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 2005-04-20

00:00:06stevenmAh.. it ain't defined if LITTLE_ENDIAN is.. okay well I hope this works because I have no way of knowing
00:02:38stevenmShagnar, http://wam.umd.edu/~stevenm/midi2wav.rock
00:02:45 Nick ac_away is now known as austriancoder (~austrianc@80.120.117.30)
00:02:52austriancoderhi all
00:03:00austriancodersome news?
00:03:13stevenmShagnar, this uses some rather experimental stuff.. I will be very surprised if that produces proper sound, but let's hope
00:03:21stevenmaustriancoder, hello. what's up?
00:03:53Shagnarstevenm allright, i just wait till the player finished (~2min)
00:04:00austriancoderremote lcd stuff is finished to 90 % - wps still needs some coding
00:04:04stevenmShagnar, ok
00:13:03Shagnarworking
00:13:44stevenmShagnar, lemme know how bad it sounds :(
00:14:00Shagnarwe'll see ;)
00:14:53Shagnar02:15
00:15:06Shagnarfor the Mi1.mid
00:15:17Shagnarlets have a look on the wave
00:15:34stevenm2:15 .. ?! woah
00:15:57Shagnarsounds nice
00:16:07Shagnarcan't hear any error
00:16:08stevenmsounds RIGHT Shagnar ?
00:16:15stevenmShagnar, are you sure this is the right .rock ?
00:16:17rasheroh boy
00:16:23stevenmdid it take longer last time ?
00:16:40rasherdoing pretty well
00:16:52Shagnarabout 02:20... started the excel-DB for the times today so i dunno... mom i'll check a second time
00:17:25stevenmShagnar, unbelievable. That was like, the most stressful 18 hours of my life
00:17:38stevenm(not really but STILL)
00:17:42stevenmbah the thing actually worked
00:17:55stevenmand I hadn't actually removed the getSample() call.. just put the code IN it
00:18:04Shagnarmom
00:18:08stevenmit's inline though, if that makes it less worse
00:18:08Shagnarin /viewers
00:18:10stevenmhm?
00:18:18Shagnarits the right one (the latest)
00:18:18stevenmmom?
00:18:23Shagnarone moment ;)
00:18:29stevenmaah
00:18:42stevenmmaybe I uploaded the wrong copy or something
00:18:50Shagnarcopied in /viewers <= right, yes?
00:18:51preglowhahah
00:19:01stevenmShagnar, right
00:19:22stevenmlooks like I uploaded what I built.. and it looks like latest build.. and I did remember to put it in ../pub ..
00:19:37preglowstart embedding build dates in the plugin :P
00:19:43Shagnarand i don't have to restart the player to make the new version affect?
00:19:59Shagnarsounds like a good idea, preglow
00:20:03Shagnar;o)
00:20:05stevenmpreglow, yea I was considering like, current time or something, just not sure how to automate
00:20:28stevenmmidi file is 1:35
00:20:38***Saving seen data "./dancer.seen"
00:20:42Shagnarmiditest is 1:34
00:20:50stevenmal right
00:21:05stevenmthat's 67%
00:21:18stevenmbeats the 50% without the endian stuff
00:21:57 Part LinusN
00:21:58stevenmnow I gotta go thru and remove the getSample CALL and actually replace it with the code.. possibly change the output buffer from chars to ints.. faster
00:22:06rasherstevenm: make it output the time of build and the time-to-decode to a file :)
00:22:21Shagnarso you're shure i run the right version?
00:22:29Shagnar*ran
00:22:30preglowstevenm: if you're just replacing all getsample calls with the getsample code, just declare getsample inline instead
00:22:45stevenmpreglow, yes that's waht it is now. and all it is is a return statement
00:22:51preglowgood
00:23:02stevenmShagnar, let me change the startup text, just in case :)
00:23:09Shagnar:)
00:23:19Shagnaror just a number at the end of the text
00:23:28 Quit ashridah ("Leaving")
00:23:29Shagnarso we can easy tell which revision we mean
00:24:02stevenmShagnar, changed the text, http://wam.umd.edu/~stevenm/midi2wav.rock
00:24:10stevenmyea I will add that in the future
00:24:23stevenmrasher, any ideas on how to measure run time?
00:24:36stevenmrasher, H100 has RTC and timers I can use?
00:24:41 Quit silencer (Nick collision from services.)
00:24:44 Join silencer [0] (~silencer@zen.via.ecp.fr)
00:24:45 Quit silencer (Nick collision from services.)
00:24:51rashernot rtc
00:24:52rasherbut timers
00:25:02stevenmrasher, that works
00:25:09rashermany of the plugins use it
00:25:23Shagnarstevenm: yes it works :)
00:25:46stevenmShagnar, wow cool
00:26:04stevenmShagnar, thanks for testing
00:26:18stevenmI'm going to go eat now, then enee review ... maybe I will be back later tonight
00:26:19stevenmbye all
00:26:24Shagnarbye
00:26:30 Quit stevenm ("Leaving")
00:28:24 Part MoosCamaro
00:29:09 Quit lolo-laptop ("Client exiting")
00:29:47 Join silencer [0] (~silencer@zen.via.ecp.fr)
00:33:12 Quit matsl (Remote closed the connection)
00:34:36 Quit cYmen ("zZz")
00:37:53austriancodergood night all
00:38:01 Quit austriancoder ("using sirc version 2.211+KSIRC/1.3.11")
00:46:01 Quit Sucka ("a bird in the bush is worth two in your house")
00:59:28 Join Camilo [0] (~chatzilla@userca029.dsl.pipex.com)
01:00
01:05:20Shagnargood night all
01:14:01 Quit Shagnar ("( www.nnscript.de :: NoNameScript 3.81 :: www.XLhost.de )")
01:22:50 Join lostlogic [0] (~lostlogic@node-4024215a.mdw.onnet.us.uu.net)
01:26:51*preglow gives up
01:27:04preglowtime for bed
01:27:27 Quit preglow ("flabb?")
01:35:03 Join edx [0] (edx@pD9EAB0DD.dip.t-dialin.net)
01:40:36 Join courtc_ [0] (~court@adsl-158-10-138.asm.bellsouth.net)
01:42:56 Quit courtc (Read error: 110 (Connection timed out))
02:00
02:10:45 Quit Bager (Read error: 104 (Connection reset by peer))
02:20:40***Saving seen data "./dancer.seen"
02:22:06 Nick courtc_ is now known as courtc (~court@adsl-158-10-138.asm.bellsouth.net)
02:24:39 Quit Camilo ("Chatzilla 0.9.67 [Mozilla rv:1.8b2/20050404]")
02:49:52 Quit Aison (Operation timed out)
03:00
03:35:14 Join TCK [0] (TCK@81-86-98-137.dsl.pipex.com)
04:00
04:05:37 Join QT_ [0] (as@area51.users.madwifi)
04:06:55 Join ehntoo [0] (~noclue2@24.177.161.77.up.mi.chartermi.net)
04:17:08 Quit QT (Read error: 110 (Connection timed out))
04:20:41***Saving seen data "./dancer.seen"
05:00
05:01:43 Join eVox [0] (eVox@ool-182dcb80.dyn.optonline.net)
05:02:03eVoxanyone out there?
05:02:29eVoxive got a problem with my ihp120
05:03:22eVoxit got plugged into a phone charger...and despite the charger being the same polarity and less power, the ihp is messed up
05:03:52Rick:<
05:03:59RickThat was silly ;P
05:04:04*Rick wonders how it messed up
05:04:13eVoxthe screen is garbled (similar to 'artifacting' on overheating video cards), but it will boot up and respond to button commands (but u gatta hold the buttons for 3x as long as normal)
05:04:22Rickah
05:04:32RickSounds like it was fried or something
05:04:32Rickheh
05:04:40Rickdoes the remote work?
05:04:46eVoxyeah i went to charge it for a trip to atlanta tomarrow, and later went looking for my phone charge and had an 'ohhhhhhhhhh shit' moment
05:04:46eVoxlol
05:05:28eVoxits still working as a harddrive (via computer connection), and the radio works, but the menus are fubar and it wont actually play mp3s
05:05:38Rick:<
05:05:43RickCan't help you there... :/
05:05:46Rickmaybe someone else could
05:05:47Rickdunno
05:05:54eVoxis there a reset to safe mode thing? like on dell jukebox's?
05:06:03eVoxah
05:06:05Rickeh, there's a reset
05:06:06Rickbut
05:06:10Rickit's not like a safemode
05:06:13eVoxyeah
05:06:21eVoxive reset it like 400 tiems
05:06:24RickI doubt that would have any effect
05:06:28eVoxnope
05:06:30RickUnless the firmware somehow fucked up
05:06:52eVoxive *tried* updating the firmware but its hard with garbled menus
05:07:42eVoxon the 'general' menu, is update firmware the 2nd one up from the bottom?
05:07:50Ricklet me check
05:08:02eVoxty
05:08:04Ricker
05:08:07Rickwhat version do you have?
05:08:23eVoxthe most current
05:08:29eVox1.6?
05:08:35eVoxthe standard...not rockbox
05:08:47Rickyeah
05:08:48Rickbut
05:08:51Rickin the newer firmware
05:08:53Rickthere's a format option
05:09:39eVoxwhat are u getting at?
05:09:46Rickwell
05:09:50Rickif your firmware is newer
05:09:57Rickthe firmware upgrade is 3rd from bottom
05:09:58Ricknot 2nd
05:10:05Rick(the last is format instead of db scan)
05:10:18eVoxwhats 2nd up?
05:10:23eVoxdb scan?
05:10:24Rickdb scan
05:10:46Rickthen you want to click
05:10:47Rick<-
05:10:48Rickclick
05:11:10eVoxyeah
05:11:27eVoxwhats 3rd up on the old one?
05:11:42eVoxcuz i have some garbled screen idk what it says
05:12:08eVoxi think upgrading with the picture of the ihp like on the boot screen
05:12:16eVoxif i squint real hard
05:12:18eVoxlol
05:13:55Ricklol
05:15:45 Quit rasher ("CGI:IRC")
05:16:40eVoxit froze
05:16:47Rick:/
05:16:52eVoxand is still fubar on reset
05:17:12eVoxhmmm...how do you take these things apart? the micro allen wrench screws?
05:17:30RickI forget the name
05:21:28eVoxdo you want to buy a paperweight?
05:25:35RickNot really.
05:26:07eVox:/
05:28:47eVoxin the newest firmware...is there a set of horizontal scrolling icons when u press the main button in?
05:29:04RickI don't know
05:29:09RickI don't have the latest
05:29:18Rickas I have rockbox on it
05:29:19Rick;P
05:29:36eVoxah
05:29:38eVoxhow is rockbox
05:29:44Rickgood
05:29:47eVoxi check the site every couple of days
05:29:56eVoxits is fully functional for playing mp3s?
05:31:19Rickno
05:31:47eVoxdoes it play mp3s?
05:36:32ehntoono.
05:37:51ehntoorockbox is just for things other than playing music at the moment.
05:38:19eVoxo
05:38:28ehntooAs I understand it, a couple codecs are in place and working, but the audio out drivers and audio backing code are still being put into place.
05:40:52eVoxah
05:57:32 Quit ehntoo (Read error: 104 (Connection reset by peer))
06:00
06:00:44 Join ehntoo [0] (~noclue2@24.177.161.77.up.mi.chartermi.net)
06:19:15 Quit eVox ()
06:20:45***Saving seen data "./dancer.seen"
06:40:17 Nick ehntoo is now known as ehntoo|ZzZz (~noclue2@24.177.161.77.up.mi.chartermi.net)
06:52:35 Join LinusN [0] (~linus@labb.contactor.se)
07:00
07:09:05 Join stevenm [0] (~steve@181-221.mam.umd.edu)
07:13:59stevenmHi all
07:15:54 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
07:21:20 Part stevenm ("Leaving")
07:21:49 Join stevenm [0] (~steve@181-221.mam.umd.edu)
07:27:35 Join courtc_ [0] (~court@adsl-33-162-147.asm.bellsouth.net)
07:28:35 Quit courtc (Nick collision from services.)
07:28:50 Nick courtc_ is now known as courtc (~court@adsl-33-162-147.asm.bellsouth.net)
07:50:53amiconnGood morning
07:51:29amiconnLinusN: I just posted an answer to the 'car adapter mode on flashed players' forum topic
07:52:16stevenmMorning? amiconn, where are you?
07:52:18LinusNah, right, silly me
07:52:27amiconnstevenm: Germany
07:52:37stevenmamiconn, ah, sweet
07:54:12LinusNd'oh!
07:54:26*LinusN just found a mistake in the dram setup
07:56:15stevenmoo lovely, 2 more shifts removed
08:00
08:00:49 Part stevenm ("Connection reset by beer.")
08:20:48***Saving seen data "./dancer.seen"
08:24:51 Join austriancoder [0] (~c1aa0259@labb.contactor.se)
08:25:25austriancodergood morning
08:25:39austriancoderamiconn: where in germany do you live?
08:29:09amiconnBerlin
08:30:19austriancoderinteresting.. i live at the bodensee
08:30:20 Quit einhirn ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
08:42:19 Quit edx (Read error: 110 (Connection timed out))
08:53:23 Join B4gder [0] (~50d8ae51@labb.contactor.se)
08:54:04B4gdergooooood morning
08:54:14LinusNbanzai
08:59:45 Join webguest65 [0] (~c31ce021@labb.contactor.se)
09:00
09:09:48 Nick Lynx_awy is now known as Lynx_ (HydraIRC@134.95.189.59)
09:21:34 Join Chamois [0] (~Chamois@champigny-5-82-226-182-23.fbx.proxad.net)
09:24:14*austriancoder has a boring networking lesson
09:24:40 Join Harpy [0] (80oRUNJF7Q@dsl-hkigw7wbb.dial.inet.fi)
09:24:42*B4gder has a boring oops in his kernel
09:27:47 Join bobTHC [0] (~foo@l05m-16-216.d1.club-internet.fr)
09:28:07bobTHCgood mornin' mates !
09:28:38B4gdermorning bob
09:32:19*LinusN just played his first seconds of realtime mp3
09:32:32austriancoderwow
09:33:26 Join cYmen [0] (~cymen@nat-ph3-wh.rz.uni-karlsruhe.de)
09:33:33bobTHC:)
09:33:42B4gderwhich song was it?
09:33:44B4gder:-)
09:34:44bobTHCwe will rock you ?
09:34:49bobTHC;)
09:38:51LinusNMachinae Supremacy - Great Giana Sisters
09:41:01B4gdergiana sisters, oooh nostalgia
09:41:08B4gder:-)
09:41:24B4gderI better go get some coffee instead
09:41:31Chamoislinus : still lot of works or will commit soon ?
09:41:33LinusNmust follow the rockbox tradition you know, first mp3 test must be Machinae Supremacy
09:41:45LinusNChamois: lots and lots
09:41:50LinusNit bugs like hell!
09:42:02Chamoisok
09:42:10*LinusN celebrates by getting a cup of coffee
09:43:13bobTHCcheeers !
09:46:51Rick:o
09:46:54Rickon iriver, LinusN?
09:47:04austriancoder;)
09:47:16 Quit austriancoder ("CGI:IRC")
09:47:39 Join austriancoder [0] (~c1aa0259@labb.contactor.se)
09:47:40 Quit Ka_ (Connection reset by peer)
09:47:51austriancoderre
09:48:59LinusNRick: of course
09:49:03Ricknifty :)
09:49:09Rickhowsit working?
09:49:10 Join ashridah [0] (ashridah@220-253-120-157.VIC.netspace.net.au)
09:49:18LinusNreally really bad :-)
09:49:29Ricklol
09:49:37*Rick has been sitting here listening to his iriver all night
09:49:38Rickbut
09:49:42RickI really should plug it in
09:49:42Rickhehe
09:49:48*Rick can't be assed to find the plug
09:52:17 Quit DangerousDan (Read error: 54 (Connection reset by peer))
09:58:58austriancoderi hate cisco routers....
10:00
10:00:59 Join edx [0] (edx@pD9522C4F.dip.t-dialin.net)
10:04:54amiconnaustriancoder: Is there a specific reason?
10:06:17 Join Ka_ [0] (~tkirk@pcp0010733332pcs.howard01.md.comcast.net)
10:08:03austriancoderamiconn: one of the routers is misconfigurated and we have now lost a lot of time to find the error and the lesson is ending in a few minutes and we had no time to do the tasks
10:08:59amiconnMisconfiguration is usually not the router's fault
10:09:40austriancoderit was the fault of the people bevore us
10:10:08ashridaheh? resetting a cisco router isn't that hard, provided you've got a console cable...
10:10:51ashridahreset it, break the bootup, change a flag so it doesn't autoload the config, reboot, ignore the config, reset the flag so it'll load your config once you've written it.
10:11:05austriancoderthats the prob.. there is only one serial cabel
10:11:28ashridahaaah, that's a pain.
10:11:33ashridahi've got my own :)
10:11:38austriancoderand my group members are not very motivated
10:12:50 Quit nozomiyume (Read error: 104 (Connection reset by peer))
10:14:31austriancoderfor wath is the eprom needed in the iriver?
10:15:52ashridahsettings storage.
10:16:10austriancoderhmmm.. settings are saved atm on harddisk
10:16:23ashridahin rockbox, yes
10:16:46austriancoderbut?
10:17:08austriancoderso we dont need the eprom inrockbox?
10:18:05LinusNno
10:18:48austriancoderoh
10:19:25austriancoderso there is no need for a driver?
10:20:52***Saving seen data "./dancer.seen"
10:22:08 Join Mirfle [0] (~chatzilla@ADSL222150.BRK.biu.ac.il)
10:22:28austriancoder:(
10:22:49*B4gder spots somone eager to write drivers :-)
10:24:02austriancoderbut i will do it.. seems to be funny ;)
10:25:33*amiconn is waiting for Zagor
10:41:46austriancoderwill be away for some minutes.. class room change
10:53:54ashridahgod. my kingdom for a decent sequence diagram drawing tool that doesn't damned well keep biting me in the ass with annoying fiddling
10:56:53 Quit austriancoder ("CGI:IRC (Ping timeout)")
10:58:14 Join sox [0] (~55e238a1@labb.contactor.se)
10:58:58 Quit sox (Client Quit)
10:59:06 Join sox [0] (~55e238a1@labb.contactor.se)
10:59:33 Quit sox (Client Quit)
11:00
11:01:51 Join Zagor [0] (foobar@h14n2fls31o265.telia.com)
11:03:57B4gderprintk printk printk
11:05:45LinusNZagor: kitchen upside down?
11:05:52Zagor:-)
11:06:08LinusNslimp3 dead?
11:06:13Zagorfinished Real Soon Now.
11:06:23LinusNah, rockbox-style time plan
11:06:28Zagoryup
11:06:36 Join DangerousDan [0] (~Miranda@newtpulsifer.campus.luth.se)
11:07:24dwihnoslimp3?
11:08:10amiconnhi Zagor
11:08:14LinusNZagor: time for my traditional "Gentlemen, we have sound" post :-)
11:08:24Zagorooh, nice
11:08:35LinusNmachinae supremacy, of course
11:08:40Zagorof course
11:09:07amiconnZagor: I've updated my fat.patch. This time it should work as intended.
11:09:24LinusN(oh, haven't we heard that one before)
11:10:03ashridahdid i hear that one right?
11:10:06Zagor:-)
11:10:35dwihnoHooray for playback!
11:10:54ashridahspeaking of machinae supremacy, i probably should get around to finding an import of their first album.
11:11:09LinusNthey rule
11:11:14amiconnI fixed 2 other problems as well. (1) Obey the maximum directory size according to the FAT spec. (2) Correctly handle the 0xE5 special character
11:11:24 Join austriancoder [0] (~c1aa0259@labb.contactor.se)
11:11:46amiconnZagor: I'd be pleased if you could have a look at it. Usual place: http://arnold-j.bei.t-online.de/Rockbox/fat.patch
11:12:22austriancoderre
11:12:43Zagoramiconn: i'll look
11:13:28 Quit Mirfle ("Chatzilla 0.9.68a [Firefox 1.0.3/20050414]")
11:15:26austriancoderLinusN: when do you do the first commit for sound playback?
11:15:55LinusNoh, not yet anyway
11:16:12LinusNit is only an experimental hack
11:17:13ashridahhurray! hack!
11:17:45B4gderwe like haxx! ;-P
11:18:00 Join [IDC]Dragon [0] (~d90a3255@labb.contactor.se)
11:18:32*[IDC]Dragon reads some success logs
11:19:08[IDC]DragonLinusN: congratulations!
11:19:33LinusNthx
11:19:53amiconnhi Jörg :)
11:20:04amiconnltnirc
11:20:08*[IDC]Dragon never heard about machinae supremacy
11:20:18[IDC]Dragonamiconn: ;-)
11:21:27 Join rasher [0] (~3e4f4094@labb.contactor.se)
11:22:32*LinusN sent the "Gentlemen..." message
11:22:47LinusNfelt really good
11:22:54rasher\o/
11:23:01[IDC]Dragonamiconn and I didn't know this protocol when doing the Ondio
11:23:01*B4gder embraces and waits for impact
11:23:17ashridahwho wants to take bets on how long an out-of-context post will be made on slashdot? :)
11:23:19dwihno\o/
11:23:40 Join preglow [0] (thomj@s183a.studby.ntnu.no)
11:23:47dwihnoThis calls for a frontpage news item!
11:24:20*preglow applaudes like a madman
11:27:44amiconn[IDC]Dragon: What protocol you're talking about?
11:30:37B4gdershouldn't it be part 3? ;-)
11:31:05B4gder1 player, 2 recorder, 3 iRiver h100
11:32:24bobTHCor 1 player, 2 recorder, 3 ondio, 4 h1xx ;)
11:32:33B4gdertrue
11:33:54[IDC]Dragonamiconn: the kind of "it plays" annoucement
11:34:25amiconnah
11:34:39B4gderwe never understood the ondio did have sound since we never got that mail B-]
11:34:50bobTHClol
11:34:59[IDC]Dragonamiconn: you've probably been the first to hear it play
11:35:12amiconnI don't think so
11:35:33amiconnYou reported the FM playing already, when my SP did not yet.
11:35:41[IDC]Dragonbecouse of the MMC part, I thought
11:35:55amiconnSmall wonder - the FM uses the same MAS as the recorder, while the SP doesn't
11:36:08[IDC]DragonFM was before mp3?
11:36:18[IDC]Dragonthis doesn't count...
11:36:41Lynx_LinusN: congrats for mp3 playback :)
11:36:53amiconnYes, FM was before mp3, at least with the Samsung tuner, but I don't mean that either
11:37:11ashridahin actual news, the human race was doomed to extinction today, as the robot revolt turned violent.
11:37:25amiconn[IDC]Dragon: After MMC was working, you almost instantly reported mp3 playback sort-of working, while my SP kept silent
11:37:46amiconnI had to adapt the configuration memory addresses first
11:39:07bobTHCi'm agree with dwihno, a new item in the frontpage news section is mandatory for this kind of milestone, imho
11:39:36preglowlet's wait until it doesn't glitch, at least ;)
11:41:13Rickhas anyone looked at iriver's fm?
11:41:44amiconnaustriancoder: If you're looking for another driver to write, you could write an i2c driver for the fm radio chip
11:42:24amiconnWhen this is done, the rest should be fairly easy. The iRiver uses the same philips radio chip as the newer Ondio FM's
11:42:32bobTHCpreglow >with linus on it , debugging dont't take ages ;)
11:43:03austriancoderamiconn: seems to be a nice idea.. will look later the day into the docs
11:43:47 Join lImbus [0] (~manuel@kernel.cycos.net)
11:44:01lImbusLinusN: CONGRATZ !
11:44:12LinusNthx
11:44:28lImbuswhy don't you update the romm topic :-) ?
11:44:34lImbuss/romm/room
11:45:02Rickwhy?
11:45:04Rickit's fine the way it is
11:46:44*rasher added a digest item
11:47:22rasherlet's see if it explodes again
11:48:27preglowthanks to NAME(lol), NAME(hehe) ?
11:48:39preglowaren't those macros? :P
11:49:36Rickhm
11:49:45Ricksomeone should add that remote pic to the RockboxShots page
11:49:45Rick;p
11:50:49Rickalso
11:51:00Rickis it just me or is the fm chip not listed in the progress stuff on the iriver page?
11:51:39rasherpreglow: yes.. no idea why that happens
11:51:50rasheryou'd have to ask Bagder
11:53:02austriancoderRick: do you want to code the driver for the fm chip?
11:53:10preglowthere is a driver for it
11:53:33preglowmight need to be modified for the h1x0, though
11:53:37Rickah
11:53:46austriancoderthough about that... porting
11:53:47Rickis it the same exact model?
11:54:04austriancoderRick: yes or no... i dont want to do dublicapted work
11:54:31Rickaustriancoder: I wouldn't mind it
11:54:35Rickbut
11:54:54Ricki'm busy this week so it would have to wait
11:55:13austriancoderah.. then i will try it
11:55:32LinusNaustriancoder: i2c-h100.c
11:55:57austriancoderLinusN: oki
11:56:46LinusNfirmware/tuner_philips.c
11:57:40LinusNyou might want to take a peek at firmware/drivers/fmradio_i2c.c
11:57:52preglowso, is the playback glitching badly?
11:58:03LinusNoh yes
11:58:21LinusNthe buffering code is not even close to finished
11:58:37austriancoderLinusN: thanks
12:00
12:00:43preglowLinusN: is it possible to say, just have 2 kb iram extra in addition to the 32kb bank?
12:02:12*HCl yawns
12:02:14HClgmorning
12:02:16rasherand again, dwihno excels in amusing ircstats quotes
12:03:10LinusNpreglow: guess so
12:03:22preglowrasher: hahaha
12:03:30HClmp3 playback isn't in cvs yet?
12:03:49Rickno, it's a hack
12:03:55HClmhm
12:03:57HClfigured as much
12:04:03HClkay
12:04:06*HCl goes to shower
12:07:40LinusNpreglow: the stack is in iram
12:08:38*LinusN goes to lunch
12:09:47 Join Shagnar [0] (~tester@p54A0F3CE.dip.t-dialin.net)
12:12:12preglowLinusN: yeah, know, was thinking more of having room for more constants and tables in iram without having to eat another 32kb bank in its entirety
12:20:55***Saving seen data "./dancer.seen"
12:24:50HClmmm
12:27:55preglowthink i'll pop out for a couple of hours time
12:28:25*B4gder grants preglow permission to leave for a while :-P
12:28:41 Quit rasher ("CGI:IRC (EOF)")
12:44:59 Quit austriancoder ("CGI:IRC (Ping timeout)")
12:50:07LinusNthe rss feed is really f*cked up
12:50:25 Quit ehntoo|ZzZz ("Leaving")
12:52:53B4gderthere's something strange in that digest thing
12:53:08B4gderwhen built automatically, it makes zero byte files
12:53:19B4gderwhen I run it manually, it just works
12:54:58B4gderI guess my lameness shines through at times like this :-)
12:57:16 Quit [IDC]Dragon ("CGI:IRC")
13:00
13:03:35amiconnZagor: Something totally FAT-unrelated: Since your server upgrade quite some time ago, logbot is no longer able to explain anything :(
13:03:49Zagoroh
13:04:58LinusNamnesia
13:10:09 Join Aison [0] (~hans@zux166-181.adsl.green.ch)
13:22:23 Quit Shagnar (Read error: 131 (Connection reset by peer))
13:22:35 Join Shagnar [0] (~tester@p54A0F3CE.dip.t-dialin.net)
13:30:18B4gder"Console: switching to colour frame buffer device 128x96"
13:30:20B4gderwooo
13:30:34B4gder(but it is still all black, but anyway...)
13:40:40 Part LinusN
13:42:16Shagnarwho actually works on grayscale ?
13:42:28B4gderno one
13:42:40B4gderthere''s an incomplete patch done
13:43:38Shagnaroh :-/
14:00
14:02:08 Quit lostlogic ("Going to the moon")
14:20:59***Saving seen data "./dancer.seen"
14:21:46 Quit Shagnar ("( www.nnscript.de :: NoNameScript 3.81 :: www.XLhost.de )")
14:23:12preglowpftpft
14:34:04 Join lolo-laptop [0] (~lostlogic@68.251.84.226)
14:35:14 Join webguest11 [0] (~d4f00fa5@labb.contactor.se)
14:35:26 Join muz [0] (~d4f00fa5@labb.contactor.se)
14:35:26 Quit webguest11 (Client Quit)
14:35:41muzhey did linus get mp3 playback to work?
14:36:09B4gderyeps
14:36:29muzlike without converting to wav etc
14:36:34 Quit B4gder ("CGI:IRC")
14:36:40 Join B4gder [0] (~50d8ae51@labb.contactor.se)
14:36:45 Quit B4gder (Client Quit)
14:36:47 Join B4gder [0] (~50d8ae51@labb.contactor.se)
14:36:53B4gdergrrr
14:37:02 Join Shagnar [0] (~tester@p54A0F3CE.dip.t-dialin.net)
14:37:03B4gderhttp://www.rockbox.org/mail/archive/rockbox-archive-2005-04/0161.shtml
14:37:54Shagnarwooha?
14:38:06Shagnari love machniae supremacy ;D
14:41:00muzis this hack "commitable" in the cvs yet?
14:42:54preglowno
14:43:09preglowof course it _can_ be commited, but we'd better wait until it works properly
14:43:48muzoh ok this is great news
14:44:09muzso the api is done?
14:44:28B4gderactually writing code is the best way to design the sound api/architecture
14:47:33 Quit muz ("CGI:IRC (EOF)")
14:54:59 Nick QT_ is now known as QT (as@area51.users.madwifi)
14:55:18 Quit B4gder ("CGI:IRC")
15:00
15:17:43 Join F1^Aison [0] (~hans@zux166-181.adsl.green.ch)
15:25:27 Quit Aison (Read error: 111 (Connection refused))
15:27:41 Join Aison [0] (~hans@zux166-181.adsl.green.ch)
15:34:16 Quit ashridah ("Leaving")
15:34:20 Quit F1^Aison (Read error: 60 (Operation timed out))
15:34:23 Join tedboer [0] (~tedboer@mtg62.upf.es)
15:34:31tedboerhi.
15:34:55tedboerthe rockboxui simulator is monochrome. is that correct?
15:37:21HClyes
15:39:54tedboerbut rockbox perse isn't?
15:40:12tedboerif the hardware is grayscale i mean?
15:41:08preglowit's not utilized yet
15:41:35Zagorrockbox currently does not support grayscale displays. we run the iriver in monochrome too.
15:42:17tedboerok.
15:42:38Zagorit's a work in progress
15:43:02tedboeryeah, i saw all these source files starting with gray_ :-)
15:43:33amiconntedboer: These have nothing to do with core greyscale support
15:43:41Zagorthat's the grayscale plugin library. it's for "faking" grayscale on monochrome displays.
15:44:02tedboerok i see.
15:44:05tedboeranyway...
15:45:09 Join MoosCamaro [0] (MoosCamaro@m214.net81-66-158.noos.fr)
15:45:19MoosCamaroHi all
15:45:32tedboerhi
15:46:07MoosCamarostill congratulations for Linus works
15:46:19MoosCamarohi tedboer
15:46:24Shagnarye
15:46:25Shagnaryes
15:47:39 Join T0mas_ [0] (~Tomas@ip503c08d1.speed.planet.nl)
15:48:28 Quit t0mas ("changing clients")
15:48:33 Nick T0mas_ is now known as t0mas (~Tomas@ip503c08d1.speed.planet.nl)
15:48:37 Quit webguest65 ("CGI:IRC")
15:48:51HCltedboer: there's a patch available to enable grayscale on iriver, its just not 100% finished yet
15:50:28preglowtell markun to stop prioritizing this foolish "school" thing
15:50:59Shagnarhcl - but it can be used with the actuel CVS release of rbx
15:51:02Shagnar?
15:51:34HClyes
15:51:43HClgrayrockbox.zip on my ftp is built with it.
15:51:58HClit has the grayscale patch applied together with an grayscale-enabled rockboy
15:52:08Shagnarweha! do you have me a link?
15:52:20HClcheck the rockboy wiki page
15:52:37tedboerok. cool.
15:52:52HClhttp://www.rockbox.org/twiki/bin/view/Main/WebHome?topic=PluginRockboy
15:56:20Shagnarthx
15:57:42ShagnarMenu Done needs more options <= yeah, more than 0 :D
15:57:49amiconnZagor: Did you do some tests with my FAT patch yet? Maybe I spotted a minor bug, have to check. It might not clear the last sector of a newly allocated cluster for >1 sector per cluster. If so, patched fat.c, line 1349 should read:
15:57:56amiconn(dir->file.sectornum <= (int)fat_bpb->bpb_secperclus); sector++)
15:59:54bobTHCHCl > the PluginRockboy is announce to be compatible with recorder ? what's the reality ?
16:00
16:00:42bobTHCs/compatible/working
16:01:09HClbobTHC: it works. very slowly.
16:01:46bobTHC@5% of RT ?
16:01:52HClsomething like that
16:01:55HCl6fps iirc
16:02:00preglowyo, hcl
16:02:02HClfaster than iriver at 11mhz
16:02:04HClbut not fast enough
16:02:05HClhai?
16:02:14preglowi think you can pretty much forget dynarec being feasible on h1x0
16:02:21preglowthe ram is so incredibly slow
16:02:25HClagreed
16:02:33preglowso _INCREDIBLY_ slow
16:03:10HCli think its better to focus on an asm cpu core
16:03:16preglowyes
16:03:17preglowi agree
16:03:20HCland asm versions of the functions that were converted to asm on x86
16:03:25preglowand using iram for important data
16:03:32preglowargh
16:03:55preglowi wish a time buble just enveloped me and allowed me to code rockbox for a couple of months straight
16:04:50HCl :P
16:04:54HClits called summer vacation :P
16:05:46preglowi'll have to get a job :/
16:06:22preglowit's good bye to long, lovely, good summer vacations for me
16:06:38HCl :X
16:12:03tvelocity[away]hire yourself to develop rockbox
16:12:21tvelocity[away]and with the money you earn you can pay for your hired developer (yourself)
16:12:55tvelocity[away]:PP
16:17:21amiconnHCl: I still think dynarec would make a performance boost, even on iriver, and compared to an asm cpu core. The slow ram performance will most likely hit an interpretive cpu core by the roughly same amount
16:17:54 Quit tedboer (Remote closed the connection)
16:18:05preglowiram
16:18:06bobTHCpreglow > and as tvelocity[away] is a good guy he donate monthly the tantamount to your wage ;)
16:18:20preglowi'm really quite convinced an asm cpu core will be faster
16:18:46preglowtvelocity[away]: hey, i like where this i going, shall we say 100$ bucks an hour?
16:19:02tvelocity[away]hahahahaah
16:19:15*tvelocity[away] hides in a corner
16:21:00***Saving seen data "./dancer.seen"
16:21:10amiconnpreglow: Why? Part of the asm cpu core code might be in iram, but it still needs to read the z80 instructions from sdram. Z80 instructions are usually single bytes, so that might be even slower than running m68k code from sdram (and having it in the icache for loops)
16:26:35HCleither way.
16:26:38HClwe can do both.
16:26:46HCldynarec can be integrated into an assembly cpu core
16:26:59HCland it won't be up to full speed without one anyways
16:27:07HClso its better to write the assembly cpu core first
16:27:13HClthen if thats not fast enough, add dynarec
16:28:34HClat the moment i'm more interested in on the fly playlist creation
16:28:56preglowafter seeing how huge a frigging boost i got from adding some table data to my imdct_l routine, i realized how bloody important it is
16:29:10preglowi even use movem for the data in question, so should be fast also in sdram
16:29:11preglowbut no
16:29:42preglowmost of the interpreter core should fit in the code cache
16:29:51preglowso don't use iram for code
16:30:13preglowhow much ram does the gameboy have?
16:30:42amiconnUp to 4 meg are possible (mostly rom)
16:30:51preglowyeah, i meant ram
16:31:10preglowor can carts add extra ram as well?
16:32:38amiconnI'm not that much into gameboy hardware, but I think this is possible
16:33:24amiconnI don't understand why you're explicitly asking for ram though
16:33:39preglowwell, if it's very little, just stuff it in iram
16:33:51preglowz80 has few registers, so it'll be accessed a lot
16:34:07preglowdoing byte and word accesses on a coldfire is slow enough as it is
16:34:32HClit can only have 16bits of adress space
16:34:34HClat the same time
16:34:43HClof which the lower 8k are rom
16:34:44preglowwhat about bank switching?
16:34:53preglowdoes it do that?
16:34:55HClyes
16:35:02amiconnDo you propose copy-on-bankswitch? Now _that_ may be slow
16:35:31preglowno, i don't
16:35:32HCli meant < 0x8000 are rom o.o
16:35:44HClthats 32kb..
16:36:19HClit has two banks
16:36:36HClwait, maybe i'm wrong about that o.o
16:37:15HCleither way, everything below 0x8000 in gb address space is rom
16:37:41HCl64kb of adress space in total
16:39:21HClpreglow: how fast is mp3 now?
16:40:20preglowvery fast, once i get linus' as well as my last changes in iram at the same time
16:40:29preglowapproaching 300% realtime
16:41:15HClat 120mhz, i assume?
16:41:19HClwhat about 48mhz?
16:41:20preglowaye
16:41:23preglowclose to 100%
16:41:27HClhrm
16:41:33HClis that enough?
16:41:46preglowoh yes
16:42:50MoosCamaropreglow: good progress
16:42:52Shagnarwill linus' mp3-play-test be implemented like the /sample.wav test?
16:43:17preglowwho knows
16:43:18preglowi doubt it
16:43:23HClpreglow: what about 320bps ?
16:43:43preglowHCl: that's what i'm talking about
16:43:46HClokay.
16:43:51HCljust checking
16:44:04t0masShagnar: I asked the same at the mailinglist ;)
16:44:17Shagnargot an answer?
16:44:22t0masnot yet
16:44:26Shagnarkk
16:44:40t0maspreglow: 300% @ 320bps? (120 mhz)
16:44:44t0masthats fast...
16:44:51preglowt0mas: it MIGHT be there once i get all my shit together
16:44:56t0masok
16:44:58preglowafter some more work
16:45:01preglowi just haven't got the time
16:57:50Shagnarrockboy with grayscale is much more better than in monochrome...^^
17:00
17:06:09 Join Psy^Dead [0] (nobby@ACC8B988.ipt.aol.com)
17:06:28Psy^Deadcongrats on the 5 seconds of iriver mp3 :)
17:06:37Psy^Deadgood choice of mp3 to use too
17:06:58CoCoLUSdid i miss something?
17:07:05Psy^Deadmailing list
17:07:24Psy^Deadiriver has mp3 test playback
17:07:29CoCoLUSwow
17:07:39 Join nozomiyume [0] (~vthakkar@ip-139-112.station.sony.com)
17:07:53Psy^DeadTo follow up on the Rockbox tradition, here is the news:
17:07:53Psy^Dead
17:07:53Psy^DeadA few minutes ago, I played the first 5 seconds of real-time MP3 audio on my iRiver H140.
17:07:53DBUGEnqueued KICK Psy^Dead
17:07:53Psy^Dead
17:07:53Psy^DeadThis was an experimental hack, so there are lots of things left to do to enjoy MP3 playback for real, but it's not far away now.
17:07:54***Alert Mode level 1
17:07:54Psy^Dead
17:07:56Psy^DeadNevertheless, we have passed a milestone in the iRiver project.
17:07:58Psy^Dead
17:08:00Psy^DeadOf course, the 5 seconds of audio was Machinae Supremacy, this time The Great Gianna Sisters.
17:08:02Psy^DeadYou can download it here: http://www.machinaesupremacy.com/downloads/machinae_supremacy_-_gianna_sisters.mp3
17:08:04Psy^Dead
17:08:06Psy^DeadKeep on rocking!!!
17:08:08preglowa link would have done graet...
17:08:08Psy^Dead
17:08:10Psy^DeadLinus
17:08:19Psy^Deadim subscribed, i dont have a link :P
17:08:22preglowhaha
17:08:26preglowthe archives are on the web
17:08:40Psy^Deadyeah, but im lazy
17:09:11Psy^Deadcongrats anyway
17:09:16CoCoLUSrockbox ml or rockbox-news ?
17:09:25Psy^Deadi know you did most of the codec porting pre
17:09:37Psy^Deadhttp://cool.haxx.se/mailman/listinfo/rockbox
17:10:00preglowi just did the optimizing
17:10:06preglowporting was done by dave chapman
17:10:07HClyay.
17:10:24CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
17:10:24*HCl yawns
17:10:46Psy^Deadtired or ungrateful? :P
17:10:52HCltired o.o
17:11:00preglowhahah
17:11:04preglowhe think he tried to diss my work!!!
17:11:25preglowi can see no other way out of this than through a duel
17:11:50preglowtrout shall be my weapon
17:12:52HCli'll take the extra thick edition of the hhgttg
17:13:35preglowi've got that, and it's a light weight
17:13:39preglowfeathery paper
17:13:45preglowyou shall fall to my trout
17:14:04preglowor maybe i'll bring my kreyzig book
17:14:14 Quit Shagnar (Read error: 145 (Connection timed out))
17:14:27 Join muesli- [0] (~muesli_tv@dsl-082-082-147-241.arcor-ip.net)
17:14:32muesli-high
17:14:32CoCoLUSargh
17:14:38CoCoLUSthree new star wars trailers :)
17:17:55***Alert Mode OFF
17:29:37muesli-l8er..
17:29:39 Quit muesli- ("ich will Kühe!!!")
17:30:29 Join einhirn [0] (Miranda@bsod.rz.tu-clausthal.de)
17:30:30 Quit einhirn (Read error: 54 (Connection reset by peer))
17:30:44 Join einhirn [0] (Miranda@bsod.rz.tu-clausthal.de)
17:40:59 Nick Lynx_ is now known as Lynx_awy (HydraIRC@134.95.189.59)
17:46:10 Quit Psy^Dead ()
18:00
18:00:26 Join rasher [0] (~3e4f4094@labb.contactor.se)
18:21:04***Saving seen data "./dancer.seen"
18:27:58 Quit bobTHC ("Smoke Weed Every Day !")
18:28:37amiconnmrf
18:31:53 Join Stryke` [0] (~Chairman8@resnet-241-86.resnet.umbc.edu)
18:34:06 Join LinusN [0] (~linus@labb.contactor.se)
18:36:03preglowi must be a masochist
18:36:37preglowi'm actually itching to try optimizing the mdct in tremor
18:36:56 Part lImbus
18:37:05preglowit's actually quite emac friendly
18:38:14preglowi do wonder how the PROD32 functions will do, though, the top bit is automatically shifted away by the multiplier
18:39:30preglowif that bit is actually used, we've got a problem
18:53:42HCli thought markun was saying something about how the top 32bits being the most important ones..
18:53:53HCland him saying about how we couldn't use the emac for that cause of it
18:53:59HCli might be wrong though
18:54:01preglowof course they are
18:54:57preglowall my imdct_l opts use the top bits with emac instructions
18:55:02preglowso he is quite obviously wrong
18:55:28preglowi'd be happy to tell him how to use the emac in such a way, just have him ask
18:56:05preglowvorbis optimizations don't look to be too hard, they often use a fixed point format that is equivalent to what the emac unit uses internally
18:57:12HClmhm.
18:57:35HClif i'm not too busy with my AI and compiler creation classes, i'll prolly take a look at converting dumb to fixed point
18:57:46 Join Sucka [0] (~NNSCRIPT@host81-156-210-120.range81-156.btcentralplus.com)
18:57:51preglowdo that, i won't have time to do it for a long while
18:57:56preglowi'll give you some pointers, though
18:58:01HClplease do :)
18:58:15preglowjust let me know when you're at it
18:58:53*t0mas is away: dinner :)
19:00
19:00:01amiconnZagor: I found another bug in my FAT code - it would have failed in the FAT16 root. Again - argh!
19:00:03amiconnAnd I checked the 'clear whole clusters' thing, this is working correctly. Phew.
19:00:07amiconnI've uploaded a fixed version of my patch to the usual place. I've added some more debug info as well.
19:00:34amiconn...would have failed in certain cases, of course
19:02:43preglowHCl: dumb will probably be a iriver only codec unless you also plan on writing fixed point functions in c
19:02:58HClmmmmmm?
19:03:03HCl*watches his cat race by*
19:03:20*amiconn is awaiting the MAS .wav 'codec'
19:10:45 Quit einhirn ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
19:14:29preglowi'm away
19:18:52HClhttp://www.rockbox.org/twiki/bin/view/Main/WebHome?topic=PluginRock
19:18:52HClboy
19:19:02*HCl slaps putty
19:24:52HClamiconn: would be interesting, but would it be of any use?
19:24:59HClaside from playing wavs o.o
19:25:05HClhm. i guess it might make rockboy sound work too
19:25:12HClif it wasn't that the archos is way too slow for it
19:47:48 Join stevenm [0] (~steve@177-96.mam.umd.edu)
19:48:46stevenmHello people
19:48:52HClhey stevenm
19:49:05stevenmI just cut off a few operatios.. Anyone feel like timing midi2wav?
19:49:06HClany progress from your side of the world?
19:49:19HClmmm, if you're prepared to wait roughly half an hour
19:49:21stevenmYea.. I cut out a few shifts
19:49:22HClbit busy at the moment
19:49:30stevenmSure..I have physics now
19:49:42stevenmthen another class after that.. if my battery lasts that long, then sure
19:49:48HClokay
19:49:50HCl :)
19:49:55stevenmyea
19:50:18stevenmI am hoping like, 75% realtime out of this
19:50:29HCl :)
19:50:41stevenmhaven't committed the code yet though, want to make sure it actualy works first
19:50:47stevenmso I'll upload a .tar
19:54:24HClk
19:59:04HClokay
19:59:04HClgot time
19:59:15HClif i can find my iriver
20:00
20:00:47HClwhich i can and have
20:01:10stevenmHi
20:01:15HClhey
20:01:38 Quit Chamois (" HydraIRC -> http://www.hydrairc.com <- Leading Edge IRC")
20:01:46stevenmokay, here is tarfile: http://wam.umd.edu/~stevenm/midi.tbz2
20:01:54stevenmthat untars into /apps/plugins
20:02:23stevenmand you'll need the soundset if you dont have it.. http://wam.umd.edu/~stevenm/patchset.tbz2 .. untars into .rockbox
20:02:37stevenmand here's the test file: http://wam.umd.edu/~stevenm/mi1.mid
20:06:18HClokay
20:06:31HClsorry, i had to rewrite a script a bit since it was malfunctioning
20:06:35*HCl gets
20:06:41stevenmsok
20:06:50stevenmmidi2wav.rock then goes to .rockbox/viewers
20:07:09stevenmmake sure its clocked to 120mhz or else it take forever
20:09:05 Join webguest12 [0] (~c10bdb25@labb.contactor.se)
20:10:07 Quit webguest12 (Client Quit)
20:13:17*HCl goes to build
20:13:42 Quit HCl ("Lost terminal")
20:13:54 Join hcl [0] (hcl@titania.student.utwente.nl)
20:13:58hclsucks.
20:14:00hclanyways
20:14:17 Nick hcl is now known as HCl (hcl@titania.student.utwente.nl)
20:14:29stevenmthere you go
20:15:27t0mashas anybody looked at radio for iRiver?
20:15:45rashernot yet, austriancoder said he might
20:16:12rasheror he may already be at it, this was 10 hours ago, and I left in the middle of the conversation
20:16:19t0masghehe
20:16:22t0mashe's pretty fast ;)
20:16:30*t0mas wanted to take a look at it too..
20:16:43rasheraround 12:00 in the logs
20:16:51t0mas(not that I think I can do it... but still :))
20:17:29rasherah, no word from him since
20:18:39t0mashmz... the chip is the same as the ondio?
20:18:50t0masso the code is there? we just need an I2C driver? :)
20:18:55rasheryeah
20:19:46 Join muesli- [0] (muesli_tv@hmln-d9b8e257.pool.mediaWays.net)
20:20:21 Join silencer_ [0] (~silencer@zen.via.ecp.fr)
20:20:33t0masto bad I know nothing of that I2C thing...
20:20:37stevenmrasher, love the IRC stats :)
20:20:44muesli-hoi
20:21:06***Saving seen data "./dancer.seen"
20:21:58 Quit silencer (Read error: 104 (Connection reset by peer))
20:22:26HClstevenm: how do i build midi2wav
20:22:26HCl?
20:22:30HClits not getting built
20:22:35HClmaybe i need to cvs update >.>
20:22:38HCl*goes to check*
20:22:43stevenmyea.. it's in SOURCES in CVS
20:23:05stevenmyea you prolly wanna get the latest CVS.. then untar that tarfile in there. I put it in sources, added to viewers.config, etc
20:23:39*t0mas is bored... and the only thing I can think of to code is to difficult... as I don't know that I2C thing
20:23:41HClyea
20:29:17HClok finally
20:29:19*HCl goes to test
20:29:26HClt0mas: you can make a start with the runtime lib
20:29:43t0maswhat should it do? :)
20:29:46HClt0mas: it pretty much involves searching for the code that gets executed when a song is done playing
20:29:47stevenmHCl, see how long it takes ..
20:29:56HClt0mas: keep track of how many times a song has been played
20:30:04t0masooooh that idea
20:30:11HCland keep track when a song was last played
20:30:11t0masfor playcount/autoplaylist?
20:30:21HClfor advanced playlist generation
20:30:28t0maswasn't rasher working with that too?
20:30:37HCl"give me all rocksongs of the 80's and 90's that last longer than 3 minutes that have been played at least 5 times"
20:30:40HClstuff like that.
20:30:45HClwell, mostly database stuff.
20:30:48HClthere's not any code yet.
20:30:59*HCl goes to test the midi
20:31:04stevenmHCl, thanks
20:31:05amiconnHCl: The MAS .wav codec should allow playing .mods as well
20:31:17rasherI'm probably not going to be of much help
20:31:22rashercode-wise
20:31:24HClamiconn: are mods low profile enough?
20:31:44amiconnThink so, at least the 4-channel protracker (-derivates)
20:31:58HClmk
20:32:09HClokay
20:32:12amiconnIf a stock 7 MHz AMiga is able to play them, an 11..12 MHz should as well
20:32:13HCltime for my timed midi run
20:32:19HCltrue
20:32:39HClthe autodetection of rolo isn't working for me properly
20:33:01HClstevenm: it crashed.
20:33:09stevenmHCl, hmm? What it say?
20:33:10HClI0B:Line-F
20:33:29stevenmHCl, what's that? How far did it get in loading?
20:33:33muesli-yepp, got this in rockboy too
20:33:44HClit showed \<midifilename>
20:33:46HCland then it crashed
20:33:51HClmuesli-: mm?
20:33:57stevenmHCl, you have the patchset in .rockbox ?
20:34:02HClyes
20:34:15*HCl goes to get a paperclip
20:34:37muesli-HCl when i started super mario i got an error msg like this and the player hang up
20:34:37stevenmHCl, that is strange.
20:35:13HClhmyea
20:35:17HClit seems all plugins are broken
20:35:20HClat least rockboy too.
20:35:27stevenmHCl, ALL plugins ?
20:35:38HClwell, at least rockboy
20:35:51stevenmHCl, Ah, I see
20:36:05HCland yes
20:36:06HClall plugins
20:36:09HClminesweeper crashes too
20:36:22stevenmHCl, we tested midi last night and it ran fine.. and the only changes I made would have caused a crash on patch load
20:36:26stevenmhmm.
20:36:30stevenmTHAt is not good
20:36:48HCleven pong breaks
20:36:55HCli'm not gonna be able to test anything with current cvs
20:36:57stevenmall with the same error?
20:37:01HClyes
20:37:08stevenmstrange
20:37:11stevenmLinusN, any ideas >
20:37:12stevenm?
20:38:11HClcan someone explain/tell me why the rockbox logo on the remote is upside down?
20:39:06amiconnHCl: (crashing plugins) Seems your plugins don't match your rockbox.iriver (plugin api mismatch), and the api version was not bumped
20:39:07rasherHCl: your build sounds brooooken
20:39:24HCli'll try to rebuild completely..
20:39:35HClbut as far as i know
20:39:37HCli already did
20:42:03rasheroh, no.. getting I0B as well
20:42:26rasheron a fresh build
20:42:47stevenmso, what happen ?
20:43:14rasherhaha, helloworld works... sortof
20:43:18stevenmSomething in recent CVS ?
20:43:56HCllinus changed the memory location thing recently...
20:43:57HCli think..
20:44:13stevenmHCl, yea I see that in the recent CVS.. would that affect plugins somehow?
20:44:18rashersounds like that could be it
20:44:20HClyes.
20:44:22rashertry "helloworld" plugin
20:44:27rasherit outputs garbage
20:44:29HClhelloworld "works"
20:44:30HClyea.
20:47:36t0mashm...
20:47:36t0mas-#define DRAMORIG 0x30000000
20:47:36t0mas+#define DRAMORIG 0x31000000
20:47:41t0mascan that be a problem?
20:47:43HClis there any way to revert certain files to the old version?
20:47:51t0mascheckout the old version?
20:47:54HClhow?
20:48:08t0masI don't know the options out of the blue
20:48:14t0masnormally use a graphical client
20:48:33HCli'll just change everything back to the 3000..
20:48:36HClsee if it helps
20:48:37t0mascvs checkout -D date
20:48:51t0masyesterday might work?
20:49:30HClmk..
20:49:36HClalready changed everything back by hand
20:49:56t0masok, is that working?
20:49:59HCldunno
20:50:01HClcompiling
20:50:34stevenmchange of classes, back in 10 min
20:51:13HClk
20:51:26LinusNall plugins must be compiled
20:51:29LinusNrecompiled
20:51:34HCli did that.
20:51:35HClit broke.
20:51:42LinusNoops
20:52:27HCllet me guess
20:52:30HClplugins.lds =]
20:52:47LinusNyup
20:53:03*t0mas is waiting for an *OOPS* commit ;)
20:54:27LinusNdone
20:55:12t0maswow that's fast
20:56:13*HCl tries steven's timing thing now
20:56:40amiconnLinusN: Btw, what was the problem with the old sdram start address?
20:56:54*HCl is curious to that too.
20:57:51LinusNthe sdram was mirrored in an unfortunate manner
20:59:24HClcausing..?
21:00
21:00:10LinusNbadness :-)
21:00:15*HCl wonders whether it causes/fixes his upside down rockbox logo on his remote..
21:00:18HClokay
21:00:32LinusNthe 32mb was suddenly two copies of the first 16mbytes
21:00:39HCleep
21:00:49HClhow'd that happen o.o
21:00:50 Quit stevenm ("Leaving")
21:01:02LinusNread the comments in crt0.S
21:01:06HClkay
21:01:08amiconnLinusN: Was that the real cause of the freezes you observed?
21:01:21LinusNoverwriting the vector table
21:01:23 Join stevenm [0] (~steve@177-96.mam.umd.edu)
21:01:28stevenmI'm back
21:01:34stevenmHCl, any luck w/ the plugin?
21:01:48HClrecompiling
21:01:57stevenmok
21:02:33 Join webguest78 [0] (~513ece6b@labb.contactor.se)
21:02:48CoCoLUSbtw, how was the 5 second mp3 playback done, like the wav test or in something that could be called a start of the multi codec architecture?
21:03:13LinusNstart of the codec arch
21:03:23muesli-kewl
21:03:29muesli-cheers linus
21:03:37CoCoLUSmade my day ;)
21:04:06HClLinusN: rolo doesn't seem to detect an rockbox firmware update on my iriver, any idea why?
21:04:27LinusNsometimes it just doesn't :_)
21:04:36HClk o.o
21:05:01HClum, steven
21:05:05HClit said could not load file
21:05:28*HCl stares at it saying "i hope this works"
21:05:47stevenmHCl, is there HD activity ?
21:05:51HClyes
21:06:14HCl1 minute 15 seconds passed, including the loading patches bit
21:06:20stevenmHCl, interesting, so it must be synthesizing. I wonder what though
21:06:28stevenmHCl, what file could it not load?
21:06:41HClum, i just clicked the select button on the midi file
21:07:18stevenmHCl, all right... I guess it looks like it is running. I am just curious, what specific error did you get?
21:07:18HCli'm just hoping it'll stop eventually o.o.
21:07:32HClcan't say, it went too fast, something with unable to load file
21:07:33stevenmHCl, at 120Mhz it should not take more than 2 and a half min
21:07:37HClthen it proceeded to load patches
21:07:37HClouch.
21:07:39HCl>.>
21:07:42HCli forgot 120mhz
21:07:43HClsorry
21:07:53stevenmHCl, it started NAMING patches?
21:08:03HClnaming? o.o.
21:08:13stevenmlike, started showing their names on the screen
21:08:34stevenmbecause it would have said "Error opening [something].pat" if it cannot find the patchset
21:08:40HClgonna start again
21:08:42HClno, not that
21:08:55HCl"Could not open file"
21:08:58HCl"LOADING PATCHES"
21:08:59stevenmHold on a sec
21:09:11HCland then it quickly says something about -7 and size
21:09:11 Quit webguest78 ("CGI:IRC (EOF)")
21:09:19HCl "/mi1.mid"
21:09:22HClit said that before everything
21:09:58 Nick tvelocity[away] is now known as tvelocity (~tony@ipa138.2.tellas.gr)
21:09:58stevenmHCl, -7 size is a .wav thing, that always happens with xxx2wav
21:10:03HClokay
21:10:27stevenmHCl, it says 'Could not open file" if the file descriptor for the .mid file is 0. I was under the impression that 0 means invalid
21:10:38stevenmHCl, so if it got changed to -1 or something, then no worries
21:10:39HClno, -1 means invalid
21:10:57stevenmHCl, ah, that explains it ... .. the sim version must be different
21:10:57HClokay, i didn't time this run though..
21:11:09HClit ended
21:11:12stevenmHCl, ah.. well you should (hopefully) get a .wav
21:11:16HCllet me try again and time
21:11:17stevenmmiditest.wav and it should be working
21:11:20stevenmHCl, thanks
21:11:36HCl9:11:28 starting time
21:13:09rasherstevenm: if you want to time, you can look at how battery_test does it .. it's super-easy
21:13:26stevenmrasher, will do
21:13:31HClheh
21:13:36HClexactly 2 minutes
21:13:40stevenmrasher, I just squashed a few shifts in this one, and precomputed a few things
21:13:43stevenmHCl, thanks
21:14:00rasherrb->get_current_tick() or something
21:14:01stevenmWow that is 78% realtime
21:14:06HClinteresting, the plugin fix also flipped my rockbox logo the way its supposed to be
21:14:09rasherjust subtract that from start/end, and you're good
21:14:14HClon the remote
21:14:27rasherthat's pretty crazy
21:14:51stevenmHCl, just out of curiosity.. the output file, miditest.wav, is it mangled or sounding ok?
21:14:56HClno idea
21:14:57amiconnstevenm: 0 is always a valid file descriptor, also for the simulator. It's just that you won't see fd == 0 on the sim very often.
21:14:58HCllet me play.
21:15:12stevenmamiconn, all right, will change
21:15:47stevenmamiconn, yea, it's -1 in other places.. I must have missed one here
21:15:56HCli know this song.
21:15:56HClo.o;
21:16:00HClecco
21:16:04HClof sega
21:16:05HCloh wait
21:16:05HClno
21:16:05DBUGEnqueued KICK HCl
21:16:05HClxD
21:16:08stevenmHCl, working?
21:16:09HClmonkey island
21:16:10HClyup
21:16:12HClworks fine
21:16:12stevenmyep
21:16:17stevenmthanks, excellent
21:16:42HCl1:34
21:17:15stevenmHCl, yea that should be file length
21:17:31stevenm94 sec / 120 sec = 0.78
21:17:45CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
21:17:45*HCl nods
21:17:52stevenmI know this is not the best way of doing it because different parts of the file have different load
21:18:03stevenmbut it works for average and gauging if it got any faster
21:18:31stevenmnext I have my eyes set on the output buffer and the huge if statements, see if that can be sped up any
21:18:47lolo-laptopstevenm: do you think there is enough room for optimization to do midi playback in battery/heat feasible time?
21:18:54 Nick lolo-laptop is now known as lostlogi (~lostlogic@68.251.84.226)
21:18:55 Nick lostlogi is now known as lostlogic (~lostlogic@68.251.84.226)
21:19:19rasherI don't think heat will be a problem
21:19:22rasheror.. it won't
21:19:41stevenmlostlogic, I am hopeful. This is only optimization in C.. after nothing more can be done, I guess I can ask someone to help me optimize in ASM, like use emac, etc
21:19:43 Join ep0ch [0] (~ep0ch@213-208-105-54.dyn.gotadsl.co.uk)
21:19:50rashernot since we run the cpu at max 120mhz
21:20:04stevenmlostlogic, heat I don't think is a big deal because we decided 140Mhz was too hot.. so then 120Mhz is fine ?
21:20:11lostlogicgotcha
21:20:27ep0chquick question, does writing to disk add much of an overhead?
21:20:32stevenmwe can always cut the sampling rate again :)
21:20:46t0masep0ch: yes
21:20:50t0masit slows things down
21:20:56stevenmt0mas, this thing writes once every 3000 samples
21:21:13stevenmt0mas, you think that causes too much slowdown ?
21:21:15HClyou cut sampling rate?
21:21:15rasherI ran my iriver for 5 hours straight at maximum cpu usage without overheating
21:21:17t0masdon't know
21:21:23ep0chso stevenm might be running at realtime with disk output off
21:21:25HClstevenm: how many times per second is that?
21:21:32stevenmHCl, yea right now it is at 22050Hz
21:21:35t0masep0ch: don't think so
21:21:38ep0chah
21:21:43t0masstevenm: you can test it?
21:21:44stevenmwell let's see here. it writes every 3000 BYTES
21:21:47t0masjust remove diskwriting...
21:21:57stevenmHCl, you wanna do a test w/o writing?
21:22:00t0masand run it with a stopwatch next to it ;)
21:22:00HClthen yes. remove diskwriting
21:22:04HCldefinately :P
21:22:08*t0mas too :)
21:22:14stevenmHCl, all right, go to midi2wav.c
21:22:27HClmhm
21:22:41stevenmComment out line 214
21:22:43t0masin plugins right?
21:22:50t0masoutputBufferPosition=0;
21:23:00HClk
21:23:07stevenmnah I was gonna say, comment out the write(...) call
21:23:09t0masbetter comment out the whole if...
21:23:30t0masor is that line I copyed needed?
21:23:36stevenmyea actually comment out everything in that loop but synthSample()
21:24:16HClick.
21:24:17HClokay.
21:24:18stevenmI actually wanna change the buffer to ints so that we don't have all the register overhead than that
21:24:48stevenmand none of that crazy incrementing crap, just assign it and stuff. then it becomes faster.. worked for getsampe
21:25:09*HCl runs to get his chrger
21:25:10HClcharger
21:26:09lostlogicwill this midi playing on a portable audio player be a first? I do not know of any others...
21:26:59HClcounting
21:27:09ep0chanyone looking at DSP effects?
21:27:22HClnafaik
21:27:47HCldoh!
21:27:50*HCl forgot 120mhz again
21:27:51HClsorry
21:27:56stevenmwhoops
21:28:02stevenmthere is no disk activity this time, right ?
21:28:08HClnope
21:28:11stevenmcool
21:28:24ep0chthere's a noise shapening plugin for foobar that would be damn nice in rockbox
21:28:27stevenmlostlogic, I guess it'd be pretty cool if it actually gets up to speed
21:28:51HClcounting
21:28:58stevenmbtw I just subscribed to rockbox mail list.. sorry about the comments and code, etc. will fix
21:29:05stevenmcoding style/tabs
21:30:13HClhm....
21:30:23HClnot realtime
21:30:38ep0chmore than realtime? ;)
21:30:46HClthats disappointing
21:30:52HClit only made a difference of 4 seconds? o.o
21:31:23HCl1:56..
21:31:49stevenmHCl, I see.. did you comment out just the write or the whole outputBuffer[...] stuff too
21:32:01HCloutputbuffer stuff too.
21:32:05stevenmah
21:32:11stevenmmaybe it is done in the background or something
21:32:24stevenmwell there is still ASM stuff.. plus stuff to be done in C
21:32:31muesli-l8er ladiez...
21:32:40*rasher tries adding a timer to midi2wav
21:34:18 Quit MoosCamaro (Read error: 104 (Connection reset by peer))
21:34:55stevenmrasher, sok, don't worry about it, I got it. Latest version isn't in CVS anyway
21:35:17t0maslol
21:35:25t0masout of total boredom I wrote a disco plugin :P
21:35:41t0masflashing both backlights :P
21:35:59rasheroh dear
21:37:05 Join asdsd_ [0] (asdsd@h-67-100-28-186.miatflad.dynamic.covad.net)
21:37:24rasherI want the plasma demo ><
21:37:28HCl :P
21:37:31preglowLinusN: the wav playing crashes were due to the sdram mirroring?
21:37:31HClwhy don't we?
21:37:41HClwe have grayscale, just need to finish the grayscale patch
21:37:49HCldoes anyone know whats lacking from the grayscale patch?
21:37:59rasherlook at the files
21:38:05rasherlots of functions are blank
21:38:10HClokay.
21:38:15rasherthen optimizations of course
21:38:18HCli need to sleep properly for once...
21:38:19HCl :/
21:38:24preglowwimp!
21:38:24HClthen i can take a look at it
21:38:35HClno, seriously, i have insomnia :/
21:38:51rasherIf you have an uptime fixation, go here: http://rasher.dk/updike/ (sortof nsfw)
21:38:56preglowi thought you meant to go to bed earlier
21:39:02preglowactually not being able to sleep sucks worse
21:39:10HClyea..
21:39:11preglowit's actually one of the suckiest things in existence
21:39:15 Join Strath [0] (~mike@dgvlwinas01pool0-a202.wi.tds.net)
21:39:24 Join GnagelRam [0] (~chatzilla@gnagelram.olf.sgsnet.se)
21:40:00t0masghehe
21:40:04t0masrasher? uptime in what?
21:40:09t0masdays? hours?
21:40:21rasherdays I guess
21:40:28stevenmtry sleeping here.... it's like 90 degrees in our rooms because the univ. is too cheap to turn on A/C
21:40:49t0mas[tomas@nigel tomas]$ uptime
21:40:49t0mas 21:40:40 up 122 days, 9:41, 1 user, load average: 0.00, 0.00, 0.00
21:40:56HCl 21:38:49 up 46 days, 8:31, 11 users, load average: 4.00, 4.00, 4.03
21:41:02stevenmrasher, what the devil .. ?
21:41:12 Quit muesli- (Read error: 60 (Operation timed out))
21:41:20*HCl flexes his stable load average of 4, meaning all 4 of his processors are used to the max :p
21:41:24rasherwe all know uptime is proportional to penis size, why hide it?
21:41:39t0maslol
21:41:46rasher 21:41:36 up 32 days, 9:57, 2 users, load average: 1.63, 0.90, 0.70
21:41:51*t0mas wins :P
21:42:12rasherwhich reminds me.. I should tape the powerswitch
21:42:17t0masbobby:~# uptime
21:42:17t0mas 21:42:12 up 243 days, 12:15, 1 user, load average: 0.07, 0.02, 0.00
21:42:19t0mas:P
21:42:23rashershut it down one night trying to turn on the light :(
21:42:41t0masand my router :)
21:42:42t0mas[root@bruce root]# uptime
21:42:42t0mas 21:42:21 up 122 days, 11:40, 1 user, load average: 0.16, 0.03, 0.01
21:42:51stevenmmine's only 16 days
21:43:02stevenmand it's about to go to zero. battery almost dead
21:43:02t0masall three debian ofcourse :P
21:43:15t0maslaptop? :)
21:43:18stevenmthis here be a gentoo laptop. gotta love suspend
21:43:19stevenmyea
21:43:27t0masmine never makes it past one day...
21:43:28stevenmI don't think I've rebooted in weeks
21:43:39stevenmt0mas, I suspend it a lot
21:43:47t0masirritating high sound of the disk kills me...
21:43:55stevenmoh yea
21:43:55t0masah... that's a way :P
21:44:08t0maslol
21:44:17stevenmmy roommate has a 64 bit AMD with some insane graphics card and 2 huge hard drives. One is a Raptor- so much heat
21:44:18t0mascheck the page with 243 days :P
21:44:24stevenmwe can't sleep in that place
21:44:30stevenmhe never turns it off
21:44:59 Join hubbel [0] (hubbel@h9n1fls302o1033.telia.com)
21:45:08*HCl 's server is located near the washing machine for exactly that reason..
21:45:29rashermine's in the kitchen
21:45:37rasherbut.. still pretty close to my bed
21:45:48rasherat least there's a door
21:45:52DBUGEnqueued KICK asdsd_
21:45:52asdsd_,00hey guys
21:45:59t0masmine are downstairs...
21:46:09t0masin a closet...
21:46:12t0masnever heard them
21:46:18HClmy server has 7 8cm fans :P
21:46:19rasherI have no downstairs :)
21:46:50hubbelcool.. my sound driver is in the cvs :) how is the audio api going?
21:46:59stevenmOkay I have to go now. THis thing is about to die
21:47:00stevenmbye all
21:47:08t0masbye
21:47:11 Quit stevenm ("Leaving")
21:47:19HClhubbel: its going, linus played the first 5 seconds of an mp3 this morning..
21:48:30hubbelHcl: nice! I saw he fixed a DRAM bug today aswell
21:48:37HClyea
21:48:55HClthere's also a rockboy sound driver, and we play the first 8mb of a wav.
21:49:36hubbelI'm almost have the motivation to tweek and get recording to work now :)
21:49:50hubbellast time it crashed horribly =)
21:50:25HCl :p
21:54:00 Quit hubbel ()
21:55:23CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
21:55:23*HCl feels like fixing up grayscale and getting it into cvs, finally
21:55:31 Quit StrathAFK (Read error: 110 (Connection timed out))
21:55:31preglowthat would be plain smashing
21:55:44rashertotally
21:55:46HClbut first i want to eat some food and finish this anime
21:55:51 Quit asdsd_ ("Trillian (http://www.ceruleanstudios.com")
21:55:57preglowfood sounds good
21:55:58rasherPlasma demo!!
21:56:15HClrasher: i'll finish the grayscale patch if you make the plasma demo :P
21:56:30rasherthere *is* a plasma demo
21:56:35HCloh
21:56:35rasherin the patchtracker
21:56:35HClo.o
21:56:39HClokay
21:56:39HClo.o
21:56:45rasherI want :D
21:56:49HClyes yes.
21:56:55HClyou want you want you need you need.
21:57:03rasherI suuuuuure do
21:57:06*HCl goes to check on food
21:57:16rasherwell, I just get excited :)
21:59:26HCl :P
21:59:35HClwell, i know how the new format works, i know how the old format works
21:59:41HCli don't see why it should be too difficult :)
21:59:46HCli'll check on it after i finish food
22:00
22:00:49 Nick tvelocity is now known as tvelocity[away] (~tony@ipa138.2.tellas.gr)
22:03:37amiconnrasher: The plasma demo will need some fixup. It uses floating point...
22:04:04 Part GnagelRam
22:04:39rasheraw
22:04:46preglowgood reason to learn fixed point math
22:04:47preglow!
22:05:11rasherI guess you could say that
22:05:12elinenbeHCl: grayscale in CVS = good
22:05:13 Quit Zagor (Remote closed the connection)
22:05:17preglowamiconn: this wav codec you are talking about accepts plain pcm data, yes?
22:05:28preglowi thought you meant .wav files, heh
22:05:29rasherwould probably be better if I did it with something I understood
22:05:55preglowno patches today either :/
22:06:41 Quit webmind (Connection timed out)
22:06:46amiconnpreglow: Yes, we should call it "PCM codec" to be precise. For some reason it is almost always called WAV codec...
22:07:24preglowamiconn: how the flaming hell did they avoid bundling such a thing to begin with?
22:07:44amiconnMost likely an IP problem
22:08:17amiconnThe MAS is developed by Micronas, and they offer a very primitive PCM pass-through
22:08:41amiconnThis PCM pass-through can only handle 44.1kHz 16 bit stereo, with no buffering
22:08:59preglowhahah, the plasma uses LONG DOUBLES
22:09:10rashersounds great
22:09:12preglowthe only float format gcc doesn't support emulating
22:09:16amiconn...and from the MAS' parallel port input, which is not connected in the archos boxes
22:09:36preglowamiconn: this'll be more flexible?
22:09:36amiconnpreglow: It actually works on archos
22:09:51 Join stevenm [0] (~steve@181-208.mam.umd.edu)
22:09:55preglowthen gcc has long double routines for sh
22:09:58preglowit doesn not for gcc
22:10:02stevenmah the joys of an air conditioned room
22:10:06preglowat least they weren't listed in the file i looked at
22:10:43amiconnAfaik, the Archos multimedia jukeboxes, also use the MAS for audio decoding, and obviously there was demand for PCM support
22:11:33amiconnSo archos developed their own PCM codec, which is said to be much more capable. Handling different sample frequencies, using buffering etc.
22:11:48preglowhahah
22:11:49amiconnWe *believe* that it also support serial transfers
22:11:59preglowa ton of code just for generating a sine table
22:12:20rasherpreglow: plasma?
22:12:22preglowyes
22:12:34rasher^_^
22:12:40amiconnpreglow: If that's the only place were floats are used, the fix should be simple
22:12:40HCldarn
22:12:51HCli really need to fix my icq so i can reach markun..
22:12:52preglowthe fix will be simple anyway
22:13:03preglowi just don't see why he doesn't use a precalced sine table in such a little plugin
22:13:16preglowif you use interpolation, you need very few points
22:13:30amiconnIirc it was for educational purposes
22:13:39stevenmwould it be possible to reverse engineer the firmware in the other models and see how the PCM codec works?
22:13:43preglow128 points with linear interpolation is sufficient for high fidelity audio use, so it's much, much more than adequate here
22:13:49amiconnOtherwise he could just have copied the sine table from cube.c
22:14:04amiconn...which is 1 degree resolution
22:14:14preglowhokay
22:14:32preglowif its educational i guess there's a point to it
22:14:41amiconnUnfortunately that one uses the odd fixed point format I once mentioned
22:14:59amiconn(divide by a power of 10 instead of a power of 2)
22:15:22preglowthere are no floats in the main loop
22:15:38preglowjust switching everything from 'long double' to 'double' should have this working on iriver as well
22:15:42*amiconn should really get up to full speed rockbox dev again
22:15:55preglowprovided someone code the grayscale lib for iriver as well, of course :V
22:16:09HCldo you have your iriver yet amiconn?
22:16:11amiconnWah, I *want* my iriver
22:16:17HClthats a no
22:16:18HCl :P
22:16:20amiconn§!$§"!
22:16:24preglowamiconn: call the shop, dress them down for being fools and get one another place
22:16:46HCli say try ebay since there aren't gonna be many shops that'll still have them
22:17:31amiconnThere are quite some H-1xxs on ebay now
22:17:36preglowi can imagine
22:17:45*amiconn missed 2 H-140s ending today
22:18:04rasherwhat did they go for?
22:18:11 Join muesli- [0] (muesli_tv@hmln-d9b8e263.pool.mediaWays.net)
22:18:15amiconnSlightly above 300,- €
22:18:15 Quit Aison (Read error: 145 (Connection timed out))
22:18:27rasherjesus christ
22:18:39rasherthat's what I paid for mine, new
22:18:53amiconnWhere?
22:18:55rasheroh wait, mine's h120
22:19:08amiconnAh, that explains it
22:19:38rasherBut mp3players.co.uk .. half a year ago though
22:19:44rasherI don't think they have them there anymore
22:19:45stevenmI wanna get one.. you guys think it's worth waiting for the 3x0 port or get an h1x0 now?
22:19:54HCli have no idea.
22:20:02HCli think in the end i might prefer 3x0
22:20:02muesli-re
22:20:17amiconnstevenm: I'd definitely prefer H-1x0
22:20:26HClamiconn: why?
22:20:27stevenmamiconn, any reason?
22:20:40HClaside from no optical out (i think?)
22:20:48CoCoLUSrockbox runs on it? :)
22:20:50amiconn..because I don't like energy-consuming colour displays on mobile devices, but maybe that's just me
22:20:51stevenm..assuming there is a rockbox 3x0 port ... ... there is going to be one, right ?
22:21:02HClthats a good point.
22:21:07***Saving seen data "./dancer.seen"
22:21:17HClstevenm: LinusN said it should be fairly easy
22:21:22*amiconn 's mobile phone is also b&w only
22:21:39stevenmawesome! color rockbox
22:22:12muesli-yepp, i am also thinking about a change to 3x
22:22:45muesli-that colour display is awesome
22:22:59t0masit just eats your batts out...
22:23:07*lostlogic is _eagerly_ waiting for H3x0 support.
22:23:17preglowamiconn: colour lcds consume almost no extra power compared to a b&w display
22:23:19rashergbc on h3x0 :D
22:23:23muesli-you dont need it, just use your remote as usual
22:23:30stevenmIs 3x0 faster?
22:23:41lostlogicstevenm: not afaics
22:23:43muesli-nope, same hardware stuff built in
22:23:49stevenmah
22:23:49muesli-afaik
22:23:51*micoo is _eagerly_ waiting for gmini400 support.
22:23:59preglowmicoo: don't hold your breath
22:23:59micoo:(
22:24:03amiconnpreglow: Not the display itself, but almost all of them *require*backlighting in order to be able to decipher anything
22:24:07lostlogicslightly different /upgraded chips here and there in the H3, but nothing major
22:24:09preglowamiconn: indeed
22:24:35preglowif you never use your backlight, that will be an issue
22:24:54stevenmI was showing my friend ajbrec w/ rockbox and video plugin. He has korean firmware on 3x0
22:25:02amiconnHowever, I know that it's possible to build colour LCDs that are readable with very little or no extra backlighting if the ambient light is strong enough
22:25:02t0maseh... the white colorlcd backlight uses more power than thow simple blue leds right?
22:25:05stevenmhe was very excited to say he was watching stargate on that at 8 fps
22:25:16amiconn...because I actually saw such a display recently
22:25:45stevenmI was like.. yea, rockbox has video at 70 fps.. on a chip 10x as slow
22:25:55preglowhaha
22:26:01preglowthe coding technology is slightly different...
22:26:06*amiconn wonders why not all mobile device manufacturers use such types of colours LCDs
22:26:14 Join ehntoo [0] (~noclue2@24.177.161.77.up.mi.chartermi.net)
22:26:31muesli-i have seen on oled on a pmp
22:26:40preglownever saw an oled
22:26:40muesli-that was really amazing!
22:26:51muesli-really phantastic
22:26:58muesli-no viewing angle
22:27:15amiconnoled is most likely not teh big power saver either
22:27:22muesli-doesnt matter how you look at it, its ALWAYS great
22:27:35muesli-no, oled needs less energy
22:27:48muesli-the hidden trap is the lifecycle
22:27:53amiconn..than colour lcd with backlight, yes
22:27:57muesli-its limited so far
22:28:10amiconn...but not compared to b&w lcd without backlight
22:28:45preglowi do prefer lighted surfaces anyway
22:28:53preglowfood
22:29:06muesli-i have seen it at the cebit..once you own you dont want to change anymore..
22:29:17muesli-it
22:29:38muesli-http://www.neosoltech.com/eng/technology/oled.php
22:29:58muesli-check the right box
22:30:52rasherWell this is a company that *sells* oleds :)
22:30:55t0mas"power consumpion under 5mW"
22:30:58t0masfor what size? :P
22:30:58muesli-nope
22:31:07muesli-http://www.neosoltech.com/eng/product/product_tmr.php
22:31:22muesli-i have seen this player on the fair
22:31:43rasherI see.. they still have good reasons to be very positive about it :)
22:31:59*t0mas starts thinking of the next rockbox port :P
22:32:24muesli-its the future..plasma and tft have sufficient disadvantages
22:32:40rasherPlasma is the devil
22:33:02rasherif nothing else, then because they weigh a TON
22:33:10HClokay
22:33:20*HCl fixed up trillian so it'll do icq again
22:33:34*muesli- yawns at hcl: great
22:33:39muesli-:)
22:33:54HCl :P well, after i finish my anime, i'll look at grayscale
22:34:01HClit mostly means that i should be able to contact markun again
22:34:35 Join Tang_ [0] (~chatzilla@ATuileries-151-1-24-83.w82-123.abo.wanadoo.fr)
22:34:35muesli-we are yearning for grayscale :D
22:34:51Tang_himuesli
22:34:53Tang_:)
22:35:08muesli-high tang :D
22:35:17Tang_:)
22:35:32t0mashi
22:35:50t0masHCl? have you ever looked at the dutch rockbox translation?
22:35:56HClno xd
22:35:57Tang_hi t0mas
22:36:01Tang_:)
22:36:03HCli absolutely hate software thats in dutch xD
22:36:06t0masghehe
22:36:14t0masI use most things in English too
22:36:30muesli-go oranje go :)
22:36:35t0masghehe
22:36:39HClanyways.
22:36:44*HCl goes to look at the grayscale patch
22:36:55*muesli- jumps
22:37:03thegeekI recently realized that I actually use english more than my native tongue(norwegian)
22:37:22 Join matsl [0] (~matsl@1-1-4-2a.mal.sth.bostream.se)
22:37:32muesli-all scandinavians and dutch speak english very well
22:37:40stevenmsweet
22:37:44Tang_lol
22:37:50stevenmis there anyone else here from the us ?
22:38:07muesli-i am yealeous on your tv program, no dubbed moviez etx
22:38:09muesli-etc
22:38:09thegeekI wish I lived in a large city
22:38:12thegeeklike london
22:38:15thegeekor even oslo
22:38:28rasherI'd die living in a large city
22:38:30thegeekdubs = evil
22:38:36muesli-yepp
22:38:47thegeekI refuse to watch anything that is dubbed
22:38:51thegeekit's just too horrible
22:38:52t0masmuesli-: we have to speak English...
22:39:00t0masnobody speaks dutch
22:39:23muesli-we do not..but those idiots transform any englisch word in to german
22:39:25t0masand dutch translations of most software have _REALLY_ bad error messages and things like that
22:39:28preglowi hate large cities
22:39:31muesli-we call it denglisch
22:39:36HClum.
22:39:36preglowbut does indeed look like i'll be moving to oslo
22:39:36t0masghehe
22:39:43*HCl stares at the current grayscale implementation.
22:40:04HCli don't think it should have the same api as the grayscale library.. simply because its a native framebuffer and simply can't do somethings a virtual framebuffer can...
22:40:22HClsome things*
22:40:36rasherCOMPLY!
22:40:41rasherer.. mcm
22:40:46rasherwrong window
22:41:01muesli-rockbox.zip 20-Apr-2005 22:40 494K
22:41:06muesli-was that you hcl?
22:41:10HClhm?
22:41:14HClwhere/when?
22:41:21muesli-right now
22:41:23HClon my ftp?
22:41:29muesli-nope, on rbx
22:41:30HCleh? o.o;
22:41:31HClno.
22:41:41 Quit Harpy (Read error: 60 (Operation timed out))
22:42:02amiconnHCl: If you're going to work on greyscale for iiver, please consider http://www.rockbox.org/twiki/bin/view/Main/WebHome?topic=GraphicsAPI as well
22:42:35muesli-i would really apreciate a software reset btw
22:42:44t0mas?
22:42:45preglowlike how?
22:42:47t0masresetting settings?
22:42:59t0mas(that's already possible)
22:43:11muesli-just press i.e 3 buttons to do a reboot
22:43:23rashernot possible
22:43:24amiconnIt's my shot at a unification of the various graphics apis present so far, trying to combine the advantages and to avoid the disadvantages
22:43:25t0mascan't press 3 buttons
22:43:26*HCl reads
22:43:46rasherplus, you can just rolo rockbox.iriver
22:43:52muesli-just one after another?
22:43:56rasherthat's as close as you'll get to software reset anyway
22:44:09muesli-rolo?
22:44:10t0masyeah, rolo rocks for testing :D
22:44:13t0mastnx LinusN again :D
22:44:27t0masmuesli-: go to a .iriver file on the player
22:44:30t0masand "play" it
22:44:33muesli-just a plugin that crashes? ;)
22:44:49t0masno, it reloads
22:44:58muesli-crashes..that causes a crash..
22:44:59muesli-;)
22:45:00t0masyou can have multiple .iriver files... and load them all
22:45:24HClamiconn: it would imply all grayscale things would be in 1byte/pixel format?
22:45:27HClbitmaps anyways
22:45:39amiconnHo, it doesn't
22:45:41muesli-what is that .river files good for?
22:45:42amiconn*No
22:45:51HClvoid xxx_gray_bitmap(unsigned char *data, int x, int y, int width, int height, int stride) 3) 3) Draw a greyscale bitmap. This uses the canonical greyscale format, 1 byte/pixel
22:46:02 Join belgarath [0] (~acd82a31@labb.contactor.se)
22:46:04HClso we wouldn't be able to use native format bitmaps
22:46:12HClor am i misunderstanding that..
22:46:23 Quit belgarath (Client Quit)
22:46:32 Join belgarath [0] (~acd82a31@labb.contactor.se)
22:46:44HClor is the blit function gonna do that
22:46:45HCl?
22:47:18t0masmuesli-: that .iriver file is the firmware loaded by the bootloader
22:47:29HClamiconn: what should the "xxx_" become?
22:47:32amiconnHCl: xxx_bitmap() ..... "Draw a *native* bitmap
22:47:43HClah. right.
22:47:44HClokay.
22:47:48HCllooks good then
22:47:58HClwhat should the xxx_ be?
22:48:45t0masamiconn? I have that bmp thing working... and from plugins its easy to use the pluginbuffer... but when I do something in core... should I just declare a static image buffer? or are there other ways?
22:48:56amiconnWell, I didn't think about that yet. Maybe it should just be lcd_ for the core implementation as it is now
22:49:15HClokay
22:49:21HCllcd_ are using the same functions..
22:49:25HClright?
22:50:06amiconnt0mas: As I already said, my idea would be that the core module which actually uses the images does its own buffering (e.g. wps)
22:50:19t0masyeah, that's what I'm doing now
22:50:56amiconnHCl: The idea is to unify the graphics apis, so some work is required for the b&w archos lcd_ implementation as well
22:50:56t0masbut in wps I can just declare a simple unsigned char images[2560]... or do something intelligent.. problem: I don't know how to do it intelligent :P
22:51:20amiconnThere is no more intelligent way
22:52:13t0mas:(
22:52:17t0masok
22:52:30amiconnHCl: Then I also have a number of ideas to optimise both speed and code footprint, which are not part of the api documentation
22:52:57amiconnThe basic idea is already present in the grayscale library
22:54:24HClmhm..
22:54:35 Quit belgarath ("CGI:IRC (EOF)")
22:54:44*HCl sighs at other people talking to him on a different chat and distracting him
22:54:55HClamiconn: so pretty much, i'd have to alter both iriver and archos to use the new api?
22:55:05amiconnFor displays with <1 byte/pixel, or when one byte in the raster memory equals more than one pixel for some other reason (planar display or such), have 2 groups of very lowlevel functions
22:55:16 Join Biptoria [0] (~51982fd4@labb.contactor.se)
22:55:31amiconnOne group is for manipulating single pixels, and the other for manipulating whole pixel blocks
22:55:40BiptoriaAnything new today my loverlys ?
22:55:41HClyou lost me o.o
22:56:05amiconnHCl: You wouldn't need to do the archos work of course
22:56:20BiptoriaA pixel is made up of 3 colours
22:56:20rasherBiptoria: Linus played first bits of mp3 on his iRiver today
22:56:23HClyea.. what about the <1 byte / pixels....
22:56:23t0masBiptoria: see mailinglist
22:56:29HCli didn't understand what you meant
22:56:41BiptoriaHow well did it play?
22:56:49rasherit didn't play well :)
22:57:03BiptoriaGet back to work then rasher
22:57:09rasher.x
22:57:44amiconnHCl: I mean, for such displays like the archos (or iriver in b&w mode) one byte matches 8 pixels, and for iriver 4-grey it matches 4 pixels
22:57:49HClyes
22:57:51HClwhat about it..?
22:58:12amiconnSo if you manipulate whole blocks at once, this should definitely boost performance
22:58:33HClum. okay *still confused*
22:58:47HCli'll just try to implement that new api as lcd_* in lcd-h100.c, right?
22:59:13amiconnlcd_bitmap() does what I mean in the b&w implementation
22:59:20BiptoriaInstead of having the CPU piss around with lots of blocks, it pisses round with larger less blocks ?
22:59:27stevenmHey guys, does anyone mind if I commit all the midi code, or is anyone else messing with that?
22:59:35HCli say go ahead
22:59:52amiconnlcd_fillrect() simply re-uses lcd_bitmap(), which is way better than doing single pixels, but not yet optimal
22:59:54HClcvs is after all a concurrent versioning system.
23:00
23:00:07amiconnThe bad example is lcd_invertrect() ...
23:00:10HClamiconn: i'll leave the optimizing for later.
23:00:15HCli'll just try to implement
23:00:22stevenmAll right, here goes
23:00:41HClofcourse
23:00:46BiptoriaWill i be able to play MP3 if i use the dailybuild ?
23:00:51HClthat was if my cat doesn't claim my lap
23:00:53rashernope
23:00:55amiconnHCl: Iiuc one reason for not yet incorporating markun's patch is its slowness
23:01:03BiptoriaWhy not :|
23:01:08ehntooLinusN hasn't committed any mp3 code to CVS yet.
23:01:10 Quit ep0ch (" HydraIRC -> http://www.hydrairc.com <- IRC has never been so good")
23:01:12rasherBiptoria: because it's not in cvs yet
23:01:19BiptoriaThe monkeh
23:01:22HClamiconn: i don't see that as a reason. rockboy was slow at the start too.
23:01:25HCljust like libmad
23:01:38BiptoriaRockyboy still is slow isnt it /
23:01:38rasherthere's no point in committing it yet
23:01:39Biptoria?
23:01:39amiconnYes, but rockboy isn't core code
23:02:10amiconn...and libmad is simply needed to get anything useful out of the unit
23:02:23amiconnGreyscale support is not really needed though
23:02:28HCli still think that it should just be put in with a comment that its slow and needs optimization o.o;;;
23:02:37amiconn(although being nice of course)
23:02:38BiptoriaWhats the diff between Greyscale and B&W?
23:02:45ehntooa world of difference.
23:02:59BiptoriaWhats it using at the mo ?
23:03:07HClrockboy, jpeg viewer
23:03:09HClmandelbrot
23:03:11Tang_sorry if i prompt your discussion
23:03:18HClplasma would use it too
23:03:38Biptoria:| i meant the menu
23:03:40HClbut okay
23:03:45Tang_just to give "request" from MR
23:03:51amiconnHCl: The biggest performance killer is the b&w bitmap drawing in 4-grey mode.
23:03:55Tang_about playlist
23:03:55HCli'll just keep petting my cat if people don't want grayscale in cvs..
23:04:01ehntoothe menu is... 4 color b&w at the moment, if I remember
23:04:12HClamiconn: i don't see why that would be a performance killer
23:04:15amiconnIirc markun's patch does that pixel-wise, which is a really bad idea because b&w bitmap drawing also means text drawing...
23:04:18BiptoriaPoor cat
23:04:24Tang_"in the futre do u think itd be possible for rbx to simply recognize wma's as audio so u could open up rbx put the wmas in a play list and save it, reboot and then use it on the normal firmware? sorry i no this prolly isnt a very desirable feature. but it seems like it wood be easy to do."
23:04:38HClBiptoria: i doubt he would've crawled on my lap for pets if he hated it
23:05:27BiptoriaIts probably hungry, or just wants to pee on you
23:05:28Tang_(request from eternalwolfman: http://www.misticriver.net/boards/showpost.php?p=197920&postcount=555 )
23:05:43ehntoobeing pee-ed on by pets isn't fun.
23:05:57HClnaw, kittys are cute
23:06:40HClTang_: should be easy
23:06:42BiptoriaDid you survive the gay ordeal the other night HCI
23:06:50Tang_:)
23:07:02Tang_i imagine indeed
23:07:03HCli tried to ignore anything i could possibly hear
23:07:06HCland close both of my doors
23:07:19HClfortunately, we have 20 cm thick soundproof walls.
23:07:24HClthat are way too hard too.
23:07:35HClthey're built so that if a fighter jet crashes into our flat
23:07:37Biptoriabut the excitment just got too much so you joined in
23:07:39HClit'll only get 3 rooms far
23:07:40Tang_i there is two HCl?
23:07:46ehntooI'm kinda wishing I hadn't read that.
23:07:52Tang_:|
23:07:52HClBiptoria: not really -.-
23:08:06*HCl has a gay flatmate who's a jerk and sometimes has "visitors"
23:08:07HCl>.<
23:08:14HClanyways.
23:08:16HClnot important.
23:08:25*HCl pets his cat and watches his anime
23:08:33Biptoria:|
23:08:39HClamiconn: i really don't see why slowness would be an argument against development
23:08:42HClits development after all
23:08:52BiptoriaIs your pet called anime ?
23:08:55stevenmHCl, you have a what?
23:09:12HClstevenm: a gay person living in my flat, in the room next to me.
23:09:24stevenmHCl, ah, I see.
23:09:45amiconnHCl: I don't say that it shouldn't be developed, I only say it shouldn't be in cvs without at least some critical points being optimised
23:09:54HClwhy?
23:10:17ehntoohmm... HCl, was that jet fighter comment a reference to the Dirk Gently novels? Just wondering.
23:10:24HClthe who? O.o.
23:10:27HClno, i was serious.
23:10:30Biptoria:| lol
23:10:31HClwe have an airforce base nearby
23:10:40HClwe even have special soundproof glass
23:10:46Tang_:)
23:10:48amiconnPerhaps it's just me, but a current cvs build should always be in a usable state. Unfortunately rockbox cvs doesn't use branches
23:11:03rasherHCl: You know hhgg but not Dirk Gently? :-O
23:11:08HClrasher: yes.
23:11:17BiptoriaSo, theoretically you could kill your flatmate and no one would hear
23:11:24HClamiconn: i don't see any slowness or "not usable" on my iriver with grayscale O.o.
23:11:38HCli usually run a grayscale version so i can play rockboy with grayscale.
23:11:39rasherHCl: *sniff* it's a DNA character from two other books he wrote
23:12:22Biptoriais the sound sorted on Rock boy yet ?
23:12:33HCldepends on what you call sorted.
23:12:35HClthe sound driver works.
23:12:41HClrockboy is just not realtime yet.
23:12:52ehntooHCl: DNA wrote Dirk Gently's Holistic Detective Agency, and The Long Dark Teatime of the Soul. The second one had something on a jet crashing through the main character's flat.
23:12:56BiptoriaYeh its slow as hell on mine
23:13:03amiconnHCl: Okay, maybe it's fast enough for the current state of rockbox on iriver. I can't check that myself, you know :(
23:13:12HClyea, i know..
23:13:25HClbut yea, i think its fast enough, and we can always just optimize it asap
23:13:34HClbut first
23:13:38HCli need to get my cat out of my lap
23:13:40HClwithout offending him
23:13:43 Quit lostlogic ("Client exiting")
23:14:02HClthere.
23:14:21Biptoriapoor thing
23:14:32ehntoolol
23:15:05Biptoriayou should of got a dog, you can sell them for more 2nd hand
23:15:21 Quit muesli- (No route to host)
23:15:25HCldogs drool, smell, and are too stupid to walk theirselves.
23:15:33ehntoowell, not always.
23:15:34HCland they don't clean theirselves either
23:15:36BiptoriaMuesli my main cereal
23:16:04HClamiconn: i should just create the lcd_* things of the new api for iriver, right?
23:16:08Tang_good night all
23:16:11Tang_cheers
23:16:16BiptoriaI know a dog that walks it self, it only has 3 legs
23:16:39rashercat's are annoyingly independent
23:16:51Biptoriayeh, there gay
23:16:52Tang_i rather cats
23:16:55Tang_:)
23:17:03Tang_cause they are independants
23:17:08Tang_and curious
23:17:12Tang_unlike dogs
23:17:14BiptoriaBoring
23:17:18Tang_:D
23:17:27amiconnHCl: yes
23:17:38BiptoriaAkk they want is food then they piss off for the rest of the day
23:17:40stevenmour hs physics teacher hated cats with all his being
23:17:53stevenmalways colliding an elephant with a cat
23:18:01stevenmcats, and his mother-in-law
23:18:14Biptoriawhats wrong with elephants
23:18:19Tang_i would want to know
23:18:30ehntootoo big, you can't shove an elephant in a cupboard when you get mad at them.
23:18:35Tang_if elephant always fall on their legs like cats
23:18:39stevenmThey have a lot of momentum when THROWN at a cat
23:19:03Tang_unfortunately no zoo directors let me try
23:19:05Tang_:D
23:19:16Tang_(time to go to bed i guess)
23:19:16HClokay
23:19:25HCli'll get on it as soon as people stop talking to me x.x;
23:19:36BiptoriaYeh im off aswell. love you HCI
23:19:44HCl*forces a smile*
23:19:45*Biptoria waves
23:19:48HClbye.
23:19:54 Quit Biptoria ("CGI:IRC")
23:20:19ehntooyou can stop smiling now.
23:20:26stevenmwoot! no more midi2wav.c warnings
23:20:50rasherthat's the first time in a while that iriver warning count has gone down :)
23:20:55stevenmeep.. there's one on normal target..
23:21:09stevenmand I know why too. I changed a parameter
23:21:15stevenmbut still.. I killed like, 9
23:21:35ehntooI see that. good work.
23:21:56stevenm:) thanks
23:22:09stevenmand I converted all the tab/space stuff and the comment style. that took a while
23:22:39HClwe need an astyle line on the wiki that people can use..
23:22:50stevenmfortunately mindlessly changing // to /* */ is only slightly less boring than listening to the guy praddle on about physics
23:25:10HClamiconn: what should the default mode be?
23:25:11amiconnhmpf
23:25:32HClthats not a valid mode :p
23:25:32amiconnsorry, wasn't directed to you
23:25:37amiconn:)
23:25:40ehntooheh
23:26:11HClwhat should it be?
23:26:15HClsolid?
23:26:37 Quit Tang_ ("Chatzilla 0.9.68a [Firefox 1.0.3/20050414]")
23:27:09ehntoo*sigh* I miss my gentoo setup... I've gotta get back at configuring it.
23:27:33 Join amiconn_ [0] (~jens@pD95D1FEA.dip.t-dialin.net)
23:27:48 Quit amiconn (Nick collision from services.)
23:27:48 Nick amiconn_ is now known as amiconn (~jens@pD95D1FEA.dip.t-dialin.net)
23:28:19amiconnblurf
23:28:26amiconnHCl: default mode should either be FG or SOLID
23:29:18stevenmanyways, for me it is food time. good bye all
23:29:22HClok
23:29:37ehntoohave a good period of nourishment ingestion.
23:29:41 Part stevenm ("Connection reset by beer.")
23:30:32HClamiconn: i could use a description on the wiki to how the mode affects each function..
23:30:51HCl_blit for example, does the mode get ignored there or what?
23:31:18 Join ashridah [0] (ashridah@220-253-122-114.VIC.netspace.net.au)
23:32:17amiconnAh, hmm, I think I made that clear enough. The description of the function should tell you whether it uses the mode or not
23:32:18 Quit courtc (Connection timed out)
23:32:27 Join courtc [0] (~court@adsl-154-41-244.asm.bellsouth.net)
23:32:47amiconnBasically, all drawing function use the mode, with the exception of the non-b&w bitmap functions
23:33:30amiconnAll other functions (scrolling, updating, specials) do not.
23:33:46HClokay
23:33:48HCldidn't see that
23:33:48amiconnlcd_clear_display() counts as a drawing function in respect to that
23:33:49HClsorry
23:33:55HClmhm
23:34:06HClokay.
23:34:58amiconnWell, lcd_clear_display() is a 'half'-drawing-function, in that it should use the background colour, but not the mode
23:36:20amiconnIt is debatable whether the xxx_blit() function is part of the graphics api at all. In fact, it's a low-level access function for the lcd that circumvents the normal graphics api...
23:36:51*amiconn realises he's talking too much and coding too little
23:38:02HCl :p
23:38:03amiconnrasher: Your stats are really great :-/ I'm the second-most talker (word count wise), while being around for far less time than Linus :-O
23:38:04*HCl works on stuff
23:39:32HCldoes anyone know how many intermediate grayscale settings the iriver lcd can have?
23:39:35HCl4?
23:39:41HClor 8?
23:40:31HClnm
23:41:17preglow4
23:41:27preglowblack, white and two user definable
23:41:38preglowif that's what you meant :>
23:41:54preglowor do you mean how many levels of gray you can have?
23:42:02HClwell, i meant what you can set the user definable too
23:42:08HClbut i found some code of markun for that already
23:42:10preglowhmm
23:42:15preglowthe data sheet has that info
23:42:20preglowaight
23:42:38HClamiconn: should i make dummy rgb functions for set_foreground etc
23:42:40HClor leave them out?
23:43:15amiconnNo rgb necessary for greyscale displays
23:43:20HClmhm
23:43:21HClk
23:43:44amiconnforeground and background should just use a single value for them
23:44:23amiconn(0..3 in case of iriver, with 0 == black and 3 == white, the 2 grey levels being equally spreaded)
23:45:13amiconnAh no, of course it should take 0..255
23:45:29amiconnThat's simple to convert anyway, just shift right by 6
23:47:12amiconnLinusN: Are you there?
23:48:23HClif something in mode complement draws on something, it should invert the foreground pixel if there is one, right?
23:48:38amiconnyup
23:49:36HClcan you explain the FG/BG mode?
23:49:56HClBG will only draw on background pixels(empty) and FG only on foreground pixels (already drawn pixels) ?
23:50:17HClshouldn't it be draw background pixels with current foreground colour?
23:50:36preglowehh? i thought fg and bg scheme was only used in text mode-ish stuff
23:50:51amiconnNo, foreground and background isn't related to what is already on the display, but what your to-be-drawn primitive looks like
23:51:01preglowright
23:51:03HCloh. right.
23:51:09amiconnpreglow: Text mode and 1-bit bitmaps
23:51:13HClso we only draw the fg that exists and the bg if exists
23:51:21HClsoo.. what should drawpixel do in that mode?
23:51:23amiconnYes.
23:51:24HCljust draw fg color?
23:51:32*preglow gets an urge to code more on his ansi viewer
23:51:51amiconnLook at the table under 'concepts'
23:51:56preglowdrawpixel shouldn't have a bg, if you ask me
23:52:27HCldoh. i'm stupid.
23:52:29HClor i'm sleepy.
23:52:32HClone of the two.
23:52:34HClthanks.
23:54:11amiconnpreglow: Yes, all non-bitmap primitives consist of foreground pixels only. So the 4 modes actually become only 2 modes
23:55:41preglowmore tea for me
23:55:43amiconn...and the mode is global, as are the foreground and background values, but I think this less of a problem than having to pass all of these parameters to each primitive
23:55:45HClwhats the bresenham algorithm?
23:55:59HClor shall i just implement drawing lines with setpixels?
23:56:09HClwe can always optimize later
23:56:23preglowwhen drawing a line, you can usually just use setpixel
23:56:29preglowthe offsets have to be recalculated anyway
23:56:36preglowbresenham algo you can find tons of implementations of on the net
23:56:48 Quit rasher ("CGI:IRC 0.5.4 (2004/01/29)")
23:56:55amiconnThe free-directioned lines are drawn with single pixels anyway
23:57:23preglowyou can have a special case for hline and vline unless they are separate functions
23:57:27preglowthen you don't use putpixel
23:57:27amiconnThere already is a working implementation of that, using bresenham for calculating which pixels to set
23:57:55amiconnpreglow: Yes, I propose separate functions for hline and vline
23:58:08preglowyes, good idea
23:58:19 Join rasher [0] (~3e4f4094@labb.contactor.se)
23:58:32amiconnI also do not propose to throw everything out and start from scratch. There's much code that is reusable
23:58:44HClyea, i'm noticing that.
23:58:50HClso far its been moving stuff and adding a switch for the drawmode

Previous day | Next day