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 2010-06-12

00:00:04TheSevenkisak: we also have thumb on lowmem targets now
00:00:35amiconnjhMikeS: Do you use the patched binutils?
00:00:44 Join stripwax [0] (
00:02:59 Join obo_ [0] (~obo@rockbox/developer/obo)
00:03:12***Saving seen data "./dancer.seen"
00:04:38 Quit anewuser (Quit: for SELL 2 by the price of 1 now!)
00:05:54pixelmaZagor: could you look up in the manuals build log what went wrong in the Iaudio manuals (they seem to be missing but work for me. I have some local changes though and would like to know if those might fix it or if there was/is something else going on)
00:06:04 Quit obo (Ping timeout: 264 seconds)
00:06:25 Quit pjm0616 (Quit: pjm0616)
00:06:34jhMikeSamiconn: I haven't gotten to setting that stuff up yet. Isn't that for thumb?
00:06:48amiconnAh yes it is
00:07:18pixelma(or Bagder)
00:07:31Zagorpixelma: I'm not sure what is an error and what is just a warning
00:07:41ZagorOverfull \hbox (2.41475pt too wide) in paragraph at lines 14−−21
00:08:03Zagor! Undefined control sequence.
00:08:03pixelmathose are warnings, did compiling stop somewhere?
00:08:19gevaertsUndefined control sequence looks like an error
00:08:22pixelmaZagor: the latter sounds more like it
00:08:25Zagor<argument> \ActionKbdPageFlip
00:08:36*jhMikeS momentarily thought pixelma was insinuating Bagder was a thumb
00:08:57pixelmaah, ok. Got that in my local changes and wanted to commit anyway :)
00:09:27pixelmalooking myself now what is wrong in the Player's manual, guess it's something similar too
00:09:49 Join perrikwp [0] (
00:10:06 Quit M3DLG (Ping timeout: 252 seconds)
00:11:23 Quit wodz (Quit: Leaving)
00:14:01CIA-6New commit by Buschel (r26802): Build iPod Video with EABI toolchain.
00:15:34CIA-6r26802 build result: All green
00:17:53gevaertsBuschel: you should also update the builds file I guess
00:18:06amiconnAre the specific things to watch out for when testing eabi builds?
00:22:22 Join pjm0616 [0] (~user@
00:23:03 Quit bertrik (Quit: De groeten)
00:23:26 Quit jgarvey (Quit: Leaving)
00:26:42Buschelgevaerts: I never did this before. Is this fine? ->
00:26:59jhMikeSamiconn: it's either worked or not so far (with only 3g not doing well)
00:27:12gevaertsBuschel: yes, looks good
00:27:20jhMikeSh10 5/6 gb seems to be ok too
00:28:23Buschelgevaerts: thanks, will submit now
00:28:49CIA-6New commit by Buschel (r26803): Build iPod Video with EABI toolchain.
00:29:15amiconnDid the bootloaders get testing too?
00:29:46Buschelif I didn't get anything wrong :o)
00:30:15Buschelbtw, I assume the 64MB iPod Videos behave the same as 32MB iPods. We will in short.
00:31:22jhMikeSummm, h10 5gb supposed to not have a separate bootloader build? (or am I overlooking something)
00:32:08gevaertsIt actually *should* have two separate bootloader builds
00:32:15 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
00:32:57jhMikeSI don't see it in the table nor in www/buildserver/builds
00:33:13CIA-6New commit by pixelma (r26804): Various minor corrections to some Iaudios' and their remote's keymap files. Fixes these manuals which were broken since r26617 and a bit more.
00:33:26amiconnNot all possible bootloaders are built by the build system
00:33:26 Quit Luca_S (Quit: CGI:IRC (EOF))
00:33:53jhMikeSwhy some and not others?
00:34:03 Part Zagor
00:34:04jhMikeStoo similar for those?
00:34:11gevaertsI suspect they got overlooked
00:34:53CIA-6r26804 build result: All green
00:35:05amiconnToo similar.
00:35:35amiconnBack then the build system wasn't *that* powerful.
00:35:57jhMikeSnow the build system builds you
00:36:50amiconnTargets obviously need to be built in all variants, but bootloaders and sims aren't offered for download, so we decided to build only as many bootloaders and sims as necessary to get good coverage of the code base
00:41:28 Quit petur (Quit: Zzzzz)
00:41:35amiconnHmm, eabi H10 6GB bootloader is larger than the old one !?
00:43:20amiconnAnd booting the OF seems to hang
00:44:38amiconnActually it seems to work, but the OF fails to access the LCD
00:45:22CIA-6New commit by pixelma (r26805): Enable the IAUDIO_RC_PAD macro for the M5 and X5 too, so that tables using this instead of the platform files' actions can be filled too. Complete the ...
00:46:01*jhMikeS just got it to load OF
00:46:41gevaertsamiconn: is that an eabi regression, or is it a regression since the last bootloader release?
00:46:49CIA-6r26805 build result: All green
00:47:18amiconnActually the eabi bootloader is smaller than an old-abi svn bootloader
00:47:23jhMikeSgevaerts: OF is playing music here so I can't say I'm experiencing anything
00:47:29amiconnBut it is larger than the released bootloader
00:48:20 Quit ender` (Quit: There are times, Sember, when I could believe your mother had a secret lover. Looking at you makes me wonder if it was one of my goats. -- David Gemmel, The King Beyond the Gate)
00:48:22CIA-6New commit by pixelma (r26806): Use the correct 'RC' macro for the remote column in the pitch screen button table.
00:49:15amiconnSeems to be an eabi regression
00:49:39amiconnAn old-abi svn bootloader works (but there's some viewport problem when displaying the messages)
00:49:49CIA-6r26806 build result: All green
00:50:02amiconnIt might be my own code though (I have some uncommitted lcd code for H10 5/6GB)
00:51:03amiconnThis code works in main rockbox with both abis
00:51:58 Quit notlistening (Quit: Leaving)
00:52:02jhMikeSamiconn: mine's unaltered SVN
00:52:22amiconnAnd the bootloader manages to start the OF so that the latter works correctly?
00:55:20 Quit stripwax (Ping timeout: 240 seconds)
00:55:24Buschelgood night!
00:55:28 Quit Buschel ()
00:55:45pixelmaAlexP: the virtual keyboard button table was even threefold (one for the Player). The latter is still there (build fails though because the note above is also included which uses an \Action macro which the Player's platform and the real c code doesn't have). Easy fix would be to just \nopt that part. More "complicated" would be to add the manual actions anyway. What do you think?
00:55:55 Quit bmbl (Quit: Bye!)
00:56:56pixelmaI'm wondering a bit what would happen to those if the \Actions really get autogenerated at some point in time and the target doesn't actually have it in c
00:59:10 Join chrisb [0] (
01:08:42amiconnThere seems to be a slight bug with the selection bar in PF
01:08:56amiconnOther than that I couldn't find problems with eabi on H10 6GB
01:12:46amiconnjhMikeS: Will you activate it?
01:14:08 Quit ranma (Read error: Operation timed out)
01:14:11 Join ranma [0] (
01:16:24amiconnLoading an eabi plugin from the running old-abi core freezes
01:24:32jhMikeSactivate? yeah, sure. I'm not having any problems.
01:24:44ssorgatemamiconn: Mmm it didn't one month ago or so
01:25:05amiconn"it" being?
01:25:30ssorgatemit = [01:16] <amiconn> Loading an eabi plugin from the running old-abi core freezes
01:25:43amiconnWell, it just did on my ipod color
01:26:03ssorgatemit isn't supposed to work, anyways
01:26:28amiconnI had the old core still running, because there was a rolo checksum error (that happens occasionally and is definitely no eabi problem)
01:27:11amiconnThen I accidentally clicked on rockbox-info.txt instead of rockbox.ipod, firing up the text viewer -> freeze
01:27:41ssorgatemMmm I don't think I tried the text viewer
01:27:56 Quit Jerom (Quit: Jerom)
01:28:05ssorgatemI remember rockboy worked
01:28:35ssorgatemor even non-eabi doom in a eabi-rockbox
01:28:39amiconnI get some failed album art images in PF with eabi
01:30:49 Quit evilnick_B (Quit: Page closed)
01:35:37 Quit ssorgatem (Remote host closed the connection)
01:37:21CIA-6New commit by pixelma (r26807): Quick fix of the Player's manual which was also broken in r26617. I overlooked that the Player uses yet another button table...
01:37:22CIA-6New commit by jethead71 (r26808): Build H10 5/6GB with EABI toolchain.
01:39:09CIA-6r26807 build result: All green
01:40:27CIA-6New commit by jethead71 (r26809): Build H10 5/6GB with EABI toolchain.
01:40:33 Quit simonrvn (Read error: No route to host)
01:41:19amiconnjhMikeS: Can you check the PF selector bar (the one in the tracklist when "entering" an album)?
01:41:36amiconnThere seem to be some wrongly coloured horizontal lines
01:41:54 Quit liar (Ping timeout: 258 seconds)
01:41:56jhMikeSI see those, yes
01:41:59CIA-6r26809 build result: All green
01:42:00amiconnHappens at least on my H10 6GB as well as iPod color
01:42:11amiconnBut it doesn't happen on arm old-abi or cf
01:42:17pixelmaah, was just confused about the commit message being the same :|
01:42:26 Join simonrvn [0] (
01:43:20jhMikeSpixelma: one to www and one to trunk
01:43:59pixelmaIMO worth altering the message a bit
01:45:02jhMikeSwith gigf, no lines...
01:47:52amiconnCan you try a cache rebuild as well?
01:48:12amiconnI'm getting "Album art is bad" for several albums on mini G2 as well
01:48:50amiconnIt *should* be all bmp album art, provided rockbox prefers that over jpg
01:49:30amiconnAnd even greylib gives those weird lines
01:50:24buckoSo I have been doing some testing on audio file formats after finding Musepack decodes efficiently on Rockbox targets.
01:51:08buckoTurns out it's also not very good −− it's encoding files 50-100% larger than Vorbis files with the same stability under my listening tests.
01:54:41bucko(On an admittedly rather small sample size.)
01:55:16jhMikeSamiconn: I have no messages about albut art, but there's not much on the 5gb anyway.
01:56:07amiconnI didn't get those on my 6GB either (which also doesn't have many albums), but both on the Color and the Mini G2
01:57:33buckoWhat's the biggest flash-based Rockbox target?
01:57:55jhMikeSgigs has no lines
01:58:51amiconnIt even happened for the same albums
01:59:57amiconnHmm, seems to be the case on old abi as well
02:01:10amiconnOkay, that's unrelated
02:02:01amiconnBut the lines are not
02:02:09amiconnThey also appear on my c200 now
02:03:14***Saving seen data "./dancer.seen"
02:04:55jhMikeSneither gigabeat has lines but e200v1 does...all are memframe lcd...
02:05:07jhMikeSh10 20GB has lines
02:05:14krazykitbucko, the various 8GB sansas with microSDHC slots would be the largest
02:05:41amiconnWell, neither color nor c200 are memframe targets, and the Mini G2 is even using the greylib
02:05:47amiconnThey all have the lines
02:06:10amiconnkrazykit: Only if not taking mods into account
02:06:29jhMikeSall with lines are arm7, while the others are arm9 and arm11...that's all I can think of
02:06:48amiconnThey're even all PP
02:07:07krazykitwell, yes. CF mods and SSD mods would take the cake
02:08:03amiconnOther than that, all 3 targets seem to work fine with eabi
02:08:10amiconnI didn't try bootloaders though
02:10:59 Join Strife89 [0] (
02:11:31jhMikeSI checked bootloaders on most of it. I didn't feel like being forced to redo the beast right now.
02:12:56pixelmakrazykit: or maybe the D2 (there is one which has 16GB built-in and they take full-size SDs) but it's not a matured port yet unlike some Sansas
02:14:29amiconnjhMikeS: 2nd Gen crashes with a stkov
02:14:51pixelmakrazykit: I mean one type of D2 model, no mods... that came out wrong
02:15:00jhMikeSI got that on 3g. tried upping the idle stack size, but it still froze.
02:15:20amiconn"*PANIC*" with the "*PAN" part overwritten, and Stkov 40004350 (1)
02:15:40 Quit FlynDice (Remote host closed the connection)
02:16:06 Join FlynDice [0] (
02:16:08jhMikeSI think that's happening when the COP tries to go into the idle loop at init
02:17:46amiconnThat address is in thread.o ibss area...
02:18:23jhMikeSthat's because it has no name, so it prints an address instead
02:18:42amiconnIt's not in the idle stack area
02:19:05jhMikeSno, it's the thread slot address, because thread->name is NULL
02:19:07amiconn.idle_stacks is at 0x40008748
02:21:05jhMikeSit initializes a thread slot for the core, then the core does a thread exit, there it remains in the idle loop until a real thread exists
02:23:40 Join funman [0] (~fun@rockbox/developer/funman)
02:27:19 Quit efyx (Remote host closed the connection)
02:31:49jhMikeSthe usage of that at startup went up for PP502x, from whatever it was to over 90% on some targets (oddly, H10 20GB only shows 68% use after init, while others show 93%), but those stacks were meant to be tiny.
02:33:28jhMikeSH10 5/6 and e200v1 both show 93% use after init
02:46:02amiconn93% on Color, Mini G2 and c200v1 as well
02:46:26 Quit simonrvn (Read error: No route to host)
02:47:03amiconn71% on 1st/2nd Gen with old abi
02:49:57amiconnWhere are the idle stacks "deadbeefed"?
02:50:30jhMikeSin crt0-pp.S, just like the main stack
02:58:32amiconnNeither doubling nor quadrupling the idle stack size helps
02:59:07jhMikeSI wonder too about where in thread_exit, it switches back to the idle stack. it's been ok so far but that could be coming to bite now since it's really not quite proper. (btw, tried that)
03:01:48jhMikeSyou might be able to get away with commenting out line 1710 just to test
03:03:30jhMikeSbtw, I stopped get the stkov when increasing it, but it just froze instead
03:03:33amiconnStill crashes
03:03:44amiconnI got the stkov always
03:03:58amiconnThe amount of overwriting varies though
03:10:15 Join simonrvn [0] (
03:15:59funmanlua overlay doesn't fit in clipv1 RAM by 10kB :/
03:19:16 Quit Horscht (Quit: Verlassend)
03:20:41 Quit togetic (Ping timeout: 245 seconds)
03:22:05 Quit GeekShadow (Quit: The cake is a lie !)
03:22:22 Join aitiba [0] (
03:22:37 Join togetic [0] (~togetic@unaffiliated/ibuffy)
03:23:16aitibaI would ike to install rockbox on my ipod but I dont know which generation can I know that?
03:25:20funmanif you run the automatic installer you can use autodetection
03:25:30 Join cjcopi [0] (
03:25:42aitibafunman, thanks
03:25:58funmanalso check
03:26:22aitibaif I have problem are there any posibilitys to keep alive my ipod?
03:27:16 Part domonoky1
03:27:49funmanyes, it's really hard to destroy ipod
03:27:53S_a_i_n_taitiba: There's no chance of damaging you're iPod installing rockbox...if that's what you mean.
03:28:49aitibaS_a_i_n_t, I dont like to damage my ipod
03:28:53aitibaI use it
03:31:16aitibaI donwload and is an exe file?
03:32:37 Join S_a_i_n_t_ [0] (S_a_i_n_t@
03:33:04funmanno it is a bz2
03:33:10 Quit steve|m (Quit: Leaving.)
03:33:12 Quit S_a_i_n_t (Ping timeout: 265 seconds)
03:33:30funmanread the page again there's a windows link
03:33:48aitibaand bz2 is a compressed file that I have to uncompress
03:33:52aitibai'm on ubuntu
03:34:46 Join steve|m [0] (
03:35:00funmanthen yes it's an exe-cutable file for linux, not a .exe for windows
03:36:18aitibafunman, do you know how can I install RockBoxUtility on linux?
03:36:53funmanjust download the url you gave an uncompress it
03:37:41aitibabut I need an emulator to can run
03:38:01aitibaon linux the way to install programs are not exe files
03:38:32aitibatwo click and taht it
03:41:09aitibaDetected an unsupported player:
03:41:09aitibaApple Ipod Nano (Third Generation)
03:41:09aitibaSorry, Rockbox doesn't run on your player.
03:42:14 Quit funman (Quit: free(random());)
03:43:32simonrvniPod 3rd Gen NO MAJOR ISSUES No Yes Yes Yes
03:43:42simonrvnTarget Current Status Notes USB Battery Charging Dual Boot OF Install with rbutil
03:44:08simonrvnaccording to it is supported
03:44:12aitibasimonrvn, the autodetect say me this
03:45:07simonrvnand on "Rockbox works nicely, but more work is needed on this port"
03:45:54simonrvni highly suggest using and reading the doc on the wiki, very useful
03:49:24simonrvnah, you didn't say ipod nano 3g (admittedly you didn't know what it was)
03:57:47 Join Cero [0] (
03:58:56CeroI have an Eclipse CL2BLK. Is it compatible for that one?
04:00:21CeroI guess i'm asking a dumb question. thanks anyways
04:00:55 Quit Cero (Client Quit)
04:02:12 Join BHSPitLappy [0] (~BHSPitLap@unaffiliated/bhspitmonkey)
04:03:15***Saving seen data "./dancer.seen"
04:11:44 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
04:30:43 Join BradC [0] (
04:45:25 Quit amiconn (Disconnected by services)
04:45:27 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:45:47 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:46:43 Quit pixelma (Disconnected by services)
04:46:44 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:47:00 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:55:10DEBUGEOF from server (Connection reset by peer) (snapshot: netstuff.c line 545)
04:55:10***Saving seen data "./dancer.seen"
04:55:12***Started Dancer V4.16
04:55:12***Connected to on port 6667
04:55:12***Logfile for #rockbox started
04:55:12Mode"logbot :+i" by logbot
04:55:13CtcpVersion from frigg!~frigg@freenode/utility-bot/frigg
04:55:14***Server message 501: 'logbot :Unknown MODE flag'
04:55:14 Join logbot [0] (
04:55:14 Join pixelma [0] (quassel@rockbox/staff/pixelma)
04:55:14 Join amiconn [0] (quassel@rockbox/developer/amiconn)
04:55:14 Join BradC [0] (
04:55:14 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
04:55:14 Join BHSPitLappy [0] (~BHSPitLap@unaffiliated/bhspitmonkey)
04:55:14 Join steve|m [0] (
04:55:14 Join S_a_i_n_t_ [0] (S_a_i_n_t@
04:55:14 Join cjcopi [0] (
04:55:14 Join togetic [0] (~togetic@unaffiliated/ibuffy)
04:55:14 Join aitiba [0] (
04:55:14 Join simonrvn [0] (
04:55:14 Join FlynDice [0] (
04:55:14 Join Strife89 [0] (
04:55:14 Join ranma [0] (
04:55:14 Join chrisb [0] (
04:55:14 Join pjm0616 [0] (~user@
04:55:14 Join perrikwp [0] (
04:55:14 Join obo_ [0] (~obo@rockbox/developer/obo)
04:55:14 Join Topy44 [0] (
04:55:14 Join Zarggg [0] (
04:55:14 Join lev_ [0] (~lev@
04:55:14 Join grndslm [0] (
04:55:14 Join elinenbe [0] (
04:55:14 Join Battousai [0] (~bryan@gentoo/developer/battousai)
04:55:14 Join sinthetek [0] (~sinthetek@unaffiliated/sinthetek)
04:55:14 Join bucko [0] (~bucko@
04:55:14 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:55:14 Join saratoga [0] (9803c6dd@gateway/web/freenode/ip.
04:55:14 Join scorche [0] (~scorche@rockbox/administrator/scorche)
04:55:14 Join panni_ [0] (
04:55:14 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
04:55:14 Join simabeis [0] (
04:55:14 Join soap [0] (~soap@rockbox/staff/soap)
04:55:14 Join Guest43341 [0] (
04:55:14 Join Kitar|st [0] (
04:55:14 Join detaos [0] (
04:55:14 Join Rob2223 [0] (
04:55:14 Join wombat23 [0] (
04:55:14 Join joeyg [0] (
04:55:14 Join shai [0] (
04:55:14 Join storm` [0] (
04:55:14 Join CIA-6 [0] (cia@
04:55:14 Join bieber [0] (
04:55:14 Join nk_ [0] (
04:55:14 Join xavieran [0] (
04:55:14 Join AlexP [0] (~ap@rockbox/staff/AlexP)
04:55:14 Join jae [0] (~jae@
04:55:14 Join bzed [0] (
04:55:14 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
04:55:14 Join elcan [0] (
04:55:14 Join jfc [0] (
04:55:14 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
04:55:14 Join r0b- [0] (
04:55:14 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
04:55:14 Join krazykit [0] (
04:55:14 Join Stummi [0] (
04:55:14 Join FOAD [0] (~dok@
04:55:14 Join blithe [0] (~blithe@
04:55:14 Join Farthen [0] (
04:55:14 Join Torne [0] (torne@rockbox/developer/Torne)
04:55:14 Join ocin [0] (
04:55:14 Join rasher [0] (~rasher@rockbox/developer/rasher)
04:55:14 Join Bagder [0] (~daniel@rockbox/developer/bagder)
04:55:14 Join mikroflops [0] (
04:55:14 Join leavittx [0] (~leavittx@
04:55:14 Join MagusG [0] (
04:55:14 Join sharp [0] (
04:55:14 Join rvvs89 [0] (ivo@pdpc/supporter/base/rvvs89)
04:55:14 Join maraz [0] (
04:55:14 Join scorche|sh [0] (
04:55:14 Join radagast [0] (
04:55:14 Join Tuplis [0] (
04:55:14 Join Beta2K [0] (
04:55:14 Join Zambezi [0] (Zulu@unaffiliated/zambezi)
04:55:14 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
04:55:14 Join guymann [0] (
04:55:14 Join Xerion [0] (
04:55:14 Join gibbon_ [0] (
04:55:14 Join n17ikh [0] (
04:55:14 Join advcomp2019_ [0] (~advcomp20@unaffiliated/advcomp2019)
04:55:14 Join jordan` [0] (
04:55:14 Join GodEater [0] (~bibble@rockbox/staff/GodEater)
04:55:14 Join avacore [0] (
04:55:14 Join Galois [0] (
04:55:14 Join whydoubt [0] (
04:55:14 Join sbhsu [0] (
04:55:14 Join gevaerts [0] (~fg@rockbox/developer/gevaerts)
04:55:14 Join BeFalou [0] (~mamutoi@unaffiliated/befalou)
04:55:14 Join yosafbridge [0] (
04:55:14 Join topik [0] (
04:55:14 Join @ChanServ [0] (ChanServ@services.)
04:55:14 Join fxb__ [0] (
04:55:14 Join Hadaka [0] (
04:55:14 Join sevard [0] (sev@unaffiliated/sevard)
04:55:14 Join Kohlrabi [0] (
04:55:14 Join ved [0] (
04:55:14 Join piroko [0] (
04:55:14 Join linuxguy3 [0] (~timj@
04:55:14 Join crwl [0] (
04:55:14 Join Kamyk [0] (
04:55:14 Join incant [0] (
04:55:14 Join MuscleNerd [0] (
04:55:14 Join kadoban [0] (
04:55:14 Join preglow [0] (
04:55:14 Join tmzt [0] (
04:55:14 Join keanu [0] (~keanu@unaffiliated/keanu)
04:55:14 Join aevin [0] (eivindsy@unaffiliated/aevin)
04:55:14 Join BlakeJohnson86 [0] (~bjohnson@2002:1876:a27b:0:227:13ff:fe65:1262)
04:55:14 Join YPSY [0] (
04:55:14 Join ps-auxw [0] (~arneb@2001:470:c807:0:1532:4e5f:2ad3:4123)
04:55:14 Join ehntoo_ [0] (
04:55:14 Join Dhraakellian [0] (
04:55:14 Join dionoea [0] (~dionoea@videolan/developer/dionoea)
04:55:14 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
04:55:14 Join mk429 [0] (
04:55:14 Join solrize [0] (
04:55:14 Join ThomasAH [0] (
04:55:14 Join lostlogic [0] (~lostlogic@rockbox/developer/lostlogic)
04:55:14 Join Utchybann [0] (
04:55:14 Join literal [0] (hinrik@
04:55:14 Join Slasheri [0] (miipekk@rockbox/developer/Slasheri)
04:55:14 Join kisak [0] (
04:55:22 Join Barahir [0] (
05:06:44 Quit radagast (Ping timeout: 240 seconds)
05:16:17 Join Nubsauce [0] (
05:16:55 Join radagast [0] (
05:20:22 Quit jae (Ping timeout: 258 seconds)
05:22:24Nubsauceanyone around I could ask some questions about rockbox and nano 2g wsod?
05:22:47 Join jae [0] (
05:23:04 Quit aitiba (Remote host closed the connection)
05:24:43 Join kaylinsigswort [0] (
05:25:55Nubsauceso I have a nano 2g, put the bootloader on it and rockbox
05:26:07Nubsaucereboot and it seems to load a different os
05:26:33Nubsaucehow can I tell? well the screen is a few shades darker and there is no clicking with the scroll wheel
05:27:04Nubsaucealso the hold button up top turns the white screen off then back on when I toggle it
05:27:07simonrvndid you see a rockbox logo?
05:27:43NubsauceI load what would be the ipod os and the scroll wheel clicks and the screen is shock white
05:27:55Nubsauceand the hold button doesn't do anythign except lock the controls
05:28:00Nubsaucei can't see anything but whilte
05:28:08Nubsauceis there a hardware white screen of death
05:28:17Nubsaucei really thought it was a firmware/software thing
05:29:19Nubsaucewhen i reformat it with the itunes software it goes back to it's original white screen (no text ever) mode
05:29:34Nubsaucenever see any logos
05:29:39Nubsaucerockbox or apple
05:30:38NubsauceI've searched all over and can't find any seemingly worthwhile information on a white screen issue being physical, have you ever heard of that?
05:30:46 Quit steve|m (*.net *.split)
05:30:46 Quit pjm0616 (*.net *.split)
05:30:46 Quit ocin (*.net *.split)
05:31:35tmztthe only thing you ever see if white or off? the shades are backlight differences between OF and rockbox?
05:31:52NubsauceI've searched your forums, ipodlinux, all over the place and nothing on a hardware wsod for nano 2g
05:32:20Nubsauceyeah, there's a shading difference between the 2 firmwares
05:32:37Nubsauceand the scrollwheel and hold button interact differently
05:33:05Nubsaucebut always a white screen, just 2 different shades, 1 for OF and a slightly darker shade for rockbox
05:35:35NubsauceI memorized how to get thru the menu on the original firmware, so I can actually play
05:35:47 Join blackromeo [0] (
05:35:59NubsauceIt's my daughters tho and the last thing I want to do is teach her how to dig down to the songs by memory
05:37:30blackromeosup again peeps
05:37:48NubsauceIs it possible it's a hardware issue, like a loose cable to the screen?
05:37:57Nubsaucesup black?
05:37:59blackromeoive finally managed to rox my box :)))
05:38:19blackromeojust one thing
05:38:25tmztany tft will fade to white if the initialization fails
05:38:36tmztbut usually through colors first
05:38:44blackromeonow that my fuze v2 is rocked boxed
05:39:04blackromeohow do i change to the orignal screen?
05:39:14 Join steve|m [0] (
05:39:14 Join pjm0616 [0] (~user@
05:39:14 Join ocin [0] (
05:39:29blackromeoi heard you can dual boot?
05:39:35blackromeohow do i do this?
05:39:39NubsauceI know how to do a nano, you hit the hold button while the apple logo is on
05:39:45Nubsauceno idea on that one
05:40:23tmztI believe it's hold the bottom part of the wheel, there are documentats on the rockbox site (the manual) that should describe this for supported targets
05:40:35tmztI have fuzev1, not v2
05:42:17blackromeoah ok
05:42:27blackromeoi shall have a look at the manual and see
05:43:05NubsauceIs there a way without looking at the screen I can verify this is actually rockbox loading?
05:43:15Nubsaucedoes it act like I've said by default?
05:43:27NubsauceI mean the lock button turning off and on the screen
05:43:34Nubsauceand no sound on the scroll wheel?
05:44:02tmztdoes the screen go directly to white?
05:44:21tmztit could be off with only the backlight on
05:44:27tmztin which case hardware is most likely
05:44:36tmztit seems both softwares are booting completely
05:44:43Nubsaucewhen it loads the original firmware, it stays a sharp white...when it seems to load rockbox it goes shock white, then drops a shade
05:44:58tmztthose are backlight differences
05:45:02Nubsaucethanks you
05:46:01Nubsaucethanks much, I'll try to find a replacement screen
05:47:37NubsauceYou know how hard it was to try to load ipodlinux manually without seeing the screen ever?
05:47:56Nubsauceit was horrible and I don't think I ever got it complete
05:50:51 Join CGL [0] (~CGL@
05:51:57 Quit blackromeo (Quit: CGI:IRC (Ping timeout))
05:52:00krazykitor you could make it use voices and generate talk clips so you don't need the screen :)
05:52:17Nubsaucethat would kick ass
05:52:41NubsauceI'll look into that
05:53:41krazykitrockbox utility should take care of everything for you, except possibly enabling voice, but that's an easy fix in the config file, i would think
05:54:28Nubsauceso does it have files preconfigured that will say the name of the current menu item?
05:54:47simonrvnit does by default
05:54:58Nubsaucethat might work well
05:55:08krazykityes, that's the voice file. the talk clips are generated for your music content
05:55:11simonrvnat least when i ran the svn "fresh install", and no config
06:04:21 Join puetzk [0] (
06:04:48puetzkmc2739: Thanks for replying, that's the impression I'd gotten
06:08:04NubsauceIt seemed to install, but when I started it back up with RB, it doesn't seem to do anything when i move the scroll wheel
06:08:12Nubsauceit's not announcing any menus or anything
06:09:11 Quit t0rc (Quit: Leaving)
06:09:23Nubsaucethese get loaded in the partition that has rb, so i can just visually verify they're there by browsing that drive letter?
06:09:28Nubsaucecheckin that now
06:10:05*Dhraakellian sees that the Sudoku scratchpad now gets saved
06:10:23*Dhraakellian also sees that (old) themes got broken
06:11:26Dhraakellianare the themes on converted, or should I check out r26640 and build it locally so I can continue using old themes?
06:12:01Dhraakellianor grab the conversion tool and play around with it?
06:14:27puetzkmc2739: If you're knowledgeable about the as3525v2 recording, there's one other behavior I noticed that seems really odd. If I record at 48kHz, it looks like every sample is doubled (as if it is just doubling from 24kHz).
06:14:44puetzkThis would be entirely reasonable behavior :-)
06:15:01puetzkHowever, the signal is audibly better at 48kHz; an actual 24kHz recording has artifacts that sound kinda like aliasing from higher frequencies,that go away in the 48kHz-but-not-really recording
06:16:45 Nick CGL is now known as [CGL]_buscando_n (~CGL@
06:17:11 Nick [CGL]_buscando_n is now known as [CGL]_WTF_ (~CGL@
06:22:20NubsauceThe only part I need to do is the "install voice file" and it should read menu's correct?
06:22:55NubsauceI see the english.voice file under langs
06:23:11Nubsauceso I know that part is done, but it doesn't seem to work
06:25:34 Quit joeyg (Quit: lions and tigers and bears, oh my!)
06:27:47Nubsaucewhat comes up on the first screen for RB?
06:28:03Nubsauceis there osmething i need to hit once it powers on that isn't announce by the voice?
06:28:52Nubsaucenevermind, got it
06:29:14Nubsauceyou guys rock
06:29:35Nubsaucecan the apple firmware even announce the menus like that?
06:34:44 Quit shai (Quit: Leaving)
06:38:15 Join flan_suse [0] (~suse@unaffiliated/flansuse)
06:38:18flan_suseAnyone in the building?
06:40:07rvvs89Nubsauce: If the advertising is to be believed, the latest generation of nano can.
06:40:26simonrvnconversion tool? why didn't i think of that...
06:42:01 Quit elcan (Ping timeout: 264 seconds)
06:42:34 Join elcan [0] (
06:45:17flan_suseI installed Rockbox, lastest current build (r26809-100611) on my Sansa Clip+. Insallation went smoothly, added fonts and themes, everything's good. But I noticed that when I plug it into my computer, nothing is detected. dmesg | tail shows that no device was plugged into the USB port. The screen on the Sansa Clip switches between a picture of a USB cable and the Rockbox logo.
06:45:43flan_suseIn other words, there's no way to add music or access the Sansa Clip+ when it's plugged into the computer.
06:47:38 Join antil33t [0] (
06:49:36puetzkflan_suse: did you redo the bootloader with the 3.6 mkamsboot?
06:49:50puetzkI know I had to do that to get the Clip+ to reboot into the OF on USB
06:49:52flan_susepuetzk, first time, so I don't think so.
06:50:10puetzkfirst time should work too
06:50:19puetzkas long as you did use the one from June 6
06:50:36puetzkif you just upgrade rockbox and not the bootloader, then that doesn't work
06:50:45flan_suseI am using: r26809-100611
06:50:46 Quit BHSPitLappy (Quit: Ex-Chat)
06:50:55flan_suseI installed the bootloader first, then rockbox.
06:51:14flan_suseBut it boots straight into Rockbox, and plugging the device into a USB port is not detected by the computer.
06:51:48puetzkMine works :-)
06:51:58puetzkYou can also get into the OF by holding down the Home button as it boots
06:52:05puetzkbut you shouldn't have to
06:52:28flan_susepuetzk, I'm supposed to see a menu selection for which firmware I want to boot into?
06:52:45puetzkbut if you hold down the Home button as you hit power, it will go to the OF instead of rockbox
06:52:54flan_susepuetzk, okay, it did.
06:53:12flan_susepuetzk, but shouldn't Rockbox work as MSC mode and the Clip+ should be detected on the computer?
06:53:32puetzkRockbox doesn't do MSC mode on the AMSv2 sansas (yet)
06:53:37flan_susepuetzk, it's not simply not mounting it, not even the kernel detects a USB device is plugged in. dmesg | tail reveals nothing.
06:53:48puetzkthough it's supposed to reboot into the OF when you plug the cable in
06:53:54puetzkwhich yours doesn't do for some reason
06:54:00flan_susepuetzk, oh.
06:54:42flan_susepuetzk, if the device is off and I plug it in, it will power on, show the Rockbox logo, then switch to a USB cable picture, then to the logo, back and forth. It looks weird when it does that.
06:54:59puetzkYeah, it's supposed to go to the USB cable picture, then reboot to the OF
06:55:15***Saving seen data "./dancer.seen"
06:55:25flan_susepuetzk, it's not though. But it does for your Clip+?
06:55:46puetzkYes. I'm running 3.6, I wonder if people have started turning on other bits of the USB driver-in-progress in the current build
06:56:10flan_susepuetzk, I might have an idea, one second.
06:56:22mc2739there is no 3.6 for clip+
06:56:44flan_susepuetzk, nope, my idea didn't work.
06:56:50puetzkmc2739: oh, right
06:56:51flan_susemc2739, you're sure? I just installed it.
06:57:00puetzkwell, I'm running SVN from about that era :-)
06:57:16puetzkforgot that it didn't get blessed
06:57:21mc2739flan_suse: yes, 3.6 is only for stable target
06:57:25flan_susemc2739, oh.
06:57:45flan_susemc2739, well, it's the current build: r26809-100611
06:57:46 Join setfilepath [0] (
06:58:21flan_susepuetzk, what is your build? I might have to grab that version instead.
06:58:35mc2739flan_suse: if you plug in your cable when the clip+ is offm it should boot to the OF for file transfers, or boot whille pressing the hold button and then attach the cable
06:58:36*puetzk mostly uses the uSD slot to move music onto it, since the OF takes several minutes of "Refreshing your media" to boot
06:58:58flan_susemc2739, I tried when it's off (and on) and the same thing happens as described above.
06:59:28flan_susemc2739, it works when I boot into OF and then plug it in.
06:59:43flan_susemc2739, I can technically "live" with that, but it's kind of awkward.
07:00:09flan_susepuetzk, so I guess yours was built on June 1st.
07:00:41puetzkyeah, older than I thought it was
07:01:06mc2739then you need to patch your firmware with the latest version of mkamsboot (from June 6)
07:01:07setfilepathSettings > General Settings > File View > Show Path > "Off" is this setting supposed to remove the path from all screens ?, if so, v3.6 doesn't do this.
07:02:22flan_susemc2739, I use the Rockbox utility to do this?
07:03:24mc2739flan_suse: the release version of rockbox utility does not have the latest mkamsboot
07:04:09flan_susemc2739, I just dump the .bin file into .rockbox and then unplug?
07:04:33mc2739flan_suse: ypu can get a development version of rockbox utility here:
07:05:37mc2739flan_suse: no, the patched bin file (with the correct name for your device) needs to go in the root directory
07:05:48flan_susemc2739, got it. Going to make it now.
07:10:32 Quit puetzk (Quit: puetzk)
07:15:06flan_susemc2739, it worked. I used mkamsboot to patch SanDisk's OF, copied the patched .bin into the root folder of the device (renamed it to clppa.bin), unplugged, and viola.
07:15:17flan_suseWhen I plug my Sansa Clip+ into the computer, it reboots into the OF.
07:15:39flan_suseEven if Rockbox was currently on.
07:17:12flan_suseThanks for all the help. You guys... rock.
07:17:15 Quit panni_ (Quit: ( :: NoNameScript 3.81 :: ))
07:17:20flan_suseOkay, that was a pathetic pun, but what else could I say?
07:18:51 Join Reaper_man [0] (
07:18:56Reaper_manhey quick question
07:19:18Reaper_manis there any way to remove the apple logo on bootup (5th gen 30GB iPod)
07:26:25CIA-6New commit by pamaury (r26810): usb-as3535v2: add more defines, reorganize/simplify things, add code for interrupt handling
07:27:45 Join pamaury [0] (
07:27:45 Quit pamaury (Changing host)
07:27:45 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
07:28:06CIA-6r26810 build result: All green
07:28:28 Quit flan_suse (Quit: Leaving)
07:45:12 Quit pamaury (Remote host closed the connection)
07:49:53 Quit elcan (Ping timeout: 258 seconds)
07:53:31 Join elcan [0] (user36@
08:10:05S_a_i_n_t_Reaper_man: Nope.
08:10:38S_a_i_n_t_is it *that* annoying?
08:10:50Reaper_manI hate Apple, so yeah
08:10:59Reaper_manat least I can sand off the back
08:11:20S_a_i_n_t_I can't remember if iPodWizzard supports that model iPod or not, but you should be able to *alter* the image, but you can't remove it, no.
08:11:32Reaper_manAlter!? :D
08:11:37Reaper_manthat works too!
08:12:38S_a_i_n_t_Google iPod Wizzard", I can't remember if it supports your model though. And be careful, some assholes decided to flood the net with versions of it loaded with spyware.
08:13:08S_a_i_n_t_If you have even halfway decent antivirus, or get it from a decent source it'll all be sweet though.
08:14:27pixelmayou know that's a *bit* off-topic here though ;)
08:15:02S_a_i_n_t_yeah, IRC wondows were switched around. I thought it was -community :/
08:15:08*S_a_i_n_t_ apologises.
08:18:14CIA-6New commit by kkurbjun (r26811): M:Robe 500 - Build with EABI.
08:19:35 Part Reaper_man ("Leaving")
08:19:41CIA-6r26811 build result: All green
08:21:24 Quit lev_ (Read error: Connection reset by peer)
08:24:23 Quit kaylinsigswort (Quit: kaylinsigswort)
08:27:24 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
08:28:15 Quit perrikwp (Quit: CGI:IRC (Ping timeout))
08:40:54Nubsaucelater, thanks for the info all
08:40:58 Quit Nubsauce (Quit: CGI:IRC)
08:55:16***Saving seen data "./dancer.seen"
09:01:41 Join wodz [0] (
09:01:47wodzpixelma: ping
09:05:21 Join flydutch [0] (
09:13:57 Join elinenbe_ [0] (
09:13:57 Quit elinenbe (Read error: Connection reset by peer)
09:13:58 Nick elinenbe_ is now known as elinenbe (
09:21:27 Join perrikwp [0] (
09:21:40 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
09:29:47 Quit S_a_i_n_t_ (Ping timeout: 248 seconds)
09:30:27 Join S_a_i_n_t [0] (S_a_i_n_t@
09:31:46 Join shai [0] (
09:32:44CIA-6New commit by bertrik (r26812): Spanish language update, FS #11394 by Adrián Cereto Massagué
09:34:18CIA-6r26812 build result: All green
09:35:39 Quit wodz (Quit: Leaving)
09:38:20CIA-6New commit by bertrik (r26813): Catalan language update (added "Pictureflow" missing string), FS #11393 by Adrián Cereto Massagué
09:40:10CIA-6r26813 build result: All green
09:53:46 Join Rob2222 [0] (
09:55:17 Join Casainho [0] (
09:57:15 Quit Rob2223 (Ping timeout: 240 seconds)
10:07:44 Quit pjm0616 (Quit: pjm0616)
10:09:05 Join stripwax [0] (
10:18:59 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
10:21:02 Join liar [0] (
10:21:36 Join pjm0616 [0] (~user@
10:21:49JdGordonbieber: tanks for moving the VIEWPORT tag, wont be able to play with it tonight though but anyway :)
10:34:06JdGordonoh new CIA-6 info :) nice
10:38:45 Join Kitr88 [0] (
10:40:13 Join CFP [0] (
10:40:43 Quit CFP (Client Quit)
10:40:49setfilepathSettings > General Settings > File View > Show Path > "Off" is this setting supposed to remove the path from all screens ?, if so, v3.6 doesn't do this
10:41:36 Join dfkt [0] (dfkt@unaffiliated/dfkt)
10:42:12setfilepathJdGordon ^
10:42:38 Quit Kitar|st (Ping timeout: 264 seconds)
10:43:15 Quit Kitr88 (Ping timeout: 240 seconds)
10:44:35 Quit [CGL]_WTF_ (Remote host closed the connection)
10:46:02 Quit setfilepath (Quit: CGI:IRC)
10:47:28bertrikI think I'll do a battery benchmark of my clip v1 now that we use eabi and thumb
10:48:13 Join Kitar|st [0] (
10:50:25literalwhat is "thumb", by the way?
10:50:34 Join DataGhost [0] (
10:50:34 Quit DataGhost (Changing host)
10:50:34 Join DataGhost [0] (~dataghost@unaffiliated/dataghost)
10:51:19bertrikit is a separate CPU instruction set in ARM processors where instructions are 16-bit wide instead of 32-bit
10:54:53bertrikBasically this makes code take up less space, thumb was enabled on some ams sansa players because they have a relatively small DRAM. For more info on thumb, see
10:55:19***Saving seen data "./dancer.seen"
10:56:34 Join bimbel [0] (~Miranda@unaffiliated/bmbl)
10:56:58 Quit bmbl (Ping timeout: 258 seconds)
10:57:14 Quit Casainho (Ping timeout: 276 seconds)
10:57:29 Quit FlynDice (Remote host closed the connection)
11:00:13literalbut you think it will affect performance, too?
11:05:18bertrikI don't know, maybe funman knows, I guess I'll find out. Most codecs probably need less than 20% of the full processor power.
11:20:47 Quit bieber (Ping timeout: 264 seconds)
11:21:02 Join bieber [0] (
11:27:56 Join DerPapst [0] (
11:35:57CIA-6New commit by amiconn (r26814): Build iPod Color, iPod Mini 2nd Gen and Sansa c200 with EABI.
11:36:00CIA-6New commit by amiconn (r26815): Buildserver: Build iPod Color, iPod Mini 2nd Gen and Sansa c200 with EABI. Also make sure m:robe500 is built on a proper client.
11:37:39CIA-6r26814 build result: All green
11:38:01 Join funman [0] (~fun@rockbox/developer/funman)
11:39:24funmanbertrik: i had tried battery life on my clipv1 but there was no difference when built with thumb
11:39:58bertrikcool, but I'll bench it anyway :)
11:44:02amiconnfunman: Do you have an arm7tdmi target?
11:44:09amiconn(PP or other)
11:45:53 Quit stripwax (Quit:
11:47:45*amiconn wonders what might cause the weird lines in PF with EABI
11:48:47funmanperhaps i could make a build with mcpu=arm7tdmi and run it on arm9 though
11:50:23 Quit S_a_i_n_t (Read error: Connection reset by peer)
11:52:53amiconnI suspect signedness problems in draw_gradient()
11:53:08amiconnBut if this is the case, why doesn't it happen on arm9 and arm11?
11:53:27amiconn(and neither on CF, or when building arm7 with old abi)
11:54:13 Join S_a_i_n_t [0] (S_a_i_n_t@
11:54:51amiconnOther than that, EABI seems to work quite well so far, except on PP5002
11:56:01 Quit S_a_i_n_t (Read error: Connection reset by peer)
11:57:50 Join S_a_i_n_t [0] (S_a_i_n_t@
12:08:00funmancan you mix eabi files and non-eabi? for example copy libfirmware.a from a non-eabi build to see where the bugs are
12:11:48 Quit jhMikeS ()
12:15:12funmanS_a_i_n_t: TheSeven: you said nano2g works fine with eabi ?
12:15:42TheSevenit did some months ago at least
12:15:54TheSevenI'm using an (outdated) eabi build right now
12:28:00 Quit S_a_i_n_t (Read error: Connection reset by peer)
12:28:51*amiconn did bootloader tests before r26814
12:32:02 Join S_a_i_n_t [0] (S_a_i_n_t@
12:32:08 Join stripwax [0] (
12:32:17 Join robin0800 [0] (
12:33:46 Quit robin0800 (Remote host closed the connection)
12:34:07 Join robin0800 [0] (
12:39:44funmanbertrik: IIUC, the meizu m6sp and m6sl are advertized as the same player (M6) ?
12:40:26funmanhm well the SP has one more button
12:40:32 Quit krazykit (Quit: drivin)
12:40:47funmanthen my suggestion of using one build for both (and support 4 different LCD types) doesn't hold :/
12:43:31 Join kugel [0] (~kugel@rockbox/developer/kugel)
12:43:37 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
12:47:56 Part Llorean
12:53:22funmangevaerts: do you know the lcd controller ID value for the other m6sl LCD type? is it 0x139 ?
12:53:29 Quit robin0800 (Remote host closed the connection)
12:55:22***Saving seen data "./dancer.seen"
12:55:52funmanwhat i read here lets me think that the m6sl supports 3 different LCD
13:01:41funmanbertrik: perhaps removing the "- 1" from LCDTCON3 setting can fix the shifted display on m6sp
13:03:52bertrikfunman, it's not always shifted for me, usually the first time it looks OK, shifted by a line later
13:04:16funmanit behaved the same here (when i had rockbox code on it)
13:05:19funmanhm well, -1 seems right in fact
13:05:48bertrikI did find some other "bugs" in the s5l8700 datasheet, so this could be one too
13:06:26bertrikThis has been too long ago to be still really familiar with the LCD registers
13:07:11bertrikIn hindsight, I should probably have documented the datasheet errors somewhere
13:07:18 Join ender` [0] (
13:14:28 Join roolku [0] (
13:16:24CIA-6New commit by roolku (r26816): Build mrobe100 with EABI
13:16:36 Quit ender` (Ping timeout: 240 seconds)
13:16:59CIA-6New commit by roolku (r26817): Build mrobe100 with EABI (buildserver)
13:17:20 Quit stripwax (Quit:
13:17:55CIA-6r26816 build result: All green
13:19:35 Join slck [0] (Venci@Slackware.SlackPix.Com)
13:19:49 Join petur [0] (~petur@rockbox/developer/petur)
13:19:57 Join ender` [0] (
13:28:22 Join halmi [0] (
13:38:12 Join stripwax [0] (
13:38:33 Quit shai (Ping timeout: 245 seconds)
13:44:41 Quit funman (Quit: free(random());)
13:48:36 Quit stripwax (Quit:
13:52:16 Join n1s [0] (~n1s@rockbox/developer/n1s)
13:52:45 Quit flydutch (Quit: /* empty */)
13:53:06 Quit roolku ()
13:56:26 Quit cjcopi (Read error: Operation timed out)
13:57:33 Quit DataGhost ()
13:59:58 Join stoffel [0] (
14:00:25 Quit stoffel (Remote host closed the connection)
14:02:17 Join robin0800 [0] (
14:03:43 Join stoffel [0] (
14:07:49 Join cjcopi [0] (
14:08:53 Quit robin0800 (Remote host closed the connection)
14:14:00 Join bluebroth3r [0] (
14:14:01 Quit bluebroth3r (Changing host)
14:14:01 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
14:14:01 Quit bluebrother (Disconnected by services)
14:25:54 Join efyx [0] (
14:43:27 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
14:47:43 Quit kugel (Ping timeout: 245 seconds)
14:51:48 Join panni_ [0] (
14:55:25***Saving seen data "./dancer.seen"
15:06:28 Join stripwax [0] (
15:08:14 Join shai [0] (
15:12:51 Join M3DLG [0] (~M3DLG@
15:20:42 Join kugel [0] (~kugel@rockbox/developer/kugel)
15:23:53bertrikThe sansa clip+ firmware seems to have special handling for FM frequencies around 96 MHz, I guess it's receiving a harmonic of one of its own internal frequencies, (like 24MHz * 4)
15:24:32gevaertsIIRC the archoses (in rockbox) do something there as well
15:25:15 Quit chrisb (Remote host closed the connection)
15:29:47kugelre: r26813, do we really want "PictureFlow" translated?
15:30:15kugelthat's the name of the plugin after all, and it doesn't show up translated in the filebrowser
15:31:49 Quit bzed (Quit: reboot time :))
15:37:19 Join bzed [0] (
15:37:23 Quit M3DLG (Ping timeout: 276 seconds)
15:43:12 Join hebz0rl [0] (
15:47:44 Nick bimbel is now known as bmbl (~Miranda@unaffiliated/bmbl)
15:47:50 Quit bzed (Remote host closed the connection)
15:48:26 Join bzed [0] (
15:51:36 Quit ender` (Quit: I do not fear death. I had been dead for billions and billions of years before I was born and had not suffered the slightest inconvenience from it. -- Mark Twain)
15:53:32 Join funman [0] (~fun@rockbox/developer/funman)
15:54:18 Quit S_a_i_n_t (Ping timeout: 276 seconds)
15:55:27 Join FlynDice [0] (~FlynDice@
15:56:34 Join S_a_i_n_t [0] (S_a_i_n_t@
15:58:51TheSevenwhat's the general opinion on committing the hardware keyclick patch? (after fixing it and cleaning it up)
15:58:58TheSevendo we want that feature at all?
15:59:18 Quit guymann (Ping timeout: 265 seconds)
15:59:26 Join kugel_ [0] (
15:59:41TheSevenpersonally, I like it, and I'm getting tired of needing to un-rot it over and over
15:59:46 Quit kugel (Disconnected by services)
15:59:50 Nick kugel_ is now known as kugel (
15:59:54 Quit kugel (Changing host)
15:59:54 Join kugel [0] (~kugel@rockbox/developer/kugel)
16:00:26TheSevenok, so what needs to be done to get it committed?
16:00:40TheSevenof course, someone with an old ipod needs to figure out where it did break
16:00:49funmanis it only external, or can you hear it in the phones too?
16:00:54TheSeventhere was also some suggestion regarding the menu options
16:01:02TheSevenonly external
16:01:36funmani don't know the patch but yeah it would be nice
16:01:57TheSeven(the patch adds an option to enable piezo clicking, however you can still (independently) enable the software clicking if you like it)
16:02:25TheSevenfunman: if you want to look at it, it's FS #5111
16:02:27funmani was thinking about different options for click, like "headphones", "speaker"
16:02:31TheSevenI'm about to resync it
16:03:04funmanperhaps ask Torne since he knows quite well ipods
16:03:39kugelTheSeven: is my first comment addressed?
16:04:51TheSevenkugel: I don't really get your point
16:05:03 Quit bzed (Remote host closed the connection)
16:05:12TheSevenfrom what I can tell, the keyclick was moved to an action_do_keyclick function which causes both clicks
16:07:08TheSevenAlso, I would like to get rid of the extra thread for it on the old ipods, if that is possible
16:07:15TheSevenon the nano2g it doesn't need a thread anyway
16:08:43TheSevenkugel: are you referring to the "#ifdef HAVE_HARDWARE_CLICK"s in apps/gui/list.c?
16:08:49TheSeventhose seem to be spurious...
16:09:31TheSeventhey cause clicks (both software and hardware) only if a device _supports_ hardware keyclick, no matter if it's enabled
16:09:41TheSeventhis doesn't make much sense to me
16:10:22kugelsome of the changes look strange to me
16:10:36 Quit S_a_i_n_t (Ping timeout: 260 seconds)
16:10:50TheSevencan you point me to one?
16:11:14TheSeven(not that I would know anything about the GUI code, I'm looking at these things from the driver side)
16:12:26 Join S_a_i_n_t [0] (S_a_i_n_t@
16:12:36kugellist.c for example
16:13:17kugelit only clicks when wrapping the lists?
16:13:44kugelthat'd make it useless for me
16:17:57 Join guymann [0] (~charlie@
16:21:42 Quit funman (Quit: free(random());)
16:25:21 Join halmi_ [0] (
16:25:55TheSevenkugel: no, it doesn't
16:26:20 Join S_a_i_n_t_ [0] (S_a_i_n_t@
16:26:27TheSevenit's in every branch of the if
16:28:29 Quit S_a_i_n_t (Ping timeout: 276 seconds)
16:28:29 Quit halmi (Ping timeout: 276 seconds)
16:30:27kugelTheSeven: so for each item?
16:30:35TheSevenapparently yes
16:30:44kugelthat would be nice
16:31:10 Join bzed [0] (
16:35:16domonokyui, das ist ja nen lustiges gehäuse.. schön platzsparend.
16:35:25domonokyups, wrong channel
16:41:45 Join kramer3d [0] (~kramer@unaffiliated/kramer3d)
16:43:18 Join Buschel [0] (
16:45:11TheSevennever touch a running system
16:45:49TheSeveni just upgraded my nano2g to the newest version, and it looks like there are some new bugs
16:48:45 Quit sevard (Ping timeout: 248 seconds)
16:49:37 Join sevard [0] (sev@
16:54:45 Quit S_a_i_n_t_ (Ping timeout: 276 seconds)
16:54:59 Join S_a_i_n_t [0] (S_a_i_n_t@
16:55:05 Join CGL [0] (~CGL@
16:55:26***Saving seen data "./dancer.seen"
17:07:58 Join robin0800 [0] (
17:08:48 Join Buschel_ [0] (
17:10:29 Quit Buschel ()
17:10:35 Nick Buschel_ is now known as Buschel (
17:14:32 Quit Buschel (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
17:17:26 Join mozetti [0] (
17:18:33 Quit mozetti (Quit: -a-)
17:19:18 Join AndIrc__ [0] (
17:19:38 Part AndIrc__
17:22:14 Quit TheSeven (Disconnected by services)
17:22:16 Join The_Seven [0] (~TheSeven@rockbox/developer/TheSeven)
17:31:29bertrikRockbox utility installs tested bootloaders, even for something unstable like clip+, right?
17:32:21 Quit FlynDice (Quit: Gotta go fly!)
17:32:28 Join FlynDice [0] (~FlynDice@
17:35:55 Join TheSphinX^ [0] (
17:37:34 Quit FlynDice (Ping timeout: 258 seconds)
17:41:01 Quit liar (Ping timeout: 258 seconds)
17:42:18 Quit DerPapst (Quit: Leaving.)
17:43:09 Join Buschel [0] (
17:43:55 Join FlynDice [0] (~FlynDice@
17:45:52BuschelJdGordon: can you take a look at FS #11395?
17:49:52 Join TheSphinX^2 [0] (
17:54:39 Nick fxb__ is now known as fxb (
17:54:50JdGordonBuschel: yeah, not really much can be done about that... that patch is slightly overkill but fine
17:58:34BuschelJdGordon: there's no sideeffect you would expect and no better solution that comes up your mind?
17:59:25JdGordonnot at 2am and in my slightly tipsy state :)
17:59:46bertrikr0b-, I have a clip+ now with the same fm radio chip that you and pamaury have, meaning I can now more easily debug it and make it work
17:59:52JdGordonWe could of course make it easier to apply those changes
18:00:12JdGordonbut for the time being either do that theme_settings_apply() or leave it open
18:00:24Buschel:o) let's keep the bug entry open unitl tomorrow then :) just in case in should have a better idea.
18:00:52kugelhow about: an event COLORS_CHANGED, and each skin would have a callback that updates every viewport which has the colors via "-"?
18:01:22JdGordonbecause we dont store that viewports are set with that colour
18:01:25kugelthough I guess %Vf/%Vb complicate it
18:01:40JdGordonyes and no...
18:01:42kugelit could be stored easily
18:02:53JdGordonchanging the background is another thing which shoudlnt require the full theme reload
18:03:14*JdGordon has fantastic plans for the skins in the future incluidng not loading all in one hit like now
18:08:03 Quit TheSphinX^ (Quit: XChat@Linux)
18:08:33 Quit robin0800 (Remote host closed the connection)
18:09:07 Quit TheSphinX^2 (Quit: Got to keep moving!!)
18:10:08 Join TheSphinX^ [0] (
18:12:09 Join funman [0] (~fun@rockbox/developer/funman)
18:12:38funmanbertrik: yep, rockbox utility only downloads released bootloaders
18:13:53 Join fdinel [0] (
18:15:27bertrikfunman, r0b, pamaury, I expect to be able to post a basic tuning patch for the clip+ by the end of the weekend or so
18:17:36funmansaratoga: i think the simplest way to have this tremor patch committed is for you to ask svn access
18:18:56bertrikfunman, is there anything else in particular you want me to take a look at (and that applies to clip+)?
18:19:31bertrikMaybe experiment a bit with PLL/clocks/boosting?
18:20:04funmanexperimenting with PLL would be nice, at least to make people not cry about playback frequency 1% too small
18:20:30bertrikDo we boost/unboost at all on amsv2?
18:20:52funmanyou can change a bit and measure the new pclk by comparing TIMER frequency and RTC, or measure new fclk by running a simple loop with known number of cycles and comparing with RTC again
18:21:17funmanno, the experiments at FS #11297 all crash for unknown reason
18:21:54bertrikI was thinking that maybe I could just put playback on a separate PLL with some kind of clicktrack (1 click/second) and see how fast it clicks for various setting
18:23:05funmanthat should work too
18:23:13funmani think there's a second PLL setting in the OF, let me look
18:24:33 Quit TheSphinX^ (Quit: Got to keep moving!!)
18:24:38 Quit Buschel (Ping timeout: 252 seconds)
18:24:45funman240MHz = 0x113b or 0x1813, and 160MHz = 0x154f
18:30:32 Quit kugel (Remote host closed the connection)
18:30:43 Join flydutch [0] (
18:34:23 Quit FlynDice (Read error: Connection reset by peer)
18:34:50funmanbertrik: about the Clip+, I think all problems are listed on SansaAMS wiki
18:35:11 Join FlynDice [0] (~FlynDice@
18:36:00funmanfeel free to pick up any item to work on, i'm not doing anything on Sansa these days (they work fine enough for me) - also feel free to ask if something you need isn't on this page
18:36:33 Join toffe82 [0] (
18:39:55funmanundef instruction handler and division by 0 are called with lr = address of the next instruction after the one which caused exception
18:40:31funmanfor undef instruction we substract 4 from the address to print the address of faulty instruction, but we could be in thumb mode
18:40:41funmanfor division by 0 we don't adjust at all
18:43:37funmanUnhelpful: can you explain how __div0* work ?
18:44:20 Quit Strife89 (Quit: Reboot to Windows.)
18:44:23funman__div0_wrap* in support-arm.S seems to set r0 to something but __div0() in system-arm.c is (void)
18:45:04 Quit domonoky (Read error: Connection reset by peer)
18:45:28 Join stripwax_ [0] (
18:46:03 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
18:46:50 Quit kramer3d (Quit: Leaving)
18:47:44 Quit stripwax (Ping timeout: 245 seconds)
18:53:15 Quit FlynDice (Quit: Gotta go fly!)
18:53:23dfktpixelma, i see you've been working on cowon x5 keymaps. may i suggest one improvement to the bootloader/power-on procedure? when flicking the power-on switch shortly, rockbox shows a "shutting down" message, and only a longer flick actually turns the player on. would it be possible to implement a short flick for "on"? i know there were some inofficial dual-bootloaders around that used short/long for loading both stock firmware and rockbox.
18:53:30 Join CaptainKwel [0] (
18:55:27***Saving seen data "./dancer.seen"
18:55:54CIA-6New commit by funman (r26818): Sansa AMS: call the exception handler with the correct address when the exception happened in Thumb state
18:56:18 Quit petur (Quit: Leaving)
18:57:30CIA-6r26818 build result: All green
18:59:39 Join FlynDice [0] (~FlynDice@
19:02:04 Quit stripwax_ (Quit:
19:08:26funmanfuck, BL are 32 bits in thumb
19:09:45 Quit n1s (Quit: Lmnar)
19:15:41CIA-6New commit by funman (r26819): ARM: __div0() takes no argument
19:15:45CIA-6New commit by funman (r26820): __div0: give the address of the instruction calling division to UIE
19:17:11CIA-6r26819 build result: All green
19:17:53 Join stroem [0] (~chatzilla@
19:18:36funmanpff, this 'All green' is boring, i'll never test my changes again until I see what CIA says when the build breaks!
19:18:43CIA-6r26820 build result: All green
19:18:50*funman kicks CIA-6
19:20:11funmandfkt: if you make a new clip keymap for recording and FM, i'll do my best to convince saratoga to revert the clip+ hold change!
19:31:12dfktfunman, is there any problem with the current clip keymap? everything works fine for me?
19:31:22funmani think it's not very intuitive
19:31:53dfkti never read a manual and could use it within a minute ;)
19:33:04dfktall the basic stuff is the same as on all otehr rockbox targets i have/had - and the specific stuff isn't very intuitive on all other players as well (eg. power switch/button for exiting from the eq, and similar)
19:33:54funmanok :/
19:33:56dfktbut saratoga's change to the keymap made it much less intuitive, in my opinion (and everybody else's who commented in the forum thread)
19:34:59dfkti mean, i can only speak for the cowon x5, cowon d2, clipv1 - i don't know how intuitive keymaps are on other targets, like ipods
19:35:23dfkti really don't think there is any usability problem on the clip+ :)
19:35:31 Join ender` [0] (
19:36:00funmanmajor usability problem of clip+ is if you play too long with the games you have pain to the fingers ;)
19:36:02dfkti even love that a short press on the down button has no function - i use it to turn on the display
19:36:22dfktlol, judging by your gameboy emu screenshots, i can feel with you ;)
19:36:36*funman discovers the effect of pressing long down button
19:36:50funmaneheh, i just checked that it played, didn't spend much time ;)
19:36:50dfktit's perfect, imo
19:37:21dfkthaving the option to have all buttons execute their function immediately, and having one button just for turning the screen on is rather convenient
19:38:48dfkti can look into a keymap for games, if you give me some specific plugin/emu i should look at (i only listen to music on my clip, but if it will bring you to fix the now broken keymap, i will check it out ;)
19:38:59dfkts/if you give me/if you tell me
19:39:12funmanwell i'm not sure if games really need someone to invest a lot of time in them
19:39:35dfktwell, if it helps you or someone else...
19:39:36funmansometimes it's funny but really both the display and the device are too small to enjoy playing a lot
19:39:41 Quit stroem (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
19:39:50funmanchopper keymap is perfect : it's all i ask ;)
19:39:50dfktyeah, it's actually stupid, imo :D
19:39:58 Quit FlynDice (Remote host closed the connection)
19:40:03dfktget a DS or PSP if you want to play games ;)
19:40:19 Join FlynDice [0] (~FlynDice@
19:41:58*dfkt checks out chopper
19:42:26funmandfkt: do you charge your clip+ with rockbox or OF ?
19:43:00dfkthahaha, it's a palm cave clone!
19:43:17dfkti've charged it with of so far
19:43:21 Join liar [0] (
19:43:58dfktis charging user selectable?
19:44:01funmani wonder if the bug where it can't power off is related to USB plug
19:44:22funmanif you press center button while plugging it doesn't reboot to OF so it charges with rockbox
19:44:26funmanis that what you meant?
19:45:02dfktah yes, at some point in time RB stayed active for charging, but now it switches automatically to OF when plugged into USB
19:45:13funmanFlynDice: < anything i should add before testing with battery_bench ?
19:45:15 Quit flydutch (Quit: /* empty */)
19:45:40funman(and test_disk, to be sure)
19:49:58FlynDicefunman: That looks like everything to me.
19:57:38dfktfunman - i see what you mean for the recording screen (i never checked it out before). i will try to come up with something easier to handle
19:59:13funmandfkt: perhaps volume buttons can be useful in this screen, they seem unused
19:59:39dfktyes, i noticed
19:59:54dfktand home- and center-button functions should probably be swapped
20:09:00Unhelpfulfunman: if i recall correctly, i based the support-arm version on how libgcc calls __div0... but i was not able to find any documentation regarding whether the parameter in r0 was used and what it was supposed to contain.
20:10:49 Join dfkt_ [0] (~dfkt@unaffiliated/dfkt)
20:11:29 Quit dfkt (Disconnected by services)
20:11:43 Nick dfkt_ is now known as dfkt (~dfkt@unaffiliated/dfkt)
20:12:03dfktfunman -
20:12:17funmanUnhelpful: both 4.0.3 and 4.4.4 eabi put nothing in r0 in the builds i have
20:12:31Unhelpfulfunman: perhaps i recall incorrectly? ;)
20:12:42funmani can't tell how you recall :)
20:13:33funmandfkt: looks good
20:14:19 Quit r0b- (Read error: Connection reset by peer)
20:14:19*Unhelpful doesn't see that in libgcc.a either. :/
20:14:20dfktfunman, splitting on the separate key without obscure dual function might be good for quick action - and start/stop should be the same button, at least that seems the most logical to me
20:14:33Unhelpfuldid you already remove it? ;)
20:14:47funmandfkt: does stop still pause without splitting recording?
20:14:56funmanUnhelpful: from your libgcc.a ? no it wasn't me ;)
20:15:38Unhelpfulfunman: no, i meant in r26819, from rockbox. :)
20:15:56Unhelpfulbut please feel free to continue not altering my libgcc.a :)
20:18:21funmani'll do my best :P
20:18:56dfktfunman, sorry, i meant start/pause, not stop. i see, tricky thing, implementing stop as well. :)
20:20:52 Nick fxb is now known as fxb__ (
20:23:19 Quit stoffel (Remote host closed the connection)
20:24:54funmani think power would do fine for stop
20:25:03 Join Buschel_ [0] (
20:26:21 Join Guest59276 [0] (
20:27:44 Quit dfkt (Ping timeout: 260 seconds)
20:28:24 Quit simonrvn (Disconnected by services)
20:28:27 Nick Guest59276 is now known as dfkt (
20:28:35 Quit dfkt (Changing host)
20:28:35 Join dfkt [0] (~dfkt@unaffiliated/dfkt)
20:29:05 Join simonrvn_ [0] (
20:29:41 Quit Buschel_ (Ping timeout: 258 seconds)
20:30:28 Nick simonrvn_ is now known as simonrvn (
20:31:02dfktfunman, two alternatives (i like the first one better):
20:39:50 Join qurvel [0] (
20:40:15funmanyeah i think the first one is more intuitive
20:40:20funmanusing volume to navigate is a bit hard
20:41:03dfktyeah - the important usability cue for me was keeping the actual functions and the navigation separate... both variants do that, but the first seems more practical
20:41:30 Part qurvel
20:42:13 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
20:44:28dfktfunman, for the fm radio i would only move the menu from the down button to the center button, to be in line with the audio player screen
20:45:37dfktthe rest is as i would expect it to be :)
20:49:53funmani guess it's fine
20:54:44*funman jumps on the occasion to point dfkt to[]=
20:55:30***Saving seen data "./dancer.seen"
20:55:47 Join moparx [0] (~moparx@unaffiliated/moparx)
20:58:47funmanFlynDice: i'm testing the SD patch on fuzev2, results tomorrow/monday
21:00:34 Quit halmi_ (Quit: halmi_)
21:02:35*dfkt thinks funman wants him to sign up to the unmanageable wiki/forum/flyspray/mailing-list behemoth ;)
21:05:14funmaneheh no problem, you can send me the patch in private ;)
21:07:40 Join Strife89 [0] (
21:09:05 Join r0b- [0] (
21:09:24 Join GeekShad0w [0] (
21:10:06 Quit GeekShadow (Ping timeout: 248 seconds)
21:19:48saratogafunman: I'm really busy this week, and waiting on my clip+ to get here anyway, so if you want to make changes don't wait for me
21:20:12saratogaas for xiph, i'll get back on them eventually
21:20:38funmani saw that sometimes tremor discussion happen on the vorbis list
21:30:45 Quit GeekShad0w (Ping timeout: 260 seconds)
21:31:28 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
21:36:26saratogaFWIW Android seems to have some really interesting codec stuff, but its all apache licensed which wikipedia says is GPLv3
21:36:35saratogatheir headers are amazing
21:36:59saratogadetailed mathematical explanations of each operation, complete with psuedocode and citations
21:39:48 Quit FlynDice (Remote host closed the connection)
21:40:39saratogaheh their mp3 synth_filter isn't any better then ours
21:40:44saratogaat least the ASM isn't
21:41:19 Join Jaykay [0] (
21:45:39funmanperhaps we could distribute the binaries as gplv3 and include some gplv3 (or compatible) code in the source
21:46:21gevaertstechnically we could, but that's a major decision
21:47:08funmantrue, and that would make inclusion of gplv2-only code in the binary not possible
21:47:08 Join wodz [0] (
21:47:22wodzpixelma: ping
21:50:08wodzwhat is this !@# $id: xxxxxx$ line in header of each file?
21:50:34funmansvn keyword?
21:50:36saratogado we have gplv2 only code?
21:50:58wodzit breaks svn diff patches
21:51:26funmanwodz: if you write "$id$" (i think) in a file, and then set svn keyword Id = "id" for this file, "$Id$" is replaced by some svn stuff when you check it out
21:51:41funmanIMO it's useless, and even harmful
21:52:16funmansaratoga: we try to avoid it, so I don't think there is any
21:52:23saratogathats what i thought
21:52:35saratogabut unless one of these android codecs is amazing its probably not worth the effort
21:53:47gevaertswodz: it does? I don't think I've ever had problems with patches due to $id lines
21:54:51funmanperhaps 'svn diff' remove them but not plain 'diff' ?
21:55:30wodzgevaerts: I had twice so far. The problem comes if You copy or rename file which contains such line already
21:55:42simonrvni did see one in the arm/as3525/usb-drv-as3525v2.(c or h, not sure)
21:56:02simonrvnactually both
21:56:17gevaertswodz: are the svn keywords set correctly?
21:57:56wodzgevaerts: honestly I don't know I usually use git (which by the way produce diff which is ok for git apply but not for plain diff)
21:58:09wodzs/plain diff/plain patch/
21:58:28funmanwodz: it's ok for patch -p1 , git adds a/ b/ in front of the path
21:58:40wodzfunman: it is not
21:59:21wodzfunman: try to apply latest patch from FS #11189 it gives rejects due to $id lines
21:59:43wodzfunman: but git apply is happy with it
21:59:52funmanoh ok
22:00:28funmanI would just remove all these $Id$ lines, they serve no purpose
22:00:44gevaertsthey do serve a purpose
22:01:17funmangevaerts: which one ?
22:01:36gevaertsbeing able to see who modified a file last and when without having to leave the editor
22:02:44funmanwhy would you want to know that?
22:03:04funman(and only that)
22:03:09gevaertsand the $id lines are not why those diffs don't work with plain patch anyway
22:03:24gevaertsfunman: why do you want to remove them?
22:03:36gevaertsi.e. what advantage does *not* having them give you?
22:03:50funmanno trouble when we move to git
22:04:00gevaerts*if* we move to git
22:04:12funmanthe 'if' was included in the 'when' ;)
22:04:39funmanand I already experienced the problems wodz describe
22:04:55gevaertsThose problems are unrelated to a potential move to diff
22:05:45gevaertsThey're purely related to using svn on the server (possibly with incorrect keywords set. That's unclear), and git-svn locally. git-svn is known to not be ideal
22:06:23 Join anewuser [0] (anewuser@unaffiliated/anewuser)
22:06:33wodzgevaerts: I had this problem also with svn when I was preparing initial commit of HD200 port
22:06:55gevaertswodz: presumably keywords were set wrong then
22:07:21wodzgevaerts: anyway if You know the remedy to this please share it
22:07:25funmanthere's no keywords substitution in git-svn : the line just shows up as "$Id"
22:07:48saratogawe should switch back to CVS
22:07:55saratogaproblem solved
22:08:07gevaertswodz: in this particular case the $id rejects are totally unimportant, they're in separate chunks anyway. The main thing is that that patch has rename lines, which are git-specific as far as I know
22:08:10funmanwodz: manually edit files you copy so they are "$Id$" and not "$Id: XXXXX$"
22:08:46funmanhm right
22:09:04funmanif someone with just svn/patch wants to apply that patch he run svn cp manually first
22:09:17gevaertsfunman: it's known that git-svn doesn't handle keywords properly. To be honest, in my view that's a problem for people who use git-svn to solve...
22:09:33funmangevaerts: well it doesn't handle them at all, so there's no problem
22:10:48funmanif i use git cp, the destination file still reads "$Id$", like the original
22:11:57gevaertsfunman: as a project we use svn, *with* keywords. If people choose not to use svn, that's perfectly fine, but it's up to them to make sure they set keywords properly
22:11:58funmanwodz: i have git-svn version (svn 1.6.6) and no problems with $Id$, did you copy files from an svn checkout?
22:12:22funmangevaerts: there's no way to set keywords from git-svn, only read them
22:13:04gevaertsfunman: do I (or the rockbox project) have to care? That's for people who use git-svn to solve or to work around
22:13:07wodzfunman: the same versions here
22:15:12funmangevaerts: work around for me is to not use keywords at all: no Id, eol-style, mime-type or whatever
22:15:39gevaertsfunman: that's not a worlaround, that's blatantly ignoring the project's conventions
22:16:29funmantrue, i'm working around the conventions
22:17:06funmanI wouldn't use svn props anyway if i was using standalone svn
22:17:06wodzWhat is the benefit of having keywords?
22:18:04funmanwhat gevaerts said at 22:01 : i.e. not much IMO, if I want to know about a file's history I'll run git/svn log anyway to differentiate between cosmetics and functional commits
22:19:59gevaertsWe could live without keywords, but as long as we use svn, we *need* the other properties. Not setting them *really* causes issues, and consciously not setting them is a serious problem IMHO
22:21:08funmanwhich problem you have with unset properties?
22:21:16wodzby other properties You mean mime-type?
22:21:18gevaertswrong line endings for a start
22:21:51wodzevery modern editor I know of handle this
22:22:10gevaertsso? That doesn't mean that e.g. make and the shell do
22:22:20gevaertsThese things *need* to be correct
22:22:23wodzthat's true
22:22:46funmanjust commit the file with unix line endings
22:23:23gevaertsjust set the properties right
22:23:29funmanwhatever works for me
22:23:44gevaertsThat's a very nice antisocial position
22:24:02funmanwell, i don't want to break other people's build either
22:24:08funmanwhat works for me works for svn users too
22:25:01funmani think we can all live fine without svn properties
22:25:14gevaertsmaybe, but that's not for you to decide on your own
22:25:37 Quit dfkt (Ping timeout: 276 seconds)
22:25:53funmani won't set any property on the files i commit / modify because i can't do that easily
22:26:12gevaertsYou can
22:26:38funmani dunno how svn:executable is needed: the executables i committed ran fine everywhere
22:27:20gevaertsIt's not *that* hard to keep a separate pure svn checkout. That's not the preferred way (because you can't set things in the same commit), but it *does* work, and it's *not* hard
22:27:32gevaerts"I can't do that easily" is not an excuse, sorry
22:29:12funmani'm certainly not having a separate svn checkout just for setting useless properties
22:31:51 Join Buschel [0] (
22:36:08 Join dfkt [0] (dfkt@unaffiliated/dfkt)
22:36:10funmanwodz: both patches got no reply :/
22:36:29wodzfunman: from what I read If You have proper config in ~/.subversion/config git-svn will grab that and set properties on dcommit of new file
22:36:51funmani can do that
22:38:11wodzwe should double check this and mention this on wiki I think
22:38:58wodzfunman: I found this here: (section Contributing)
22:41:14funmani added
22:43:08wodzit is against
22:43:33wodzit states that binary files should be mime-type application/octet-stream
22:43:45saratogaBuschel: FWIW android has a lot of mp3/aac with arm asm available under gplv3 compatible license (including dct) if you're still looking at that
22:43:58funmanhm i thought viewvc used image mime types to display them in the browser
22:44:36funmani'll just set eol for .c/.h/Makefile
22:44:53wodzwhat eol we use?
22:45:00 Join Jerom [0] (~Jerom@
22:45:23Buschelsaratoga: do you have link at hand?
22:45:25funmanwell i'd say unix
22:45:48wodzso setting it to native is probably wrong
22:45:51funmandamn i'll just do nothing
22:45:53gevaertsI'd expect native
22:45:59funmaneasiest way to not break anything
22:46:55wodzgevaerts: do You have auto-props set in Your subversion config file?
22:47:17gevaertswell, not consciously
22:47:43wodzso what do You set *by hand*?
22:48:41saratogaBuschel: i was looking at this distro:
22:48:45 Quit jae (Ping timeout: 240 seconds)
22:48:51 Quit moparx (Ping timeout: 258 seconds)
22:49:05saratogai'm not sure if its different then the official source (i have trouble finding things in git based projects)
22:50:15gevaertswodz: native
22:50:27funmani think you can see the commits between 'snapshot of eclair' and HEAD
22:50:36 Join jae [0] (
22:50:47CIA-6New commit by Buschel (r26821): Fix FS #111395. Call settings_apply_skins() after changing theme colours.
22:51:57funman looks rather empty of comments :/
22:52:21CIA-6r26821 build result: All green
22:52:45funmanah i can see other files have commented C code
22:53:23saratogait looks like the c versions of the asm code have the comments
22:53:38saratogabut i do wish they'd inline the c code as comments, it makes it much easier to follow
22:54:17funman < is that what you need?
22:54:56 Join S_a_i_n_t_ [0] (S_a_i_n_t@
22:55:06saratogathe mdct stuff is probably equivalent to what we already have
22:55:20Buschelthe dct16 stuff may be of interest
22:55:23saratogawas mostly wondering about the dct, but looking at the asm it doesn't look all that clever
22:55:28saratogano ldm at all
22:55:31***Saving seen data "./dancer.seen"
22:55:35saratogawhich is probably a bad sign for arm7tdmi
22:55:37 Quit S_a_i_n_t (Ping timeout: 264 seconds)
22:55:43funmancopyright says 1998 so it might be quite old code
22:55:49saratogado cortex cpus not need ldm or something
22:56:22wodztime() to sleep()
22:56:23saratogai'm a little surprised the built in android codecs aren't better optimized
22:56:30 Quit wodz (Quit: Leaving)
22:56:41saratogamaybe the open source ones are jsut fallbacks and OEMs include binary decoders?
22:57:32funmanquite possible
22:59:05Buschelhmm, the android decoder uses just another (not before seen) way to perform the filterbank synthesis...
22:59:57rasherWhy does rockboxdev run make -j8!?
23:00:05funmanbecause zagor is evil!
23:00:23gevaertsevil or just crazy? :)
23:00:25*rasher tells everyone about MAKEFLAGS
23:00:29funmanless than Bagder, who used make -j
23:01:03rasherSurely it should not set any -j?
23:01:28funmannope, perhaps make the -j option an argument of the script
23:01:45The_Sevensaratoga: maybe those things are just fast enough that nobody cares any more?
23:01:46rasherfunman: you can already do MAKEFLAGS="-j8" ./
23:02:04rasherThat is, if we weren't overriding it
23:02:26Buschelsaratoga: the dct16 is a dct variant of the formerly used mpc dct32. seems like androids 2xdtc16() and the following merge_in_place_N32() equals our implementation (merged to on single dct32 function).
23:03:01saratogaso nothing worth taking then?
23:04:50Buschelsaratoga: I don't think so. so dct16 asm looks like just direct translation of c to asm. nothing special. neverhteless: maybe we should try to do the same with our dct? did you take a look at the gcc results?
23:05:13saratogano i haven't looked much at rockbox in the last week
23:06:01saratogawas mostly looking at android because i wondered what arm NEON code looked like, but it seems they don't use it
23:06:35funmanffmpeg has some neon code
23:08:47CIA-6New commit by rasher (r26822): Don't explicitly set -j for make. Inform the user that he can set MAKEFLAGS instead.
23:09:28saratogathe neon gcc intrinsics look neat:;a=blob;f=ndk/apps/hello-neon/project/jni/helloneon-intrinsics.c;h=35367c1702f899ae8ccfe2e819db2dfe9afae185;hb=693ad65e24cefe3d60d3702e6adab13d18b121ba
23:10:18CIA-6r26822 build result: All green
23:11:59saratogai hope this android rockbox thing works out
23:12:06saratogai kind of want to try this stuff out
23:12:27funmangigabeast can do NEON ?
23:12:47 Nick fxb__ is now known as fxb (
23:13:09saratogano just vfp i think
23:15:19funmanhm why is the mention on USB mode setting only for fuzev1 & clip+ in the manual
23:16:20funmanah no, steps for other players are just above, only fuzev2 is missing
23:16:30 Quit Jaykay (Ping timeout: 245 seconds)
23:16:33saratogayeah looks like cortex is where neon starts
23:17:05CIA-6New commit by funman (r26823): fuzev2 manual: mention the USB OF setting
23:18:39CIA-6r26823 build result: All green
23:37:45 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
23:43:29 Quit funman (Quit: free(random());)
23:47:26 Quit Topy44 (Ping timeout: 248 seconds)
23:50:03 Quit CaptainKwel (Quit: Ex-Chat)
23:52:59 Join Topy44 [0] (
23:56:07 Quit bmbl (Quit: Bye!)

Previous day | Next day