi vote for having memcpy and friends in the plugin lib
08:22:20amiconnTry wine?
08:22:49amiconnLinusN: I already thought about that, and it seems linking memcpy and frieds to the plugins may be the only solution
08:23:05amiconn...either in the plugin lib or the gcc lib version
08:23:10B4gderbut they should be fairly small
08:23:18LinusNbuto how do we avoid duplicating source code?
08:23:32B4gderperhaps we make a memlib
08:23:46LinusNa clib perhaps
08:23:47B4gderthe codecs need them too, right?
08:23:48amiconnI would suggest to add memcpy.o and friends to libplugin.a
08:24:23amiconn*memcpy.o from firmware/common/
08:24:35B4gderis memcpy enough?
08:24:57amiconnHmm, that may or may not work... depending on which optimisation is used
08:25:10amiconnFor asm, there is no memcpy.o but memcpy_a.o
08:25:57amiconn14.24.19 # <amiconn> Quote from "Most of the compiler support routines used by GCC are present in libgcc, but there are a few exceptions. GCC requires the freestanding environment provide memcpy, memmove, memset and memcmp."
08:27:52amiconnHmm, I should've read that thorougly. No mem* in libgcc...
08:28:18amiconnThe libmem idea sounds good to me.
08:28:26amiconn...but we have no memmove yet
08:28:46B4gderwe could easily add one for now, and optimize it later
08:28:51amiconn...and I would really like to *replace* memcpy with memmove
08:29:38amiconn...but this doesn't work the simple way (#define memcpy memmove) because of these requirements
08:29:51amiconnIt should work to define an alias
08:30:28amiconnI wouldn't add memmove right now as long as gcc is happy without it
08:30:48B4gdertrue, but it would make it more complete
08:30:57B4gderand the codecs need and use it
08:39:10 Quit oxygen77 ("CGI:IRC")
08:39:13 Join oxygen77 [0] (
08:40:02*B4gder has gotten doubts on gcc 4.0.1 on x86
08:41:48B4gderall of a sudden several curl tests fail
08:42:01B4gderafter the gcc upgrade
08:42:31B4gderhaven't confirmed that gcc is guilty yet though
08:43:06 Join cYmen [0] (
08:45:10B4gderand lots of new rather annoying warnings
08:45:28B4gderlike on assert()
08:47:03 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
08:53:55 Join matsl [0] (
09:07:24***Saving seen data "./dancer.seen"
09:09:12 Join webguest62 [0] (
09:10:49 Join El_Gringo [0] (
09:11:21 Part El_Gringo
09:14:33 Join einhirn [0] (
09:15:10 Join El_Gringo [0] (
09:15:15 Quit El_Gringo (Client Quit)
09:29:37 Quit oxygen77 ("CGI:IRC")
09:33:13 Join oxygen77 [0] (
09:37:55 Join ashridah [0] (
09:45:05 Quit Lynx_awy (" reboot")
09:47:49 Join Lynx_ [0] (
10:03:47 Nick QT_ is now known as QT (as@area51.users.madwifi)
10:03:59 Join alxcm [0] (
10:04:45 Join webguest39 [0] (
10:05:27 Quit MrStaticVoid ("Be right back")
10:16:30 Join Coldtoast [0] (
10:17:35 Join MrStaticVoid [0] (
10:20:17 Join Harpy [0] (
10:24:26 Join Aison [0] (
10:28:15B4gdercalm day today
10:39:15webguest62massively off topic, but *damn* google earth is impressive!
10:41:41Lynx_not for europe
10:46:15webguest62Well they've not done the 3d Buildings for anywhere but the US. But the satellite imagery of my house is pretty good!
10:46:37webguest62Managed to catch London on a sunny day ny the looks of it.
10:47:36Straththat a rare occurrence? :)
10:47:47webguest62well not lately - but usually yes ;)
10:48:06webguest62it's pretty sunny here today anyway
10:50:37 Join webguest29 [0] (
10:50:56 Quit webguest29 (Client Quit)
10:50:59 Quit webguest39 ("CGI:IRC")
10:52:57 Join Lear [0] (
10:53:51 Quit matsl (Remote closed the connection)
11:00:05 Quit Lear ("Chatzilla [Firefox 1.0.4/20050512]")
11:01:34 Nick Strath is now known as StrathAFK (
11:07:28***Saving seen data "./dancer.seen"
11:13:26 Quit oxygen77 ("CGI:IRC")
11:14:36 Join oxygen77 [0] (
11:17:18 Quit jamesshuang (Read error: 110 (Connection timed out))
11:17:19 Join F1^Aison [0] (
11:21:35 Quit webguest11 ("CGI:IRC")
11:33:53 Quit Aison (Connection timed out)
11:34:21 Join Aison [0] (
11:35:52 Quit F1^Aison (Connection timed out)
11:40:17 Join jmlfr [0] (~chatzilla@
11:41:08 Part jmlfr
11:52:25 Join Bigribo [0] (
12:00:11 Join hicks [0] (
12:04:03 Quit webguest62 ("CGI:IRC (Ping timeout)")
12:11:17 Join Febs [0] (
12:20:34 Join webguest24 [0] (
12:31:43 Quit webguest24 ("CGI:IRC")
12:57:27 Quit Coldtoast (Read error: 113 (No route to host))
13:04:09*amiconn found a way better method to decide whether a face is visible/invisible in cube.rock
13:06:25 Join webguest99 [0] (
13:07:32***Saving seen data "./dancer.seen"
13:07:35webguest99any chance of being able to select random album playback in file tree mode
13:08:14webguest99thats very definite
13:08:16B4gderit already does that
13:08:21webguest99no it doesnt
13:08:22B4gderwithin a dir
13:08:30webguest99no it doesnt
13:08:48B4gdermine does
13:09:12webguest99I have 378 albums in a music file and it doesnt play them back randomly
13:09:33webguest99it plays them consecutively, and the songs within the albums are random
13:09:41B4gder"albums in a music file"?
13:09:52webguest99music directory
13:11:08 Join Moos [0] (
13:12:41B4gderuse a playlist
13:12:55webguest99LOL, the answer to everything
13:13:16B4gderrockbox will never play random over multiple dirs
13:13:23B4gderwithout playlist
13:13:34 Quit edx (Read error: 110 (Connection timed out))
13:13:36webguest99not possible to code ?
13:13:55B4gderit is possible to code
13:14:12B4gderbut it doesn't fit the rockbox concept
13:14:42B4gderwhen you've read up on rockbox internals, you'll see
13:14:56webguest99right, thanks for the info
13:15:19 Part webguest99
13:20:34amiconnWith my new algorithm it is possible to draw the wireframe cube with hidden lines removed :)
13:22:14LinusNamiconn: using the sign of the Z-component of the normal vector?
13:22:32 Quit hicks (Remote closed the connection)
13:26:10*LinusN is reviewing the updated a-b patch
13:26:33LinusNit adds an event system for the playback
13:27:06LinusNevents are currently TRACK_CHANGE and POS_REPORT
13:27:45LinusNthe event dispatcher calls the registered event handlers when an event occurs
13:29:13LinusNthe a-b repeat code registers a handler for the POS_REPORT event, and calls audio_ff_rew(<A pos>) when the B pos is reached
13:29:50LinusNwe should be able to port this to the software codecs as well, but the latency is a problem
13:29:52B4gdersounds like a rather fine approach
13:32:59 Quit oxygen77 ("CGI:IRC (EOF)")
13:34:42amiconnLinusN: I redefined the faces[] array to have all faces winding clockwise when looking from the outside
13:34:50LinusNthe buzzing sound on boot seels to be fixed by my 1380 reset
13:35:20amiconnThe I use a simplified cross product to detect the winding, and only draw if it is > 0 i.e. we're looking at the front face
13:35:41LinusNthat's what i would do too
13:36:02amiconn(after projecting to 2-D, so it is very simple, all z components are zero)
13:36:42 Quit Bagder_ (Read error: 110 (Connection timed out))
13:36:46amiconncoob's approach was sorting the 6 faces by z-order and then draw the 3 frontmost ones only
13:36:54amiconn...but this has disadvantages
13:37:37amiconn(1) It always draws 3 faces, even if one or two of them are invisible and get overdrawn
13:37:50amiconn(2) It needs sorting
13:39:30amiconn(3) it only works for solid faces, not for wireframe
13:40:42LinusNpretty much lame, in other words :-)
13:47:51B4gderportalplayer has 96K "internal sram"
13:47:56B4gderfeels familiar ;-)
14:02:55 Join RotAtoR [0] (
14:03:36Bigribois there any player at all that can play MOD or SID ?
14:04:18B4gderipodlinux can apparently play mods
14:04:40B4gderpowered by MikMod
14:06:04B4gderSIDs are even harder
14:08:14Bigribothere is libsidplay but it would have to be ported?
14:08:48Bigribomikmod is built to be ported :)
14:09:09B4gderfeel free to make it happen
14:09:19Bigriboyes I'd like that
14:09:29Bigribomaybe on my vacation ;)
14:10:42B4gder mentions all of these
14:10:48B4gderwith some pros and cons
14:11:23Bigribobesides, you can always get C64 remixes as ogg. and it's not limited to the C64 soundchip
14:12:04Bigriboah.. I'm gonna check this
14:14:32BigriboSPC would be pretty cool too
14:14:56Bigriboand if this thing worked on the iFP 390 :)
14:16:33CassandraNew hardware platforms are the trickiest things.
14:16:53CassandraSomeone has to write the low level bootloader and driver support for the hardware.
14:20:44amiconnLinusN: Do you have a simple solution how to avoid drawing visible edges twice, if both adjacent faces are visible?
14:20:45 Join webguest28 [0] (
14:21:56 Part webguest28
14:23:56BigriboSID is a funky thing with illegal OPcodes
14:24:37Bigribothere was some guy that built an exclusive SIDpod
14:24:37B4gderthey don't normally use illegal opcodes
14:25:12*B4gder wrote a music editor/playback code for the C64
14:25:42SlasheriB4gder: i will commit soon and after that it should be very easy for you to add the linux audio playback support :)
14:25:56*B4gder feels some preasure now :-)
14:26:10B4gderso much to do, so little time
14:26:30amiconnB4gder: Thinking about adding greyscale to the x11 sim? ;)
14:26:34B4gderI thought I'd start with contributing my tiny alsa-app, in case someone else wants to start
14:26:52B4gderamiconn: yeps, I've downloaded some sample code to figure out how to do it
14:26:52Bigribohe used a PIC processor for it
14:27:58B4gderand I have the lang2 scripts in the pipe too
14:28:10amiconnAn, the nice ones :)
14:30:06Bigribogrrr. I want a sidpod
14:30:38LinusNamiconn: you mean not drawing the same line twice in a wireframe?
14:31:32LinusNlet each line have an id
14:31:47LinusNand keep record of what you draw
14:31:50 Join edx [0] (
14:31:56amiconnFor simple wireframe this is trivial, since there is an array holding the 12 lines
14:32:24amiconnHowever, when dealing with faces, there is an array of the 6 faces, which holds the 4 corners of each face
14:32:51amiconnI'm not sure how to link that to the lines without too many indirections
14:32:56LinusNeach face should have a list of which lines it uses
14:33:32amiconnHmm, you mean in addition to the list of corners?
14:34:18amiconnThen each line would need a tag indicating whether it was already drawn
14:34:32LinusNor you have a bitmask
14:34:52LinusNone bit per line id
14:35:23amiconn...and these tags need to be reset for every new frame
14:35:59amiconnHmm, the bitmask idea is very good :)
14:36:01LinusNan array of bool, indexed with the line id
14:36:28amiconnThen every face would include the bitmask of all lines it includes, as one mask
14:37:57LinusNand you don't need all corners of each face, only 3 of them
14:38:13LinusNfor the cross product
14:38:22amiconnYes of course
14:38:31amiconn...but I need all of them for drawing
14:38:48LinusNno, the line has the coordinates
14:39:12amiconnThat would add a layer of indirection...
14:39:20amiconnIs there an official way to write binary constants in C source?
14:39:52amiconnSomething like 0b11011010 ?
14:40:00LinusNamiconn: is indirection a problem?
14:40:11amiconnIt's slower...
14:40:24LinusNis speed a problem?
14:40:39amiconnNot on iriver...
14:41:08amiconn...but I intend to use the grayscale lib on archos later
14:41:43LinusNi think proper line and face elimination saves you a lot more than avoiding indirection
14:42:12LinusNbesides, i have a few models i want to include, that are defined this way :-)
14:42:19amiconnHowever, I can have that without the indirection, with just one additional field in the face definition
14:42:39LinusNi was this >.< close to updating cube.c a year ago
14:42:55amiconn...the bitmask which lines it uses in addition to the 4 corners
14:44:53LinusNwhat if a face doesn't have 4 corners?
14:45:06amiconnThey all do
14:45:17amiconn...for a cube
14:45:20LinusNnot in my models
14:46:03amiconnIt's the same with lines. If a face doesn't have 4 corners, it does also not have 4 lines
14:46:13LinusNof course
14:46:16amiconnHow would you design a structure to accomodate that?
14:46:25LinusNwith lists
14:46:49LinusNa list of all lines in the model, containing coordinate triplets
14:47:16amiconnThis means you would store each point twice...
14:47:23LinusNno wait
14:47:33LinusN1) a list of all points
14:47:42amiconnCurrently, cube.rock uses another approach
14:48:03LinusN2) a list of lines, containing two point id:s
14:48:03amiconn(1) There is a list of all points (8), unrotated
14:48:27LinusN3) a list of faces, containing the line id:s
14:48:33amiconn(2) This is transformed by the transformation matrix into another list of the same structure
14:48:48LinusN4) a list of the same faces, with 3 point id:s for the cross product
14:49:10LinusNSlasheri: what is committed?
14:49:25SlasheriLinusN: i separated low level pcm drivers and pcm buffering
14:49:32LinusNSlasheri: good
14:49:37Slasherinow it crossafading etc. will work nicely with simulator also :)
14:49:42amiconn(3) Then the projection to 2D happens, generating a list of the same length, but with 2-D points
14:52:23 Quit TCK (Read error: 110 (Connection timed out))
14:52:37LinusNamiconn: this is the structure we use for the models in The Last Traktor III
14:54:07*B4gder feels the wave of nostalgia coming in
14:54:24 Join TCK [0] (
14:55:37SlasheriB4gder: btw, now you can find all the necessary pcm driver stubs to be implemented in uisimulator/common/stubs.c
15:01:43Slasheriamiconn: Hmm, the backlight dimming is not smooth with the latest cvs.. I don't know when that behaviour has changed
15:03:32amiconnLinusN: The second list of points is redundant the same way as my extra line bitmask would be
15:05:31Slasheriamiconn: ah, never mind. That was caused by one unnecessary logf call i already removed
15:05:38Slasheriit was just called a way too often
15:06:04amiconnWhy would calling logf() influence the backlight dimming?
15:06:12amiconnDid you call it from the ISR?
15:06:57Slasheriyes :D
15:07:35***Saving seen data "./dancer.seen"
15:07:48Slasherioh, i think my commit failed to build..
15:08:31amiconnHmm, why?
15:08:55Slasherifixing that.. i didn't commit debug_menu.c because i had done other changes to that file too i didn't want to commit
15:16:09LinusNamiconn: yes, you can find the 3 points in the line list, if the lines are sorted clockwise in the list
15:16:48LinusN(in the face definition)
15:21:54 Quit Febs (Read error: 110 (Connection timed out))
15:30:23*LinusN likes the a-b repeat
15:32:02LinusNso, what do people use a-b repeat for?
15:32:26Maxime`MrnI liked the "play next" function in the original firm'
15:32:39LinusNplay next?
15:32:53Maxime`Mrnyeah, where you choose the next song to be played
15:33:28LinusNand you don't like the rockbox queue?
15:33:55crwlwhat's the difference between "insert next" and "queue" in rockbox anyway?
15:33:58Maxime`Mrnhum, it's accessible from one touc ?
15:34:06MoosLinus: me i used the A-B with th original fw for listened a part of big size songs or for repeat just 2 or 3 tracks :)
15:34:12crwli'd suppose queue is something you would want to do for stuff that's already in the playlist, but i coudln't find out how
15:34:45LinusNcrwl: "insert" inserts the file in the dynamic playlist, "queue" will remove the file after it has been played
15:34:53LinusNso "queue" is temporary
15:34:55 Join R3nTiL [0] (
15:35:14crwlLinusN, ah, ok, that sounds useful
15:35:34LinusNMaxime`Mrn: no, not with one touch
15:36:00LinusNso the a-b repeat in the iriver firmware could span several tracks?
15:36:33SlasheriMaxime`Mrn: Hmm, how you can queue songs in iriver firmware?
15:36:49crwli didn't know that's possible...
15:36:57Maxime`MrnSlasheri: when you're in the file selection, push "a-b" in the song you want to be played next
15:37:16Slasherioh, that must be an undocumented feature..
15:37:33Maxime`Mrnit's documented lol
15:39:44MoosLinus: me i liked very much to do the A mark to the start of one song, and if i wanted to replay the 2 or track song i didn't B mark for just listen this 2 or 3 tracks...
15:41:30Moosand especially for my albums in 1 song not splited, good to select just one part or much of this foee.g
15:44:08LinusNthe a-b patch for archos can only mark within one song
15:44:20LinusNbut i guess that's a start
15:45:48Moosvery good start :)
15:46:18*amiconn has no use for A-B
15:47:27LinusNneither do i
15:48:11 Join AlbyTott [0] (~trotter@
15:48:36amiconnHmm, it seems the redundancy for the lines and points is necessary :/
15:48:48Moosi think anyone use it daily, but it can will be ocasionaly usual
15:49:22amiconnYou can't sort the lines in a way that they're oriented clockwise for all faces
15:50:15LinusNno, but if you have a list of lines for each face instead of a bitmask, you can sort that list
15:50:44 Quit Maxime`Mrn (Read error: 54 (Connection reset by peer))
15:50:54LinusNand an array of bool for the "already drawn" tracking
15:51:17 Join Maxime [0] (
15:51:25 Nick Maxime is now known as Maxime`Mrn (
15:57:50 Quit B4gder ("go go go")
16:00:48amiconnLinusN: How would you do that?
16:01:22amiconnThe problem is that the firt point of every line is sometimes the starting point of a new segment, and sometimes the end point
16:01:52LinusNah of course
16:01:57amiconnEssentially, when using the line list you'll get every point of the face twice, and have to check which one to use
16:02:05LinusNso we're back to my original idea with a list of 3 points
16:02:55amiconnThis still has the same problem, as I need _all_ points of the face for filled drawing I prefer the face list to contain all points (and all lines for wireframe with hidden line removal)
16:05:09amiconnQuite some data duplication...
16:05:10LinusNi forgot that you have a different filling algorithm
16:05:20LinusNthan in tlt3
16:05:27amiconnHow would you fill using lines?
16:05:47amiconnBasically I'm filling triangles, using the 3 corners
16:05:53LinusNtlt3 used xor filling
16:06:13LinusNmuch like the blitter
16:07:01LinusNso we drew all lines and then filled the entire frame buffer afterwards
16:07:41 Quit R3nTiL ("CGI:IRC (EOF)")
16:08:27amiconnI think that's significantly slower than drawing the filled triangle directly, at least when done with the cpu
16:08:39amiconnDid you have a look at my filltriangle() ?
16:08:46LinusNi saw it
16:09:18LinusNit depends on the size of the frame buffer and how many faces you have to draw
16:09:24amiconnThe all lines approach might be faster for faces with many points...
16:09:55amiconnThe problem is that filling an outlined area requires to read back from the frame buffer
16:10:12 Quit ashridah ("sleep")
16:10:31amiconn...and that's non-trivial with packed pixels like those in our LCD framebuffer
16:11:05amiconn...and you need to calculate a starting point inside the face
16:11:10LinusNthe filling routing in tlt3 was an unrolled EOR xxxx STA xxxx ...
16:12:14amiconnThe filling itself isn't hard I think, but how did you find the edges?
16:12:15LinusNthe trick is to carefully draw the lines, so that you always plot the complement of the colors on each face
16:12:59LinusNand only one pixel per scanline, so you don't undo the effect of the eor
16:13:23LinusNno need to find any edges
16:14:11amiconnHow did you fill then?
16:14:38LinusNyou draw all lines so that the pixels you plot will make the filling routine invert the pixels from that location
16:15:04*amiconn doesn't understand at all
16:15:05LinusNexample (from left to right)
16:15:35LinusNthe 1's are the two pixels plotted by two lines on the face
16:16:03LinusNthe filling now runs from left to right, xoring every pixel with the ackumulator
16:16:53LinusNjust like the blitter in the amiga
16:16:54amiconnIf I xor everything, I get 111111111110111111111110111111
16:17:08LinusNno, you have an ackumulator
16:17:23LinusNwhich is xored by the pixel value
16:17:31LinusNit starts at 0
16:17:56LinusNthen xor it with the current pixel and store the ackumulator at the pixel position
16:18:17amiconnAh, understood
16:18:27LinusNexcept you would to this bytewise
16:18:38amiconnWhat I cannot imagine is how this would work with multiple colours...
16:19:05LinusNit worked beautifully with the multicolor mode in the c64 because the colors were stored in pixel pairs
16:19:11amiconn...especially if some faces are already drawn, and I add the next one...
16:20:38LinusNso you would plot each pixel with the complement value of the colors on each face
16:20:55LinusNso that the xor produces the correct color
16:21:44*amiconn still can't imagine how this would work :(
16:21:49LinusNand you have to draw all lines first, then fill the frame buffer afterwards
16:22:15LinusNall lines of all faces
16:22:30amiconnHmm. Then how would you make sure every face is fully surrounded by pixels of its colour?
16:22:48amiconnThat means the lines need to be drawn twice, but not at the exact same location...
16:22:55LinusNoh no
16:23:00amiconn....or do I miss something here :/
16:23:10LinusNyou must make sure avery line is only drawn once
16:23:25amiconnIn which colur ???
16:23:42LinusNwith the complement color of the two faces on each side of the line
16:24:04amiconnHmm? The faces most likely have different colours...
16:24:38amiconnAnd why I do need the complement? Complement of what, if there are two colours...
16:24:40LinusNif the color on the left side is 01 and the right side is 10, you plot color 11
16:25:17amiconnIn your monochrome example it looks like thie inside will get the same colour as the line...
16:25:20LinusNso that color1 xor plot color == color2
16:25:36LinusNi have to run now
16:26:01amiconnThis actually might work quite fast, with our 4-pixel bytes
16:26:15amiconn...*if* I manage to understand the method correctly :/
16:26:17LinusNyes, exactly the same format as the c64 iirc
16:26:26LinusNgotta go, cu later
16:26:29 Part LinusN
16:30:46 Quit Moos (Read error: 104 (Connection reset by peer))
16:34:20 Join Moos [0] (
16:35:10 Quit alxcm (Remote closed the connection)
16:35:11 Join alxcm [0] (
16:37:13 Join redcow_ [0] (
16:37:56redcow_hi, is it possible that the delete function doesnt work on iriver unite? or at least nor for me
16:39:00redcow_if i press the play button to confirm, it goes back to the file list and does nothing
16:39:38Moospress the joystick not the play butom
16:41:16redcow_k i try
16:41:49redcow_uhm ur right, thanks
16:42:01Moosnp :)
16:42:30redcow_but a bit confusing PLAY = Yes, it means for me the play button and that is normally not the joystick
16:42:51redcow_or both should work, the joystick and the "play" button
16:43:02Moosthe keys are not the same in Rockbox
16:43:45amiconnredcow_: This is a problem of rockbox being multiplatform
16:44:14amiconnIt will be solved when the new localisation mechanism is introduced, allowing for per-model strings
16:44:28redcow_ah i understand :), another question what about the fm tuner patch from (austriancoder (?)), does it allready work?
16:47:02redcow_damn :/ fm and working navigation on remote are the only really big things that are at the moment missing
16:50:03Moosit will be done one day or other :)
16:54:37redcow_does rolo works only with rockbox patched firmwares or with originall iriver firmware ? i mean is it possible to boot into original iriver firmware while beeing into rockbox , if u understand what i mean
16:59:35*amiconn notices Slasheri didn't adhere to the long policy...
17:07:37***Saving seen data "./dancer.seen"
17:10:38 Join hicks [0] (
17:11:47 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
17:14:23 Quit redcow_ ("Verlassend")
17:15:10Slasheriamiconn: Hmm, true.. there are at least some ints that have to be changed to long if we are going to use software codecs on other that 32 bit platforms
17:17:14amiconnThe long policy should be made a part of the 'contributing' document...
17:18:15 Quit godzirra (Connection timed out)
17:23:28 Quit AlbyTott ()
17:26:03 Join asdsd____ [0] (
17:26:35bill20r3has anyone taken apart thier iriver remote? I want to shorten the cord.
17:26:52 Part asdsd____
17:40:23 Join bagawk [0] (1000@bagawk.user)
17:44:08Slasheriamiconn: Hmm, can i use gray shades in the wps bitmap files?
17:44:39amiconnNo, that's not yet implemented
17:44:44Slasheriah, ok :)
17:45:02amiconnI'll probably adapt the bmp loader, but then someone will still need to adapt the wps code
17:45:56 Join aegray [0] (
17:57:03 Nick StrathAFK is now known as Strath (
18:03:14 Join einhirn [0] (
18:05:03 Join hardeep [0] (
18:13:04 Join alxcm_ [0] (
18:13:04 Quit alxcm (Read error: 104 (Connection reset by peer))
18:13:55CassandraThis feature request about a powered down off state where we just keep the clock ticking over. Does it seem plausible?
18:18:35bagawkCassandra, the clock is always running or did I missunderstand?
18:18:36 Quit alxcm_ (Read error: 104 (Connection reset by peer))
18:23:24 Join alxcm [0] (
18:30:21Cassandrabag - not when the unit is powered off. They're asking for a software off state so the clock can be used as if it had an RTC.
18:30:41 Join cYmen_ [0] (
18:31:41 Quit cYmen (Read error: 104 (Connection reset by peer))
18:32:07bagawkCassandra, ahh so this is for irivers
18:32:45CassandraConceivably for players and recorder v1s too.
18:33:00CassandraAlthough the poster was suggesting for iRivers.
18:33:29Cassandra(Although with the Archos models, it's trickier because of hardware poweroff)
18:34:49CassandraRTC function is something I really miss on the iRiver. I like using my Jukebox as my alarm clock. So I'm wondering if it's feasible.
18:36:50SlasheriHmm, is there any way to tell where to print out the text in wps? That would allow to create a fixed size statusbar
18:37:59SlasheriI mean when i have something like "%pv %bl %?ps<on|off> %fc" on a line
18:38:24Slasheriif %pv (current volume) for example changes from 99% to 100%, the whole remaining text will misalign on the line
18:39:09CassandraWell, there's a patch in the tracker at the moment that does multiple %a tags.
18:39:18CassandraThat'll sort of do what you want.
18:39:21Slasheriah, that sounds good
18:39:42CassandraBut no, there's no specific way to place text output at a particular co-ordinate.
18:40:05CassandraIf you ask me, the entire WPS needs a rewrite.
18:40:06Slasheriok, then i will just wait for that patch :)
18:40:49CassandraSlasheri: Don't think anyone's taken it up. This could be your big chance. :)
18:47:27 Join TCK- [0] (
18:59:39amiconnI'd guess Slasheri is still busy with playback...
19:00:03amiconnBtw, playback of the last track in playlist still stops prematurely...
19:00:14CassandraWorks for me.
19:00:54*Cassandra is becoming tempted to turn all the status bar icons into placeable widgets that could be put in the WPS like graphics.
19:02:52Slasheriamiconn: yes, that's true
19:02:53amiconnI think it will be not that simple
19:03:13Slasherii just made my first wps, i will upload a screenshot :)
19:03:13CassandraI suspect the update is the tricky bit.
19:03:16amiconnCassandra: The wps would have to know when to refresh what
19:04:05CassandraI'm waiting for someone to make a WPS that doesn't make me want to throw up.
19:04:07Slasherijust modified one existing wps :)
19:04:56CassandraThat one's not so bad. I do wonder what all this obsession with knowing what the next track is is all about.
19:05:13amiconnFor me, the current wps capabilities are more than sufficient, but I agree that some people would like it more configurable
19:05:53amiconnWhat I don't understand at all is why people clutter the already small display with background graphics to have even less space for the actual information
19:05:57 Quit TCK (Read error: 110 (Connection timed out))
19:06:37CassandraEye candy.
19:06:40amiconnThat doesn't mean I'm against graphics in the wps, but I'm against useless graphics only taking space
19:06:59CassandraThe same reason my PC case is full of pointless LEDs.
19:07:28amiconnI prefer plain, simple and clear
19:07:29Cassandra(They are mostly blue though. Everyone accepts that blue LEDs are cool, right?)
19:07:38***Saving seen data "./dancer.seen"
19:08:08bill20r3of course.
19:08:28CassandraI like simple but not crowded. It should theoretically be possible to create a graphical WPS that I like.
19:08:38amiconnI think having conditional graphics to show folder/artist, filename/title etc. makes sense
19:08:58 Quit bagawk ("Leaving")
19:09:19CassandraYou mean as in building some into Rockbox?
19:09:24amiconn...and also being able to position & align by pixel
19:09:34amiconnCassandra: No, that's not necessary
19:10:02amiconnThey could be part of the wps, or of a wps bitmap
19:10:10CassandraMost people seem to have gone for the simpler approach of laying out their graphics on a BMP.
19:10:29amiconnI think about a skinning method used in a number of other programs
19:11:16amiconnInstead of having each tiny graphic as separate .bmp, have one bmp containing all of them and the wps could then use clips from it
19:12:03amiconnOtherwise a (imho) useful wps would need a whole number of tiny BMPs...
19:12:33Cassandra*nods* That's a nice idea.
19:12:53amiconnUsing parts of a loaded bitmap is a piece of cake with the new lcd_bitmap_part() function
19:17:34amiconnMy overall idea is that the new wps code should allow pixel wise positioning, and be based on a box concept
19:18:08 Part aegray ("Leaving")
19:18:19amiconnEvery element would have 2 coordinates, x and y, and each element would have a 'natural' width and height
19:19:09amiconnThe width of text elements would additionally have a limit. If the text length exceeds this limit, the box would not be made wider, but the contents would scroll instead
19:20:04amiconnThe point within the element where the x and y coordinates are bound to is variable. For each direction, it could be left aligned, centered, or right aligned
19:20:16amiconn(top / center / bottom accordingly)
19:21:41amiconnI can't think of a wps that would be impossible with that concept...
19:22:39amiconnThis concept would also work on the player, only that the positioning would happen charcell-wise
19:25:29amiconnThis would even allow to do some things on the player that are impossible now, like having the scrolling <trackno - title> beside a 1-char progress 'cup'
19:33:35Cassandraamicon: You've sold me. When are you writing it? ;)
19:37:29 Nick Lynx_ is now known as Lynx_awy (
19:39:35CassandraOh, are we likely to get 4 colour bmp support in the WPS soon?
19:40:45CassandraNew cube looks cool, btw. (Although it's still making me feel slightly dizzy.)
19:41:28amiconnThere'll soon be a new mode - wireframe with hidden lines removed - on all targets
19:42:31CoCoLUSwe'd need 2 more grayscales for a rubik cube :)
19:42:33amiconnI intend to get solid greyscale cube working on archos too, but that requires working on the grayscale lib first
19:42:53amiconn...which is my next project
19:43:04amiconn...perhaps with some intermediate bmp loader hacking
19:43:26amiconnGrr, days do definitely have too few hours :(
19:44:15 Join Bluechip [0] (
19:45:31 Join godzirra [0] (
19:46:05 Join thegeek_ [0] (
19:48:23 Join webguest40 [0] (
19:49:14webguest40with the lastest bedge, my H140 hangs on the Rockbox boot screen, I have reset, samething
19:50:43webguest40ok on the thord reset it loaded and booted normally
19:51:10 Join solex_ [0] (
19:53:14 Quit thegeek (Read error: 110 (Connection timed out))
19:55:05 Quit solex (Read error: 145 (Connection timed out))
19:55:59 Part webguest40
19:57:13 Quit hardeep ("[BX] Pretzel Boy uses BitchX. Shouldn't you?")
19:58:43 Join Stryke` [0] (
20:01:39 Quit Moos (Read error: 110 (Connection timed out))
20:08:43 Quit courtc (Read error: 110 (Connection timed out))
20:09:19 Join courtc [0] (
20:10:20 Join Febs [0] (
20:14:23 Join Moos [0] (
20:14:30 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
20:14:40 Join Stryke` [0] (
20:19:31 Join TCK [0] (
20:27:27 Join zezayer [0] (~zezayer@
20:36:45 Quit Stryke` (Read error: 110 (Connection timed out))
20:41:45 Quit TCK- (Read error: 110 (Connection timed out))
20:54:42 Join Maxime [0] (
20:55:14 Join Febs_ [0] (
20:55:40 Quit Febs (Read error: 131 (Connection reset by peer))
20:55:40 Quit Maxime`Mrn (Read error: 131 (Connection reset by peer))
20:55:41 Nick Febs_ is now known as Febs (
20:57:55 Quit Bluechip ()
21:01:48 Join StrathAFK [0] (
21:06:41SlasheriUpdated the wps, maybe even too much graphics now:
21:07:40***Saving seen data "./dancer.seen"
21:12:21 Quit zezayer (Remote closed the connection)
21:16:08 Quit Strath (Read error: 110 (Connection timed out))
21:33:48 Join zezayer [0] (~zezayer@
21:40:35 Join alxcm_ [0] (
21:40:36 Quit alxcm (Read error: 104 (Connection reset by peer))
21:51:23 Join Coldtoast [0] (
21:52:14 Quit TCK (Read error: 110 (Connection timed out))
21:55:25 Quit courtc (Read error: 113 (No route to host))
22:04:14 Quit RotAtoR ()
22:13:49 Join Bagder [0] (
22:32:05BagderSlasheri: here?
22:33:04SlasheriBagder: yes :)
22:33:32Bagderhow is the pcm_play_data() called?
22:33:43BagderI now have a function that plays a pcm buffer
22:34:09Slasherioh :)
22:34:20Slasherijust give the callback function as parameter
22:34:39Slasherithen the pcm_play_data will call that callback automatically when necessary (at start three times in a row)
22:35:40Slasherinever mind.. so you have to call get callback whenever you need to fill the audio fifo
22:35:55 Nick StrathAFK is now known as Strath (
22:36:32Bagderok, I'll fiddle around a bit and see...
22:37:09SlasheriBagder: just save the callback pointer you get with pcm_play_data()
22:37:40Slasheriand then use that callback to request more data
22:37:55 Quit zezayer ("Chatzilla [SUSE 1.0.4-1.1/20050511]")
22:37:55Bagderis there a thread started already or should I create one?
22:38:15Slasheriwhen the size is zero, there is no more bytes to give and you should stop calling the functioj
22:38:27Slasheriyou should create one as needed
22:38:35Bagderwill do
22:42:34 Quit pabs (Read error: 104 (Connection reset by peer))
22:42:56amiconnNow I have a problem: Should I change the cube.rock description in the wiki to reflect cvs, or leave it valid for the latest release version?
22:43:39BagderI say update now
22:44:15thegeek_btw, nice job amiconn:)
22:44:21thegeek_looks smooth
22:44:32amiconnthegeek_: I just committed another update
22:45:30amiconnVisible difference: Another mode (hidden lines), and the ability to pause
22:46:28amiconnThe only target where I don't need button combos for any function is the archos recorder
22:47:45amiconnI made sure that you can switch mode while paused, and the display will update according to the mode :)
22:48:58BagderSlasheri: and I start play at the first call to pcm_play_data?
22:49:56SlasheriBagder: yes, you should do that
22:50:21amiconnMaybe I should try to implement xor filling...
22:50:30Slasheriand initially you can request three chunks with the callback function if you need to
22:50:56Slasheribut you have to consume one chunk before requesting more
22:50:56BagderI'll wait and try to get sound first
22:51:06amiconn...nah, it's fast enough for now as it is. It's already very blurry on the iriver LCD at normal speed, let alone highspeed
22:51:50*amiconn goes to shoot some screen... oops
23:03:17Bagderalsa programming is not very documented
23:03:55 Join hardeep [0] (hardeeps@norge.freeshell.ORG)
23:06:22Bagdermy stand-alone program works, but the same function fails when used from within the sim ;-/
23:06:48amiconnHmm, now I have a problem with the recorder screenshots :/
23:07:17amiconnI adjusted cube.rock to the LCD aspect, but the screenshots are 1:1 so they are distorted
23:07:32Bagderit never ends ;-)
23:07:43***Saving seen data "./dancer.seen"
23:07:53amiconnI have 2 options: (1) correct the aspect with image processor, but that may add some blur
23:08:18amiconn(2) Take screenshot in sim, because aspect correction isn't done there
23:08:36amiconn...but that gives a little different impression
23:09:13Bagderhey, the ipod guys got a movie! ;-)
23:09:13amiconn+(3) Accept the distortion
23:09:20amiconnI know
23:09:27amiconn...but I have no video cam
23:17:00 Join pabs [0] (
23:25:20 Quit MrStaticVoid ("Lost terminal")
23:35:17BagderI already dislike alsa
23:40:01Bagderdoesn't work yet for me
23:45:07amiconnNice screenshots there are :)
23:47:42Bagderlinux sound
23:47:49HClfor rockbox?
23:48:01HCloh right
23:48:03HClfor the sims
23:48:04HClnice work
23:48:15Bagderit'll be great to debug things with
23:48:32Bagderlike the runtimedb ;-)
23:48:34HCl :p
23:48:40HClbut i'm gonna go shower then go to bed
23:52:52 Quit Aison ("( :: NoNameScript 3.72 :: )")
23:57:04*Bagder cries for help on the list
23:57:04 Quit thegeek_ (Read error: 104 (Connection reset by peer))
23:57:41 Join thegeek [0] (

Previous day | Next day