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).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2005-12-08

00:00:21MoosHave a good night
00:02:00dpassen1night Lear, thanks for fixing it
00:03:00 Join TCK [0] (
00:06:30Bagder"SFV (Simple File Verification) is used for creation the CRC32 checksum files"
00:07:57 Quit Lear ("Chatzilla [Firefox 1.5/undefined]")
00:09:23amiconnmarkun: One odd thing I noticed: arabjoin.c is located directly in firmware/ (would expect it in firmware/common/ like unicode.c), and arabjoin.h is directly in firmware/ too (would belong in firmware/export/ or firmware/include/ depending on where arabjoin.c is located)
00:12:37 Quit muesli_- (Read error: 110 (Connection timed out))
00:16:40 Quit Rick (Read error: 104 (Connection reset by peer))
00:17:25 Join Rick [0] (
00:18:03 Quit |Bger| (Read error: 110 (Connection timed out))
00:19:16 Quit dpassen1 ()
00:20:53 Join Cassandra- [0] (
00:23:34markunamiconn: yes, you are probably right.
00:24:20peturI just added a patch (1375787) that eleminates cpu hogging while keeping smooth playback on the win32sim. It replaces patch 1373083 which should be removed.
00:24:34peturIt's a small patch, if anybody has time...
00:32:07 Quit ender1 (Read error: 104 (Connection reset by peer))
00:32:10peturOh well, it's there for anybody to use. I'm off... goodnight!
00:32:30 Quit petur ("here today, gone tomorrow")
00:32:33Midgey34petur: patch seems to work, the sim doesn't stutter as much now (still has a few stutters here and there but I am running a number of processes)
00:36:46XavierGrwhat do I have to change to make the sim to have audio again?
00:38:12 Join JonSenior [0] (
00:39:06 Quit Cassandra (Read error: 110 (Connection timed out))
00:39:07 Nick Cassandra- is now known as Cassandra (
00:44:04 Join San [0] (
00:48:32 Join sa0 [0] (
00:48:41sa0hi all
00:48:56sa0just a question about IRiver port ?
00:49:02sa0is it ok in here ?
00:50:41sa0i'm looking for themes other than those included (h100)
00:50:46sa0can someone help ?
00:52:34XavierGrforums maybe?
00:53:03XavierGrthere is no official place which all sort of themes reside AFAIK.
00:54:45Paul_The_Nerd"Themes" as a whole are a relatively new feature
00:55:16XavierGrsa0: You can find wpses in the wiki, but not themes.
00:55:24Paul_The_NerdThere's the WpsGallery in the wiki, where you can get individual WPSes, but most of the ones that are a whole theme are getting submitted and included it seems anyway.
00:56:25sa0ok thx
00:58:16 Join SimonB [0] (
00:58:34SimonBhi all
00:59:31 Quit xmixahlx ("blah blah blah")
01:00:33 Part JonSenior
01:01:08Midgey34XavierGr: enable it in autoconf.h
01:02:05SimonBbit of a silly question, if i may?
01:02:48SimonBim not much of a developer but would like to get involved, how could i contribute other than reporting bugs etc that i encounter.
01:03:32Midgey34you could write WPSs
01:03:52Midgey34or help with documentation at the wiki
01:05:08SimonBahh cool, wps's are pretty well docymented, so its a definbate possibilty, plus documentation is something i could/would get invovlved with
01:05:19SimonBdocumented rather.
01:06:05SimonBsorry bout the spelling, ironic isnt it, interested in documenting but cant type a coherant sentance on irc, heh :/
01:06:46lostlogicx Profiling patches now available in two flavors. On to actually profiling Tremor and figuring out how to optimize it for me.
01:08:53Midgey34excellent work good luck
01:10:14XavierGrIs there any script that will make me search for tabs and replace them with spaces?
01:11:09***Saving seen data "./dancer.seen"
01:11:15 Quit sa0 ("( :: NoNameScript 3.81 :: )")
01:11:28XavierGrlostlogic: Any chance optimizing ilbmad?
01:12:04lostlogicxXavierGr: I made on device profiling patches... I'm only interested in vorbis personally, so it's up to others to use the profiling patches to optimize other codecs.
01:12:12 Quit tim66 (Read error: 113 (No route to host))
01:12:52XavierGrI am trying to optimize midi, but I am a louser optimizer.
01:13:19lostlogicxhehe, we'll see if I'm any good at this with Tremor... the profiling was just a couple of week distraction from failling to optimize it ;)
01:16:28 Part SimonB ("bbl")
01:17:46 Nick Lost-ash is now known as ashridah (n=ashridah@
01:19:00amiconnArgh, it seems the tagdb tools need to be adjusted to write utf8 strings into the db ...
01:20:06Paul_The_NerdAnother point on the list, then, 'eh?
01:20:26amiconnpreglow: Clicky resampler alert:
01:31:24 Nick ashridah is now known as Lost-ash (n=ashridah@
01:32:16 Quit Kohlriba (Read error: 104 (Connection reset by peer))
01:38:41MoosNight all
01:38:47 Quit Moos ("Glory to Rockbox")
01:49:19 Join Membrillo [0] (
01:50:20 Quit edx ()
01:50:56 Quit Membrillo (Client Quit)
01:51:52markun needs some updating :)
01:59:28 Join max [0] (
01:59:52 Quit Paul_The_Nerd (Excess Flood)
02:03:00 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
02:16:51 Quit Rick (Read error: 104 (Connection reset by peer))
02:22:07lostlogicHmm, can combine nm output and .map files to locate more symbols.
02:35:54XavierGrokay battery_bench plugin is now updated to latest unicode changes. If anyone is interested:
02:49:28 Quit YouCeyE (Remote closed the connection)
02:54:27 Part DJDD_
02:56:59 Join EoS [0] (
03:01:29 Join Rick [0] (
03:02:33 Join YouCeyE [0] (
03:08:11EoShello anyone there?
03:08:34Lost-ashwhat's up?
03:08:36 Nick Lost-ash is now known as ashridah (n=ashridah@
03:11:10***Saving seen data "./dancer.seen"
03:11:33EoSi was wondering, is it safe to install the iriver rockbox port?
03:14:38ashridahdepends how you define safe. it's generally been fine for people, but i wouldn't go so far as to say it's perfectly safe.
03:14:42ashridahwhich model iriver?
03:15:27EoSwell can it kill my unit?
03:15:31ashridahah, that works fine. i've got one, had rockbox on it for ages.
03:15:52ashridahthe bootloader's been working for almost a year now for the H1xx, it's fairly safe.
03:16:05ashridahstill no garuntees yet, but i doubt they'd ever really be in the position to offer one
03:16:05EoSoh goody, i loved rockbox on archos, been waiting for it to come out on irirver for so long
03:16:19ashridahheh, you could have joined the party on the H1xx months ago :)
03:16:41ashridahsome of the features rockbox adds for the iriver players are much nicer than the stock firmware
03:17:50ashridahbut the install process is fairly painless. you just need an unmodified ihp_120.hex
03:17:59EoSyeah i got a guide
03:18:42ashridahthe daily builds have just had a fairly major patch go in, so there's still some cleanup from the unicode stuff
03:18:59ashridahbut that's only affecting trivial stuff, and is rapidly being squashed.
03:19:14ashridahmostly a few unfixed plugins
03:21:17EoSyay, rockbox up and running :)
03:21:33ashridahhope you like the features it adds to the iriver units
03:21:54ashridahthey're certainly more capable than the archos platforms
03:22:00EoSim pretty sure i will, i remeber how much it enhanced the old jukebox6000
03:26:04EoSok thx for the info and thx for making this fine piece of software!
03:26:27 Quit EoS ("Peace and Protection 4.22")
03:28:28 Nick ashridah is now known as Lost-ash (n=ashridah@
03:41:27 Join amiconn_ [0] (
03:54:44 Quit hardeep ("[BX] Man I'm *SLEEPY*!!!! My keyboard is slipping away!")
03:54:56 Join yngwi [0] (
03:55:16 Quit YouCeyE (Remote closed the connection)
03:55:46 Quit yngwi (Client Quit)
03:55:56 Quit DreamTactix291 (Read error: 110 (Connection timed out))
03:57:30 Join YouCeyE [0] (
03:59:58 Quit amiconn (Read error: 110 (Connection timed out))
03:59:58 Nick amiconn_ is now known as amiconn (
04:13:58 Join Jungti1234 [0] (n=jungti12@
04:16:29 Quit Jungti1234 (Client Quit)
04:18:09 Part YouCeyE ("Leaving")
04:33:41 Part max
04:37:00 Quit Dima202 (Read error: 104 (Connection reset by peer))
04:40:57 Join Noel_sad_song [0] (
04:48:27 Join YouCeyE [0] (
04:52:06 Quit Noel_sad_song ("CGI:IRC (EOF)")
05:03:18 Join Rob2222_ [0] (
05:11:12***Saving seen data "./dancer.seen"
05:18:39 Quit Rob2222 (Read error: 110 (Connection timed out))
05:20:11lostlogicmuch progress in symbol naming is made.
05:26:26lostlogicnow all I'm missing is inlines, I think.
05:28:47lostlogicmaybe it's statics that are the problem
05:29:00lostlogicwhatever, 468 lines of Java to parse as much as I'm currently getting.
05:34:46 Join DreamTactix291 [0] (
05:40:34 Join Symbob [0] (n=Symbob@
05:43:43 Part Symbob ("bye")
06:02:26lostlogicwoot! it was a bug! Profile parsing works, all symbols accounted for!
06:35:14 Quit Rob2222_ (Read error: 104 (Connection reset by peer))
06:35:53 Join Rob2222 [0] (
06:37:50lostlogicmoving the most called functions to IRAM does not help. caching must already be better
06:53:32 Join Mrono [0] (
06:54:11Mronoanyone here
06:57:10Mronooo kay
07:00:52 Part Midgey34
07:11:14***Saving seen data "./dancer.seen"
07:12:26*Mrono is away (auto-away after 15mins of inactivity) - (11:12 pm)
07:32:22 Quit YouCeyE ("Leaving")
07:51:58 Join Jungti1234 [0] (n=jungti12@
07:53:03*Mrono is Back from (auto-away after 15mins of inactivity) - (Gone: 40mins 36secs)
07:54:03Slasheri(but please disable that auto away message thing)
08:07:09 Join ender` [0] (i=ychat@
08:08:33*Mrono is away (auto-away after 15mins of inactivity) - (12:08 am)
08:12:40 Join Coldtoast [0] (
08:13:06Coldtoastanybody else have a huge pause problem?
08:13:50Coldtoastwhen I pause and it fades out, instead of a smooth fade, the audio fluctuates a few times between loud/soft
08:14:30Coldtoastthen when I unpause, it just constantly pulses between loud/soft and the player becomes unresponsive
08:14:49Jungti1234Which player?
08:15:10Coldtoastjust disabled "Fade on pause/stop" and it's fine
08:15:25Jungti1234I will check.
08:15:25preglowmy, i'm tired
08:15:35 Join elinenbe_ [0] (
08:15:46Coldtoastand it crossfades between tracks perfectly normally
08:16:36 Quit elinenbe (Read error: 104 (Connection reset by peer))
08:16:53 Join Tigren [0] (
08:17:14Tigrenso is the rockbox firmware stable to install on an iriver h340?:
08:17:33Bagderdefine stable
08:17:39Coldtoasthere's something else I just noticed
08:18:03TigrenBagder: useable without crashing... and wont render it unusable ie killing the hd
08:18:22BagderTigren: yes, then I think we can say it is stable
08:18:24 Nick elinenbe_ is now known as elinenbe (
08:18:39ColdtoastI re-enabled Fade on Pause/Stop, paused, it did the volume flutter, I managed to get it to power off and when I powered back on, the Fade on Pause/Stop setting was disabled
08:18:49TigrenBagder: what about it would make you hesitate to say it is stable?
08:19:21BagderTigren: only the lack of long time testing and the small amount of developers devoted to h3x0-specific development
08:19:43Bagderit has a cap on cpu freq, it just got usb support added
08:19:47Tigrenwould you suggest installing rockbox or the new firmware at this point?
08:20:07Bagder*I* would not hesitate to use rockbox
08:20:14Coldtoastyou can still boot the irivr formware you know
08:20:17Bagderbut then I am a notorious rockbox user
08:20:29Tigreni used rockbox when i had an archos jukebox
08:20:43Tigrenand swore by it
08:20:45BagderTigren: aaah, then you'll feel like coming home again! ;-)
08:20:48Coldtoastinstalling the bootloader gives you a safer player Tigren
08:20:56 Join elinenbe_ [0] (
08:21:00Coldtoastcos if the iriver fw bricks your player, you can most likely recover
08:21:18TigrenBagder: it seemed like in all the pictures, only half of the screen was being used... whats up with that?
08:21:39BagderTigren: lots of stuff is not yet adjusted to the larger screen size of the h3x0 series
08:21:43 Quit elinenbe (Read error: 104 (Connection reset by peer))
08:21:43 Nick elinenbe_ is now known as elinenbe (
08:22:00Tigrenso right now only part of the screen is useable?
08:22:01Coldtoastyou can just create a wps to fix that tho, right?
08:22:25BagderTigren: all of the screen is used, just not all over as good as you'd like it to
08:22:37Tigrenoh i see, i can change that?
08:23:29Bagderyou can change the WPS to use the screen as you like
08:23:36Tigrenok cool
08:23:41Tigreni can make it pretty
08:23:57Bagdercheck out the WpsGallery wiki page for inspiration
08:24:44Coldtoastso anybody with a h100 getting this pause bug?
08:25:15Bagderhaven't used my iriver in days...
08:25:27Coldtoastwhy not?
08:25:34Tigrenthe daily builds are good enough to use?
08:25:42Tigrenooohh, pretty
08:25:56BagderTigren: they are bound to be the best builds more or less every day ;-)
08:26:15Tigren is the right download?
08:27:33Bagderthat URL is the bleeeding edge build
08:27:50Tigrenoh boy
08:28:01Bagderthat's the daily
08:28:09Jungti1234I surprised!
08:28:26Coldtoast is my wps
08:28:36Jungti1234it did the volume flutter
08:28:37Coldtoasth100 tho. heh
08:28:53preglowColdtoast: nice track
08:29:06Coldtoastyeah. I listen to Opeth when I'm at the gym
08:29:20Jungti1234Flutter continuously.
08:29:41Coldtoastyeah. hold STOP in til it shuts off Jungti1234
08:29:49Tigrenno color in those WPS?
08:30:02ColdtoastI have a h140. no colour
08:30:07Jungti1234It doesn't work normal.
08:30:23Tigrenoh i see
08:30:25Coldtoasthold STOP til your player shuts off
08:30:37Coldtoastthen disable "Fade on Pause/Stop" in the prefs
08:31:17Jungti1234But, H300 is normally.
08:33:13 Join Tigren_ [0] (
08:33:17 Quit Tigren (Read error: 104 (Connection reset by peer))
08:34:02 Nick Tigren_ is now known as tigren (
08:36:40tigreni can use the rockbox bootloader with the new us firmware right?
08:37:12BagderI don't know
08:37:24Jungti1234This bug loaf!
08:39:18Coldtoastbug loaf?
08:39:59tigrenhmm, i cant get to any of the new firmwares...
08:44:18Jungti1234 : See this.
08:44:59Jungti1234 <- Why is this included?
08:49:45preglowwhy is what included?
08:51:01 Join _FireFly_ [0] (
08:51:31BagderJungti1234: don't assume any non-ascii charsetss in IRC
08:51:38Bagderit won't work
08:53:50preglowman, i love all the shitty code that msvc eats, but i have to fix to not use msvc
08:55:02tigrenwhat are the .rwps files?
08:55:08preglowremote wps
08:55:19preglowthat is, a wps for your remote
08:55:22tigrenoh ok
08:55:34BagderLCD remotes (of course)
08:55:46tigrenyes yes
08:56:44preglowBagder: so, how's your new boss treatin' ya?
08:56:48preglowBagder: lemme guess, an asshole, right?
08:57:04Bagderhe's an evil bastard
08:57:17preglowforces you to sit around in your underwear?
08:57:28preglowthe bastard
08:57:49preglowwe usually shoot people like that in norway
08:57:53preglowor feed them to the polar bears
08:58:42tigrenso color hasn't been built into the 340 yet?
08:58:56Bagdertigren: yes it has, but not extensively all over
08:59:00preglowcolour support might be a bit off yet
08:59:02Bagderfeel free to send in patches
08:59:38Bagderlike for example the bmp loader doesn't do colors...
09:00:42tigrenso how do i boot the iriver firmware if i have the rockbox firmware installed?
09:01:03Bagdertigren: keep play and rec pressed when you boot
09:06:47 Quit tigren ("Client Exiting")
09:11:17***Saving seen data "./dancer.seen"
09:19:59 Join LinusN [0] (
09:20:19LinusNwoo, the fade-on-stop is a real eardrum killer!
09:22:44amiconnLinusN: me fix
09:22:44preglowa dap player os that kills eardrums == hardcore
09:22:55amiconnSee my ml post
09:23:00LinusNtinnitus, here i come
09:25:07 Join MrShlee [0] (
09:32:18*Mrono is Back from (auto-away after 15mins of inactivity) - (Gone: 1hr 23mins 45secs)
09:32:18Mronohey guys'
09:32:31MronoI was wondering
09:33:16Mronowhen will there be better support for h300
09:33:22Bagdersee topic
09:33:35Bagderyou can make it happen sooner by helping out
09:33:54Mronothat wasn't there when i came in earlier
09:34:02Bagderyes it was
09:34:06Bagderbeen there for many days ;-)
09:34:19Mronowell then i'm a dumbass
09:34:59MrShleeMrono.. Better Support?
09:35:18Mronoa less buggier build
09:35:21MrShleewhat are you talking about foo.. if you don't like something.. fix it
09:35:42Mronowell then
09:35:48Bagderthings actually don't fix themselves
09:36:54 Quit TCK ("I shall not cease!")
09:37:00 Quit Mrono ("The outside world is calling me")
09:37:12Bagderit seems most h3x0 users just wait for someone else to do the work
09:37:27Bagderand most Rockbox devs don't have any h3x0 players
09:37:50Bagderwhich might make the waiting long
09:37:58LinusNexactly, most h300 users are *users*
09:38:58amiconnLinusN: Your eardrums will be safe again with the next build. Sorry for that :/
09:39:11Bagder"VERY LOUD and painful bug" / MR
09:39:56Zagorperhaps now they'll stop complaining that rockbox isn't loud enough ;-)
09:40:30Bagderyou have to say that louder, most of the users are now almost-deaf :-)
09:43:33 Join Vladoman [0] (
09:48:55 Quit San (Read error: 110 (Connection timed out))
09:54:23*Bagder decided to leave the neuros google group
09:55:46Jungti1234a hahahaha
09:56:08 Quit tvelocity ("Leaving")
09:57:11*LinusN drives to bagder's place
09:57:17 Part LinusN
09:57:20*Bagder gets dressed
09:58:05Jungti1234Do you know who he is?
09:58:51Jungti1234 <- Master of picture inside.
09:59:29Rob2222Hya all. Does anybody know, why with every new bleeding build my settings are overwritten? Where are the settings stored?
09:59:40BagderJungti1234: sorry, I don't see why that is fun?
09:59:57BagderRob2222: they are not overwritten at all
09:59:57preglowman, i've got a d/a converter the size of a vhs player here
10:00:09BagderRob2222: and they are not reset with every build, only with some
10:00:23Bagderthe settings are stored in sector 61
10:00:31Jungti1234He is North Korea's the Minister of National Defense.
10:00:44 Quit Vlad0man (Read error: 110 (Connection timed out))
10:01:13preglowi like the grin on the face of the fellow behind him
10:01:17_FireFly_Rob2222 you can save your settings in a cfg-file
10:01:42Bagderif you use bleeding edge and care about your settings, you *really* should use cfg files
10:05:15 Join lamed [0] (
10:06:44lamedhey. any help with displaying hebrew on the new unicode codepage? -I guess we're not quite there yet, right?
10:07:43preglowi'd guess you just need a proper font
10:08:44markunlamed: what is the problem, there were other reports in the forum..
10:09:07Rob2222hmm ok
10:09:35markunlamed: does the language file work with one of the ?x? fonts?
10:10:23lamedmarkum: a lot! I can't display the filetree in hebrew, I can only see hebrew on the wps, it's not in the right direction...
10:10:51 Join KN|stiff [0] (i=phhome@
10:11:06lamedall hebrew fonts have been unicode compiled
10:11:07Jungti1234hi markun
10:11:34markunmorning Jungti1234 :)
10:11:42Jungti1234Unicode has some bugs.
10:12:01markunlamed: no, they haven't. Hebrew*.bdf are still iso-8859-8
10:12:07lamedmarkum - don't you have any problems with it?
10:12:55lamedwe are both right. only hebrew-8x13B is unicode
10:14:01 Join B4gder [0] (
10:14:09markuneven hebrew-8x13B is not hebrew. btw, it shouldn't be there anymore after my commit of yesterday.
10:14:18markun.. is not unicode :)
10:14:18Jungti1234 : markun, why is this included 'L'?
10:14:41Jungti1234(It reverses.)
10:15:24lamedmarkum: ?
10:15:26markunJungti1234: maybe a UTF-8 marker that was saved with your editor?
10:16:01markunlamed: yes
10:16:06_FireFly_markun markers should be ignored
10:16:10_FireFly_in wps
10:16:13markunof course
10:16:21Jungti1234I changed it by UTF-8 by notepad.
10:16:33markunlamed: so which font are you using now?
10:16:34_FireFly_Lear are you here ??
10:17:38_FireFly_markun there is already a fn which ignores the marker but it seams it doesn't work correctly or there is a place left who this fn should also be used
10:17:53_FireFly_or the marker is different as the current checked one
10:17:58markunJungti1234: please upload your WPS to me
10:19:10lamedgodamn! a unicode font (the first one afaik) with hebrew iso set on the default codepage: NO CHARECTERS what so ever on screen!
10:20:48markunvery strange.. you could remove the font in USB mode, that would bring you back to the default font.
10:22:20preglowdoes spdif out work in rockbox?
10:22:36markunDoes anyone else have problems with the unicode fonts?
10:23:00 Join uncledrax [0] (
10:23:37uncledraxwoot!! just recieved my h120!!! yay (i know im a bit behind the times :-) )
10:23:52markunI just tried the russian and hebrew language files with 6x13.fnt, works fine here.
10:24:27 Quit lamed ("CGI:IRC (Ping timeout)")
10:24:28uncledraxfinally i can playback all my musepack and wavpack files
10:25:02 Join lamed [0] (
10:25:28uncledraxi also recieved the h10 with it. Im going to look into a possible rockbox port
10:26:09B4gderh10 is portal player
10:26:28MrShleeTHe h10? sure it uses the same chip as the pod.. but memory might be a little low for it?
10:26:32B4gdera true challange
10:26:56uncledraxwell once i have a bit of a play, ill start looking into the tech
10:26:58MrShleeworth checking out.. I think its got sweet fa when it somes to ram
10:27:00preglowuncledrax: you might want to find some guy calling himself 'chaosd' (i think), he also had a look at the decryption function
10:27:15preglowi also had a look at it, and it didn't look very hard
10:27:18uncledraxpreglow: thanks will do
10:27:20markunB4gder: could you try test russian translation for me with one of the ?x? fonts?
10:27:32B4gdernot right now
10:27:36preglowit uses the pp5020, yes?
10:27:43preglowif so, you'll share a lot of code with the ipod port
10:27:53uncledraxi think it has the same as a ipod mini
10:28:43lamedI was just lucky enough to to succeed loading an older .cfg file. I think it might be a major problem that one might cause all characters to disapear.
10:29:05uncledraxgive me a couple of days to get over this infection i have and then ill be gun-ho
10:29:09lamedmarkum: what font and default codepage are you using?
10:29:25markun_FireFly_: jungti's WPS starts with char 0xfeff, should it have been ignored?
10:30:01markunlamed: 6x13.fnt, codepage latin1, hebrew.lng
10:30:09preglowcrashd, it was
10:30:16 Join Bagder2 [0] (
10:30:26 Quit Bagder2 (Read error: 104 (Connection reset by peer))
10:31:24 Join _`3nergy [0] (
10:32:17 Quit `3nergy (Read error: 104 (Connection reset by peer))
10:32:25 Quit Sando (Read error: 104 (Connection reset by peer))
10:33:06 Nick _`3nergy is now known as `3nergy (
10:33:51lamedhebrew.lng: do you mean the cvs-language file?
10:34:47lamedmarkum - yes you were on top of things. it works. weeee-ha
10:34:49markunit's markuN btw :) Just the hebrew language file where all the other language files are..
10:35:21_FireFly_markun i will look if
10:35:23markunlamed: files and translation are working?
10:35:26lamedYES! So I've translated it not _Just_ to myself!! Then I have a great reason to update it! Lol....
10:35:49_FireFly_markun: yes 0xfeff is also an unicode BOM
10:36:05markunlamed: try the WPS again..
10:36:14markunmight now work now
10:36:23markunDid you have some hebrew hard-coded in the WPS?
10:37:53lamedmarkun: everything is working properly with latin iso. but why is that? shouldn't the default codepage be set to hebrew? (that messes things up). btw i was talking about the language file translation-my work
10:39:36markunlamed: the codepage should only be to translate non-unicode id3 tags to unicode. In your case hebrew codepage should be no problem.
10:40:13lamedyes, but somehow it mess up the id3 tags.
10:40:54markunlamed: can you test if the russian language files also works for you?
10:41:15markunsomeone in the forum told me russian is not working, only with the unifont
10:42:36lamedWoops! acctually i'm not getting the id3 tags right on latin codepage.
10:43:01lamedand it _is_ working with hebrew
10:43:37markunlamed: any idea what the problem was? I suspect the people complaining in the forum have the same problem..
10:45:13lamedYes. I was hurrying. do You have to load a new file to get the id3 translated, right? - if you change codepage without changing files, it won't update the current file.
10:45:37markunyes, that's right. Also something to fix.
10:46:10markunlamed: could you post the tings you did here
10:46:11lamedAND, when you come to think about it, you have no idea what font to use now, as the **hebrew** font won't work, and nither most of the others
10:46:40markunlamed: I'm fixing the fonts little by little. Soon they should all work again.
10:47:49lamedmarkun: but i was trying many fonts that were uploaded by you that didn't worked. what fonts _should_ work?
10:48:02MrShleeI've noticed that fadein/out doesnt change the pause/stopped value right away until its finished.. not really a huge bug but would be good to change the status right away..
10:48:16lamedMrshlee it was fixed
10:48:33 Quit B4gder (Read error: 110 (Connection timed out))
10:48:40markunThe 5x7 (etc) fonts, elt16-unicode, unifont.
10:48:48lamedthat was because of the DB volume system change.
10:49:11markunthe other fonts only cover latin1 or don't work at all
10:49:51lamedgodamn. markum, i tryied elt16-unicode once again. I got no charcters apearing once again.
10:50:11markunlamed: yes, there was a bug in it. Should be fixed with my latest commit
10:52:00lamedah. It's very tierding to wait for a cvs update command & build & make & unzip all the time.
10:52:11lamed(although i have a script)
10:52:57lamedmarkum: then wha's the " 8 Dec 00:50Marcoenreplace or update some fonts by there unicode equivalents" activity?
10:54:29markunlamed: I replaced the iso8859, ast116-8 fonts with etl16-unicode and Hebrew-8x13B with 8x13B
10:54:55markunthe adobe fonts cour* and tim* where updated
10:56:18 Join PaulJ [0] (
11:02:31markunlamed: you can see the affected fonts also in the log cvs log..
11:03:12lamedah... I was using update -dP but I was just overwriting the .rockbox folder... so i've got doubles.
11:03:47 Join B4gder [0] (
11:04:51lamedMarkum: I _am_ looking at the fonts. and i was thinking strange of that 404 not found error respond... there should be a better representation of a removed from cvs files.
11:05:57lamedmarkum: where are you from?
11:07:05Jungti1234He lives in Netherlands.
11:07:06Lost-ashcvs has substandard support for removed files
11:07:13 Join Sando [0] (
11:09:15*preglow cheers SVN! SVN!
11:09:57 Join tvelocity [0] (
11:11:11lamedlost-ash were you referring to me?
11:11:21***Saving seen data "./dancer.seen"
11:12:01lamedjungti1234 - are you always answering for him? Is he a native speaker then?
11:12:08lamed(of hebrew)
11:12:45Jungti1234He is not now.
11:13:05Jungti1234So, answer instead of.
11:13:29lamedhe does speak hebrew, isn't he?
11:14:01Jungti1234I don't know.
11:14:25Jungti1234But, he is Dutchman.
11:15:24lamedwhat does he think of the political situation in Afganistan?
11:17:04markunlamed: no, I don't speak hebrew :)
11:17:14preglowsometimes i really hate the one pass linker tendencies of ld
11:17:15markunWas only using it to test unicode
11:17:20preglowlike when i need to use twenty libs
11:17:22_FireFly_preglow yeah switch to svn ;)
11:18:20lamedmarkum: then it's back to just me using hebrew.lang.
11:18:52lamedi should start a poll on that.
11:18:59markunidanm in the forum uses hebrew I think
11:19:18markunOr is that you?
11:19:37lamedyes but probably not the menu translation... I'm road_runner or lshachar in the forums.
11:19:52 Join edx [0] (
11:20:50Jungti1234It's 2GB now.
11:22:33lamedmarkum: lemee get it stright (posting that thread), currently the cour* fonts are unicode translated, but they don't yet support heberew charcters, but they will, right?
11:23:12 Join LinusN [0] (
11:23:48 Join Tandoc [0] (
11:24:38t0masany "code-devs" around?
11:24:43Tandocokay... so I set my langauge on my H120 to japanese... and now nothing is labelled.. can someone walk me back to the language selection menu...
11:24:54t0masCan I commit this patch:
11:25:04t0masit looks like a way better solution
11:25:08t0masthan battery_test
11:25:50markunlamed: I don't think all fonts will support hebrew, but you can help searching for unicode bdf fonts that do.
11:26:20LinusNt0mas: go ahead
11:26:42lamedt0mas: I think you _much_ better submit
11:27:02t0masI can do all of em now...
11:27:06t0masbut I'll first have to test those 2
11:27:15lamed2cents adv...
11:27:39Tandocseriously, can someone walk me back to the language selection menu .___.
11:27:40 Quit Sando (Read error: 104 (Connection reset by peer))
11:27:44lamedpreglow: why are you getting all angry today?
11:27:54t0maslamed: it doesn't compile for recorder (the first one)
11:27:59lamedtandoc: just happend to me!
11:28:00markunTandoc: menu -> down 3 -> right -> up 2 -> right
11:28:01Jungti1234hey Tandoc
11:28:05TandocTHANK YOU
11:28:07*Tandoc bows
11:28:22Tandoci got as far as down 3
11:28:25markunTandoc: I was wrong
11:28:32preglowlamed: sucky assembler, annoying linker and slow platform
11:28:33markunThe first one is down 2
11:28:48Tandocso down 2 == 3rd one down?
11:28:53Tandocwhich is what i got to >.>
11:28:54amiconnLinusN, t0mas: I didn't have a close look yet, but I think this plugin should _replace_ the old battery_test.rock
11:28:59markunyes :)
11:29:09 Join San [0] (
11:29:12t0maslamed: the first one is not ready.... it has 2 problems: 1. breaking recorder (serious problem) 2. not checking scroll boundary
11:29:22t0masamiconn: ok, shall I update the old one with this?
11:30:11markunTandoc: not many fonts support japanese
11:30:28markunat least unifont should work
11:30:30lamedt0mas: you _are_ fast. where you reading the log? 1. it shouldn't break the recorder. (should be a VEry minor problem), 2. yes, but it might be submitted until someone cures it, couldn't it? I mean - it has a fake boundary
11:30:52Jungti1234'Dotum' supports Japanese. :)
11:30:54Tandoca pity my tags aren't in unicode .___.
11:31:01amiconnt0mas I'd say yes, after testing on archos. Did someone already check this?
11:31:01t0maslamed: no, just checked out the link you gave
11:31:04markunwe should make a table of the ranges each font supports
11:31:14t0masamiconn: I am compiling a few simulators now
11:31:16*amiconn has his H140 with him today
11:31:51markunTandoc: you can set a codepage to convert them to unicode: gen. set. -> display -> def. codepage
11:32:15Tandocwhere can one get said 'dotum' font D=
11:32:15amiconnI'm not sure whether testing on the sim will help much
11:32:34markunJungti1234: can you find out if we can distribute the font?
11:32:39amiconn(for this plugin)
11:33:16lamedt0mas: It shouldn't break the recorder(compiling now)... should I submit a newer diff? second, it has a fake boundry limit, list.c:278. boundry set to 1000 pixels.
11:33:43Jungti1234markun: The word is difficult. :(
11:33:57t0maslamed: if you're sure it compiles on all targets... and you think the boundary is ok to fix later...
11:34:02t0masthen just email me a new diff
11:34:04t0masand I'll commit
11:34:23t0mas(see privmsg for my address)
11:36:02Tandoci google for dotum and it saus it korean?
11:36:14 Join Sando [0] (
11:36:34Tandocor does it support jap script
11:37:21t0mas<amiconn> I'm not sure whether testing on the sim will help much <−− do you have a recorder?
11:37:30Jungti1234It supports most Japanese.
11:37:42amiconnt0mas: Yes I do, but @home...
11:38:02t0masI'll hold the patch untill you've tested
11:39:46markunMaybe we should make a check to detect if a font and a language file are incompatible.
11:40:45DBUGEnqueued KICK Tandoc
11:40:45Tandoc12(10 Jungti1234 12): i'd be looking for a .fnt extension or something else?
11:40:50 Quit tvelocity (Read error: 110 (Connection timed out))
11:41:45markunTandoc and lamed will not be the only ones who try a lang file and end up with a unreadable menu.
11:43:38 Join tvelocity [0] (n=tony@
11:45:43lamedmarkun: it wasn't the lang. file. I was using english.
11:46:05lamedIt was using latin codepage with the corrupted font that you fixed.
11:46:14markunlamed: which font then?
11:46:20markunah :)
11:46:53 Part LinusN
11:46:55 Join LinusN [0] (
11:47:04 Part LinusN
11:47:22Jungti1234It's no copyright.
11:47:28 Join LinusN [0] (
11:47:56 Quit San (Read error: 110 (Connection timed out))
11:49:15lamedQ: is there a switch to tell diff to neglect any styling changes like adding newlines? It was amiconn that told me to both post a file that's better readeble and a regular one iirc.
11:49:50uncledraxpreglow: you have done an amazing job with musepack!!
11:50:18Tandocit'd be nice if i could read korean =__=
11:50:29Jungti1234They said 'Special copyright is no requirement.'.
11:51:10markunSomeone must have the copyright. I can't just rely on someone else saying it doesn't matter..
11:51:37lamedmarkun: what fonts should those thread guys use for russian?
11:52:01Jungti1234hey markun
11:52:24Jungti1234A person who it makes Dotum said.
11:53:00Jungti1234Hanyang I&C of Dotum made.
11:53:39Jungti1234And they said 'Copyright is no requirement'.
11:54:35Jungti1234Is it that I understand wrongly?
11:55:56amiconnmarkun: Maybe the problem with missing umlauts etc in the database view is that the database strings should be utf8 now?
11:56:10 Join Guest229461 [0] (i=phhome@
11:56:17amiconnIf so, both songdb.jar and need updating...
11:56:47Tandocthx for the help guys, too lazy to lurk/idle on freenode
11:56:49 Quit Tandoc ()
11:58:49lamedI'll ask again: is there a way to tell 'diff' to neglect styling changes like newlines added every here and there?
11:59:11Zagorlamed: -b
11:59:46markunamiconn: yes, they should be converted
12:00:10markunlamed: same fonts for russian. But you can try it first of course.
12:00:28lamedmarkun: i don't read russian...
12:00:28Jungti1234hmm.. markun
12:00:35 Quit KN|stiff (Read error: 110 (Connection timed out))
12:10:01 Join hshah [0] (
12:17:30 Join Zak1392 [0] (
12:18:54Zak1392hey guys
12:23:46Zak1392what's with the volume in the latest rockbox build?
12:25:26lamedit's so cool.. we've moved to a DB scale!
12:25:51lamedit's so complex that most users won't ever know what it means!! but it's the only real way to measure anything.
12:25:54XavierGrYeah but now I will have to set the option of volume to graphical.
12:26:15XavierGrIt is so akward to have full volume and see a 0 for it.
12:26:23Zak1392is there any way to set it back?
12:26:47XavierGryou will get used to it after a while.
12:28:13Zak1392what about an option to set it back? it makes my .wps look 'not' perfect now ;)
12:28:17lamedW#$WEFSG how do i tell diff to ignore lines such as "+ " (that's a plus and some white spaces) I've tied both
12:28:46lamedcvs diff -u -B -b and cvs diff -u -B -w
12:29:57XavierGrwhat about search and replace in the diff?
12:30:12XavierGrsearch and replace +"space" with nothing
12:30:17lamedthen you have to change the "@@ -66,10 +66,16 @@" lines
12:38:46lamedcan anyone use the diff −−strip-trailing-cr command? all i get is the help message
12:45:47 Quit uncledrax ("CGI:IRC")
12:47:25lameda cry for help?! anyone knows how do I recursively replace a whitespaced line with a blank one?
12:54:32XavierGrwhat is whitespace and what monospace?
12:55:37markunXavierGr: very different. whitespace: space and tab. monospace: same width for all the glyphs in a font
12:58:20XavierGrlamed why did you used space AND tab in the first place?
13:01:37 Join ]RowaN[ [0] (
13:02:15CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
13:02:15*lamed wonders what are you guys blabbering about. he was just adding some lines to make the code look better. sometimes probably leaving some white spaces.
13:02:20]RowaN["Removed the various sound scaling options" ... so no more original-firmware-like bass adjusting, back to volume capping?
13:03:03Zagor]RowaN[: no volume capping either, afaik
13:03:42]RowaN[but will that option disappear if i upgrade to todays build?
13:04:02Zagorif the change was committed yesterday, then yes
13:04:20]RowaN[in the words of Moloko... Bring It Back! =p
13:04:40ZagorI'm interested in hearing why you want the player to adjust your volume for you. I find the concept very strange.
13:04:50]RowaN[not adjust the volume.. adjust the bass
13:04:56Zagorsame thing
13:04:59]RowaN[like it does at the moment... 36hours old build iirc
13:05:15]RowaN[same as the original iriver firmware is what i mean, thats how i like it
13:05:38 Join webguest61 [0] (
13:05:38Zagori've never seen any audio equipment do that. if you crank the volume too loud, you may get distorsion. works the same on all stuff.
13:05:51]RowaN[i dont think you understnad what im saying
13:05:56Zagorprobably not :-)
13:06:23]RowaN[i think i'll keep a backup of my current rockbox.iriver file =]
13:06:35webguest61will I damage hardware if I unplug the earphone jack with the player on ?
13:06:43Zagorwebguest61: no
13:08:08lamedwas he serious?
13:08:29webguest61the new volume db scale is cool :)
13:09:28 Part webguest61
13:09:29preglowZagor: one of my main points as well
13:09:45preglowZagor: people are used to the concept of overdrive, even though they might not know exactly what it is
13:10:18preglowcan anyone tell me if the spdif output works in rockbox before i go hunt for a cable?
13:10:48XavierGrit works but not to record.
13:11:00amiconns/pdif out should work, s/pdif in not (yet)
13:11:24***Saving seen data "./dancer.seen"
13:12:35preglowok, i'll go see if someone's got a cable, then
13:14:31Zagor]RowaN[: I'm serious in my non-understanding. why do you want to player to adjust the sound settings for you?
13:15:42lamedwhy is there a language configuration question on the configure script? what difference does it makes?
13:16:28Zagorit chooses which language is compiled in
13:18:37 Join Febs [0] (
13:18:56preglowno cables with jacks
13:19:40lamedzagor: so... does that means there _is_ a hebrew default font?
13:22:00Zagornot font, language
13:22:14Zagorthe default font is always the same
13:22:37 Quit B4gder (Read error: 110 (Connection timed out))
13:23:04 Nick solexx__ is now known as solexx_ (
13:23:06FebsI was just reading the logs. Someone compained earlier of his settings being lost each time he updates to a new build.
13:23:31FebsThis is true for the H300s. Settings are lost even if there is no bump in the config version.
13:23:31 Join webguest00 [0] (
13:23:31lamedso it just set the default language. well that's not much of a help then. :/
13:24:24solexx_does anybody know whether it is advisable to enable clipping prevention in the replaygain menu *and* in the settings menu?
13:24:46preglowthere is no clipping prevention in the settings menu
13:24:53solexx_err, sound settings
13:24:59preglownot there either
13:25:24MrShleeThere are about 3 different types of clipping prevention tho
13:25:32preglowupgrade your builds
13:26:22solexx_preglow: i did this morgning. I just didn't notice that the setting vanished from the sound settings
13:26:43MrShleeholycrap batman.. its true
13:27:01solexx_it gave me some very bad pops anyway
13:29:27markunlamed: we'll have to update the default font to unicode somehow
13:29:31 Join San [0] (
13:30:05 Part webguest00
13:30:47lamedmarkum: would that be more compicated then just updating it?
13:33:05*amiconn just closed an ool feature request :)
13:33:49markunI think the default font is compiled in and not loaded from disk, so a unicode font would take up too much memory
13:35:16amiconn...and another one :)
13:35:44amiconnmarkun: Yes. But I already suggested something...
13:41:51Slasheriah, the structs are so nice.. i had something like entry = (struct tagcache_entry *)(hdr + sizeof(struct tagcache_header)); but instead of that sizeof, i should have had 1.. it take long time to find that problem :D
13:42:18preglowwhy don't we do typedef struct?
13:43:20Jungti1234Good night!
13:43:30 Quit Jungti1234 ("bye")
13:44:58 Join B4gder [0] (
13:45:48 Join linuxstb [0] (
13:48:38Zagorpreglow: we don't do typedef
13:48:42linuxstbZagor: What the iriver users are looking for (IIUC) is simply a setting that gives them high bass at low volumes, and flat bass at high volumes.
13:49:32preglowlinuxstb: of course that's what they want, that's what the iriver firmware uses
13:49:42preglowwhich we should ignore
13:49:53preglowi just find not typedefing structs a major bother
13:49:57Zagorlinuxstb: but ... why?
13:50:03preglowhaving to repeat 'struct' bloody everywhere
13:50:15Zagorpreglow: it makes the code soooo much easy to read
13:50:26Zagormuch more*
13:50:36preglowcan't say it's much of a difference to me
13:50:40preglowtakes more space if you don't
13:50:53preglowbut then again, it's not something i'm going to waste time arguing about
13:50:56Zagoras soon as you start typedefing very soon the code gets filled with different types that you have to look up the definition of
13:50:57 Join leftright [0] (
13:51:05 Quit San (Read error: 110 (Connection timed out))
13:51:11preglowZagor: i don't typedef other stuff, just structs
13:51:35 Quit Zak1392 ()
13:51:37Zagorstill, how would we know that without looking up the definition?
13:51:47preglowbesides, if you need to look up defintions of types, you're doing something wrong
13:51:50Zagorit might just as well be an int or a char*
13:52:13preglowbut anyway
13:52:15preglowi don't care much
13:52:27ZagorI do :-)
13:54:07leftrightpreglow: with regrads to your unmounting difficulties, have you tried disabling the system restore feature for that external drive under 'System properties', and disabling the garbage bin as well
13:55:14leftrighterr, I'm not suggesting that *you* normally have problems mounting things :)
13:55:19markunhm.. I did a typedef in arabjoin.h
13:56:47]RowaN[zagor: are you saying you dont like replaygain? that adjusts sound settings
13:56:48preglowleftright: will try later
13:58:01linuxstbZagor: Because if they push the volume high enough for clipping to start, then the way they will want to resolve the problem is to reduce the bass. So they want Rockbox to do it automatically for them.
13:58:24leftrightpreglow: then 'Left' click on the sys tray icon its quicker, my apologies if you already knew this
13:58:53SlasheriHmm, now the tagcache works much better.. When it's ready, it allows also browsing by genre (and with ramcache function enabled, everything is accessed from ram)
13:58:56]RowaN[linux.. why was that feature introduced about a day ago, and now removed (as i understand it)
13:58:59Zagor]RowaN[: imho replaygain is different. it's something you do to the songs you have, not something the player magically does.
13:59:24preglowleftright: that's what isn't working
13:59:30Zagordoes the iriver models have very weak headphone amplifiers?
13:59:36preglowZagor: no
13:59:43preglowZagor: not very weak, at least
13:59:54Slasherii think the HP amp is quite powerful
14:00:09]RowaN[anyone? =]
14:00:10preglowit is more than good enough for me, at least
14:00:12preglowand i like my wattage
14:00:29preglow]RowaN[: because amiconn came up with a solution that more devs liked
14:00:31linuxstb]RowaN[: There was a long discussion on the rockbox-dev mailing list
14:01:11Zagorif the amp is powerful enough, why is clipping a major concern? it's odd that over all these years I've never heard the problem raised before.
14:01:13Slasheriiirc, the amp was 20 mW to both channels
14:01:18]RowaN[so whats the favoured solution? could the OPTION not just be left in, for people like me who want it
14:01:45preglowjust pure old habit
14:02:21]RowaN[with the old/current rockbox way i find myself continually having to go into the bass menu either to increase it, or to decrease so i can get a decent volume, unlike the original firmware style
14:02:44]RowaN[i know not everyone is like me, but it would be nice to keep that option present =[
14:02:52Zagorthat sounds to me like a headphone amplifier problem
14:03:06Zagorwhat am I missing here?
14:03:16]RowaN[well no, its just a fact that with constant high bass, some tracks will get a lower volume
14:03:21]RowaN[to avoid clipping
14:03:38Zagorwhy would they clip if the amp can handle the power?
14:03:55Zagorare we talking digital clipping here?
14:04:03]RowaN[for SOME tracks, i like to have a really high volume, at the expense of volume..... thats why the old iriver firmware style is suited to me
14:04:14]RowaN[because then i dont have to keep going into the bass menu
14:04:22]RowaN[oops lol
14:04:29]RowaN[i mean really high volume at the expense of bass
14:04:29Slasheriit might be an internal overflow inside the amp.. i am not sure but at least the bass boosting allowed by the amp is huge
14:05:14]RowaN[ok well you've lost me now.. but all im saying is that yesterdays solution to my problem has now been taken away from me
14:05:19preglow]RowaN[: lower volume with high bass? eh?
14:05:25preglow]RowaN[: that's how rockbox used to work
14:05:31]RowaN[higher volume with low bass
14:05:36linuxstb]RowaN[: What if the bass was as easy to change as the volume? So you can tweak them yourself?
14:05:47PaulJZagor: i opened a thread at misticriver on this topic, maybe it helps you to understand why some people (at least me) would like to see this feature.
14:06:20]RowaN[linux.. thats another option, but like i say, the solution to my problem was already in place
14:06:34]RowaN[i think we've run out of buttons for bass adjustments on the wps =]
14:07:04leftrightwhy dont you use the qick menu for bass adjustment from wps
14:07:05linuxstbPaulJ, ]RowaN[: Something I think most devs feel strongly against is Rockbox adjusting settings invisibly in the background.
14:07:10Zagoreverything I hear points to the problem being a too weak amp
14:07:23 Join ep0ch [0] (n=ep0ch@
14:07:24Zagorso iriver made a software kludge to work around it
14:07:25]RowaN[because i prefer it to happen automatically, instead of going though a load of menus every time i switch tracks
14:07:36preglowweak amp?
14:07:39leftrightwhy dont you use the sound setting quick menu for bass adjustments
14:07:42preglowit's a digital filter
14:07:47preglowit's got nothing to do with the amp
14:07:48]RowaN[<]RowaN[> because i prefer it to happen automatically, instead of going though a load of menus every time i switch tracks
14:08:12linuxstb]RowaN[: Why do you change things when you switch tracks?
14:08:12preglowbut anyway
14:08:26preglowwe already have a 'prevent clipping' option
14:08:38Zagora stronger amp would give you enough volume without lowering bass
14:08:39preglowwhy can't we just move this from out of the replaygain menu?
14:08:42]RowaN[prevent clipping method options arent removed?
14:08:52Slasheriand amp is limited to the line out level.. So with volume 0 dB and bass boosting digitally, it must clip
14:08:56]RowaN[let me backup my install and try todays build
14:09:24linuxstb]RowaN[: So you haven't even tried the Rockbox you are complaining about? :)
14:09:49preglowZagor: the bass and treble controls are digital filters
14:10:04preglowZagor: it clips if you apply too much, it has nothing to do with amp wattage
14:10:05ep0chpreglow: what do you want to test with spdif output btw?
14:10:19PaulJpreglow: afaiu the prevent clipping option from the replaygain menu doesn't prevent clipping that happend because of too much bass boost
14:10:21preglowep0ch: not test, it's just ease of connecting my player to my current rig in the office here
14:10:27ep0chah ok
14:10:28leftrightthe current method is a "pure" method of volume adjustment, there is no trickery in the background,
14:10:29preglowPaulJ: no, but we could fix that
14:10:49]RowaN[linuxstb: i never said that i'd tried it
14:11:02]RowaN[i just pasted above what i saw in the wiki to discuss it
14:11:35Zagorpreglow: then how can it be related to volume?
14:11:48Zagorif the digital signal clips, it will clip at any volume
14:11:57preglowZagor: because at full volume, you have to assume that the numerical range is almost saturated
14:12:08Zagorvolume is not digital
14:12:19preglowZagor: are you sure about that?
14:12:44preglowZagor: i'd guess the dac ran full amp all the time, with volume controls being digital
14:12:55Zagoroh my god!
14:13:05Zagorthat's a nice guarantee for horrible sound
14:13:06Slasherii think the uda could be a digital amp
14:13:12preglowZagor: nah, not really
14:13:21preglowif you can implement it digitally, that's exactly what they'll do
14:13:26preglowlower component count, plus cheaper electronics
14:13:33Zagoryeah, and crappy sound.
14:13:36preglowZagor: why?
14:13:48Zagorbecause you have lower dynamic range at lower volumes
14:13:49Slasherii would say digital amps can have more accurate sound
14:13:58]RowaN[oh, as i suspected, the feature is removed. so my point stands.. yesterdays solution to my (and other peoples) problem has now been reversed
14:14:55Zagorbut ok, if the iriver uses digital attenuation then finally I understand the problem.
14:15:19linuxstb]RowaN[: Yes, and a possible different solution implemented - i.e. no clipping prevention at all. So Rockbox's old type of clipping prevention which no-one seemed to like (limiting the volume) has gone.
14:15:29preglowZagor: btw, it doesn't matter, you have a prescaler before the volume, so you can crank the volume down before eqing, then up (either digitally or analogue, doesn't matter) after eq
14:15:43preglowif you're already at the max amplitude in the amp, then you can't use this method anymore
14:15:47preglowand have to start limiting something
14:16:06]RowaN[linux: so in the replaygain menu menu, what does the prevent clipping option do now?
14:16:08preglowZagor: prescaler before the EQ, sorry
14:16:30Zagora decently dimensioned amp should never be at full power
14:16:58linuxstb]RowaN[: I'm not sure, but I guess it makes sure that the replaygain adjustment doesn't cause clipping.
14:17:14preglowZagor: it might not even be, but they still have to draw the line for the control somewhere
14:17:41Zagoronly when you reach 0db in the amp and still want to go louder
14:17:55Zagor...which means you want a bigger amp
14:18:04Zagornot mangle the input
14:18:08preglowyes, which happens, if they had designed some headroom in the amp they could only use internally, then this wouldn't be an issue
14:18:11]RowaN[so now there is no way to prevent clipping.. not even an option?
14:18:12 Join KN|stiff [0] (i=phhome@
14:18:24preglowbut they have the same range internally as the user sees
14:18:59MrShleethe best option is.. "got a problem.. buy a T-AMP"
14:19:01 Join tucoz [0] (
14:19:18preglowso when the user maxes it, they can no longer compensate for the eq prescale in the amp/volume control after the eq
14:19:38B4gder]RowaN[: and this is what we talk about all over
14:19:58tucozHi, just happened to notice the clipping-discussion again. As far as I understand this, the only use of the iriver way is that it's nice when used as line out.
14:20:08preglowif we're going to do this, i want just one new option for it
14:20:26preglowand i'm afraid some users are going to start complaining about splitting the different 'prevent clipping' effects into different options
14:20:40preglowone of the effects will be replaygain bsaed, another eq based, etc
14:21:12SlasheriHmm, can somebody send me a working tagdb file so i could try the existing implementation to see how it looks/works?
14:21:28linuxstbpreglow: The complication is that not only do people want the bass to be automatically reduced when they increase the volume, but they want it to automatically increase again when they decrease the volume.
14:21:48Zagorso it's the classic "we skimped on the hardware, tell the engineers to compensate for it in software"
14:22:00 Join mikearthur [0] (
14:22:05mikearthurhey guys
14:22:06tucozZagor, it's a flaw?
14:22:09mikearthurI thought you would want to know
14:22:10 Quit KN|stiff (Client Quit)
14:22:20preglowZagor: i guess you'll find this a lot in dap's. it's pretty understandable they don't want to put in over-designed amps
14:22:22Zagortucoz: from where I sit, yes clearly
14:22:23mikearthurI RMA'd my rockboxed iRiver, the hard driver died
14:22:27preglowZagor: you'll be wasting both power and money
14:22:37 Join KN|stiff [0] (i=phhome@
14:22:40mikearthurit came back with iRiver firmware installed, but no complaints about the firmware
14:22:49mikearthurand it is the same unit, same little dinks
14:23:05]RowaN[linux: you're saying its complicated.. but it was previously implemented so whats the prob
14:23:13preglowmikearthur: they probably didn't even see it
14:23:14tucozZagor, hehe, interesting. Come to think of it, it is in fact quite counter intuitive the way iriver-fw handles it.
14:23:14mikearthurhowever, a bit scratch in the paintwork, maybe they do this to all Rockbox users, as they are jealous of your firmware ;)
14:23:33preglowlinuxstb: what's so complicated with that?
14:23:37mikearthurit was in Germany, their EU office
14:23:40mikearthurjust thought you'd want to know
14:23:44Zagorpreglow: using a standard analog headphone amp is not over-designing. it's the standard.
14:23:48mikearthurin case anyone asks
14:24:04linuxstbpreglow: Making the bass setting reflect what the actual setting being used is.
14:24:07B4gdermikearthur: ;-)
14:24:12preglowZagor: yeah, but they'd have to limit the output power of it so that extreme eq cases would work with no surprises
14:24:25Zagoruhh, no
14:24:37tucozbut it's nice when I plug it to my HiFi-equpment, I only need to raise the volume to get a flat eq (when using iriver fw).
14:24:47preglowlinuxstb: i'd say they can pretty much forget such feedback
14:24:58preglowlinuxstb: it would have to be invisible for us even to bother implementing it, i'd say
14:25:09preglowlinuxstb: and that's how the iriver firmware works, they're used to it
14:25:13 Part ep0ch ("Kopete 0.10.3 :")
14:25:36B4gderand these guys want the bass to go up again when the volume is lowered
14:26:22B4gderwhich makes the bass have one "preferred" level and one "actual"
14:26:27leftrightis it possible to ever satisfy everyone ?
14:26:36]RowaN[yes, with enough options? =]
14:26:38PaulJlinuxstb: you could show 2 values in the bas boost menu. the value the user has set and in brackets teh value that is applied e.g 24 (12)
14:26:49B4gder]RowaN[: no, because no one likes many options
14:27:13tucozleftright, I believe we're almost there. Why not provide an option for this?
14:27:18]RowaN[how about an option to disable "complicated" options? hehe
14:27:33]RowaN[hide all the options from people who are confused by them
14:27:34leftrighttucoz, hehe
14:27:37B4gderan option to disable silly users would be neat...
14:27:42mikearthurleftright: no, never
14:27:44preglowB4gder: as long as the volume routine also sets the bass, i can't see a problem. if (volume + bass) > 0 { internalbass = 0 - volume } else { internalbass = bass } setbass(internalbass);
14:27:50]RowaN[are you calling me silly badger?
14:28:13mikearthurleftright: until my ihp140 will do my coursework for me, I'll never be totally happy with Rockbox's features
14:28:16preglowall badgers are silly
14:28:17B4gderpreglow: I agree, I was just thinking about the problem you mentioned about the change reflecting in the bass value
14:28:39preglowB4gder: i'd say forget that kind of feedback
14:28:50preglowB4gder: it complicates the settings system
14:28:51B4gder]RowaN[: I said silly users, if you feel hit then sure, that's you
14:28:53leftrightso keeping inline with Rockbox's "Design Philosphy" is is a primary factor
14:29:10B4gderif not, then it isn't you
14:29:15]RowaN[rudeness noted
14:29:17B4gderpreglow: I agree
14:29:26B4gder]RowaN[: being overly sentive today?
14:29:38preglowhey, you allowed him to choose himself
14:29:43linuxstbpreglow: How does this all fit in with the software equaliser you're planning?
14:29:47preglowif he thinks he's stupid himself, then we can't argue
14:29:54preglowlinuxstb: just dandy
14:30:02]RowaN[im a sensitive guy badger, but apolagy accepted
14:30:10mikearthuris there a feature to adjust voice volume? I haven't used rockbox for a month
14:30:14preglowlinuxstb: only difference is that the prescaler is software based
14:30:22 Join Jungti1234 [0] (n=jungti12@
14:30:25B4gder]RowaN[: ?
14:30:28preglowlinuxstb: i was thinking i'd just combine it with the replaygain gainer
14:30:54 Quit Guest229461 (Success)
14:30:57PaulJpreglow: this might be a bad idea
14:31:01Jungti1234hey markun
14:31:03preglowPaulJ: why?
14:32:05PaulJpreglow: if some files have replaygaun tags ond others don't, the ines with tags would play on much lower volume
14:32:27preglowPaulJ: why?
14:32:45preglowPaulJ: the extra lowered volume will be compensated after the eq anyway
14:32:47preglowPaulJ: just like now
14:33:43amiconnpreglow: Replace the '0' in your formula with sound_max(SOUND_VOLUME)
14:34:18preglowamiconn: it was just a concept thing :)
14:35:03t0masare you around?
14:35:21LinusNt0mas: yup
14:35:36MrShleeWell, I don't think there is any question about it. It can only be attributable to human error. This sort of thing has cropped up before and it has always been due to human error.
14:35:45tucozSo, the clipping handling has gone from no limiting -> limiting -> loads of options -> no limiting?
14:35:47t0masthe lcd driver, what is the second data argument (data2) in lcd_write_command_ex?
14:35:58B4gdertucoz: no
14:36:00preglowtucoz: no
14:36:02B4gderthe limit will come
14:36:10preglowtucoz: from volume limiting, to loads of options, to no limiting
14:36:36tucozpreglow, I was thinking of the first implementation, were iirc there were no limiting :)
14:36:41preglowcan't remember
14:37:01t0masLinusN: I'm trying to disable the LCD, for amiconn's low-power-mode idea... and I've figured how to disable all kind of things, and in which order... but I don't understand the second data argument... what is it for?
14:37:07LinusNt0mas: there is one command that takes 2 arguments
14:37:37t0masno, 3
14:37:44t0mascmd, data1, data2
14:37:56LinusNi meant *lcd* command
14:38:12t0masso in this command:
14:38:12t0maslcd_write_command_ex(LCD_CNTL_POWER_CONTROL, 0x16, -1);
14:38:18t0masthe -1 is just a placeholder?
14:38:41t0masok, that's why I didn't understand what it did from the datasheet :)
14:38:54PaulJpreglow: most of my mp3s have ~98db, and some of them even clip (according to replaygain) at 89db, so such an mp3 would be plaed at less than 89db, if it hadn't replaygaintags it would be played without a decrease in volume ~98db.
14:39:35preglowPaulJ: 98db is more than 16 bit audio can handle...
14:39:56preglowi don't get your db numbers
14:40:04lamedargh! it took me 15 minuts to find the one line puts_ functions of the h300! I haven't noticed that lcd-16bit driver.... arghhhhhh
14:40:15]RowaN[the volume % on my wps now goes from -84 to 0... is that normal for the newest build?
14:41:36lostlogic]RowaN[: it's in dB now not %
14:41:57]RowaN[0db sounds pretty loud hehe
14:42:03lamedThat's today topic acctually.... someone should chage the topic
14:42:03 Join judas20 [0] (
14:42:06Jungti1234markun, I fixed problem of Korean translation again.
14:42:22lostlogic"seriously we do not estimate volume levels any more"
14:42:25PaulJmaybe i'm wrong. but i have the impression that mp3s with replaygaintags play not as loud as mp3s without.
14:42:30 Part judas20
14:42:39*leftright thinks that amiconn's wake up timer will be rally handy
14:42:45preglowPaulJ: that is correct
14:42:50 Join Moos [0] (
14:42:52preglowPaulJ: replaygain almost always gains a track down
14:43:29lamedlostlogic rofl :D
14:43:44 Quit actionshrimp ("a bird in the bush is worth two in your house")
14:44:14preglowplease, please, anyone is there any way to make ld be multipass?
14:44:30preglowi can't handle this retarded single pass functionality with more than twenty lib files to include
14:45:03leftrightthe default value of 89 used by Replay Gain software was a "compromise" level, it can be adjusted i the software you use, but then you modify the headroom availble, which could influence the dynamic range of an album
14:45:16lostlogicLinusN: My profiling patch is getting to the point where I would consider submitting it, but I don't know how to properly whack it into the configure script, and my current parser is all Java (not sure if that's a problem).
14:45:21preglow89db references to what?
14:45:34preglowit can't be dBfs, and it can't be dbspl
14:46:22ColdtoastI reckon the peakmeter should be made conditional
14:46:27preglowit is
14:46:30Coldtoastso we can make our own animated peak meters :)
14:46:31preglowjust remove it from your wps
14:46:52Coldtoastwas that done in the past few days?
14:47:04Coldtoastlast time I tried it, it didn't work
14:47:06]RowaN[someone skool me - why is 0db not silent?
14:47:12preglowi wish someone went ahead and implemented this:
14:47:38LinusNlostlogic: parser in java?
14:47:55preglowLinusN: we'll fix that
14:48:17preglowi can only stand by so long and see java infiltrate rockbox!
14:48:19lostlogicLinusN: Java's my 'career language' so when I was presented with the problem of mapping PCs -> symbol names, the easiest way to do it for me was in Java
14:48:29lamedthere most be a wiki page for the new db scale.... ]rowan[ - 0db is max volume.
14:48:37preglowlostlogic: but anyway, how's the profiler working?
14:48:42lostlogicpreglow: if you want to port it to something else, be my guest, it really isn't _that_ hard.
14:48:50lostlogicpreglow: seems to work fine, RETARDEDLY slow, but fine.
14:48:51preglowlostlogic: i will, probably
14:49:03preglowlostlogic: sounds a bit strange that it's so slow
14:49:07preglowlostlogic: btw, is all profiling code in iram?
14:49:11*B4gder whispers peeeeeerl...
14:49:16preglowlostlogic: it should be, so it doesn't thrash the code cache
14:50:01lostlogicpreglow: good think, no I haven't put it in IRAM, can things in system just be popped over to IRAM like that?
14:50:43preglowlostlogic: btw, i don't think 'register' has much of a function anymore
14:50:53preglowlostlogic: well, sure
14:50:57 Quit Jungti1234 ("bye")
14:51:04preglowlostlogic: any function can be popped over to iram if there's enough space left in iram
14:51:11lostlogicalso, profiling as written takes up uhh 24k of data space (looke like it can be reduced a fair amount w/o getting close to running out on a common sized codec) is there enough free IRAM for the data to go there too?
14:52:01preglowlostlogic: no
14:52:07preglownot by a long shot
14:52:12preglowi think
14:52:25preglowbut anyway
14:52:27lostlogic*sad face*
14:52:29preglowi don't care about the profiler being slow
14:52:33*lamed preglow get a room or something
14:52:41lostlogicyou care about it trashing the code cache
14:52:52preglowi care about it not disturbing the profiled process
14:53:07lostlogic*sees if it can be moved to IRAM*
14:53:18preglowwhich it very certainly will if the code isn't in iram
14:53:36 Join Kohlrabi [0] (
14:54:11lostlogicshould only be the function enter and exits that need to be in IRAM, the context switch handling shouldn't matter, eh?
14:55:14lostlogicpreglow: BTW, make sure you have the latest"> if you are going to port my Java to a better language for the task. It changed a fair amount as of upload at about midnight CST
14:55:37preglowi'll let you finish it before it do anything to it
14:55:57lostlogicthe parsing and file format seem to be relatively static now
14:56:05lostlogicand the big part of parsing is loading up the symbols from libraries and map files
14:56:16lostlogicwhich is all in one .java file and can be ported at any time.
14:58:14*lostlogic reprofiles with profiler in IRAM
15:00:41preglowb7ut yeah
15:00:45preglowcould you post some example results?
15:01:15lostlogicin that lil tarball is example pre and post parse of a 1/2 song profile run on Tremor.
15:02:33preglowlostlogic: quite a lot of unresolved symbols
15:02:58lostlogicshite, I must not have uploaded the latest
15:03:30preglowi was looking at the double one
15:03:37preglowthe single one's got something different again
15:03:57lostlogicI haven' tupdated double in a while, I think I've decided to trash that design and not try to build call traces.
15:05:01preglowi don't get this
15:05:08preglowXPROD31 is supposed to be inline
15:05:11preglowwhy is it showing up
15:05:34lostlogicI wonder if -finstrument-functions de-inlines things, thereby totally fucking up results?
15:06:08 Quit tvelocity ("Leaving")
15:06:18lostlogiccould mark it unprofilable to prevent that if it is the case.
15:06:31preglowlostlogic: that would make sense, i guess
15:06:43preglowlostlogic: i don't know, i guess it's better to have it this way
15:06:59 Part leftright
15:07:35preglowlostlogic: but this looks totally decent, great work
15:07:40ColdtoastI found something interesting today
15:08:31Coldtoastif I create a playlist and the iriver caches tracks, if I then go and delete one of the songs in the playlist off of the player, it's still in buffer so the track plays
15:08:52preglowlostlogic: thought of any way to integrate this in rockbox?
15:09:03preglowlostlogic: i'd like to see this in cvs as soon as possible
15:09:06Coldtoastseems odd when it happens
15:09:11Coldtoastis that avoidable?
15:09:57preglowsomeone would need to code around it
15:10:25Coldtoastit's probably a very rare situation anywa
15:10:53PaulJi like this behaviour, this way i can rename a song that is currently playing
15:10:55Coldtoastmaybe not worth botherign about? but I thought I'd mention it
15:11:04preglowPaulJ has a good point
15:11:14ColdtoastI'm talking about a song you delete from the player
15:11:20preglowit's a fringe case, and i don't know if we should waste code space on handling it
15:11:25***Saving seen data "./dancer.seen"
15:11:27Coldtoastwhat does that have to do with renaming?
15:11:29preglowColdtoast: yeah, but it's the same thing
15:11:38preglowColdtoast: the file is changed, and isn't where it used to be
15:11:55preglowi think this should stay as is
15:11:59Coldtoastit might result from the same thing, as far as codign goes, but it's hardly related in usage terms
15:12:11 Join leftright [0] (
15:12:25Coldtoastwell, as I said I just thought I'd mention it. heh
15:12:34preglowhaven't thought about it myself
15:12:37preglownever tried that
15:12:41Coldtoastme either til today
15:13:14PaulJi rename the extension of faulty files to .xxx
15:13:16Coldtoastit could cause a prob tho if a track is half-cached when you delete it maybe?
15:13:16leftrightfixing wiki entry, on Archos players can tone adjustments be made <0 for bass and trable ?
15:14:08LinusNleftright: yes
15:14:25ColdtoastI' wonder what Rockbox will do if I delete a track while it's playing and cachign at the same time and there are no tracks in teh playlist efter it
15:14:37preglowtry it
15:14:43ColdtoastI will now
15:15:05Coldtoasthopefully Rockbox will auto change to the next dir, as I have that set
15:15:17 Quit Febs (Read error: 110 (Connection timed out))
15:19:02Coldtoastyep. it did
15:19:41 Quit MrShlee ("Changed by popular Request - "I am the heart of your darkness, you stinking oompa-loompa morons!"")
15:19:55amiconnColdtoast: This behaviour is there and known for long
15:20:02amiconnImho it's not a problem at all
15:21:05amiconnRockbox simply plays the cached data even if the file gets deleted before. If a track is only half-cached (much more likely on archos), it plays the cached part, then advances to the next song
15:21:22 Quit KN|stiff ("( :: NoNameScript 3.81 :: )")
15:21:25lostlogicpreglow: to build profiling support is pretty easy −− Makefile just needs to define RB_PROFILE, that's all good. To profile a specific codec, ya just need to add -finstrument-functions (adn I think remove -fomit-frame-pointer) for that specific codec.
15:21:48Coldtoasthow much is a 40GB drive for the h100 these days?
15:22:16amiconnlostlogic: Imho it would make sense to make it selectable in configure for (D)eveloper builds
15:22:29preglowlostlogic: ouch, then it will affect us
15:22:34preglowlostlogic: since we lose a register
15:22:45]RowaN[my 30gb one was about 짙118, 4months ago (including postage from germany to UK)
15:23:15ColdtoastI think the HDD noise is affecting audio
15:23:16lostlogic*ponders* lemme see if it works w/-fomit-frame-pointer −− I don't think so, but I haven't actually tried -finstrument-functions and -fomit-frame-pointer.
15:23:22Coldtoaston my player
15:23:29lostlogicamiconn: configurable per-codec in the dev builds?
15:23:48preglowlostlogic: why not enable it for all codecs?
15:24:00amiconnAh, this is per codec. Hmm, that could be difficult
15:24:13*amiconn thought the profiling stuff is general
15:24:22preglowit's got to be per thread
15:24:27preglowor things get mixed up
15:24:39lostlogicpreglow: because a profiling session on more than one codec will have overlapping symbols.
15:25:04amiconnah, right
15:25:09preglowlostlogic: yeah, yeah, but you have to stop after one, of course
15:25:23preglowlostlogic: but let's just make that the responsibility of the coder, not the build system
15:25:31linuxstblostlogic: But does it matter if you only actually run a single codec?
15:25:41preglowwhat linuxstb says
15:25:43 Quit B4gder ("Time to say moo")
15:26:05lostlogic*nod* I'll work on making it pretty with the build system today.
15:26:22 Part tucoz ("Leaving")
15:28:00preglowlostlogic: would be smashing
15:28:13 Quit _FireFly_ ("Remember, you're unique, like everyone else")
15:30:18Coldtoastwow ]RowaN[. that was EXPENSIVE!!
15:30:36Coldtoastjust looked and I can buy a 40GB one for $125 without shpping
15:30:56lostlogicit builds with -fomit-frame-pointer.
15:31:05preglowlostlogic: does it work? :-)
15:31:09lostlogicabout to find out
15:31:48preglowlostlogic: looks to me like it should work fine without the frame pointer
15:31:54preglowit's the instruction pointer it tracks
15:32:32 Join actionshrimp [0] (
15:32:34lostlogicyarr, as long as GCC's __builtin_return_address(1) function didn't require the frame pointer.
15:32:52preglownah, that just accesses the stack frame and finds the return address
15:33:13preglowi think frame pointer are only used for debuggers to find the argument list and so on
15:33:35lostlogicgprof requries it so I just assumed (which should never be done)
15:34:00preglowbut then it's all just peachy
15:34:00 Join San [0] (n=Test@
15:34:06preglowprofiling can't get much better than this
15:34:08]RowaN[coldtoast: it was the cheapest at the time.... or was that 118 euros? cant remember now hehe
15:34:13lostlogicit runs noticeably faster w/-fomit-frame-pointer
15:34:41]RowaN[thats reminds me ive still got a half dead 20gig drive
15:34:46]RowaN[i ebayed it but no one bidded
15:34:49Coldtoastwell, the good thing is it means it's worth replacing the drve when it DOES die
15:35:03Coldtoastrather than replace the player
15:44:42lostlogicam going to have to adjust profiling code for -fomit-frame-pointer, otherwise it goes smashy smashy.
15:46:04 Join webguest93 [0] (
15:46:36 Quit actionshrimp ("a bird in the bush is worth two in your house")
15:47:08lostlogicyeah, changes what values I get for call_site which results in using many more spaces in memory (to the point that it overflowed), can switch back to using __builtin_return_address again which should work (as you said)
15:47:46webguest93hi, i'm having a bit of trouble with codepage settings. can anyone help me?
15:47:55preglowjust ask
15:48:16lostlogicbut that made it slower again. oh well.
15:48:39preglowlostlogic: well, as i said, i don't care much about that, what matters is that the code is as identical as possible to the optimised case
15:49:11 Part LinusN
15:49:48 Part leftright
15:49:51webguest93whenever i set the codepage option, it resets back to the default latin after restarting the player
15:51:10webguest93this doesnt happen with the build i've been using 12/05 build
16:00:09lostlogicgah, parser doesn't handle statics in IRAM right.
16:00:12lostlogic*shoots self*
16:01:35lostlogicalso bbl.
16:07:01 Quit San (Read error: 110 (Connection timed out))
16:25:26 Part mikearthur ("Konversation terminated!")
16:26:57amiconnwebguest93: The codepage setting isn't saved on disk. I'll fix this tonight.
16:27:28amiconnThe 2005-12-05 build didn't have this setting yet, it was introduced with unicode support
16:30:17webguest93huh? then why do i have it?
16:30:29webguest93oh probably because i patched it
16:36:04 Join webguest80 [0] (
16:36:37webguest80OMFG ROCKBOX OWNS (that is all)
16:36:43 Quit webguest80 (Client Quit)
16:38:08*t0mas stares at his screen in amazement...
16:39:46amiconnmarkun: r u there?
16:42:09lamedguys... i really nead help. I'm getting weird
16:42:09 Quit webguest93 ("CGI:IRC (EOF)")
16:42:14lamedLD rockboxui.exe
16:42:32lamedhome/guest/rockbox-devel/sh3/apps/plugin.o(.text+0x68): In function `rockbox_api':
16:42:39lamedhome/guest/rockbox-devel/apps/plugin.c: undefined reference to `_lcd_puts_style'
16:42:45lamedhome/guest/rockbox-devel/sh3/apps/screen_access.o(.text+0x2dc): In function `screen_init':
16:42:50 Quit Zagor ("Client exiting")
16:42:51lamedhome/guest/rockbox-devel/apps/screen_access.c:139: undefined reference to `_lcd_puts_offset'
16:43:02lamedhome/guest/rockbox-devel/apps/screen_access.c:139: undefined reference to `_lcd_puts_offset'
16:43:03 Join muesli_- [0] (
16:44:11lamedwhen i'm trying to build for a h300 sim. it didn't encounterd the problem just up until trying to make the target file. (there weren't any warnings before it crashed)
16:44:37lamedwhy would it ever tell me that the .o file has problem?
16:45:43 Quit Coldtoast (Read error: 104 (Connection reset by peer))
16:46:56lamedYES! god @#RWLDKF da2##$it i can't believe it...
16:47:07lamedi'm on it for couple of hours
16:48:30hshahcan someone PLEASE commit my WPS because TiMiD hasn't gotten round to it...
16:49:10merbananwho was working on faad here ?
16:50:19linuxstbmerbanan: I did the initial work to get it working in Rockbox. But I've done nothing since then.
16:51:02 Join DangerousDan [0] (
16:51:56merbananlinuxstb: ok, there is a patch to speed up some things on mplayer-dev
16:52:01markunamiconn: yes, but busy
16:52:12merbananaccessable through gmane
16:55:16linuxstbmerbanan: Do you have a URL handy?
16:55:58preglowoptical out works
16:56:52merbananlinuxstb: the fft is optimized abit, some inlining and other tweeks
16:56:59merbananrename it to gz
16:57:29linuxstbDo you have a link to the email message that went with the patch?
16:58:05linuxstbmerbanan: OK, I've found the thread.
16:58:11preglowurl, ples
16:58:12merbananlinuxstb: <−− sorry for being stupid :)
16:58:56amiconnmarkun: The codepage setting isn't saved to disk. Maybe you didn't commit all unicode changes to cvs? Just asking, will fix it tonight...
17:00:59lamedsomeone shoot me. please.
17:01:09 Join muesli- [0] (
17:01:20lamedanyone willing to review my code and let me know why won't it compile?
17:05:27merbananlinuxstb: 30% faster on x86 was claimed, but that number was a rough estimate
17:05:40XavierGrI would help you but I must run off to work see you later lamed...
17:05:57XavierGrmerbanan: 30% faster what?
17:06:18linuxstbmerbanan: I can't test it now, but it looks a very useful patch, and we should be able to apply it to our version of libfaad.
17:06:19merbananinteger faad on x86
17:07:30 Quit lamed ("CGI:IRC (EOF)")
17:07:41XavierGrlinuxstb: libfaad is usde for flac codec?
17:08:21linuxstbXavierGr: No, we use the FLAC decoder from ffmpeg
17:08:43XavierGrso where libfaad is used?
17:08:47linuxstbFor AAC
17:08:51XavierGrI mean which codec?
17:10:29linuxstbmerbanan: If preglow doesn't get there first, I'll have a look in a few hours time when I'm at home. Thanks for pointing it out.
17:11:13merbananlinuxstb: anytime, hope it speeds it up abit
17:11:27***Saving seen data "./dancer.seen"
17:11:50linuxstbDo you know if mplayer now maintain their own fork of libfaad?
17:11:52preglowi'm a bit busy
17:12:07merbanannote that it removes the rounding for some things
17:12:40merbananmplayer has their own snapshot/fork
17:13:06merbananshould be in mplayer cvs
17:13:40 Join lamed [0] (
17:13:59lamedxaviergr thanks I hope i won't need your help later. :((((
17:21:21 Join actionshrimp [0] (
17:25:09 Quit muesli_- (Read error: 110 (Connection timed out))
17:30:16 Join EoS [0] (
17:32:15 Join Benacool [0] (
17:35:30EoSanyone here?
17:36:43Benacoolme ;-)
17:38:05EoSare u using rockbox on iriver?
17:38:34Benacoolon my US H320
17:39:08 Join Cassandra- [0] (
17:39:56EoSdo uave this bug when u stop playback the player will make very loud crackling noises on the headphones?
17:40:50amiconnEoS: Update to the latest bleeding edge; there was a bug in the fade on stop/pause function
17:41:14amiconn(or deactivate said function)
17:43:03EoSah ok thx!
17:47:24 Join eli_sherer [0] (
17:47:25 Join Nixsos [0] (
17:49:20 Join San [0] (
17:50:09 Join Febs [0] (
17:50:25EoSi have a h140, do i get the h100 or the h120 rockbox?
17:52:17linuxstbmerbanan: I've just naively applied that patch from mplayer, and it does give a noticable speedup. It's still not quite realtime on my 128kbps test file, but it's very close.
17:52:35preglowcloser than usual?
17:52:43preglowit's very close here as well, but i've got some uncommited stuff
17:53:19*preglow is looking forward to profile data from faad
17:53:29linuxstbpreglow: Yes, closer than the CVS version.
17:53:56EoSah ok i just disabled the fade on stop/pause on no crackling sounds anymore
17:54:26linuxstbThe patch adds some small x86 asm optimisations, so it may be useful to do EMAC versions of them.
17:54:32 Quit elinenbe (Read error: 110 (Connection timed out))
17:54:58preglowis that .bin file a patch?
17:55:06linuxstbYes, it's a diff.gz file.
17:55:14preglowi'll download and have a quick look
17:55:41linuxstbrename it faad_speed.diff.gz if you want to give it the original name
17:57:12preglowat least the memcpy opts are applicable to us
17:57:15preglowour memcpy should be pretty fast
17:57:38 Quit Cassandra (Read error: 110 (Connection timed out))
17:57:39 Nick Cassandra- is now known as Cassandra (
17:58:04*amiconn smiles :)
17:58:04preglowlinuxstb: only asm optimisations i see are ones we've already done
17:58:14linuxstbOK :(
17:58:20preglowlike the MUL functions
17:58:33preglowamiconn: i hope you've got a good reason for smiling in here!!
17:58:44amiconn[17:57:43] <preglow> our memcpy should be pretty fast
17:58:54preglowahh, rightie, heh
17:58:59]RowaN[sorry can someone explain to me why rockbox no longer has the option to prevent clipping?
17:59:24preglow]RowaN[: we didn't like it
17:59:52preglowwell, at least i didn't
18:00:07]RowaN[do you enjoy music that clips?
18:00:21preglowno, i just don't push it so hard it clips
18:00:35preglowi like to decide these matters myself, instead of letting something do the choices for me
18:00:42preglowanyway, didn't we discuss this a couple of hours ago?
18:01:10 Join muesli_- [0] (
18:01:18]RowaN[it would be nice to have the option, for people like me who cant magically tell if they've set the volume too high until its too late
18:01:28preglowyes, and you'll probably have it
18:01:43]RowaN[what do you mean?
18:01:43preglowi'd like the 'prevent clipping' option we've got in replaygain to prevent that kind of clipping as well
18:01:56preglowbut i don't want a zillion options for this
18:02:09preglowit'll be one behaviour, and i think the iriver behaviour is what it should be
18:02:50]RowaN[you mean where the bass gets stripped/applied as you increase/decrease the vol?
18:03:10]RowaN[thats what was in place yesterday though??
18:03:19preglowyes, but with a shitload of other options as well
18:03:21preglowand no db volume scale
18:03:45]RowaN[so rockbox is just in going though a "phase" thats its going to grow out of? =p
18:04:08preglowa phase of what?
18:04:19]RowaN[a phase of no control over clipping
18:04:28preglowcall it what you want
18:04:34preglownew features come and go all the time
18:04:38preglowthis is just another case of change
18:04:55]RowaN[rockbox usually changes for the better
18:04:59 Nick Lynx_ is now known as Lynx_awy (
18:05:01preglowi think it just did
18:05:04FebsYou know, the whole "behind the scenes" issue could be solved if the bass boost setting actually changed as volume was increased in prevent clipping mode.
18:05:08preglowi'm not fully satisfied with the volume control
18:05:10markunafter replacing so many fonts with the etl16-unicode font I discover it was the base font for unifont :)
18:05:21]RowaN[and who are these people who hate options? where are they?
18:05:28*Febs doesn't even undertand what he just wrote.
18:05:32preglowFebs: that's what i have planned
18:06:15lamedfebs, try 8x13B
18:07:06FebsSorry, I'm not being very articulate today. I meant that even *I* didn't understand my own comment about clipping prevention.
18:07:19lamedCan anyone please try a patch of myne to help me trace the problem with it?
18:07:27preglow]RowaN[: try maintaining your own fork of rockbox where you just stuff all the options you can think of into it, and we'll see which rockbox is the easiest to use in one year
18:07:39preglowand not least: which rockbox has the least bloat
18:07:56preglowFebs: well, at least i think i understand it
18:08:02lamedIt's as simple as making a simulator with my patch installed.
18:08:06linuxstb]RowaN[: Options don't just make the user interface more complicated, they make the code more complicated and harder to maintain.
18:09:04Febspreglow: I think that you do. One of the complaints about a prevent clipping mode is that too much goes on "behind the scenes" and the actual bass boost no longer represents the bass boost setting.
18:09:21 Join Noel_sad_song [0] (
18:09:22FebsIf the setting were to change as the bass boost were limited by the prevent clipping function, it would mitigate that complaint.
18:09:39preglowFebs: that people will just have to live with, i can fix the bass adjustments behind the scene, but i will not complicate the settings code just so people can see if they're trying to overload the dac
18:09:50linuxstbFebs: But the problem is that users want the bass to return to the original high setting when the volume is lowered again.
18:10:05preglowlinuxstb: that's only a problem if you want the settings values to reflect the internal values
18:10:09preglownot even the iriver fw does this
18:10:15preglowi don't even view it as an issue
18:10:20 Quit San (Read error: 110 (Connection timed out))
18:10:26amiconnThe question is whether we should have clipping prevention at all. So far the iriver H1x0 with original fw is the only device I know of that does
18:10:37preglowamiconn: i sure as hell think it's just fine as it is
18:10:54preglowi'd like to just let the current code for a couple of weeks and see if complaints die down
18:11:04amiconn...but *if* we do, then we should (1) avoid having a bazillion options for it and (2) make visible what's going on
18:11:36preglowamiconn: i'd like _just_ the iriver firmware way of doing it (imho, volume should be the last parameter to be capped), and i'd like all the clip prevention mechanisms rolled into one option
18:11:46]RowaN[preglow: forks wha?
18:11:48preglowthe make visible part will be difficult
18:11:50amiconn(2) should be way easier than before since we have volume in dB now
18:11:53preglowit's clutter the settings code
18:12:17]RowaN[well yesterdays build with all those extra options (3 options) is simple to use, for me, so i'll continue to use it
18:12:31preglowand you're free to do that
18:12:51 Join LinusN [0] (
18:13:40LinusNpersonally, i think the fletcher-munson style clipping prevention is a good thing to have
18:13:44FebsThe irony here is that I will probably never use the prevent clipping function myself. I'm just advocating a function that I think others will find useful.
18:13:56EoSi think rockbox is really awesome... im impressed as hell that u guys pulled that off!
18:14:10LinusNespecially since the clipping will sound rather nasty, since it's digital
18:14:20preglowLinusN: how the hell can we implement that?
18:14:23LinusNEoS: thx
18:14:35LinusNwhat i meant was the bass scaling
18:14:48preglowwhat does that have to do with fletcher-munson? :>
18:15:12amiconnBass + treble it is
18:15:13preglowfletcher-munson clipping prevention sounds like a fancy name for replaygain
18:15:56lostlogicxpreglow: Fixed the java version of my parser in the latest tarball so that statics in iram work. ;) *cheers "Go Java"*
18:16:03preglowlostlogicx: wooot
18:16:14preglowlostlogicx: come on! impress us with a test profile of faad! :PPP
18:16:32LinusNi was under the impression that those were the guys with the theory about the human ear hearing bass frequencies better at high volume
18:16:42Febs"Fletcher-Munson" style clipping prevention would probably be a lot more palatable to many here than "iriver-style." ;)
18:16:44 Quit muesli- (Read error: 104 (Connection reset by peer))
18:16:47preglowbut anyway, my five cents, if we are to implement one clip prevention mode for the hardware eq, i say we do it like iriver fw does it, that seems to be the functionality the greatest number of people want
18:16:48lostlogicxdon't have any faad songs, actually don't even have my iRiver on me ATM, should be doing my job or something
18:16:52eli_shereri'm having trouble with songs atart playing 1sec before they need to...
18:17:43FebsLinusN: Yes, that's right. E.g.
18:17:50preglowLinusN: ahh, ok, you could say that, i'm just used to talking about fletcher-munson curves, which describes how the ear works over the entire frequency range
18:18:07preglowaacs can be provided
18:18:29preglowbut anywho, i need a small break, brb
18:18:50LinusNpreglow: A-weighting?
18:19:41LinusNeli_sherer: you mean it starts 1s before the previous song ends?
18:21:11]RowaN[gotta go
18:21:14]RowaN[keep it reel guys
18:21:19 Quit ]RowaN[ ("byee")
18:22:25eli_shererLinusN: no...before the song even starts it plays a sec and then play it again normally...
18:22:52LinusNeli_sherer: how are your crossfade settings?
18:23:04eli_shererLinusN: don't know...
18:23:30eli_shererLinusN: just checked..."off"
18:23:30LinusNplease give me an example how you make this happen
18:23:44LinusNfrom when you turn it on
18:24:14LinusNwhich build are you running, and on which platform?
18:24:47eli_shererdaily build from 07/12 on h300
18:25:12preglowLinusN: a-weighting curve is based on one of the fletcher-munson curves, yes
18:25:26LinusNpreglow: i guessed that
18:25:38preglowLinusN: one of the curves for low intensities
18:25:48preglowLinusN: so it's a bit misused
18:28:20lostlogicxgah, developer builds disable -fomit-frame-pointer, which we decided was a bad idea for profiling (and breaks dsp build any way)
18:28:48preglowlostlogicx: i know, and there's no reason it should be disabled for iriver dev builds
18:29:00preglowat least linus can fix his own builds when using the bdm :-)
18:29:08LinusNdamn! the iaudio x5 has no bdm debug port!
18:29:14preglowLinusN: wah?
18:29:24preglowLinusN: the pins are still there?
18:29:26preglowoh, bga...
18:29:41preglowLinusN: well, that'll be a fun solder job
18:29:44LinusNwell, no risk, no fun...
18:30:42 Quit Noel_sad_song ("CGI:IRC")
18:31:40lostlogicxLinusN: good luck :-\
18:32:02LinusNhehe, thx
18:32:06lostlogicxpreglow: -fomit-frame-pointer back on for profiling then.
18:32:12lostlogicx*makes a stern face*
18:32:42LinusNi believe the scheduler will choke if you enable it
18:33:31lostlogicxLinusN: better idea of where to put the profiling option in the build system then?
18:33:46LinusNi'm not sure, i haven't given it much tought
18:35:55 Quit lamed ("CGI:IRC (EOF)")
18:35:55 Quit DreamTactix291 (Read error: 104 (Connection reset by peer))
18:36:24 Join DreamTactix291 [0] (
18:36:33 Quit linuxstb ("Client Exiting")
18:42:17lostlogicxpreglow: got a static volatile struct pfd_struct * and want to stick it into a struct pfd_struct *, get warning, how to make go away?
18:42:40preglowwhat warning
18:43:27lostlogicxassignment discards qualifiers from pointer target type
18:45:15preglowlemme see the code
18:45:20 Quit Cassandra (Read error: 104 (Connection reset by peer))
18:47:57lostlogicxLinusN: so Profile and Debug would be incompatible because DEBUG needs frame pointers?
18:48:22preglowit does only for targets that have hardware debuggers
18:48:39lostlogicxwhich H300 does
18:49:26eli_shereralmost implemented support for all BMP bpp types...yay!
18:49:29LinusNpreglow: or gdb serial port debugger
18:49:42preglowLinusN: doesn't that count as hardware?
18:49:53LinusNpreglow: nitpick
18:50:12lostlogicxok, I'll make them mutually exclusive.
18:50:30LinusNlostlogicx: the bdm debugger doesn't need frame pointers
18:50:51LinusNin fact, i have never built with -DDEBUG
18:51:01LinusNthat is an archos legacy
18:51:05lostlogicxin the configure script Debug turns on -g which requires frame pointers, what is Debug used for?
18:51:19 Join petur [0] (
18:51:36LinusNthe debugging would be much easier if we had frame pointers, but we manage without them
18:51:39 Join dpassen1 [0] (n=dpassen1@resnet-233-61.resnet.UMBC.EDU)
18:51:53lostlogicx*nod* then making them exclude each other won't kill anything.
18:51:55LinusNor am i in the blue here?
18:52:09preglowif you look at the disassembly, you can figure out where the stack frame is anyway :-)
18:52:13LinusNdon't we always build with -g?
18:52:24preglowi wouldn't think so
18:52:31lostlogicx-g is for gprof profiling
18:54:02LinusNi was under the impression that -g was for adding debug symbols
18:54:07preglowit is
18:54:21LinusNand rockbox is always built with -g
18:54:52LinusNGCCOPTS=-W -Wall -O -nostdlib -ffreestanding -Wstrict-prototypes -g -m5200 -Wa,-m5249 -malign-int -mstrict-align -fomit-frame-pointer
18:55:20lostlogicx*ponders* do I ahve the wrong option in mind?
18:55:33LinusNperhaps you do
18:55:53LinusNplease refresh my memory, does profiling need the frame pointer?
18:56:32lostlogicx-pg gprof profiling needs frame pointer
18:56:35lostlogicxmy profiling does not.
18:56:50LinusNok, so what is our problem?
18:56:58*LinusN is confused
18:57:06lostlogicxwant to add -fomit-frame-pointer to developer builds when profiling is selected
18:57:25LinusNbut it's already there
18:57:32preglowit didn't use to be
18:57:35preglowsure someone changed it?
18:57:49lostlogicx debug=""
18:57:49lostlogicx echo "Normal build selected"
18:57:49lostlogicx GCCOPTS="$GCCOPTS $GCCOPTIMIZE"
18:57:57LinusNlook at the GCCOPTS above
18:58:27lostlogicx-fomit-frame-pointer is in GCCOPTIMIZE which is not concatenated on devel builds unless I'm reading wrong.
18:58:45LinusNnobody has ever used the "developer builds"
18:58:54LinusNon the iriver
18:59:12LinusNit's a legacy from the archos builds
18:59:34LinusNso you can do whatever you want with it
18:59:38LinusNon the iriver
19:00:07 Join NightCat [0] (
19:00:17 Part NightCat
19:01:29 Join NightCat [0] (
19:01:36 Part NightCat
19:02:22preglowbut it's a fact that fomit-frame-pointer wasn't enabled not too long ago
19:02:28 Quit EoS ("Peace and Protection 4.22")
19:02:31preglowpeople complained that my crossfeed opts didn't work in debug builds
19:02:40lostlogicxpreglow: still isn't if you select develoepr build and not profiling.
19:02:42preglowturns out it was because gcc didn't allow clobbering of fp in those builds
19:04:02 Join NightCat [0] (
19:04:03*eli_sherer doesn't have enough motivation to make support for 1bpp BMPs :_(
19:11:14preglowmy, coding for proprietary platforms are fun
19:11:30***Saving seen data "./dancer.seen"
19:14:31amiconnLinusN: Afaik 'developer builds' allow to choose several developer options, one of which is logf. The latter is used by a number of devs afaik, and it was introduced especially for iriver...
19:16:41 Quit muesli_- (Read error: 110 (Connection timed out))
19:17:54LinusNamiconn: true, i was referring to the "debug" build only
19:19:58amiconnThere's someone in the forum who seems to have a future rockbox build ;)
19:21:17LinusNamiconn: ?
19:21:41amiconn <=
19:23:37*eli_sherer did && instead of & and now realises it...Bits are a b****
19:24:16*eli_sherer thanks LinusN for giving him USB support (makes the testing alot more faster).. ;)
19:24:17 Join muesli_- [0] (
19:25:12preglowi don't get this
19:25:22preglowsome people seem to take for granted that iriver and rockbox sounds different
19:25:24preglowin default configuration
19:25:32*amiconn would like to know why there is && and || but not ^^ available in C
19:26:03LinusNamiconn: me too
19:28:20eli_sherer|| = OR | = XOR ... :)
19:28:59eli_shererlogical meaning...
19:29:31eli_sherera=true b=true (a|b)=FALSE
19:30:41preglow|| = logical or
19:30:45preglow= bitwise or
19:30:53eli_shererand logical
19:31:13HClbitwise xor is ^
19:32:11eli_shererthat's right...
19:32:38 Join _FireFly_ [0] (
19:32:50eli_sherer & bitwise AND
19:32:51eli_sherer ^ bitwise exclusive OR
19:32:51eli_sherer | bitwise inclusive OR
19:32:51DBUGEnqueued KICK eli_sherer
19:32:51eli_sherer && logical AND
19:32:51eli_sherer || logical OR
19:32:51***Alert Mode level 1
19:32:53eli_sherer AND-expression & equality-expression
19:32:55eli_sherer exclusive-OR-expr ^ AND-expression
19:32:57eli_sherer inclusive-OR-expr | exclusive-OR-expression
19:32:59eli_sherer logical-AND-expr && inclusive-OR-expression
19:33:01eli_sherer logical-OR-expr || logical-AND-expression
19:33:05CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
19:33:05*eli_sherer sorries for the flood
19:41:12LinusNis there a reason that the codepage setting isn't saved?
19:42:52***Alert Mode OFF
19:45:12preglowhe's going to fix it, i think
19:48:39 Quit Benacool (Read error: 110 (Connection timed out))
19:52:24 Join San [0] (
19:56:05San[16:50] <EoS> i have a h140, do i get the h100 or the h120 rockbox?
19:56:05San[16:51] <Nixsos> h120
19:56:12Sanwhats the difference^^?
19:57:07 Join xmixahlx [0] (n=xmixahlx@
19:57:47preglowh100 is for models with 16 megabytes of ram
19:58:07preglowlike h100, h115 and h110
19:58:26lostlogicxcompiling libalac with profilling causes an internal compiler error
19:59:14*preglow hands lostlogicx the golden hat of success
19:59:47 Join Lear [0] (
20:00:38Learmarkun: how come you didn't add the codepage strings to the end of the language file?
20:00:42 Join muesli- [0] (
20:01:03preglowhmm, it actually looks like libmad runs at 45mhz even with 192kbit files
20:02:50LinusNLear: my guess is that the patch was so old that the strings ended up in the middle
20:03:03_FireFly_LinusN: i don't think so
20:03:22_FireFly_because then this part of the patch gets rejected
20:03:26LearI suspect that too. But doesn't this break compatibility in some ugly way?
20:03:34LinusNLear: yes it does
20:04:04Learpreglow: done some optimization?
20:04:13preglowLear: no
20:04:52LearHm.. When I've looked at the audio thread, it seemed to almost, but not quite, be realtime at 45 MHz...
20:05:41 Join elisherer [0] (
20:08:11Slasheripreglow: indeed playlist_next(1) is the problem with queued songs..
20:08:28Slasherithere is even a comment in the playlist code "We need to delete all the queued songs".. weird
20:08:44Slasheriso it has to load the filename from disk after it has been removed from memory.. :D
20:09:33Slasherior something like that.. still the next song is already buffered
20:09:50Slasherii will put more logfs
20:13:41 Join linuxstb [0] (
20:16:32Slasheripreglow: Hmm, i don't see other clean/simple solution for this but to extend the threading in playlist..
20:17:12Slasherior maybe there is something
20:17:49 Join hardeep [0] (i=hardeeps@norge.freeshell.ORG)
20:17:53preglowthis problem hasn't always been there
20:17:56 Quit muesli_- (Read error: 110 (Connection timed out))
20:17:57preglowso something must have gone wrong some place
20:19:33Slasheripreglow: playlist_next is called before switching to the next track inside playback engine.. And if playlist_next blocks, there isn't much we can do..
20:19:55linuxstbMore patent madness:
20:20:30Slasheripreglow: and i don't know the playlist code well enough to fix it (without learning the code better first)
20:20:31lostlogicxalac.c: In function `readbits_16':
20:20:31lostlogicxalac.c:103: internal compiler error: in cgraph_expand_function, at cgraphunit.c:540
20:21:27preglowbut like i said, i don't think this always behaved this way
20:21:29preglowwhat changed?
20:22:03linuxstblostlogicx: I fixed a lot of libfaad ices by changing variables from signed to unsigned (or the other way around). Mixing the two can cause problems.
20:22:26lostlogicxlinuxstb: *nod* I'll take a look, ty.
20:22:39 Quit eli_sherer (Read error: 110 (Connection timed out))
20:22:54hardeepSlasheri: what's the problem?
20:23:27 Quit elisherer ()
20:23:29 Quit _FireFly_ ("Leaving")
20:23:38 Join eli_sherer [0] (
20:25:31Slasherihardeep: remove_track_from_playlist function inside playlist engine blocks and causes queued songs not to be gapless
20:25:39linuxstblostlogicx: Also, libalac uses -O3 which I think inlines most functions.
20:26:14LinusNwhy do we call remove_track_from_playlist?
20:26:40hardeepto remove queued tracks
20:27:00 Quit San (Read error: 110 (Connection timed out))
20:27:08lostlogicxlinuxstb: I think it's the O3 conflicting with profiling options. good call.
20:27:41 Join yngwi [0] (
20:27:54LearBtw, is libfaad realtime yet?
20:28:44hardeepSlasheri: Are you calling any other playlist functions at the same time? There's a mutex to protect playlist control file writes
20:29:12LearBtw, just commited a fix for the convbdf crash (which was caused by a bad defaultchar value), but it just struck me that perhaps convbdf should fail for such an "error". Any opinions?
20:29:35Learlinuxstb: how far from (roughty speaking)?
20:29:44hardeepother than that, the only thing I can see what could potentially block is the settings_save and that would only be called if the queued song is the first track in the playlist
20:30:25linuxstbLear: It's probably at about 80%-90% with my 128kbps test files.
20:30:39hardeepactually, correction, it would be called if it's before the "first_index"... which might not be the first one in case of shuffled dirplay
20:30:52amiconnLinusN: While it's correct that the language strings should have been at the end, it's not that big an issue in this case
20:31:08LinusNmaybe it isn't
20:31:11amiconnUnicode itself does break compatibility
20:31:17LinusNah, yes
20:31:17 Join JonSenior [0] ( I'll take the opportunity and clean up all deprecated strings, then bump the .lng version
20:32:13Learamiconn: plans on sorting the strings a bit too? (So related stuff comes close to each other, though I realize it can be hard to judge sometimes...) :)
20:32:17amiconnI'll also try to group strings together which are spread
20:32:24LearGood. :)
20:32:26Slasherihardeep: no, the problem is that disk spins up (it writes something to it)
20:32:53 Quit muesli- (Read error: 104 (Connection reset by peer))
20:33:01eli_shererWOO HOO!!! 16 colors (4bit BMPs) supported...
20:33:26hardeepSlasheri: We need to update the playlist control file with the deleted track
20:34:03preglowso queues are pretty much useless, then
20:34:19hardeepif you want to avoid disk access, yeah
20:34:35preglowwell, spinning up disk after every file played more or less makes it completely useless, if you ask me
20:34:45preglowespecially when it introduces gaps in the output
20:35:07preglowwhy isn't the playlist control file cached and only written when the disk is spun up?
20:35:16JonSeniorA few days back you announced working H300 usb, was that the end of it, or should I continue trying to disassemble the original firmware?
20:35:19LearCould be large?
20:35:48LearBut it could cache up to n changes, I guess...
20:35:49hardeepalthough, it's essentially a playlist so we're not talking more than a few kilobytes in most cases
20:36:11LinusNJonSenior: i don't think we need to disassemble that much more. have you found anything interesting in the disassembly?
20:38:56 Join _FireFly_ [0] (
20:39:19markunamiconn: I don't know why the codepage isn't saved, do you?
20:39:43amiconnNo, I think it was just forgotten to implement it
20:39:52amiconnI'll add it
20:40:32markunIt doesn't exist in the hd_bits
20:41:58markunbtw, what is the Starmap font used for?
20:42:09hardeepactually, it would pretty easy to cache the last N playlist control "commands" and only flush when needed or shutdown
20:42:15 Quit NightCat (Remote closed the connection)
20:42:26hardeepinstead of dealing with caching the whole file
20:42:27amiconnmarkun: [16:59:23] <amiconn> markun: The codepage setting isn't saved to disk. Maybe you didn't commit all unicode changes to cvs? Just asking, will fix it tonight...
20:43:00amiconnhardeep: Yes, with one (minor) drawback
20:43:16amiconnIt might be that the delayed write fails because of low batt
20:43:17markuneverything got committed, so I just forgot to add it.
20:43:34LinusNamiconn: i just added it
20:43:42LinusN(codepage saving)
20:44:27preglowhardeep: well, just cache 16 or so last entries, and the feature will be usable again
20:45:00preglowjust make sure that spinning up for rebuffering will mostly happen more often than spinup to update playlist control
20:45:26peturbeing a developer with a H340, any topic I can focus on? Don't know what missing parts are being worked on...
20:46:24hardeeppreglow: yeah, that's pretty easy... i'll change
20:46:36amiconnpreglow, hardeep: Perhaps this will become easier when the possibility to register an 'on_spinup' handler gets added (necessary for switching to file-based settings)
20:46:57hardeepamiconn: definitely
20:47:14amiconnThat's on my todo list (file based settings)
20:50:32LinusNpetur: you could continue the work with the color bmp loader
20:51:15peturwhat are the plans related to usbotg, radio and line-in (recording)?
20:51:53lostlogicx*sigh* profiling also breaks libfaad, and changing compiler opts doesn't seem to have any effect
20:52:18LinusNpetur: the plans are to make it work
20:53:15amiconnpetur: usbotg will be a very interesting (and complex) area. Afaik, nobody has started working on it yet, but iirc Bger is interested in doing some usbotg work
20:53:37peturshould I also sign up on the mailing list to stay up-to-date?
20:53:42 Part coob
20:53:48amiconnRadio and line-in should work (LinusN: do they for H3x0 ?)
20:53:55JonSeniorLinusN: I never made much progress. I've been trying to get a working emulation system up and running
20:53:58amiconns/pdif in is still missing
20:54:25JonSeniorI do now have a sandboxed windoze system (Thankyou VMWare) so I have access to a wider variety of tools.
20:54:46JonSeniorOn the off-chance that I do make some progress, are there any other areas that need looking at?
20:55:14_FireFly_JonSenior: for which tools do you need windows ??
20:55:42JonSeniorFirefly: Working emulator / simulator for the Coldfire architecture.
20:55:59JonSeniorFirefly: Currently getting my head around Codewarrior
20:55:59_FireFly_what is wrong with the linux one ??
20:56:15JonSeniorThe linux one doesn't work. Other than that... nothing! ;-)
20:56:30JonSeniorOn my machine it seg faults when attempting to run any code.
20:56:48JonSeniorIt'd also be good to have 5249 support without having to bodge it.
20:57:02_FireFly_the rb sim or a generall sim for the coldfire-arch
20:57:39amiconnmarkun: Ahahaha, default_char == 8192
20:58:08 Quit edx (Read error: 110 (Connection timed out))
20:58:21JonSeniorFirefly: General sim for the coldfire-arch
20:58:27_FireFly_JonSenior: ah ok
20:58:48JonSeniorFirefly: The plan was to run the original firmware to find out how they did things like USBOTG
20:59:28JonSeniorAnyway. I have to dash, but I'll read the logs later. Linus... if you have any areas that you would like investigating, let me know.
20:59:36JonSeniorSee ya.
20:59:37LinusNamiconn: radio and recording is not yet working on h300
20:59:40 Part JonSenior
21:00:32amiconnLinusN: I thought the radio chip is the same? And the ADC is definitely the same... Different hookup?
21:00:47LinusNamiconn: different hookup
21:01:22LinusNnothing hard, it just has to be traced on the pcb
21:02:26lostlogicxpreglow: think the profiling patch would get merged if it didn't support all codecs yet due to ICEs?
21:02:39preglowlostlogicx: why not
21:02:45preglowit's definitely better than nothing
21:02:55preglowand it's functionality we could really use
21:03:30 Join Slyck [0] (
21:03:39Slyckanyone know how to get un g-lined?
21:03:41 Nick Slyck is now known as Acksaw (
21:04:03lostlogicxkk, should be ready this afternoon sometime, patch tracker it or just give it to you?
21:04:48*eli_sherer has just uploaded the new beta version (exclussivly to MR) on ""
21:05:08preglowlostlogicx: patch tracker
21:05:12Acksawnice one eli
21:05:14preglowand tell me about it
21:05:35amiconnmarkun: One of the feature request I was able to close today was utf-8 support. The request was more than 2 years old :)
21:05:51preglowsee, who said we don't treasure our users' input
21:05:55LinusNeli_sherer: beta version of what?
21:06:13_FireFly_color-bmp support afaik
21:06:21_FireFly_and an bmp-viewer
21:06:51amiconnLinusN: The first-note skip fix can be closed, right?
21:06:59preglowit was fixed, so...
21:07:08LinusNamiconn: yes
21:07:17LinusNah, rockpaint
21:08:09lostlogicxis there something different about how wavpack is built compared to other codecs? It seems to be having trouble linking against the profile symbols.
21:08:38AcksawMy volume has gone dodgy...
21:08:47 Join solexx___ [0] (
21:08:48Acksawsaying -20 (should be 80 i think)
21:09:03LinusNAcksaw: the scale has changed from % do dB
21:09:17preglowlostlogicx: nothing i can think of, what does it say?
21:09:30AcksawAnyway to change it back?
21:09:48lostlogicxwhen it goes to link, it just can't find the symbols
21:09:57lostlogicx/home/lostlogic/rockbox/apps/codecs/libwavpack/metadata.c:62: undefined reference to `__cyg_profile_func_enter'
21:10:14_FireFly_that is an cygwin fn or not ??
21:10:36_FireFly_at least i would say cyg stays for cygwin
21:10:50lostlogicxthat may be it's origin
21:11:15lostlogicxbut it's a general purpose profiling function generated by -finstrument-functions option to GCC now.
21:11:35***Saving seen data "./dancer.seen"
21:11:45LinusNAcksaw: no
21:12:20lostlogicxpreglow: ahh, I see the problem −− the plugin API doesn't have the symbols yet, and it's in the wav2wv plugin
21:12:41kkurbjunamiconn: I have color working on the H300 stable for the main unit, but I've had to disable the Remote control. I was wondering if there's a wiki on your thoughts for implementing color support correctly as right now the functions all expect the same type for the display which is the problem with the remote.
21:12:51kkurbjunin the WPS that is
21:12:57lostlogicxguess I'll have to at least define the symbols for plugin profiling too.
21:13:03Acksawon a wps im using (San's) on his picture the title and track etc are in the middle of the screen, mine are on the left hand side with "%ac" infront, why is this?
21:13:45_FireFly_Acksaw: wps-code ??
21:13:56_FireFly_can i have it to look at
21:13:59Acksawhold up
21:14:22amiconnkkurbjun: There was some discussion going on about this, mainly TiMiD and me
21:15:15Acksawfirefly, i will pm it you
21:15:24amiconnThe idea was that we need to define the 'device capability', i.e. the capability of the most capable display if there's more than one. (b&w -> grey -> colour)
21:15:52amiconnFor the less capable display(s) we'll then need conversion function(s)
21:15:58 Quit linuxstb ("CGI:IRC (Ping timeout)")
21:16:11Acksawmy ears.
21:16:25Acksawi just paused a song and got 3 short blasts of the song...
21:16:48Acksawwhich hurt.
21:16:52LinusNAcksaw: update to the bleeding edge
21:16:54kkurbjunso you're suggesting to store the image data in the main unit's display and then scale down to the sub displays?
21:17:17amiconnImages should always be stored in the native LCD format
21:17:19Acksawthanks linus
21:17:36amiconnThe bmp loader will have to support the native format of all LCDs, plus monochrome
21:17:58 Join lamed [0] (
21:18:16amiconn'all' means all LCDs present in the respective target of course
21:19:12 Join linuxstb [0] (
21:19:14lamedamiconn: home?
21:19:15peturheh, I'm away from my system for 10 minutes and somebody already has colour bmp working...
21:19:58kkurbjunahh, so different typedefs for the main display and then the sub displays? I was thinking of making the bmp loader output based on the display depth, only problem is how to decide which display to load the data for
21:20:20kkurbjunI guess taht could be fed as a parameter
21:20:21_FireFly_maybe with an parameter ??
21:20:36kkurbjun: )
21:20:43amiconnkkurbjun: The bmp loader needs an additional parameter
21:20:49 Quit solexx_ (Read error: 110 (Connection timed out))
21:21:48*amiconn was a bit slow
21:22:06amiconnkkurbjun: What typedefs, btw?
21:22:19kkurbjunthe display format
21:22:31kkurbjunI had been converting from chars to fb_data
21:22:42amiconnHmm, where?
21:22:45kkurbjunfb_data a typedef for the main display
21:22:50amiconnI know
21:22:57kkurbjunok, bmp loader
21:23:02amiconnAh, yes
21:23:05kkurbjunthe wps code
21:23:13kkurbjunin the image buffers
21:24:18 Quit linuxstb ("CGI:IRC (Ping timeout)")
21:24:26lostlogicxwhat's teh equivalent to codeclib.c in plugins?
21:24:40kkurbjunso gui_img and wps_data
21:25:00kkurbjunthat's where I was running into a bit of trouble though because of the multiple display issues
21:25:21kkurbjunand I guess that would be compounded if remotes ever came out in color and b/w displays
21:26:17amiconnThere should be no problem yet, because the remote display is b&w. It will become more of a problem on units with 2 non-b&w displays which have different pixel formats
21:26:41 Quit Acksaw ()
21:27:14kkurbjunwhat about H300 w/ color and b/w
21:27:42lamed_firefly_ didn't you liked numrical options jumping from lowest to highest setting?
21:27:46amiconnThat's what I meant. One b&w and an arbitrary other display isn't a problem
21:28:09_FireFly_lamed: not really i was only the patch creator :)
21:28:10preglowbut ok, i've got to start thinking about leaving, later all
21:28:11amiconnTwo non-b&w displays are, but we don't yet have such a target
21:28:42_FireFly_lamed: have you any "problems" about this ??
21:28:51_FireFly_at least for the volume this is a problem
21:28:58 Quit eli_sherer (Read error: 110 (Connection timed out))
21:29:06kkurbjunwell, true, it's not a problem if the wps is kept in black and white
21:29:17_FireFly_lamed when the volume jumps directly from 0 to full
21:29:49amiconnkkurbjun: The remote wps will definitely stay b&w
21:29:55amiconnThe main wps should not
21:30:06amiconnStill I don't see the problem
21:30:17kkurbjunso should there be two different types of wps_data?
21:30:37kkurbjunright now I've converted the wps_data to store in the main unit's format
21:30:43kkurbjunbut the remote is b/w
21:30:47_FireFly_could be useful
21:30:56kkurbjunor shorts vs. chars
21:31:01lamed_firefly_, at first i thought it was a bad idea but it gives you the ability to check the two opposites of a value: say for bass & treble it's a very good idea. iirc it stoped on the last value before scrolling back hasn't it? -if it hadn't then stopping at the end might be a good compremise.
21:31:17kkurbjunwhich leads to problems feeding that data to display
21:31:26_FireFly_lamed it has
21:31:31kkurbjunstruct wps_data
21:31:32kkurbjun#ifdef HAVE_LCD_BITMAP
21:31:32DBUGEnqueued KICK kkurbjun
21:31:32kkurbjun struct gui_img img[MAX_IMAGES];
21:31:32kkurbjun fb_data img_buf[IMG_BUFSIZE];
21:31:32***Alert Mode level 1
21:31:32kkurbjun fb_data* img_buf_ptr;
21:31:34kkurbjun int img_buf_free;
21:31:51kkurbjunthat's what I've changed in ... gwps.h
21:32:01_FireFly_but linuxstb and some others wanted to have it not wrap around
21:32:39_FireFly_lamed: but what is when someone doesn't hold the button while change the volume
21:32:57kkurbjunam I incorrect in assuming that both displays use the wps_data struct?
21:33:01_FireFly_then he will become shocked when the volume jumps directly from 0 to full
21:33:06amiconnThe wps shouldn't need to know the actual bitmap format, just some parameters
21:33:19_FireFly_kkurbjun: yepp
21:33:39_FireFly_amiconn: yes but what is with the image-buffer
21:33:55_FireFly_the type must be different
21:34:33_FireFly_afaik for b/w displays the image-buffer is char but for color it is fb-data
21:34:33amiconnIt's just data...
21:34:54_FireFly_hmm right
21:34:54lamed_firefly_ well then, only my 0.02$. something way more important is making the list pointer to be kept at the 2/3 of the middle screen... now it scrolls down to the one line before last, and the pgup/down doesn't acts the same when scrolling up and down...
21:35:17lamed_firefly: i don't recal volume at the wps delt the same way?
21:35:31kkurbjunwhat struct does the remote use?
21:35:39_FireFly_kkurbjun: ??
21:36:13kkurbjunright now I was storing the data in the format of the player as that was what the loader was originally doing.. I think to the img_buf
21:36:29kkurbjunthe img_buf doesn't contian raw bmp data
21:36:58kkurbjunsorry, was I correct that both displays use wps_data
21:37:05_FireFly_kkurbjun: both screen uses the struct declared in gwps.h
21:37:58amiconnBah, hidden surprises :(
21:38:04 Join linuxstb [0] (
21:38:28_FireFly_lamed: what ?
21:38:42kkurbjunso the problem I had was that img_buf was not the right type.. I guess I could store a short as 2 bytes and then have the routine in lcd_16.c account for that
21:38:45amiconnIn the code...
21:38:56CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
21:38:56*amiconn is cleaning up the languages
21:39:07kkurbjunumm, but it's still not the right format for the remote
21:39:50_FireFly_would it work when the data is stored as short and converted to char when the data gets used ??
21:40:19_FireFly_so only fb_data is the type be used in wps_data
21:40:35amiconnI'd simply store the data in a char array. Just make sure the 16 bit data is properly aligned in the buffer
21:40:40_FireFly_and an additional var which defines the pixel-format for the display
21:40:48_FireFly_or that
21:41:33***Alert Mode OFF
21:41:59kkurbjunok, I think I see a way that might work
21:42:06kkurbjunI'll just give it a shot
21:42:21kkurbjunthanks for the thoughts
21:42:29lamed_firefly_ you just haven't replied to my q's, but nm it was a boaring subject :)
21:43:26lamedanyone willing to help me out on debugging something?
21:44:43lamedi guess i'm doomed forever
21:45:56peturdebugging what?
21:48:07lostlogicxfirst run test with profiling integrated into the build system *crosses fingers and toes*
21:55:09 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
22:01:05_FireFly_lamed: about the list-pointer problem it seams that a define is used to define on which line the scroll should start when moving the line-selector
22:01:13 Join San [0] (
22:01:15_FireFly_but this is uses for both screens
22:01:40_FireFly_so on the remote is looks bad when using a values higher then 2 and using a big font
22:03:31 Quit Lear ("Chatzilla [Firefox 1.5/undefined]")
22:06:43 Join Midgey34 [0] (
22:07:26lostlogicxworthy of note: trying to profile codeclib caused player lockup for the time being.
22:10:14_FireFly_lamed: but i have no glue currently how it doesn*t work correctly when page-down
22:10:33_FireFly_is used
22:10:34_FireFly_to scroll down quickly
22:15:21lostlogicxwhy is there no mp3 or file?
22:15:56lamed_firefly_ it's not too complex, and we can check the _reallly_ old sources to get the idea. but i don't find the time nor the urgence to do that...
22:18:53lostlogicxmuch better, thanks. profiling other codecs also works now (at least mad and vorbis) I'm submitting my patch.
22:19:39 Join NightCat [0] (
22:23:45 Part LinusN
22:24:22 Join idanm [0] (i=idan@
22:25:10idanmanyone here?
22:28:38 Join andy [0] (
22:31:13_FireFly_lamed: ?
22:31:43lamed_firefly_ i was talking to idanm
22:33:11idanmhas anyone have the h300 lcd remote working with rockbox?
22:33:22idanmor knows how to do it?
22:33:45lamedidanm: are you using the hebrew language file on rockbox?
22:33:56NightCatYes hi is :)
22:34:08NightCat(I mean hi=he...)
22:34:30lamednightcat... how do you know?
22:34:53NightCatFrom forum :)
22:35:12NightCat"it works great!" - He said :)
22:35:26NightCatabout hebrew I think :)
22:35:41NightCatidanm? Are you here?
22:35:56NightCatI have an H300 LCD :)
22:36:13NightCatIt's not H300's one but it works.
22:36:37lamedno i was talking about the lanugage file
22:37:10idanmI use the hebrew files
22:37:37lameddo you use the hebrew language translation?
22:37:39idanmNightCat: do u have the h100 remote?
22:37:50idanmI can check the trans files, sec
22:38:04lamedidanm never mind
22:38:10 Quit _FireFly_ ("Leaving")
22:38:13idanmwhy do u ask?
22:38:25NightCatidanm: I have a remote from iRiver IMP-550 (with LCD) and it works :)
22:39:05idanmthe h300 lcd remote display is good, but the keys mapped wrongly.
22:39:31 Join edx [0] (
22:39:38lostlogicxprofiling patch submitted to tracker.
22:39:43NightCatlamed what about bug with russian lang in quick menu?
22:40:00lamedbecause i wrote the hebrew translation and i was wondering if anyone is ever in the world using it.
22:40:03idanmhebrew traslation ok.
22:40:28idanmhmm.. I still prefer the english :)
22:40:41NightCatidanm, what about mapping?
22:41:19idanmif I press the play button, then it stops, left is up, and the rest does't work.
22:41:25NightCatI think they give you a patch...
22:41:47NightCatAre you use it?
22:41:54idanmi'm not sure how to use it... why dosen't it in the main build?
22:42:15idanmcan u help me apply and compile it? i've never done it
22:42:21lamednightcat: it will take some time. it's problematic because of the internal font that rockbox is using. i have the same problem with hebrew...
22:42:42 Quit hshah (Read error: 110 (Connection timed out))
22:43:07NightCatlamed: OK.
22:43:08idanmwhat problem?
22:43:34NightCatidanm: About compiling: how is your english language?
22:44:10NightCatidanm: take the patch for first:;topic=2079.0;id=318
22:44:41NightCatidanm: then go here
22:45:23NightCatThere is a simple manual how to do what you want :)
22:45:29 Join einhirn [0] (
22:45:39NightCatI use it to make my own simulator :)
22:46:00 Join Gacel [0] (
22:46:08Gacelhi everybody
22:46:13NightCatHi :)
22:46:22NightCatidanm: Just do it step by step :)
22:46:31Gaceljust installed rockbox on my ancient ihp120... great work :)
22:46:31idanmtakka said in the forum that I'll need to re-write the table
22:46:34NightCatidanm: Note
22:46:36 Join RotAtoR [0] (
22:47:13NightCatidanm: For first try to install this patch...
22:47:33Gaceli might do a .wps or help some programming (if i find some time between my other projects)
22:48:08idanmis the simulator binary available?
22:48:13NightCatGacel: I think it will be good if you help with programming :)
22:49:01GacelNightCat: why? are programmers needed? (looks quite active the project)
22:49:02NightCatidanm: it will be binary available when you compile it from source (it's in that page too) :)
22:49:29NightCatGacel: Yes, cause there is still much to do!
22:50:08NightCatGacel: Look for ex. on the feature requests
22:50:08NightCatbug reports
22:50:14idanmI think you gave me a link to a wrong place.
22:50:36idanmdid u mean the " Setting up a cygwin Rockbox development environment" page?
22:50:48NightCatGacel: for ex. we need a rockboy upgrade :)
22:51:03Gaceli am just scanning all the docs ;)
22:51:27NightCatidanm: No I mean "The Simplified Guide To Compiling The Rockbox Source Code".
22:51:36Gaceli am have some projects which have to be finished first, so no quick involvement
22:52:08NightCatGacel: Ok, come back when you could :)
22:52:50Gaceli will lurk a bit :)
22:53:42NightCatidanm: How is it?
22:54:10idanmdon't I need cygwin?
22:54:31NightCatidanm: I think, no...
22:55:23NightCatidanm: There is the section on that page named "What You Will Need"
22:56:34NightCatidanm: ok, good luck, I must finish my study work...
22:58:02 Quit Febs ("CGI:IRC (EOF)")
23:07:45 Quit dpassen1 ()
23:08:17idanmrbconf command doesn't work, any ideas?
23:08:30Bagderidanm: yes, use configure directly
23:08:50Bagderrbconf is evil
23:09:06 Nick wubbla_ is now known as wubbla (
23:09:16idanmshould I edit this file?
23:09:27Bagderwhat file?
23:09:44Bagderrun it
23:10:17Bagderit'll prompt you for what you want to build
23:10:18idanmshould the build folder be under the source folder or home/guest/build?
23:10:42 Join LinusN [0] (
23:10:57Bagderidanm: I mostly make the build folder next to the app and firmware ones
23:11:17Bagderbut you can put it next to the source too IIRC
23:11:37***Saving seen data "./dancer.seen"
23:12:46idanmI maked simulator, how do I use it?
23:12:59NightCatrun it :)
23:13:10Bagderrun make install first
23:13:53NightCatOh, sorry, I think he mean that he already done this :)
23:14:10idanmwhat's the diff between the normal and devel builds?
23:14:24NightCatDevel for developers :)
23:14:39NightCatI used normal.
23:14:39Bagderdevel lets you select a few more options
23:15:03amiconnBah, these language files are really longish
23:15:27idanmI need to applay the remote patch now..
23:16:17 Join Paul_The_Nerd [0] (
23:16:17idanmwill I need to compile the whole source again afterwards?
23:16:33NightCatI think...
23:16:38Bagderyou run make again
23:16:40peturmake will only build what it needs to
23:16:49idanmhow much time is it to build all?
23:17:05Bagderidanm: do 'make clean all' and you'll find out
23:17:16NightCatAh, studying is a good think :)
23:17:26andyanyone looked into the release source for Iriver PMP-1xx ?
23:17:27NightCatIRC is a good thing to :)
23:18:12idanmcan I cancel the making for applaying the patch, then delete the build folder and start over?
23:18:32 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
23:18:38peturjust cancel, do the patch and make again
23:19:21idanmcontrol+c ?
23:19:23NightCatPeople, could some of you help me with WPS?
23:19:31NightCatidanm Yes
23:19:41idanmok so how do I applay remote_type.patch ?
23:21:52NightCatPeople??? Some help with WPS, please?..
23:22:42Gaceli dont think i can help ;)
23:22:48Paul_The_NerdWhat's the problem, NightCat?
23:23:30NightCatIt blinks...
23:23:52Paul_The_NerdOn H300?
23:23:56NightCatI have some picts in my WPS and they blinks...
23:24:03NightCatYes H320...
23:24:09Paul_The_NerdIt's a bug
23:24:14Paul_The_NerdA known bug, rather
23:25:01idanmwhy do I need to set the language file in configure if can change it on the rockbox menu?
23:25:23Midgey34anyone know who id working on the vertical rockblox code and why it was never committed?
23:25:28Bagderidanm: you select the built-in one
23:25:40idanmi.e. default ?
23:26:25NightCatPaul_The_Nerd: It's bad... I want a normal WPS without blinking...
23:26:34BagderMidgey34: I think it was never committed just because no one ever gave it attention enough
23:26:57idanmis the simulator has a built in build or it uses external build ?
23:27:05Paul_The_NerdNightCat: Remember, H320 Rockbox is in a VERY early state.
23:27:24BagderNightCat: we'll appreciate you fixing the bug! ;-)
23:27:26Paul_The_NerdMusic playback isn't even perfect yet, there are stability issues, I don't believe the processor is running at full speed yet...
23:27:27NightCatPaul_The_Nerd: I know, just my dreams :)
23:27:54Midgey34Badger: Alright I was using it as a base to add color and a few more features to the rockblox code and wanted to make sure my work wasn't completely worthless
23:28:10NightCatBagder: Yes :) Only if RB moves from C to Delphi or PHP :)
23:28:34 Join b0br [0] (
23:29:14NightCatPaul_The_Nerd: It runs on 90 instead of 140 MHz...
23:29:44Bagderwell, the h1x0 version only runs at 124
23:29:47NightCatBagder: I have to learn C but have not much free time yet... :(
23:30:28 Join Jungti1234 [0] (n=jungti12@
23:30:29NightCatBagder: Please answer to idanm? :)
23:30:29idanmdo u think it be possible to connect usb keyboard to the h300 ?
23:30:57NightCatidanm: I think yes but if someone impliment it...
23:31:09NightCatJungti1234: Hi.
23:32:10NightCatMidgey34: Could you add some colors to snake game?
23:33:14markunJungti1234: I updated the korean language file
23:34:07Midgey34NightCat: Snake or Snake2?
23:34:50idanmwhy does it says 'no wps module present'?
23:35:04Bagderidanm: 'cvs co wps'
23:35:32 Join eli_sherer [0] (
23:35:37eli_shererSan: are you here?
23:36:11idanmI downloaded the tar.gz file
23:36:40Bagderidanm: daily source?
23:37:03NightCatMidgey34: snake 2 :)
23:37:28NightCatMidgey34: And do it fit the H300 screen :)
23:37:40NightCatMidgey34: Please :)
23:37:58idanmbleeding edge
23:38:00Jungti1234hi eli_sherer
23:38:08lostlogicxidanm: I think it should be possible to get a USB HID driver for the H340, but don't quote me on that. There is also _no plan_ to do so at this time that I am aware.
23:38:12 Join San||Away [0] (
23:38:42idanmI build it successfuly, but the remote still doesn't work right
23:38:48 Join Amar [0] (
23:39:09NightCatidanm: Have you change the remote option in the menu?
23:39:29b0brpls, do you know if anybody works on recording for h3xx at the moment?
23:39:47Bagderb0br: I'm quite sure no one does
23:39:52idanmhmm... sec
23:39:57Bagderfeel free to step forward
23:43:25idanmnow it works !!
23:44:06NightCatidamn: Congratulations!
23:44:18b0bri can't do that, but my good friend is pretty good at c/c++ and he's buying h340 in few days. he's also musician and promised to write enhanced metronome plugin with dialects features etc..
23:44:46idanmI didn't understand why isn't it on the main build ?
23:44:50NightCatb0br: it will be very good :)
23:45:52NightCatidanm: Hm: "The patches are not made by the Developers so there is a chance that they could contain bugs..."
23:46:43NightCatidanm: I think you have to ask LinusN... I think he could insert this patch...
23:47:01Bagderall cvs committers _can_ insert patches
23:47:22 Join mirak_ [0] (
23:47:33idanmwhere is the list of csv committers?
23:47:35Bagderit is often a question of lack of time
23:47:36NightCatYes but Linus works on H300,,,
23:47:45LinusNamiconn: there?
23:48:01BagderNightCat: on the bootloader and hw setup problems, yes
23:48:24NightCatLinus, what about inserting remote patch to the build?
23:48:28mirak_idanm: I want a joystick to play snake :D
23:48:29BagderI don't think you should expect LinusN to mess around with everything related to h300
23:48:46b0bryea :) but i would like to persude him to work on recording feature for h3xx, but he doesn't want to mess with low level code. does anybody know what is needed to be done to make h1xx recording work on h3xx?
23:48:53LinusNNightCat: the patch that lets you use the h300 lcd remote?
23:49:15NightCatLinus: Yes.
23:49:32mirak_is there a pda plugin or something like that ?
23:49:34NightCatBagder: Actually I don't know who else work on H300 :)
23:49:43LinusNdoes it work for both remote types?
23:49:45mirak_an agenda or something
23:49:59BagderNightCat: you don't have to "work on h300" to commit a patch
23:50:01idanmhey LinusN! can u insert the h300 remote patch to csv?
23:50:12lamedok this is me signing off for today. goodnight
23:50:16 Quit lamed ("CGI:IRC")
23:50:29 Quit mirak_ (Client Quit)
23:50:57NightCatlamed: goodnight :)
23:51:02 Quit akaidiot (Read error: 110 (Connection timed out))
23:51:14 Quit San (Read error: 110 (Connection timed out))
23:51:32NightCatLinus: this patch just help idanm :)
23:51:39LinusNdoes the patch work correctly for both lcd remote types?
23:52:08NightCatidanm: how does it works?
23:52:08idanmI checked the H300 lcd and non-lcd, the non-lcd still seems as if it always on hold
23:52:25LinusNnon-lcd is totally different
23:53:06idanmit has a menu for choosing the non-lcd remote, not working.
23:53:17amiconnLinusN: somewhat
23:53:50LinusNamiconn: it looks like the h300 cpu frequency instability is in fact the spurious OFF event bug
23:53:53NightCatMaybe there have to be something keymapping like in games?
23:53:58idanmabout the h300 lcd, I would change the up-down keys to replace the left-right on browsing. and set left-right to up-dir and go into a dir.
23:54:09amiconnLinusN: Uh? Do you have an idea what causes it?
23:54:33LinusNnot yet, but it might be a bug in the button driver, or a wild pointer overwriting the button status
23:55:02amiconnThis is really puzzling; why OFF, and OFF only?
23:55:15amiconnI also never observed such behaviour on Ondio
23:55:46NightCatOk, people, have to go... Goodnight :)
23:56:12*amiconn ???
23:56:37 Quit NightCat ()
23:58:27amiconnLinusN: Does the FM recorder also suffer from spurious OFF events?
23:58:36LinusNi believe so
23:58:50idanmgood night eveyone! thanks for your help !!!

Previous day | Next day