Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2009-06-24

00:00:53notlisteningjust an idea though
00:01:02 Part CaptainKwel_
00:01:25notlisteningseen it implemented else wwhere and it was useful but that device did not have speech feedback
00:04:43 Quit bubsy ("Panic.")
00:06:01 Nick tmzt_ is now known as tmzt (
00:10:17notlisteningdomonoky, can you direct me to the place on the wiki where it instructs you to make you own voice menu files as i want to try and test the encoder independently from rbutil
00:12:26domonokyyou have to edit and maybe configure to add your tts engine before, ofcourse :-)
00:13:49notlisteningthanks it was just to test the encoder as it is failing and wanted to get the output on the command line from the encoder on the files tthat are failing
00:14:32domonokydid you check if the wavfile really is created ?
00:14:59notlisteningyup wav file is created with correct fileame and contents
00:15:18notlisteningand the .talk file is created but from there i am a little bit stumped
00:15:53notlisteningchanging the remove extension option also stops it from failing :P
00:19:24 Quit bmbl ("Bye!")
00:20:44 Part safetydan ("Leaving.")
00:21:14 Join safetydan [0] (n=deverton@rockbox/developer/safetydan)
00:21:56 Quit lymeca ("KVIrc 3.4.0 Virgo")
00:22:18 Quit Cory` (Read error: 54 (Connection reset by peer))
00:22:36 Join bubsy [0] (i=Bubsy@
00:23:53 Quit bertrik ("De groeten")
00:26:04 Quit Rondom ("Ex-Chat")
00:26:44 Join Cory` [0] (
00:26:56CIA-71New commit by bagder (r21481): support sdl as arch, and make "lethal" errors stop the looping in
00:30:12rasherBagder: probably should have scrubbed
00:30:24Bagderme fix
00:30:36saratogai've still got that running from the other day, do i need to restart if i want to be available to test?
00:30:58Bagderno, it still participates fine when I start the server
00:31:22CIA-71New commit by bagder (r21482): eeek, clear off my dummy data
00:31:54Bagderbut now we can add 'sdl' as arch as well
00:31:58 Quit AlexP (Read error: 110 (Connection timed out))
00:32:11Bagderand it would be nice with a few clients having that enabled
00:32:18*gevaerts updates
00:34:21gevaertsBagder: should now have sdl as well
00:34:44BagderI started one myself too and it has sdl as well
00:34:46gevaertsand a different password. I hope you don't check that yet
00:43:46JdGordon|Bagder: if you want another client feel free to use rbclient's account on my comp to do whatever you want with...
00:44:26Bagderthanks, I'll give that a go tomorrow or so
00:45:09Bagderhaving 9 new clients still is slower than the 20+ running in the old system ... :-)
00:45:27JdGordon|cut down the amount of builds to run?
00:45:33JdGordon|duh :p
00:45:34Bagderyeah, I'll do that
00:45:39linuxstbRemove all ipods...
00:46:05Bagderbut with sdl just added, it'll be great to see a full run complete first
00:46:26 Quit domonoky (Read error: 104 (Connection reset by peer))
00:46:28***Saving seen data "./dancer.seen"
00:46:47Mikachudo you have any stats on how many of each are actually downloaded regularly?
00:46:52rasherBagder: It doesn't seem to be connecting here
00:47:04Bagderno, the server isn't running
00:47:19BagderI'm done for tonight
00:47:24rasherThat would explain it
00:47:24 Quit notlistening ("Leaving")
00:48:04BagderMikachu: yeps, and it's quite a lot of the most popular ones at least
00:48:43Bagdersansa e200, 7400 times during June so far
00:49:00Bagderoff the build server this is
00:49:25BagderI don't have any stats for the download mirrors
00:50:08*Bagder runs off to bed
00:54:19 Join fenugrec [0] (
00:54:27Mikachuhm, vmware image downloads have stayed fairly constant from me, around 40-50 unique ips per month
00:54:49scorche|shspeaking of, when i get back, i should probably try to get the new one down in size and replace it..
00:55:02Mikachumake sure you let me know so i can update mine
00:55:27fenugrecI formatted my Fuze (form the OF), compiled a bootloader, copied fuzeA.bin, extract to the fuze ; on bootup Rockbox show but says "loading firmware... file not found"
00:55:27Mikachuit's from february '07 (the file timestamp) :)
00:55:36fenugrecThis is with SVN 21478
00:56:45fenugrecwhat's missing ?
00:59:51 Quit robin0800 ("Leaving")
01:00:06fenugrecwould running fsck.vfat from linux check more things than chkdsk.exe on windows ?
01:01:36Tornenot particularly.
01:01:45Tornei couldn't tell you they were definately equivalent
01:01:53Tornebut i'd be stunned if there was much in it.
01:01:59 Join robin0800 [0] (
01:02:07TorneFAT is just not that complicated, there's only a small number of well-understood things to check
01:02:50Torneit wouldn't hurt though if you had a linux box around
01:02:59fenugrecI'd have to reboot, is all
01:03:31fenugrecfor some reason I can't remember, I'm compiling Rockbox from colinux under winXP ...
01:03:43Mikachufsck.vfat is very bad
01:03:53Tornechkdsk in NT-based windows is a perfectly good check, anyway.
01:04:01Torneunlike old chkdsk :)
01:04:04Mikachuwhen you have errors and run it once and fix everything, running it again still finds errors
01:04:14Mikachulast time i had to run it three times
01:04:33TorneMikachu: oh, it's just not very bright about how it checks, some problems mask other problems and it dosn't re-evaluate it
01:04:47Torneif you run it until it says it's ok it's probably just as good
01:04:57Mikachuvery trustworthy :)
01:05:16fenugrecok... but I guess there probably was no problem since it was freshly formatted, and chkdsk didn't find a thing
01:05:23fenugrecso what could the problem be ?
01:05:57Torneunfortunately i know nothing about how the fuze boots so i can't really help you there :)
01:07:57fenugrecit's not the most descriptive of error messages, either
01:08:47Tornepresumably it's the rockbox bootloader printing that, so it's gotten that far at leat
01:09:02fenugrecindeed, and I can still boot to the OF
01:09:07fenugrecso I'm not bricked yet
01:09:10Torneso it must be the main rockbox binary it
01:09:13Torneit's unable to load.
01:09:35fenugrecit's the third bootloader + rockbox build I do, and the first to fail... I just don't understand
01:10:16fenugrecI notice there's a file, .rockbox/bootlader-fuze.sansa
01:10:24Mikachuthe others were on other devices?
01:10:25fenugrecisn't that redundant ?
01:10:32fenugrecNo, this same Fuze
01:10:35 Quit at0m (Read error: 60 (Operation timed out))
01:10:37gevaertsdid you run make clean properly?
01:10:59fenugrecstarted a new build/ directory as well
01:11:19rasherMaybe you can just use an older bootloader?
01:11:23Torneis this based on significantly newer code than last time? :)
01:11:49fenugrectotally too late, I dont' have any of the older ones. I'd have to revert to a previous code revision
01:12:06Tornehow old was your previous revision, though? :)
01:12:24fenugrechm, was around 211xx
01:12:31fenugreca few weeks
01:12:37rasherThen it's not really too late, is it? Honestly though, I don't know anything about the Fuze, so maybe you do need an updated bootloader.
01:13:09fenugrecyes, I was trying with a fresh-built BL from 21478
01:13:49Tornetry the release bootloader? or the new bootloader with a release build?
01:14:04gevaertsfenugrec: I definitely don't have a .rockbox/bootlader-fuze.sansa file on a fresh build
01:14:04fenugrecFuze isn't released yet I fear
01:14:17Torneah. too many platforms for me to keep track :)
01:14:41fenugrecok, maybe my SVN tree is hosed
01:14:48*gevaerts tests some more
01:15:13fenugrecgevaerts: do you run tools/configure after building the bootloader and before running "make zip" / fullzip ?
01:15:42gevaertsif you don't, you get a bootloader build
01:15:46Torneare you using the same build directory for bootloader and regular?
01:15:49fenugrecas I thought
01:15:50Tornemaybe you shouldn't? :)
01:16:07fenugrecTorne: maybe not, but it "used" to work
01:16:15gevaertsif I run make zip in a bootloader dir, I do get a that has .rockbox/bootlader-fuze.sansa in it
01:16:16Tornewell yah, but makefils change
01:16:28Torneand it's harder to make a mistake that way
01:16:47Torneand i can't think of a reason *not* to use two build directoryes :)
01:17:29gevaertsfenugrec: try make clean again and configure again. I suspect that you accidentally made a bootloader build instead of a normal build
01:17:39fenugrecthe only reason to use one build directory is "not knowing any better", hehe..
01:18:30fenugrecgevaerts: that's probably what happened on my first try, but I'm sure I did a Normal on the last try. I'll try two build directories right away.
01:18:44fenugrecdoes configure affect anything outside the build/ directory ?
01:18:54fenugreci.e. taint the source tree
01:18:58gevaertsfenugrec: it's also always a good idea to check if make clean did indeed clean up things
01:19:11Torneconfigure doesn't, make does but only under tools which should be the same for both
01:20:28 Quit timc (Remote closed the connection)
01:20:45fenugrecGoodness, are we already at #21482 ?
01:21:09Tornezagor keeps changing the new build system :)
01:23:08fenugrec"cat: rockbox/buildRB/apps/features: No such file or directory" (this is during "make fullzip")
01:23:37fenugrecbut I do get the file
01:23:43gevaertsyou're still doing a bootloader build
01:24:21fenugrecgevaerts: possible, but if so there's a problem : I'm using two different build dirs, and I re-run configure
01:24:36gevaertswhich files do you have left after make clean?
01:24:39fenugrecof course specifying a (N)ormal build
01:24:55 Quit jgarvey ("Leaving")
01:25:10fenugrecmake clean in buildRB/ leaves {Makefile autoconf.h}
01:25:38fenugrecmake clean in buildBL/ leaves {Makefile autoconf.h fuzeA.bin-21482}
01:25:43 Quit JdGordon| ("Miranda IM! Smaller, Faster, Easier.")
01:25:47fenugrec(last file is the one I created with mkamsboot)
01:26:04gevaertsok. Have a look at Makefile in both. Do they have -DBOOTLOADER in EXTRA_DEFINES?
01:26:51fenugrecBL/Makefile : yes RB/Makefile : no (EXTRA_DEFINES is empty)
01:27:53gevaertsDo you have a bootloader-fuze.sansa in RB after running make there?
01:28:41fenugrec(earlier I was using one build/ for both BL and RB - that must have been (one of) the problems
01:29:14 Quit Thundercloud_ (Remote closed the connection)
01:29:46gevaertsDo you have rockbox.sansa there?
01:30:22fenugrecnot in RB/
01:30:32gevaertsso what do you have there?
01:30:48fenugrecnow {Makefile autoconf.h make.dep}
01:30:59gevaertsdid you run make there like I asked?
01:31:07fenugrecjust did
01:31:19fenugrecrather, I ran "make fullzip"
01:31:21fenugrecnot "make"
01:31:27gevaertsthat's *not* the same
01:31:41fenugrecok, take 2
01:35:03fenugrecomg, rockbox.sansa == rockbox OS
01:35:09fenugrecthat would be a problem, not having it
01:37:45fenugrecwell, it works now
01:38:17gevaertsgood :)
01:38:35fenugrecthanks for your time guys.
01:39:00fenugrecWould it be a nice thing to have the fullzip target depend on "make"
01:39:16rasherThere's some reason it doesn't. I don't remember it
01:39:18gevaertsthis has been debated several times
01:39:25fenugrecI probably won't be the last clueless user to forget
01:39:36fenugrecah, ok
01:39:55Mikachuyou can run "make all fullzip"
01:40:03gevaertspeople want to be able to run make zip if some plugins failed. This is pretty important when doing big all-over-the-place changes or new ports
01:40:44fenugrecbut if fullzip depends on 'all', you can run "make all zip" if you want..
01:41:16 Quit Cory` ("( :: NoNameScript 4.22 :: )")
01:41:33fenugrecanyway, if it's been discussed already then so be it, it's not a big problem anyway
01:42:00 Quit robin0800 ("Leaving")
01:42:05 Join Cory` [0] (
01:46:16 Nick fxb is now known as fxb__ (
01:47:45fenugrecxcopy *.wv f:\MUSIC /P /S
01:47:58*fenugrec types in the wrong window again
01:48:09 Quit Cory` ("( :: NoNameScript 4.22 :: )")
01:48:34Mikachuyou typed all that without looking? :)
01:48:41 Join Cory` [0] (
01:53:37 Quit evilnick ("Page closed")
02:02:11 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
02:05:07 Quit PaulJam (Nick collision from services.)
02:05:14 Join PaulJam_ [0] (
02:06:59 Part wincent ("Kopete 0.12.7 :")
02:07:26 Join kkurbjun [0] (
02:12:05 Quit PaulJam_ (".")
02:18:21 Join slam_ [0] (n=grndslm@
02:19:46 Quit tessarakt ("Client exiting")
02:19:50 Quit HellDragon (Client Quit)
02:20:04 Join HellDragon [0] (
02:20:35 Quit fenugrec ("Leaving")
02:24:51 Join fenugrec [0] (
02:25:16fenugrecwow, this is awesome - I'm playing wavpack files with 0% boost rate on my Fuze
02:28:21fenugrecalthough 62MHz is the lowest clock ...
02:30:00 Quit BlakeJohnson86 ("Leaving.")
02:46:32***Saving seen data "./dancer.seen"
02:49:33 Join BlakeJohnson86 [0] (
02:49:50 Quit BHSPitMonkey (Remote closed the connection)
02:52:56 Join webguest64 [0] (n=4b5ce8b8@gateway/web/cgi-irc/
02:55:04 Quit Cory` ("( :: NoNameScript 4.22 :: )")
02:55:25 Quit Rob2223 (Read error: 110 (Connection timed out))
02:56:08webguest64i am having trouble with usb transfers using my sansa e200 and windows 9
02:56:14webguest64i mean windows 7
02:56:14 Join Cory [0] (
02:56:45 Nick Cory is now known as Guest52664 (
02:57:40JdGordonare you an admin?
02:58:20webguest64the transfers just freeze at some random point, also freezing either foobar or windows explorer(whatever is being used to transfer)
03:00:48webguest64i'm using rockbox 3.3
03:05:58 Quit webguest64 ("CGI:IRC (Ping timeout)")
03:06:38 Join webguest46 [0] (n=4b5ce8b8@gateway/web/cgi-irc/
03:10:02 Quit webguest46 (Client Quit)
03:18:24 Join at0m [0] (
03:20:29 Join dmb [0] (n=dmb@unaffiliated/dmb)
03:21:58 Join cool_walking_ [0] (i=cb3b81c3@gateway/web/freenode/x-912a1877a3d6d270)
03:22:08 Quit kadoban (Read error: 54 (Connection reset by peer))
03:25:00 Join kadoban [0] (
03:32:25 Quit faemir (Remote closed the connection)
03:37:29 Quit fenugrec (Read error: 110 (Connection timed out))
03:43:30 Nick Guest52664 is now known as CoryClaxon (
03:47:35 Quit pixelma (Nick collision from services.)
03:47:35 Quit amiconn (Nick collision from services.)
03:47:36 Join pixelma_ [50] (n=pixelma@rockbox/staff/pixelma)
03:47:38 Join amiconn_ [50] (n=jens@rockbox/developer/amiconn)
03:47:53 Nick pixelma_ is now known as pixelma (n=pixelma@rockbox/staff/pixelma)
03:47:56 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
03:49:35 Join fenugrec [0] (
03:50:50 Quit fenugrec (Client Quit)
03:52:01 Join timc [0] (n=aoeu@
04:16:50 Join KBH [0] (
04:20:35 Join Hillshum [0] (n=4ba5ebce@gateway/web/cgi-irc/
04:21:10 Join HBK- [0] (
04:34:27 Quit JdGordon ("Leaving.")
04:35:11 Quit HBK (Read error: 110 (Connection timed out))
04:39:11 Quit KBH (Read error: 110 (Connection timed out))
04:40:32 Quit Hillshum ("CGI:IRC (Ping timeout)")
04:41:19 Join JdGordon [0] (
04:46:35***Saving seen data "./dancer.seen"
04:48:44 Quit HBK- (Read error: 110 (Connection timed out))
04:51:02 Quit slam_ (Read error: 110 (Connection timed out))
05:07:46 Join theglades00 [0] (n=48e6d685@gateway/web/cgi-irc/
05:09:33 Quit theglades00 (Client Quit)
05:10:08 Quit BlakeJohnson86 ("Leaving.")
05:12:31 Join AndyIL [0] (i=AndyI@
05:14:03 Quit kadoban (Remote closed the connection)
05:16:38 Join kadoban [0] (
05:19:58 Join BlakeJohnson86 [0] (
05:26:03 Quit AndyI (Read error: 110 (Connection timed out))
05:35:10 Quit fred_2 (
05:35:10 Quit Hadaka (
05:35:10 Quit lostlogic (
05:35:10 Quit kadoban (
05:35:10 Quit Zambezi (
05:35:32 Quit Horscht ("Verlassend")
05:35:39NJoinfred_2 [0] (
05:35:39NJoinHadaka [0] (
05:35:39NJoinlostlogic [50] (n=lostlogi@rockbox/developer/lostlogic)
05:35:44NJoinkadoban [0] (
05:35:44NJoinZambezi [0] (
05:36:13 Quit planetbeing (
05:36:13 Quit tmzt (
05:36:13 Quit Llorean (
05:36:13 Quit dlctestnt (
05:36:13 Quit trisiak (
05:36:13 Quit Bombe (
05:36:18 Quit JdGordon (
05:36:18 Quit tarbo (
05:36:18 Quit Beta2K (
05:36:18 Quit liiwi (
05:36:18 Quit Galois (
05:36:41NJoinplanetbeing [0] (
05:36:41NJointmzt [0] (
05:36:41 Join Llorean [0] (n=DarkkOne@rockbox/user/Llorean)
05:36:41NJoindlctestnt [0] (n=dlctestn@
05:36:41NJointrisiak [0] (
05:36:41NJoinBombe [0] (n=droden@freenet/developer/Bombe)
05:36:41 Join JdGordon [0] (n=Miranda@rockbox/developer/JdGordon)
05:36:41NJoinBeta2K [0] (
05:36:41NJoinGalois [0] (
05:36:41NJoinliiwi [0] (
05:36:47NJointarbo [0] (n=me@unaffiliated/tarbo)
05:52:12 Quit martian67 (Read error: 104 (Connection reset by peer))
05:53:10 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
05:53:27 Join joshme17 [0] (n=a2282be7@gateway/web/cgi-irc/
05:53:36joshme17hello fellow rockbox users
05:55:12joshme17is anyone on the chat?
05:56:25advcomp2019_joshme17, yes
05:56:33 Quit joshme17 (Client Quit)
05:57:24 Join joshme17 [0] (n=a2282be7@gateway/web/cgi-irc/
05:57:33joshme17Im back
05:58:09advcomp2019_o ok.. you have a question?
05:58:16evilnick_homeHi joshme17
06:00:53joshme17I'm trying to make the scroll speed faster with my e200
06:01:10joshme17I broke the wheel on my sansa and now its hard to scroll
06:01:19joshme17So I want to make it more sensitive
06:01:51joshme17are there any sensetivity buttons?
06:02:18Unhelpfuli don't believe there are
06:02:42 Join HBK [0] (
06:03:27joshme17I was trying to "mod" my mp3 to make the scroll wheel smoother
06:03:38joshme17but instead its way harder and rougher now
06:17:19CIA-71New commit by kkurbjun (r21483): M:Robe 500: RTC is now working, Added some SPI flexibility per end device and modified the interrupt handler for the tsc2100 which should make it more ...
06:21:49 Quit Lss (Read error: 104 (Connection reset by peer))
06:28:46 Quit joshme17 ("CGI:IRC (EOF)")
06:30:01CIA-71New commit by kkurbjun (r21484): Fix Red
06:37:03 Quit martian67 (Connection timed out)
06:38:26 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
06:46:37***Saving seen data "./dancer.seen"
07:06:45 Quit kkurbjun (Read error: 110 (Connection timed out))
07:08:14 Join _lifeless [0] (n=lifeless@
07:08:57 Quit __lifeless (Read error: 60 (Operation timed out))
07:12:46 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
07:16:01 Nick fxb__ is now known as fxb (
07:26:30 Quit at0m (Remote closed the connection)
07:27:49 Quit CoryClaxon ("( :: NoNameScript 4.22 :: )")
07:38:37daurnGodEater: lua is on rockbox now yeah? what are people doing
07:41:06 Quit Zarggg ()
07:48:29 Quit JdGordon (Read error: 104 (Connection reset by peer))
07:55:38 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
07:59:03 Nick fxb is now known as fxb__ (
08:10:31 Join AlexP [0] (n=alex@rockbox/staff/AlexP)
08:10:34 Join Zarggg [0] (
08:10:46GodEaterdaurn: the idea is to let people use it as another way to write plugins
08:11:12GodEatermorning AlexP
08:12:14daurnGodEater: nice, I'll make something up if you want
08:12:29GodEaterdaurn: Llorean has a project for you it would seem :)
08:15:53 Quit kachna|lappy (Read error: 113 (No route to host))
08:22:07 Join einhirn [0] (
08:24:36 Join slam_ [0] (
08:25:29daurnGodEater: where is lua info on wiki etc?
08:25:38GodEatergood question :D
08:26:07 Join lee321987 [0] (n=chatzill@
08:26:07daurnhmmm, trying to compile simulator for fuze:
08:26:08daurn/usr/bin/ld: /home/daurnimator/rockbox/svn/build/apps/codecs/libspeex-voice.a(bits.o): Relocations in generic ELF (EM: 40)
08:26:09daurn/home/daurnimator/rockbox/svn/build/apps/codecs/libspeex-voice.a: could not read symbols: File in wrong format
08:26:09DBUGSent KICK daurn to server
08:26:09daurncollect2: ld returned 1 exit status
08:26:10Kick(#rockbox daurn :No flooding!) by logbot!n=bjst@rockbox/bot/logbot
08:26:13 Join ender` [0] (
08:26:37 Join daurn [0] (n=daurnima@unaffiliated/daurnimator)
08:26:42GodEaterdaurn: oops :)
08:26:51GodEaterdaurn: pastebin next time ?
08:27:06daurnpastebin for 4 lines seems a bit much
08:27:19GodEaterlogbot disagrees it would seem
08:27:39GodEateranyway :
08:27:45GodEaterthat's all there is right now
08:27:54GodEaterI suspect the helloworld.lua is probably more helpful
08:28:05lee321987"Sound quality on the Sansa e200" −− Any thoughts on whether v2 sounds better than v1 when Rockboxed?
08:28:14daurnGodEater: I'll figure it out, shouldn't be ahrd
08:28:23daurnGodEater: so, able to help with the sim?
08:28:42GodEaterdaurn: let me try building it too
08:30:21GodEaterlee321987: "sound quality" is very very subjective
08:31:04lee321987I'm not real picky. Is there a big difference in quality to you?
08:31:15GodEaterI've never listened to either
08:31:18GodEaterso no ;)
08:31:50GodEaterbertrik or kugel probably have both though
08:33:54 Join slam__ [0] (
08:34:49 Join n1s [0] (n=n1s@rockbox/developer/n1s)
08:34:53daurnGodEater: build it?
08:35:12GodEateryep, just finished
08:35:16GodEaterworked fine
08:37:19GodEaterdaurn: which distro you on ?
08:37:58Sir_Brizzwhat kind of videos can my ipod g4 with rockbox play?
08:38:55daurnGodEater: xubuntu jaunty
08:39:15GodEaterSir_Brizz: Mpeg1/2 with MP1/MP2/MP audio.
08:39:22GodEaterer MP3 :)
08:39:26GodEaternot MP
08:39:34*GodEater always forget you can't capitalise numbers
08:40:29Sir_BrizzI wonder how I can get a video in that
08:40:58GodEaterread that page, we have "recommended" conversion tools at the bottom
08:41:13GodEaterso you can take whatever videos you already have, and turn them into Rockbox playable ones
08:41:32Sir_Brizzoh great
08:41:41Sir_Brizzffmpeg *smacks forehead*
08:41:59GodEaterWinFF is our favoured front end
08:42:06GodEaterit comes with a number of presets for Rockbox targets
08:42:21Sir_Brizzoh nice
08:42:27GodEaterwe thought so :D
08:46:40***Saving seen data "./dancer.seen"
08:47:02daurnGodEater: and lua is now in default build? I don't have to add anything?
08:47:55 Quit slam_ (No route to host)
08:49:50GodEaterit's a viewer
08:50:07GodEaterso you should be able to just browse to a .lua file, click on it, and the plugin should execute it.
08:50:43 Join bertrik [0] (
08:52:31daurnyes... should
08:52:34daurnfailed to load /.rockbox/rocks/viewers/lua.rock
08:52:34daurndlopen(simdisk/.rockbox/rocks/viewers/lua.rock): simdisk/.rockbox/rocks/viewers/lua.rock: invalid ELF header
08:53:29*GodEater downloads helloworld.lua to try it
08:55:31GodEaterhmm, this could be tricky
08:55:39GodEatermy X Display forwarding isn't working for some reason
08:56:44bertrikmarkun, I tried i2c communication on the meizu, using the built-in i2c controller. It looks like I'm getting an ACK at address 0x0A (but only once so far, next time I try, my request is NACKed)
08:57:08bertrikI couldn't find anything in the datasheets that is at i2c address 0x0A
08:57:09markunwell, at least something is listening, right?
08:58:15bertrikyes, maybe I should now try more useful requests to that address
08:59:12markunin the schematics of the M6 you can see that the UDA1380 is connected to p10.0 and p10.1
08:59:30lee321987Can I donate money to RB? Who does it go to?
08:59:53GodEaterlee321987: there's paypal button on the site, it goes to the "Rockbox fund", not an individual
09:00:17bertrikwhat I've seen is that a read from a codec is usually something like <START> <ADDRESS-w> <REGISTER> <START> <ADDRESS-r><DATA from codec> <STOP>
09:00:20lee321987thanks guys
09:00:30GodEaterthank *you*
09:00:35markunlee321987: that page tells you what to expect from a donation :)
09:00:46GodEaterdaurn: helloworld.lua works fine here
09:00:53markunlee321987: which player do you use?
09:01:00bertrikso with a repeated START I mean, this is not mentioned in the datasheet but can be done I think
09:01:18lee321987c200v1, but currently looking for a e200v1
09:01:25daurnGodEater: I rebuilt, and this time when I choose the lua file in the fileviewr, nothing happens
09:01:51GodEateris it one you've written, or is it the sample one ?
09:01:55daurnthe sample one
09:02:03daurn - if I go into .rockbox/rocks/viewers/lua.rock and try and run it, it has that ELF header error
09:02:06GodEaterdefinitely works here =/
09:02:19GodEaterwhich svn version you built from ?
09:02:29markunbertrik: how did you find out there was something at 0xA? trial and error?
09:02:42 Join flydutch [0] (
09:02:56bertrikyes, just write 0 bytes to each address and check if I got an ACK
09:02:58GodEaterdaurn: you'll have to ask one of the better devs, I'm at a loss
09:03:13daurnwho would that be?
09:03:35GodEaterwell safetydan is the guy who wrote the lua plugin, but any of the regulars should be able to help if they have time
09:03:47daurnhmm actually
09:03:52daurnthat happens for ALL plugins
09:03:53GodEaterlinuxstb, Bagder, etc.
09:03:59GodEateryou have a broken build
09:04:16GodEater<stating the obvious since 1973>
09:04:31daurnbut i just built it in an empty dir, and it had no errors
09:04:48*daurn starts again
09:05:17 Quit lee321987 ("ChatZilla 0.9.84 [Firefox 3.0.11/2009060215]")
09:06:55Sir_Brizzgah I somehow got the subtitles encoded on this video...
09:08:06Sir_Brizzdarn you DVDFab!
09:10:03 Join kachna|lappy [0] (
09:12:16daurnGodEater: awesome, and now the build has put things in the wrong dir
09:12:21markunSir_Brizz: wrong channel?
09:12:56GodEatermarkun: I think he's having fun with MpegPlayer
09:13:05GodEaterfor various values of "fun"
09:13:06Sir_Brizzprobably :/
09:13:07 Join petur [50] (n=petur@rockbox/developer/petur)
09:15:50daurnGodEater: finaly, working now
09:16:11GodEaterI have to shoot out to a meeting, biab
09:24:19 Quit bertrik (Read error: 113 (No route to host))
09:25:47 Join Thundercloud [0] (
09:28:29 Quit planetbeing ()
09:28:43 Quit flydutch (Read error: 104 (Connection reset by peer))
09:29:57 Join flydutch [0] (
09:33:21 Quit BHSPitMonkey ("afk: aids")
09:37:01CIA-71New commit by rob (r21485): Fix an error in the TCC NAND driver which caused a panic on startup for certain users.
09:37:13CIA-71New commit by rob (r21486): TCC78x: Enable interrupts/threading in the bootloader (required now that the storage driver yields).
09:39:36daurnis O_TRUNC the same as "w+" ?
09:45:36Sir_Brizzis it possible to usb charge in rockbox on an ipod and NOT sync?
09:45:42Sir_Brizzor is there just no way to usb charge?
09:46:57 Quit mt (Read error: 110 (Connection timed out))
09:47:00 Join planetbeing [0] (
09:47:06 Quit planetbeing (Remote closed the connection)
09:48:58linuxstbdaurn: Compare the descriptions of "man fopen" and "man 2 open". You need to specify more than just O_TRUNC.
09:51:08daurnso "w+" is O_RDWR | O_TRUNC ?
09:52:06 Quit Thundercloud (Remote closed the connection)
09:52:14linuxstbdaurn: Looks about right.
09:52:51daurnwhats the difference between a and a+ ? >.<
09:53:31daurnah, its WRONLY, not read and write
09:54:11markuntrying to edit a wiki page I'm sent to TWikiRegistration, where should I login?
09:55:06linuxstbmarkun: I normally get a pop-up window whenever I try to do something that requires me to login...
09:55:14markunlinuxstb: me too
09:55:42linuxstbI've just tried it, and it's still working normally for me - i.e. I got a prompt
09:56:06linuxstbReboot and try again ;)
09:56:17markunhehe :)
09:56:34 Join mt [0] (n=MTee@rockbox/developer/mt)
09:56:41linuxstbI would probably try restarting my browser though.
09:59:37markunlinuxstb: that did the trick :) (I had upgraded firefox but not restarted the browser)
10:06:31Tuplanollamorning. can I use a 16GB microsdhc card with my e280 with rockbox?
10:10:30 Join jessor [0] (
10:13:20 Join mrkiko [0] (
10:13:27 Part mrkiko
10:14:10 Join mrkiko [0] (
10:15:11markunso it should work
10:15:30Tuplanollaah yes
10:15:32Tuplanollathanks a lot
10:16:02markunhopefully the info in there is correct :)
10:16:43Tuplanollayep. we'll see it in a couple of days when I get mine :)
10:23:47 Join funman [0] (n=fun@rockbox/developer/funman)
10:27:05 Quit tchan (Read error: 110 (Connection timed out))
10:27:46 Quit kugel (Read error: 110 (Connection timed out))
10:29:00 Nick zitune[afk] is now known as zitune (n=zitune@bearstech/zitune)
10:30:13 Join raymanism [0] (n=40d7d101@gateway/web/cgi-irc/
10:35:35 Join __lifeless [0] (n=lifeless@
10:42:12daurnwell, I've just wrapped the rb fileoperations for lua to work like normal
10:43:25 Join ruskie [0] (i=ruskie@sourcemage/mage/ruskie)
10:43:49Unhelpfuldaurn: you're the new lua plugin maintainer? ;)
10:46:41linuxstbdaurn: What do you mean "work like normal"? Aren't the RB file operations already normal?
10:46:42***Saving seen data "./dancer.seen"
10:48:34Unhelpfulprobably "more like in other lua interpreters"
10:50:05webmindrockbox supports lua?
10:50:42 Quit _lifeless (Read error: 113 (No route to host))
10:53:10linuxstbwebmind: There's a "viewer plugin" to run lua scripts, yes.
10:53:13daurnto work like the normal lua io functions
10:53:32daurnUnhelpful: just did it in lua, could probably be done easier in C though :p
10:54:09linuxstbdaurn: Ah, OK. So currently (in svn) the standard lua io functions don't work?
10:55:17daurnlinuxstb: correct
10:55:41daurnlinuxstb: you have to use strange and foreign to me, rb.write, etc
10:55:41 Quit raymanism ("CGI:IRC (EOF)")
10:56:47funmanthat is the case for C plugins as well
10:56:57 Join barrywardell [0] (
10:57:07GodEaterfunman: yes, but the rb.functionname works just like the C counterpart
10:57:38daurnwho is the lua maintainer?
10:57:40 Join robin0800 [0] (
10:57:48linuxstbdaurn: Those are the standard POSIX file I/O functions. But yes, if you don't know C, they will be foreign to you.
10:57:53GodEaterwell, he's the original author
10:57:59GodEateranyone can be a maintainer :)
10:58:07funmandaurn: do you volunteer?
10:58:22linuxstbdaurn: We don't really have maintainers - people come and go and their interests change.
10:58:46daurnI'm not volunteering for anything C related
10:59:03funmandaurn: you could just send your patch to rockbox-dev@
10:59:30linuxstbIt's not really a patch though, is it? Unless we can have a library of lua code included for every lua plugin.
10:59:53daurnits something that people could require from lua
10:59:59daurnrequire"stdio" or something :p
11:00:30linuxstbI would say post it to a flyspray task (as a patch), and try and start a discussion there about how that (and similar libraries) could be incorporated into Rockbox.
11:01:33linuxstbMaybe we should have something like .rockbox/lua/ where such files can live?
11:02:34daurnwell in the standard lua interpreter, it looks for lua libs in /usr/local/share/lua/VERSION/...
11:02:35GodEaterwell we already have "actions.lua" living in the viewers directory
11:03:00linuxstbGodEater: That could be moved then. I don't think having lots of .lua files there is nice.
11:03:07GodEaternor I
11:03:21linuxstbSomething for MrSomeone...
11:04:10linuxstbdaurn: Are you suggesting we add a lua/VERSION/ hierarchy?
11:05:06*GodEater suspects we're unlikely to have multiple versions of the lua interpreter available at one time
11:05:31 Join kugel [0] (i=kugel@rockbox/developer/kugel)
11:06:22 Quit cool_walking_ ("Page closed")
11:06:36daurnlinuxstb: no, but there needs to be a standard lib dir
11:07:42 Nick blippe is now known as lazysod (n=none_of_@
11:11:30linuxstbdaurn: Yes, that's what I suggested - .rockbox/lua/
11:12:20daurnlinuxstb: and I was agreeing
11:13:48 Join lucent [0] (i=lucent@unaffiliated/shadows)
11:13:49linuxstbOK ;)
11:14:20lucentfunman: heya, I'm running into a curious abort error with 8gb fuze and 8gb uSD card
11:14:46lucentonly seems to happen with some files and not others
11:14:50daurnlua segfaults rockbox when you try and run a file that doesn't exist
11:15:08daurntry running this as a .lua file: dofile("doesntexist")
11:15:36funmanlucent: not curious, the storage driver is broken
11:15:55lucentoh okay
11:17:16lucentfunman: know by-chance if the other sansas (c200 or e200) are any more stable under rockbox? I'm not freaking about my fuze, it works well enough and I'm happy...
11:17:33lucentthough, I would like something reliable enough that can accept uSD cards so I can deejay with that soon
11:18:11CIA-71New commit by lenzone10 (r21487): Updated italian translation.
11:18:15funmanafaik it's the same problem on all sansa
11:18:46*lucent grins
11:18:54funmanc200v2 still lacks access to some buttons, m200v4 clip & c200v2 suffer from a problem with playback
11:18:58lucentit's so close!
11:20:27lucentwell any tests you want I can try here
11:20:29 Join PaulJam [0] (
11:20:53lucentI love the work you've done already, and the other contributors as well
11:21:54lucentfunman: where would I look, I want to make a quick tweak so that the fuze and rockbox assumes it just wants power when connected to USB, not "USB data" mode
11:22:13lucenti.e. automatically like I'm holding the select button when inserting usb
11:22:28funmanholding the select button is simpler :)
11:23:01lucentsometimes I mess it up and then i have to unplug that usb cable, it's a mess if I'm driving a car and doing that ;)
11:24:10lucentit would be nice if I could press select button to revert out of the usb data mode, and into charge mode
11:24:59mtsaratoga : The patch you made, is it to be applied after applying my latest patch ?
11:25:37 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
11:25:58funmanlucent: make usb_power_button() in firmware/usb.c always return true
11:26:07lucentfunman: perfect, thank you
11:26:54 Nick zitune is now known as zitune[afk] (n=zitune@bearstech/zitune)
11:26:59 Join Grahack [0] (
11:38:15mtlinuxstb (when you have time to test the patch) : Could you check the bug I talked about before about not being able to get metadata tags from id3v1buf[0] and -buf[1] ?
11:39:01linuxstbmt: OK. Did you see saratoga's patch that made some changes to your patch? He posted it in IRC after you left yesterday.
11:39:15mtYes I saw it.
11:39:39funmani found audiohw_set_freq() to change playback frequency, but i can't find the equivalent for recording frequency
11:40:05linuxstbIt's not the same function?
11:40:14mtlinuxstb : I have just asked him here about whether it should be applied after my patch, or is it intended as a replacement patch .
11:40:38linuxstbmt: So you did ;) I missed that.
11:41:11mtlinuxstb: ;)
11:41:37mtanyway, if it's a replacement, I'll have to modify it a little because it misses out some files.
11:43:31daurnwho can tell me why the whole of rockbox seg faults when you loadfile a file that doesn't exist with lua
11:43:36mtlinuxstb : Is it common to have sim vs target problems like the current one ?
11:44:00funmandaurn: segfault means data abort ?
11:44:24daurnsegfault meanign the simulator quits with "Segmentation fault"
11:44:28linuxstbdaurn: Because there's a bug in Rockbox's lua interpreter...
11:44:46*linuxstb doesn't know what answer daurn was looking for
11:45:37daurni guess I just don't understand how rockbox works; its not normal for the whole o/s to crash when one app - an interpreted one at that - tries to open a file
11:45:59linuxstbmt: Yes. For example, the sim manages memory differently, so if you're overwriting something you may not notice in the sim, but you do on target.
11:46:24funmandaurn: the simulator is special
11:46:52linuxstbdaurn: The sim _is_ one app, and Windows/Linux/etc is the OS. Does your OS crash?
11:47:13linuxstbAnd on target, there's no memory protection, so yes, one part of Rockbox can bring everything down.
11:48:03daurnok :)
11:48:42daurnhow can I debug (print to sim console)?
11:49:00funmanif the simulator segfaults you should use gdb
11:49:06linuxstbIn C or in lua?
11:49:14daurnC first, then tell me about lua
11:49:29linuxstbYou use the DEBUGF() macro, which works like the C printf() function.
11:49:57linuxstbAnd I don't know about lua...
11:51:08linuxstbDoes that dofile function work if the file exists?
11:51:22 Quit jkl (Operation timed out)
11:51:47daurnthe problem is somewher around line 600 of apps/plugins/lua/lauxlib.c
11:58:53linuxstbdaurn: I don't have time to debug it now. If you can't find the problem, file a bug report, with an example .lua program that demonstrates the bug.
11:59:14daurnok, found the problem line:
11:59:24daurnline 557 in lauxlib.c const char *filename = lua_tostring(L, fnameindex) + 1;
11:59:55daurnthe +1 should be INSIDE the bracket
12:00:31lucentfunman: is bug reporting on the errors I see (SD related) useful to you and others?
12:00:42funmanlucent: nope, unless it comes with a patch :)
12:00:46lucentoh okay
12:01:03lucentsorry I don't do that kind of thing well, I wish I could help in some way
12:01:36funmanlucent: the manual needs some work if you want to help on something else than code
12:01:53lucentfunman: hm, yeah I may do that
12:02:22lucentso many distractions, but I really do feel motivated to try and help
12:02:54daurnor not >.<
12:04:59 Quit timc (Read error: 54 (Connection reset by peer))
12:08:55 Join goffa__ [0] (n=goffa@
12:09:13lucentfunman: my rockbox is configured to resume playback on power-on, and I get those data errors sometimes. If I pause playback just after power-on (repeatedly pressing PLAY like a crazed monkey), there's no crash, and then I just hit play and it keeps oging
12:09:40lucentstill random SD driver error?
12:09:53lucentjust hoping there was some info I could give you that would be helpful
12:10:05lucentmaybe there isn't anything to it :(
12:11:32funmani don't know what is going wrong and i still didn't look at it carefully
12:12:41lucentI wish success for you if you get time to look at it :)
12:12:43daurnflyspray server is SLOW?
12:19:13daurnbug submitted.
12:19:32 Quit goffa_ (Connection timed out)
12:20:35 Quit n1s (Read error: 110 (Connection timed out))
12:28:56 Join timc [0] (n=aoeu@
12:39:11CIA-71New commit by mcuelenaere (r21488): Fix FS #10378: Lua crashes when loading non-existing files
12:41:23 Part ruskie ("...")
12:41:48GodEaterbug fixed
12:46:05Grahackhi, has mcuelenaere a different nick on IRC ? I'd like to chat about the Lua plugin.
12:46:46***Saving seen data "./dancer.seen"
12:47:05funmanno he has the same nick (but is absent from irc now)
12:47:13 Quit funman ("free(random());")
12:47:25GrahackThanks. Is there someone involved in the dev of Lua plugin here then ?
12:48:13 Join mcuelenaere [0] (
12:48:15GodEaterjust ask your question
12:48:21mcuelenaereGrahack: you called?
12:49:11Grahackgreat, I wanted to talk about buttons and the name of the file actions.lua
12:49:20 Join simjoko [0] (
12:49:31Grahackdo you plan to expose button codes to lua ?
12:50:21mcuelenaereGrahack: it isn't technically difficult to do that, but I don't really see a point in doing so
12:50:33mcuelenaerehow is a script going to know which buttons it must use?
12:50:34 Quit markun (Read error: 104 (Connection reset by peer))
12:50:38mcuelenaerethey differ in naming across targets
12:51:09Grahackbut could lua.make be aware of the target ?
12:51:16Tornemcuelenaere: this is true but the actions are somewhat useless on some targets
12:51:22Tornee.g. ipod where select and right are the same action
12:51:38 Join markun [50] (n=markun@rockbox/developer/markun)
12:51:46Torneso it seems like totally avoiding target specific stuff in nontrivial lua scripts is unlikely anyway :)
12:51:56mcuelenaereGrahack: what do you mean?
12:51:58mcuelenaereTorne: hmm true
12:52:27mcuelenaerelast time I asked about this, the general opinion was 'actions are the way to go'
12:52:28GrahackI see $(TARGET) in lua.make, we could write a that could create a lua table if we tellhim which file to process
12:52:46Tornemcuelenaere: oh, it is, if what you're writing can make do with them
12:52:53GodEaterGrahack: well volunteered
12:52:54mcuelenaereGrahack: that's what I planned in doing if we were going to do this
12:53:05Tornebut if you're writing a game and you want "up down left right select" that's not easy to do with the action lib
12:53:20Unhelpfulbeing able to use get_custom_action would be nice if you want to do games and the like :)
12:53:23Grahackspacerocks couldn't work only with actions...
12:53:41Tornei guess there could be an argument for having actionlib support several sets of bindings, as it were, actually
12:53:52Torneatm it tries to reflect the standard usage of each button in the main rockbox ui
12:54:03mcuelenaereUnhelpful: get_custom_action requires a C callback, which isn't easy to implement in Lua
12:54:09Tornebut maybe you could have a "game mode" or similar where it tried to give you direction buttons instead, or similar
12:54:15GrahackI'm not a perl expert so I talked about this to aperl-friend, we could work on this a bit
12:54:40Grahackgenerating a lua table with target specific button codes
12:54:55mcuelenaereGrahack: it isn't really difficult, I thought regex'ing button-target.h for BUTTON_* lines and then preprocessing it would do the job
12:55:04mcuelenaerea bit like currently does
12:55:30Tornemcuelenaere: probably close enough
12:55:42Grahackok so we're looking in the samedirection, great
12:55:43Torneyou'll pick up things like BUTTON_MAIN BUTTON_REMOTE that way but that's probably a feature :)
12:55:49Unhelpfulmcuelenaere: a fixed callback wouldn't be that difficult, though, maybe something along the lines of the way pictureflow does it - there's a table of action contexts, and it just strips the flags from the context and indexes the table.
12:55:56mcuelenaereTorne: heh, we could filter that out :)
12:56:06Torneyah, but it might be good actually :)
12:56:20Grahackthe second question is: would it be niceto rename actions.lua to if it contains actions, contexts and button codes ?
12:56:24Tornethough actually evaluating their values might be fiddly
12:56:33Unhelpfuli don't think doing the callback in lua is needed... i agree it sounds awful to have to do :)
12:57:06mcuelenaereGrahack: I'm not sure whether we should split that or put it in one big file
12:57:13mcuelenaereremember, these targets don't have much RAM
12:57:30mcuelenaerethe less files we would need to include, the better
12:57:47Torne..wait, what i was describing above is what pluginlib_actions already does, mostly :)
12:58:05mcuelenaereTorne: wasn't pluginlib going to be deprecated?
12:58:19*mcuelenaere thought we should try to move away from it as much as possible
12:58:54Unhelpfulif you want up/down/left/right *and* some other actions you'll need to combine, and then it all falls apart (with PLA)
12:59:20Tornemaybe we just need a better design for PLA that lets you combine things to some degree
12:59:41mcuelenaerehmm so basically, Lua scripts will need to behave the same as C plugins and (runtime) check what keys are available and use those
12:59:53mcuelenaereTorne: pluginlib isn't exposed to Lua (yet)
13:00:15Tornemcuelenaere: sure, but i can't see it being particuarly hard to do :)
13:00:18Torneit's just a suggestion
13:00:33TornePLA is not a universal solution, anyway
13:00:45daurnmcuelenaere: hey
13:00:49Torneultimately some script somewhere is going to want raw buttons
13:00:59Tornebut then it's gonna just have to runtime decide what target it's on
13:01:02mcuelenaeremm yes, probably
13:01:04mcuelenaeredaurn: hi
13:01:15Tornes;not really enough to just detect what buttons exist because that ells you nothing baout their physical placements
13:01:28Torneyou'll end up doing the equivalent of the ifdef pile that C plugins do ;)
13:01:30Tornebut at runtime
13:02:02*mcuelenaere wonders whether the script should decide what keys to use based on the player (eg new constant) or based on what keys are available
13:02:15Tornebased on player, almost certainly
13:02:30Unhelpfulexpose PLA in lua, or pluginlib in general? that could be problematic as we'd need to decide what pluginlib functions we want linked into the interpreter...
13:02:33daurnmcuelenaere: I wrapped file operations with a more lua-like syntax
13:02:38Tornehow else is it going to know that on ipods, BUTTON_MENU and BUTTON_PLAY are th ebest choices for up and down in 4-way controls? :)
13:02:41daurnmcuelenaere: though you might want to implement it in C
13:02:54mcuelenaeredaurn: I saw that, and I'm trying to mimic it in C using IO lib
13:03:12daurnmcuelenaere: newer:
13:03:22TorneUnhelpful: add stuff as people request it probably ;)
13:03:28mcuelenaeredaurn: ah, thx
13:03:33GodEaterperhaps we need aliases for buttons too ?
13:03:45TorneGodEater: isn't that more or less what PLA *is*? :)
13:03:51Tornea bunch of alias mappings
13:03:59Tornes'just not a very flexible one
13:04:01GodEaterhehe - I guess so
13:04:06daurnmcuelenaere: I'm sure you can thing of a better userdata object, I just used what first came to mind
13:04:08*mcuelenaere should look into how PLA works
13:04:09*GodEater shall further engage brain before speaking again
13:04:48mcuelenaeredaurn: I'm not really familiar with Lua, but I'll just port the Lua IO lib using Rockbox's file functions
13:04:48Tornemcuelenaere: it's just a bunch of get_custom_action tables
13:04:59mcuelenaereso a lot of C callbacks?
13:05:14Tornemcuelenaere: the plugin picks which tables it's interested in and a priority order for them and passes it over
13:05:17Tornejust one callback
13:05:30Tornethe code is really short, not hard to understand :)
13:05:48*mcuelenaere doesn't spot a callback
13:06:07mcuelenaerepluginlib_getaction, right?
13:06:41Torneit's get_context_map
13:07:15 Join nibbler_ [0] (
13:07:32Tornepluginlib_getaction pretty much is just get_custom_action, except it just sets up the static variables that the callback is going to use as well
13:07:36 Quit rvvs89 (Read error: 104 (Connection reset by peer))
13:08:25mcuelenaereTorne: isn't that the preferred way of using PLA?
13:08:37Tornei'm not sure what you mean
13:08:38Tornethat's what it is
13:09:03 Join n1s [0] (n=n1s@rockbox/developer/n1s)
13:10:31mcuelenaereTorne: never mind, I think I understand how PLA works
13:10:51Torneit's just a convenience wrapper for rb->get_custom_action to save you writing your own callback
13:10:59Torneand a set of mapping tables that might be useful
13:11:34mcuelenaereyes, so I don't see why you should do that rb->get_custom_action() call in your code; just use pluginlib_getaction() and it works
13:11:36UnhelpfulTorne: a general-purpose callback, and the ability to define tables from your script, would be nice.
13:12:17Tornemcuelenaere: hm? i didn't sya you did..
13:12:40Grahackall this is beyond my knowledge... would it mean dropping the lua actions table ?
13:13:36TorneGrahack: at the moment lua calls the default get_action which uses rockbox's standard set of action mappings
13:13:37mcuelenaereGrahack: no, it would mean adding some extra Lua functions
13:14:09mcuelenaerealso, BUTTON_* constants will probably get added
13:14:11TorneGrahack: there's also other sets of action mappings in PLA which you can use instead of the default mappings, which will solve the problem you had of select and right not being distinguishable on the ipod, but may introduce other problems ;)
13:15:12Torneyah, either way you'd want to add the BUTTON_* stuff as well
13:15:24Grahackand would it be possible (or maybe it's already the case) to expose the platform to lua, like = "iPod Video" for example ?
13:15:31Tornefor lua scripts that are willing to take the awkwardness of having to make it target-specific to get flexible button config
13:17:15TorneGrahack: yah, that would be required to make BUTTON_* useful
13:17:20UnhelpfulTorne: presumably they can include a second script in a platform-dependent way?
13:17:29TorneUnhelpful: yes, that'd work nicely
13:17:33Tornesaves memory
13:17:49daurnI wouldn't recommend that
13:18:13daurnall you need is all the button state exposed in a table
13:18:29daurnand make a lua function/library that does it all for you
13:18:31mcuelenaereGrahack: I don't think using strings is a good way to define targets
13:19:00daurnso have the standard buttons; and whatever target you're running on has extra keys in the tabel
13:19:03Tornedaurn: the point is that for button_get to be *useful* you need to know what target you are on
13:19:04Grahackor anything else! what could it be ?
13:19:10Tornebecause otherwise you don't know what button to map to what
13:19:30 Join funman [0] (n=fun@rockbox/developer/funman)
13:19:38daurnTorne: I'm saying that you can abstract that away ==> know the actual target that is
13:19:46Tornethat's what pluginlib_actions does
13:19:52Unhelpful...badly. :/
13:19:54Tornebut the point is that it's not sufficient for every purpose
13:19:56daurnand just check for buttons.left and buttons.up and buttons.right etc
13:20:09Tornedaurn: yes. that's exactly what pluginlib_actions already does.
13:20:11daurnand if you have all of them; have one control set
13:20:16Unhelpfuldaurn: not really, the layout is important.
13:20:19Tornebut if you want a configuration that PLA doesn't have, it's tough.
13:21:26daurnalso, I haven't looked at the lua api yet
13:21:33Torneit doesn't matter what the lua api is, to be honest
13:21:33daurnbut can you play/pause/etc?
13:21:44Torneoh, sorry
13:21:48Torneyou're talking about something els enow :)
13:22:39*mcuelenaere wonders whether he should define stdin/stdout/etc to 0 or just remove it from the Lua IO lib
13:22:57daurnmcuelenaere: I'd remove stdin
13:22:57Tornemcuelenaere: just drop them, i'd suggest
13:23:03Torneif they aren't implemented.
13:23:08daurnmcuelenaere: but make stdout a file that gets created somewhere
13:23:08mcuelenaeresame for popen then I guess?
13:23:10 Join stoffel [0] (
13:23:11Tornethen scripts will fail more promptly :)
13:23:12mcuelenaerethis breaks compatibility though
13:23:14daurneg, luaoutput.txt
13:23:22Tornedaurn: ah, yes, that's possibly a good idea
13:23:33Tornebut yeah for things that aren't implemented, unless lua already has a defined way to say they aren't implemented, remove them
13:23:40Tornethat way scripts which try to use them will fail early
13:23:42Torneand people will notice
13:24:04 Join linuxguy4 [0] (
13:24:08 Join Boogerhead [0] (
13:24:33Grahackmcuelenaere: and what about the math lib ? do you plan to expose it ?
13:24:47daurnTorne: I don't know of any lua progs that actually use io.input or io.ouput
13:24:49mcuelenaereGrahack: not all functions have C equivalents in Rockbox
13:24:49BoogerheadPainful question −− How do I tell the database to re-initialize or update? I've got nothing in the menu to do so
13:25:04mcuelenaereGrahack: you do know there's no floating point, right?
13:25:08daurnand I can't think of any cases where they would be useful that major tweaking wouldn't be necessary anyway
13:25:24Tornedaurn: ..yes, so if you remove them then most people won't notice, and the people who are trying to use them will realise quickly that they're not implemented :)
13:25:39Tornewheras if they just do nothing or similar people may wonder why it's not working
13:26:06TorneBoogerhead: settings -> database -> update now, if my memory serves
13:26:15daurnfor most things, I agree; but we're talking about stdin and stdout here: its not obvious that rockbox doesn't have those things?
13:26:19 Quit nibbler_ (Remote closed the connection)
13:26:20GodEateror the context menu for the database entry will do it too
13:26:42Tornedaurn: i'm not sure why that's an argument for not removing them
13:26:51Boogerheadtorne: thanks .. settings: general: database ... thanks! =)
13:26:59Grahackmcuelenaere, so ok there are some issues that are beyond my knowledge... I'll try to implement what I need in the math lib in pure lua
13:27:11daurnTorne: remove stdin, stdout should be replaced by something where you can read it still
13:27:47Tornedaurn: yes. I said that was a good idea. I said remove things which are *not implemented*
13:27:55Torneif you make stdout go to a file then it's implemented :)
13:28:07mcuelenaereGrahack: there are some fixed-point math functions, see apps/plugins/lib/fixedpoint.c, they just aren't wrapped in Lua yet
13:28:16 Quit Boogerhead (Client Quit)
13:28:47TorneGrahack: what actual math functions are you after?
13:29:14Grahackmin, max, trigo
13:29:38Grahack(ok min and max are easy :-D)
13:29:47daurnoh, theres no math lib?
13:30:18Torneyou'll need the fixed point library for trig
13:30:26Torneotherwise everythhing is going to return 0 or 1 a lot :)
13:30:43Tornethe lua math lib implementations won't work because we're using ints
13:30:50Grahackdaurn: mcuelenaere said "not all functions have C equivalents in Rockbox. You do know there's no floating point, right?", I'm not sure (and maybe don't want yet) to understand this fully
13:31:05TorneGrahack: lua's numbers are all whole numbers, in rockbox
13:31:16Tornethere's no such thing as 0.1 for your script
13:31:21Torneso, trig is difficult. :)
13:31:40Grahackso that's why I can't splash for a non-whole duration
13:31:43Torne3/2 == 1
13:31:48TorneIndeed :)
13:31:50 Nick J-23 is now known as pciupak (
13:31:57Tornedivision will just truncate the result
13:31:59 Nick pciupak is now known as J-23 (
13:32:03*mcuelenaere thought this was explained in the manual
13:32:03Torneso 1/2 is 0
13:32:23Torneso that's why we can't just provide the lua math library easily
13:32:43Tornewe'd have to implement it in terms of fixed-point arithmetic instead, and you would have to have your script deal in fixed point and convert to/from integer as needed.
13:33:13daurnthere is already lua fixed point math libraries I'm pretty sure
13:33:42Grahackmcuelenaere: sorry I read the RB manual thorougly but don't remember this one
13:33:42Tornedaurn: yes, but you wouldn't want to use them
13:33:50Torneyou'd want to wrap our C fixed point math library up :)
13:34:09mcuelenaereGrahack: it is mentioned in the Lua plugin description, but if you don't know what floating point means you've probably missed it :)
13:34:54GrahackI thought I knew:-( I guess I maybe skipped this technical "pointu" thing
13:34:57 Quit linuxguy3 (Read error: 110 (Connection timed out))
13:35:13daurnmcuelenaere: you've looked at luaconfig.h?
13:35:40mcuelenaeredaurn: I think so, yes, why?
13:36:27daurnI thought you just tell it you don't want to use floating point, and it fixes everything for you
13:36:34daurn(and you don't have to get rid of the math lib)
13:36:53Grahackok in fact I didn't even think about reading the manual here:
13:37:01Grahackwhere it's explained
13:37:50mcuelenaeredaurn: I had some trouble getting Lua to use fixed point, so I hacked some places to get it working right
13:38:07Grahackthanks guys for everything, bye
13:38:12 Quit Grahack ("Leaving.")
13:38:29Tornedaurn: i think the math library still "works" when lua numbers are ints, but only for the minimal definitoin of works
13:38:36Tornei.e. things like sin() will jus tend up returning 0 or 1
13:38:45Tornei don't think the math library code actually adapts to deal with it
13:38:50daurnmcuelenaere: can you enable the math lib?
13:39:00*mcuelenaere tries
13:39:26Tornebut things like math.abs or math.min should work
13:39:40daurnTorne: yeah, possibly
13:39:46Torneactually having it return fixedpoint would be an API change
13:39:54Tornesince fixed point numbers actually mean something different
13:39:58daurnbut alot of code uses math.floor :)
13:39:59Tornewhereas float/int is just a rounding thing :)
13:40:19Tornewell again math.floor might Just Work
13:40:21Tornedepends how it's implemented
13:40:28Torneit would do nothing, but that's correct :)
13:40:34daurnwell, with ints, you don't even do anything :
13:40:50Torneyah, but it may not be implemented that way :0
13:41:22Tornebut yah. trigonometry is *not* going to work, even if some of the rest does.
13:42:02 Quit simjoko ("KVIrc Insomnia 4.0.0, revision: , sources date: 20090520, built on: 2009/06/06 12:36:26 UTC")
13:42:45mcuelenaerehmm I get a PANIC: unprotected error in call to Lua API (attempt to index a string value)
13:43:10daurnmcuelenaere: see that page mcuelenaere, its a library to get lua to work with non floating point numbers :)
13:43:41Tornedaurn: i don't think it help swith math
13:43:45Torneand you can already make lua use int trivially
13:43:51Tornei think LNUM is just for using other types
13:43:59Tornelike single-precision float, or unsigned int, or so on :)
13:44:27Torneooh, or complex numbers. that's kind of cool. :)
13:45:06daurn"The LNUM patch, also known as "integer patch" allows Lua 5.1 to treat pure
13:45:07daurninteger operations with full accuracy and more speed on non-FP platforms,
13:45:07daurnwithout changing the external interfaces."
13:46:08 Quit kugel (Read error: 110 (Connection timed out))
13:46:23TorneAhh, i see
13:46:38Torneit's supporting *multiple* number types
13:46:56Torneso numbers can be integers until such time as they get divided by three or whatever :)
13:48:06mcuelenaeredaurn: Rockbox's printf() functions don't support %f/%g/..
13:48:18mcuelenaereand Lua floating point (including this patch) depends on that
13:49:25daurnand that can't be changed/fixed implemented I take it? (or without major effort)
13:49:47mcuelenaereit can, but that takes core changes
13:50:01mcuelenaereor it would require a separate sprintf() implementation in the plugin
13:50:54 Quit FlynDice (Remote closed the connection)
13:52:25mcuelenaeredoes Rockbox somewhere use software-floating point?
13:52:28 Quit funman ("free(random());")
13:52:31*daurn looks over at Bagder , Llorean , linuxstb , etc
13:53:45daurnmcuelenaere: also, any reason that you don't have the package library?
13:54:15mcuelenaeredaurn: haven't looked at that, safetydan didn't include it so I suppose it didn't compile
13:56:02mcuelenaereah I see, that requires some sort of dynamic library linking framework to work
13:56:21mcuelenaere(which obviously Rockbox doesn't have)
13:58:52daurnstill, can you add a subset of it
13:59:04daurnloading lua packages is a very important feature
13:59:24daurn(which is still possible with dofile, and you can implement require in lua completely)
13:59:39preglowmcuelenaere: software-floating point in rockbox? no
13:59:39daurnbut a C implementation would be nice
13:59:43preglowwe very much try to avoid that
13:59:47preglowsince we don't have an fpu
13:59:58mcuelenaereso I suppose it never has been tested?
13:59:58preglowwell, the gigabeast does, i think, but that's the exception
14:00:19 Quit trisiak (
14:00:19 Quit Bombe (
14:00:19 Quit Llorean (
14:00:19 Quit tmzt (
14:00:19 Quit dlctestnt (
14:00:31preglowit depends what you mean by it, floating point will work, but there is no library support
14:00:40NJointmzt [0] (
14:00:40NJoinLlorean [0] (n=DarkkOne@rockbox/user/Llorean)
14:00:40NJoindlctestnt [0] (n=dlctestn@
14:00:40NJointrisiak [0] (
14:00:40NJoinBombe [0] (n=droden@freenet/developer/Bombe)
14:00:43preglowan it'll be painfully slow on almost all our targets
14:00:45preglowso please do avoid it
14:01:02mcuelenaereyes I know that :) I just wasn't sure whether it was ever tested on-target
14:01:22mcuelenaeredaurn: so you mean require() should just be an alias for dofile() in Rockbox Lua?
14:01:29preglowsure, many of our codecs used floating point before i fixed them to not do so :>
14:01:51preglowlike i said, floating point will work, just don't expect sqrt() or cos() to exist
14:02:30daurnmcuelenaere: not an alias, but it should check in the package.loaded table to see if its already been loaded, and if now, loadfile is from a list of locations specified by package.path
14:03:20daurnpreglow: I want to use the LNUM patch - it allows you to use ints most of the time, but floating point nums when absolutely necessary
14:03:25mcuelenaereah, I hacked dofile() to do the 'search in list of locations for file'-part :)
14:04:52daurnmcuelenaere: no, dofile should NOT do that
14:05:09daurnmcuelenaere: that is require's job :)
14:05:52daurnmcuelenaere: and can you tell me what you're doing with the io library?
14:06:31mcuelenaereI'm basing myself on the original io library, stripping away stuff we don't have and using Rockbox's file functions
14:06:51daurnok :)
14:07:11daurnmcuelenaere: and I hope you notice theres the file library, and the io library.
14:07:27daurnthe file library being the important bit
14:08:36mcuelenaereah I already wondered about the weird *a, *n, *l, .. part :)
14:09:39preglowdaurn: well, then you need to find a libm
14:09:43preglowfor use in the lua plug
14:10:11*Torne suggests you nick it from newlib
14:10:46preglowbut like i said, floating point operations are painfully slow for almost all of our targets, i hope you're not going to run anything that does a lot of computation
14:11:01daurnpreglow: lua is for running user scripts, its up to them what they run
14:11:06mcuelenaeredaurn: yes, looking at that page :)
14:11:31daurnpreglow: if they want to run a script that uses fp, and is damn slow, thats their own fault
14:12:14mcuelenaeredaurn: how does LNUM know if you want to use a integer or a number?
14:13:52 Quit trisiak (
14:13:52 Quit Bombe (
14:13:52 Quit Llorean (
14:13:52 Quit tmzt (
14:13:52 Quit dlctestnt (
14:14:06NJointmzt [0] (
14:14:06NJoinLlorean [0] (n=DarkkOne@rockbox/user/Llorean)
14:14:06NJoindlctestnt [0] (n=dlctestn@
14:14:06NJointrisiak [0] (
14:14:06NJoinBombe [0] (n=droden@freenet/developer/Bombe)
14:15:10daurnmcuelenaere: and while you're at the file library, lnum adds:
14:15:15daurn- 'file:read()' has a new mode "i" for reading in integers. Reading as "n" is still valid, but may cause accuracy issues (floating point accuracy used).
14:16:04Tornemcuelenaere: integer literals start as integers, float literlas start as float, and if an integer operation's result is inexact it gets promoted to float
14:16:25Tornekinda like how python converts between int and long
14:17:40 Quit stoffel (Read error: 113 (No route to host))
14:23:46 Nick fxb__ is now known as fxb (
14:31:13daurnmcuelenaere: how you going? :)
14:33:24mcuelenaerestill working on it
14:37:36 Join LambdaCalculus37 [0] (i=44a0430d@gateway/web/freenode/x-3c8f0734ee1f9d8d)
14:37:52mcuelenaerehmm how could I easily test if I'm at the EOF in Rockbox? Compare lseek()'s output with filesize()'s one?
14:38:06mcuelenaere(with lseek(fd, 0, SEEK_CUR) of course)
14:46:48***Saving seen data "./dancer.seen"
14:46:49 Join stoffel [0] (
14:50:02 Quit robin0800 ("Leaving")
14:52:03 Part LinusN
14:56:09LambdaCalculus37I've been wanting to use FS #7505 on my Gigabeat F for a while, but there are no instruction on where to place gigabeat_flash.c on the tracker.
14:57:15GodEaterLambdaCalculus37: in apps/plugins
14:57:20GodEaterand then edit SOURCES
14:57:20gevaertsLambdaCalculus37: in apps/plugins/. You also need to edit SOURCES there
14:57:27*GodEater beats gevaerts to it
14:57:36gevaertsnot entirely!
14:57:50*GodEater went all fancy with his SOURCE editing and used proper #ifdefs and everything
14:57:54 Join SirFunk [0] (
14:58:14GodEatergevaerts: on my screen both my lines came out before yours :P
14:58:29gevaertsGodEater: let's ask LambdaCalculus37
14:58:30daurnGodEater won on both counts
14:58:35Torneirc is only partially ordered
14:58:44Tornegodeater won for me too but that may not be true from all servers :)
14:59:56 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
15:00:11*gevaerts will accept LambdaCalculus37's judgement, as he was the intended audience
15:00:40LambdaCalculus37GodEater told me first, and gevaerts gave the extra bit in info to go with it.
15:02:12 Join kkurbjun [0] (n=kkurbjun@rockbox/developer/kkurbjun)
15:02:34LambdaCalculus37Okay, so put it in apps/plugins, edit SOURCES, then how do you proceed with flashing the Gigabeat?
15:02:47GodEaterrun the plugin
15:03:11GodEateruse it to backup your flash
15:03:16gevaertscopy bootloader.bin from FS to the root of your player, build everything, and run the plugin
15:03:31*GodEater is willing to bet it won't work first time
15:03:42LambdaCalculus37GodEater: I live dangerously. ;)
15:03:44GodEaterunless LambdaCalculus37 uses -DENABLE_DEV in the Makefile
15:04:50*LambdaCalculus37 will remember to
15:05:10GodEaterI was impressed at the number of warnings kkurbjun managed to fit in there
15:05:21GodEaterit's full of "dire portents" and "here be dragons"
15:05:33GodEater"abandon hope all ye who enter" etc.
15:05:56gevaertsand yet you went on :)
15:06:07kkurbjunI don't want to be held acountable for anyone breaking their player :-D
15:06:09GodEaterI can't read
15:06:28kkurbjunLambdaCalculus37: could you dump your flash before patching it if it fails the checksum. I want to figure out what is different in all these flashes
15:06:43LambdaCalculus37kkurbjun: Sure thing.
15:07:01GodEaterthe patch is building up a nice collection of those now
15:07:05*LambdaCalculus37 wonders if he should...
15:07:24LambdaCalculus37I wonder if I should borrow my girlfriend's Gigabeat X and dump the flash from that as well for kkurbjun.
15:07:32GodEaterhell yes
15:07:39gevaertssure. Dumping the flash isn't risky
15:07:51CIA-71New commit by alle (r21489): Correct the slight misalignment of the clock (FS #10334 by Andre Lupa)
15:07:58LambdaCalculus37GodEater: I'll ask her if she'll lend me her Gigabeat X for that.
15:08:03kkurbjunLambdaCalculus37: yeah, I would be very interested to see tha tflash
15:08:20LambdaCalculus37kkurbjun: Also get ahold of toffe82. He has a Gigabeat X as well.
15:08:28kkurbjunif you have the GBSYSTEM for that too I wouldn't mind taking a look at that as well
15:12:57 Quit LambdaCalculus37 (Ping timeout: 180 seconds)
15:13:05 Join LambdaCalculus37 [0] (i=44a0430d@gateway/web/freenode/x-5465da4bc661ce69)
15:16:42LambdaCalculus37Crud. Got bumped.
15:17:02LambdaCalculus37Anyway, kkurbjun: I'll get the GBSYSTEM folder from my girlfriend as well.
15:17:59GodEater"bumped" ?
15:21:01saratogamt: that patch was just an SVN diff against my local checkout with your patch applied and some random changes
15:21:15saratogait didn't work though so I'm not sure its too important for you to apply it
15:21:40saratogamostly i just cleaned up some of the problems in bitstream.c/h so that it would compile on target and throw a few less warnings
15:21:47saratogai'm not even sure it was all correct
15:21:56 Join robin0800 [0] (
15:25:04LambdaCalculus37GodEater: The Freenode webchat bumped me off.
15:25:20LambdaCalculus37Anyway, I didn't get to ask.
15:25:37 Join Lss [0] (
15:25:38LambdaCalculus37What exactly is gigabeat_flash writing to the flash?
15:32:14 Join jgarvey [0] (
15:37:35 Join nibbler_ [0] (
15:41:54 Join Strife89 [0] (n=nds@
15:42:05 Nick linuxguy4 is now known as linuxguy3 (
15:42:23 Quit patmulchrone (Remote closed the connection)
15:42:29 Quit stoffel ("No Ping reply in 90 seconds.")
15:42:50 Join stoffel [0] (
15:54:46 Join notlistening [0] (
16:01:57 Quit flydutch ("/* empty */")
16:07:11 Quit HBK ()
16:11:06notlisteningwhat is battery drain like when you using lossless formats?
16:11:51notlisteningin comparison to mp3 lets say? Do you guys have stats on it?
16:12:54gevaertshave a look at the CodecPerformanceComparison page on the wiki
16:13:35 Quit Strife89 ("Away!")
16:13:35gevaertsThat doesn't reflect real battery drain yet, you need to still take disk spinups into account
16:13:59GodEaterLambdaCalculus37: did you get an answer ?
16:14:03notlisteningyeah it is more general information that specifics i want
16:19:29LambdaCalculus37GodEater: Not to what exactly gigabeat_flash writes to the flash.
16:19:46GodEaterit writes the contents of bootloader.bin
16:19:55 Join Hillshum [0] (i=cd7ae8fa@gateway/web/freenode/x-9bbd29871bec6c3c)
16:23:31 Nick J-23 is now known as kuciap (
16:23:44 Nick kuciap is now known as J-23 (
16:24:33LambdaCalculus37GodEater: Of course. ;)
16:26:02 Join BryanJacobs [0] (
16:26:25GodEaterof course what ?
16:26:30n1s notlistening: flac uses very little cpu but needs to spin the disk more than mp3 so it should be somewhat close toor slightly worse than mp3 in battery runtime depending on the player
16:28:01*Hillshum thought flac used about the same cpu as mp3
16:28:20LambdaCalculus37GodEater: gigabeat_flash writes the contents of bootloader.bin... I should've guessed. :)
16:28:41GodEaterI actually missed bootloader.bin in the FS# at first
16:28:42LambdaCalculus37Hillshum: But remember that FLAC has to access the drive much more to buffer data.
16:28:48GodEaterand wondered why nothing was happening
16:29:03HillshumLambdaCalculus37: right, though on flash that really doesn't matter much
16:29:06LambdaCalculus37GodEater: So I have to build my own bootloader.bin to flash, right?
16:29:18GodEaterit's attached to the FS task
16:29:27gevaertsuse the latest one there
16:29:27LambdaCalculus37GodEater: Ahh, okay.
16:29:29GodEaterI would recommend NOT building your own
16:31:01*GodEater wonders if LambdaCalculus37 also has his charger with him
16:31:32 Quit __lifeless (Read error: 113 (No route to host))
16:39:53 Join faemir [0] (n=faemir@
16:41:30 Join funman [0] (n=fun@rockbox/developer/funman)
16:43:10HillshumOh, I think I found a specific issue with the Clip keymap
16:44:03 Quit timc (Read error: 60 (Operation timed out))
16:44:08HillshumThe "Erase dynamic playlist?" screen says play to confirm, any other to go back, but select is what makes it confirmed
16:45:42funmani think this problem appears on other screens as well (and other targets : fuze, c200v2, ..)
16:46:11funmanin my opinion the message should say press select to confirm
16:46:39HillshumYes, I've looked at the code to find it some haven't found it yet
16:46:50***Saving seen data "./dancer.seen"
16:47:59LambdaCalculus37GodEater: I have one.
16:48:38rasherHillshum: apps/lang/english.lang, id LANG_CONFIRM_WITH_BUTTON
16:49:26rasherIn fact, this should be changed with langtool, since it can be done for all languages
16:50:32Hillshumhmm. Once I get home (with acces to a unix box) I'll do it
16:51:13rasher../../tools/ −−changetarget −−from "yh*,ipod*,x5,m5,gigabeat*,e200*,c200,h10,h10_5gb,mrobe100" −−to "yh*,ipod*,x5,m5,gigabeat*,e200*,c200,h10,h10_5gb,mrobe100,clip" −−id LANG_CONFIRM_WITH_BUTTON −−inplace *.lang
16:51:42rasher−−to "yh*,ipod*,x5,m5,gigabeat*,e200*,c200*,h10,h10_5gb,mrobe100,clip,fuze" I guess
16:52:11funmanc200 should be changed to c200* everywhere
16:52:55rasherWell that's easy :)
16:53:01Hillshumeverywhere in that area anyway
16:53:18 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
16:53:39rasherIsn't the c200v2 identical (physically) to the c200?
16:53:45LambdaCalculus37It is.
16:53:53LambdaCalculus37As is the e200v2 to the e200.
16:53:55rasherThis was already done for e200
16:54:14 Quit martian67 (Remote closed the connection)
16:54:24rasherI'll do the c200 if you like?
16:54:32funmanyes please
16:54:33 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
16:55:36 Quit martian67 (SendQ exceeded)
16:55:55CIA-71New commit by rasher (r21490): Change c200 to c200* everywhere since the c200v2 is identical to the c200.
16:57:07rasherIt seems that LANG_CONFIRM_WITH_BUTTON should have the logic reversed (such that * is SELECT, and recorder,player and whichever else are PLAY)
16:57:07 Join martian67 [0] (n=martian6@about/linux/regular/martian67)
16:57:46 Join CaptainKwel_ [0] (i=2669ecc2@gateway/web/freenode/x-2720dd0cb2928388)
16:59:10 Quit PaulJam (Nick collision from services.)
16:59:19 Join PaulJam_ [0] (
17:00:43rasherShould mean ever so slightly less work for new targets, since most seems to use SELECT
17:01:46 Join timc [0] (n=aoeu@
17:04:29 Quit LambdaCalculus37 (Ping timeout: 180 seconds)
17:04:50funmanplayer recorder*, ondio*, meizu* av300, ifp7xx, iaudio7, cowond2, m3, mrobe500, logikdax, cretive*, c100, m200*, tpj1022, sa9200, hdd1630, lyre_proto1 fall into the * category
17:05:00funman+ onda*
17:05:16rasherI wonder how many of those actually use play
17:05:30funmanor buttons :)
17:12:33rasherPossibly not worth the bother
17:13:07 Quit PaulJam_ (Nick collision from services.)
17:13:09 Join PaulJam [0] (
17:13:34 Join slam_ [0] (
17:14:13daurnps, note to mcuelenaere / safetydan
17:14:18daurnin helloworld.lua
17:14:25daurnuse "or"
17:14:39 Join LambdaCalculus37 [0] (i=44a0430d@gateway/web/freenode/x-70f33b601778c430)
17:14:47daurneg, local func = rb.lcd_bitmap_transparent_part or rb.lcd_bitmap_part or rb.lcd_mono_bitmap_part
17:14:55daurninstead of your long sequences of ifs
17:17:31*daurn fixes
17:17:40daurnrasher: you able to give me svn perms?
17:18:24daurnor GodEater?
17:18:35CIA-71New commit by alle (r21491): Ensure that the file handle is always closed in text editor plugin (part of FS #10138 by Teruaki Kawashima, minor modifications by me)
17:19:07GodEaterdaurn: hahahahaha - I've been a project member for nearly 5 years, and I don't have svn access :)
17:19:17AlexPdaurn: No, you need to post the patch on flyspray
17:19:23HillshumGodEater: have you every really needed it?
17:19:41GodEaterI've not missed it no
17:20:00GodEaterI've too many other things on the go to work on code for RB unfortunately
17:20:04daurnah ok :(
17:20:07GodEaterand AlexP is far better at breaking the manual than I
17:20:22AlexPIt is always deliberate :)
17:20:29GodEaterwell so you claim
17:26:11 Quit slam__ (Success)
17:30:32 Quit linuxguy3 (Remote closed the connection)
17:30:42 Join linuxguy3 [0] (
17:32:12*daurn submits his first ever rockbox patch
17:32:36LambdaCalculus37daurn: \o/
17:32:41LambdaCalculus37daurn: What's the FS#?
17:32:56daurnnot submitted yet
17:33:26daurnuh, the file headers have copyright mcuelenaere in them... should I change/remove that?
17:34:03funmanif you reclaim copyright on it yes
17:37:36daurnfunman: huh?
17:37:43gevaertsdaurn: if it
17:37:58gevaerts's a small patch, no. If you rewrite most of the file, probably yes
17:38:28HillshumAt most, add yourself along with mcuelenaere
17:38:35 Quit PaulJam (Read error: 60 (Operation timed out))
17:38:54daurnand whats the whole $Id line?
17:39:21gevaertsthat's the last committer
17:39:48 Join PaulJam [0] (
17:40:02daurndo I change/edit that? or is it automatic
17:40:21 Quit nibbler_ (Read error: 113 (No route to host))
17:41:02funmanyou should change it to "$Id$" (or is it "$Id:$" ?)
17:41:12 Nick fxb is now known as fxb__ (
17:41:23gevaertsit should be automatic
17:41:26funmansubversion will expand it when checking it out provided the correct svn property is set on this file
17:41:45gevaertsnormally you don't touch it
17:42:11*daurn now asks: how to submit the modifications?
17:42:22funmanpost a patch on flyspray
17:42:46daurnbut how to I submit it? make a diff?
17:43:04Hillshumdaurn: yes
17:43:08gevaertsyes. svn diff should produce the right format
17:43:10Hillshumsvn diff
17:43:19*Hillshum lost
17:44:13funmansaratoga: did you look at Helix (Real Networks) he-aac fixed point decoder?
17:44:42CtcpPing from gevaerts!n=fg@rockbox/developer/gevaerts
17:46:08daurnyay FS #10379
17:46:31rasherfunman: The helix stuff is not gpl compatible. Can't touch
17:47:01funmanrasher: perhaps there is is a way to redistribute it unmodified?
17:47:23funmanor redistribute the modifications with the original licence
17:47:29gevaertsfunman: without linking?
17:47:41funmanoh, hmm
17:47:43rasherNot without turning Rockbox inside out
17:47:46BryanJacobscould do it user side
17:48:03BryanJacobshave a script that retrieves the sources and builds them on the user's machine
17:48:11BryanJacobsthus, not redistributing
17:48:11Torneyuck :)
17:48:23BryanJacobswell, gevaerts asked
17:48:35gevaertsthat was a rhetorical question!
17:48:51funmangevaerts: linking to the helix decoder is permitted by the licence
17:49:00gevaertsfunman: by the GPL?
17:49:07rasherfunman: But we'd be breaking the GPL
17:49:12gevaertsThis works both ways
17:49:20BryanJacobsif the helix decoder has restrictions > GPL, you can't link with it
17:49:42Tornefunman: everything linked to rockbox has to be GPL licenced, or someting less restrictive
17:49:51rasherUnless they changed it since I looked at it last, it does
17:50:10funmanso we couldn't distribute rockbox as gpl, and helix-aac.codec as the helix/real licence?
17:50:19BryanJacobsnot linked together
17:50:35Tornethe .codec files are already linked to rockbox
17:50:39BryanJacobsyou could try for the "system software" exception
17:50:44BryanJacobsbut you'd have a hard time making that one fly
17:51:07BryanJacobsas rasher said, you'd have to turn Rockbox inside out
17:51:07rasherBryanJacobs: especially considering how monolithic Rockbox is
17:51:10funmanthe solution would be to use a separate codec loader and distribute this one under lgpl
17:51:48rasherYeah, I think at that stage we've just about reached the "inside out" stage
17:51:57*BryanJacobs agrees
17:51:59Torneyou'd have to abstract every API
17:51:59funmani.e. helix-aac.codec as lgpl and helix-aac.bin (loaded by helix-aac.codec) as helix licence
17:52:08 Quit n1s (Read error: 104 (Connection reset by peer))
17:52:15*gevaerts considers that sort of tricks to be cheating
17:52:15Tornefunman: no, that doesn't work either
17:52:19 Join n1s [0] (n=n1s@rockbox/developer/n1s)
17:52:25Tornebecause helix-aac.codec would become under the terms of the GPL by linking to rockbox
17:52:25funmanTorne: oh?
17:52:27BryanJacobs"functionally complete" clause
17:52:33TorneLGPL is less restrictive
17:52:38Tornebut the combined result is then GPL'ed
17:52:50funmanwhich combination ?
17:52:54BryanJacobsthe GPL requires distribution of complete sources anyhow
17:52:56BryanJacobsso it doesn't matter
17:53:01Tornehelix-aac.codec being linked to rockbox's core.
17:53:12BryanJacobshelix-aac.bin is part of the source
17:53:15Tornethat brings the less restrictive part under the full GPL.
17:53:18BryanJacobsthus, must be GPL licensed
17:53:25BryanJacobsTorne: no because it's not redistributed
17:53:38rasherYou *could* distribute source-only, in which case the GPL wouldn't really come into play, but... I'd personally fight that tooth and nail
17:53:51BryanJacobsyou can link GPL to LGPL, can't you?
17:54:02TorneBryanJacobs: Yes, but the combined result is *all* under the GPL.
17:54:07BryanJacobsI think it might invoke the upgrade clause of the... oh yeah
17:54:10Tornethe LGPL'ed bit doesn't *stay* LGPLed
17:54:16BryanJacobsgot it :-)
17:54:38saratogafunman: very very briefly
17:54:41Tornehtat's exactly how the upgrading clause works :)
17:54:44saratogaits not GPLed so we can't use it directly
17:54:53saratogaits supposed to be quite fast though
17:54:56funmani am a bit lost :/
17:55:20Tornefunman: "there is no practical way to do it" is the summary :)
17:55:30funmansaratoga: yes
17:55:34BryanJacobsyou could have GPL'ed dynamic codec loader
17:55:46BryanJacobswhich then loads the helix-licensed he-aac codec
17:55:55BryanJacobsso there's no linked distribution
17:55:58TorneBryanJacobs: only under the system software exception.
17:56:06BryanJacobsRockbox is an OS, no?
17:56:06 Join tessarakt [0] (
17:56:17Torneso that *might* just about count.
17:56:19BryanJacobsso this might work
17:56:57BryanJacobsdoes sound like butchery, though
17:56:59funmanTorne: BryanJacobs could you point me to specific chapter of gplv2 for these exceptions ?
17:57:01gevaerts*only* if we don't distribute that codec, and even then I'm not entirely sure
17:57:36gevaertsThe system software exception doesn't work if you distribute the whole lot yourself
17:57:48funmanah i found 3. and i agree with gevaerts
17:58:05Torneyah, i have i tbackward still
17:58:08 Join evilnick [0] (i=0c140464@gateway/web/freenode/x-fde9ccb111979dcd)
17:58:17BryanJacobssecond-last para of 3
17:58:57BryanJacobsI suppose the last para of 3 nixes this, eh?
17:59:04notlisteningokay can someone confirm my understanding if i wanted to make a player tht was looking to perform well with lossless formats is a larger ram better and how big?
17:59:17rasherPlease stop being scary. Everyone of you!
17:59:20BryanJacobsit says if you make it available from a designated spot then it counts as redistributing - I'd forgotten that
17:59:27funmanso i could say develop a gpl helix-he-aac codec loader for rockbox and distribute it separately
17:59:32funman(from rockbox)
17:59:51gevaertsfunman: I think that would be getting deep into lawyer territory
18:00:06Tornenotlistening: larger ram is always better for eveyrthing
18:00:13gevaertsnotlistening: depends. Hard disk or flash?
18:00:14funmangevaerts: well i am not sure
18:00:25gevaertsfunman: that's why :)
18:00:25saratogayou can't really load a codec like that to get around the GPL because the codec is still used in the OS's process
18:00:37BryanJacobsthe codec loader itself would be A-OK, no?
18:00:46notlisteninggevaerts, can you answer for both?
18:00:50BryanJacobsit's only the people who use it who'd be in violation
18:00:59Tornesaratoga: I think you can, actually, it's not any different from the nVidia shim to load their proprietary DRI module
18:01:06BryanJacobsI mean, the codec loader doesn't even have to load he-aac, any codec could use it
18:01:08Tornebut it would have to be distributed seperately
18:01:18BryanJacobsTorne: why, if it's GPLed?
18:01:22gevaertsnotlistening: on hard disk, as much as you can get, on flash, it doesn't matter as much
18:01:24Torneand it would have to completely provide the API the non-gpl bit used
18:01:39BryanJacobsTorne: agreed on the second part there
18:02:06saratogai'm not really that interested in trying to get Helix's decoder to work
18:02:18saratogai'd rather borrow any good ideas they have an implement them in GPLed code
18:02:27BryanJacobsthat would be stealing
18:02:37BryanJacobsyou know, copyright infringement?
18:02:44gevaertsBryanJacobs: you know, ideas?
18:02:48gevaertsnot code
18:02:57BryanJacobsgeneral ideas can be patented, code is copyrighted
18:02:59TorneBryanJacobs: not if you clone by separation of responsibility, actually :)
18:03:07BryanJacobsbut where does your work become a derivative?
18:03:08funmanthe general idea of a aac decoder is already patented ;)
18:03:20BryanJacobsI mean, if I change the variable names is my code now different?
18:03:20LambdaCalculus37And MP3 decoder as well.
18:03:31funmanas a french citizen i don't really care (for now)
18:03:35Hillshumand mp3 encoder
18:03:36BryanJacobswhat if I take their code and run it through an optimizing compiler, then disassemble that?
18:03:37TorneBryanJacobs: for copyright, it's sufficient to have one guy read the code and write a description of the tricks used to get good performance, then have soeone else who hsa never read the code write a new codec
18:03:49TorneBryanJacobs: no, that's not ok, you are tainted by having read the code
18:03:50BryanJacobsTorne: I agree that clean rooms work
18:04:01funmanTorne: in US perhaps
18:04:05BryanJacobsbut it didn't sound like saratoga was suggesting that
18:04:16Tornefunman: cleanrooming works just about everywhere
18:04:27BryanJacobsin the US we've got the famous IBM case
18:04:30funmanTorne: disassembling works here, that's all i want to know
18:04:44Torneer, what?
18:04:56BryanJacobsthe original reverse engineering!
18:05:00Torneyou can disassemble it and look anywhere :)
18:05:06Tornebut you can't copy what you find, anywhere :)
18:05:09Torneas far as I know :0
18:05:14funmani can study, reverse engineer, and write a new implementation
18:05:41funmanno need to pass my findings to another person who would write the implementation
18:05:44Tornei disbelieve that there is *no* case whree that would be considered a derived work, in *any* jurisdiction that's subject to WIPO
18:05:59 Join Jaykay [0] (
18:06:09Tornesince, for example, i could study, reverse engineer, then reproduce the same code from memory. :)
18:06:14Tornethat would obviously be a derived work
18:06:19Torneand i'm pretty sure that would be illegal in france too :)
18:06:31Torneat which point it becomes a matter of actually getting an opinion on it to be safe, is the point
18:06:33BryanJacobs says that disassembly is OK
18:06:35gevaertsTorne: are you allowed to design a car if you have a drivers license?
18:06:43Tornewhereas i fyou cleanroom it's guaranteed to be ok
18:06:52funmannot a derived work, but an interoperability measure
18:07:19funmanfor example libdvdcss
18:07:29Torneyes, but my point is that's a sliding scale
18:07:34Tornewith no clearly defined point where it's ok
18:07:40Tornewhereas cleanroom implementations are not
18:07:53Tornethings like libdvdcss are 'ok' by most definitions, yes
18:08:06BryanJacobsexcept under the DMCA :-P
18:08:16Tornebut if they'd disassembled a dvd player and then 'accidentally' written a program that was the same exact instructions, that would probably be considered copyright infringement
18:08:31Torneeven if the people who did it claimed it was a coincidence and they only disassassembled it for research
18:08:32BryanJacobsyou could argue that he-aac encoding is a copy prevention algorithm, and there'd be no way to prove you wrong
18:08:33funmanyou can check that the same exact instructions aren't generated
18:08:46Tornefunman: yes. so exactly how different doe sit need to be to count?
18:08:51Tornether eis *no particular point*
18:08:53Torneit's up to a court to decide
18:08:57Tornethat's my point, nothing more
18:09:01*gevaerts thinks that this is getting off topic
18:09:05Torneso just being in france doesn't magically make it ok
18:09:06funmanTorne: linux and windows are both operating systems, but i doubt they share a lot of binary code
18:09:30*rasher agrees with gevaerts
18:09:31funmana decoder is something complex so i doubt 2 implementations share a lot of binary code as well
18:10:00funmanrasher: gevaerts hmm not sure .. licensing and reverse engineering are related to rockbox
18:10:45rasherBut not this general talk
18:10:48gevaertsfunman: dvdcss definitely isn't, and neither are windows and linux
18:11:01LambdaCalculus37Yes, we've just veered way off-topic here.
18:11:08funmansorry then
18:12:50 Part jessor
18:19:08 Quit kachna|lappy (Connection timed out)
18:19:40funmanHillshum: did you see fs#10290 (about lang, there is another string to be changed apart LANG_CONFIRM_WITH_BUTTON)
18:19:48 Part Hillshum
18:19:52saratogathe code is copyrighted, and the implementation is patented, but the underlying ideas are not
18:19:59 Quit petur ("work->home")
18:19:59LambdaCalculus37funman: Too late. :P
18:20:00 Join Hillshum [0] (i=cd7ae8fa@gateway/web/freenode/x-9bbd29871bec6c3c)
18:20:08saratogai can read their solutions to problems, understand them, and then use the general idea if I like
18:20:11Hillshumwhat was that?
18:20:26saratogaclean room disassebly is not needed here since the code is already open source, so no one will care to sue over it
18:22:12LambdaCalculus37Hillshum: funman wanted to know if you saw FS #10290.
18:24:25 Quit Hillshum (Ping timeout: 180 seconds)
18:25:15 Join tessarakt2 [0] (
18:25:29 Quit linuxguy3 (Remote closed the connection)
18:25:53 Quit tessarakt (Read error: 110 (Connection timed out))
18:26:57 Join Hillshum [0] (i=cd7ae8fa@gateway/web/freenode/x-8620d8252111eb6e)
18:28:37HillshumSorry about that, did anyone want me?
18:30:16LloreanHillshum: there are public logs.
18:30:35*Hillshum thanks Llorean for the reminder
18:30:43 Quit timc (Remote closed the connection)
18:32:51 Join timc [0] (n=aoeu@
18:33:15Hillshumfunman: No, I'll take a look
18:34:22 Quit mrkiko ("ofhfhfhfhfhfhfhf")
18:37:28 Join bertrik [0] (
18:40:48LambdaCalculus37The keymaps defined in apps/plugins/lib/pluginlin_actions.c are for rotated screens, I take it? (e.g. BUTTON_RC_FOO)
18:41:14 Quit Hillshum ("Page closed")
18:45:34 Join pyro_maniac [0] (
18:45:45 Join nibbler_ [0] (
18:46:54***Saving seen data "./dancer.seen"
18:48:05 Join flydutch [0] (
18:51:42 Quit evilnick ("Page closed")
18:55:24 Quit Jaykay ("ChatZilla 0.9.84 [Firefox 3.0.11/2009060215]")
18:58:45bertrikmarkun, the MeizuM6Port suggests that some thing like RTC and FM radio are already half done, but I see no code for it in the tree, is there some other place?
18:59:42GodEateris anyone else running the beta build server ?
18:59:54rasherI am
18:59:59GodEateris it doing anything ? :)
19:00:02bertrikGodEater, I'm running the build client
19:00:12rasherErr, yeah, I'm running the client as well
19:00:15GodEaterbertrik: that's what I meant :o
19:00:22*gevaerts runs it as well
19:00:34BagderI plan to do further server work later tonight
19:00:46GodEaterto make it do what ? :)
19:00:49bertrikoh just give us some random builds :P
19:00:59Bagderverify that all things work
19:01:06Bagdertrigger on svn commits
19:01:12*GodEater wants to know who is the bogomips king currently
19:01:22Bagdergevaerts eats us all
19:01:42 Join Hillshum [0] (i=cd7ae85b@gateway/web/freenode/x-a778bcd0cb2a2207)
19:01:49*gevaerts doesn't tell about this bind mount on /proc/cpuinfo
19:02:26BagderI also need to put all the results of a build round together so that build tables can be made from them
19:03:10markunbertrik: maybe it's still in my tree and I forgot to commit it (since it was untested)
19:03:18HillshumCould a build server be powerpcMac?
19:03:23*Bagder goes to hunt down some dinner
19:03:30BagderHillshum: a build client, sure
19:03:32markunI'll try to find it, but it's a bit of a mess
19:03:44scorche|shBagder: how much is second best =P
19:04:02Bagderwell, actually the server can also be on a mac but I doubt we'll run many servers ;-)
19:04:15bertrikI don't think we should commit untested code, but I sure would like to have a look at it
19:04:50 Join evilnick [0] (i=0c140464@gateway/web/freenode/x-36611f53de995700)
19:06:44HillshumBagder: I meant mac, so I'll set it up when I get home
19:06:58*Hillshum needs another USB keyboard for his new mac
19:09:27markunbertrik: in fact, I did commit the tea5760uk.c file. It's just a copy of tea5767.c with some changes by comparing the datasheets
19:10:34bertrikah, ok
19:10:35markunthe LV24020LP might be compatible with the LV24000, but you would have to check the datasheets again
19:10:57markunlet me see if I can find the RTC related files
19:12:48bertrikmarkun, I agree about the logarithmic brightness curve
19:14:14bertrikit currently changes a lot at the low end and very little at the high end of the brightness curve
19:14:49rasherBagder: thoughts about a fallback upgrade plan for cygwin clients? Such as moving it to a temporary location, which will move into place?
19:15:43 Join at0m [0] (
19:16:33markunbertrik: the values I gave you should give enough steps, don't you think? 15 + off
19:17:38bertrikmarkun, I think so, I should try it out and see if it's enough
19:17:59markunbertrik: can't find the RTC files. Someone made them when he asked me if there was anything simple he could do, I forgot who it was. I can't find them in my normal rockbox tree.
19:20:00markunwell, unfortunate, but no too much work lost
19:20:30 Join JdGordon| [0] (n=Miranda@nat/microsoft/x-343ad55677c8489d)
19:20:40 Join Cory` [0] (
19:21:03 Quit Cory` (Client Quit)
19:22:56 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
19:27:21 Nick fxb__ is now known as fxb (
19:39:45 Quit barrywardell (Remote closed the connection)
19:41:29 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
19:43:11 Quit PaulJam (Nick collision from services.)
19:43:20 Join PaulJam_ [0] (
19:43:27 Join Cory` [0] (
19:45:04bertrikit looks much better now with the logarithmic conversion for the backlight
19:47:06 Join kugel [0] (n=kugel@rockbox/developer/kugel)
19:47:33 Join jkl [0] (
19:48:51mcuelenaerebertrik: I'm also interested in logarithmic steps for the Onda VX747's backlight, are you going to commit your changes?
19:50:32 Quit Lss (Read error: 104 (Connection reset by peer))
19:52:43 Quit Cory` ("Ex-Chat")
19:52:46bertriknot very soon I think, but it's quite simple. It's basically just a lookup table with approximately exponential values in it that maps 16 brightness levels to 256 brightness levels
19:53:16mcuelenaereonly 16 brightness levels?
19:53:39kugel16 is enough imo
19:53:54bertrikmarkun gave me a list of values, you could calculate a table with more entries though
19:54:01mcuelenaereoh, the M6 has 256 brighness levels?
19:54:30 Quit markun (Read error: 104 (Connection reset by peer))
19:54:32mcuelenaere(ranging from 0->255)
19:54:34 Join markun_ [50] (n=markun@rockbox/developer/markun)
19:55:24bertrikwith PWM you can basically make as many levels as you like
19:55:27kugelthe gigabeats have a logarithmic curve too I think
19:55:59markun_bertrik: we could have used powers of 2 (-1) but I used sqrt(2) to have twice the number of levels
19:57:39 Nick markun_ is now known as markun (n=markun@rockbox/developer/markun)
19:58:19 Join qwebirc1009538 [0] (i=44a0430d@gateway/web/freenode/x-9cee5fb2a52d6272)
19:58:53 Quit LambdaCalculus37 (Ping timeout: 180 seconds)
19:59:01 Nick qwebirc1009538 is now known as LambdaCalculus37 (i=44a0430d@gateway/web/freenode/x-9cee5fb2a52d6272)
19:59:44 Quit PaulJam_ (Nick collision from services.)
19:59:49 Join PaulJam [0] (
20:02:41bertrikmcuelenaere, the vx747 also uses PWM I see
20:02:52mcuelenaerebertrik: yes, hence I asked :)
20:04:06bertrikhow does it work? it seems to vary the PWM frequency rather than the PWM duty cycle
20:04:19 Quit LambdaCalculus37 (Nick collision from services.)
20:04:22 Quit kugel (Remote closed the connection)
20:04:39 Join cmwslw [0] (
20:04:58 Part cmwslw ("Ex-Chat")
20:06:54 Join LambdaCalculus37 [0] (i=44a0430d@rockbox/staff/LambdaCalculus37)
20:06:59mcuelenaerebertrik: true, it only varies the PWM frequency; haven't played much more than that with it
20:08:01mcuelenaereit always stays off for x cycles and then powers on for 1 cycle (repeat)
20:09:59markunbertrik: ftl.c in openiboot (iphone linux) is over 1000 lines :( I hope we can do it in less.
20:10:36*bertrik too
20:11:46mcuelenaerehmm can plugins access errno?
20:12:31 Join Cory` [0] (
20:13:03saratogais the flash translation layer understood on the M3?
20:13:19bertrikmcuelenaere, it shouldn't be too hard to convert the 747 backlight code to duty cycle variation with a logarithmic curve
20:13:37bertriksaratoga, no, not as far as I know
20:14:01mcuelenaerebertrik: and staying at the same PWM frequency then?
20:14:02 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
20:14:20 Join kugel [0] (i=kugel@rockbox/developer/kugel)
20:14:22gevaertssaratoga: it's probably the same as in newer apple products like the iphone, and people are working in that
20:14:34bertrikmcuelenaere, yes
20:14:44mcuelenaerebertrik: out of interest, what frequency do you use on the M3?
20:14:45 Quit funman (Remote closed the connection)
20:15:05bertrikmcuelenaere, I'm not entirely sure :P
20:15:10 Join kugel_ [0] (
20:15:17 Join funman [0] (n=fun@rockbox/developer/funman)
20:15:42bertrikPCLK / (256 * 101 * 64), but I'm not sure what value PCLK is in the M3
20:15:48mcuelenaereah :)
20:16:30mcuelenaerewhat's it supposed to be according to the datasheets? (what range?)
20:16:31bertrikhm, for 50 MHz PCLK this would be about 30 Hz, it does flicker a bit now
20:17:43bertrikPCLK is max 100 MHz but I don't know what the current value is
20:19:21markunbertrik: according to firmware/target/arm/s5l8700/crt0.S it's 50MHz
20:19:38 Quit PaulJam (Nick collision from services.)
20:19:38markun // FCLK_CPU = 200MHz, HCLK = 100MHz, PCLK = 50MHz, other clocks off
20:19:40 Join PaulJam_ [0] (
20:20:13 Join petur [0] (n=peter@rockbox/developer/petur)
20:25:11 Quit JdGordon| ("Miranda IM! Smaller, Faster, Easier.")
20:27:29 Join JdGordon| [0] (n=Miranda@nat/microsoft/x-9151c2f63a25c0c5)
20:30:05bertrikmcuelenaere, I'll try to make a patch for the vx747
20:30:20mcuelenaerebertrik: thanks :)
20:30:21 Join Thundercloud [0] (
20:30:35mcuelenaerebut you don't need to, I can base myself upon your code too :)
20:31:05 Quit PaulJam_ (Nick collision from services.)
20:31:13 Join PaulJam__ [0] (
20:31:48mcuelenaerehmm does having errno exposed to plugins cause problems?
20:33:02 Quit kugel (Read error: 110 (Connection timed out))
20:33:10bertrikerrno is weird because it is a global variable but needs to be unique for each thread IIRC
20:33:25 Quit AndyIL (Read error: 104 (Connection reset by peer))
20:33:59mcuelenaerehmm does that apply to Rockbox too?
20:34:34 Quit nibbler_ (Read error: 110 (Connection timed out))
20:35:11bertrikI have no idea
20:35:38 Quit LambdaCalculus37 (Ping timeout: 180 seconds)
20:35:44*mcuelenaere will probably just strip all errno related stuff from Lua
20:37:55 Join merbanan [0] (
20:38:50bertrikmcuelenaere, I think this should work
20:39:25pyro_maniacdomonoky: ping
20:40:29pyro_maniacdomonoky: there is a "-silent" after lrelease in the but lrelease has none.
20:40:49 Join nibbler_ [0] (
20:40:51pyro_maniaci mean, lrelease has n silent option
20:41:20*mcuelenaere fetches his DAP
20:41:21domonokymy lrelease has one..
20:41:35domonokywhat does lrelease -? show ?
20:43:59 Join kachna|lappy [0] (
20:44:36 Join toffe82 [0] (n=chatzill@
20:44:40domonokywhat version is this ? (strange that it has the inverted switch -verbose)
20:44:45mcuelenaerebertrik: it seems to work nicely :)
20:44:55pyro_maniacdomonoky: kubuntu seems to use still the 3.3 version
20:45:33domonokyah, but thats probably also too old to build rbutil.
20:46:09pyro_maniaci have to use the qt4 verion manually checks for at least Qt4.3, and i think we need even 4.4 as minimum now.
20:46:47domonokyah, so it should use lrelease-qt4 ?
20:46:57***Saving seen data "./dancer.seen"
20:47:04 Quit stoffel (Remote closed the connection)
20:47:05pyro_maniacthats the way it worked for me
20:47:58 Join LambdaCalculus37 [0] (i=44a0430d@rockbox/staff/LambdaCalculus37)
20:48:11pyro_maniacand in lrelease-qt4 seems a silent options seems to exists
20:48:37domonokyits the same for qmake itself, some distros try to support 3.x with qmake and 4.x with qmake-qt4. dont know how we could fix that in the .pro file.
20:49:57pyro_maniacme either
20:51:15pyro_maniacmaybe grep for the version of that tools?
20:51:24 Nick kugel_ is now known as kugel (
20:51:56 Quit LambdaCalculus37 (Ping timeout: 180 seconds)
20:53:34 Nick J-23 is now known as pciupak (
20:53:46 Nick pciupak is now known as J-23 (
20:54:00domonokywe could also try lrelease-qt4 first, and if it failes/ doesnt exist, use normal lrelease. But dont know how make this in a .pro file :-)
20:56:18 Quit petur ("even linux needs its reboot")
20:57:33domonokyis there is still a normal lrelease binary in the 4.x Qt directory ? ( then we could use the path we know from qmake to call the correct lrelease.)
20:57:39 Quit saratoga (Remote closed the connection)
20:57:39 Quit Hillshum (Remote closed the connection)
20:57:39 Quit evilnick (Remote closed the connection)
20:57:39 Quit CaptainKwel_ (Remote closed the connection)
20:58:47 Join Hillshum [0] (i=cd7ae85b@gateway/web/freenode/x-6ad74b23b5a1d885)
20:59:39 Join petur [0] (
20:59:58 Quit funman ("free(random());")
21:00:39pyro_maniacthe name is still lrelease
21:02:50 Quit robin0800 ("Leaving")
21:03:13domonokyhm, the .pro file already tries to todo this. does qmake- query "QT_INSTALL_BINS" return the correct Qt path for you ?
21:04:09domonokyqmake-qt4 -query "QT_INSTALL_BINS" ofcourse for you :-)
21:08:10 Join evilnick [0] (i=0c140464@gateway/web/freenode/x-9dac9ec83f87a331)
21:08:22 Quit petur ("and another reboot :(")
21:08:54 Quit PaulJam__ (Nick collision from services.)
21:09:02 Join PaulJam_ [0] (
21:10:29pyro_maniacit returns the /usr/bin but there is the wrong lrelease
21:11:13 Join robin0800 [0] (
21:11:39 Join petur [0] (
21:12:44domonokyso both qmake and qmake-qt4 return /usr/bin ? :-/
21:17:09pyro_maniacqmake returns: "**Unknown**"
21:18:34domonokyso i am out of ideas. you could create bug tracker entry, so we dont forget about this issue.
21:22:30 Join Llorean1 [0] (
21:23:26 Join LambdaCalculus37 [0] (i=44a0430d@rockbox/staff/LambdaCalculus37)
21:24:06 Quit Llorean (Nick collision from services.)
21:24:09 Nick Llorean1 is now known as Llorean (
21:25:28pyro_maniacdone in FS #10380
21:34:51 Join FlynDice [0] (n=FlynDice@
21:34:53 Quit BryanJacobs ("Java user signed off")
21:36:37 Join mirak [0] (
21:41:41 Quit Hillshum ("Page closed")
21:49:20 Join ender [0] (
21:50:03 Join fml [0] (n=4fd3ca47@gateway/web/cgi-irc/
21:50:35 Join foaly [0] (
21:52:08CIA-71New commit by mcuelenaere (r21492): Use a logarithmic scale for the Onda VX747 backlight (thanks to Bertrik Sikken)
21:54:03foalyhey all, trying to build rb but have a problem, I've done all the rockboxdev stuff and ran ../tools/configure, but when I do make it gives me an error with what should be a logo file I think with a .c extention and immediaetly exits. It won't even tell m what he problem is
21:56:59foalywill try making again and give the exact error, hang on
21:57:02 Quit PaulJam_ (Nick collision from services.)
21:57:09 Join PaulJam__ [0] (
21:58:33 Quit LambdaCalculus37 ()
22:00:34foalymake: *** [/home/bigd/rbtrunk/build/apps/bitmaps/native/rockboxlogo.240x74x16.c] Error 1
22:00:40 Quit robin0800 ("Leaving")
22:02:10linuxstbfoaly: Typing "make V=1" may give further clues.
22:03:23foalyBMP2RB rockboxlogo.240x74x16.bmp
22:03:40foalyThen it gives me that erro ragain
22:03:45fmlJdGordon, kugel: you are the viewport & status bar experts, right? I'd like to commit this: Any objections?
22:04:29foalythen this: rm /home/bigd/rbtrunk/build/apps/bitmaps/native/rockboxlogo.240x74x16.c
22:05:31JdGordon|fml: I'm at work so cant look at the patch now... if you want to bug me again in 5 hours if you're still around...
22:05:54 Quit ender` (Read error: 110 (Connection timed out))
22:06:28Lloreanfoaly: You might want to mention which target you're building for
22:07:06foalya gigabeat f, target 40
22:07:36LloreanThe last official build for that returned fine.
22:07:45LloreanDo you have any patches applied?
22:08:24foalynope, straight from svn with nothing created except the build folder and config stuff
22:08:35 Part Llorean
22:09:05 Join Llorean [0] (n=DarkkOne@
22:09:32LloreanWell, as I said, the official build returns fine. Did you do *anything* else other than ../tools/configure, choose build 40, then "make"
22:09:48 Quit HellDragon (Read error: 104 (Connection reset by peer))
22:10:00foalyyeah I ran /tools/
22:10:17foalyand selected arm
22:10:18fmlJdGordon|: it's about changing (and restoring back at the end) of "global_settings->statusbar" in a plugin. It's set to false at start and restored before exit. The patch removes those lines.
22:10:45Lloreanfoaly: So you didn't also add the cross-compiler to your path?
22:10:54foalyalso modified my path to include :/usr/local/arm-elf/bin
22:10:55JdGordon|my immediate reaction is that that change shuold eb fine
22:11:12JdGordon|fml: does it even work currently?
22:11:26Lloreanfoaly: Anything else? Something is different about your setup, so I'll need you to figure out what it is.
22:11:53rasherfoaly: try running V=1 make
22:11:55*JdGordon| hopes everyone knows that he hates the current statusbar system and punches kugel in the face to hurry and with his custom list vp patch so it can be fixed
22:12:18foalynothing else I can think of, I'm running on Debian 5.0 if it helps.
22:13:32fmlJdGordon|: that's the point! Applying the patch does not change the appearance! I.e. even without explicitly setting status bar to false (i.e. after applying the patch) the status bar is not displayed in the plugin. So I think it's safe to remove those lines.
22:13:48 Quit slam_ (Read error: 60 (Operation timed out))
22:14:16JdGordon|ok, yeah so go for it
22:14:22JdGordon|which plugin is this for?
22:14:36fmlJdGordon: properties and text_editor
22:14:49JdGordon|(press tab twice) :)
22:15:06JdGordon|ok, yeah should be fine
22:15:21fmlJdGordon|: it won't work in the web irc :-)
22:15:43foalyshould I try deleting and getting the svn repo all over again?
22:16:10kugelfml: looks ok, I think the plugin loader handles preserving the statusbar settings
22:17:20CIA-71New commit by alle (r21493): Do not unnecessarily change and restore the status bar setting (part of FS #10138 by Teruaki Kawashima)
22:17:24gevaertsfoaly: can you provide the full output of "make V=1"?
22:17:27CtcpPing from gevaerts!n=fg@rockbox/developer/gevaerts
22:17:39kugelJdGordon|: you already hate the statusbar system you managed to rework it some month ago with a commit rage?
22:18:01JdGordon|hehe yes... well its better than what it used to be
22:18:08JdGordon|it still needs work
22:18:27rasherI still get statusbars where they shouldn't be, or missing statusbars once in a while
22:18:28foalyBMP2RB rockboxlogo.240x74x16.bmp
22:18:39rasherfoaly: don't paste here
22:18:47foaly make: *** [/home/bigd/rbtrunk/build/apps/bitmaps/native/rockboxlogo.240x74x16.c] Error 1
22:19:15foalyur sorry
22:19:24gevaertsDid you do earlier builds in the same build directory?
22:19:25foalywas in another console and didn't see
22:20:17foalyI haven't built before
22:20:53 Join HellDragon [0] (
22:22:09kugelmcuelenaere: is 0 in that log array off?
22:22:20kugelI mean, does it represent backlight off?
22:22:26mcuelenaereno it doesn't
22:22:43mcuelenaereit is very dim
22:22:54kugel"#define MIN_BRIGHTNESS_SETTING 0" doesn't make much sense then ?
22:23:03mcuelenaereisn't it defined to 1?
22:23:11kugelnot anymore :)
22:23:30mcuelenaereeh sorry, I confused it with DIMNESS_SETTING
22:23:45 Join Llorean1 [0] (
22:24:01gevaertsfoaly: can you run "make clean", and then again "make V=1" and paste the output of that one on a pastebin somewhere (e.g.
22:24:04 Quit Llorean (Nick collision from services.)
22:24:05 Nick Llorean1 is now known as Llorean (
22:24:08mcuelenaerehmm yes, that should probably be changed
22:24:23kugelgenerally, you're supposed to not make backlight off available through brightness setting
22:24:41mcuelenaereit isn't
22:24:54mcuelenaere0 is just very dim :)
22:25:15kugelthat's fine, although it shouldn't say 0 then
22:25:36fmlHrm... Applying star_no_statusbar.patch from FS #10138 does not have any visual effect (in the sim a t least)
22:26:11kugelwell, some weeks ago the statusbar popped up regularly in the menu there, I couldn't reproduce it recently though
22:27:18PaulJam__JdGordon|: btw, tried the new patch from FS #9886 and it showed the same problems like the last versions (i guess that was expected, since it was only a resync). and i had a few warnings during compiling.
22:27:59CIA-71New commit by mcuelenaere (r21494): Onda VX747: make backlight setting range from 1->16 instead of 0->15
22:28:09JdGordon|yeah, I was hoping it maggically got fixed, but i have the same on my mini2g
22:28:22JdGordon|when I get motivation again im going to restart that damn patch
22:28:37fmlkugel: what does the parameter hide_bars in do_menu mean? True: Hide if the status bar is displayed, False: do not change anything?
22:28:39*JdGordon| has a hige todo list for rockbox, and little motivation atm unfortunately
22:29:05kugelfml: ask JdGordon| :)
22:30:07JdGordon|yeah, that sounds about right
22:30:10JdGordon|also the button bar
22:30:15JdGordon|which is a seperate beast :/
22:30:51mcuelenaerehmm does anyone know a way to expand macro's when doing gcc -dM -E <file>?
22:31:38 Quit ender (" Kids. You gotta love them. I adore children. A little salt, a squeeze of lemon--perfect. -- Harry Dresden")
22:32:18fmlBut then I don't see why we should apply it. Why should we hide the status bar when the menu is displayed?
22:32:54JdGordon|because the menu used to draw the statusbar manually inside its viewport
22:33:00JdGordon|it gets fiddly now...
22:33:26CtcpPing from gevaerts!n=fg@rockbox/developer/gevaerts
22:34:45fmlbertrik: I'm disappointed by the comment in config-ondavx747.h: MAX is defined as 16 but the value 10 is commented as "full brightness." Is that correct?
22:35:18 Quit Llorean ("Leaving.")
22:35:24fmlmcuelenaere: this was for you, not for bertrik!
22:36:13mcuelenaerefml: bertrik changed that, I don't know what effect that define has..
22:36:14 Join robin0800 [0] (
22:38:05fmlmcuelenaere: I don't know what it all is about, just read the comment and was slightly disappointed :-)
22:38:46mcuelenaerefml: well I initially used the full brightness for that define, but then bertrik changed it and I assumed he did that with a good reason :)
22:38:53 Quit jfc (Remote closed the connection)
22:39:32foalyput the output through, I guess I give you the page url?
22:40:16 Join jfc [0] (
22:41:16*gevaerts doesn't see what's wrong
22:41:54fmlgevaerts: I thought that "full" and "max" are the same
22:42:06gevaertsfml: ?
22:42:27foalyit almost sounds like it wants to compile an image?
22:43:15gevaertsin a way it does
22:44:16foalybut why isn't it actually giving an error?
22:44:26gevaertsfoaly: the only thing I see is that you have v=1, not V=1, so we don't get all verbose output
22:44:53foalyah oops
22:45:02foalyshould I redo that and paste again?
22:45:09bertrikfml, we can either change the comment of the value
22:46:32fmlbertrik: if all makes sense then we don't have to change anything. But does it make sense as it stands now?
22:46:43 Join funman [0] (
22:47:00***Saving seen data "./dancer.seen"
22:47:01fmlbertrik: just removing the comment would be fine IMO
22:48:11 Join n00b81 [0] (n=taylor@unaffiliated/n00b81)
22:49:15 Join Hillshum [0] (i=4ba5ebce@gateway/web/freenode/x-636601184354ed2d)
22:49:18rasherBagder: like this?
22:49:26 Part n00b81 ("Leaving")
22:49:40rasherBagder: sorry,
22:49:42bertrikthe patch I made was a proof-of-concept that I could not test myself, so it didn't bother to be 100% accurate on the comments
22:49:58bertrikmaybe the default should be max brightness, like it was before
22:50:04Bagderrasher: yes, looks fine to me!
22:50:12 Quit merbanan ("Leaving")
22:50:28rasherI guess that means isn't optional anymore
22:50:54Bagderand thus that script should take the arguments
22:50:56fmlbertrik: or that way. But it's weird as it stands now
22:50:57rasherIt could try to do the move inside perl
22:51:01obooh joy - the View seems to have runtime detection during it's lcd_init code for two types of unknown renesas controllers
22:51:23Hillshumwhat's the best way to get gcc on osx for rockbox?
22:51:42kugelbertrik: I think it shouldn't
22:52:14rasherBagder: how about adding rename("$", "$perlfile"); That way is only really mandatory on cygwin or other weird archs
22:52:45Bagderit's still needed as it must restart the perl script after an update
22:52:47fmlShall I just remove the comment?
22:52:48funmankugel: do you know why with the fuze, we can skip cards only one by one in solitaire ?
22:53:15rasherBagder: but the user could be doing that in other ways
22:53:21bertrikfml, do you have a vx747?
22:53:26kugelfunman: yea
22:53:58fmlbertrik: no. But IMO either the default value should be changed or the comment removed
22:54:04rasherBagder: so this:
22:54:44bertrikfml, better discuss it with mcuelenaere then, is this really such a big deal?
22:54:46kugelfunman: should be changed to also include the fuze
22:54:51bertrikI can't be bothered to be honest
22:54:59kugelbut solitaire's way is definitely weird
22:55:08 Join Rob2222 [0] (
22:55:08rasherCompletely untested, since I don't have a build server around, but surely there are no bugs!
22:55:23kugelalso, I can't understand why the ipods doesn't use the wheel as the fuze and e200 do
22:55:56fmlbertrik: it's not a big deal at all, just a little bit confusing for someone who reads the code
22:56:00Bagderrunning a build server is actually just a matter of running the script
22:56:05Bagdernot that you have to
22:56:16 Quit nibbler_ (Read error: 110 (Connection timed out))
22:56:23rasherShould I just commit this, or can we do some sort of test?
22:56:31gevaertsYou mean there's *two* scripts for the client and only one for the server? That's not fair!
22:56:53fmlbertrik: like you say in a comment "And now we'll assign 2 to it", but have "x = 5;" as code
22:56:56Bagderrasher: commit it, and we'll test it later
22:57:41CIA-71New commit by rasher (r21495): Make the update feature two-step. Hopefully this should be more likely to succeed.
22:58:09mcuelenaereis DEFAULT_BRIGHTNESS_SETTING also the default setting for apps/ or does that have another value?
22:59:11kugelit's also the intial value of the backlight_brightness variable
23:00:28foalyformatting probably broke though
23:01:07gevaertsfoaly: what does 'ls -l x16.bmp > /home/bigd/rbtrunk/build/apps/bitmaps/native/rockboxlogo.240x74x16.c
23:01:10gevaertsmake: *** [/home/bigd/rbtrunk/build/apps/bitmaps/native/rockboxlogo.240x74x16.c] Error 1
23:01:25gevaertsfoaly: what does 'ls -l /home/bigd/rbtrunk/tools/bmp2rb' say?
23:01:31CIA-71New commit by mcuelenaere (r21496): Onda VX747: fix DEFAULT_BRIGHTNESS_SETTING
23:02:05kugelmcuelenaere: why did you change that?
23:02:29mcuelenaerekugel: because 10 is quite dim
23:02:41kugelimo backlight brightness shouldn't be at maximum by default, I can't tell a target that does it also
23:03:03foaly-rwxr-xr-x 1 bigd bigd 25213 2009-06-24 07:48 /home/bigd/rbtrunk/tools/bmp2rb
23:03:11 Join n00b81 [0] (n=taylor@unaffiliated/n00b81)
23:03:27kugelmcuelenaere: I assume that if 10 is too dim already then the log table isn't so good?
23:04:06 Quit petur ("Zzzzz and only half the work done, thanks to linux :(")
23:04:22foalyis that file perms?
23:04:31gevaertsyes. They're ok
23:04:35gevaertsDo you have /home/bigd/rbtrunk/apps/bitmaps/native/rockboxlogo.240x74x16.bmp ?
23:04:42 Join _lifeless [0] (n=lifeless@
23:05:21mcuelenaerekugel: no, but I think first-time users would prefer full brightness over 'a more dim one'
23:06:01foalyyep I do
23:06:12 Join nibbler_ [0] (
23:06:29foalyoh, this could be it
23:06:37foalyI think root has it or something
23:06:55gevaertsyou checked out as root?
23:07:17rasherOr you tried to build as root at one point?
23:07:53foalyyeah, but I also did chmod -R 777 rbtrunk
23:08:17foalyshouldn't that have made it accessible to anyone?
23:08:52*Hillshum sees rasher also changed LANG_SELECT_WTH_BUTTON to use select for most targets
23:09:12rasherHillshum: did I? Didn't I just talk about it?
23:09:20*rasher panics
23:09:29gevaertsfoaly: if you have a good connection, maybe it's easier to just do the checkout again. No need to re-run rockboxdev,sh though
23:09:48gevaertsThis time do everything as the same (normal) user though
23:09:50foalyah ok. will run as bigd this time heh
23:11:16funmankugel: perhaps the people who wrote the ipod keymap didn't like to play solitaire this way
23:12:20Hillshumrasher: I'm pretty sure you did
23:12:47 Quit HellDragon (Read error: 54 (Connection reset by peer))
23:12:47funmanHillshum: i'm pretty sure he didn't, check the svn logs ;)
23:13:22rasherHillshum: No, just c200
23:13:25rashereh, c200v2
23:14:23*Hillshum cant' get the right view online to tell and can't remember for sure what it was to begin with
23:14:31*Hillshum also has to go
23:14:32CIA-71New commit by alle (r21497): Delete the save file immediately after loading it to avoid false tries at the end (part of FS #10138 by Teruaki Kawashima, minor modification by me)
23:15:58funmankugel: r10984 "solitaire: better layout for ipods, allowing one-handed operation during the game"
23:17:12funmananyone with an ipod willing to test a solitaire keymap change?
23:19:20CIA-71New commit by alle (r21498): Clear the display before showing the splash (part of FS #10138 by Teruaki Kawashima)
23:19:21kugelfunman: well, it didn't remove scrollwheel usage though
23:19:47 Join saratoga [0] (i=9803c6dd@rockbox/developer/saratoga)
23:20:01funmankugel: nope, it was added though, perhaps amiconn didn't think of the wheel to change between cards stacks?
23:20:25kugelit makes perfectly sense imo
23:20:51funmani don't want to make a change to a target i don't own
23:21:22kugelI asked lamda to test such a patch a while ago, not sure if he actually did
23:21:30funmanif this change is ok we could use #define SOLITAIRE_USE_WHEEL and put 2 additional cases in the button switch (BUTTON_LEFT|REPEAT)
23:21:35 Part n00b81 ("Leaving")
23:21:53fmlI have committed all but one patches from FS #10138. The remaining patch (star_no_statusbar.patch) shouldn't be committed IMO. Could anybody look into it and confirm that? Then I'd close the FS task.
23:21:55kugelwe have HAVE_SCROLLWHEEL for that
23:22:31 Quit Hillshum (Ping timeout: 180 seconds)
23:22:47 Quit flydutch ("/* empty */")
23:23:09kugelalso, I rather have SOL_LEFT_REPEAT added instead of hacking the return value of button_get
23:24:28funmanand define it to BUTTON_NONE for ondio for example ? (where BUTTON_RIGHT|BUTTON_REPEAT is mapped to something else)
23:24:53kugeleither that, or just don't define
23:25:06kugelthe case can be #ifdef'd
23:25:10 Quit n1s (Read error: 110 (Connection timed out))
23:27:07amiconnfunman: There is a reason why the wheel moves through cards within the stack and not between stacks on the ipod. The ipod has no "natural" up/down buttons, but only left/right buttons. Hencec those need to be used for horizontal movement, hence leaving the wheel for vertical movement
23:27:19amiconnAlso, this matches how lists use the wheel
23:28:02funmanthanks, so kugel i think your solution is good
23:28:35Mikachui have a patch for sudoku which makes the direction you move when scrolling dependant on where you initially touched the wheel :)
23:28:44*amiconn wonders what problem needs a solution in solitaire
23:29:13kugelamiconn: no natural what? just because of the label?
23:29:19pixelmaMikachu: isn't there a "switch direction" button in Sudoku now (press select or something)?
23:29:26 Quit nibbler_ (Read error: 110 (Connection timed out))
23:29:27kugelyou can hardly compare solitaire to lists, really
23:29:32Mikachumaybe, but the patch is from 2007 or so
23:29:47kugelshould you also use left/right in pf then?
23:30:19*kugel doesn't agree to that reasoning
23:30:45CIA-71New commit by bagder (r21499): Allow a "commander" to connect to the server and instruct it what to do - ...
23:31:03pixelmawell you go through lists vertically and you go through pics on pf horizontally - so I can't follow that reasoning...
23:31:44funmanfuze has play/menu buttons, they are not mapped to up/down in the OF
23:31:44*rasher doesn't understand this discussion
23:31:45pixelmaas in... I think the way you navigate there is different
23:31:46kugelhence I'm asking if we shouldn't rather use left and right for PF
23:31:54rasherSurely a wheel can be seen as both horizontal and vertical motion?
23:31:58 Quit pyro_maniac ("Leaving.")
23:32:04kugel(ironically that is)
23:32:14amiconnkugel: Solitaire needs horizonal and vertical movement, like many other places. Also, did you ever try moving down on an ipod with 'Play'?
23:32:32kugelno, is that a problem?
23:32:41funmanamiconn: horizontal movement is just more used than vertical
23:32:42Mikachui had to change the vkeyboard to use menu/play for up/down, it's unusable with the scrolling imo
23:32:42amiconnThe ipod will switch off if you hold too long
23:32:45 Join ender` [0] (
23:32:46*Bagder is glad about his new "commander" system to the buildmaster
23:32:55amiconnfunman: So what?
23:33:00Bagderallows me to connect from remote and start a build
23:33:02Mikachuamiconn: you don't use the jpeg plugin much? :)
23:33:14funmani would use the wheel for the most used movements
23:33:23kugelamiconn: so? that surely is a general problem, not really related to solitaire
23:33:34kugelfunman: exactly
23:33:41amiconnMikachu: I do, but I don't use the ipods that much. And I don't like up/down on Menu/Play there either, just that there is no better way when you also need zoom
23:33:49amiconnSolitaire needs no zoom though
23:33:52 Join Zarggg_ [0] (
23:33:59Mikachuamiconn: what i mean is you have the shutdown problem in the jpeg plugin
23:34:11amiconnI know...
23:34:20 Join slam_ [0] (
23:34:21amiconnfunman: I wouldn't
23:34:23Mikachu(but i changed that to select+play) :)
23:34:37funmanamiconn: isn't it easier to use the wheel ?
23:35:07funmanis the ipod wheel different from sansa e200 ?
23:35:14 Quit Zarggg (Read error: 60 (Operation timed out))
23:35:20pixelmasure - it's a touch wheel
23:35:53amiconnThe wheel is crap for precise control. You can move fast - but nothing more
23:35:53 Quit Zarggg_ (Read error: 104 (Connection reset by peer))
23:36:09funmanthat's what i thought : does it make navigation much different from the e200? because i find it much more easy to use than buttons in that case
23:36:32 Join kugel__ [0] (
23:36:38funmanamiconn: you can always go back.. and that is faster anyway than using buttons
23:36:38 Quit kugel (Nick collision from services.)
23:36:46 Nick kugel__ is now known as kugel (
23:37:00 Quit _lifeless (Read error: 60 (Operation timed out))
23:37:03amiconnFor at least up to 5 steps, I'm certaily faster using buttons
23:37:28 Quit Rob2222 ()
23:37:36kugelamiconn: so it's boils down to your personal opinion about the scrollwheel?
23:37:59*Mikachu gets crazy thoughts about the sensitivity of the scrollwheel being set dynamically by where you make initial contact
23:38:15amiconnWith buttons, I just press e.g. 4 times if I want to take 4 steps. With the wheel, I touch, move, notice that I moved to far, have to stop a bit, look how much too far it went, move back etc
23:38:34notlisteningthe e200 and speech is a bit of a pain but you ahve to be light with it
23:38:42amiconnIt is not an opinion, it is an experience
23:39:07pixelmaalso, at least on the e200 the buttons are seperated from the wheel unlike on the Ipod
23:39:36*kugel thinks that the wheel should be used for the most common movements, consitently
23:39:43rasherBagder: cure for hung builds?
23:39:55funmanpixelma: not on the fuze though
23:40:09*Mikachu got a sim build
23:40:19Mikachuis "gigabeats" a sim or some model is "S"?
23:40:29Bagderrasher: nah, I hope to use it to trigger builds, manually and automated
23:40:43Bagdereven from remote I mean
23:40:45amiconnkugel: The wheel should be used consistently wrt directions. On the ipod it means to use it for vertical movements, unless zooming is also needed
23:40:53BagderMikachu: a model
23:40:57kugelamiconn: aren't you actually usually doing less steps for going through 1 stack?
23:41:09notlisteningwould you not have a scroll wheel if given a choice and what else would yoy use?
23:41:25kugelamiconn: so left/right for pf too, yes?
23:41:29amiconnPlain, simple buttons, in a directional cross
23:41:42pixelmafunman: I know, but it still is a bit different I guess if the wheel movement is also mechanical
23:41:43amiconnkugel: ?
23:42:03kugelwhat are you refering to?
23:42:10rasheramiconn: Why should it? If there's only one direction available..
23:42:18amiconn[23:41:25] <kugel> amiconn: so left/right for pf too, yes? <== /me doesn't understand
23:42:23notlisteningi had a scroll wheel on the side of a device with push and pull and click in movements that worked like a crea for doing everything
23:42:35kugelyou don't scroll vertically there in the album viewe
23:42:35Mikachulike a what?
23:42:51amiconnnotlistening: Actually several targets have those, and I prefer those targets for several reasons, one of which is the button layout
23:42:55kugelyou just said the wheel should only be used for vertical scrolling and zooming
23:42:58 Join Zarggg [0] (
23:43:28 Join ladanz [0] (
23:43:42*kugel doesn't see the relation between scrollwheel and vertical scrolling
23:43:55ladanzhi there; just updated to version 3.3; where is the sleep timer???
23:43:59amiconnAs there is no vertical movement in pf, the wheel can be used for horizontal movement. I would map the Left/Right buttons in addition to that though (dunno whether it is already the case)
23:44:27amiconnkugel: There is no relation in general, but there is on the ipod when you need both directions
23:44:32rasheramiconn: I think kugel assumed you meant the wheel should never be used for horizontal movement
23:44:41rasherAs did I, for a moment
23:45:07gevaertsladanz: which target?
23:45:13kugelamiconn: and why vertical for that case?
23:45:25*amiconn doesn't want to repeat himself
23:45:37ladanzsansa e260; when u mean that gevaerts
23:45:38kugelI still don't see a relation, not even for cases where both directions are needed
23:45:56 Join Llorean [0] (
23:46:16amiconn23:27 ...
23:46:20CIA-71New commit by mcuelenaere (r21500): Lua: expose BUTTON_* defines
23:46:22gevaertsladanz: the manual says it's in System -> Time and Date
23:46:40rasherkugel: Makes sense to me. When both directions are available, you can pick either "wheel = horizontal" or "wheel = vertical". One is easier to operate than the other, so you *always* make it do that.
23:47:21kugelone is easier to operate?
23:47:26 Join barrywardell [0] (n=barrywar@
23:47:34ladanzyes it is ^^; thanks alot gevaerts ! next time i will look up the manual myself :-D
23:47:47rasherkugel: According to amiconn, it is easier to use the buttons for left+right, rather than up+down.
23:47:52Mikachuwhen you need two directions, i think using the wheel for one of them is _really_ annoying and hard to use
23:48:06rasherMikachu: I tend to agree
23:48:07LloreanMikachu: Depends on whether you "need" two directions or not
23:48:13LloreanFor example, the Jewels grid seems to be 2D
23:48:13ladanzbut it seems to me, that this is an unusal spot; because i set it every time....
23:48:15Mikachulike in the keyboard entry
23:48:21LloreanBut scrolling 1D with the wheel can be just as effective
23:48:30Mikachusokoban does it right :)
23:48:45LloreanAs long as "leave the right edge" means "enter on the left edge, one line lower" it can be a nice way to navigate a grid
23:49:01kugelMikachu: that's also true for many places. not solitaire though imo
23:49:09gevaertsladanz: it's been debated... If it's something you set while using the player normally, does it belong in settings?
23:49:26Mikachukugel: solitaire is more an array of arrays rather than a matrix :) so i think it is alright there
23:49:33amiconnrasher: It is more precise, not necessarily always easier. But it is also more logical
23:50:09*Llorean would prefer in all 2D situations to have button-based controls at least as an option
23:50:45 Join HellDragon [0] (
23:51:13kugelconfigurable buttons?
23:51:27rasherJust that both works, I guess
23:51:34gevaertskugel: living dangerously? ;)
23:51:51 Quit PaulJam__ (Nick collision from services.)
23:51:53kugelgevaerts: No, I was wondering of Llorean meant that
23:51:57 Join PaulJam_ [0] (
23:52:01rasherIt gets a bit tricky in the vkeyboard since play accepts, currently
23:52:14rasherThe vkeyboard is horrible to use on e200
23:52:33Lloreankugel: No, not configurable buttons. :-P
23:52:51LloreanThe vkeyboard ranges from "mildly annoying" to "horrible" on basically every target I've used it on.
23:52:54Mikachui did this for vkeyboard (bottom hunk);a=commitdiff;h=5ed41f1660a
23:53:00LloreanThere's no way to make it intuitive. You either *have* to read the manual, or do some guessing.
23:53:07 Join Rob2222 [0] (
23:53:09kugelI was having a patch making it a bit better, up/down/left/right choosing characters, long select for accapting, scrollwheel for moving the cursor in the line
23:53:10Mikachubut i guess you want to avoid "combos"
23:53:27 Quit fml ("CGI:IRC")
23:53:34pixelmakugel: in solitaire the long play is a problem though as someone who wants to go through the cards vertically might also want to hold the button to go down repeatedly..
23:53:45*foaly cheers!
23:53:45rasherkugel: long select for accepting seems a bit dangerous
23:53:57foalyit works!
23:53:58LloreanThe main problem with vkeyboard that I have is, without manual, you have to guess what will be "accept" and "cancel"
23:54:11kugelI used it in my custom build days, I was very happy with it
23:54:17Mikachuusually menu+select is cancel on ipods
23:54:18 Join fml [0] (n=4fd3ca47@gateway/web/cgi-irc/
23:54:22Bagderhah, in the end I got 68 builds back ;-O
23:54:32GodEaternot bad out of 22 :D
23:54:40Bagderout of 59 actually
23:54:43Bagder9 bonus
23:54:50Mikachuhow many from me? :)
23:55:05ladanzgevaerts, is there a way to write it down in some config; so it's enabled when starting?
23:55:14GodEaterBagder: so that happens if two clients finish the same build at once ?
23:55:21rasherBagder: Tried to kill a build while it was already uploading?
23:55:27 Quit barrywardell ()
23:55:30gevaertsladanz: I don't know, sorry. I've never used that feature
23:55:32Bagderrasher: that's a good theory, yes
23:55:48rasherBagder: Maybe a "CANIUPLOAD?" "NOALREADYGETTINGTHAT!"
23:56:13Bagderno, we need to allow parallel uploads
23:56:13gevaertsrasher: no. What if this upload is real slow?
23:56:35rasherGood point
23:56:37Bagderthe server gets told once the zip+log is done uploading, not before
23:56:38amiconnLlorean: Imo accept should somehow relate to the standard select, and cancel to the standard cancel
23:56:41rasherAnd it's not going to be a problem anyway
23:56:43ladanzgevaerts, alright; thanks anyways. bye :-)
23:56:47 Quit ladanz ("Leaving")
23:56:49rasherSince it'll start building regardless
23:56:50rasherIgnore me
23:57:10 Quit fml (Client Quit)
23:57:14Lloreanamiconn: That would help, yes.
23:57:26 Join barrywardell [0] (n=barrywar@
23:57:56rasherHow about an OK "key"?
23:58:01rasherIn the keyboard, that is
23:58:07LloreanReally, *most* of vkeyboard could be standardized to 6 buttons (up, down, left, right, select, cancel) with long-select being "accept" and long-cancel being "cancel", short cancel being "backspace"
23:58:14LloreanBut that doesn't give you pageflipping or anything, just very basics.
23:58:23*GodEater proposes all commands used by the build server are done in LOLSpeak
23:58:32*Mikachu vetoes
23:58:41*rasher double vetoes
23:58:53*scorche|sh vetoes the veto

Previous day | Next day