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 2006-02-07

00:00:15 Quit Henrico ("[BX] Did somebody say gtkBitchX?")
00:00:29 Part ep0ch ("Kopete 0.11.1 :")
00:01:08*petur is installing colinux
00:01:25peturshould I try Debian or Gentoo Deluxe?
00:01:36peturthose are the options I get...
00:01:41*BHSPitMonkey installs petur into a wood chipper
00:01:47BHSPitMonkeywhy's that
00:03:04 Join XavierGr [0] ( be used as build environment for RB
00:04:01lostlogicNicoFR: shouldn't be too hard −− there is actually already code in the seeking part of the WPS code that is supposed to handle it, but it's kludgy code, there should be a better way.
00:04:20lostlogicNicoFR: perhaps manually calling the set_elapsed callback at the end of seek
00:04:24JdGordonpetur: id go deb, only coz gentoo is a pita
00:04:47 Quit ashridah ("Leaving")
00:04:49JdGordonbut its personal preference really...
00:04:59JdGordonalso, if its only for rb then id say stay with cygwin
00:05:51peturI figured installing colinux would take less time than building a crosscompiler under cygwin :)
00:06:45JdGordonjust use the cygwin packages?
00:06:59JdGordonand y? ud still need to compile them under colinux
00:08:05JdGordonalso, it means u need to use vim or something in colinux to edit the files.. unless u use a seperate partition for it
00:08:10JdGordonnot a hdd image..
00:08:50peturtimeout! more tomorrow...
00:08:53 Quit petur ("here today, gone tomorrow")
00:18:08 Join sucka [0] (
00:18:48 Quit actionshrimp (Nick collision from services.)
00:18:56 Nick sucka is now known as actionshrimp (
00:19:37 Join RotAtoR [0] (
00:25:33 Join t0mas [0] (n=Tomas@unaffiliated/t0mas)
00:27:08JdGordonwhy was the wps never done to allow text to be positinoed by co-ords? would have made the whole thing much siompler...
00:27:25BHSPitLappythat's what I've always said
00:27:35BHSPitLappyprogress bars, too
00:27:52LinusNJdGordon: take a guess
00:27:58JdGordontoo lazy?
00:28:05LinusNguess again
00:28:20 Join webguest58 [0] (
00:28:21BHSPitLappynot lazy enough?
00:28:48LinusNthe answer is 1) the wps was once developed for a 2-line character lcd
00:28:49MoosJdGordon:think to a minimum of userfriendly things, that *complicate* things
00:29:05LinusN2) we felt it was good enough for the archos recorder lcd as well
00:29:18JdGordonah, ok
00:29:20LinusN3) "lazy"
00:29:21JdGordonmakes sense...
00:29:25BHSPitLappywell it's not user friendly anymore!
00:29:46Moosmore than with coordinates for all
00:29:54LinusNBHSPitLappy: in what way?
00:29:54BHSPitLappynot really
00:30:00Moosthink to simply user
00:30:01 Quit webguest58 (Client Quit)
00:30:04BHSPitLappyit's a beyotch to make a wps
00:30:05 Quit DreamTactix291 (Read error: 110 (Connection timed out))
00:30:12LinusNBHSPitLappy: no it isn't
00:30:14BHSPitLappygoing off of guesswork
00:30:28LinusNBHSPitLappy: it is a pain to make a wps woth lots of pretty pictures and colors
00:30:34JdGordonhow many targets still only have charachter lcds?
00:30:41LinusNJdGordon: one
00:31:05LinusNi'm not saying that we should keep the current system
00:31:38BHSPitLappyJdGordon: did your color thing ever get committed?
00:31:40LinusNi'm just responding to your question why the wps is like this
00:32:02MoosLinusN: historical cause :)
00:32:08LinusNas always
00:32:14Mooshehe :-)
00:32:20JdGordonhmm.. ok
00:32:31BHSPitLappy"the country was still at war, and hitler's forces were growing ever stronger..."
00:32:34JdGordonwhat im thinking could be very nice.. and workable on the char lcds..
00:32:47LinusNi tend to trigger on the "why isn't it it done like this" questions
00:32:48 Quit t0mas ("brb, forgot a little kernel option... again...")
00:32:49BHSPitLappy"the rockbox community knew we needed -something- to get them by on the players of the time"
00:33:09MoosLinusN: I'm very happy that with those colors LCD, Rockbox tend to change the way a bit
00:33:27Moosin the *good* side :-)
00:33:29BHSPitLappyJdGordon: ?
00:34:22JdGordonwould something like this work? have the wps format like... tag|x|y|<conditions..> and u put the text in the conditions part.. and ud need a new tag for regular text... just have a long list of that...
00:34:23LinusNthe problem with placing text in arbitrary pixel positions is the scrolling
00:34:26JdGordonif that makes sense...
00:34:34 Join webguest40 [0] (
00:34:38BHSPitLappyJdGordon: did your color thing ever get committed?
00:34:46JdGordonnot yet..
00:35:04BHSPitLappydoes it work on 176x132?
00:35:09*Cassandra wonders where on earth these people who post feature requests come from.
00:35:19LinusNCassandra: hell
00:35:20JdGordonLinusN: that could be got around if u allow for either line x or pixel x...
00:35:30Cassandra"Please add a pointless feature to rockbox in order to help me break the law with impunity."
00:35:40JdGordonBHSPitLappy: should do.. it resizes for the screen
00:35:44LinusNCassandra: that's a good one
00:36:10LinusNJdGordon: you mean scrolling text on lines and non-scolling in pixel positions?
00:36:20CassandraI have a very short response to that.
00:36:21MoosCassandra: that sounds more like philosophical meaning :)
00:36:33JdGordonwould it work? should i play today?
00:37:02BHSPitLappyJdGordon: actually I think the coords should be rectangle dimensions
00:37:31JdGordonBHSPitLappy: why? use the fonts to figure it out...
00:37:39amiconnLinusN: I don't see the principal difference in placing scrolling text at pixel position vs. lines
00:38:07XavierGrI have to agree with amiconn
00:38:15amiconnEach scrolling line will need a set of attributes anyway (foreground colour, perhaps font, style...)
00:38:25LinusNamiconn: there is no principal difference, except that we need to rewrite the scroll thread for pixel positioned scrolling text
00:38:32amiconnx and y position (and width) would just be an extension of that
00:39:02JdGordonis there any way to scroll small bits of lines? or only whole lines?
00:39:10LinusNi'm all for a rewrite of the scroll thread
00:39:31LinusNJdGordon: there is a patch for that
00:39:33amiconnOf course, if someome positions two scrolling lines so that they partially overlap, the result will look strange at best,
00:39:45amiconnbut I don't think this is a problem
00:39:48 Quit midkay (Read error: 104 (Connection reset by peer))
00:40:38JdGordonfine.. i gtg.. ill have a play with the wps screen when i get back
00:42:55 Quit hardeep ("BitchX: made with real honey.")
00:45:29BHSPitMonkeyit could be like the Bounce plugin :P
00:48:17XavierGrLinus: Are you busy with something at the momoent?
00:49:15Mooshopefully with iAudio things :-)
00:49:31XavierGryes I know that's why I am asking
00:50:40 Join midgey34 [0] (
00:51:21 Join Chamois [0] (
00:51:21 Quit perldiver (Connection timed out)
00:52:05 Join perl|bbq [0] (
00:56:54LinusNyes, i'm working with the x5
00:57:36ChamoisLinusN :do you have good results ?
00:58:05LinusNnot yet
00:58:53JdGordonb4 i go.. updated the yesno screen for latest cvs
00:58:58 Quit __peer__ ()
01:01:45 Quit ender` (" The Japanese make comics about everything from ninja to samurai to superheroes to magical girls to ordinary high school life)
01:02:03LinusNJdGordon: does it work for the archos player?
01:03:53 Join perldiver [0] (
01:05:11 Part Chamois
01:07:53***Saving seen data "./dancer.seen"
01:10:45 Quit imphasing ("Lost terminal")
01:19:48 Quit perl|bbq (Read error: 110 (Connection timed out))
01:19:50 Join imphasing [0] (
01:25:27jaebirdlinuxstb: I'll give it a try!
01:32:26 Join mymomthelush [0] (
01:35:02JdGordonLinusN: i dunno...
01:38:45JdGordonLinusN: crap.. nup
01:40:27LinusNguessed that :-)
01:40:42JdGordonis that the 2 line char lcd?
01:46:52NicoFRJdGordon: would you have a screendump of your patch ?
01:47:13JdGordon i think
01:47:40NicoFRthx... it's quite nice :)
01:50:11 Quit youngcereal (Read error: 110 (Connection timed out))
01:54:28 Join youngcereal [0] (
01:55:30JdGordonwhats the context button on the archos?
01:56:55JdGordonwow... those keys r wierd :p
02:00:47jaebird4g grayscale boots
02:03:08 Quit Moos ("Glory to Rockbox!")
02:03:17 Join mikearthur [0] (
02:03:55 Quit mikearthur (Remote closed the connection)
02:04:34 Quit NicoFR ()
02:04:34 Join mikearthur [0] (
02:06:03jaebirdI'm a little confused...I build rockbox-devel, what in addition to the rockbox.ipod to I copy?
02:06:14JdGordonu need
02:06:20JdGordontry make install...
02:06:58jaebirdmake: *** No rule to make target `install'. Stop.
02:07:25JdGordongrab it off the daily builds page then...
02:07:28JdGordoni dunno
02:08:04jaebirdI guess the 4g grayscale can use the from the Color?
02:08:54JdGordondunno, maybe not.. coz the colour defines in the code..
02:09:37 Join DreamTactix291 [0] (
02:09:41jaebirdit built just fine, but there is no install.
02:11:01PaulJtry: make zip
02:11:26 Quit actionshrimp ("( :: NoNameScript 4.01 :: )")
02:12:13 Quit slimx (Remote closed the connection)
02:12:38jaebirdAhh yes, I was about to say that...looking at the makefile suggested make zip
02:13:11JdGordonLinusN: done.. works fine on the archos now
02:13:20JdGordonuses the old style
02:13:32PaulJthe command "make install" is afaik for the uisimulator
02:14:33jaebirdWhat does error -5 mean when loading rockbox
02:15:00jaebirdI also got a screen size warning on WPS during the make zip
02:15:48 Join midkay [0] (n=midkay@
02:18:37 Quit Kohlrabi ("Fast alle Menschen sind Regenw├╝rmer")
02:21:47 Part XavierGr
02:21:54 Part LinusN
02:22:49 Join XavierGr [0] (
02:22:55 Quit ModernExecutive ()
02:24:36 Join Lynx_ [0] (
02:26:46jaebirdIpod 4g shot here:
02:26:56jaebirdsorry for blurry :)
02:27:38JdGordonyouve got an odd battery icon?
02:27:46JdGordonand y is the verison shown on the bottom?
02:29:04 Quit goa (Read error: 104 (Connection reset by peer))
02:29:05 Join hd [0] (
02:29:24 Quit mymomthelush ("Trillian (")
02:29:42jaebirddon't know...don't know much about rockbox
02:30:16 Nick hd is now known as goa (
02:30:41jaebirdinteresting, I added some music files and now I only get a blank screen
02:30:56jaebirdbacklight still works, but nothing there.
02:31:39 Join ashridah [0] (
02:33:14*Cassandra crosses fingers.
02:33:21 Join Jungti1234 [0] (n=jungti12@
02:33:24*JdGordon cuts em off
02:33:33JdGordonah.. sorry.. shouldnt play with nives
02:33:41CassandraI came up with an evil plan to get my hands on toys for Rockbox development.
02:34:00CassandraNo, you might accidentally chop off your 'k's.
02:34:20CassandraIt may not work though. Wait and see time now.
02:35:12JdGordonis it a letter like "Dear Mr jobs.. your software sux.. please send a dozen 5g ipods to the good ppl at rockbox so we can put some good softwar eon it. Thanx.." ?
02:35:31 Join YouCeyE [0] (
02:35:35 Quit YouCeyE (Client Quit)
02:36:04jaebirdnow I can't get anything on the screen..hmm
02:37:13CassandraSomething like that.
02:37:52CassandraMore like "Dear X, you aren't offering enough money to employ a sysadmin. Would you like to employ a fifth of one."
02:38:04 Quit Lynx_awy (Read error: 110 (Connection timed out))
02:38:04 Nick Lynx_ is now known as Lynx_awy (
02:39:29Cassandra(Even though you really *ought* to employ a whole one.)
02:42:34 Quit midkay (Remote closed the connection)
02:44:56 Join midkay [0] (n=midkay@
02:45:09 Quit midkay (Read error: 104 (Connection reset by peer))
02:45:54 Quit muesli- (Read error: 110 (Connection timed out))
02:49:38 Join YouCeyE [0] (
02:50:44 Join midkay [0] (n=midkay@
02:50:47 Quit midkay (Client Quit)
02:56:30 Quit ashridah ("gone")
03:01:48 Join cool2bdave [0] (
03:02:05YouCeyEI would like to know how to make svcd files using command line from a set of vob files
03:02:13YouCeyEwrong window
03:04:21 Join DJDD___ [0] (
03:07:54***Saving seen data "./dancer.seen"
03:11:09cool2bdaveHello I have my first bit of working rockbox code. It reads ID3 Lyrics USLT format (media player, musicmatch, realplayer) It exports the text to a text file and displays it with the text viewer. I added a Show Lyrics to the onplay menu
03:11:44cool2bdaveI havnt used cvs before so I dont know how to upload the patch
03:12:00cool2bdaveOr do the diff operation
03:28:14jaebirdHas anyone else tried the 4g 2bpp stuff in cvs?
03:28:38 Quit DreamTactix291 (Read error: 110 (Connection timed out))
03:34:56cool2bdaveWill do it 2morrow ttfn
03:34:59 Quit cool2bdave ("CGI:IRC")
03:44:29 Join quobl [0] (n=quobl@tor/session/x-0205fe7da21d8ecf)
03:50:01 Part PaulJ
04:00:16 Join kkurbjun [0] (
04:01:26 Join webguest88 [0] (
04:02:19 Part webguest88
04:05:04kkurbjundoes anyone know much about the midi plugin?
04:06:30BHSPitLappysupposedly it works, but takes forever
04:07:23kkurbjunhas any profiling been done on it?
04:07:43*BHSPitLappy shrugs
04:19:33Jungti1234hi all
04:19:51 Quit JdGordon ("Miranda IM! Smaller, Faster, Easier.")
04:24:42 Quit Jungti1234 ()
04:29:14 Quit webguest40 ("CGI:IRC (EOF)")
04:35:55XavierGrkkurbjun: You want to know about the midi plugin?
04:36:34XavierGrWell I don't know much. It will not work currently.
04:36:52XavierGrIt needs the old xxx2wav files
04:36:57XavierGrthat are stored in the atic
04:37:19XavierGr^attic. If you want I can compile to you a working version.
04:37:49XavierGrTHe plugin (when you complete the files) will load a midi file and output the sound in a wave file.
04:38:58XavierGrIt is 78% realtime at 22khz and very slow on 44khz
04:39:50XavierGrThe guy that programed it does no longer work for it, due to lack of time and that he doesn't have a target to test it.
04:40:45 Quit Paprica (Read error: 110 (Connection timed out))
04:40:55XavierGranyway I hope that if you are interested to make it better, I will be glad to help as I can.
04:41:06XavierGrOk time to hit the sack. Later all
04:54:26 Join JdGordon [0] (
04:54:39 Join gtkspert [0] (
04:54:43 Quit JdGordon (Client Quit)
04:54:49 Join JdGordon [0] (
04:54:59 Join elinenbe [0] (
04:55:17 Join Paprica [0] (
05:01:32 Join Rob2222_ [0] (
05:04:05 Quit mikearthur ("Konversation terminated!")
05:07:40 Join DreamTactix291 [0] (
05:07:57***Saving seen data "./dancer.seen"
05:12:24 Join hardeep [0] (i=hardeeps@SDF.LONESTAR.ORG)
05:12:47 Join perl|bbq [0] (
05:19:50 Quit Rob2222 (Read error: 113 (No route to host))
05:25:35 Quit perldiver (Read error: 110 (Connection timed out))
05:29:15 Quit DreamTactix291 (Read error: 110 (Connection timed out))
05:29:48 Quit ehntoo (Remote closed the connection)
05:33:45 Quit perl|bbq (Read error: 110 (Connection timed out))
05:48:15 Join perldiver [0] (
06:02:26 Quit YouCeyE ("Leaving")
06:04:00 Join YouCeyE [0] (
06:07:50 Quit imphasing (Remote closed the connection)
06:08:47 Join optimistik [0] (
06:09:51optimistikhey this has nothing to do with rockbox but does anyone kno if its tru that pod dealers can't do service on ipods if the customer still has the one year free from apple
06:10:52*BHSPitMonkey hits the offtopic n00b with a trusty cluebat
06:11:20JdGordonwouldnt surprise me
06:13:40BHSPitLappyJdGordon: is the patch difficult to apply
06:13:46BHSPitLappyI've never used one
06:14:27BHSPitLappy(color picker)
06:15:05 Quit optimistik ()
06:15:37JdGordonits not a finshed patch
06:15:40JdGordonits only the gui
06:16:01JdGordonthere is a colour settings patch that i was gonna "fiish" with my gui, but that patch didnt work
06:16:10JdGordonso we wait for some1 to fix it
06:19:08 Join imphasing [0] (
06:19:31BHSPitLappyall I want is the color value
06:19:40BHSPitLappywill it show a color and its value?
06:20:48JdGordonye, in that case.. all u need to do is put the .c and .h in the apps/gui folder.. add the to apps/SOURCES and then add a menu entry somewhere to load it
06:20:53JdGordonor make it a plugin..
06:23:13BHSPitLappywould the plugin be quicker? :P
06:24:11JdGordonmaybe... ye shold be
06:24:29JdGordonyou just need to change the displa-> to rb-> and add rb-> infront of heaps of functions
06:24:59BHSPitLappyoh, and what's the patch called
06:25:06BHSPitLappycolor settings [2]?
06:25:10 Join perl|bbq [0] (
06:25:17JdGordonno.. 1 sec
06:32:02 Join Paul_The_Nerd [0] (
06:38:38BHSPitLappyJdGordon: sorry to be stupid, but what header file were you talking about?
06:38:40BHSPitLappythe .h
06:38:58JdGordonye, is it there? i cant rememebr if i uploaded it...
06:39:13JdGordonu dont really need it.. all it has is the funcion definition so u can call it from another file
06:39:19BHSPitLappythe only attachment on that link you gave me was the .c
06:39:55JdGordonye, ok... so just grab the function definition and add a ; onto the end somewhere above where u want to call the function
06:41:03 Join DarthLappy [0] (
06:41:38 Quit perldiver (Read error: 110 (Connection timed out))
06:44:46BHSPitLappyok, so the draw screen or the get color?
06:47:00 Join ashridah [0] (
06:49:29 Nick colin_sleep is now known as colin_ (
06:51:57 Quit RotAtoR ("zzzzzzzz")
06:55:38JdGordonBHSPitLappy: got it working?
06:56:08BHSPitLappydistracted by movie lol
06:59:59BHSPitLappyok, just need to figure out the menu entry
07:04:12*BHSPitLappy uses his wonder twin powers to call forth the powers of JdGordon
07:08:00***Saving seen data "./dancer.seen"
07:08:21 Join DreamTactix291 [0] (
07:14:18 Join Jungti1234 [0] (n=jungti12@
07:21:56JdGordonim back
07:22:13JdGordonBHSPitLappy: got it yet?
07:22:26BHSPitLappydamn cygwin
07:22:59JdGordondoes anyone here use the radio ever?
07:26:40JdGordondo u find the radio screen sluggish?
07:27:01JdGordonit just feels slow for me...
07:27:35Paul_The_NerdI'm on an H120, so it's possible that ours perform differently
07:29:16 Quit DreamTactix291 (Read error: 110 (Connection timed out))
07:39:21BHSPitLappyJdGordon: for some reason I'm not able to build :(
07:39:30JdGordonhmm whats the error?
07:39:41BHSPitLappyit stops after the first few "up to date" messages
07:40:16BHSPitLappyI don't know what's different, I've built enough before
07:43:20 Part midgey34
07:46:17 Quit Thus0 (Read error: 104 (Connection reset by peer))
07:47:43 Quit Paul_The_Nerd ("Leaving.")
07:54:31JdGordonweee my build is fucked :D
07:54:55JdGordonfor 'VOICE_EXT_RWPS'
07:54:55JdGordonn of 'VOICE_EXT_RWPS' was here
07:55:04JdGordon./home/Administrator/rockbox-devel/h300-sim/lang.h:615: error: conflicting types
07:55:05JdGordonfor 'VOICE_EXT_RWPS'
07:55:05JdGordon./home/Administrator/rockbox-devel/h300-sim/lang.h:615: error: previous definitio
07:55:05DBUGEnqueued KICK JdGordon
07:55:05JdGordonn of 'VOICE_EXT_RWPS' was here
07:55:26JdGordon?? that repeated for probably all the lines in lang.h
07:56:01BHSPitLappydid you do something or is it cvs?
07:56:28JdGordonive been playing with the radio code... other than that its cvs...
07:56:58BHSPitLappywell we had 2 red builds all across the board it looks like, but that was like 10 hrs ago?
07:57:42JdGordonjust did an update... c what happens
07:58:36BHSPitLappyI think I was already up-to-date when I tried these
07:59:01BHSPitLappywhere are preglow / linuxstb_ today :P
07:59:25BHSPitLappythey're not even in my scrollback
07:59:27JdGordonisnt it like 2am for them now?
07:59:37BHSPitLappyit's 1am for me!
07:59:54Bg3rJdGordon it's like 8-9a.m. for tnem :)
08:00:02BHSPitLappywake up you guys!
08:00:06JdGordonah, so they be around soon enough
08:00:59JdGordonsorry... my bad... ????
08:01:03Bg3ruf ... was the int 2 bytes or 4 bytes ?
08:01:08BHSPitLappyhit a nerve
08:01:18JdGordon4 isnt it?
08:02:51Bg3ranyway, 16 bits are enough...
08:03:17Bg3rfor anyone :D
08:04:40 Join midkay [0] (n=midkay@
08:08:45BHSPitLappyapparently power.c fucked the ipod builds
08:09:00midkayBHSPitLappy: HAHA!!...
08:09:30BHSPitLappyROFL IRON! err, wait
08:09:43midkay.. no idea.
08:11:00 Join perldiver [0] (
08:11:38 Join DangerousDan [0] (
08:12:18 Quit imphasing (Remote closed the connection)
08:13:47 Join imphasing [0] (
08:14:19 Join BHSPitLappy2 [0] (i=Steve-O@
08:14:51BHSPitLappy2did my "ugh" get through
08:15:05 Quit imphasing (Remote closed the connection)
08:15:07midkayah, it did not.
08:15:11BHSPitLappy2damn internet's choking and dying
08:15:22 Quit BHSPitLappy (Nick collision from services.)
08:15:28 Nick BHSPitLappy2 is now known as BHSPitLappy (i=Steve-O@
08:15:37midkayis it, now? i think that's a lie. i think you're STRANGLING it.
08:16:03 Join imphasing [0] (
08:17:18 Quit DangerousDan (Read error: 104 (Connection reset by peer))
08:17:57 Quit Jungti1234 ()
08:20:26 Join ender` [0] (i=ychat@
08:29:22 Quit perl|bbq (Read error: 110 (Connection timed out))
08:30:23JdGordonyay, my new radio code is working :D
08:35:33 Join BHSPitLappy2 [0] (
08:36:39 Quit BHSPitMonkey (Read error: 110 (Connection timed out))
08:37:00 Quit perldiver (Read error: 110 (Connection timed out))
08:38:36 Quit BHSPitLappy (Read error: 110 (Connection timed out))
08:38:58BHSPitLappy2haxoring an unsecured wifi network
08:39:20imphasinguh huh
08:39:24imphasingthe ramada inn?
08:41:56Bg3rbtw, is there any ondio with 8MB mod ? :)
08:43:54 Join needleboy [0] (
08:46:34Bg3rJdGordon :) what's the new part in it ?
08:47:13JdGordonnothing new just yet... cleaning it up, and gonna make it wpsable...
08:47:43Bg3rsounds good idea :)
08:48:06 Join BHSPitLappy [0] (
08:48:47 Join BHSPitMonkey [0] (
08:49:55 Join Paul_The_Nerd [0] (
08:49:59amiconnBg3r: I don't know of one. Wouldn't make sense runtime-wise, only for things like viewing larger JPEGs.
08:50:48 Join aznstyle [0] (i=aznstyle@T66-154.DATANET.NYU.EDU)
08:51:04aznstylei have an idiot question
08:51:15aznstylewhy do i keep getting ROLO?
08:51:24aznstyleon my ipod nano
08:51:42JdGordonr u runinig the rockbox.ipod file?
08:51:55aznstylei think so
08:51:58aznstyleits booting with that
08:52:00Paul_The_NerdYou don't want to.
08:52:27aznstyleso it wont read files that are in the itunes database
08:52:40Paul_The_NerdIt boots *into* rockbox. That file is to sorta reboot, but that feature isn't working right yet.
08:52:45 Quit BHSPitLappy (Nick collision from services.)
08:52:55aznstyleokay so i have to drag and drop into that directory
08:53:01aznstylebasically delete my itunes database files
08:53:04 Join BHSPitLappy [0] (
08:53:05Paul_The_NerdActually, that's the root directory.
08:53:19JdGordonwhy the heck doesnt the backlight turn on automatticlyy on a button press?
08:53:30JdGordonin my button loop...
08:53:33Paul_The_NerdIt automatically hides folders beginning with a period, and hidden folders. Which is why you only see the rockbox.ipod file
08:54:21aznstylewhy wont it go into ipod_control?
08:54:44Paul_The_NerdBecause ipod_control is a hidden folder
08:54:46JdGordondoes cpu_idle_mode(true); kill the backlight auto-turning-on code thingy?
08:54:53Paul_The_NerdHit the Menu button
08:54:59Paul_The_NerdChoose General Settings
08:55:16Paul_The_NerdThen File view, then Show Files, and set that to all. Then hit select.
08:55:16BHSPitLappythe internet sucks
08:55:27BHSPitLappydoes it build now?
08:55:58aznstyleso if i unhide it it should work rite
08:56:04Paul_The_NerdWell, define "right"
08:56:13Paul_The_NerdYou can browse to the files stored on it, and play them
08:56:15aznstyleplay music
08:56:27Paul_The_NerdBut it won't parse the database, so you'll see the strange filenames it renames them to.
08:56:52aznstyleand booting back into original firmware
08:56:57aznstylei hold menu but it doesnt work
08:57:08Paul_The_NerdHold play/pause to turn the unit off
08:57:19Paul_The_NerdOnce it's completely off, press menu, release, and immediately press and hold it, before anything even happens
08:57:35Paul_The_NerdHold it until you see that it's loading the original firmware. It may take a few tries to get the timing right.
08:57:36 Join LinusN [0] (
08:58:26aznstyledamn its pretty loud
08:58:31aznstylehm it work tho eh
08:58:35aznstylenot as pretty but w/e
08:59:20 Part LinusN
08:59:35BHSPitLappywe talking ipods?
08:59:42aznstylethe gui
08:59:44 Join LinusN [0] (
08:59:49 Join perldiver [0] (
08:59:56BHSPitLappythere's a tiny window where the lcd is completely off that you need to hit your key
09:01:01Paul_The_NerdThe whole "pretty" debate is vague anyway. Since the while playing screen is *very* customizable, you can easily reproduce (or improve upon) the basic ipod. The menu structure is basic text based, but you can load a background image. I like that option a lot more than that crazy sideways slide thing.
09:03:11Bg3ramiconn (8MB ondios) asked because there is an option in the configure ...
09:03:31aznstylewhat about the color
09:03:34aznstyleon the nano
09:04:10amiconnBg3r: Yes. This option is there for all SH-based archoses, because the architecture is the same.
09:04:39aznstylewhy is there spindown time for the nano...
09:04:54amiconnIt would also be possible to 8MB-mod a player etc, but I only know of 8MB-modded recorder v1's and FM recorders
09:05:02 Join einhirn [0] (
09:05:05Bg3raha :) k
09:05:47Bg3rsuch mod on a flash based palyer just seemed strange for me
09:05:51Bg3rto me
09:06:47amiconnIt's not *that* strange, as it would allow things to run that need more RAM at once, like
09:07:04Bg3rjpeg viewer :)
09:07:10amiconn(mentioned) viewing large JPEGs, running doom...
09:07:19Bg3rah, yes, doom
09:07:25Bg3rthis is a good point!
09:07:38amiconnOh, and one thing possible right now - a high quality voice file
09:08:03***Saving seen data "./dancer.seen"
09:08:18Paul_The_Nerdaznstyle: the "spindown time" on the nano affects nothing. It just hasn't been set not to show yet, I believe, as the ATA controller for the flash can be told to sleep, but right now I don't believe it does (unless something changes)
09:08:38Paul_The_Nerdaznstyle: And what was your question about the color on the nano specifically?
09:08:52aznstyleno way to enable it yet right
09:09:02aznstylecolor skins
09:09:05 Join DreamTactix291 [0] (
09:09:14Paul_The_NerdColor's enable
09:09:34aznstyleis there a setting i dont see?
09:09:48BHSPitLappyuse stuff that's color?
09:10:01BHSPitLappytry the Boxes 176x132c wps
09:10:11 Join Sinbios [0] (
09:10:17Paul_The_NerdYeah, browse themes to Boxes 176x132c
09:10:32aznstyleand also theres like a lag between songs
09:10:47 Quit BHSPitLappy2 (Success)
09:10:56Paul_The_NerdThat's a very vague statement
09:10:56BHSPitLappyif it's an mp3, it has to be set gapless when encoded
09:12:05 Quit BHSPitMonkey (Remote closed the connection)
09:13:40 Quit aznstyle ()
09:16:30 Join B4gder [0] (
09:22:02 Join safetydan [0] (
09:26:55 Quit needleboy (Read error: 110 (Connection timed out))
09:30:21JdGordondoes any1 know if cpu_idle_mode(true); is stopping the backlight turning on when a button is pressed?
09:30:42LinusNJdGordon: it's certainly not the intention
09:31:33JdGordonhmm... i have a button loop and that is the only thing different to all the other button loops ive ever done, and the backlight doesnt turn on anymore..., untill i set it to fals
09:32:09amiconnJdGordon: H300?
09:33:29amiconnAnyone tried the FM context menu on H300?
09:33:42LinusNcan't say i have
09:33:43amiconnIt's sloo-oow
09:33:44JdGordonnot i
09:34:05LinusNah, yes, we need to boost the cpu on interactivity in the fm screen
09:34:38amiconnThe mono bitmap drawing really needs optimisation
09:35:01 Quit hardeep ("[BX] Been around the world and found that only stupid people are breeding.")
09:35:58amiconnLinusN: For memset16() (and possibly memset32() for X5, if you're going for 32 bit gfx) Bagder suggested having a memory.h
09:36:36amiconnWould that go into firmware/include, and the implementations into firmware/common, or do you think these aren't 'standard' enough
09:36:45B4gderjust adding a memory.h header for the proto
09:37:33LinusNthe standard implementation would go to firmware/common and the optimized versions to firmware/<cpu>/memory.c or something
09:38:36LinusNdo i have a "go" for starting with the target/<cpu>/<brand>/<model> tree?
09:38:53B4gderyou have my go at least
09:39:34 Quit Sinbios ("If the definition of a klutz is someone who doesn't have eyes on their ass, then yes, I suppose I am a klutz.")
09:39:42 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
09:39:55 Join einhirn [0] (
09:41:22 Join needleboy [0] (
09:42:20LinusNi take the silence as a "yes" :-)
09:42:57JdGordonhmm.. ok im fiddling with the radio code... ive got it to stay on while the usb is connected, but is there any way to allow the controls to work? i mean not goto the usb mode screen?
09:42:59B4gderI would actually say that this is the only sane way forward
09:43:09preglowand i at least think i agree
09:43:13preglowbrain doesn't work today
09:43:38B4gderI've checked out and worked with multi-target projects such as u-boot and Linux
09:43:48B4gderand they both use a similar approach
09:44:15B4gderu-boot supports 140 boards, using 29 different CPUs...
09:44:31 Join Sinbios [0] (
09:44:46B4gderwe aren't quite there yet
09:45:11amiconnI'm not sure whether it will actually improve things, but I think this should be tried
09:45:29 Quit Paprica (Read error: 110 (Connection timed out))
09:46:01amiconnI suppose that the SOURCES mechanism will be changed a bit for this to work?
09:46:17LinusNnot in the beginning
09:46:21amiconnLike, adding the appropriate directories to the list?
09:46:44amiconn(instead of individual files)
09:47:07LinusNfirmware/SOURCES could still be used like today, with #ifdefs for the cpu's and models
09:47:32B4gderI have another idea for that
09:47:38amiconnfirmware/SOURCES is one of the messy #ifdef files
09:47:49B4gderwe move the SOURCES into the firmware/<cpu>/ for the cpu-specific ones
09:48:04B4gderand in firmware we build the cpu-specific ones with make -C <cpu>
09:48:34LinusNi will begin with using firmware/SOURCES as usual, and then we can extend that if we wish
09:48:43B4gderI agree with that approach
09:48:56amiconnI would think that we won't need a SOURCES in the cpu-specific and target-specific dirs
09:49:18LinusNlet's just solve the problems as we go
09:49:25B4gderamiconn: the point of using that there would only be to-reuse the same kind of make logic/scripts
09:49:30JdGordonthats a novel idea///
09:50:43B4gdersince SOURCES is used all over, it would possibly feel odd to not use it in some dirs
09:51:26amiconnLinusN: Hmm. Imho, the SOURCES #ifdefing is one of the messy places. The ifdefing within the actual sources feels less messy
09:51:46JdGordoni tinhk the oposite...
09:51:53CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
09:51:53*JdGordon hates all the #ifdefing
09:52:07B4gderthe main reason for the SOURCES-#ifdef is the slooooow cygwin
09:52:18amiconnNot only
09:52:22LinusNamiconn: SOURCES is messy, but that's not where i spend most my devbeloping time
09:52:52LinusNand i believe even SOURCES will be cleaner with this approach
09:52:56amiconnWe need a mechanism to *not* build common/ files in favour of optimised versions
09:53:24LinusNyes, in the #else part of the target #ifdef chain
09:53:53amiconnOtherwise the iriver binary may e.g. end up containing the C version of memcpy even though there is an asm version
09:53:57LinusN#ifdef CPU_COLDFIRE
09:54:17LinusN#elif defined(ARM)
09:55:07amiconnUnfortunately there is not implicit way to tell the build system to build the generic version only if there is no optimised version
09:55:26LinusNno, but does it have to be one?
09:55:41LinusNi like explicit ways, feels more KISS in some respects
09:55:45amiconnIt would avoid the SOURCES ifdefing as in your example
09:56:21Paul_The_NerdI think the explicit way might be better anyway.
09:56:22amiconnOh, and btw, no one wrote optimised mem*() functions for arm, but for SH1 :)
09:56:51Paul_The_NerdIn terms of, when someone first looks at the sources, they'll know very clearly what's going on once they read that chain.
09:57:40amiconnOne place where it will be hard to separate by cpu and target are the plugins
09:57:57B4gder"ASM optimised string functions"
09:58:22amiconn...unless we switch to one dir per plugin, with the same ./cpu/target sub-structure
09:58:39amiconnThat would make the number of files and dirs explode...
09:58:42B4gderthe plugins are and should be mainly app-level anyway
09:59:05amiconnYes, but that doesn't stop them to contain target or cpu dependent stuff
09:59:18B4gderwe won't get rid of all #ifdefs
09:59:34B4gderthat dir separation is mainly for separating drivers and lowlevel stuff
09:59:34amiconnSeveral gfx plugins need to differentiate whether to use the grayscale lib or native gfx
09:59:53amiconnThen mandelbrot contains SH1 and coldfire optimised math functions
10:00:28 Quit DreamTactix291 (Read error: 110 (Connection timed out))
10:00:43 Quit Strath (Read error: 104 (Connection reset by peer))
10:00:46amiconnB4gder: *The* example for gfx #ifdefing is cube.c
10:00:47B4gderpreglow: linux/arch/arm/lib/ contains a whole set of asm-optmized functions for ARM, you might enjoy checking them out sometime
10:01:00B4gderamiconn: yes?
10:01:14B4gderbut what are you saying
10:01:36B4gderthat ifdefs will remain? oh yes they will
10:02:31amiconnOn the player it uses playergfx, hidden line and wireframe. On recorder it uses native gfx for hidden line and wireframe, and grayscale lib for solid mode. On H1x0 it uses native gfx for everything, as it does on H3x0 and iPod, but in colour
10:03:02amiconnI tried to make it more readable with macros, but it's still messy
10:03:03 Join Strath [0] (
10:03:08LinusNi'm not trying to solve *all* #ifdef problems, only the most obvious ones, where each platform has it's own way of doing things, like ata_enable() for instance, where each model has the enabble signal on a different pin
10:03:25JdGordonhow do i get make to dump errors to a file? make > err.txt isnt working..
10:03:36B4gderJdGordon: 2>err.txt
10:03:51B4gderthat is "make 2>err.txt"
10:04:00 Part LinusN
10:04:27B4gderor "make >err 2>&1" to get both stdout and stderr to the same file
10:04:28 Join LinusN [0] (
10:05:33amiconnLinusN: I didn't look at ata.c, but I think that e.g. the solution in backlight.c isn't all that messy
10:06:05amiconn(concentrating the manifold of targets in a few small functions)
10:06:18LinusNbacklight was my next example
10:06:23LinusNi think it's a mess
10:06:59 Join petur [0] (
10:07:54LinusNbacklight.c would be a lot cleaner if the actual port pin fiddling was done elsewhere
10:08:03preglowB4gder: ooooh
10:08:26LinusNthen the actual logic would be a lot more obvious
10:08:41LinusNand porting to a new target would be simpler
10:09:09 Join slimx [0] (n=slimx@
10:09:32 Quit Sinbios (Read error: 104 (Connection reset by peer))
10:09:57JdGordonbuttin in again.... is there a proper way to call the main menu from the radio screen? or do i just do main_menu();?
10:09:59preglowsafetydan: but yeah, looks like a graphing eq wont be too hard, so i hope you're prepared for more gui work :PP
10:10:04amiconnLinusN: Hmm, to me backlight.c is one of the less messy source files.
10:10:21 Join BHSPitLappy2 [0] (
10:10:25LinusNi give up
10:10:29 Quit BHSPitLappy (Nick collision from services.)
10:10:33amiconnAnyway, if you're going to do the split, some more ideas
10:10:33 Nick BHSPitLappy2 is now known as BHSPitLappy (
10:12:15amiconn(1) I would consider the sim an additional architecture. (2) For the functions that could/should be inlined, it could be useful to have .h files in the cpu/target specific dirs. Then these dirs should be made part of the include search path, depending on cpu/arch
10:12:30 Join josh__ [0] (
10:12:30 Quit josh_ (Read error: 104 (Connection reset by peer))
10:13:04LinusN(1) good idea
10:13:13LinusN(2) good idea
10:13:26needleboyLinus, goo dmorning
10:13:36needleboygood even
10:13:42needleboyyou or anyone else...
10:13:45amiconn-> if you then #include "backlight_control.h" it would use the file in SH1/recorder for recorders, coldfire/H1x0 for H1x0 etc
10:13:51LinusNi'll wait with (1) for now
10:13:57needleboyi compiled my own bootloader... what do i do with the BIN file?
10:14:02LinusNamiconn: yes
10:14:07needleboyhow do i patch the iriver firmware?
10:14:15BHSPitLappyneedleboy: use it with ipod_fw to generate a firmware
10:14:22BHSPitLappyneedleboy: whoops
10:14:24needleboyi got iriver h300
10:14:26amiconn(1) is e.g. for sim_backlight_on() etc. Would save us from a lot of stubs...
10:14:29BHSPitLappytoo accustomed to that answer, lol
10:14:31LinusNneedleboy: which version did you build?
10:14:44needleboylatest CVS with the Remote Control patch
10:14:48needleboyfor H300
10:14:57preglowjust be aware the compiling your own bootloader is very unsupported
10:15:04preglowmight easily brick
10:15:05needleboyI want to check the 'HOLD is ON' issue when turning on the unit with the remote
10:15:11needleboyi know...
10:15:26LinusNneedleboy: you shouldn't use latest cvs when buildind the bootloader
10:15:39LinusNit's untested
10:15:48needleboyok... what then?
10:16:10needleboyhold on... instead
10:16:20LinusNcvs co -rbootloader_h300_v5
10:16:24needleboywould any of you mind looking at the new remote patch and see if it's good for CVS?
10:16:42needleboymaybe fix the whole bootloader remote control issue?
10:16:52needleboythe guy who made the patch said it works
10:17:01Bg3rneedleboy what's its date ?
10:17:06LinusNthat won't be fixed unless we auto-detect the remote in the bootloader
10:17:28needleboyeven if it's selcted in the firmware?
10:17:35needleboyyeah... i get why not...
10:17:35LinusNof course not
10:17:58LinusNthe bootloader doesn't use the settings in the firmware, for obvious reasons
10:18:58amiconnI only have the H1x0 remote, otherwise I would try some things
10:19:14LinusNi have all remotes, but no time
10:19:37linuxstb_What's the problem with the h300 bootloader and remote controls?
10:19:43 Nick linuxstb_ is now known as linuxstb (
10:19:50LinusNthe non-lcd remote is wired differently
10:20:00LinusNso the bootloader thinks Hold is on
10:20:11amiconnThe H300 LCD remote too afaik?
10:20:25preglowB4gder: and here i was, worrying how far i'd have to push amiconn before he did mem* functions for arm...
10:20:29needleboynot any of the LCD remotes
10:20:31amiconnNo? But the buttons are different, right?
10:21:04amiconnpreglow: I don't have an arm-based target, and most probably won't get one
10:21:16B4gderlinux contains m68k versions too of optimized mem functions
10:21:26preglowB4gder: amiconn's are better
10:21:30B4gderI'm sure
10:21:36preglowB4gder: the ones in linux are part c part asm
10:21:42B4gderyes, I noticed
10:21:46preglowamiconn: but yeah, linux comes with pure asm versions of the mem* boys
10:21:54B4gderI guess not many people really care for linux on m68k
10:22:41preglowmemset doesn't use the stack at all, it seems
10:22:42amiconnI tried it once on Amiga, didn't manage to get it to work...
10:24:17preglowarm memset() seems to write 64 bytes at a time
10:24:25*amiconn doesn't know what linux/ dir Bagder and preglow are talking about :(
10:24:28safetydanpreglow: gui work suits me just fine
10:24:45safetydanyou people with your emac unit's and assembly based memmove's scare me :)
10:24:59B4gderamiconn: try this linux-2.6.15/arch/arm/lib/
10:25:31B4gderthat's the dir with generic cpu-specific functions
10:25:41B4gderlike memset, memcpy etc
10:26:31preglowB4gder: yes, quite a lot nice stuff in there
10:26:40preglowi think i'll just go ahead and help myself to a lot of it
10:27:18amiconnBagder: where??
10:27:37B4gderamiconn: I checked in my locally downloaded linux source
10:29:56preglowsome people really do know how to utilise the conditional support on arm...
10:30:27*amiconn isn't accustomed to the linux directory tree yet :/
10:31:06amiconnender`: r u there?
10:32:15 Join TCK- [0] (
10:32:51 Quit BHSPitLappy ("Dropped my laptop.")
10:32:57preglowi wonder what the hell is the meaning of the LOADREGS and RETINTRS macros.......
10:33:06Bg3rhuh m68knommu's memset is in C...
10:37:20 Quit quobl ("Leaving")
10:37:26preglowmemzero() is not standard, no?
10:38:13Bg3risn't it just memset(...0..)
10:38:47preglowwith the potential to be slightly faster, since it might require fewer regs, but no
10:38:54preglownot when you burst, so just forget it
10:41:04amiconnHmm. If the arm implementations are the same quality as the m68k ones, then I'd say forget it and write your own...
10:41:32preglowi think we'll just use these in the meantime
10:41:39amiconnThe SH stuff is unusable for SH1, as it contains SH2+ instructions
10:42:00preglowthe arm memset seems ok to me
10:43:56preglowbut yeah, i don't think i'll ever attempt to write them myself anyway
10:46:09amiconnHaha, the SH memchr.S is cheap...
10:46:22Bg3ramiconn : send a patch! :)
10:46:24amiconnNot using cmp/str on SH should be punished.
10:48:11safetydan15 dB gain should clip like mad... yet someone says it "sounds so good"
10:49:45preglowi saw someone using just the pk filters and the hardware eq filters with maxed settings, and claiming it sounded totally sweet
10:49:58amiconnThe m68k memcpy is only partially asm, and doesn't care about source alignment, only destination
10:50:20safetydanpreglow, I think we're talking about the same guy
10:50:24preglowsafetydan: me too
10:50:26amiconnMy experiments have show that source alignment is more important than destination alignment, at least on our coldfire
10:50:48amiconnOh, and it doesn't burst, as it doesn't use movem
10:53:56amiconn...but it gave me some ideas :)
10:54:19pregloweq graphing will require 20 long muls, 10 adds, one long div and two cos() operations per graph pixel per filter :>
10:54:58preglowbut gotta go, back in an hour
10:57:04 Join muesli__ [0] (n=muesli_t@
11:00:31 Join amiconn_ [0] (
11:00:46 Quit amiconn (Nick collision from services.)
11:00:46 Nick amiconn_ is now known as amiconn (
11:04:03 Quit TCK- (Read error: 110 (Connection timed out))
11:06:13 Quit darkless (Read error: 104 (Connection reset by peer))
11:08:04***Saving seen data "./dancer.seen"
11:17:57 Quit perldiver (Read error: 110 (Connection timed out))
11:18:26 Quit youngcereal (Read error: 110 (Connection timed out))
11:18:44 Join youngcereal [0] (
11:22:36 Join webguest54 [0] (
11:23:04linuxstblostlogic: Is your pcm buffer patch on sourceforge up to date?
11:25:41 Join darkless [0] (n=darkless@
11:26:31 Join Membrillo [0] (
11:26:55 Join Lost-ash [0] (
11:27:20 Quit ashridah (Nick collision from services.)
11:27:22 Nick Lost-ash is now known as ashridah (
11:32:05 Quit webguest54 ("CGI:IRC (EOF)")
11:40:35 Join DreamTactix291 [0] (
11:42:03 Join Moos [0] (
11:43:16Membrillohas anyone got Goldeneye to work on the rockdoom engine?
11:48:28 Part petur
11:53:38 Join NicoFR [0] (
11:54:49ender`amiconn: still need me?
11:55:07amiconnLinusN: I observed a nasty effect on H3x0: If you repeatedly press BUTTON_RIGHT, it will eventually register as BUTTON_OFF.
11:55:14amiconnAny idea what might cause this?
11:55:48amiconnender`: Wasn't it you who got rockbox to build on SFU?
11:56:33ender`you need gcc 4.0.2 from tools warehouse, then you can build m68k-gcc and the rockbox utils
11:56:47amiconnWhat about 'make'?
11:57:10amiconnWhen I tried it, the supplied 'make' didn't even support the -C option
11:57:45ender`it's also in the tools warehouse
11:58:18ender`2 very useful links for SFU:
11:58:29amiconnHmm. How do I get & install tools from the warehouse?
11:58:47amiconnThey are listed, but I don't see links?
11:58:58ender`register, then use the pkg_install script
12:00:38LinusNamiconn: wow, interesting
12:02:17amiconnender`: I will try SFU again then. Hopefully speed is similar to linux on the same box (unlike cygwin)
12:02:39amiconnAn SFU-Howto in the wiki woul dbe ideal...
12:02:58ender`gimp builds much faster than in msys, though with rockbox speed difference isn't that much
12:03:06amiconnLinusN: That dropped me out of mandelbrot several times when scrolling around...
12:03:30amiconnrather, when scrolling right
12:03:50*Cassandra wonders if anyone's tried to build Rockbox using mingw32.
12:03:51ender`i can do that - as soon as i find out why suddenly half of my interix stuff segfaults
12:04:01amiconnCassandra: On linux?
12:04:07*linuxstb was wondering the same thing as Cassandra...
12:04:08CassandraNo, Windows.
12:04:15CassandraYou'd need MSYS as well, I suspect.
12:04:43ender`Cassandra: try compiling m68k-gcc for windows
12:05:02 Join Matze [0] (
12:05:06Paul_The_Nerd Cassandra: I tried very briefly, but I remember running into some problem relating to Make, and deciding it wasn't worth the hassle at the time
12:05:14CassandraNot surprising.
12:05:30Cassandraender`, no thanks. I have a working UNIX build environment.
12:06:08ender`amiconn: don't forget to read the FAQ on the interop forum, it contains a lot of useful info
12:06:24CassandraI prefer the mingw32 approach to cygwin. Cygwin is to busy doing its own UNIXy thing. Windows just wasn't cut out to be UNIX.
12:06:37LinusNamiconn: maybe the A/D value thresholds are too narrow, try tweaking them in button_read()
12:10:38ender`Cassandra: seen SFU/interix?
12:10:47amiconnWhat readings do you get for Up/Left/Down/Right?
12:11:06amiconnI get 0x20/0x43/0x66/0x87
12:12:04 Quit XavierGr ("Trillian (")
12:12:32amiconnSometimes also 0x88...
12:13:19Cassandraender`, nope. I'd be amazed if they'd actually managed to address the fundamental paradigm differences much better than cygwin though.
12:13:28CassandraIf I want UNIX, I'll use UNIX.
12:13:37amiconnUnfortunately I can't get the Off reading, as that quits the screen..
12:13:50CassandraIt's better at being UNIX than Windows is. ;)
12:15:49amiconnI'll fix that. It's sufficient to leave the screen with BUTTON_OFF|BUTTON_REL so you can check the reading :)
12:17:21 Quit dwihno (Read error: 104 (Connection reset by peer))
12:17:45linuxstbCassandra: Do you know if anyone is planning to work on the contents of the new manual?
12:18:04 Join dwihno [0] (n=dw@
12:19:58amiconnLinusN: Off = 0xA9 here
12:22:04LinusNso we could use for example 0x98 instead of 0x90...?
12:22:21 Join tucoz [0] (
12:22:21 Join Lynx_ [0] (
12:22:35LinusNtry it
12:22:40amiconnI just committed that tiny change to the debug menu
12:22:53tucozlinuxstb: I am, not just yet. I believe Febs have expressed a will to help out as well.
12:23:03Bg3ramiconn which byte ?
12:23:11Bg3rAA here
12:23:23tucozlinuxstb: as he has written a good deal of the wiki docs
12:23:38amiconnI'd like to know your Off value, then I'll adjust the values to best-match the gaps with both sequences
12:23:51linuxstbtucoz: Has anyone looked at producing html output from the latex source?
12:23:55Bg3ramiconn my off is AA
12:24:03amiconnThe upper threshold for Off should be lowered as well, to lower the risk of catching transients
12:24:22tucozlinuxstb: not that I know of. Could use pdf2html as a start maybe
12:24:40 Join Paprica [0] (
12:25:26tucozlinuxstb: also, Henrico from the forums seems to know his LaTeX and I think he has started with the plugins
12:25:29linuxstbI'm not convinced that would be able to produce decent output - I'm sure there are latex2html tools around.
12:26:05linuxstb:) Just found that as well...
12:26:09 Join DarthL4ppy [0] (
12:26:21tucozprobably not, but I wonder how that will be done...
12:26:48linuxstbIt could be worth looking at latex2html now - so we make sure we don't do anything which makes things hard for it.
12:27:07tucozFor an online doc?
12:27:11 Join Kohlrabi [0] (
12:27:36linuxstbI don't think it's worth maintaining a wiki manual, as well as the cvs latex manual.
12:27:39tucozok, I'm installing it now
12:27:59tucozme neither, the wiki is too hard to navigate
12:28:10tucozthings tend to be forgotten in the wiki
12:28:43safetydanThe wiki manual does make it easier to update things when they happen though, and makes it easier for users to correct things
12:28:45 Quit DreamTactix291 (Read error: 110 (Connection timed out))
12:28:59tucozCassandra, do you know how to invoke latex2html instead of pdflatex?
12:31:07linuxstbsafetydan: Now that the manual is in CVS, developers (and patch submitters) should be encouraged to update the cvs latex manual when adding new features. A single patch can now change both the Rockbox source and the manual.
12:31:50linuxstbAnd I wouldn't object to those users active in writing documentation being given cvs access
12:32:19LinusNamiconn: 0xa7
12:32:42Bg3rso, for now 0xa7, 0xAA, 0xA9
12:33:00safetydanlinuxstb, good point
12:33:50LinusNlinuxstb: i like the sound of that
12:34:18 Join lonux [0] (
12:35:22lonuxany chance that video is on its way to the H1XX, or any new features? it just seems that nothing really happens to this port anymore, mature enough for an official release? cheers!
12:37:19*safetydan looks at adding manual entries for the equalizer UI
12:37:34preglowsafetydan: there already is something in the wiki manual
12:37:41preglowbut don't think it's very much
12:37:54JdGordonis the eq ui commited yet?
12:38:13safetydanJdGordon, not commited yet
12:38:20tucozlinuxstb: I ran latex2html on the rockbox.tex file. You can see the results here.
12:38:22safetydanpreglow, I was going to just copy that in to the manual
12:38:27 Quit Membrillo ()
12:38:28JdGordonso shouldnt it stay out of the manual untill it is?
12:38:32lonuxis the EQ coming for the h1xxs? that would be great :)
12:38:38safetydanit's been a while since i've written LaTeX though
12:38:51tucozsomethings wrong with the logo though
12:38:52preglowi'm going to commit the eq soonish now
12:39:01 Nick colin_ is now known as colin_sleep (
12:39:02linuxstbtucoz: Not bad for no work :)
12:39:15safetydanWhat's the preferred convention for referring to a feature that is for software codec platforms only?
12:39:30tucozlinuxstb: and I got some errors. But the end result is not too bad
12:39:44preglowi think 'software codec platform' is a good enough term
12:39:45tucozlinuxstb: exactly. LaTeX really is cool
12:39:47preglowcan't think of anything else
12:39:57 Join TCK [0] (
12:40:23 Quit Lynx_awy (Read error: 110 (Connection timed out))
12:40:23 Nick Lynx_ is now known as Lynx_awy (
12:40:42Paul_The_NerdIs there some way to tell it *not* to use the 3540 pixel wide PNG thougH?
12:41:03tucozComment out the include?
12:41:04preglowsafetydan: the current sourceforge patch is the newest one?
12:41:15linuxstbPaul_The_Nerd: I think we should have some build-system magic using something like ImageMagick to convert bitmaps to different sizes for different output formats.
12:41:53preglowdid the original artist never supply an .eps or something of it?
12:42:04tucozit's available as a .tif
12:42:09amiconnLinusN: Didn't help much. It's a bit better, but it still happens.
12:42:10linuxstbrasher produced a svg version - I've not seen the output of it though.
12:42:13preglow'cuz thats what the pdf manual should use, a bitmap will make for slooow pdfs
12:42:28linuxstbIt's somewhere on
12:42:39amiconnIn fact, when the glitch occurs, Right registers as Off->Right->Off sequence
12:42:40preglowyeah, but that is a traced version
12:42:42tucozpreglow, do you think we should go for eps's instead of png's?
12:42:52LinusNamiconn: :-(
12:42:56linuxstbpreglow: Yes, but it may be good. As I said, I've not seen the output.
12:42:58preglowtucoz: only if we the graphic is eps in original format
12:43:05preglowa traced bitmap isn't always good
12:43:08preglowlinuxstb: no, me niether
12:43:14tucozthere will be lot's of screenshots in the docs
12:43:20linuxstbI'm just saying we shouldn't reject it before seeing it.
12:43:22preglowtucoz: yeah, but they'll be fine
12:43:31LinusNamiconn: so we might not debounce correctly...
12:43:38 Quit DarthLappy (Connection timed out)"> specifically
12:43:55LinusNwe have a debouncing problem on the v1 recorders as well...
12:43:55preglowLinusN: doesn't the pcf do the debouncing?
12:44:07tucozLinusN, didn't you find a .tif of the logo the other day?
12:44:18LinusNpreglow: how could it? it's only an a/d converter
12:44:25preglowLinusN: that's right
12:44:27linuxstbtucoz: For the screenshots, I think we should just store them at the native resolution of the target LCD - and resize in the build system.
12:44:34preglowforgot about that
12:44:38LinusNtucoz: which logo?
12:44:44tucozrockbox logo
12:44:56tucozthe huge t-shirt competition logo
12:45:10LinusNcheck the irc log
12:45:20lonuxwhile i understand you have more important things to do, i think its pretty rude that you guys arnt even dignifying my question with a response
12:45:54Paul_The_Nerdlonux: The question may not have even been noticed. It happens sometimes.
12:46:00LinusNlonux: i think preglow answered you
12:46:24 Join webguest93 [0] (
12:46:35LinusNlonux: 12.38.52 # <preglow> i'm going to commit the eq soonish now
12:46:40tucozpreglow, here is the .tif-logo
12:46:47amiconnLinusN: Debounce should be in effect for all platforms. I wonder what's goiung on here, taking into account the spurious Off events on archos recorder...
12:47:05LinusNamiconn: badness indeed
12:47:25 Quit lonux ("CGI:IRC")
12:47:34LinusNtalk about rude
12:47:39preglowhe msged me saying thanks
12:47:40preglowbut yeah
12:47:50Cassandralonux, this is a pretty busy channel. I think it's pretty rude of you to assume that people didn't just miss your question. Especially when someone *did* answer it.
12:47:52linuxstband I was just about to answer his original question....
12:48:12Paul_The_NerdI think he was referring to the first question, if Rockbox on iRiver was release-ready.
12:48:40preglowwhich is a good question, really
12:48:53amiconnImho it's not
12:48:54Paul_The_NerdOut of curiosity, isn't there supposed to be a "Rock" to go with the "Box" in that tif?
12:49:00LinusNamiconn: i agree
12:49:04amiconnSo many incomplete things...
12:49:09preglowwhen should we consider it finished?
12:49:19preglowwe should put a number of points in the wiki
12:49:20LinusNPaul_The_Nerd: huh?
12:49:57tucozPaul_The_Nerd: yes, something went wrong
12:49:59CassandraNope. I think we're going to need a good 2-3 months of feature freeze and bug squashing before it's ready to go out the door.
12:49:59Paul_The_NerdLinusN: Well, when I look at the tif, I see a white background and the word "Box" without "Rock"
12:50:21LinusNPaul_The_Nerd: i assuem you use the silly "fax viewer" in windows, right?
12:50:25Paul_The_Nerdpreglow: There's a few things mentioned on the iRiverStatus page.
12:50:33amiconnSome things that come to mind: recording *from* radio screen, peakmeters in radio screen, complete multi-screen UI, properly working voice UI...
12:50:39CassandraWe're probably at the stage where we can do that, although it might be wise to give it another month or so to allow colour WPS support to mature.
12:50:47tucozPaul_The_Nerd: are you talking about the .tif, or the page I generated with latex2html?
12:50:51LinusNamiconn: s/pdif recording
12:51:04Paul_The_NerdLinusN: Well, just as a note, the .png that's in CVS looks the same way, when displayed in Firefox.
12:51:07Paul_The_Nerdtucoz: Both ,actually
12:51:16amiconnAnd recording settings: Frequency, stereo/mono...
12:51:39preglowi'd like to see the latency issue resolved
12:51:46Paul_The_NerdAre you gonna throw in that shine MP3 encoder, for when quality is of much less concern that space?
12:51:57amiconnTalking about channels: Playback channel settings, resampler that's actually usable for lower sample frequencies...
12:52:21tucozPaul_The_Nerd: I think it's firefox. Looks great in imageMagick
12:52:21amiconnOh, and fix that crossfeed for proper volume
12:52:28linuxstbPaul_The_Nerd: There is an encoder codec system in the patch tracker - but I don't think anyone's working on committing it.
12:52:37preglowhey, you're the one who swore linear interpolation would be good enough for lower sampling rates :P
12:52:44Paul_The_Nerdtucoz: Aye, even looks fine in windows paint.
12:53:02preglowbut no, the resampler has other issues as well
12:53:09Paul_The_Nerdtucoz: But windows "Image/Fax viewer" or whatever, doesn't like it. 's odd.
12:53:16preglowwhich i just can't figure out where comes from
12:53:24linuxstbIs anyone able to look at this SVG image?">
12:53:40safetydanfirefox 1.5 just shows the source
12:53:50Paul_The_Nerdlinuxstb: I only see the code behind it. Thought Firefox 1.5 was supposed to support SVG
12:54:00safetydancontent type is text/plain
12:54:01linuxstbIt's probably just an incorrect mime type on the server.
12:54:02safetydanthat's why
12:54:04tucozsame here
12:54:18linuxstbMaybe if you save it locally and then do file->open
12:54:50safetydanlooks fine in Inkscape
12:55:03linuxstbIs it good enough to use in the manual?
12:55:08amiconnOh, and for release, non-functional settings should either be removed or made to work
12:55:59CassandraRelease management is one area where we could really stand to have a project manager type.
12:56:07safetydanlinuxstb: looks very smooth and high-res
12:56:10safetydanI'd say it's suitable
12:56:12 Quit DarthL4ppy (Success)
12:56:18 Part tucoz
12:56:22linuxstbsafetydan: Can inkscape convert it to eps?
12:56:23CassandraAlthough I don't envy the catherder that is willing to take that on.
12:58:03 Join petur [0] (
12:58:04ashridahokay. i thought that said catheter when i first saw it :)
12:58:05preglowif anyone feels they can do it, go on!
12:58:11*ashridah has been watching too much scrubs
12:58:15safetydanlinuxstb, yes, I'll give it a shot
12:58:52preglowamiconn: i can fix those settings with a similar hack to what i did for playback speed...
12:59:00CassandraRealistically all it needs is someone willing to thoroughly test Rockbox and document everything that doesn't work.
12:59:02Paul_The_NerdI think the rockbox.svg needs a little work still, personally
12:59:07safetydanlooks like it converted fine
12:59:13CassandraShouldn't take more than ten minutes, right?
12:59:44amiconnpreglow: Which settins & hack do you refer to?
13:00:03amiconnThere are more than just audio things non-working
13:01:24CassandraThere was an attempt at building a wiki page of test scripts once, ISTR.
13:02:20preglowsafetydan: line 805 dsp.c, under an if, sure you don't lack an indent there?
13:02:54preglowamiconn: channels, stereo width
13:03:40preglowamiconn: just implementing the functions in dsp.c instead, and ifdef the original functions to not compile for iriver
13:04:20preglowamiconn: what other stuff than audio things don't work?
13:04:30preglowanything you can think of should in iriverstatus
13:04:49JdGordoncough cough... yesno patch cough cough...
13:05:37amiconnOn H1x0, only 'Car adapter mode'
13:05:44amiconnOn H300 there are more
13:05:45LinusNJdGordon: with that patch, how do you answer "yes"? by clicking left and then select?
13:06:10amiconnCar adapter mode won't work for H1x0 as it doesn't power on when plugging power
13:06:15JdGordonon all targets cept the archos 2 line 1 ye, on that one its the same as before
13:06:36amiconnFor H300 it's possible, but the bootloader has to be hooked into this entry point as well
13:06:38JdGordonthe default setting could be usersettable... or hardcoded to yes instead of no
13:06:51Paul_The_NerdI have a question: Since the ideal is simplicity (one button press for yes, rather than two) and all targets have left/right buttons, why not just make one of those "Yes?"
13:06:52safetydanpreglow, it looks fine to me
13:07:04amiconnLinusN: Would it be possible to make the bootloader pass a parameter to rockbox, indicating which way it started?
13:07:20LinusNamiconn: sure
13:08:07***Saving seen data "./dancer.seen"
13:08:31amiconnPaul_The_Nerd: The 'Right' button isn't always called Right, so there would still be the localisation problem
13:08:34safetydanpreglow, it's spaces in my copy, 8 in front of the if, 12 in front of the eq_process
13:08:34preglowsafetydan: it's lacking an indent here, the eq_process line comes straight under the if
13:08:39JdGordonPaul_The_Nerd: well.. it could... just _someone_ needs to decide for sure which option to use... i personally think left and right and ok to choose...
13:08:53preglowsafetydan: 8 in front of both here...
13:09:07amiconn...and furthermode, right isn't always the most logical, or convenient, button for 'yes'
13:09:19amiconnSomeone should really start working on l10n v2 :/
13:09:27safetydanpreglow, I know what it is, it's the -b flag I use when generating the patches
13:09:30JdGordonamiconn: left is yes
13:09:31Paul_The_NerdJdGordon: The problem with that is that it's still two button presses.
13:09:39JdGordonso? bg deal...
13:09:40safetydanpreglow, it means it's ignoring my extra indent for eq_process since it's just whitespace
13:09:46JdGordon.5sec will kill you?
13:09:54amiconnJdGordon: Now _that_ would be illogical
13:09:58preglowsafetydan: well, we want the extra whitespace
13:10:02LinusNthe thing about yesno is that it sometimes should be harder to select "yes" by mistake
13:10:09amiconn'Left' is usually 'back out', quite the opposite of 'yes'
13:10:34safetydanpreglow, yes, yes we do. I had trouble wrangling my editor in to leaving the existing whitespace alone so I used -b in the diff options
13:10:34amiconn...and on archos player, the left button is labeled '-'
13:10:36JdGordonamiconn: ?? look at every message box on linux/windows/mac yes is always on the left
13:10:58safetydanJdGordon, not in gnome
13:11:05JdGordonok.. so maybe not :p
13:11:34JdGordonLinusN: it can be made so the caller chooses the default? but that might not be the best option
13:11:36CassandraThose wacky Gnome guys and their need to be different.
13:11:44preglowsafetydan: have you tried compiling it for any other targets, btw?
13:11:51safetydaniPod sim
13:12:09Paul_The_Nerdpreglow: I've used v11 on iPod, and H120. Haven't tried v13.
13:12:18amiconnJdGordon: Oh, you're talking about this widget thing. Unfortunately I have to say I don't like it at all...
13:12:20preglowipod works fine
13:12:24preglowi was thinking more about archos
13:12:37Paul_The_NerdDon't have any of those.
13:13:10peturmaybe I sound stupid but why not 'play' for yes and 'stop' for no? would at least be better than 'navi' on H3xx which makes it easy to accidentally accept...
13:13:26safetydanpreglow, I haven't installed the SH compiler so I didn't try the archos platforms
13:13:50preglowok, no biggie
13:13:52preglowi'll do it
13:14:02safetydanarchos player sim built okay
13:14:25safetydanhrm... the SDL sim does not work well with the charcell LCD units
13:16:06linuxstbWhere else is the yes/no widget needed apart from the delete file screen?
13:17:25JdGordondunno.. but it would be nice to be able to acces it in plugins...
13:17:31JdGordonu never know when ud need it
13:17:40amiconnE.g. if 'recursively insert directories' is set to ask, and then you create a playlist from a directory
13:18:24amiconn..or when 'create bookmark on stop' is set to 'ask'
13:18:42Paul_The_NerdDoes the recent patch that "warns" when you're about to overwrite a playlist in ram use it?
13:18:55Paul_The_NerdI saw it in the changelog, but it's not a situation I encounter.
13:20:20amiconnAll yes/no dialog boxes on computers also have means to operate them with the keyboard, without the need to navigate to the appropriate button
13:21:17linuxstbamiconn: Do they? I thought you had to use TAB to navigate to the appropriate button?
13:22:13peturALT-Y / ALT-N (in windows)
13:22:20Paul_The_Nerdamiconn: You mentioned that the buttons aren't always called Left/Right. I'm curious what the others are called?
13:22:37linuxstbpetur: That's windows...
13:22:46amiconnEsc is almost always 'No' as well, or 'Cancel'
13:22:49peturyeah.. sorry ;)
13:23:16amiconnPaul_The_Nerd: In the code, we always have BUTTON_LEFT and BUTTON_RIGHT, for simplicity
13:23:33amiconnThe player Left and Right buttons are labeled '-' and '+'
13:23:50amiconn..and the iPod has them mapped to Rew and Ffwd afaik
13:24:05Paul_The_Nerdamiconn: Do BUTTON_LEFT / RIGHT function as Prev and Next on all targets though? (During playback, I mean)
13:25:05Paul_The_NerdIf they did, you could say "Next" rather than "Right" since you're basing it on the function rather than the label.
13:25:32 Join NiX8916Lightning [0] (
13:25:40 Part NiX8916Lightning
13:25:41amiconnIf it's only for labeling, I repeat my icon idea..
13:26:00JdGordonwhats ur icon idea?
13:26:06 Join San_Benedetto [0] (
13:26:13amiconnWe could have ICON_YES and ICON_NO, resembling what's actually printed on the button
13:26:13linuxstbI'm not sure I like the hard-coding of BUTTON_LEFT/BUTTON_RIGHT though - on the ipod we sometimes want to use the scrollwheel when other targets use left/right
13:26:17Paul_The_Nerdamiconn: How does the icon idea mesh with voice?
13:26:25San_Benedettocan someone tell me where can i get DOOM files for rockbox?
13:26:40amiconnPaul_The_Nerd: The yes/no requests aren't voiced (yet)
13:27:09Paul_The_Nerdamiconn: In that case, why aren't there alternate strings #ifdeffed for the various targets, out of curiosity?
13:27:12amiconnI don't know, but I'd think a blind user will be able to remember which button is yes and which is no
13:27:23Paul_The_NerdAs a stopgap until language v2?
13:27:31JdGordonamiconn: its nice... but a PITA to do...
13:27:53amiconnJdGordon: ?
13:28:02JdGordonsomeone has to draw the icons...
13:28:56amiconnNow that should be the least problem...
13:30:15linuxstbWouldn't you need to adjust the icons based on the font size?
13:30:32JdGordonfont and screen size...
13:30:48preglowSan_Benedetto: rockbox doom uses the same files as ordinary doom
13:30:55amiconnThe screen size is no problem, as the icons will be target specific anyway
13:31:39amiconnFont size is another problem, but then it is a problem for all icons
13:31:57Paul_The_NerdI don't know if it's such a problem, if it's *just* icons.
13:32:27amiconnPaul_The_Nerd: You can't #ifdef in the .lang files
13:32:50amiconnThey are processed by perl scripts
13:33:32linuxstbAnd is there a single voice file which works on all targets?
13:33:33Paul_The_NerdSomehow, "not being spoken by the voice" equated to "doesn't use the .lang files" in my head. I'm not sure why by this point.
13:34:29Paul_The_Nerdamiconn: In that case, my vote goes with icons as well. I don't think they need to be adjusted in terms of font size though. Most people will only need to glance at them very briefly anyway.
13:34:31amiconnlinuxstb: Currently all voice files are target independent
13:34:44amiconnThat's another thing that should be solved with l10n v2
13:35:16B4gderI do have some minor things moving in the langv2 corner
13:35:29amiconnEach platform only uses a subset of .lng and .voice
13:35:51amiconn...but both the .lng files and the .voice files contain all strings/clips
13:36:04linuxstbWhat are the other issues for l10n v2 that we couldn't simply solve by adding #ifdefs to the lang file, and producing per-target voice files? Or is adding #ifdefs not that simple?
13:36:05amiconn...which is a waste of space
13:36:05 Quit webguest93 ("CGI:IRC (EOF)")
13:36:38 Join Zagor [0] (
13:36:46B4gderuplang wouldn't work with ifdefs without a major hack
13:36:47amiconnOn swcodec, quite some more menu items could be voiced, but I wouldn't do that before l10n v2
13:37:18amiconnWe have strong space constraints, especially on archos, and it's getting worse with the latest code additions...
13:40:47B4gdermy langv2 script can actually generate a lang.[ch] set
13:40:59B4gderand I have a v1tov2 conversion script somewhat working
13:41:18B4gderthe next step is a v2 version of uplang
13:41:42B4gderand fix a few bugs
13:41:50amiconnHmm. I should probably finish my .lang cleanup before v2...
13:42:06B4gderthen we can take the plunge
13:42:16B4gderand fix translated plugins as a next step
13:42:36amiconnYes, chunked .lng files
13:43:21amiconn..and .voice too. I'm not sure though what memory the plugins should load voice chunks into
13:51:02San_Benedettocan someone tell me where can i get files for playing doom on rockbox?
13:51:32Paul_The_NerdSan_Benedetto: It uses the same WAD files as PC doom.
13:53:54San_Benedettowhere can i get
13:53:57San_BenedettoWAS files?
13:54:59Paul_The_NerdGoogle for "Doom shareware" and get the shareware version if you don't own a full copy.
13:55:20San_Benedettook thanks
13:55:32B4gderso there's no free wad files?
13:55:48Paul_The_Nerdb4gder: Well, there's "Freedoom"
13:55:50*B4gder admits to never having played doom
13:56:09*San_Benedetto too
13:56:17safetydanSan_Benedetto, in the release notes for the H300 optimised build there's a link to a copy of the shareware doom WAD files
13:56:19Paul_The_NerdThey released the program code under the GPL, but the art resources, etc, aren't "free" in any sense.
13:56:41amiconnB4gder: There's a free (as in beer) .wad file, containing the first episode of doom 1
13:56:44Paul_The_Nerd <−− This is a full replacement for all resources with true "free" ones though.
13:57:23B4gderI guess I'll continue living in doom ignorance anyway ;-)
13:58:51B4gderI honestly rather spend my time programming
13:59:28amiconnouch! Looking at misticrister is pure eyestrain
13:59:47B4gderamiconn: if you register you can select a less offensive "theme"
13:59:59muesli__amiconn agree
14:00:04amiconnI prefer not to register...
14:00:16muesli__in cognito surfer? ;)
14:00:19B4gderyou won't miss much
14:01:53JdGordonamiconn: u said u use the radio right?
14:02:22amiconnIn fact today I used radio on H300 for the first time
14:02:43 Join t0mas [0] (n=Tomas@unaffiliated/t0mas)
14:02:50amiconn...but I used it more often on Ondio FM before
14:03:00JdGordonim thinking it makes more sense to only let u tune the radio if ur in scan mode.. not preset mode... use the <> arrows to goto the next preset...
14:03:20JdGordoninstead of having 2 extra buttons.. agree?
14:03:59amiconnThis is how it works right now...
14:04:13JdGordonnot according to the source...
14:04:50JdGordonmeh, dunno... deleted..
14:07:02 Join mikearthur [0] (
14:07:08 Join linuxstb_ [0] (
14:07:20 Quit linuxstb (Nick collision from services.)
14:07:22 Nick linuxstb_ is now known as linuxstb (
14:08:24 Join DreamTactix291 [0] (
14:09:10Paul_The_Nerdpreglow: Is Q supposed to represent a wider range as it goes up, or down? People seem to be saying on the forums that Q is working opposite how it's described in some things they're linking. (As a note, I don't know enough about this to be sure I'm paraphrasing them properly, except the "opposite of what it describes" bit)
14:09:48amiconnHigher Q means a narrower filter
14:10:04preglowthis should come as no surprise to anyone ever having used an eq
14:10:05Paul_The_NerdSee, as far as they're saying, Higher Q is resulting in a wider one.
14:10:11preglowq is a really common unit
14:10:16preglowit does not
14:10:21preglowq behaves exactly as it should
14:10:58preglowi've listened to it myself, and i know what it should sound like
14:12:46Paul_The_NerdAs I said, I don't know. I just know I've seen it mentioned a few times now in threads relating to the UI. Though it could just be one person, going off their own ears and judging poorly for all I know. :)
14:13:41preglowpeople should also stop trying to do centerfreq/bandwidth calculation using q in digital filters
14:13:47preglowit won't always be right
14:13:55preglowthanks to frequency warping
14:14:20preglowi'll add a proper bandwidth parameter one day i'm feeling particularily nice
14:14:25safetydanthe GUI could probably do with some better labels too
14:14:39safetydanprobably should be calling centre frequency cutoff for the peak filters
14:14:44safetydanshouldn't even
14:15:47B4gderwell, we can work out such issues after commit too
14:16:04preglowi'm justing waiting for a patch that fixes the whitespace issue, and i'll commit it as is
14:16:43safetydandidn't think you were waiting for me on that :)
14:17:09 Quit San_Benedetto ("soon ;x")
14:24:48preglowoh, but i am
14:24:53preglowi'd like it in cvs as fast as possible
14:25:15safetydanworking on it now
14:25:23safetydanlousy soft-tabs settings
14:26:12 Join sam^ [0] (
14:26:12JdGordoncan the h300 record from radio?
14:26:29peturfirst go to radio, then recording screen
14:26:35sam^hi there, I just read that RB is available for ipod nano. Is it possible to install this alternative on linux?
14:26:40peturand select line-in
14:27:34Paul_The_Nerdsam^: By alternative to, do you mean "in place of" or "alongside?"
14:27:44sam^I mean from within linux.
14:27:54sam^on my ipod nano.
14:28:20preglowsam^: sure
14:29:26preglowwhat, that seems to be windows only now :>
14:30:07sam^then I wouldn't try
14:30:43Paul_The_Nerd Most of the instructions are here.
14:31:09sam^ah, ok
14:31:11Paul_The_NerdYou can use the bootloader-nano.bin from the IpodInstallation page if you don't want to create a cross compiler
14:31:16 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
14:31:27Paul_The_NerdBut you will still need to compile the ipod_fw utility, I believe. I don't think a binary's around.
14:31:38preglowbinaries for linux is kind of frowned upon
14:31:44preglowwith reason, it doesn't work universally
14:34:04linuxstbWe could put a link directly to checkout the C source using viewcvs.cgi - it can just be compiled by typing "gcc -o ipod_fw ipod_fw.c"
14:34:10sam^If I follow the steps on the ipodboot-page, I would compile everything necessary for installing the bootloader?
14:34:43 Quit JdGordon ("Miranda IM! Smaller, Faster, Easier.")
14:34:57linuxstbsam^: You can just get ipod_fw.c from here:*checkout*/tools/ipod_fw.c
14:35:08linuxstbAnd then compile using "gcc -o ipod_fw ipod_fw.c"
14:35:21linuxstbThat's all the compiling you need to do, unless you plan on changing Rockbox yourself.
14:35:40sam^I checked out the whole CVS thingy now
14:35:55linuxstbThen you need to type "make ipod_fw" in the tools/ directory.
14:36:28linuxstbYou can then follow the Windows instructions, but instead of using the ipodpatcher.exe utility, you can use "dd".
14:37:02linuxstbYou should find which /dev/sd? device your ipod is attached as, as then read/write to the first partition - e.g. dd if=/dev/sda1 of=bootpartition.bin
14:39:54*amiconn suspects that 'Crossfeed' actually doesn't work in FM mode, despite it's available in 'Sound Settings'
14:40:13amiconnSame will probably be true for EQ
14:41:06 Join perldiver [0] (
14:42:04peturFM goes via the UDA mixer...
14:43:41Papricalinuxstb, little bug in the backdrop on the wps..
14:43:59Papricawhen you change wps and it without backdrop
14:44:07Papricathe old backdrop show...
14:44:17preglowamiconn: well, we could start sampling the fm radio, process it, and then play it back again :)
14:44:23preglowso it's not impossible, heh
14:45:03peturyour effect box, right?
14:45:18peturneeds bi-dir I2S
14:45:52preglowi think someone said it's supported
14:45:59preglowand yeah, for my fx box, heh
14:46:03sam^linuxstb: I guess, "dd if=/dev/sda2 of=bootpartition.bin" copies a file which is way too large
14:47:01Paul_The_Nerdsam^: sda2 would be the contents of the Fat32 partition.
14:47:16Paul_The_NerdIf it's sda, you need sda1.
14:47:22sam^ah .. okay
14:48:06sam^I didn't knew that there are two partitions mounted
14:48:17Paul_The_NerdWell, sda1 isn't generally mounted.
14:48:24Paul_The_NerdIt has no file system.
14:48:50sam^-mounted ;)
14:49:01*Paul_The_Nerd grins.
14:49:27*safetydan curses real work getting in the way of rockbox
14:49:36linuxstbsam^: Your ipod has two partitions - sda1 is the 80MB firmware/boot partition, and sda2 is the main FAT32 partition where your music goes (and where you unzip to).
14:50:58amiconnpreglow: Sampling the radio would in fact make sense, imho. We would autmatically have a working peakmeter
14:51:11amiconnRecording requires it anyway
14:51:41sam^I now extracted the partition and I have this ipod_fw file. where do I get the rockboot.bin ?
14:52:03linuxstbKeep reading the instructions - that's the file you create using the ipod_fw utility.
14:52:15 Join NiX8916Lightning [0] (
14:52:50sam^the ipod_fw executable even contains the rockbox bootloader? then it must be very small.
14:53:22preglowamiconn: yes, i thought about that the other day when i noticed that peak meter doesn't work for radio
14:53:27preglowamiconn: and won't either, afaik
14:53:31preglowamiconn: unless we sample it
14:53:49linuxstbsam^: Which instructions are you following?
14:54:01sam^I'm following
14:54:56linuxstbsam^: You need to download bootloader-nano.bin from the IpodInstallation page, and you can then follow step 3 in IpodBoot
14:55:44sam^thank you
14:57:25preglowexample of what the eq graph will look like on h1x0
14:57:28preglownot exactly plenty of space
14:59:23Paul_The_NerdDo any other DAPs allow a graph at all?
14:59:34preglownot that i know of
14:59:46preglowi'm hoping for another rockbox first for our collection
15:00:21preglowyou wont be able to manipulate it, it just plots what the frequency response of your current setup looks like
15:00:22sam^I put rockbox.ipod in the root of my ipod, but the bootloader fails. should I put it in .rockbox?
15:00:23petura red plot line on the H1x0 - that 'll be a first ;)
15:00:58 Join jaebird_web [0] (
15:01:39preglowit'll also be pretty slow, and we won't have enough screen space to keep it on the same screen as the sliders anyway
15:01:44preglowso it'll be in a separate screen
15:02:11preglowi was thinking just binding it to a spare key in the eq screen or something
15:02:16preglowso you can flash it up when you want to
15:02:22jaebird_weblinuxstb: I tested the 4g 2bpp from cvs...there is a screen shot somewhere in the irc logs
15:02:43peturjust imagine the band indicators below it, left/right selects the band, up/down changes it....
15:02:43Moospreglow: that sounds promising
15:02:56jaebird_webhowever, now I'm only getting a blank screen with backlight...not sure what happened.
15:03:07Paul_The_Nerdpreglow: I kinda expected it to be that way anyway. Maybe though, draw vertical lines corresponding to 0-4?
15:03:16jaebird_webIt still boots up with standard rockbox logo, but then goes blank
15:03:28linuxstbjaebird: As I said in the forums, it needs someone with a 4g to debug it....
15:04:14jaebird_webnot sure where to start...the bootloader works with the lcd
15:04:36sam^linuxstb: where do I put this rockbox.ipod file which contains the firmware?
15:04:51preglowPaul_The_Nerd: weeelll, it just takes a quick peek back at the sliders screen to see where the different filters are anyway
15:04:52sam^everytime I copy it on the ipod, it is deleted afterwards
15:04:58preglowPaul_The_Nerd: we'll see how much we can cram in
15:05:37 Part NiX8916Lightning
15:05:37preglowPaul_The_Nerd: that thing doesn't take all the screen space it could, there should still be room for some axis, like a frequency axis at the bottom and a amplitude axis at the left
15:05:38Paul_The_Nerdpreglow: It was just a thought. I mean, it could easily be a negative effect, but I was thinking on 2bpp screens that using the lightest shade of gray wouldn't very negatively affect the visual.
15:05:41linuxstbsam^: Download the and then extract directly to the root of your FAT32 partition. This will put rockbox.ipod and the .rockbox directory in the right places.
15:05:46sam^I guess it was a fault of konqueror sorry.
15:06:11linuxstbThe safest way is on the command-line - "unzip -d /mnt/ipod/" (or wherever you have mounted the fat32 partition)
15:06:15preglowPaul_The_Nerd: we'll see, i'll again just take the route of commiting the low-level math stuff, and let whoever wants to do the gui do that
15:06:20 Join lostlogi1 [0] (
15:06:28perplexitywhy not use the plot as a background and draw the sliders over the top alphablended :)
15:06:33preglowpeople who can do dsp is pretty rare around here, so i'll try to just stick to that
15:07:04preglowbut now i've gotta do some more Other Work
15:08:03Paul_The_Nerdpreplexity: The plot is apparently "slow." I'm guessing that means a noticeable amount of time to render.
15:08:08***Saving seen data "./dancer.seen"
15:08:20preglowwell, yeah
15:08:25 Quit lostlogic (Nick collision from services.)
15:08:28preglowevery pixel needs a significant amount of calculations
15:08:30Paul_The_Nerdperplexity: Also, it'd be nice to have it work the same on B&W screens.
15:08:30 Nick lostlogi1 is now known as lostlogic (
15:08:43jaebird_weblinuxstb: I'll start looking at debugging tonight...just need a little place to start
15:09:06perplexityI sorta had my tongue firmly jammed in my cheek.. thus the smiley :) but anyway
15:09:06preglowi _could_ do it cheaper than that, but graphing routine is very general right now, you can use it for any kind of filter, and i don't want to give that up
15:09:16 Join San_Benedetto [0] (
15:09:20San_Benedettohey can someone help me
15:09:27San_Benedettohow do i play doom
15:09:41San_Benedettoi put doom2.wad on the player
15:09:43Paul_The_Nerdperplexity: Ah. I tend to interpret :) as hopeful and ;-) or ;) as tongue-in-cheek. But that's a personal bias on the meanings, I guess. Heheh
15:09:44San_Benedettonow what? :\
15:09:45ashridahthe aliens don't want to be friends.
15:09:57*preglow cheers for the EQ commit!
15:10:03perplexityahh.. yes.. fair enough.. forgot the wink I guess
15:10:03 Quit perldiver (Read error: 110 (Connection timed out))
15:10:10Paul_The_NerdSan_Benedetto: Are you running the H300 optimized build?
15:10:21San_Benedettoit works now!!
15:10:25San_Benedettonever mind, thanks! :)
15:10:25linuxstbjaebird: Are you sure that Rockbox is displaying a blank screen, and it's not simply the file browser displaying no files?
15:10:49Paul_The_Nerdlinuxstb: That seems unlikely, since it should by default show rockbox.ipod, right? Well "Rockbox"
15:11:13linuxstbBut others have reported the 4g working fine - apart from the button driver not working.
15:11:14jaebird_webwell pretty sure...the first time I booted up I saw a battery indicator stuff at the top
15:11:21jaebird_webnow I get nothing on the screen
15:12:38jaebird_webhmm...I got the latest CVS about 13 hrs ago, has something changed that would affect 4g?
15:12:47preglowlinuxstb: button driver not working? perhaps our delay removal affects it worse?
15:13:16*safetydan watches build anxiously
15:13:24linuxstbpreglow: I tried that - didn't fix the problem.
15:13:25jaebird_webwhen I did get something on the screen I was able to click the menu button and it seemed to change stuff.
15:13:45preglowsafetydan: it'll take a while
15:14:56Rob2222_amiconn: My values are 21/43.44/67.68/89 (unit 1) and 1F.20/42.43/65/86 (unit2) and I have the problems with down recognizes as left sometimes at unit 1
15:15:01linuxstbjaebird_web: It could be a contrast problem. Try editing firmware/drivers/lcd-ipod.c to change the lcd_set_contrast() function to have the line "val=96;" as the first line.
15:15:36linuxstbThe initial version in CVS had the contrast set too low - and maybe that value is written in your settings.
15:18:31jaebird_webok...good, I'll try that, unfortunately we must be in vastly different timezones, because I'm not home now.
15:18:47sam^when I enable "crossfeed" in current ipod-nano, there is an annoying noise
15:18:57preglowcrossfeed doesn't work, and i don't know why
15:19:12preglowi've given up that crossfeed anyway, i'm waiting for jlo to post a description of his new one
15:19:12B4gderjaebird_web: us euro people aren't at home either right now ;-)
15:19:26*linuxstb works from home....
15:19:33*preglow doesn't...
15:19:49*petur wished he could
15:19:51B4gderoh, I want to do that again as well
15:19:51lostlogiclinuxstb: any news on ipod playback?
15:20:10lostlogicI can work from home if I beg my boss real nice every now and again
15:21:21sam^sometimes, it seem to scroll without the clickwheel is being touched
15:22:11linuxstblostlogic: Yes - I'm looking at your patch now. I've re-written the ipod parts of pcm_playback, and the current status is that it seems to work perfectly, apart from the fact that there is no sound coming of my headphones....
15:22:32jaebird_webthis is very exciting. :)
15:22:59linuxstblostlogic: But skipping, seeking etc is fine, and the PCM buffer seems to be going up and down normally.
15:23:19 Join hydrahead [0] (
15:23:38lostlogiclinuxstb: ahahaha apart from that lack of sound :( :)
15:24:25youngcerealto all
15:24:29 Join zhilik [0] (
15:24:36linuxstblostlogic: Yes, that's minor flaw...
15:24:38youngcerealwhat ist the problem with the battery on ipod
15:24:47preglowyoungcereal: lack of code
15:24:52linuxstbyoungcereal: Simply that no-one has implemented it.
15:24:52youngcerealah ok
15:24:59hydraheadhey ppl, any progress on the 4g buttons not responding?
15:25:23sam^it isn't currently possible to teach rockbox to use the ipod database?
15:25:24youngcerealthe same with charging i thing
15:25:46youngcerealif i want too charge the reboot to data transfare
15:25:54linuxstbyoungcereal: We think charging works whilst in Rockbox. Hold down the MENU button when you insert USB.
15:26:02preglowi'm pretty sure it does on nano
15:26:53linuxstblostlogic: Is your last patch on Sourceforge the one that you are waiting to commit?
15:27:19lostlogiclinuxstb: yes
15:27:41preglowdoesn't that break voice ui?
15:27:46lostlogicpreglow: no, that's fixed
15:27:51preglowoh, excellent
15:27:59preglowthen i have no reservations and eagerly await your patch
15:28:04lostlogicwas the same problem as teh pitch down
15:28:05youngcerealand I do what a men must do too see his Battery power I look too change it self (i hope i have a chance ) ;)
15:28:07 Part LinusN
15:28:44sam^w00t, rockbox supports ogg vorbis ...
15:28:51preglowhahah, yes
15:29:13sam^would like to have it support the db as well ;-)
15:29:23lostlogicsam^: so write a parser for it :)
15:29:24B4gdersam^: then you better start working
15:29:34sam^when I switch on file filtering on ID3 databaste, then there's nothing currently
15:29:39preglowi hope slasheri commits the tagcache soon
15:29:46sam^seems to be for some other players
15:29:57lostlogicsam^: it's a rockbox specific database
15:30:30sam^k ... what I like about the bootloader is that I can use my old firmware as well
15:30:42linuxstbWhy would you want to do that? :)
15:30:57B4gderoriginal firmwares are overrated
15:30:57*ashridah shudders
15:31:01sam^more eye candy :p
15:31:03Paul_The_Nerdlinuxstb: At least on the Nano, charging seems to fill up the battery.
15:31:41ashridaheye candy cannot be heard
15:31:53hydraheadhas anyone with debugging knowlege taken over the greyscale 4g ipod?
15:32:17linuxstbsam^: Try a backdrop image.... Copy a 176x132 .bmp file to your ipod, browse to it, hold down the select button until the "on-play" menu comes up, and select "set as backdrop".
15:32:22Paul_The_Nerdsam^: <−− Make your own While Playing Screen with all the eyecandy you want.
15:32:35linuxstbsam^: You can also change the font using the Browse Fonts option in the menu.
15:32:45sam^can those background be coloured?
15:32:49*linuxstb thinks Rockbox needs some nicer defaults...
15:32:59linuxstbsam^: Yes, they can be 24-bit bitmaps
15:33:58sam^which format are those fonts in?
15:34:06linuxstbRockbox format.
15:34:08 Quit Paprica (Read error: 110 (Connection timed out))
15:34:18B4gder.. but converted from bdf
15:35:30sam^what a pity that a lot of them is so small
15:36:07 Part San_Benedetto
15:37:59 Nick Lynx_awy is now known as Lynx_ (
15:38:37Lynx_Hehe, with talk about syncronized lyrics display, a Karaoke plugin can't be far away. Great for train rides and such ;)
15:40:01 Quit hydrahead ("CGI:IRC (EOF)")
15:41:10 Nick paugh is now known as AliasCoffee (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
15:42:35 Part jaebird_web
15:43:12 Join lostlogi1 [0] (
15:43:17peturhave a part of the wps screen scroll the lyrics :)
15:43:22 Quit lostlogic (Nick collision from services.)
15:43:27 Nick lostlogi1 is now known as lostlogic (
15:44:23sam^seems intresting, this project
15:45:38 Join ep0ch [0] (n=ep0ch@
15:46:19 Join ^BeN^ [0] (
15:46:43 Nick ^BeN^ is now known as Paprica (
15:46:51hyarionis there any photos on rockboxified iPods? (can't find any on the wiki)
15:46:57*hyarion wants to see how the UI looks like
15:47:28B4gderhyarion: you can build a simulator and try it out on host
15:47:29sam^hyarion: have a look at
15:48:01Paul_The_NerdMan, that review doesn't even show the Boxes theme
15:48:38peturdoesn't even show any nice WPS
15:50:22hyarionthat brickmania looks nice :D
15:50:45lostlogicspeaking of brickmania...
15:50:55sam^I put a new theme in the wps directory, but it doesn't show up in the menu?
15:51:28lostlogicPaprica: I found a real bug this time... I think... it's minor −− with glue paddle, if you get the ball stuck to the end of the paddle, and release it against a wall, the ball stays in the wall until it hits some bricks along the wall, is that intended?
15:51:50linuxstbsam^: A theme (the .cfg file) goes in .rockbox/themes. The wps itself (.wps and bitmaps) goes in .rockbox/wps/
15:52:22sam^apple now ships a 1 GB nano as well.
15:52:22Papricareally not...
15:52:39sam^it costs 159$/Ą, the shuffle prices were reduced
15:52:40Papricai know about this bug and dont know when it happens
15:52:45Papricathank u!
15:52:49Papricawill fix it
15:54:52ep0choh someone just got 60 points ;)
15:55:06safetydanphew, only a couple of more errors
15:55:19safetydaniAudio X5 Sim doesn't matter right?
15:55:25B4gdernot a lot, no
15:55:41B4gderat least not right now
15:56:20preglowsetting a couple of button defines isn't that hard to do, provided you know the layout of an x5, of course...
15:56:29Bg3rhaha the build table looks like a cross
15:57:21Bg3rlostlogic ?
15:57:23ep0chcrosses are a good sign :)
15:57:45ep0chunless its a diagonal cross
15:57:45B4gderI see at as a plus
15:57:50B4gderdouble-plus good
15:57:56B4gder1984 for everyone
15:58:25peturnext competition: creating figures on the build table :D
15:58:38B4gdernow that is a worthy challange
15:58:56Bg3rin fact it's easy :)
15:59:19ep0chnot if you dont have commit access :s
15:59:30safetydanalmost looks like the Swedish flag if you look at it sideways and mentally change the colours
15:59:34linuxstbRockbox logo anyone? Or do we need more targets?
15:59:35*Bg3r doesn't have patience to be scored :)
15:59:55ep0chpreglow: i was thinking a single "dsp on/off" option might be an idea...
15:59:56Bg3rto get ...
16:00:01preglowep0ch: why?
16:00:31amiconnAnother bunch of .lang strings, voiced :/
16:00:51ep0chwell in future for codec output testing, an easy method to switch the dsp off
16:01:18preglowep0ch: a right, that should be easy with lostlogicals proposal
16:01:27preglowbut i want dsp to be always on for ordinary use
16:01:37preglowit doesn't have to do anything but convert to 16 bit, though
16:01:52ep0chhehe i'd like to manually switch the whole thing on and off :)
16:02:22peturpreglow: does the eq give any performance hit when you just run it flat?
16:02:46preglowpetur: a tiny, tiny bit
16:03:03ep0chpetur: i think the testing that i did i couldnt notice any extra boostage
16:03:03preglowpetur: five extra ifs that never trigger
16:03:13preglowpetur: per block
16:03:44preglowbands that have 0 gain are never calculated
16:04:18preglowsafetydan: did you fix that gain 0 reset history thing, btw?
16:04:33safetydanpreglow, no
16:04:37preglowi'll do it
16:05:02safetydanI'm supposed to be doing real work at the moment
16:05:11preglowme too :-)
16:05:23preglow'tis nice to be paid for hobbies
16:06:09preglowafter so many exceptions thrown and gtkmm oddities, one needs a break
16:06:09 Join leftright [0] (
16:07:23leftrightthe EQ is great thank you, the GUI is very nice, amazing what you guys can do with zero's and one's
16:07:28safetydanJava and XML weirdness is my work today
16:08:08safetydanleftright: I'm glad people find it usable. I never actually use the EQ so I just sort of made something up for the GUI :)
16:08:38leftrightumm, the defaults that the EQ loads with are a bit weird -25 for band 3
16:08:55Paul_The_NerdThe EQ should've probably bumped the config version
16:09:00preglowso the eq has been codec entirely by people who don't use an eq
16:09:29preglowi guess 'codec' has kind of embedded itself into my fingers...
16:09:34leftrightperhaps you should try your own invention
16:09:50leftrightyou'll be pleasantly surprised
16:10:34preglowi've tried it, sure
16:10:42preglowbut i never use it for anything else than testing
16:11:38leftrightthen you are blessed with ears thst have a flat freq response, very fortunate
16:11:53preglowthat i'm not
16:11:56preglowi'm just blessed with not caring
16:11:57ep0chimho thats the way its meant to be heard
16:12:06leftrightheh, that sorts it
16:12:44leftrightah but us old farts need all the help we can get
16:12:51ep0chits fun to play with though :)
16:13:21preglowit would be with button acceleration
16:13:22peturleftright: so you've got a response curve for your ears... :D
16:13:55peturimho, it all comes down to a matter of taste
16:14:05leftrightI have a response curve for woman as well, but wont go into that here :)
16:14:47lostlogicBg3r: yar −− saw your commit... does it cause any problems with USB charge state reporting on H3x0?
16:15:02leftrightthanks folks
16:15:05 Part leftright
16:15:12 Join ]RowaN[ [0] (
16:18:02amiconnB4gder: Did the cvs display on the front page change?
16:18:07ep0chdid any other ideas come about for decreasing the dsp latency? i.e. having a really small buffer for dsp
16:18:24amiconnIt used to only display changes when the automatic build finished
16:19:15preglowit does a bit of both now
16:19:26preglowif a new build has just been started and you commit something, it will end up on the front page at once
16:19:41B4gderit still works like before
16:19:58Zagorafaik it updates periodically, independent of the builds
16:20:13B4gderit updates after the builds
16:20:42ep0chwhat happens when someone commits during a build? does building stop and restart?
16:20:53]RowaN[we need some cool EQ presets =]
16:20:58B4gderep0ch: it just builds again after the current build
16:21:25B4gderzagor is right, it builds the table even if no builds are made
16:21:46ep0choh ok, it might never actually get a build done if it had to keep stopping
16:22:11 Quit needleboy ()
16:22:15 Quit TCK ("well, if you say so.")
16:22:39B4gderit simply checks for changes every 60 seconds and rebuilds everything if there was one
16:23:51lostlogicB4gder: any further progress on using distcc for some of the builds OR makign the builds parallel build friendly?
16:23:52 Quit DreamTactix291 (Read error: 104 (Connection reset by peer))
16:24:12B4gderhaven't worked on that anymore
16:24:15 Join DreamTactix291 [0] (
16:24:39B4gderI'm not so sure distcc will speed up things very much though
16:25:00ep0chyou know the IRC webpage, it would be nice if there was a seperate page for each year, and the order of months reversed
16:25:07preglowB4gder: how come?
16:25:10 Join XavierGr [0] (
16:25:25B4gderpreglow: because of the transfer times compared to the local compile time
16:25:43preglowyeah, that's a valid point
16:25:49preglowit all depends how parallel you can make it
16:25:56lostlogicB4gder: but you could run 3 of the builds at once, to the different hosts perhaps?
16:25:56Zagorwe've got a pretty slow net connection
16:26:14amiconnWhat about distributing the different builds to several build boxes?
16:26:20B4gderwe'd need to get "make -j" to work first
16:26:29XavierGrpreglow do you know iPod's 4g y screen size?
16:26:34preglowXavierGr: 'fraid not
16:26:38preglowbut can find out
16:26:40amiconnE.g. one would build all archos builds, one would build all iriver builds etc
16:26:49t0masamiconn has a point there...
16:27:05lostlogicor what amiconn said, you could put a scriptlet on my boxen if I gave you shell access that would update and build some targets and then just send the logs and zips back
16:27:12t0masand are we using ccache yet?
16:27:14B4gderwe could do that, but the point of distcc is that it takes very little admining for the admins of the server
16:27:15XavierGrpreglow: ok then could you tell me 2 buttons that are free to change picture (up or down) in the Jpeg plugin?
16:27:15t0masthe samba thing
16:27:24preglowXavierGr: colour or non-colour?
16:27:26t0masthat would save a lot of recompiling of things that aren't changed
16:27:46preglowgrayscale is 160x128, colour/photo is 220x176
16:27:52XavierGrwell the jpeg viewer is currently for ipod 4g
16:27:57B4gderlostlogic: yes, that would make sense
16:28:16XavierGrhm okay then I should use the 160
16:28:27XavierGrand free buttons on the greyscale?
16:28:33preglowisn't that exactly the same size as h1x0?
16:28:49XavierGr2 free buttons to chage pics?
16:29:03 Quit B4gder ("time to say moo")
16:29:07t0masB4gder: and the ccache thing?
16:29:13preglowthey're using that
16:29:16t0masah ok
16:29:23t0maswhy is compiling still taking so long then?
16:29:24XavierGr(except for the default ones that scroll the pic)
16:29:27preglowfree buttons? can't you use any of them?
16:29:44amiconnXavierGr: There is no grayscale lib for the greyscale iPods yet
16:29:48preglowipod has scrollwheel left, scrollwheel right, up (menu), down(play), left(prev), right(next) and select button
16:29:54XavierGrlet me describe it:
16:29:55Paul_The_NerdAre there even *any* buttons mapped for the JPEG viewer on iPod yet?
16:30:07Paul_The_NerdI think he's asking which buttons aren't already used.
16:30:37XavierGrin the filescroller if a pic is small one and fits the screen without zoom, up and down will change pics, but if the pic is zoomed (large pictures) then we must have 2 other buttons for it.
16:30:42XavierGrWhat do you recommend?
16:31:03XavierGramiconn:really I saw some changes for it.
16:31:43XavierGrI thought it is now supported. Anyway preglow just tell me 2 buttons that are not used in jpeg plugin
16:31:43 Join leftright [0] (
16:31:52linuxstbXavierGr: You could use left/right and select - i.e. (BUTTON_RIGHT|BUTTON_SELECT)
16:31:52Paul_The_NerdXavierGr: (BUTTON_SELECT|BUTTON_RIGHT) is what I'd recommend.
16:32:01linuxstbGood idea...
16:32:13Paul_The_NerdMan, wish I'd thought of that linuxstb. ;)
16:32:38leftrightIf I load another EQ config from Saved EQ presets, it doesn't update the GUI
16:32:53XavierGrso BUTTON_RIGHT | BUTTON_SELECT for next picture. and what BUTTON_LEFT |BUTTON_LEFT for previous? is that ok?
16:33:09Paul_The_NerdSeems good to me at least.
16:34:18leftrightmaybe I need to check my config file first incase you hve changed the labels there
16:36:23 Join ModernExecutive [0] (
16:36:30 Quit ModernExecutive (Remote closed the connection)
16:38:00 Quit ashridah ("Leaving")
16:39:01 Nick josh__ is now known as josh_ (
16:40:37amiconnBg3r: yellow builds...
16:43:38leftrightsafetydan, the GUI doesn't update If I load another EQ config from Saved EQ presets, anyone else ?
16:44:44Bg3ramiconn i saw them
16:44:53Bg3rat least they aren't red :)
16:44:57leftrightthe sound does change with a different cfg, but gui no
16:45:38XavierGrOkay I updated again the jpeg filescroller. If any dev good take a peek and comment it I would be really pleased.
16:45:50XavierGrgood = could
16:45:54safetydanleftright, it seems to work fine here in the sim
16:46:00safetydanI'll take a closer look after work
16:46:48leftrightthanks, I have tried several saved configs, and it changes the sound but not the display
16:50:29 Part sam^ ("Konversation terminated!")
16:52:39 Join Mark_ [0] (
16:52:52 Quit Kohlrabi (Read error: 104 (Connection reset by peer))
16:54:39 Join damaki [0] (
17:01:13leftrightsafetydan, the cfg layout is defferent now, I re-did my files and its working fine now
17:01:51 Quit ]RowaN[ ("CGI:IRC (EOF)")
17:04:08Bg3ramiconn ?
17:04:11Bg3ri don't get it
17:08:12***Saving seen data "./dancer.seen"
17:08:25 Join spiralout [0] (
17:08:52spiraloutHAPPY BIRTHDAY ! LinusN ! :D
17:09:28PapricaHappy birthday =]
17:10:00Bg3rwhat ? :)
17:10:25spiralout;) yeah check the forum ...
17:10:40Bg3rhip hip hooray for linus :)
17:11:20youngcerealhappy birthday
17:11:57peturhe's not here, probably gone drinking already :D
17:12:23safetydanleftright, okay that's cool. thanks for the followup
17:12:25Bg3rhe should:)
17:13:21spiraloutor am i wrong ? "Upcoming birthdays" *g
17:14:55 Quit goa (Read error: 131 (Connection reset by peer))
17:15:59 Join goa [0] (
17:17:06 Part leftright
17:17:25 Quit goa (Client Quit)
17:17:35 Part Paul_The_Nerd
17:17:39 Join goa [0] (
17:19:12 Join hardeep [0] (
17:19:59 Quit DJDD___ ("Trillian (")
17:25:52Bg3rBagder ?
17:27:46petursaid moo a while ago
17:30:37 Join perldiver [0] (
17:30:46Bg3ruf, i can't locate the fscking ifdef problem ...
17:31:25 Quit linuxstb (Read error: 113 (No route to host))
17:31:39 Join linuxstb [0] (
17:32:03 Join ghode|afk [0] (
17:32:18 Join ]RowaN[ [0] (
17:32:28 Quit ]RowaN[ (Client Quit)
17:32:35 Join ]RowaN[ [0] (
17:33:41]RowaN[i just found out that my WPS was spazzing up because i used a mid dot character ( .... maybe one of your beautiful people can fix the WPS engine to make it behave better when mid dots are used =]
17:33:48 Join ModernExecutive [0] (
17:34:35]RowaN[&middot; ... its the middle one of these 3: '┬Ě.
17:35:07 Quit ModernExecutive (Client Quit)
17:35:49 Join ModernExecutive [0] (
17:36:53linuxstbBg3r: In firmware/export/power.h, charger_inserted() is only defined if HAVE_CHARGING is defined. But in apps/tree.c, it is checked based on RECORDER_PAD and !HAVE_SW_POWEROFF
17:38:04Bg3rlinuxstb i saw it already ...
17:38:17linuxstbI'm now agreeing with you - it doesn't make sense...
17:38:20Bg3rwas looking a few lines below ...
17:38:45Bg3rHAVE_CHARGING isn't defined for the sims
17:39:11Bg3radded it to the define
17:39:22*linuxstb goes away again....
17:39:39 Join tvelocity [0] (
17:40:06Bg3r10x :)
17:40:43 Quit ]RowaN[ ("CGI:IRC (EOF)")
17:43:07 Quit spiralout ("ChatZilla [Firefox]")
17:44:45 Part petur
17:45:42Papricayou will be amazed
17:45:53Papricano no, its not an ipod =]
17:46:00Bg3rwhat's tis :)
17:46:06Papricajust look
17:47:47 Join Kohlrabi [0] (
17:50:53Moosthat looks good for the one liking ipod screen
17:51:05Moosnot my case ;)
17:51:35Papricamm its a bitmap scroll bar..
17:51:43Papricanot only for the ipod....
17:51:58MoosI noticed it, that looks good
17:52:21Moosdo you have patch for this yet?
17:52:44Papricamm i do it now to the grayscale lcd too..
17:52:53linuxstbPaprica: What's the "-25%" mean at the bottom?
17:53:12Papricahah it need to be db..
17:53:28Moosgood spot linuxstb
17:53:33linuxstbDoes that use a backdrop image?
17:53:57Papricammm the progressbar not, its another tag
17:54:27NicoFRPaprica: very nice indeed
18:01:14 Join DangerousDan [0] (
18:01:20ep0chwhatever next eh a bitmap peakmeter?
18:01:48ep0chnow i'm jealous not having a colour screen
18:03:39Papricayapiiii working wonderful on the player
18:07:14lostlogicPaprica: the progress bar bitmap must be a 1px wide bitmap, or how does it work?
18:07:36ep0chah its seems not all 128 kbps mp3s are equal. some play without boost, some do. wierd.
18:08:55lostlogicep0ch: you're jealous of color screen, I'm jealous of better performance
18:09:18Papricammm you give full length bitmap and it takes from it the part it needs
18:11:44ep0chha when looking at the pcm debug screen, holding a button down causes the pcm buffer to drop down much quicker
18:11:57lostlogicep0ch: yeah, I've noticed that too
18:13:30lostlogichmph, I think that the pcmbuf_get_latency calculation is off with my patch...
18:13:42lostlogicep0ch: does the progress bar look further off with my patch than it used to?
18:14:13ep0chlostlogic: looks fine
18:15:00ep0chlostlogic: oh, do you mean progress progress bar or pcm buffer progress bar?
18:15:09lostlogicprogress bar in wps
18:15:18lostlogicI think it's ~1/5s off
18:16:23ep0chcant really tell
18:16:30 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
18:16:53NicoFRPaprica: do you have a patch for your bitmpa progress bar ?
18:17:24Papricamm not yet. i do it for the greyscale lcd's now
18:18:10NicoFRlooking forward to it :)
18:20:00 Join _FireFly_ [0] (
18:25:53lostlogicPaprica: you're my hero −− doing all the pleasant GUI things that are important to the project, but that I have no interest in :)
18:27:09lostlogicSpeaking of WPSs anyone have opinions they'd like to share on how to offer clock display in the WPS? I have a patch on the tracker that does it, but I think the tag setup is objectionable.
18:27:19 Join bluey [0] (
18:29:11 Nick Lynx_ is now known as Lynx_awy (
18:29:18Papricalostlogic, hihi =]
18:29:28Paprica10x ;]
18:29:46 Quit DreamTactix291 (Read error: 110 (Connection timed out))
18:33:12 Join yeahx [0] (
18:36:22 Quit DangerousDan (Read error: 104 (Connection reset by peer))
18:36:59 Join DangerousDan [0] (
18:41:11 Quit Matze (Nick collision from services.)
18:42:48 Join LinusN [0] (
18:43:45Bg3rLinusN :) happy birthday :) wish u all the best ;)
18:43:59Bg3rand more free time :P
18:44:19*Moos is joining to Bg3r :-)
18:44:21LinusNBg3r: psst, it's tomorrow :-)
18:44:26Bg3rah ?
18:44:28Mooshehe :)
18:44:45Bg3rsomeone here said that it's today ...
18:44:54Bg3rsorry :(
18:44:59Mooshow du you know this guys?
18:45:05Moosspare or what? :)
18:46:00Moosoops, spy I mean$
18:50:34 Join _Kingstone [0] (n=sg1@
18:50:34 Quit Kingstone (Connection reset by peer)
18:50:38 Join PaulJ [0] (
18:50:46 Nick _Kingstone is now known as Kingstone (n=sg1@
18:54:25 Join shriek [0] (
18:55:05 Quit yeahx ()
18:55:54kkurbjunXavierGr: I would be interested in a compiled version of the midi player for the H300 if you get a chance (midi2wav) that is. I'm curious how it sounds on some files I have.
18:57:43shriekdoes wett no longer frequent the chatboard?
19:01:08 Join tucoz [0] (
19:01:26tucozHi, anyone had a look at the encoding patch?
19:02:16tucoza few weeks old, but is there anything wrong with the approach except the problems stated in the comment to the patch?
19:03:36 Quit ModernExecutive (Read error: 110 (Connection timed out))
19:06:11linuxstbtucoz: I don't think anyone's looked at it seriously.
19:06:34 Join ModernExecutive [0] (
19:06:39tucozok, would be cool if the patch is ok
19:08:15***Saving seen data "./dancer.seen"
19:08:56tucozlinuxstb: this page seems quite obsolete. Should I write that, and link to the ManualHowto instead?
19:09:40Cassandratucoz - archive it off somewhere. I'd been meaning to do that.
19:09:51tucozCassandra, how do I do that?
19:10:05CassandraErm, copy it into a new page?
19:10:32CassandraIt does contain some potentially useful info about the old manual and the tools needed to work with it.
19:10:32linuxstbClick on "more" at the bottom, and you have a rename option.
19:10:45tucozok, I do that then
19:10:54CassandraAh, yes, that's what I *meant* to say.
19:11:00*Cassandra grins.
19:11:03safetydanNow that the EQ UI is committed, can someone close please
19:11:14safetydanor can I do that?
19:11:29CassandraI think that's seperate, but you should be added.
19:11:35CassandraSpeak to bagder.
19:11:37CassandraI'll close it.
19:11:49tucozDocBoxObsolete, is that ok?
19:12:06safetydanapparently I can do that
19:12:42CassandraFair enough. He must've done it when he gave you cvs access.
19:13:17 Join einhirn [0] (
19:13:23Papricanew patch
19:13:43*Cassandra saw it on the patch tracker. Looks interesting.
19:13:55 Join einhirn_ [0] (
19:13:59CassandraEspecially combined with your other pb patch.
19:14:34Papricayou saw the video?
19:14:43Papricaor you try?]
19:14:48 Join webguest18 [0] (
19:14:54 Part webguest18
19:14:55Papricammm if someone can try it..
19:15:23amiconnBah, more bitmapping :/
19:15:34tucozCassandra, do you think I should add the ManualHowto as a CategoryFrontpage instead?
19:15:51CassandraYeah. Sounds like a plan.
19:16:13Cassandraamiconn: Rockbox's own Grumpy Old Man.
19:16:21amiconnHaha :/
19:16:30Papricaamiconn, what bad?
19:16:38CassandraThe amusing thing is that he did a lot of the groundwork that made all this graphical stuff possible.
19:16:43amiconnI would really like to see the port completed instead of adding more and more new stuff
19:17:07amiconnMost of this new stuff I'd call non-essential at best
19:17:18CassandraWhich is why I favour a feature freeze in the near future.
19:17:32 Join damaki_ [0] (
19:17:46amiconn'completed' of course in the sense of 'supports everyting that's supported on the other targets, as far as technically possible'
19:18:09tucozlike the encoder arch
19:18:13amiconnI would also like to see more of the existing plugins adapted
19:19:11amiconnI don't have that much time this week, but I'll see what I can do. Jpeg for the colour targets shouldn't be hard
19:19:28amiconn(in greyscale as a first step. I don't know enough about jpeg decoding to add colour myself)
19:19:36 Join Paul_The_Nerd [0] (
19:19:37preglowi can have a small look at it
19:19:37NicoFRPaprica: i'm compiling with your patch right now...
19:19:44preglowsince i at least know how jpeg works
19:20:12amiconnpreglow: I hope that I'm able to make things easy for you.
19:20:28NicoFRpreglow & amiconn: that would be awesome :)
19:20:31PapricaNicoFR, good =]
19:20:36amiconnI'll add xcld_color_bitmap() in addition to xlcd_gray_bitmap()
19:20:54amiconnSo switching to colour should be easy as soon as the decoder is extended
19:21:07preglowamiconn: and i _think_ just ifdefing the colour parts of jpeg out of jpeg.c should work well for targets that don't support it
19:21:27 Part tucoz
19:21:33preglowas in "isn't incredibly messy"
19:21:42amiconnIn fact that has to be done, otherwise I'm afarid the plugin will become too large for archos
19:21:45NicoFRPaprica: seems to work fine
19:22:01preglowamiconn: i know it has to be done, i'm just hoping it would be a bloody mess. i'm no big fan of ifdefs in code
19:22:10preglowwon't be...
19:22:42amiconnDo you have an idea what's the difference between standard and progressive jpeg?
19:22:56amiconnBecause, progressive isn't supported by the current decoder
19:23:09shriekrom what i know a progressive jpeg is one that is optimized for the web
19:23:22amiconnI'm wondering how much additional code it would mean to add that
19:23:48amiconnshriek: I know. I mean the internal differences of the format
19:23:53preglowi think the memory requirements are worse than the additional code
19:24:29amiconnIn what way?
19:24:38shriekamiconn: haha, i thought you might
19:25:08amiconnI mean, the jpeg plugin takes over the audio buffer anyway, and does some sort of dynamic management
19:25:33preglowyeah, you're right
19:25:49preglowit might not be a big deal, i've never made a progressive jpeg decoder
19:26:56Paul_The_NerdOut of curiosity, while poking in there, how hard would it be to change it to "Image viewer" and add BMP support?
19:27:29amiconnIf we do pseudo-dynamic allocation for some arrays which are now declared statically, we could almost double the code size before running out of plugin ram on archos
19:27:57amiconnjpeg.rock takes 26KB of the available 32KB, but code+rodata+data is only ~16KB
19:29:28 Quit shriek ("CGI:IRC")
19:29:31 Quit damaki (Read error: 110 (Connection timed out))
19:29:31Slasherilostlogic: Hmm, if you find the pcmbuf patch you made is working, please commit that.. i am also planning little modifications to the code at the moment
19:29:34 Join shriek [0] (
19:30:00 Join LuffarJoh [0] (
19:30:21preglowSlasheri: he's going to, i think the only thing holding back is some ipod problems
19:31:05amiconnHah, now this is interesting :)
19:31:13Slasheripreglow: ah, ok.. hmm, maybe i will then commit first and let lostlogic resolve some conflicts (only a little modification to reduce the eq latency when entered to the configuration screen)
19:31:34 Quit goa (Read error: 104 (Connection reset by peer))
19:31:49 Quit einhirn (Read error: 110 (Connection timed out))
19:32:04*amiconn always thought the file transfer Laptop<->Amiga via AmigaSamba is slow because of the slowness of AmigaSamba. Turned out the old router was the cause :)
19:32:29amiconn100KB/s -> 450KB/s
19:32:37safetydanYou still use an Amiga?
19:32:57preglowSlasheri: how do you do it?
19:32:58safetydanwow... I haven't used one of those in years
19:33:12safetydanGood times with an A2000 then an A1200
19:33:16amiconnIt's mainly my emaill machine now
19:33:17safetydanI even bought a CD32
19:33:35Slasheripreglow: limitting how much there can be data in the pcm buffer (something like faking the pcm buffer to be smaller than it is)
19:33:46preglowSlasheri: yeah, i actually considered that a while back
19:33:48amiconnA4000/060, 80MB RAM, 60GB HD, gfx card, 18" TFT
19:33:54amiconn(shared with my new linux box)
19:34:14preglowSlasheri: and it might actually work, but i don't think it's a permanent solution to the latency problem
19:34:35preglowSlasheri: how hard is it to make all other sound settings behave the same?
19:34:47Slasheripreglow: Hmm, the other solution would require storing 32 bit samples and that would double the space required for crossfading
19:35:02Slasheripreglow: easy, just call pcmbuf_set_low_latency(true);
19:35:18preglowand then the pcm buffer is allowed to empty until it reaches the new watermark?
19:35:39Slasheriit will be flushed immediately until the new limit is reached
19:35:42preglowso there will still be a period where there is subtantial latency
19:36:12Slasheriyes, and that latency will be exactly the pcm buffer current data size (- 0.5s or something like that)
19:36:22preglowbetter than nothing
19:36:26 Quit AliasCoffee ("Leaving")
19:36:36kkurbjunXavierGr: never mind, I have it outputting on my player now
19:37:00preglowbut i'd really like a good solution to it, and the only other solutions i can think of is: 1) 32 bit pcm buffer, 2) 16 bit pcm buffer, and reduced calculation precision (bad), 3) a very short pcm buffer (bad)
19:37:03Paul_The_Nerdkkurbjun: Did you happen to see the message in the log I left about Doom?
19:37:13kkurbjunPaul: what was that?
19:37:22kkurbjunwhere it was crashing?
19:37:36kkurbjunI did see that. is it the same spot for both wads?
19:37:49linuxstbpreglow: If you go with 1), we can just send 32-bit data to the DAC on the ipod.
19:37:54Paul_The_Nerdkkurbjun: It's *about* line 216 in r_things.c -> if (*(int *)lumpinfo[l].name == intname)
19:38:18Paul_The_NerdIt's very definitely *exactly* that line for Doom2 wad.
19:38:28preglowlinuxstb: for real?
19:38:35Paul_The_NerdFor shareware Doom, I didn't trace it yet.
19:38:43Paul_The_NerdIt was still in the Init Refresh Daemon portion
19:39:25kkurbjunPaul: I can give you something to test that's in idoom to see if it works. I want to replace that function eventually with the re-written code from Prboom.
19:40:13 Join damaki__ [0] (
19:40:28 Quit hardeep ("Chatzilla 0.9.70 [Firefox]")
19:40:37Slasheripreglow: in fact i think i might try to design some dynamic chunked buffer allocation for pcm and codec data.. that would allow buffer adjustments on the fly and enable the possibility for system wide malloc.. :)
19:40:42kkurbjunPaul: here is what they have:
19:41:04*amiconn slpas Slasheri
19:41:09amiconn*slaps even
19:41:12Slasherihehe :D
19:42:40Slasherior maybe that wasn't so good idea.. :)
19:42:41preglowthe only kind of malloc i want is a method to reallocate the mp3 buffer on the fly
19:42:50preglowso we can enable disk cache without rebooting
19:42:50 Quit Seed (Nick collision from services.)
19:42:57 Join Seed [0] (
19:43:10preglowall other approaches are pointless, i don't miss malloc
19:43:13Slasheripreglow: if we do that, then we get also the system wide malloc on the other hand..
19:43:23preglowSlasheri: well sure, but it can't be used very often
19:43:26linuxstbme shivers
19:43:29Paul_The_Nerdkkurbjun: That's not the right part of the code though...
19:43:30preglowSlasheri: since playback will be interrupted each time
19:43:43Slasheripreglow: hmm, why would it be?
19:43:43preglowwhich is fine! so no one will try to use it
19:44:07preglowSlasheri: well, you _COULD_ rearrange the entire contents of the mp3 buffer too, so it's seamless...
19:44:10Slasheriusing that malloc, could cut the used or unused portions of memory from the codec buffer..
19:44:12preglowbut that's not keeping it simple
19:44:13Slasheriso no interruptions
19:44:24Paul_The_Nerdkkurbjun: <−−- Line 46 of this
19:44:34amiconnI'm with you on the dynamic audio buffer reallocation - meaning all buffers participating in the scheme have to be reallocated in order to leave no gaps
19:44:56 Join matsl [0] (
19:45:02kkurbjunPaul: I know, but it affects that later code
19:45:10 Quit bluey (Remote closed the connection)
19:45:25amiconnI'd think it would be good for playlist buffer, tree buffer, dircache, and voice buffer on swcodec
19:45:53amiconnNothing more; I don't want rockbox to have a real malloc(). It would open the door to complexity explosion
19:46:20amiconn...since you would have to check the result of every single allocation
19:46:32Paul_The_Nerdkkurbjun: Where is NUMSPRITES defined?
19:46:47amiconnI don't want to see rockbox spitting 'malloc errors', as ipodlinux reportedly does...
19:47:00preglowyes, completely agreed
19:47:07kkurbjuninfo.h or info.c
19:47:17preglowi don't want a general malloc to be available at all
19:47:41amiconnYes. Imho we should try harder to get rid of malloc in the codecs
19:47:46Slasherihmm, maybe some sort of special/simplified malloc for those few buffer reallocations would be good then
19:47:51amiconnWOuld give back 512KB of RAM
19:47:53preglowi think so
19:48:29amiconnThere already is buffer_alloc(), but it's not restartable
19:48:57linuxstbI think libfaad is the main culprit for malloc - the others are quite well tamed already.
19:49:07*preglow slaps libfaad
19:49:11*preglow slaps libfaad again
19:49:13*preglow slaps libfaad again and again
19:49:14Slashericodec buffer could act as malloc pool, so malloc/realloc would always take some memory from the codec buffer and keep it contiguous
19:49:18preglowfeels damned good
19:49:21*linuxstb listens to aac on his ipod...
19:49:25amiconnHmm, and faad isn't even usable (on iriver)
19:49:49preglowit's a bitch to optimise
19:49:56linuxstbNo-one seems to care about faad. Maybe once more ipod developers come along, one of them will take it up.
19:50:06preglowit's a great ugly hung
19:50:08linuxstbI can't imagine iriver users having large aac collections.
19:50:22amiconnlinuxstb: I wonder why faad is realtime on ipod while alac isn't. The rellation is quite the opposite on iriver...
19:50:55linuxstbMaybe it's gcc4 - aac was slightly faster on iriver with gcc4. And it's only just realtime on the ipod.
19:51:02amiconnDoes faad come with arm asm optimisations?
19:51:06preglowi suspect alac has something else wrong with it
19:51:13preglowlinuxstb: did you test it with null output driver?
19:52:01*amiconn would is curious whether Tremor is realtime on iPod
19:52:29preglowit is
19:52:32amiconnAfaik ipl can't play ogg >128kbps
19:52:47Paul_The_Nerdamiconn: Q8 vorbis stutters *sometimes*. Very rarely.
19:52:48preglowworks fine for all my oggs
19:52:57linuxstbIt seems libfaad does have arm optimisations, but we're not using them....
19:53:23*Paul_The_Nerd blinks
19:53:28preglowamiconn: any reason why havent split more plugins into several files?
19:53:33preglowjpeg.c should be split...
19:53:33 Quit shriek ("CGI:IRC (EOF)")
19:53:42kkurbjunPaul: did that code do anything?
19:53:56*Cassandra suspects this to be connected to the fact the Linux on an iPod is total overkill.
19:53:59Paul_The_Nerdkkurbjun: Just about to test it. Sorry, had some other issues to work out with my build.
19:54:08Paul_The_NerdI'm compiling now
19:54:21preglowlinuxstb: btw, if you're just talking about the general 64 bit math macros, gcc might be wise enough to do that itself
19:54:31linuxstbI think it probably does.
19:54:52preglowanything else would make me seriously disappointed
19:55:15Paul_The_Nerdkkurbjun: Okay, still having build errors. Give me a few minutes, I've messed up my tree somewhere.
19:55:28kkurbjunPaul: no problem
19:55:31 Quit damaki_ (Read error: 110 (Connection timed out))
19:56:16*amiconn suspects that Cassandra suspects right
19:56:33preglowlinuxstb: btw, what makes you think the opts aren't used? isnt __arm__ defined on all arm platforms?
19:56:53amiconnpreglow: A standard plugin is single file. Multiple file plugins are exceptions
19:57:08 Join Henrico [0] (
19:57:29preglowamiconn: any reason for this?
19:57:46preglowi don't like navigating around huge c files
19:59:00linuxstbpreglow: Yes, you're right. They are being used. Ignore me.
20:03:13amiconnpreglow: First the build system supported singe-fle plugins only. Support for multi-file plugins was added later, specifically for rockboy
20:03:44amiconnAnd I think it's easier for a beginner to start a plugin when it's a single file
20:03:55amiconnNo other reasons.
20:04:06preglowyeah, sure, but that's a build system issue
20:04:25preglowit's no harder if all files are collected in one dir with the plugin name, for example
20:04:46amiconnIt is
20:05:19amiconnYou'll have to know how to reference functions in other modules
20:05:33preglowwell, ok, it's hard for people who haven't mastered basic c...
20:05:39preglowlet's not make that a design issue
20:05:46*amiconn remembers that he didn't understand this at all when he started coding in C
20:06:07Paul_The_Nerdkkurbjun: Still Data Aborts at the exact same spot with those changes.
20:06:11amiconnWe can of course keep both ways
20:06:46preglowkkurbjun: didn't you say something about packing the structs one time?
20:07:14amiconn..but I don't think the single-file plugins existing today are too large, except those that currently contain excessive amounts of bitmaps. They should be converted to use the bmp build system instead
20:07:25kkurbjunpreglow: yes, but paul removed all the structure packing I did, turns out it wasn't needed
20:07:42kkurbjunadded some later, but it was only what Idoom had
20:07:49preglowdata has to be aligned to its natural size on arm
20:07:52preglowso it'll data abort
20:08:01preglowamiconn: well, 2k lines in a c file is too much, if you ask me
20:08:04linuxstbamiconn: Some of them are. Sudoku for example is getting unmanagable - and will be even more so if a game generator was added.
20:08:26amiconnSudoku is just 1000 lines
20:08:50amiconnjpeg is almost 2000
20:10:00kkurbjunpreglow: yes, linuxstb was telling me that.
20:10:00 Quit einhirn_ ("Miranda IM! Smaller, Faster, Easier.")
20:10:01amiconnStill manageable, imho
20:10:24kkurbjunPaul: does the code ever get into the if statement?
20:10:29 Join DreamTactix291 [0] (
20:10:38kkurbjunor does it always crash immediately there?
20:10:40amiconnTake twice the size, then it starts to become inconvenient...
20:10:47NicoFRi just saw you were talking about libfaad and aac... is there any hope for aac on iriver ?
20:11:12 Part ModernExecutive
20:11:33Paul_The_Nerdkkurbjun: It never gets in. I have a printf( immediately inside it that never is called.
20:11:42Paul_The_Nerdkkurbjun: Oddly, the code there is exactly the same in iDoom...
20:12:59preglowNicoFR: sure, but going to take loads of work
20:13:09preglowfaad is so bloody bloated
20:13:20preglowan aac decoder can be really small, but faad isn't
20:13:58kkurbjunPaul: I think the problem is in w_wad.h that is the section I changed from idoom
20:14:12kkurbjunit's code to speed up finding lumps
20:14:59Paul_The_NerdI noticed your lumpinfo_t struct is very different from iDoom's
20:15:00kkurbjunand lumpinfo wouldn't be alligned on a boundary because of name (I'm guessing)
20:16:13 Join webguest08 [0] (
20:16:37 Join damaki [0] (
20:18:39kkurbjunPaul: I think the Prboom code will fix it, I'm trying to get the midi player running right now though. I may be able to work on it tomorrow.
20:19:10NicoFRpreglow: aren't there any alternatives to faad ?
20:19:22preglowyes, but not any with licenses we can use
20:19:27preglowthe helix one is very nice :/
20:20:16Paul_The_Nerdkkurbjun: Well, no rush. Let me know next time you need something tested.
20:21:12safetydanpreglow, how's that EQ visualisation going?
20:21:31preglowsafetydan: oh, didn't do anything further on it, i'll come back to it soon
20:21:48kkurbjunPaul: will do. I'll try and integrate the diff you had: what was the link to it again?
20:21:56preglownow you get to work on button acceleration :>>>
20:21:57safetydanrighto, guess I'll just work on the rest of the TODO list then
20:22:00kkurbjunI forgot to save it
20:22:06Paul_The_Nerdkkurbjun: Just a sec.
20:22:10preglowthat's the most annoying part for me now
20:22:16linuxstbpreglow: Did you choose the ipod button mappings for the eq?
20:22:20 Join sam^ [0] (
20:22:25preglowlinuxstb: nope, safetydan did
20:22:26safetydanlinuxstb, that was me
20:22:34preglowlinuxstb: i'm still not finished thanking god it even works on ipod
20:22:35Paul_The_Nerdkkurbjun: Let me know when you've saved it, don't like to leave apache running.
20:22:43linuxstbI was thinking the scroll wheel and left/right should be swapped.
20:22:47preglowlinuxstb: me too
20:22:51sam^is it possible to configure the bootloader to boot into apple's OS per default?
20:23:00preglowsam^: of course
20:23:04safetydannot having an iPod I picked keys after playing around in the sim
20:23:13linuxstbsam^: vi bootloader/ipod.c
20:23:30sam^okay, so I would have to compile it by myself -thx
20:23:47kkurbjunPaul: have it, thanks
20:23:50Paul_The_Nerdsafetydan: I second linuxstb's opinion, by the way.
20:24:11safetydanrighto, fixed next time I commit some stuff
20:24:19Paul_The_Nerdkkurbjun: The button definitions *may* not work, as apparently you need to use button_hold() instead of button_get() or something like that...
20:24:57preglowlinuxstb: i'm wondering if the clickwheel driver perhaps shouldn't insert an event if there already is one in the queue
20:25:12linuxstbIt's worth testing.
20:25:13preglowlinuxstb: it's a bit annoying have a thousand of them building up when the cpu is busy
20:25:59 Join Lear [0] (
20:26:05Paul_The_NerdI had to hard-reboot to get out of about 5-million zoom ins on the mandelbrot plugin when I brushed against the scroll wheel too quickly.
20:26:20ep0chsafetydan: what kind of button acceleration are you thinking of using? a button combo or time based?
20:26:34linuxstbI'm currently being driven mad by the ipod audio driver. I've got it working perfectly with lostlogic's reworked pcm routines - apart from the fact that the output is silence...
20:26:45preglowyeah, saw that
20:26:48preglowsounds like a bunch of fun
20:26:52linuxstbStill haven't fixed it...
20:27:24linuxstbBut when audio reappears, all the ipod playback glitches should be gone.
20:27:50safetydanep0ch, probably both
20:29:02 Quit webguest08 ("CGI:IRC (EOF)")
20:30:36 Quit thegeek (Read error: 104 (Connection reset by peer))
20:31:08linuxstbsafetydan: If you switch the eq to use the scroll events on the ipod, you won't need any accelleration. I assume that's what you're talking about.
20:31:30 Quit damaki__ (Read error: 110 (Connection timed out))
20:31:31 Join thegeek [0] (
20:31:39 Quit thegeek (Remote closed the connection)
20:31:43 Join thegeek [0] (
20:31:55safetydanlinuxstb, also need acceleration for other targets though
20:32:08linuxstbOf course - I'm just saying not to accellerate the ipod.
20:32:29preglowamiconn: actually seems like jpeg.c has some chroma tables already
20:32:40 Quit thegeek (Client Quit)
20:32:43 Join thegeek [0] (
20:33:01safetydanlinuxstb, ah okay, scrolling with the wheel is fast enough then?
20:33:05safetydanhard to tell in the sim
20:33:40linuxstbI think it will be, yes. It's not very well simulated.
20:34:02preglowamiconn: don't think this should be too hard
20:34:24preglowi think we'll need acceleration for the wheel as well
20:34:28preglowbut we'll see
20:34:33 Join Thus0 [0] (
20:36:36safetydanAre the iAudio X5 button definitions safe to use?
20:38:35*linuxstb has sound again :)
20:38:46preglow\o/ _o/
20:39:34 Join damaki_ [0] (
20:40:00linuxstbAnd it seems glitch-free as well. Where's lostlogic to commit his patch?
20:40:01 Join Bger [0] (n=Bager@
20:40:02*Paul_The_Nerd cheers.
20:40:12Paul_The_Nerdpreglow: Sound dance?
20:40:14preglowlinuxstb: skipping bugs and shit gone?
20:40:21preglowPaul_The_Nerd: deed
20:40:40linuxstbSeems that way. I've completely rewritten the pcm_playback.c code for the ipod.
20:40:43 Quit markun (Read error: 54 (Connection reset by peer))
20:41:34linuxstbBut it's based on lostlogic's changes - so he needs to commit first.
20:42:07linuxstbIf I mention lostlogic enough times, will he come?
20:43:04 Join CrookedRain [0] (
20:43:21preglowyeah, lostlogic'll come then, surely
20:43:40preglowi'm just counting on his computer highlighting like mad, lostlogic
20:44:09lostlogicpreglow: ?
20:44:11 Join damaki__ [0] (
20:44:14linuxstbIt worked!
20:44:14CrookedRainis it possible to set the backlight to turn off after a little bit in bootloader usb mode on h300?
20:44:15lostlogicI need to commit?
20:44:19preglowlinuxstb: wants to :-)
20:44:25preglowyes, that worked out
20:44:26preglowdisregard me
20:44:40lostlogickk, gimme 5 ticks
20:44:47linuxstblostlogic: Yes, all ipod problems with your new pcm routines seem to be fixed now.
20:44:52XavierGrpreglow I see you are looking to expand the jpeg viewer. Could you take a glance at my jpeg patch first? Because as I understand it the code will be seriousl changed in some parts. Just give it a look and tell me your opinion. This patch also adds viewing pics while playing music (if the buffer can fit it)
20:45:03lostlogiclinuxstb: awesome work, I look forward to seeing your changes too :)
20:45:38perldivercan somebody post a link to a new GUI design, it was white/black/yellow'ish?
20:45:49preglowXavierGr: i'll see, will be a while before i do it anyway
20:46:37XavierGrI was looking ways to expand the jpeg plugin, but right when I started my H300 broke and I send it to iriver.
20:46:49XavierGrAs you said the chroma tables are there.
20:46:58XavierGrWe must expand the decoder
20:47:14XavierGrand here is the link of the patch, I updated it today.
20:47:42preglowXavierGr: it extends the plugin api, yes?
20:48:09XavierGryes because in order to optimise the directory routines I need the pointer
20:48:13preglowdo you know what the other devs think of that?
20:48:14XavierGrof the dir buffer
20:48:46preglowdoes it add much to the core of rockbox?
20:48:49ep0chlostlogic: i think slasheri wants to make changes too if he hasn't done so to cvs
20:48:56preglowep0ch: he has
20:49:06XavierGrpreglow only 1 function is added to the api
20:49:27lostlogicbagh, conflicts!/
20:49:37preglowto be expected, yes, heh
20:50:29linuxstbXavierGr: Does the fact the plugin size has been reduced to 512KB affect your patch?
20:50:33XavierGrpreglow: This patch was created 6 months ago. In this interval I changed the approach 3 times. Adding 1 function to the api make things more easy, less clattered and faster.
20:51:02*lostlogic eats slasheri's brain
20:51:12XavierGras long as the jpeg will fit in the buffer it will load it there, else a pop up appears if the user wants to stop music
20:51:21lostlogicI wish he could have coordinated that with me since he knew I had refactored most of pcmbuf.c :(:(:(
20:51:28preglowanyone who's got any shorten files want to commit the new shorten patch?
20:51:38XavierGr(the plugin buffer of course is not used when there is no playback)
20:51:40preglowXavierGr: ok, i'll have a look at it
20:52:06preglowlostlogic: you didn't answer him when he asked for you :>
20:52:12XavierGrok thank you very much
20:52:39linuxstbpreglow: I can take care of shorten
20:52:47lostlogicah, I was at lunch
20:52:55 Quit damaki (Read error: 110 (Connection timed out))
20:53:15preglowXavierGr: it's not a very big patch, nice
20:53:26preglowlinuxstb: patch looks fine to me, i just can't test it
20:53:31Slasherilostlogic: sorry, those changes were only pretty minor..
20:53:37XavierGryou should see the first version. It was huge! I am glad that I did it again
20:53:38Papricalinuxstb, the backdrop needs little fix
20:53:41Slasherilostlogic: but please commit it as soon as you can :)
20:53:54Papricawhen you switch to wps without a backdrop
20:54:08Papricathe previous backdrop stays..
20:54:09linuxstbPaprica: I know, but it's not easy - unless we don't allow the main backdrop to be displayed in the WPS.
20:54:29Papricai try to do somthing with it
20:54:38lostlogicSlasheri: yeah, unfortunately, major conflictiness, because I'd significantly changed how chunks are sized and stuff as well
20:54:45lostlogicI'll get it figured out shortly
20:54:57Slasherilostlogic: ah.. :/
20:55:15linuxstbPaprica: Please do.
20:55:24Slasherilostlogic: in fact i have had in mind to remove the chunks entirely and just create one very long chunk..
20:55:36Slasherii could look into that after you commit your changes
20:55:57lostlogicSlasheri: we need the chunks to feed them to the DMA, I'd think...
20:56:13Slasherilostlogic: we can feed only one chunk to the dma
20:56:26Slasheriand after buffer wraps, then feed it again..
20:57:17Slasheribut indeed, that would create some problems..
20:57:18XavierGrpreglow, note: The patch changes the names of 1 function in the dbinterface.c and .h. Because there is also a tree_get_context in there. So I renamed it to compile cleanly. (They shouldn't have the same name in the first place, they are different functions)
20:58:07linuxstblostlogic: I don't know if it's your changes or mine, but ALAC seems to be much happier now - it's realtime even when the disk is filling.
20:58:08XavierGrah no it is not the tree_get_context it is the struct entry which is declared twice.
20:58:08 Quit damaki_ (Connection timed out)
20:58:30preglowlinuxstb: expected that
20:58:37lostlogiclinuxstb: cool
20:58:55 Quit _FireFly_ ("Leaving")
21:01:17lostlogicSlasheri: your comment : FIXME: Crashes unless we use chunk size here, what would you like that value to be in seconds −− it won't crash now no matter the size
21:02:19 Join actionshrimp [0] (
21:02:26preglowXavierGr: is it really necessary to increment the plugin api version?
21:02:54XavierGrwell I am not sure? Linus told me to do so if I add a function in the end
21:03:05XavierGrOr I misunderstand?
21:03:08preglowi thought you only needed to do that if you added it in the middle
21:03:22XavierGrthat's the mini api version, no?
21:04:00preglowok, might be
21:04:05preglowXavierGr: and what is the JPG_LIST define for?
21:04:46XavierGryeah sorry that's a left over from the 2nd aprroach
21:05:01XavierGrit used a file to store the filenames temporary back then.
21:05:13Bgerpreglow the idea is to check whether the api is the same as when the plugin is compiled and if it's lower, then to reject running
21:05:23Bgers/api/api ver
21:05:34preglowi know that
21:05:38Bgerokay, sorry for my english, but you know it ...
21:05:58preglowpoint is just that adding functions at the end doesn't break anything
21:06:25XavierGrbut imagine that my compiled rock to be in an older compiled rockbox version
21:06:30Bgerbut if you compile it with api ver 53 and run it on api ver 51 ...
21:06:31XavierGrit would crash no?
21:06:44Bgerthe api's at the end won't be there
21:07:52preglowalso, what's going on with the "gray buf error" line in the patch?
21:07:59Paul_The_Nerdpreglow: The comments say increase API version when you add something, min version when you add in the middle, if I recall
21:08:04preglowseems like you replace the quotes with something strange
21:08:16***Saving seen data "./dancer.seen"
21:08:55preglow352 in the patch
21:09:46XavierGrah yes I changed all splash error intervals to 0.5 secs instead of 2 secs. It is annoying to have to wait 2 secs imho?
21:10:06XavierGrbut that is my opinion if you think it must be longer we should change it.
21:10:54Learbetter be able to read the splash at least...
21:12:25XavierGrwell the splash is a very small message in my tests 0.5 is plenty of time. It is annoying to wait 2 seconds, especially on occassion with many non-compatible files
21:12:59 Join Daishi [0] (
21:13:12Papricalinuxstb, just add the line
21:13:24Papricain line 837
21:13:30Papricaat gwps.c
21:13:43Learpreglow: you mentioned something about eq graphing. does that mean it would show the frequency response of the eq? you know of any (windows or linux) programs for that?
21:14:03Papricaand the backdrop will not stay..
21:15:17Papricathe menus back drop stay
21:16:34preglowLear: matlab?
21:16:43preglowLear: and yes, it means that
21:17:08Paul_The_NerdPaprica: If the WPS doesn't have a backdrop set, it's *supposed* to show the menu backdrop instead.
21:17:10LearI was hoping for something smaller/easier... :)
21:17:30Papricaso it works
21:17:32Paul_The_NerdThe idea being that if you want the WPS to have a blank screen behind it, you use a blank BMP. Otherwise, you can create a WPS with NO bmp, and have it match your menus
21:17:38preglowLear: there are some filter design programs around, but i haven't used them since i got my hands on matlab
21:17:51LearHm... maybe gnuplot would work...
21:17:58preglowXavierGr: i think 0.5s is a bit short, the user should be able to see them
21:18:01preglowLear: it would
21:18:15preglowLear: i've used octave/gnuplot for making filter plots
21:18:16preglowworks fine
21:18:23LearBut if it could show the frequency response of an x-band parametric equalizer, and allow me to play with the parameters, I'd be happy.
21:18:43safetydanpreglow, is your graph code usable (i.e. could I work on integrating it?), or is too rough for that?
21:18:57preglowsafetydan: if you give me a sec to convert it to fixed point, we'll see
21:19:06preglowit's floating point right now
21:19:10XavierGrpreglow: okay but I think 2 seconds is too much what about 1?
21:19:17preglowXavierGr: better
21:19:40safetydanpreglow, no rush
21:21:51 Quit gromit` (Remote closed the connection)
21:24:45 Join ModernExecutive [0] (
21:24:48Learheh, nice daily build cross... :)
21:24:50safetydanany objections to a %me WPS tag for "is EQ enabled?"
21:25:28 Join adamd [0] (
21:25:38preglowi don't see the need
21:25:45 Join gromit-- [0] (
21:25:46preglowwouldn't you know if it is enabled?
21:25:53 Join NightCat [0] (
21:25:56 Join markun [0] (
21:26:00NightCatHi all!
21:26:05NightCatPaprica here?
21:26:17adamdhello hello. am I right in thinking that the usb boot loader mode on the iRiver is there to allow access to the drive? because my PC doesn't pick anything up
21:26:38NightCatDo you know, that there is a problems with WPS?
21:26:51Papricawhat is the problem?
21:26:57safetydanpreglow, it's a feature request from the forums, I'll leave it for the moment
21:26:59lostlogicok, gotta test to make sure nothing broke from combining Slasheri's and my changes, and then I'll commit
21:27:08NightCatPaprica: But I don't know is it from margins patch or not...
21:27:16Paul_The_Nerdsafetydan: How 'bout storing the name of the last loaded EQ preset file (and if the EQs are tinkered with, replace that string with 'Custom') and use that in the tag, if you're gonna do a tag.
21:27:39NightCatPaprica: When I use WPS with tag %wd it hides status bar...
21:27:48NightCatPaprica: but
21:28:02NightCatPaprica: it writes strings like there is a status bar...
21:28:18Papricagive me the wps..
21:28:26safetydanPaul_The_Nerd, I'll put that on the TODO list
21:28:50Papricathe wps
21:28:54preglowi still don't see the reason, wouldn't you bloody know what you've set your eq to?
21:29:14Paul_The_Nerdsafetydan: No rush, it just seems like a better way to me. ;)
21:29:17safetydanpeople are odd
21:29:19preglowi don't want to add every bell and whistle to the wps code just because the misticriver people likes crowded wps screens
21:29:26Learsmall changes aren't _that_ audible, maybe?
21:29:28safetydanif it can be customised, they want it to be customised
21:29:54lostlogicSlasheri: do we really need to hold the buffer all but 1/2s full for crossfade!?
21:30:53NightCatPaprica: And I found another bug: when you use a change color patch when you plays song it becoms the default colors...
21:31:00NightCatHere the:
21:31:25Papricathe color patch isnt mine
21:31:33NightCatPaprica: I've don't change my WPS and it was working erlier...
21:31:45safetydanheh, button acceleration + low latency eq == fun with sound
21:31:46*Paul_The_Nerd still uses the Rockbox_Default WPS. With the peakmeter stripped.
21:31:47Slasherilostlogic: hmm, maybe not.. you can try to remove that delay :) it's only there to make sure we don't miss start of the next song
21:31:49NightCatPaprica: I know (about colors)...
21:32:36preglowsafetydan: oooooohh
21:32:40preglowsafetydan: gimmegimmegimmegimme
21:32:42lostlogicSlasheri: nod, I think I've got it working, will be committing shortly
21:32:45NightCatPaprica: But today I update cvs and patched it with margings pathc (new version from tracker)...
21:32:56Slasherilostlogic: sounds great :)
21:33:37Papricamm i know that the patch is buggy
21:33:57NightCatYou think it is a patch problem?
21:34:11Papricamm i think
21:34:16Papricai need to work on it
21:34:20NightCatBut erlier it was work...
21:34:25Papricamm try without the margins
21:34:36NightCatI tried - the same...
21:34:47NightCatOr you mean without patch at all?
21:34:49Papricaand with clear source?
21:35:08NightCatWith clear source - no...
21:35:16NightCatI'll do it now...
21:35:17 Quit Paul_The_Nerd ("Leaving.")
21:35:50safetydancurrently it's just key combo acceleration on iRiver (ON + LEFT or ON + RIGHT)
21:35:53safetydanbut good enough
21:36:24preglowso it just uses larger steps when you press on?
21:36:43safetydanpretty much
21:37:03safetydan100 Hz instead of 10Hz, 1dB instead of 0.1 dB, 1 Q instead of 0.1 Q
21:37:37preglowi think a scheme that just makes the step delta dependent on a timer delta would be better, so the step gets bigger the longer you keep a button pushed, up to a limit of course
21:37:44preglowi haven't tested such a scheme, of course :>
21:37:57safetydansee, that sounds like effort
21:37:59safetydanthis was easy
21:38:06preglowi can try it myself
21:38:15preglowbtw, this graphing won't be so expensive after all
21:38:29safetydanyou can swing from 10 Hz to 1000Hz in about 5 seconds
21:38:35safetydan10000 Hz I mean
21:38:37NightCatWho is running_amar?
21:38:44NightCatAnybody knows?
21:39:09lostlogiclinuxstb: I'm committed if you want to get your ipod fixage in
21:39:10safetydanpreglow, did you find a faster way?
21:39:23lostlogicSlasheri: might want to review what all I've just changed... but it's a lot, so luck :)
21:39:32NightCatColor patch is his...
21:39:40 Quit Siku ()
21:40:01linuxstblostlogic: Thanks. I'm just giving my ipod a short charge - I've worn it out
21:40:13NightCatI want working WPS and color changer...
21:40:50preglowsafetydan: no, i just stopped being braindead
21:41:01safetydanthat's always nice
21:43:09safetydanpreglow, button acceleration is committed
21:44:02 Join XavierGr_ [0] (
21:44:20XavierGr_sorry for that power outage for 5 minutews
21:44:51XavierGr_preglow if you commented something while I was away, please tell me....
21:45:20preglowXavierGr_: nope
21:45:31*preglow does the obligatory fast filter sweeps
21:48:06NightCatPaprica: I've tried: It works without your patch...
21:48:07ep0chhehe we need a couple of lfos ;)
21:48:31preglowthe first thing i did with the eq was giving it an lfo, haha
21:48:49Papricaok i'll take a look on that tomorrow
21:49:19linuxstbNew iPod audio driver in CVS. This one works :)
21:49:25linuxstb(I hope....)
21:49:36ep0chwahoo time to cvs up
21:49:56lostlogicep0ch: enjoy... the buidl system is like 4 behind so we'll never know what builds and what breaks
21:50:09NightCatPaprica: There is another patch, look:
21:50:16linuxstblostlogic: We all get to share each other's points...
21:50:36 Join JoeBorn [0] (
21:51:30PapricaNightCat, try it, its my patch
21:51:46Papricahe do it for the todays cvs
21:54:31 Quit Kohlrabi ("Fast alle Menschen sind Regenw├╝rmer")
21:54:40 Join Siku [0] (
21:55:13 Quit XavierGr (Read error: 110 (Connection timed out))
21:55:52 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
21:59:07 Nick XavierGr_ is now known as XavierGr (
21:59:30DBUGEnqueued KICK XavierGr
21:59:30XavierGrDinner Time!
21:59:44CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
21:59:44*preglow cheers for lostlogic
22:00:09safetydanseems that the low latency pcm function won't be called if you access the EQ menu from the onplay menu
22:00:40preglowfeel free to fix
22:01:19ep0chjust loaded my eq preset from disc while sound was playing. playback stopped?
22:02:07safetydanpreglow, just looking at the best way to do that
22:02:18lostlogicep0ch: sounds broken :)
22:02:20safetydandoesn't seem right to go to low latency mode for the whole onplay menu
22:02:33t0maslostlogic: already fixed
22:02:44t0mas"iPod: Re-written audio driver. This brings it in line with changes to pcm playback system, and also appears to fix all the ipod-specific playback glitches"
22:02:45Slasheriep0ch: that might be possible because the pcm runs pretty low when low latency mode is active..
22:03:08Slasheriwe could probably increase the buffer size a bit
22:03:15Slasheribut now time to go sleep ->
22:03:19ep0chi vote eq as patch of the month
22:03:52ep0cheven though i wont use it :)
22:04:07safetydanthe eq is nothing
22:04:10safetydannow eq graphing
22:04:13safetydanthat's something
22:04:16 Join solexx_ [0] (
22:04:30safetydanwho doesn't love graphs on their mp3 player?
22:04:38ep0chrealtime eq graphing settings screen...
22:04:53preglowyeah, i think the realtime eq graphing might be realistic now
22:05:09preglowwe'll see
22:05:14Learepoch: loading a config file always stops playback at the moment.
22:05:25*safetydan contemplates "Super Graphical EQ UI" mode
22:05:32Learsome settings mess with buffer sizes and the like, so to be on the safe side...
22:07:46YouCeyEare there any preset eq in rockbox?
22:07:54YouCeyEfor iriver
22:08:50safetydanYouCeyE, you can download some
22:08:58YouCeyEwhere from?
22:09:47ghode|afk - somehere
22:09:49 Quit hshah (Read error: 104 (Connection reset by peer))
22:10:06safetydanbeaten to it
22:11:27NightCatPaprica: The second patch works fine. All's ok with marging now.
22:11:46Papricagood to know
22:12:16NightCatPaprica: Now need to solve the problem with colors...
22:12:25lostlogicman, we ate CVS today, high fives all around :)
22:13:18 Join RotAtoR [0] (
22:13:30PapricaNightCat, what do you need? foreground color or background?
22:15:32LearBtw, saw on hydrogen audio that 3GPP has the source for a fixed-point HE-AAC encoder/decoder. Don't know about licensing yet though...
22:15:46LearAnd encoding was quite limited in bit-rate range, I think...
22:16:30 Quit solexx (Connection timed out)
22:16:37NightCatPaprica: foreground
22:16:54preglowoh my, how little i care about aac encoding
22:17:08NightCatWhen go to WPS (start playing) colors backs to default.
22:17:28NightCatBut it works if go to menu and select it again...
22:17:36Papricawps tag is ok too?
22:17:44 Quit Moos (Read error: 104 (Connection reset by peer))
22:17:45NightCatThen they go default if again stop and play...
22:17:46 Join hshah [0] (
22:17:52 Join Moos [0] (
22:18:02NightCatPaprica: Wich tag?
22:18:28YouCeyEhi the eq are fox h3xx
22:18:28NightCatThere is such tag?
22:18:36YouCeyEi have h120
22:18:43YouCeyEwill they work?
22:18:47Papricai have the start
22:18:54Bagderthe eq is for rockbox
22:19:05NightCatI have no such tag in my wps...
22:19:21Papricai know
22:19:32Papricai will do it.. =]
22:20:36pregloweq works on all current rockbox platforms without hardware decoders
22:20:59NightCatPaprica: I just can't understand, why colors flashs when go to wps...
22:21:12Papricai dont know too
22:21:13lostlogicBagder: I remember you not much being a fan of how I'd done the real time clock tags for WPS... how do you think it should work?
22:21:44BagderI doubt I've said anything about it
22:21:47lostlogicI think so... it was a fairly long while ago before I had CVS, one of my early patches
22:21:51NightCatOh, for realtime clock: why it always flashs?
22:22:02lostlogicNightCat: flashes?
22:22:24NightCatMmm... Back to 00:00...
22:22:52NightCatI have to set it again very often...
22:23:37lostlogicNightCat: anytime the iRiver firmware boots up it will get messed... otherwise... I dunno
22:24:05NightCatAh... Mmm... Maybe... I'll check :)
22:24:46NightCatYes :) I know... I recharge batteryes every night...
22:24:54NightCatWith iriver firmware...
22:25:38NightCatThere is no recharge code i RB for H300 yet...
22:25:47lostlogicNightCat: huh? yes there is...
22:25:49NightCat*in RB*
22:25:52lostlogicNightCat: charging works fine
22:26:08NightCatWithout booting?
22:26:12 Join bazz [0] (
22:26:37lostlogicNightCat: why do you need it to be bootloader charging?
22:26:39lostlogicit doesn't matter...
22:26:56CassandraH300 has a hardware charge circuit, doesn't it?
22:27:01lostlogicCassandra: yeah
22:27:06NightCatI mean if I start it and stop playing it will shut down in some time yes?
22:27:15lostlogicNightCat: not if the charger is connected
22:27:25NightCatAh, ok.
22:27:44bazzis it a known bug that if you specify %?mp<Stop|Play|Pause|Ffwd|Rew> it will display Ffwd for everything other than play (pause,ffwd and rwd) except for one moment when you let go of the ffwd/rwd button when it displays rewind. (pause always displays Ffwd)
22:28:06 Nick Paprica is now known as Paprica[sleep] (
22:28:41lostlogicbazz: never seen that behavior
22:29:08bazzlostlogic: hrm, i see it both on my player and in my sim, but let me play a bit, it could be a bug in my wps file
22:29:13 Join damaki_ [0] (
22:33:43bazzlostlogic: nm, i'm an idiot :)
22:36:07 Quit hyarion ("Lost terminal")
22:36:53 Part ModernExecutive
22:36:56*amiconn is curious what 'super graphical' is ;)
22:37:09 Quit zhilik ("")
22:38:52 Quit Lear ("Chatzilla 0.9.70 [Firefox]")
22:38:57amiconnsafetydan: When adding strings to the .lang files, *always* add at the end.
22:39:02 Join BHSPitLappy [0] (
22:39:26amiconnLikewise, deprecated strings must not be removed, they need to be deprecated
22:39:38 Join hardeep [0] (i=hardeeps@SDF.LONESTAR.ORG)
22:40:09lostlogicamiconn: hmph, good to know.
22:40:13preglowwhy does this matter so much for lang files? we change other settings around often enoughj
22:40:24ep0chthis for voice files?
22:40:49Bagderlang files survive more
22:40:54amiconnIf strings are added in the middle, all subsequent strings will be mixed up if someone uses older .lng files
22:41:06Bagderand english is the master for all other languages
22:41:39ep0chthats a bold statement ;)
22:41:40preglowwhy do lang files survive more?
22:41:56Bagdersince we stick to these rules
22:41:58amiconnThe deprecated strings can only be removed and strings sorted when we bump the binary .lng version. This will render all older .lng files unusable
22:42:18preglowappreciate the recursive turn this just took
22:42:35Bagderwell, there's no reason to break old installs
22:42:40amiconnThis is something I am doing right now; I'm (still) about 85% thorugh english.lang
22:42:44Bagderand by adhering to these simple rules, we don't
22:43:03safetydanamiconn, sorry about that, I just remembered that restriction after committing
22:43:21amiconnThe other .lang files can be adapted by a script (which has to be written, but shouldn't be too difficult)
22:43:31*Cassandra realises quite how big this channel's got lately.
22:43:42preglowBagder: break old installs? these are development builds here...
22:43:53amiconnOoh, >100 users now...
22:43:54preglowanyway, it doesn't matter
22:43:59 Quit ep0ch ("Kopete 0.11.1 :")
22:44:28amiconnpreglow: The lanugage files are shared across all platforms
22:44:42 Quit damaki__ (Read error: 110 (Connection timed out))
22:44:46CassandraFrankly a prune of unused string's would be good. Less work for me when I record them.
22:45:05Bagder*cough* langv2 *cough*
22:45:14amiconnCassandra: All unused strings should be deprecated anyway
22:45:50amiconnI found 2 or 3 that were not, I cleaned them from my local tree already
22:46:13 Quit bazz ("Leaving")
22:46:25*preglow sings the fixed point dirge
22:46:25lostlogicHave people gone and recorded their own voice instead of using T2S?
22:46:35preglowcassandra has
22:47:32 Join Rondom [0] (
22:47:33amiconnpreglow: What?
22:48:23CassandraI'm due to do it again shortly, since I have much better quality equipment now.
22:48:34preglowCassandra: good luck
22:48:38preglowamiconn: what what?
22:48:50amiconnwhat fixed point thingy...
22:49:05preglowjust me lamenting about fixed point never working the first time around
22:49:11Henricodo we include sound in the manual?
22:49:13CassandraThanks. I have volunteers to do the editing job, so it's just a matter of setting up the mic and sitting down for an hour or so.
22:49:23*amiconn is wondering what preglow is doing
22:49:32BHSPitLappywhat's new in here? something always is :D
22:49:34amiconnResampler? Stereo width? Speex?
22:49:35safetydanamiconn, eq graph I think
22:49:44CassandraBehringer, providers of reasonably cheap decent quality audio equipment are my stars.
22:50:16CassandraHenrico, huh? This is for the voice UI.
22:50:16preglowamiconn: eq graphing
22:50:37Henricosure, but if we include screenshots, why not soundsamples?
22:50:43preglowthought i'd do something that would be quick
22:50:48preglowbut no, no quickie for me today
22:50:49BHSPitLappyeq graph would be cool
22:51:23BHSPitLappywould that look more like a cpu monitor or something?
22:51:27preglowspeex _still_ has tons of floating point code :/
22:51:28BHSPitLappyrather than a visualization
22:51:46preglowBHSPitLappy: what? it will just plot what the eq frequency response looks like
22:52:15preglowlike that
22:53:31CassandraWhat'd be really cool is something that takes a frequency response graph for some headphones and works out an EQ setting that gives as flat a response as possible on them.
22:53:50preglownot exactly a trivial problem
22:53:55preglowbut entirely possible
22:54:07BHSPitLappypreglow: kewl, would there be ways to throw that into a wps?
22:54:23preglowBHSPitLappy: i can't imagine why anyone would want that in a wps
22:54:30*preglow is against wps bloat
22:54:46*Cassandra imagines it'd be possible to get detailed frequency response data out of manufacturers, but I don't understand the maths well enough to write one.
22:54:49lostlogicspeaking of −− anyone heard from NathanH about his token parser?
22:55:03CassandraI'd love to have my EQ settings custom tuned to my headphones though.
22:55:30CassandraAnd that really would be a Rockbox unique feature.
22:55:57preglowCassandra: you should never trust frequency responsen acquired from manufacturers
22:56:01BHSPitLappypreglow: somebody will do it :)
22:56:06CassandraTBH all we really need is an app that will generate EQ files. Doesn't even need to run on target.
22:56:16preglowonce you have that, though, it's just a simple matter of making an inverse filter
22:56:25preglowwhich is only "trivial" if you allow it to be an fir filter
22:56:38preglowwhich would be kind of in the heavy league for a dap
22:56:52Cassandrapreglow, you have a point. But I'm not sure how else you'd get the data without professional audio testing equipment.
22:57:12Cassandrawhat's one of them then?
22:57:15 Join Febs [0] (
22:57:15preglowno, no you can't :-)
22:57:32preglowyou'd pretty much need a very high quality mic and an anechoic room
22:58:47 Join hyarion [0] (
22:58:58preglowan fir filter is one of the two basic kinds of digital filters you have
22:59:16preglowstands for 'finite impulse response', the current eq uses iir (infinite impulse response) filters
22:59:36preglowboth because that's what traditional eqs use, and because iir filters are much cheaper than fir filters
22:59:46preglownewer eqs are starting to use linear phase fir filters, though
23:00:39BHSPitLappyhyarion: boo
23:00:52 Part hyarion
23:01:32 Join hyarion [0] (
23:01:36hyarionis he gone?
23:01:45preglowi killed him
23:01:57josh_14:00 -!- hyarion [] has left #rockbox []
23:01:57josh_14:01 < BHSPitLap> lol
23:01:57josh_14:01 -!- hyarion [] has joined #rockbox
23:01:59t0mashm? are you alergic to cows?
23:02:10preglowcows say 'boo' around your place? :>
23:02:11josh_I don't see any "-!- BHSPitLappy [...] has left #rockbox [...]" in there...
23:02:18preglowi'd hate to go there
23:02:21CassandraSo we couldn't really use an FIR fillter for Rockbox then?
23:02:24FebsCassandra, if you need additional help with the editing project, I can pitch in.
23:02:40*BHSPitLappy shoots hyarion, in tears
23:02:51t0masghehe preglow: yeah... boo pronounced by dutch people sounds like boe... wich is similar to the way we describe the sound of cows
23:02:55preglowCassandra: could, yes, but i have yet to try implementing one to see how long filters we can use before everything just vomits
23:03:26*lostlogic keeps EQ safely in the off position
23:03:40CassandraFebs: Help always welcome. Patches to the TeX, screenshots, plugin docs, I'll take whatever's offered.
23:03:55t0maslostlogic: iriver?
23:03:58Cassandrapreglow, would it be worth doing, assuming it worked?
23:04:14lostlogict0mas: H3x0 −− I prefer performance, and original audio ;)
23:04:17t0masCassandra: are you sure you want flat sound?
23:04:22CassandraI really ought to look at the maths of audio manipulation at some point.
23:04:31t0masmost people use EQ's to pump up time freqs...
23:04:33FebsActually, I was talking about the audio editing with respect to the voice files.
23:04:40hyarionBHSPitLappy: it's not nice to shoot at people like that :/
23:04:41Cassandrat0mas, why wouldn't I? I tend to believe it's the artist's intent.
23:04:54FebsBut I DO want to work on the manual.
23:04:57BHSPitLappyhyarion: it's not nice to make people cry :'(
23:05:16CassandraFebs, oh right. Thanks. I'll keep you in mind.
23:05:23t0masCassandra: I assume the artist and his or her tech people listened to it on some setup too...
23:05:30t0masand tuned it to sound good on that
23:05:46preglowCassandra: sure, i'll do one once, but it's not a priority
23:06:01Cassandrat0mas, well, yes, but I'd assume producers try to get as flat a setup as they can.
23:06:04t0masso if you build the perfect "flat" player... you will not hear what they heard when mastering...
23:06:13t0masCassandra: no, they don't...
23:06:24t0masthey try to get the sound perfect for the greatest common divisor...
23:06:34FebsWhen I've mixed audio, we generally do the mixing on nearfield monitors, but we then listen to mixes on a small radio, headphones, and in several different cars.
23:06:36Cassandrapreglow: Cool. I'm in no great need of it.
23:06:36preglowthe greatest common divisor doesn't always hit you
23:06:39t0massometimes even just for radio...
23:06:42FebsAs well as on home systems.
23:06:43preglowfuck what they intended, listen to music like you want it
23:07:14preglowbesides, not all music is mastered by pros
23:07:19FebsThe mix ends up being a compromise. Things can sound great on the studios nearfields but complete crap on some other system.
23:07:29lostlogic*admits that pitch adjustment amuses him greatly*
23:07:30Cassandrapreglow, if you like stuff like prog, the artist's intentions can often be quite important.
23:07:35t0masFebs: and the finally masterd version is tuned to sound good on radio and home systems... and I assume now a days on ipod earbuds :)
23:07:57preglowyou can't have it sound good on everything
23:08:03preglowthere are too many nuances involved
23:08:07FebsWe had loonnngg discussions on my band's CD about how much limiting to apply.
23:08:15preglowlimiting :/
23:08:17***Saving seen data "./dancer.seen"
23:08:21CassandraPersonally I don't care long as they make it sound good on my Arcam Solo. ;)
23:08:21t0masfor cars you can't compress enough...
23:08:31t0masso if you're making a "car mix" cd... off you go
23:08:44t0masbut it ruines the dynamics on a high end home theatre system
23:09:11FebsYeah, I know,, preglow. We had a mix that had *great* dynamics, but listening in the car, you had to have your hand on the volume control constantly.
23:09:17adamdwhat exactly does compressing do? quieten the louder bits and raise the quieter bits?
23:09:32CassandraI'm not entirely convinced of the benefits of listening to music on a home theatre system.
23:09:38*amiconn doesn't completely understand why cars need audio compression
23:09:40Cassandraadamd, exactly.
23:09:40t0masadamd: jup
23:10:00CassandraIf you've ever tried to listen to Pink Floyd on a car system, you'll know why.
23:10:09FebsAmiconn, because the noise floor is so high. If you turn the music up loud enough to hear the quiet parts, the loud parts are overwhelming.
23:10:10t0masamiconn: because of the high amount of background noise
23:10:10amiconnThe car is practically the only environment where I can use higher volumes without disturbing others
23:10:24lostlogicanyone other than ep0ch done boost ratio comparisons before and after my commit today?
23:10:34t0mastime to watch CSI
23:10:36t0masbye bey
23:10:46preglowwatching csi makes your brain fall off
23:10:50 Nick t0mas is now known as Ts|Away (n=Tomas@unaffiliated/t0mas)
23:10:59*Ts|Away catches his brain
23:11:09Ts|Awayit's a great show to watch before you go to bed..
23:11:14BHSPitLappyusing iPods makes your ears fall off
23:11:17preglowhaha, probably
23:11:24adamdit's time for bands to start releasing music in a mix-it-yourself type format too aswell as standard CDs :P
23:11:36CassandraI thought it made you a fashion victim.
23:11:53BHSPitLappy2020: cancerous iPod tumors plague 65% of earth's population
23:11:55ghode|afkouch, ogg -7 with eq = 100% boost :/
23:12:03lostlogicadamd: they should release it on CD 96khz/32bit FLAC'd for us, right?
23:12:09CassandraWho except professional engineers and DJs can be arsed to mix dynamically.
23:12:19lostlogicghode|afk: do you also use crossfade?
23:12:21 Quit midkay (Read error: 110 (Connection timed out))
23:12:37 Quit Rondom ("I'm leaving on a jetplane, don't know when I'll be back again...")
23:12:46 Part safetydan ("Leaving")
23:12:52ghode|afkits off atm
23:12:53amiconnBHSPitLappy: I don't know about the iPod output, but the old archoses have higher max volume than the irivers. Not that I would need it; my usual listening volume (with earphones) is -40..-30 dB ...
23:13:14BHSPitLappyamiconn: it's not the volume, iPods just plain create tumors. it's as simple as that.
23:13:24*Cassandra is quite surprised by how much louder the Nano is than the iRiver.
23:13:32BHSPitLappyjiggly, apple-shaped tumors.
23:13:35ghode|afkwhat does "pcmbufdesc" mean?
23:13:44amiconnCassandra: Interesting...
23:13:50preglowamiconn: ipods have up to +6db
23:14:06lostlogicghode|afk: I should give that a better label
23:14:10preglowit does clip rather badly, though
23:14:12BHSPitLappyit means "preglow comes mondays but understands fully disk erasion so cool"
23:14:20*Cassandra listens to the Nano set about 1/2 way up the volume scale, and the iRiver at about 70%-80%.
23:14:24ghode|afklostlogic: hovering at around 68% now
23:14:26Cassandra(This is with replaygain.)
23:14:37Febsadamd, the typical user can't even figure out how to use a 5 band parametric EQ. There is NO WAY most people could do their own mixes! ;)
23:14:41amiconnpreglow: If you push the archos to +12 dB it will clip as well.
23:14:48lostlogicghode|afk: that is how many chunks of the buffer are used (since it now has variable sized chunks, I needed to monitor how many were needed so I added that to the debug)
23:14:53adamdwell I know it sounds stupid and I know it's nigh impossible but I'm sure it'd be excellent fun for there to be a release of maybe a hd-dvd full of all the different tracks... for home-mixing. now that would be fun.
23:14:57amiconnBut even at 0 dB it's louder than the iriver, for whatever reason
23:15:07lostlogichmm, I'm using EQ on my test track @59-60% boost
23:15:18lostlogicwhich is 14% higher than w/o EQ
23:15:27*Cassandra hasn't played with EQ yet.
23:15:28adamdnaw, not for everyone :P but I for one would love to do some kind of surround sound mix experimentation and I know a lot of others would. it'll never happen, but it'd be fun if it did.
23:15:32FebsMixing is an art. It would be like giving someone some cans of paint and telling them to make their own Mona Lisa.
23:16:02CassandraLike most arts, the majority of people who do it are shite.
23:16:03Bagderdevcon2006 getting mentioned on the neuros list ;-)
23:16:27CassandraDamn - must book tickets and get organised.
23:16:53adamdI just think that for anyone who fancied having a go, either to start up learning and make a career out of it or just for a bit of fun, there's no way for them to do it unless they have access to a band to record or something
23:17:04lostlogicI gotta make sure I can get a passport in time and then book tickets.
23:17:21adamdI had a bit of a surround sound craze a while ago and got bored quickly because I had nothing to play around with myself
23:17:45preglowsurround mixing sure as hell isn't easy
23:17:59preglowordinary mixing requires tons of mixing to get right
23:18:08CassandraThing is, if you think about it, surround is kind of an odd idea for music.
23:18:26CassandraI mean if you go to a gig, usually the sound is all coming from the same direction.
23:18:29preglowwell, yeah, but so was stereo ;)
23:18:35ender`still, practically everything produced in the last 10 years uses surround
23:18:45preglowender`: no?
23:18:46Cassandrapreglow, stereo makes more sense, I think.
23:19:00preglowCassandra: not all music is recorded at a gig
23:19:07ender`if nothing else, you'll hear the vocals on the center most of the time, while music plays on all speakers
23:19:48ender`then eg. during refrain, the vocals are expanded to all speakers (just connect your CD player to prologic decoder to hear this)
23:20:01adamdI think it's a fantastic idea for music. it means you're not having all the instruments mushed together in two tracks [which in most cases are nearly the same anyway, with minor stereo effects], they're more discrete, surely that means more of the information from the original multi-track master is delivered to the listener
23:20:16Cassandrapreglow, this is true - I'm just saying that surround seems a little contrived. I mean there's no form of art that involves sitting totally surrounded by musicians.
23:20:54Cassandraadamd, frankly, if you can't hear the instruments as discrete sounds, then your stereo isn't good enough.
23:20:59ender`Cassandra: at least to me, it feels better when the music is coming from all around, rather than just front
23:21:03preglowsure, no, but it's worth experimenting with
23:21:16FebsCassandra, there is if you ARE a musician!
23:21:31Febs(sitting around surrounded by musicians, that is)
23:22:13adamdlistening to a CD shouldn't be an imitation of a live gig, or "sitting surrounded by musicians". why can't it take on its own form? the feeling when you are completely immersed in the sound is, in my opinion, so much more moving than listening to anything in stereo
23:22:17adamdFebs: good point
23:22:28CassandraFebs: Well true.
23:22:52CassandraTBH the best way to enjoy a live musical performance is to take part in it.
23:23:03 Quit NightCat (Remote closed the connection)
23:23:25FebsWhether a CD should be an imitation of a live gig really depends on the type of music. Generally the goal of classical recordings is to recreate the soundstage of a concert hall.
23:23:25CassandraNow that'd be a use for the unmixed version I could relate to - cut out which ever track you wanted to perform. ;)
23:23:39preglowthey fail
23:23:49FebsOn the other hand, almost no rock recordings ever have a natural soundstage. The soundstage is created by the mixing engineer.
23:24:06ender`Febs: in that case, you can put just the echoes on the rear speakers, that'll be as close to the concert hall imitation as you can get
23:24:24preglownot all echoes come from the rera...
23:24:26preglowrear, too
23:24:26adamdI just reckon that people would have a much better opinion on surround sound music if they thought of it as a different way of listening, rather than comparing it back and forth to stereo mixes. I honestly believe it's a superior experience than listening on a stereo.
23:25:51FebsGetting back to the topic of designing EQ curves that make headphones completely flat, generally, unlike speakers, a completely flat response is not usually the goal of headphone manufacturers.
23:26:29lostlogicdo any of us have iPod 5gs −− didn't someone mention having a driver for the codec in the 5g that needed looking at?
23:27:22adamdbtw, my original question - should the usb boot loader mode make the iRiver hard drive available to windows? because if so, something's buggered here...
23:27:32preglowlostlogic: linuxstb has access to one
23:27:33adamd*usb boot loader mode
23:27:43linuxstbadamd: I think it's known that usb bootloader mode isn't perfect on the h300.
23:27:49lostlogicadamd: yes, it should
23:27:52adamdaha, okay
23:27:55lostlogicadamd: it does about 80% of the time for me
23:28:00CassandraFebs, I believe it was in the case of mine, but what's the usual goal?
23:28:06adamdit hasn't worked yet here
23:28:08XavierGryes I had problems with the USB bootloader on H300 too. (when I had an H300)
23:29:43adamdanother thing I thought I'd just ask about [not trying to be impatient.. just curious] is whether the channel selecting on the H300 is anywhere near being enabled? [swapping output from stereo to mono to "custom" etc]
23:30:01 Join midgey34 [0] (
23:31:18lostlogicok, seriously, what the hell is wrong with IP laws. "Any chat you have using AOL Instant Message is the property of AOL, you may not redistribute the results of any such chat"
23:31:36adamdthat's nuts
23:31:51lostlogicsorry, that was a satirical non-quote
23:32:03adamdlol okay
23:32:04lostlogicbut it's basically what AT&T's Natural Voices license amounts to
23:33:02lostlogicadamd: as for the channel selection, software codecs currently only have stereo far as I can see... you can crossfeed if you want to hear both channels in both ears
23:33:15adamdwhat do you mean by crossfeed
23:33:36adamdthat's the aim, it'd be useful for when sharing earphones with anybody
23:33:46lostlogicin playback settings, crossfeed mixes the channels with each other
23:33:51preglowcrossfeed wouldn't work for that
23:33:52adamdexcellent, thanks
23:34:11preglowit feeds so little from each channel into the other
23:34:26lostlogicah, my mistake.
23:34:41preglowi'm probably going to implement channel settings and stereo width soon
23:35:03adamdahh that'd be very useful
23:35:19lostlogicwill that require changes to the pcm handling code, or is the codec always stereo, we just mix and feed identical data?
23:35:19preglowman, if this grapher would just bloody start working _now_...
23:35:27preglowmix and feed, i'd say
23:35:50preglowi think dsp may process pure mono streams
23:36:05preglowi had to do a fix for pitch change screen when i assumed i always got two channels
23:37:27Cassandralostlogic, I don't believe if it went to court AT&T could enforce their license.
23:37:38CassandraI must ask my lawyer friend about it.
23:38:20*linuxstb discovers that it one year ago yesterday that the "iriver h1x0 bootloader flashing party" happened:
23:38:23CassandraThe problem is the fact that it's almost certainly an illegal contract term (at least under British law) and therefore unenforceable, the amount it'd cost to prove that would be huge.
23:38:57Bagder"our only recourse is to notify the AT&T attorneys of what we believe may be unauthorized use of their intellectual property."
23:38:59lostlogicCassandra: yeah :(
23:39:10Bagderis that a threat?
23:39:20preglowahh, what a day
23:39:24*preglow remembers the initial panic
23:39:31FebsWhere was that letter from? U.S.A? What state?
23:39:32preglowBagder: i believe it is
23:39:44CassandraBagder, more like intimidation, really.
23:39:57BagderFebs: "Wizzard Software Corporation", sounds US to me
23:40:24CassandraI still think it might be a good idea to ask the EFF and ORG what they think.
23:40:41preglow Pittsburgh, PA 15213
23:40:53CassandraSeems we should be making a noise about this.
23:41:04lostlogicI think we give Wizzard the finger, and talk to AT&T about it
23:41:06Cassandra"AT&T shafts blind people" is a great headline.
23:41:13FebsWhat a coincidence. I am licensed to practice law in PA.
23:41:28lostlogicif _they_ have a problem then we _might_ want to listen
23:42:08Cassandralostlogic, certainly worth a try, I'd say.
23:42:25pillthe red lines on daily builds page look like a christian cross upside down
23:42:34pillrockbox is evil
23:42:46CassandraOh no, Rockbox has hidden satanic messages.
23:42:55Bagderwe are cursed!
23:43:04preglowFebs: ok, so you've got this wrapped, then? :)
23:43:33 Part LinusN
23:43:55XavierGrlol indeed
23:44:35FebsWell, if Rockbox wants to challenge it, I may be able to help. I would need to run a conflict check at my firm to make sure that the company is not a client of my firm.
23:45:13 Join damaki__ [0] (
23:45:31lostlogicFebs: I guess the question is, whether the terms of the license are legal in the first place, and who therefore is legally in the right.
23:45:58CassandraFebs: That'd be cool.
23:47:49BagderI'd say "Rockbox" could have a hard time to act legally, since it isn't a formal organization or company
23:48:01Bagderjust a bunch of people spread out
23:48:01FebsVery true.
23:48:09FebsI remember discussing that once before.
23:48:42FebsOf course, it also makes it difficult for the licensor to enforce the license, if in fact it is enforceable.
23:49:06preglowoh, sweet moses
23:49:07preglowi am such a fool
23:50:49Cassandrapreglow, well, we didn't like to mention it, but since you have ... ;)
23:51:01 Quit Bger ("BitchX: try our lowfat flavor too!")
23:52:53preglowit's _always_ stupid bugs
23:53:20lostlogicpreglow: isn't that kinda like "It's always the last place I look"?
23:53:23 Join midkay [0] (n=midkay@
23:56:44preglowi'm even stupider than i thought
23:57:31preglowfinally got the bloody graphing code going in fixed point
23:58:17BHSPitLappydo you think the ipod is powerful enough to make visualizations?
23:58:30BHSPitLappy(not counting the 5G)

Previous day | Next day