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 2005-03-06

00:00:16rasheramiconn: excellent
00:01:25 Quit Schee4 ("Leaving")
00:01:59amiconnBagder: Two more things, regarding the plugin Makefile. (1) What do you think about the way I added sub-dir support? (2) Is there a reason why for the simulators, the plugins don't depend on the plugin lib (which I just stumbled upon), but on plugin.h instead?
00:02:53Bagder1) haven't checked how the subdirs support works, 2) mostly because it would make all plugins to get rebuilt when the lib is changed which could be painful
00:04:03amiconn(2) is done for the target.
00:04:28amiconn...and I don't understand the dependence on plugin.h
00:04:55Bagderplugin.h sets the API
00:05:23rasheramiconn: that is most excellent :)
00:05:28rasheryou scary person
00:05:51amiconnBagder: But then the target plugins don't depend on plugin.h ?
00:05:57Bagderthey should
00:05:59 Quit frank_ ("Leaving")
00:06:12amiconn$(OBJDIR)/%.elf: $(OBJDIR)/%.o $(LINKFILE) $(OBJDIR)/libplugin.a
00:07:17amiconn...and this works better than the simulator plugin build, imho
00:07:42Bagderwell it rebuilds more than necessary
00:07:46 Join muesli- [0] (
00:07:49Bagderwhile the sim does fewer than it should
00:08:26amiconnI stumbled upon this when testing my playergfx library. When I change the library only, the library gets rebuilt, but not relinked to the plugins
00:08:44 Quit sox ("Snak 4.13 IRC For Mac -")
00:08:55BagderI don't mind seeing it fixed
00:09:09amiconnBagder: Is there a way to make only those plugins depend on the library that actually use it?
00:09:34Bagderwe would have to come up with a way to "detect" those plugins
00:09:51Bagderand generate dependencies based on that
00:10:09amiconnHmm. Same goes for the codec test plugins. The plugins don't pickup codec changes
00:10:13Bagderfor now, I think we should make them depend on plugin.h and the lib
00:10:39Bagder... but we simply cannot make all plugins depend on all codecs
00:11:11 Quit muesli- (Read error: 113 (No route to host))
00:11:27Rickis iriver little endian or big endian?
00:11:46 Join [Zmaj] [0] (
00:12:29 Join amiconn_ [0] (
00:12:34[Zmaj]hi all
00:14:07 Join matsl [0] (
00:14:40 Quit amiconn (Nick collision from services.)
00:14:40 Nick amiconn_ is now known as amiconn (
00:15:26amiconnBagder: The codec plugins could be detected by a pattern match. They all match *wav.c
00:16:21amiconn..and the plugins depending on the plugin lib have one thing in common - they include headers from the lib directory
00:17:00Bagderall that still forces us to do hairy scripts to generate dependencies
00:17:16Bagderbut I guess we will end up there sooner or later
00:17:29 Join condor9 [0] (
00:18:30amiconnThe compiler can do this for compiling the .c files. Couldn't that mechanism be 'misused' for the linkage stage somehow?
00:18:59Bagderno, the compiler only does it for include files
00:19:10preglowisn't there some other way of hard coding what a codec needs? not right now, but when they're codec plugins, not ordinary plugins
00:19:40pregloware there anyone at all working on getting the codec api up and going, btw?
00:19:58Bagderpreglow: not that I know of
00:20:04preglowwe really need that :/
00:20:26preglowlinuxstb said he'd take a look at it, but seems like he's loaded with other work these days
00:20:40condor9amiconn: I fixed my partition table problem w/2.4. Thanks for your help.
00:21:07preglowbeing able to actually use the codecs would be a big motivating factor
00:21:23amiconnBagder: Anyway, when all plugins depend on the plugin lib, they all get relinked when the lib changes, but linking is fast... The way it is now, I often end up rebuilding *much* more than needed, to avoid unwanted effects: "make clean; make"
00:21:48Bagderlet me repeat: "for now, I think we should make them depend on plugin.h and the lib"
00:23:04amiconnYeas, okay. Hmm, I still don't get why they should depend on plugin.h This is the linkage stage, and if plugin.h changes, the .o gets rebuilt anyway, by the generated dependencies
00:23:22Bagderhm, true
00:23:30Bagderthe dependency on plugin.h will never make a difference
00:23:57amiconnbut it's one more check for 'make'
00:24:51amiconnokay, fixing...
00:29:52 Quit jyp ("poof!")
00:31:05 Join muesli- [0] (
00:33:41*rasher watches logo.rock on playersim
00:33:58amiconnWell, it's a bit flickery and slow on the sim
00:34:25preglowmpegdec is also a rather noisy decoder
00:34:30preglowi seriously doubt it is compliant
00:35:14rasheramiconn: looks fine here (x11 sim)
00:35:20amiconnpreglow: How does one test compliance?
00:35:41preglowamiconn: by using a reference waveformen and computing rms
00:35:42amiconnrasher: I tried all of win32 sim, x11 sim and target. Target looks best :)
00:35:56rasherof course
00:35:57preglowamiconn: i just did a simple waveform subtract, there's lots of noise left
00:36:18amiconnpreglow: Is there a test program and reference .mp3 ?
00:36:33preglowamiconn: yes, libmad links to them on the compliance page
00:38:21amiconnHmm. Iirc mpegdec is the Stephane Tavenard thing, so it is based on mpega.library for the amiga (which is pure 68k asm).
00:38:44preglowtavenard, yes
00:38:52amiconnI could perform a compliance test with my Amiga, to sort out possible porting problems
00:39:22preglowwell, is there much point? it's slower than libmad
00:40:34 Join muesli_ [0] (
00:40:41preglowand the source code is a nightmare
00:40:42amiconnI don't konw, you're the emac man. Libmad was slower than mpegdec is now before you added emac, right?
00:40:59preglowi can check that in a jiffy
00:41:59 Quit methangas (" HydraIRC -> <- 100,000+ downloads can't be wrong")
00:42:02amiconnI know that mpega.library (~mpegdec) is twice as fast as libmad on Amiga.
00:43:22preglowwell, small wonder, it's asm
00:43:37 Quit muesli- (Read error: 60 (Operation timed out))
00:44:38preglowit's slower, about 5% so
00:46:19preglowmpegdec is also much smaller
00:46:29preglowso it might of course be good, i don't know
00:51:14preglowthere's also some 68k asm bundled here, but it's written for an amiga, so probably won't work on coldfire
00:52:34 Join coob [0] (
00:52:35 Join Bonkers [0] (
00:52:36Ricktry it ;P
00:52:40Rickgot nothing to lose.. ;0
00:54:59preglowit won't compile with gas, don't worry
00:55:03preglowi'll have to rewrite it
00:56:25 Quit condor9 ("Leaving")
00:57:12 Join ripnetUK [0] (
00:57:21preglowbut i don't know, someone else have a look at the code and comment
00:57:43preglowi don't quite feel like optimizing libmad if we're not going to use it
00:57:49coobpreglow: did you see that ALAC stuff?
00:57:56preglowcoob: ehh?
00:58:07amiconnpreglow: I found the compliance test files, but for the software: "Because this software was developed under ISO copyright, it is not possible to open this directory for public access." :(
00:58:32preglowamiconn: ...........
00:58:51preglowcoob: don't ask me, i don't care about that
00:58:51BagderI added that alac link to the wiki already
00:58:59coobah k
00:59:05coobthought i might be a bit late.
00:59:06preglowi've got my hands full with libmad, flac and tremor
00:59:17preglowthat's the ones i personally care about, and that's more than enough work
00:59:30Bagderamiconn: I think I have a working approach for better dependencies for the plugins
01:04:49***Saving seen data "./dancer.seen"
01:06:18 Quit muesli_ (Read error: 110 (Connection timed out))
01:06:57 Quit ripnetUK ()
01:07:39amiconnBagder: Now I know why the common simulator lib gets checked twice.
01:08:36amiconnOnce it is referenced in apps/Makefile, and once in the respective simulator makefile. However, I don't know which one can be removed
01:10:29 Join muesli- [0] (muesli_tv@
01:25:30 Part amiconn
01:38:47 Join Bluechip [0] (
01:46:33 Join Sucka [0] (
01:51:59preglowi'm off
01:52:07 Quit preglow ("bam")
01:55:43 Quit stevenm ("Leaving")
01:56:41RickI just realized
01:56:56Rickthere's a "bug" with markun's grayscale patch
01:57:08Rickit doesn't reset fg/bg back to 11/00
01:57:15 Join webguest38 [0] (
01:57:15Rickafter plugin exits
01:59:58 Quit webguest38 (Client Quit)
02:00:09 Join nobby [0] (
02:07:24 Quit toolmanwv ("—I-n-v-i-s-i-o-n— 2.0 Build 3515")
02:13:27 Join Sando [0] (
02:13:40 Join Strath [0] (
02:18:25Ricksimple enough ;P
02:20:03 Join ashridah [0] (
02:20:38Rickgreets ashridah
02:25:55 Part Bluechip
02:30:02 Quit matsl (Remote closed the connection)
02:34:11 Join midk [0] (
02:34:54nobbyanyone know how iriver audio progress is going?
02:36:01nobbyrockboy seems to be getting more attention.. :(
02:37:09Rickthat's a bad thing?
02:38:33ashridahnobby: only by one or two people. some of the bits in rockboy needed to be attended to anyway (decent LCD driver, for instance)
02:38:39midkgaming vs an os for the iriver? i'd say. :P
02:38:48ashridahbut the simple matter is that optimising codecs takes time.
02:39:24Rickis HCl even working on codecs?
02:40:38 Join AC [0] (
02:45:34 Part nobby
02:46:28Camilohi AC
02:48:10ACany news?
02:48:17 Quit Sucka ("a bird in the bush is worth two in your house")
02:48:23Camilonot from me :) I'm a spectator
02:49:57ACi am wokring on remote lcd
02:50:16Camiloreally, what's the state of that?
02:50:29ACvery beginning
02:50:41ACfirst thing is to get the backlight flashing
02:50:52Camilohow does it work? serial comms?
02:51:34Camilowhat speed? does it matter
02:52:01ACspeed.. i dont know, but it will be fast enough
02:52:17CamiloI take it the player is the master?
02:52:57ACclock comes from cpu
02:53:19CamiloI wonder if both ends can be masters
02:53:43ACno.. lcd is slave
02:53:54CamiloI had to debug some spi driver at work last week
02:54:03Camiloit was a pig
02:54:24ACi have never done spi.. my first try :)
02:54:40Camilohave you got a scope?
02:55:44 Join geoff_o [0] (
02:55:56ACscope? englisch is not my motherlang
02:56:00_Lucretia_hi all, I've started an Xclef MT-500 wiki page
02:56:35ACoh.. now i dont have one
02:56:54Camilowell I hope you manae without then :)
02:57:13 Quit Ka (Read error: 110 (Connection timed out))
02:57:15ACi hope too
02:58:26ACi must emulate spi, because the lcd is connected directly to the cpu
02:59:15Camilothat sounds painful
02:59:37ACpainful but interesting
02:59:40CamiloI was using mpc8270, with spi controller onboard
03:00:17Camiloso you are just using GPIO to drive the LCD?
03:00:47ACthere is no other way..
03:00:57ACi was told
03:01:11Camilowhich GPIO pins are you using?
03:02:40Camilothanks, I'm looking
03:04:41 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
03:04:51***Saving seen data "./dancer.seen"
03:13:11 Quit midk (Read error: 54 (Connection reset by peer))
03:13:13 Quit ze (Read error: 104 (Connection reset by peer))
03:15:43_Lucretia_I've just been reading something in the FAQ #17 about the MAS chip, anyone interested in this?
03:17:36Camilogood luck AC, I'm off now
03:18:23 Join ze [0] (
03:18:35 Quit Camilo ("Chatzilla 0.9.67 [Mozilla rv:1.8a6/20050111]")
03:27:33 Join stevenm [0] (
03:30:22 Join bagawk [0] (~Lee@bagawk.user)
03:30:29 Part geoff_o ("Kopete 0.9.1 :")
03:39:32 Join jpburton5150 [0] (
03:41:20jpburton5150question: is there some reason the grayscale patch hasnt been applied to cvs yet?
03:46:07jpburton5150cus i think it should be...
03:47:11 Quit AC ("CGI:IRC")
03:50:01bagawkWho knows
03:50:07bagawkMaybe it is not stable
03:51:21ashridahjpburton5150: there are issues to resolve for the older platforms first.
03:55:28jpburton5150i see.. thats what i figured actually
03:59:10 Quit bagawk ("Leaving")
04:05:23 Join QT_ [0] (as@area51.users.madwifi)
04:17:13 Quit rasher ("CGI:IRC")
04:22:13 Quit QT (Read error: 110 (Connection timed out))
04:26:31 Join Ka [0] (
04:28:49 Quit muesli- ("ich will Kühe!!!")
04:40:40 Quit jpburton5150 ()
04:50:39Ricklooks like my bitmap loading... could work
04:53:44ashridahheh. that's it, stay positive :)
04:53:57stevenmGoddamnit. Not again.
04:54:22stevenmCan someone please explain to me why a musician would want to patch something into channel 9 when making a MIDI ?
04:56:10ashridahrofl. for a second then, i thought you meant the local free-to-air tv channel here called 'channel 9'
04:56:13stevenmYeah. Drum release, for the MIDI codec. Disable looping and release for channel 9, to just let it run out and shoot down the voice. But no, somehow there's a program event there.. halfway thru this file, you get a SEA of sound
04:56:49stevenmI just disabled all patching of ch9... it's not even supposed to program it if such a drum exists
04:57:21stevenmYet, halfway I get a sound overflow, with all notes on ch9 / note=27, 31, etc. Yeah, the first valid drum here is 38
04:58:02*ashridah blinks
04:59:26*stevenm bangs head against wall repeatedly, causing a nonlooped precussive sound
05:03:12stevenmthere, finally. this is nuts.
05:04:53***Saving seen data "./dancer.seen"
05:10:33Rickmade two 'splash' bitmaps for my game
05:12:03stevenmdoes anyone knows when linus or stripwax will be back ?
05:12:54RickI don't think linus was even on today
05:15:02 Join ze__ [0] (
05:16:41Rickhow is the bitmap stored internally? from bottom left or bottom right?
05:16:44Ricki'm guessing bottom right
05:26:02Bonkersnormally the two choice are top left or bottom left
05:26:53 Quit ze (Read error: 110 (Connection timed out))
05:26:54 Nick ze__ is now known as ze (
05:35:28Rickwell, someone earlier told me that the bitmaps internally on the iriver are stored upside down
05:36:49Bonkerswell upside down depends on your point of view
05:37:40Rickmy point of view = what I see on the screen. ;P
05:39:02Bonkersheh, well I meant that still doesn't clarify if the origin is the top left or bottom left
05:40:26Ricki'll just see how it turns out when I draw the bitmap then
05:43:31stevenmRick, I remember at some point running into a problem where regular .BMPs were stored upside-down
05:44:48 Join Lurkski [0] (
05:47:36stevenmJust not sure if it's all of them, or what. that was a long long time ago
05:49:18 Part Lurkski
06:01:18Rickgot my bitmap exporter going
06:26:28Rickwhoo, stupid bug prevented it from finding the loaded bitmap
06:26:31*Rick kicks it
06:27:15*Rick goes to make a fix
06:30:26 Join ze__ [0] (
06:31:53RickI feel silly now
06:32:09RickI generated a grayscale bitmap instead of a b&w one
06:41:47 Quit ze (Read error: 110 (Connection timed out))
06:41:47 Nick ze__ is now known as ze (
06:41:48 Quit DMJC (Read error: 54 (Connection reset by peer))
06:44:53 Nick [Zmaj] is now known as [zmaj] (
07:03:17Rickthere's an internal limit on how many file handles can be opened?
07:04:12ashridahwell, yeah, since file handles consume memory
07:04:55***Saving seen data "./dancer.seen"
07:05:00ashridahsince there's not really a malloc available to rockbox, there's probably a set number defined at compile time.
07:05:11ashridahthe malloc that's been added is relatively new.
07:06:19RickI was just wondering, I ran out because I forgot to close one at the end of my load sequence
07:06:40Rickso eventually I couldn't run my plugin anymore
07:10:13Rickokay, from what I see
07:10:49Rickinternal format is made up of 8x8 blocks (each 8 bits representing a vertical line of 8 pixels), going vertically then horizontally
07:48:35Rickit works
08:21:16 Nick [zmaj] is now known as [Zmaj] (
08:38:07RickI thought button_get(true) implied a yield?
08:39:21RickI think I confused something
08:59:14 Join methangas [0] (
09:04:58***No seen item changed, no save performed.
09:11:16 Join jyp [0] (
09:32:29 Quit stevenm ("Leaving")
09:48:41 Quit jyp ("poof!")
09:50:32 Nick [Zmaj] is now known as [zmaj] (
10:00:41Rick'sup HCl
10:02:15 Join lostlogic [0] (
10:19:28 Join R3nTiL [0] (~zorroz@
10:21:31 Nick Strath is now known as StrathAFK (
10:26:40 Join amiconn [0] (
10:31:40 Join Patr3ck [0] (
10:37:42 Join ze__ [0] (
10:39:59Rickla la la
10:48:33 Join webguest55 [0] (
10:48:41 Quit ze (Read error: 110 (Connection timed out))
10:48:42 Nick ze__ is now known as ze (
10:50:08 Join muesli- [0] (muesli_tv@
10:55:03 Quit R3nTiL ()
11:05:00***Saving seen data "./dancer.seen"
11:11:25 Nick [zmaj] is now known as [Zmaj] (
11:32:34 Part amiconn
11:33:54HClis there any way to tell cvs to not commit any changes in a certain file?
11:34:03*Rick is working the kinks out of rbounce
11:34:08Rickhmmm... dunno
11:34:57Rickmy game is progressing well
11:35:10HClwhat game is it?
11:35:16Ricki'll take a dump
11:35:21Rickscreendump that is
11:36:57BonkersHCl: never for a certain file or jsut one time?
11:38:18HClnever for a certain file...
11:38:28HClbut i guess i can just do cvs commit <filenames i want to commit >
11:38:33BonkersI think there's a file called .cvsignore that you can create
11:38:41RickI need a paused key
11:38:44HClokay :)
11:38:48Rickwould make taking screenshots easy
11:39:10Bonkersyep, .cvsignore
11:43:10*HCl gets confused to what game it is o.o;;;
11:43:18Rickit's a game where balls bounce around
11:43:23Rickand you have to fill up the area
11:43:29HClokay o.oo.
11:43:30Rickwithout them hitting the new wall
11:43:44Rickthere's a kde version of the game called kbounce
11:43:44HCli'll see when its in cvs, i guess ^^;
11:43:49muesli-Rick are you working on rockboy?
11:43:54Rickthat's HCl
11:44:13HClyea, thats me.
11:44:14HCl :p
11:44:29muesli-ah ;-) so HCl are there improvements of speed possible?
11:44:33Ricktrying to figure out a bug atm that causes the walls to stop prematurely
11:45:22Rickoh, haha, I think I know what it is now
11:45:45HCli'm working on a prototype of dynarec
11:47:36Rickyep, I think I figured it out
11:48:49Rickokay, that was a very stupid bug
11:50:25Ricknow I need to figure out an algorithm to search for holes and close them up
11:55:22BonkersRick: jezzball?
11:56:10 Quit webguest55 ("CGI:IRC")
11:56:12RickBonkers: that was the name of a TI calc version, yeah
11:56:22BonkersI think it was the windows version's name first
11:56:27RickI dunno ;O
11:56:32Bonkersit was an old microsoft game that came with a gamepack
11:56:41Bonkersgreat game though
11:58:23Rickif you want to see
11:59:03BonkersI'll have to check it out tomorrow, I've to to head to bed, it's 6am here
11:59:18Rick(It's 3am here ;P)
12:00:00muesli-12am ^^
12:06:47 Quit [FlaT]Heidel (Read error: 60 (Operation timed out))
12:20:30 Join jyp [0] (
12:21:07 Join [FlaT]Heidel [0] (
12:21:11 Nick [FlaT]Heidel is now known as Heidel (
12:22:17 Join cYmen [0] (
12:23:54 Join [IDC]Dragon [0] (
12:24:08[IDC]Dragonhi there
12:24:18[IDC]DragonHCl, do you read?
12:25:54 Nick QT_ is now known as QT (as@area51.users.madwifi)
12:34:23HCli do
12:34:26HClwhats up?
12:34:42[IDC]DragonI've leeched all your ROMs ;-)
12:34:59ashridahknew i forgot something
12:34:59[IDC]Dragontrying out an emulator called bgb now
12:35:15[IDC]DragonI was about to ask you for a good one
12:35:23HClthat reminds me, *deletes interstella 5555 off his ftp*
12:35:25HClfor windows?
12:35:30[IDC]Dragonbut this looks pretty good
12:35:31HClvirtual boy advance
12:35:36HClis the best
12:36:14HClor was it called something else :x
12:36:32coobvba is neat
12:36:33HClvisual boy advance
12:36:35HClnot virtual
12:36:36HCl :)
12:36:38ashridahinterstella 5555?
12:36:43*[IDC]Dragon googles
12:36:44coobneeds scale2x though
12:36:54coobgrab the snes9x with scale2x stuff in it you can
12:38:07[IDC]Dragonbeta or release?
12:42:21[IDC]Dragon1.8 beta works for me
12:43:07Rickfroze my h120
12:43:11*Rick looks for a thumbtack
12:43:18HClashridah: its a movie of daft punk, its their entire album discovery describing an anime story
12:43:59webmindHCl, oe...
12:44:02webmindHCl, wanna ?
12:44:08webmindcan I have a copy?
12:44:10HClwhat? o.o
12:44:14HCloh. sure.. its on cnet..
12:44:28*webmind logs in
12:44:44HClits pretty nice since it has no speech whatsoever
12:44:54HClaside from the lyrics of the songs, which aren't much, really.
12:45:13HClyes it still describes a nice story
12:47:22 Join Sucka [0] (
12:50:50 Join SeeSchloss [0] (
12:59:38 Quit [IDC]Dragon ()
13:05:02***Saving seen data "./dancer.seen"
13:13:24DeadManone more time....
13:13:39DeadManI have the movie
13:14:06DeadManGreat oldskool style anime
13:14:23DeadManReminds me of Battle of the Planets :)
13:14:45RickI think I just damaged my reset button
13:14:55DeadManhow so?
13:15:00Rickpushed on it too hard
13:15:04Rickit 'popped'
13:15:18Rickout of place
13:15:20DeadManmight need to open it up and take alook
13:15:33DeadMangot a torq screw driver?
13:16:51RickI guess that's enough testing until tommorow when I can see it (and open it up, hopefully)
13:17:09DeadManthe reset is really poorly designed
13:17:33RickI figure I should be able to see it by just taking off the bottom
13:17:45DeadManit should be a recessed button not a button inside with room to stick things like needles either side of it
13:18:12RickI really hope I can fix it :(
13:18:28DeadManprobably can if it's not broken off
13:19:02DeadManthey should supply you with the proper sized tool....silly iRiver
13:19:04Rickit feels like it was just pushed aside
13:19:10ashridahheh. i think i've used that reset button once up until i installed rockbox
13:19:56Ricki've been trying to track down an infinite loop bug in my code
13:20:25crwli needed the reset quite often but then I fixed those files that caused hanging
13:20:35crwlthe iriver firmware crashes with ogg files done with a certain oldish encoder
13:20:55crwlsome cvs version between libvorbis rc2 and rc3 didn't set the nominal bitrate header, which causes the crash :P
13:21:06ashridahcrwl: interesting. haven't had that myself, but most of my ogg encoding is relatively recent
13:21:25crwlashridah, it might be that no release version of libvorbis created files like that
13:21:35ashridahi think i've run across one mp3 that caused the playback to hang, although not the entire player
13:21:37crwlbut I happen to have a couple of albums
13:21:39DeadManit's best to use a metal rod that is thing enough with a flat head rather than a needle. A paperclip should work though
13:21:50DeadManthin not thing
13:21:54Rickgood idea
13:21:57Ricki'll do that tommorow though
13:22:04Rickhopefully I can take the bottom off and take a look
13:23:10Rickloooks like the reset button is that white thing on the bottom 'right center'
13:24:00Ricki'm off to sleep
13:24:05DeadManlittle black plastic push in
13:24:20Rickhave you taken yours apart?
13:24:41DeadManlooks like it though from that pic
13:25:05Rickhopefully it's like a plastic bit that can be positioned
13:25:19Ricknot something that snapped/bent off
13:25:35 Join amiconn [0] (
13:25:41Rickanyway, gnight
13:38:35 Join funkymonkey [0] (
13:38:36 Quit methangas (Read error: 104 (Connection reset by peer))
13:39:36 Join Schee [0] (
13:43:44 Quit ashridah ("sleep")
13:47:49 Quit SeeSchloss (Read error: 60 (Operation timed out))
13:59:16 Join Tomas2 [0] (
14:03:21Tomas2Mostly American's here?
14:03:22HClpeople on internet who think its fine to be rude and mean cause its "just the internet" are really frustrating and annoying -.-
14:03:28HClmostly european, i think.
14:03:34Tomas2oh ok...
14:03:35HCli'm dutch, like you.
14:03:40Tomas2I tought everybody was sleeping ;)
14:03:46Tomas2A cool
14:03:55HClbut at the moment i'm just rather frustrated
14:04:16 Join ze__ [0] (
14:04:26HClpeople being mean and rude to me, cause they feel its "just the internet"
14:05:06Tomas2yeah... I'm irc-operator on activenet... It are mostly little kids.. 13 or 14 y/o
14:05:27Tomas2They don't understand that there are real people with real feelings on the other side
14:07:38Tomas2But I was searching for a rockbox plugin developper?
14:08:01Tomas2Is there anybody here who devved a plugin/app for rockbox jet?
14:08:59HCli made rockboy...
14:09:03HCli guess that counts? :3
14:09:50Tomas2(I saw your hostname is ending in, what studie are you doing? some technical?)
14:10:33Tomas2I am a reasonable c++ coder... so it must be possible to write some C plugin's for rockbox...
14:11:10HClhehe :) computer science..
14:11:29Tomas2Yeah, tought so
14:12:14Tomas2I'm in VWO-5 so looking for some intresting studie..
14:13:06HClto be honest. i think compsci is sometimes rather boring.
14:14:00Tomas2well.. I live close to delft, so I have been looking there... and they have "Technische informatica" but that's starting to code all over again.. so i was more looking at aerospace design...
14:15:05Tomas2Officially called "Aerospace Engineering"
14:15:45Tomas2but I was searching something completely differend...
14:15:47 Quit ze (Read error: 110 (Connection timed out))
14:15:48 Nick ze__ is now known as ze (
14:15:57Tomas2I wanted to ask you some things about rockbox plugins
14:16:02HClwell, compsci takes it slow at first, but thats mainly just the first semesters of the first year
14:16:06HClit gets complicated quickly
14:16:14HClwith compiler design and artificial intelligence in the second year
14:16:44Tomas2hmz... sounds more intresting than "technische informatica" in delft...
14:17:02Tomas2maybe I'll have a look at the tutwente...
14:20:31Tomas2But is there an example small plugin? To find out how to interface works, and things like that?
14:20:40amiconnHehe... Mario on speed now in the win32 sim.... I think too fast to be playable
14:21:11muesli-am just playing sm, but its a little bit too slow
14:21:38HClTomas2: helloworld.c
14:21:39coobjust clcoked mario with no warps :<
14:21:52coobmario award, woo.
14:21:54HClTomas2: there's docs/PLUGINAPI that describes the interface
14:21:58amiconnI optimised lcd_update() and lcd_update_rect() to invalidate only the window portion that actually changed. It used to always invalidate the whole window before
14:22:01Tomas2ok, missed that one
14:22:20Tomas2I'll have a look at the API doc's
14:22:26CoCoLUSis that only for the win32 sim or for the iriver too?
14:23:09Tomas2what emulator are you using? Iriver? Or just an Archos?
14:23:26amiconnWin32 simulator only. The target always did that
14:26:29Tomas2a full cvs checkout takes some time...
14:31:53Tomas2HCl? What hardware are you using yourself? Iriver of Archos?
14:32:08HCli didn't see much of practical use of an gameboy emu on archos
14:32:11HClbut ami ported it
14:32:34Tomas2H120 or H140?
14:32:45HCli got an h140, the one in the pics at
14:33:21Tomas2ah ok
14:33:26Tomas2I have the other colored H120
14:33:57*HCl goes to continue playing mario, got to the air level
14:34:10Tomas2running in simulator or on your iriver?
14:34:25HClobviously, i shouldn't run the testing dynarec version for that.
14:34:31HCli don't have a simulator
14:34:53Tomas2ok, so you're testing all code on iriver?
14:35:05HClrockboy actually ran on the iriver
14:35:10HClwhile it broke on the simulators
14:35:17HClbefore it was fixed
14:37:00HClit only really became feasable to test and work on iriver when usb support got added
14:38:41HClah, nevermind, not important o.o
14:43:10HCli finished super marioworld
14:43:11HCli think
14:43:54HClyup :)
14:43:56HClending credits
14:44:22Tomas2hm.. I don't get the gnu version to work on windows :X
14:46:56coobthere's mario world on gameboy?
14:47:09coobthe one with yoshi?
14:47:23HCldunno, i have a rom named mario & yoshi, haven't tried it yet
14:47:37HCltop score
14:47:41coobsame :D
14:47:51coobexcept i was playing the colour version
14:47:57coobsuper mario dx or whatever
14:48:45HClfor some reason i can't seem to be able to get past the title screen in mario & yoshi
14:50:14 Quit Sando (Read error: 113 (No route to host))
14:51:32 Join preglow [0] (
14:53:08muesli-is it already possible to use the rockbox built in fw-flash option?
14:53:22HClnot on iriver
14:56:03muesli-for comfortable flashing an software reset option would be great
14:56:28HCljust stop and power on again..
14:56:42muesli-otherwise yepp
14:56:43HClsoftware reset is for lazy people :p
14:56:51muesli-exactly :D
15:05:05***Saving seen data "./dancer.seen"
15:16:06_Lucretia_are there any flash-based players supported? I can't find any information on the website...not HDD players, pure flash, thanks
15:17:00amiconn_Lucretia_: The Archos Ondio is flash based & supported
15:18:41_Lucretia_amiconn: is that the only one?
15:18:58_Lucretia_amiconn: was there any help from the manufacturer?
15:19:04amiconn(so far)
15:19:25Bagderno manufacturer ever helped Rockbox
15:19:30_Lucretia_what about iRiver, I think I read somewhere a while back that they had set up an OSS site
15:19:38_Lucretia_Bad: doesn't surprise me
15:19:48_Lucretia_Bad=Bagder :-)
15:20:10Bagderwell, we aren't surprised either, we don't expect help
15:20:19Bagderfrom them
15:23:08amiconnBagder: A question concerning file naming: I'd like to add a kernel.c equivalent for the x11 sim, but I'm not sure how to call it. The win32 sim calls it's kernel.c equivalent kernel.c, while it calls the thread.c equivalent thread-win32.c. The x11 sim has a simple thread.c ...
15:24:25BagderI have no real opinion on that
15:24:53BagderI think I'd prefer the same name without prefix
15:25:05amiconnHmm, okay. I'll just use kernel.c then
15:26:41amiconnAnother thing: The simulator dependencies are twisted. If I change something in uisimulator/win32 (or uisimulator/x11 respectively), then just 'make', libsim.a does get rebuilt *after* the executable gets linked. Very odd.
15:26:58Bagderyes the deps are not working completely
15:28:47_Lucretia_well, like I mentioned yesterday, I'm looking at the Xclef MT-500 flash player
15:29:59_Lucretia_given that the player can be flashed over the USB cable, what kind of possibility do you think there is of getting more info from the device over the USB cable and a some sort of server program on a Linux box?
15:30:46_Lucretia_The firmware starts at 0x2000, and the interrupt vectors start at 0x0, plus there's probably more fw after the vectors
15:31:11*_Lucretia_ is not a hardware reverse engineer, just a programmer
15:31:19*Bagder detects those now removed questions in lcd-win32.c... :-)
15:33:02amiconnBagder: Nice, eh? lcd_update() is also faster now. Before this change it used to invalidate the whole window, not only the whole lcd rectangle
15:33:37amiconnWhen I now run mario in win32 sim recorder rockboy, it's so fast that it isn't playable...
15:34:36BagderI guess the Sleep() was in there to slow it down
15:35:21amiconnI'm currently looking at the x11 sim, first thing is to make it more realtime (correct tick and sim_sleep() )
15:35:49amiconnWhen I got this to work here, someone should test on a linux box....
15:36:14*_Lucretia_ shrugs
15:36:20amiconnNext will be the button handling, then I'll take a look at how x11 handles bitmaps
15:36:33Bagder_Lucretia_: those things are very hard just to guess
15:36:51Patr3ckif someone is interested, i submitted a patch for manual profiling for the ...2wav plugins / libraries
15:41:45BagderPatr3ck: I'm sure it'll be useful
15:42:19Tomas2amiconn: I can test on linux if you want
15:42:37amiconnIt's not ready yet...
15:42:51Patr3ckBadger: as far as I understand there is still need to optimize the codecs
15:42:51Tomas2no, hilight / privmsg me if it is..
15:43:09BagderPatr3ck: yeps
15:43:37amiconnl8erz... gotta leave.
15:43:44 Quit amiconn (" bye")
15:44:42Patr3ckBadger: The people optimizing the codecs will see the patch?
15:45:18Bagderwe get mails from the tracker when new patches are submitted
15:46:05Patr3ckok, fine
15:46:09Tomas2hmz... how do I use the simulator?
15:46:19Tomas2make install
15:46:26Tomas2of something totally different? :)
15:46:36Bagderthen start it ./rockboxui
15:46:52Tomas2ok, I'll build it.. for if amiconn is ready :)
15:47:31Patr3ckI am wondering what I could do to help more
15:47:43Tomas2Are you good at C?
15:48:22Tomas2Maybe we can make a plugin request page or something like it.. and start making the usefull one's
15:48:42Patr3ckMe? What is good ;-) I did a lot of c++ during the last 1,5 years
15:49:27Tomas2yeah, me too... but C isn't to difficult to do then...
15:49:53Patr3ckyes, except from the stuff that directly programms the hardware
15:49:59 Join gargantua [0] (~vinceteam@
15:50:37Bagderthere are also hundreds of open bugs and hundreds of feature requests...
15:50:42Tomas2yeah, that's why I wanted to start with plugin's
15:50:50Tomas2then we don't have to know much about the hardware
15:51:04Bagderplugins are very easily made
15:51:35Patr3ckI thought more about helping the progress of the iriver port, even if its hardware related
15:51:56Tomas2Well.. I think you should start at the bug list then?
15:52:03Bagderthat ThingsTodo page lists numerous iriver related things
15:53:17Tomas2yeah... Can I use the simulator to adjust the screen todo's?
15:53:30Patr3ckI could imagine the "Firmware-patching GUI application" would be something
15:53:34BagderTomas2: yes
15:54:07preglowPatr3ck: i'm working on that
15:54:09Tomas2I'll take a look at the games that need to be changed for iriver
15:54:15preglowPatr3ck: and it's done
15:54:21Patr3ckpreglow: :-)
15:55:14Patr3ckSo only games left... hm
15:55:52Patr3ckpreglow: do you use a GUI toolkit?
15:55:54Bagderthe sound codec architecture is the biggest job still not done
15:56:05Tomas2and the hardest?
15:56:25BagderTomas2: possibly, optimizing the codecs is a pretty complicated work too
15:56:48Patr3ckWell the design is there:
15:57:11Patr3ckNa, wrong page:
15:57:14preglowPatr3ck: win32 api directly
15:57:48Tomas2I'll take a look at that flipit game..
15:58:11preglowi figured windows would be the platform most people would be on, and i already had a framework lying on my disk
15:59:08Tomas2maybe someday find someone to do in in java?
15:59:13Patr3ckpreglow: sure, cross-plattform is always difficult
15:59:29CoCoLUS<fn~Bagder> the sound codec architecture is the biggest job still not done
15:59:30preglowPatr3ck: and you'd have to bundle along large dlls and shit for a small patcher
15:59:35CoCoLUSwho's working on that one?
15:59:42preglowCoCoLUS: no one, afaik
16:00:17preglowlinuxstb said he'd look at it, but that's over a week ago, and nothing's happened, so i assume other work caught up to him
16:00:31Patr3ckpreglow: wondering if this is true for wxwindows as well
16:00:56preglowPatr3ck: i don't know, i didn't even investigate it, as i had most of the code just lying around
16:01:00preglowright now the patcher is 50kb
16:01:08preglowwith bootloader.bin in the exe
16:01:31preglow57kb, i mean, it got larger with md5sum checker
16:02:05Patr3ckthis is really small
16:02:21CoCoLUSmaybe some more people should look into sound codec architecture? :)
16:02:37preglowwell, it's not a lot of code, just basic gui handling, iriver.c and mkboot.c from tools/, and md5sum code from some other place
16:03:28Patr3ckCoCoLUS: This is such an integral, important part...
16:04:01Tomas2where was that wiki plugin overview?
16:04:06CoCoLUSso everyone is afraid of tackling it? :)
16:04:09Tomas2can't find it anymore :)
16:04:25Patr3ckpreglow: I could imagine if there is need for a linux app, it could be as well be implemented native
16:04:42Bagderafraid, no, lack of time to do it yes
16:05:14preglowCoCoLUS: it's also one of those things that has to be well thought out
16:05:44preglowbut feel free to think up suggestions and put them in the wiki
16:07:33CoCoLUSisn't there some rather well "thought out" variant in the wiki?
16:07:43CoCoLUSwith the double buffering system?
16:07:54Bagderits is not complete
16:08:00Patr3ckMaybe a quick and dirty prototype of the design would be helpfull
16:08:04Bagderand there probably needs to be 3 buffers
16:08:34BagderI'd say most things would be helpful
16:08:48Bagderjump in!
16:09:50Patr3ckBadger: :-), starting to really think about some sort of prototype
16:10:21BagderI'm a firm believer in iterative development
16:10:27 Quit _Lucretia_ (Read error: 60 (Operation timed out))
16:10:33Bagderso anything done now can always be improved
16:10:40Patr3ckMe too, but were are the unit tests ;-)
16:11:43Bagderiterative doesn't require unit tests
16:11:56Bagderbut it would be useful, sure
16:12:27Patr3cksure, but the feeling of did I break something when changing is almost gone
16:12:42Patr3ckwell, not much to break if there is nothing there yet
16:12:53Tomas2Who devved the flip it game?
16:13:19Tomas2ow sorry..
16:13:26Bagderthere's also the newly introduced bug with viewers
16:13:27Tomas2"Copyright (C) 2002 Vicentini Martin
16:13:29Bagderthat should be fixed
16:13:52Tomas2is "Vicentini Martin" ever on IRC?
16:13:57Tomas2And what's his nick?
16:13:58BagderI don't think so
16:14:57Patr3ckBadger: Is it in the bug tracker?
16:15:08BagderPatr3ck: no
16:15:26Bagderbut you'll spot it yourself with the latest cvs
16:15:47Bagderjust pop up the (long-press-on-select) menu on a file
16:16:51Tomas2Bagder? For that wikiname... you ask firstname and lastname concatinated... but my lastname is Salfischberger wich makes TomasSalfischberger.. wich is a little long isn't it?
16:17:57BagderTomas2: we have many long names, it doesn't matter
16:18:47CoCoLUSsounds german :)
16:19:01BagderPatr3ck: you'll see duplicate plugin names in the menu
16:19:10Tomas2My family is from austria...
16:19:47Patr3ckBadger: Thanks, I will take a look at it
16:20:03Tomas2Is it ok to edit the ThingsTodo page? And put my name in for the flipit puzzel?
16:20:12BagderPatr3ck: the problem comes from the fact that a single plugin can be a viewer for more than one extension
16:20:28BagderTomas2: sure, that's what the wiki is for!
16:21:03Patr3ckBadger: So the plugin is listed e.g. twice when it is registered for more than one file type?
16:21:12BagderPatr3ck: correct
16:21:26Patr3ckBadger: ok, thanks
16:25:19 Join _Lucretia_ [0] (
16:35:07Patr3ckBadger: Favour a quick local or a clean fix that has more impact?
16:35:43BagderI think I'd prefer the clean one
16:36:30Patr3ckok, so for each plugin there should be a list of supported extensions
16:36:37 Join frank [0] (
16:37:05BagderPatr3ck: not necessarily, since the list of extensions => plugin already exists
16:37:55Patr3ckBadger: I see (now)
16:37:58Bagderthe problem is more that the code showing "open with..." doesn't take into account that the same plugin may be listed more than once
16:39:28 Join cYmen_ [0] (
16:40:03 Quit cYmen (Read error: 104 (Connection reset by peer))
16:46:39preglowBagder: when they said they have it running on 90mhz coldfire, they didn't mention quality settings, did they?
16:46:58 Quit gargantua ()
16:47:31preglowyou can force mono mode, there are three quality settings, also half and quarter bandwidth decoding
16:50:36preglow#ifdef MPEGAUD_DEMO \n quality = 0; #endif
16:50:41*preglow dies a little inside
16:52:25Tomas2Bagder? what did you do with all those makefiles?
16:52:44 Join newnick [0] (
16:52:44BagderI modified them
16:53:05Bagderuse cvs diff, the web interface or read cvs commit mails
16:53:55*Bagder runs away
16:54:20Tomas2yeah, I received the mails ;)
16:54:26Tomas2that's why I asked
16:54:42HClpreglow: did you get it to stop outputting garbage?
16:55:18 Quit newnick (Client Quit)
16:57:11preglowHCl: yes
16:57:21preglowHCl: it works fine now
16:58:17preglowhole god damn shit
16:58:32preglowi just commented away the disk writing, and libmad actually runs at 195% realtime
16:58:58HCldisk writing?
16:59:05preglowwell, yes, mpa2wav dumps the result to disk
16:59:13*HCl grins.
16:59:17HClyes, that would make a difference.
16:59:32preglowyes, so i figured, but 100% difference
16:59:56preglowthere's got to be something wrong here, it can't be that large
17:00:07HCli think it would be a more accurate test to write it to mem somewhere in a buffer
17:00:12HCland keep overwriting it to that mem
17:00:23HClthough, sound can use dma? so you can keep decoding while its playing
17:00:24preglowwell, that's what it does right now
17:00:43HClsounds good to me.
17:00:53HClpersonally, i'm curious to the lowest mhz we can run it at
17:01:00HClwhile it still gets > 110%
17:02:08preglowthe open with list is really broken now
17:02:20preglowit scales almost linearly
17:02:25preglow120/2 = 60
17:02:29preglowbut we need more headroom
17:05:08***Saving seen data "./dancer.seen"
17:05:18preglowi don't bloody believe this
17:13:00pabspreglow: is that committed?
17:13:20*pabs woudl like to listen to some music for a change :o
17:13:47preglowpabs: how can you listen to the music?
17:14:03*Tomas2 luistert naar: Enya - Shepherd Moons (3:40)
17:14:04preglowif you've got rockbox working with sound, i'd very much like you to commit that first
17:14:07*Tomas2 is listening ;)
17:14:07pabspreglow: i thoughtthat's what you were working on
17:14:21preglowi'm fiddling around with codecs, i don't know squat about the sound hardware
17:14:38pabsit's on linusn then!
17:14:49*pabs looks around for hte little bugger
17:15:14preglowlinus usually doesn't come on in weekends
17:15:31preglowhe's more than likely got his hands full with more important stuff
17:16:04preglowbut if libmad actually runs at 200% realtime...
17:17:09crwlis anyone looking at tremor?
17:17:34 Join Renko [0] (
17:18:14preglowcrwl: how looking at?
17:18:44crwlpreglow, porting it to h1xx, or something... like you seem to be doing with libmad :)
17:18:57preglowremoving the disk write contributed a VERY handsome boost
17:19:07preglowcrwl: i'm planning on turning my attention to it after a while
17:19:14preglowi haven't got too much time these days either
17:19:43preglowi'm decoding music at 200%
17:20:14crwlat 120 MHz?
17:20:33crwldoes someone know at what speed the original iriver firmware runs?
17:20:39preglowgood news:
17:20:47preglowit's almost bloody realtime at 48mhz
17:21:13preglowcrwl: varies
17:21:26crwlwell, when playing :)
17:21:30preglowcrwl: varies
17:21:40preglowdepends on codec
17:21:44preglowand i don't really know
17:21:44crwlwhen playing mp3?
17:21:45preglownor care
17:21:53muesli-sorry for that dully question, but what does it mean when you're decoding at 200% is it playing twice as normal?
17:22:17preglowmuesli-: it produces data at double the rate it can be consumed by the audio hardware
17:22:26crwli'm just wondering what's the level of optimization there in iriver firmware - and would rockbox use more battery, ...
17:22:37preglowcrwl: level of optimization is probably decent
17:22:46preglowcrwl: rockbox more battery: probably
17:22:51preglowcrwl: in the start, at least
17:24:11crwlpreglow, how much more?
17:24:20preglowcrwl: how would i know?
17:24:25Tomas2depends on what you're running?
17:24:26preglowcrwl: it's not even finished yet
17:24:30crwlyes, i understand that
17:24:31crwli'm just curious
17:24:57Tomas2Is the kernel capable to change cpu speed when it's used/not used?
17:25:14preglowTomas2: yes
17:25:19preglowTomas2: in a way
17:25:36preglowTomas2: right now we have a mechanism of boosting cpu frequency when we need it
17:25:38Tomas2well... then in the end it might use less battery than iRiver did..
17:25:47preglowwell, how do you know it doesn't do that?
17:25:53crwlthe battery in h1xx is quite good anyway
17:26:14crwlmost people probably would be ok even if rockbox initially consumed almost double the power
17:26:36crwlmost of the ipod people are, after all :)
17:26:38Tomas2yeah... mine normally works for 6 to 7 days...
17:27:00Tomas2listening music in train and bus...
17:30:11 Join Tomas|laptop [0] (
17:30:18Tomas|laptopback in linux :)
17:31:02 Quit Renko (Remote closed the connection)
17:35:24 Nick [Zmaj] is now known as [zmaj] (
17:37:06 Join XShocK [0] (
17:37:41XShocK200% realtime... congrats.. :)
17:37:54Tomas|laptopmake[2]: Entering directory `/home/tomas/dev/rockbox-devel/uisimulator/common'
17:37:54Tomas|laptopmake[2]: *** No targets specified and no makefile found. Stop.
17:38:09Tomas|laptopsomething to do with your makefile edit's?
17:38:18Tomas|laptopor do you know what's wrong?
17:38:18DeadManwe need MP3 playback api :)
17:39:22XShocKTomas, did you get the uisimulator from cvs?
17:39:32XShocKbesides only rockbox
17:40:01Patr3ckBadger: I have created a patch for the duplicate plugins shown
17:40:14Tomas|laptopXShocK: yes
17:40:29Tomas|laptopI checked-out rockbox-devel
17:40:31 Join matsl [0] (
17:40:36Tomas|laptopand updated it just a few minutes ago
17:41:32Tomas|laptopit's an error after this:
17:41:32Tomas|laptopAR /home/tomas/dev/rockbox-devel/uisimulator/librockbox.a
17:41:56XShocKyou didn't get the uisimulator
17:42:14XShocKcheck that uisimulator folder is in the rockbox directory
17:42:33Tomas|laptopyes it is
17:42:56Patr3ckthe patch can be downloaded at
17:43:09 Quit _Lucretia_ ("Leaving")
17:43:39Tomas|laptopXShocK: I checked-out rockbox-devel, then got to rockbox-devel/uisimulator and hit ../toold/configure
17:43:44Tomas|laptopanswered the questions
17:43:46Tomas|laptopand did make
17:43:54Tomas|laptopbut the make fails with the error I copyed above
17:44:16XShocKhmm. is uisimulator something like 1.8 MB large?
17:44:49XShocKi always was getting rockbox+uisimulator. not the "-devel" one
17:45:02preglowPatr3ck: dead link
17:45:14Tomas|laptopI'll try
17:45:29Tomas|laptopmaybe the -devel one isn't totally working
17:45:43Patr3ckhang on, looking into it
17:46:06XShocKand you don't need to build it in the uisimulator directory, it can be you own newly created folder
17:48:11XShocKdid anyone test the performance of ATA driver? it really looks like it is very slow, reading a megabytes takes several seconds... will need some real optimisation..
17:48:26XShocKs/megabytes/1 megabyte/
17:50:36Tomas|laptophow did you test it?
17:51:49Tomas|laptopto memory?
17:52:28XShocKyes, when i was playing the WAV file, 10 megabyte buffer was filling quite a big time...
17:52:59 Join _Lucretia_ [0] (
17:52:59Tomas|laptopHave you eliminated all other slow link possibility's>
17:53:04Tomas|laptoplike slow loading to mem?
17:53:08Patr3ckuse the webserver is doing something stupid with the .patch extensions
17:53:58XShocKwhat i did is very simple call to read(..) function.
17:54:40XShocKthere might be a problem with slow ram... but still there is nothing i could do for it..
17:54:54XShocKit needs to be optimized.
17:55:14Tomas|laptopwell... there is 32 mb of ram... and the iRiver firmware starts filling a part of it with mp3 cache when playing
17:55:21XShocKit doesn't use DMA, which would be a good choice.
17:55:27Tomas|laptopmy player takes about a minute befor it stops loading from disk..
17:55:44Tomas|laptopso maybe it's just te limit for the hardware?
17:55:51Patr3ckpreglow: "open with" will now only show those plugins that are registered for the corresponding filename extension
17:56:51XShocK22-13= it took 9 seconds to fill the buffer. which is certainly less than what i got in rockbox
17:57:00preglowPatr3ck: excellent
17:57:23preglowthat was sorely needed
17:57:26XShocK9 seconds is on irirver original firmware
17:57:55Tomas|laptopdoing what?
17:58:03Tomas|laptop9 seconds buffer filling?
17:58:45XShocKyes, then it shut the horddisk. howeve i don't know if it fills all 32 megabytes of ram or not, thou it should
17:58:51preglowXShocK: you could check if it can do dma, right now it uses pio mode
17:59:17Patr3ckpreglow: will you add it to cvs, if you conclude it is ok?
17:59:29Tomas|laptopXShocK: I'm almost sertain it takes longer on mine...
17:59:48preglowPatr3ck: i might, just not right now
17:59:58XShocKpreglow: i will try to.
18:00:07 Quit funkymonkey (" HydraIRC -> <- The professional IRC Client")
18:00:14Patr3ckpreglow: no problem just wanted to make sure it is not forgotten
18:00:21preglowPatr3ck: i'll have a look at it in half an hours time
18:00:50Patr3ckpreglow: thanks
18:01:23Tomas|laptopXShocK: in iRiver firmware... loading an audio book (sive > 500 mb) it run's disk for 42 seconds
18:02:13Tomas|laptopso about 30/42 mb/s
18:03:30XShocK51-33=18 seconds, it was a 80mb audio file.
18:05:00XShocKit depends, but still i am quite sure that it fill the whole 30 mb buffer. and if it does in in about 20 seconds, it is 1.5 mb/sec, while rockbox now performs much less than that...
18:05:03Tomas|laptopstrange... time to find a Fat32 defragment tool for linux...
18:05:17Tomas|laptopmine is much slower...
18:05:24Tomas|laptopdoing 0,75 mb/sec
18:05:49XShocKmaybe it is full of data?
18:05:57Tomas|laptopyes it is...
18:06:12Tomas|laptopwell.. full.. 10 gb used
18:06:17XShocKi have something like 4 gigs left, and once put i almost never deleted stuff.
18:06:32XShocKout of 20.
18:07:28Tomas|laptopI have 8 gig left.. also a 20 gig model
18:08:12Tomas|laptopbut i think mine is more fragmented then... I deleted and moved a lot of things when testing a few scripts last week
18:09:31 Join Tang [0] (
18:09:38TangHi Rockbox
18:10:22TangHi Thomas
18:10:25muesli-hi Tang :)
18:10:35Tomas|laptop*kuch* H?
18:10:36TangHi Muesli
18:10:45TangHi everyone
18:13:07_Lucretia_anyone know about the hirose type connectors? Do they enforce a particular communication (e.g. parallel/serial)?
18:13:10TangAnyone can tell me if something is normal with Rbx?
18:15:20Tomas|laptopwhat Tang?
18:15:23TangI've changed my iHP remote for an iMP one recently
18:15:33Tangand i've installed Rbx two days ago
18:15:42XShocK_Lucretia_: isn't it a company? not the type of connector
18:15:53TangWhen i push iHP play it boot on Rbx okay
18:16:12Tangbut with the remote "play button" i've two cases
18:16:23Tangwith a long push it start iRiver fw
18:16:42Tangand with short push (on the remote play button)
18:16:48Tangit load Rbx fw
18:16:53Tangis it normal?
18:17:09Tomas|laptopno idea...
18:17:18Tomas|laptopI tought Rbx didn't support the remote jet...
18:17:19Tangi thought the remote
18:17:30Tangwas to load only iRiver fw
18:17:40TangQuite strange no?
18:17:51Tomas|laptopwell.. I think i can explain...
18:18:19 Join Chamois [0] (
18:18:21Tomas|laptopa short push only activates the device... but when the bootloader load's, you have already released the key... so there is no way for rbx to tell how it was started
18:18:38Tomas|laptopand so it thinks it was normally started.. and starts rbx?
18:18:54Tomas|laptop(somebdoy correct me if I'm wrong)
18:19:04preglowTomas|laptop: hmm, you'd have to release it _Very_ fast
18:19:04Chamoisyou're true
18:19:22Tomas|laptoppreglow: maybe Tang is very fsat?
18:19:40Tomas|laptopwhere in the boot process is the key check located?
18:19:41ChamoisLinusN knows this problem
18:19:44Chamoisand will correct it
18:19:50 Join sox [0] (
18:19:54Tanghum strange
18:20:00Chamoishi Tang ;-)
18:20:02Tang$hi preglow and Chamois
18:20:09soxbagder: there's one place where you havent changed the makefile:
18:20:10soxmake[2]: Entering directory `/Users/svante/rockbox/rockbox-devel/apps/plugins'
18:20:10soxmake[2]: *** No rule to make target `#pragma', needed by `/Users/svante/rockbox/rockbox-devel/build/dep-plugins'. Stop.
18:20:29Tangokay i was quite surprised since i read nothing about this
18:21:03Tomas|laptopTang? what kind of remote did you have before?
18:21:10Tangi thought it was a normal procedure
18:21:10Chamoisin the IRC log two weeks go
18:21:17Tomas|laptopI just checked a picture of the imp remote... that's the default one in Holland...
18:21:19Tangi owned the original iHP1xx remote
18:21:24Tangbut i broke it
18:21:33Tangso i had to purchase another one
18:21:34Tomas|laptopIn holland the imp is the original...
18:21:50Tangand i only found iMP550 one in France (at low price)
18:21:50Tomas|laptopthat one
18:21:52_Lucretia_XShocK: dunno, I'm following links here :-|
18:22:19Tangokay that's the old mine (thought mine is champaign since i own an iHP120)
18:22:30Tomas|laptopyeah, mine too...
18:22:42TangNow i use an iMP550 one but it's said to be fully compatible
18:22:50Tangand it work very safe
18:23:01Tomas|laptopyeah, this one is advertised as an iMP550 one :)
18:23:06Tangjust the strange process i said about fw loading
18:23:20Tangbut in fact it didn't annoy me
18:23:50 Quit Chamois ("CGI:IRC")
18:23:52Tangi didn't understood very quick what make the remote start Tbx or iRiver
18:24:18TangNow it's okay and for me the actual process is even more practical
18:24:36Tangjust wanted to check if it was normal or not
18:24:57Tomas|laptopmaybe it would even be smart to make this default... long push (whatever on button you use) is iriver, short one is rbx
18:25:16Tomas|laptopso you have to hold the key for 2 seconds for iRiver or something...
18:26:53XShocKpreglow: did you check how disabling the writing to a disk would affect the performance of FLAC?
18:29:53 Quit matsl (Remote closed the connection)
18:31:50preglowXShocK: no
18:31:57preglowXShocK: my guess is: quite dramatically
18:32:03preglowwe'll see
18:32:11preglowi'm working on other stuff right now
18:33:48XShocKmmmm.. 200% is awsome... :)))
18:35:11preglowyes, and there lots more to do
18:35:24preglowflac is already almost 175%
18:35:34preglowat least on my file
18:35:37preglowit varies drastically
18:38:21jypHas sound been committed yet ?
18:39:53jypbtw, good job preglow ;)
18:39:54Tangi tried flac on my iHP
18:39:59Tangand i got only 13%
18:40:05preglowjyp: i didn't do much
18:40:09Tangdue to CPU freq?
18:40:20 Quit Patr3ck ()
18:40:23preglowTang: if you didn't clock it up, then yes, because of that
18:40:39Tangokay thanks
18:40:40*Tomas|laptop is away: eating
18:40:42Tangwhat i thought$
18:40:45jyppreglow: All the better ;P
18:40:55preglowTang: Info->Debug->View I/O ports then press up, and it'll run at 120 mhz
18:41:19jyp... it would mean that I haven't much to do for the gmini either
18:41:27TangOkay in fact i'read something about recently
18:41:40Tangbut i hadn't Net connection to check yesterday
18:43:27preglowjyp: oh, you've got your hands full, alright :D
18:48:42preglowcan anyone think of a reason to have a plugin accept all file types?
18:49:21 Join Stryke` [0] (
18:50:23 Join stevenm [0] (
18:52:50preglowXShocK: one file is 260% realtime with flac then
19:00:07stevenmWhat are the keybindings for the Simulator ?
19:00:29stevenmor where can I read about them
19:01:12preglowthey're printet at start up here
19:01:56stevenmaah ok, thank you
19:02:04preglowthat's the x11 sim
19:05:11***Saving seen data "./dancer.seen"
19:06:11 Join amiconn [0] (
19:06:49XShocKpreglow: cool. :)
19:07:04preglowflac isn't as optimised as libmad yet
19:07:31preglowbut ok, no one here against the idea that Open With menu should only list plugins capable of handling the current file type?
19:07:38preglowi'll commit patr3ck's patch, then
19:07:39XShocKinteresting... how does it do without frequency boost?
19:08:31preglowXShocK: on one file, exactly 100% realtime
19:08:47preglowXShocK: i can easily push it below that
19:08:51preglowabove, i mean
19:08:59XShocKhehe.. and it is not optimized..
19:09:07preglowone part of it is
19:09:22preglowi've written lpc reconstruction routine in asm
19:11:59amiconnpreglow: Do you still think about converting the mpegdec asm version to coldfire/gas syntax? If yes, maybe this is helpful:
19:12:28preglowamiconn: i'm really not thinking about it, no, but i'll have a look at it
19:16:27Tomas|laptoppreglow: aren't there plugins like a favorite list and things like that wich will accept all files?
19:16:46preglowTomas|laptop: yes, that's why i asked
19:16:50preglowsomeone'll have to fix that, then
19:17:04preglowthose plugins should have a special extension
19:17:19 Nick [zmaj] is now known as [Zmaj] (
19:17:19amiconnpreglow: 'open with' should offer some plugins for all file types (even non-recognised ones), e.g. the text viewer
19:17:39preglowamiconn: how do you propose to implement that?
19:18:33amiconnI did not yet think about that. I only want to mention it in case someone is going to implement plugin handling depending on file type in the 'open with' menu
19:18:45preglowi've already commited it
19:18:49preglowthe old behaviour was broken
19:18:56preglowit listed stuff several times
19:19:11amiconnYes... but that's a different problem.
19:19:15preglowlike, i'd get one mpa2wav entry for 'mp3', and one for 'mp2'
19:19:28preglowwhich is really ugly
19:20:00amiconnActually, offering all viewer plugins for 'open with' is the intended behaviour. If you just want the default action for the file, you simply play it.
19:20:49preglowwell, ok, then i'll have to revert that patch
19:21:00preglowbut having several plugin entries is really ugly, imho
19:21:36amiconnYes... the new behaviour should be reverted, but a fix for the duplicate plugin names is still needed
19:21:47preglowyou can risk having mpa2wav listed three times
19:21:49Tanglol indeed
19:21:55Tangnoticed that too
19:22:53preglowi personally think a better approach would be having just some plugins accept all files, and let those plugins have a special file extension entry in viewers.config
19:23:06preglowbut ok, how to revert a commit? do i need to just reverse the patch and commit again?
19:25:02 Join Renko [0] (
19:25:35Tomas|laptopmaybe you can put your patch in optional somewhere?
19:25:53preglowwell, how?
19:26:01preglowit's not my patch
19:26:10preglowso i don't really want to spend a ton of time on it
19:26:12Tomas|laptopI personally think your way is much easier... and I don't want for example a textviewer or a mp3 split tool to show up for a wma file
19:26:19TangAbout the main display
19:26:37Tangit's normal to see nothing in play mode for now?
19:27:25preglowyes, i imagine this would be a fine setting
19:27:47preglowi don't really want a text viewer listed on an mp3 file either
19:28:38amiconnpreglow: Yes, but text files may have almost any extension, or even no extension at all. I want to be able to view all of them
19:29:32Tomas|laptopmaybe a simple compromise? check if a file extension is listed for some plugin, then display only those, and the all-extension plugins
19:29:38amiconnIn fact, this may also happen for other file types, and you simply can't associate all possible extensions.
19:29:54preglowamiconn: yes, i agree in that particular case
19:29:56Tomas|laptopand if no plugin is defined for a filetype then use some list of show-when-unknown plugins?
19:30:02preglowamiconn: but having a flac file called something.mp3 doesn't make sense
19:30:03amiconnANother example: JPEGS may have .jpg .jpeg, .jpe, .jff, .jfif
19:30:21Tomas|laptopbut only .jpg and .jpeg are used commonly
19:30:32preglowyes, but they should be supported
19:30:46Tomas|laptopwell.. then my compromise is working?
19:30:50preglowi think it would be a good optional thing anyway
19:30:58Tomas|laptopadd all possible jpeg extensions to the jpeg plugin?
19:31:04amiconnTomas|laptop: Yes, and that's why displaying all viewer plugins in 'open with' is useful
19:31:07preglowTomas|laptop: currently, that's not possible
19:31:50Tomas|laptopmaybe we can make it a configuration option? to show only registerd or to show all?
19:31:58amiconnThe way viewers.config works currently is far from optimal for plugins that are listed several times.
19:32:30preglowi think an extension list for each would be great
19:32:43preglowand extension * for plugins that match all extensions
19:32:51amiconnAll info is registered n times, even the info that is identical. Plugin name, icon, and extension (which of course is different)
19:33:20preglowand an option for switching on/off listing of only relevant plugins
19:33:24amiconn...and the buffer space is limited, and should not be extended too far without reason.
19:33:43Tomas|laptopwell preglow's idea sounds useful?
19:33:45Tomas|laptopsomething like:
19:34:09Tomas|laptop[plugin name] [info] [extensionlist like: a; ab; jpg; mp3]
19:34:15preglowmy idea would probably be cancelled out by the extra needed code to parse the file, but it's more elegant
19:34:25Tomas|laptopand * in place of the extension list if it's for every file
19:34:55amiconnI'd still vote for always showing all viewer plugins in 'open with' (but every plugin only once of course). That's what 'open with' is about. Windows does it the same way (okay, that's probably not the best example)
19:35:04Tomas|laptopit isn't to difficult to parse a a; ab; sdf; sdf list...
19:35:24Tomas|laptopamiconn: windows XP does it a different way now...
19:35:40Tomas|laptopit shows the programs you use for that filetype... and an select others option
19:35:42amiconnWinXP is what I'm using...
19:36:19Tomas|laptopnever noticed it display's the used programs in a context menu? and the option to open the full list when your program isn't in there?
19:36:58amiconnIf I select the generic 'open with..' it always show all registered programs. Okay, it offers an additional 'open with' sub-menu in the context menu, where it shows only those programs I already used once for that file type
19:37:24amiconn...but that one I almost never use
19:37:49Tomas|laptopok.. well I alway's used that...
19:38:30amiconnThe programs I want to use when I do 'open with' are most likely not yet present in that sub-menu, that's the reason
19:38:31Tomas|laptopgnome does the same here...
19:38:58Tomas|laptopgives some program's it thinks you want... and an extra option to get a list of everything...
19:39:42amiconnI generally dislike it when the os tries too hard to guess what I want.
19:39:46preglowyes, indeed
19:40:03preglowit always fails miserably
19:40:10amiconnMost of the time, yes
19:40:12preglowmore often it just plain forgets what i usually use
19:41:50Tomas|laptopwindows users :P
19:50:12 Part amiconn
19:50:32XShocKi think the best choice will be to add line for "*" files in viewer.conf, and also have an option in setting for showing all possibilities.
19:51:58XShocKand also the format of current viewer.conf should be changed to someting like Tomas proposed with several extensions
19:52:21preglowthat would be best
19:52:30preglowsomeone code support for it :P
19:52:39 Quit _Lucretia_ ("Leaving")
19:52:41Tomas|laptopwell i can try..
19:52:49Tomas|laptopbut what exactly is needed? and where?
19:52:51preglowthe * support is not really needed right now
19:52:53preglowsince there's no option
19:53:22Tomas|laptopSo a simple check function? wich returns al plugin's having a given extension?
19:53:44 Join matsl [0] (
19:54:00 Quit Bagder (Read error: 110 (Connection timed out))
19:54:13 Join Bagder [0] (
19:55:04preglowTomas|laptop: i'd say support for a extension list in the plugin struct
19:55:14preglowTomas|laptop: and code to parse the extension list in viewers.config
19:55:24preglowTomas|laptop: then change onplay.c to support it
19:55:28preglowit'll be a bit of work
19:55:40Tomas|laptopok, then i need a little help
19:56:01Tomas|laptopI just started reading things about rockbox this morning..
19:56:09preglowi have none to offer right now
19:56:17Tomas|laptopam i allowed to use an array in that struct?
19:56:27Tomas|laptopof.. say.. 10 extensions?
19:56:45Tomas|laptopthen parse the first 10 extensions in viewers.conf and put them in the array?
19:56:57Tomas|laptopand edit onplay.c to loop throug the array?
19:57:55preglowa linked list would be better, i think
19:58:00preglowso we don't waste space
19:58:30preglowapart from that, it sounds good
19:58:40preglowput you must never waste space, use exactly what's needed
19:58:47preglowthere's very little ram on the archos players
19:58:58Tomas|laptopok.. I don't think this is gonna work in a few minutes... but I'm going to try
19:59:03preglowof course not
19:59:06preglowtake your time
19:59:20Tomas|laptopyeah, first have to ready some of the code..
20:00:59 Nick StrathAFK is now known as Strath (
20:01:27preglowfiletypes.c is more important, btw
20:01:31preglowonplay.c has nearly no needed code
20:02:15Tomas|laptopok, do you know someone who helped writing the old system?
20:02:31Tomas|laptopI think I'm gonna need some support some times :)
20:03:07preglowi have no idea who wrote it
20:03:26Tomas|laptopwell... I'll just scream here if I can't find something..
20:04:15 Join webguest49 [0] (
20:04:59 Quit webguest49 (Client Quit)
20:07:57 Join Bagder_ [0] (
20:11:35Tomas|laptoppreglow? When i use your linked list idea I should use a struct containing a string and a pointer to the next struct right?
20:12:01Tomas|laptopand that as a part of the plugin struct?
20:12:13Tomas|laptopfilled with the extensions for that plugin?
20:16:29 Quit Bagder (Read error: 60 (Operation timed out))
20:16:57 Quit jyp ("poof!")
20:22:59 Quit Seed (Nick collision from services.)
20:23:07 Join Seed [0] (
20:25:22 Quit Tomas|laptop ("Leaving")
20:28:08 Join markun [0] (
20:28:19 Join Tomas|laptop [0] (
20:28:32Tomas|laptopwrong button...
20:29:48Tomas|laptopcalculator.c:1072: internal compiler error: in reload_cse_simplify_operands, at reload1.c:8365
20:30:10markunpreglow: I wanted to move some of the mdct functions of Tremor to icode. I now get 'IllInstr' at mdct_backward. What should I do to move code to iram?
20:30:47preglowmarkun: have you copied the code in question to iram?
20:31:04preglowif not, have a look at the memcpy in mpa2wav that does that
20:31:15markunpreglow: ok, I'll take a look at it.
20:36:00markunpreglow: Just moving the mdct functions to iram doesn't give a speed improvement.
20:36:53preglowmarkun: no surprise
20:37:12preglowmarkun: there should be a small difference, but the greatest gain is found in just optimizing code and placing data in iram
20:37:17preglowmarkun: there already is a code cache
20:37:37markunI placed the data in mdct_lookup.h into idata
20:38:01preglowthat should have yielded a improvement
20:39:19markunwait, it does improve!
20:39:23markunDidn't reboot..
20:39:48markunI get 71% real-time with a q 3 ogg file.
20:40:05markunI've disabled output to vorbistest.wav and disabled the printing in malloc.
20:41:31markun72% now. If I have time for it this week I would like to start on implementing the mdct with prerotation+fft+postrotation.
20:42:34preglowyou certain are you that will yield a gain?
20:42:37 Join midk [0] (
20:42:44preglowis there a good coldfire fft available?
20:43:44markunThere is the 1024 points ref.. I will try to start from there.
20:44:25preglowit should be modifyable
20:44:32markunDid you read the master thesis of the guys who optimized Tremor to run on a DSP?
20:44:35preglowmost larger ffts are iteratively based on smaller ffts
20:44:55 Join _Lucretia_ [0] (
20:45:08markunIt's very interesting. The initial test file decoded in 27 seconds. After optimizing in 8.
20:45:42preglowi haven't got time to read it now
20:45:58markunMaybe you can bookmark it and have a look at it later.
20:46:33markunThey made a patch for Tremor low-mem, but it's not available anymore.
21:05:13***Saving seen data "./dancer.seen"
21:14:08CoCoLUSlibflac is way > 100%, isn't it?
21:21:12Bonkersflac is faster than ogg or it's just been more optimized by now?
21:22:16preglowwell, it's not that computationally intensive from the start
21:22:26preglowi've only done one opt in libflac
21:22:44Bonkerssomehow xmms likes to take a large chunk of cpu when playing flacs occasionally
21:22:51CoCoLUSall we need is the sound arch. :)
21:22:51Bonkersthat's the only reason I ask
21:31:58 Nick [Zmaj] is now known as [zmaj] (
21:32:50Tomas|laptopok.. stupid question..
21:33:00Tomas|laptophow do I install rockbox on the simulator?
21:33:23Tomas|laptopI have a build folder.. (iRiver build) but what do I copy when I want to run the simulator?
21:35:46 Join Camilo [0] (
21:39:45Tomas|laptop*kuch* really stupid *kuch* .. I have it :) copy the build folder..
21:44:29 Join HenrikB [0] (
21:47:25Tomas|laptopanybody has an example viewers.conf for me?
21:47:55Tomas|laptopit should be here: /.rockbox/viewers.config
21:48:19 Quit edx (Read error: 110 (Connection timed out))
21:49:35markunYou can do 'make zip' in your build folder. This gives you a which you can unzip in the archos folder.
21:49:55markunIt includes viewers.config
21:50:44Tomas|laptopeh.. I have iRiver build options in use:)
21:50:50Tomas|laptopso it doesn't make zip..
21:50:58Tomas|laptopdo you know the format of viewers.config ?
21:51:43preglowviewers.config is in plugins/
21:59:23 Quit Camilo ("Chatzilla 0.9.67 [Mozilla rv:1.8a6/20050111]")
22:01:55Tangi've jsut installed Rockboy with greylevel thanks to Chamois (who gave me some links)
22:02:05Tangit's quite cool
22:02:14Tangseems to be real time ;)
22:02:22Tang(with Mario)
22:03:15Tomas|laptophave you tried game boy color roms?
22:03:29CoCoLUSthey are working
22:03:42CoCoLUSbut... you know... its not like the iriver display has colors :P
22:03:53Tomas|laptopno, that's not the problem ;)
22:03:57preglowanyone know if plugins use the main stack?
22:03:58Tomas|laptopI only want the rom's to work
22:04:34preglowwell, yes, of course they do
22:04:51CoCoLUSmost of the color roms include the color aspect somewhere in the gameplay
22:05:00CoCoLUSwhich makes them somewhat unplayable :)
22:05:12CoCoLUSjust look at the color dungeon in zelda dx...
22:05:40Tomas|laptopyeah, but for some games there is a non color and a color version.. and I have only a color rom..
22:05:55Tomas|laptopand it's not a problem to not see the color of a pokemon ;)
22:06:14 Quit sox ("Snak 4.13 IRC For Mac -")
22:06:36 Join jyp [0] (
22:10:07 Quit markun ()
22:18:11*Tomas|laptop is leaving... Have to work tomorrow :(
22:18:15 Quit Tomas|laptop ("Leaving")
22:20:14 Quit Tomas2 ("Guess what... leaving too :)")
22:28:12 Join Soul_Eater [0] (
22:34:40Soul_Eaterso whats new with the project
22:36:39 Quit Strath ("Client exited")
22:37:45 Join Strath [0] (
22:44:23DeadManmore games ;)
22:46:42 Nick [zmaj] is now known as [Zmaj] (
22:49:47 Join DMJC [0] (
22:52:34Soul_Eaterhaha oh really?
22:52:39Soul_EaterWe got Legend of Zelda goin yet?
22:53:28DMJChow's sound going?
22:53:55Soul_Eatertight. LoZ uses alot of controls. I wonder how the hell that works on the iriver
22:54:12 Join amiconn [0] (
22:54:16HClnot well.
22:55:44TangI leave the place
22:55:48Tangbye erveryone
22:56:09 Quit Tang ("Chatzilla 0.9.66 [Mozilla rv:1.7.5/20041108]")
23:01:34 Part DMJC ("Leaving")
23:03:36amiconnThe x11 sim now locks up after some time, and it's definitely not the x11 mt problem :(
23:05:14***Saving seen data "./dancer.seen"
23:06:39preglowamiconn: how long a time?
23:07:16amiconnWell, it happens when a scrolling line is displayed, and depending on the scroll speed 'some time' ranges from seconds to minutes
23:07:23preglowXShocK: does the codec eat data in big endian format?
23:10:16HClCoCoLUS: actuall, cgb roms have a backwards compatibility mode that should make them play fine on a grayscale gameboy too.
23:10:39HClits just that gnuboy has a palette bug that causes it to have wrong palettes at times.
23:10:44HClits on the tofix list.
23:11:09HClunfortunately, in order to be able to fix it, you have to traverse and understand that horrid ugly goo of code they wrote.
23:12:26Soul_EaterGnuboy can save games right? Does it save them directly to a directory on the hd?
23:18:46XShocKwas out for a while.
23:18:56XShocKwhat codec?
23:19:07XShocKyou mean the DAC ?
23:19:45XShocKyes, DAC takes big-endian, i had to byte-swap the data from WAV to get it working
23:20:42 Join nobby [0] (
23:22:31amiconnpreglow: What also happens - as soon as the x11 sim hangs, it starts to consume all available cpu time.
23:23:05 Join edx [0] (
23:24:46nobbydoes anyone have a copy rockbox for iriver with the greyscale patch and rockboy, and could they send me it? I have the bootloader allready flashed
23:25:12preglowamiconn: "interesting"
23:25:27preglowXShocK: goodie
23:25:30 Quit YouCeyE ("Leaving")
23:25:41 Join YouCeyE [0] (
23:40:34nobbydoes anyone have a copy rockbox for iriver with the greyscale patch and rockboy, and could they send me it? I have the bootloader allready flashed
23:42:45Soul_Eatertechnically, could the rockbox ihp1xx series firmware run on an h3xx or not
23:43:01 Nick Bagder_ is now known as Bagder (
23:43:02nobbycompletely different hardware
23:43:23Bagdernot "completely" different, but different enough
23:43:59Soul_Eaterdifferent like more advanced?
23:44:00nobbywhats the similarities? they both have stereo minijacks for headphones? :P
23:44:13Bagderthe same CPU for example
23:44:24nobbyis it? i thought it was different
23:44:47preglowwould not be surprised to learn the codec chip is the same either
23:45:21Bagderpreglow: indeed the same
23:46:22Bagderthe same FM tuner
23:47:24preglowanyone know what the usb host controller is?
23:47:41preglowmotorola has issued a data sheet on how to make the mcd5249 work well with a specific phillips chip
23:47:46Bagderit isn't listed on the wiki page
23:47:56preglowso wouldn't surprise me if that's it
23:52:35 Quit nobby ("hurry up and get those codecs working! i want them by tomorrow morning! :P")
23:54:55 Join DrRick [0] (
23:55:57muesli-g'night ladiez

Previous day | Next day