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

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

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

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

#rockbox log for 2010-04-30

00:00:25 Quit Llorean (Quit: Leaving.)
00:11:45 Join Llorean [0] (~DarkkOne@
00:11:48***Saving seen data "./dancer.seen"
00:20:50 Quit efyx (Remote host closed the connection)
00:22:24 Quit dantje (Quit: Ex-Chat)
00:22:56 Quit lifeless_ (Remote host closed the connection)
00:23:46 Quit M3DLG (Quit: RAGE QUIT)
00:25:21 Quit kugel (Remote host closed the connection)
00:26:37 Quit scorche (Ping timeout: 264 seconds)
00:30:17 Join scorche [0] (~scorche@rockbox/administrator/scorche)
00:31:39 Quit liar (Quit: Verlassend)
00:34:18 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
00:35:20 Quit merbanan (Ping timeout: 246 seconds)
00:38:21 Quit Schmogel (Read error: Connection reset by peer)
00:43:07 Quit jgarvey (Quit: Leaving)
00:53:11 Quit Llorean (Quit: Leaving.)
01:03:53 Quit MethoS- (Remote host closed the connection)
01:12:44 Quit mapi1 (Remote host closed the connection)
01:21:40 Quit grndslm (Quit: Leaving)
01:23:05 Join mikroflops_ [0] (
01:26:23 Quit mikroflops (Ping timeout: 246 seconds)
01:38:58 Quit DerPapst (Quit: Leaving.)
01:48:40 Part toffe82
02:00:02 Quit planetbeh (Ping timeout: 248 seconds)
02:08:22 Join Rob2223 [0] (
02:11:51***Saving seen data "./dancer.seen"
02:12:02 Quit Rob2222 (Ping timeout: 245 seconds)
02:21:05 Join komputes_ubuntu [0] (~komputes@ubuntu/member/komputes)
02:27:44 Nick fxb__ is now known as fxb (
02:28:22 Nick fxb is now known as fxb__ (
02:36:36soapJust to be sure - there is no Rockbox bootloader on the iAudio ports, correct? iAudio bootloader launches Rockbox directly? This is why dual-boot is not there?
02:36:52soaptake that back
02:37:19soapRockbox bootloader instead of iAudio firmware. I misread.
02:39:07 Part kkurbjun
03:09:28 Quit komputes_ubuntu (Ping timeout: 240 seconds)
03:14:17 Join mapi [0] (
03:14:39 Join CGL [0] (~CGL@
03:27:29 Quit adnyxo (Ping timeout: 265 seconds)
03:33:59 Join saratogalab [0] (~9803c20d@gateway/web/freenode/x-dwsokubopewrvppo)
03:44:14 Join chrisb [0] (
03:45:35 Join MagusG [0] (
04:07:29 Join kramer3d [0] (~kramer@unaffiliated/kramer3d)
04:08:10 Join n1s [0] (~n1s@rockbox/developer/n1s)
04:10:19 Quit Barahir_ (Ping timeout: 246 seconds)
04:11:55***Saving seen data "./dancer.seen"
04:12:03 Join Barahir [0] (
04:12:25 Quit S_a_i_n_t (Ping timeout: 276 seconds)
04:12:46 Join S_a_i_n_t [0] (S_a_i_n_t@
04:13:05saratogalabsoap: would you ban user "kindly" for spam?
04:14:12soapdone, saratogalab
04:25:23 Quit amiconn (Disconnected by services)
04:25:25 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:25:42 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:25:42 Quit pixelma (Disconnected by services)
04:25:47 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:26:02 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:26:47 Quit anewuser (Quit: What do you know...THE WORLD'S first NTRQ (that's for NES/FAMICOM) tracking compo. Have powerpak? Try it out! Otherwise ROM IMAGE.)
04:33:26 Quit _arbingordon (Quit: `)
04:35:24 Join arbingordon [0] (~w@unaffiliated/arbingordon)
04:38:58 Quit panni_ (Read error: Connection reset by peer)
05:03:51 Join CaptainKewl [0] (
05:04:42 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
05:05:24 Quit n1s (Quit: Lmnar)
05:18:51 Join Llorean [0] (
05:20:58 Quit CaptainKewl (Remote host closed the connection)
05:21:23 Quit Llorean (Changing host)
05:21:23 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
05:23:16 Join robin0800 [0] (~quassel@
05:41:27 Quit Horscht (Quit: Verlassend)
05:42:16 Quit robin0800 (Quit: No Ping reply in 180 seconds.)
05:42:44 Join robin0800 [0] (~quassel@
05:56:04 Join CaptainKewl [0] (
05:57:23 Join Ryn`cha [0] (
06:00:17 Join robin0800_ [0] (~quassel@
06:10:59 Quit chrisb (Ping timeout: 265 seconds)
06:11:58***Saving seen data "./dancer.seen"
06:21:56 Quit robin0800 (Ping timeout: 264 seconds)
06:22:06 Quit robin0800_ (Ping timeout: 265 seconds)
06:30:57 Part Ryn`cha
06:33:38 Quit CaptainKewl (Remote host closed the connection)
06:36:18 Join shai [0] (
06:50:59 Join BHSPitMini [0] (
07:25:09 Quit CGL (Remote host closed the connection)
07:38:53 Join qurvel [0] (
07:40:25 Join phanboy4 [0] (
07:42:05 Part qurvel
07:46:32 Quit Boldfilter (Quit: Boldfilter)
07:56:08 Quit BHSPitMonkey (Remote host closed the connection)
08:03:16 Join B4gder [0] (~daniel@rockbox/developer/bagder)
08:03:53 Quit rvvs89 (Ping timeout: 252 seconds)
08:06:00 Quit elcan (Ping timeout: 248 seconds)
08:11:20 Quit saratogalab (Quit: Page closed)
08:11:59***Saving seen data "./dancer.seen"
08:16:34 Quit kramer3d (Quit: Leaving)
08:18:01 Join rvvs89 [0] (
08:19:41 Join elcan [0] (
08:29:13 Quit n17ikh (Ping timeout: 240 seconds)
08:30:12 Join n17ikh [0] (
08:33:14 Join ender` [0] (
08:40:10 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
08:54:06 Join grndslm [0] (
08:54:12 Join bimbel [0] (~Miranda@unaffiliated/bmbl)
08:55:35 Quit bmbl (Ping timeout: 252 seconds)
08:55:49 Join Luca_S [0] (
09:01:34S_a_i_n_tJdGordon: Ping.
09:03:11S_a_i_n_tThe .sbs no longer flashes...but during the database init, several differently sized splashes are visible.
09:03:16S_a_i_n_tIs this intended?
09:03:46JdGordonyeah, I decided to fix that as a seperate commit
09:03:51S_a_i_n_tMy understand was they should all be the same size if they appear within one second of each other.
09:03:59JdGordonmostly because I couldnt be bothered making it perfect last night :p
09:04:00S_a_i_n_tOhhhh....I see.
09:04:36S_a_i_n_tRight, that's cool...I just thought it was supposed to fix the .sbs *and* the splash sizes.
09:04:42S_a_i_n_tNever mind then ;)
09:06:18S_a_i_n_tHaving the splashes all one size (if there are quickly alternating splashes, such as the database init) would be a pretty big aesthetic improvement IMO.
09:06:41S_a_i_n_tI guess it didn't look so bad on the slower targets ;)
09:07:41JdGordonno, I couldnt get my head around the centering code needed to make it work :p
09:08:12JdGordonmy test target was e200-sim with a stupidly narrow ui viewport (like 50px wide)
09:08:43 Join flydutch [0] (
09:09:01S_a_i_n_tIMO, even if the text was left aligned and the splashes were a consistent size it would look bettter than it does now...
09:09:32JdGordonthe db init is about the only place where that is shown, which probably means it should be redone to not use splashes
09:09:42S_a_i_n_tIt just looks weird when a smaller splas is drawn overtop of a larger, previous one...then you get two outlines, and it just looks...weird.
09:10:25S_a_i_n_tAs long as I got a splash at the end of the database init, telling me it had completed, that would be fine with me.
09:10:42S_a_i_n_tI don't need to see irrelevant numbers of files found during init.
09:11:13S_a_i_n_tBut, I suspect that may cause some debate with some people.
09:12:21S_a_i_n_tBut, yeah...I for one would be happy to see a reduction in the ammount of splashes used during database init. One (at the end, upon completion) is enough.
09:22:03 Join DerPapst [0] (
09:25:10 Join petur [0] (~petur@rockbox/developer/petur)
09:35:17 Join solexx_ [0] (
09:38:46 Quit solexx (Ping timeout: 258 seconds)
09:47:57 Join einhirn [0] (
09:53:46 Quit BHSPitMini (Ping timeout: 265 seconds)
10:03:25 Join funman [0] (~fun@rockbox/developer/funman)
10:12:03***Saving seen data "./dancer.seen"
10:19:22 Quit einhirn (Read error: Connection reset by peer)
10:19:31 Join lpereira [0] (
10:21:43 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
10:22:50 Join esperegu [0] (~quassel@
10:24:33 Join efyx [0] (
10:26:09 Quit lpereira (Ping timeout: 268 seconds)
10:27:55 Join lpereira [0] (
10:34:24 Quit phanboy4 (Quit: Leaving)
10:35:25 Join pamaury [0] (~c2c7a50a@rockbox/developer/pamaury)
10:57:57CIA-5New commit by 03funman (r25762): sd-as3525v2: cosmetics
10:58:25 Quit lyngaas (Quit: Leaving.)
10:59:21funmanFlynDice: shouldn't you have your copyright in this file ?
10:59:43 Quit Luca_S (Quit: CGI:IRC)
11:02:22 Join hebz0rl [0] (
11:08:16 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:12:57 Join hebz0rl_ [0] (
11:23:28 Join mischasworld [0] (
11:23:44 Quit hebz0rl_ (Quit: Ex-Chat)
11:24:50funmanFlynDice: < my not-working attempt at using PIO reads. The linux driver 'schedules' 'tasklets' for PIO read but i've no idea how this works. Perhaps partial PIO reads are triggered with some interrupt ?
11:25:55funmanah yep it seems they are triggered with RXDR/TXDR
11:30:09 Join wodz [0] (
11:37:46funmanFlynDice: with at least 3 interrupts RXDR are triggered (stopped counting after 3) but rockbox stays stuck at the logo
11:39:04funmanoops count is in sectors
11:39:08wodzI think my next port will be rockchip 270x - quite nice SoC
11:39:36wodzI found looking complete datasheet yesterday
11:42:09funmanFlynDice: (count is in sectors, not bytes) - still not working and might even crash the clip+ because backlight thread doesn't work anymore
11:44:08bertrikwodz, nice to hear
11:44:29funmanwodz: you have a rockchip player already?
11:46:26 Quit pamaury (Quit: Page closed)
11:47:19wodzno but this is ~50$ cost so I think I will ask for some as a my birthday gift :-)
11:48:27 Quit esperegu (Read error: Connection reset by peer)
11:49:07wodzGenerally it is arm7ej-s + ZSP400 combo with integrated lcd driver, codec, sdram controller, adc and lots of other cool stuff onboard
11:49:35 Join esperegu [0] (~quassel@
11:49:47wodzFrom what I know for now DSP core is not used for audio
11:50:42funmanhopefully, i think those are largely undocumented
11:50:49 Quit funman (Quit: free(random());)
11:51:55wodzThere is full documentation for ZSP40x core familly from LSI. The problem is there is no free toolchain for it I am aware of
11:52:58B4gderarm7 ?
11:53:23B4gdera bit unusual these days
11:53:23wodzBut there is "entry point" toolchain which can(?) be downloaded from on of LSI spinoffs site after registration which I belive is gcc based
11:53:40wodzarm7ej-s != arm7tdmi
11:54:22B4gderI know
12:06:06 Part PRINCESS_FLUFF
12:08:22wodzamiconn: ping
12:12:05***Saving seen data "./dancer.seen"
12:32:52 Join JohannesSM64 [0] (
12:54:54 Join DerPapst1 [0] (
12:55:29 Quit DerPapst (Ping timeout: 265 seconds)
13:01:34 Join funman [0] (~fun@rockbox/developer/funman)
13:02:56 Quit flydutch (Read error: Operation timed out)
13:06:33 Join flydutch [0] (
13:10:35wodzhow are plugins loaded?
13:10:58wodzI mean how can I debug plugin with gdb
13:11:58CIA-5New commit by 03mt (r25763): Add libwmapro to apps/codecs. These files comprise a set of unmodified files needed from ffmpeg's libavcodec and libavutil to compile and use the wma ...
13:12:23 Join mt [0] (~chatzilla@
13:12:44funmanwodz: you can use simulator if you break on plugin_start()
13:13:29wodzI can use BDM+gdb to debug in system
13:14:03wodzbut if I load symbols from .elf for plugin addresse do not match
13:14:07 Join b0hoon [0] (~quassel@
13:15:14funmanis there a small shift?
13:15:29wodzI didn't investigated this much
13:15:37funmanperhaps plugin_load() can be helpful
13:16:00wodzok I'll look
13:16:23funmanthere's a 0x18 bytes header but it's included in the elf file
13:19:19 Join LinusN [0] (~linus@rockbox/developer/LinusN)
13:20:51 Quit arbingordon (Read error: Connection reset by peer)
13:21:25 Join arbingordon [0] (~w@unaffiliated/arbingordon)
13:21:35 Join rajendrau [0] (~chatzilla@
13:23:15rajendrauwhen i copy videos into video folder, I am not able find them in Sansa e260 v2 player through rockbox interface
13:24:31soapamiconn, why does the X5's line-out mute when the volume is set to minimum but doesn't change volume otherwise?
13:25:16 Join DerPapst [0] (
13:25:45 Join japc [0] (~japc@
13:26:36 Part LinusN
13:26:48 Quit DerPapst1 (Ping timeout: 246 seconds)
13:29:31 Join anewuser [0] (anewuser@unaffiliated/anewuser)
13:32:03funmanrajendrau: use 'show all files' setting
13:32:35rajendrauwhere is that setting
13:32:42rajendrauin the player ?
13:33:06funmancheck the manual
13:33:29rajendraufunman: ok thank you
13:35:12funmanmpegplayer only opens .mpg, .mpeg, .m2v and .mpv
13:35:57linuxstbI'm not sure it should still open m2v and mpv - those are probably left from when it only played raw video streams.
13:36:12 Join mikroflops [0] (
13:36:14linuxstbAnd I'm guessing it won't play those any more.
13:38:09funmanlinuxstb: raw mpeg2 works fine
13:38:33soapsoap, I _assume_ mute is at the software mixer level?
13:38:47funmanthey're probably not common at all but it doesn't harm to having them associated
13:38:49linuxstbfunman: Are you sure about that? You mean just a raw mpeg-2 video stream (no container) ?
13:38:57funmansoap: yes, rockbox doesn't mute speakers/headphones
13:39:07funmanlinuxstb: yes
13:40:05soapfunman, I think you may have been lead to taking that line of mine out of context. On the Cowons the line-out is at a constant volume (as it should be) /except/ for mute, where it is, well, muted.
13:40:32 Join Schmogel [0] (
13:40:49 Quit mikroflops_ (Ping timeout: 276 seconds)
13:41:11soapLooking at the DAC datasheet I'm /assuming/ Rockbox is muting at the software mixer level because I didn't see a way to mute the line out in hardware. I was looking for confirmation or denial on that specific case.
13:41:45funmanif any muting is performed it's done by the DAC driver
13:42:04funmanapps/* can't mute
13:43:14funmanso for cowond2 it would be in drivers/audio/wm8985.c : i only see muting when enabling recording though. Perhaps output is muted when volume is 0 ?
13:43:44linuxstbfunman: Sorry, looking at the source, it looks like you're right. I don't understand why jhMikeS bothered to do that though (I _think_ it was part of his work, not mine...)
13:44:19funmani suppose raw streams are easier to parse than PS
13:45:34funmani didn't look at the source but i tried it on target :)
13:46:42wodzWolfson codecs usually mute when volume is set to 0
13:46:50wodzbut mute in analog domain
13:52:39 Quit mischasworld (Ping timeout: 240 seconds)
13:59:17soapwhere, wodz, in the analog domain would such a mute happen. Looking at a block diagram (I know, not the schematic) I see very little in the analog path of the line-out.
14:00:09 Quit shaggy-h (Ping timeout: 240 seconds)
14:01:09wodzsoap: wait a sec I'll look at datasheet
14:04:22kisak_funman: morning, remember me from 2 days ago?
14:06:55 Join mitk [0] (~mitk@
14:07:07funmanyou got registered on wiki ?
14:07:11kisak_I did
14:07:30kisak_my user is DwayneAsbury
14:08:46funmanit should work now: i added you to WikiUsersGroup
14:09:04wodzsoap: if OUT3/4 is used You are right than the only way to mute such output is to disable signal path in mixer
14:09:43 Join robin0800 [0] (
14:09:47 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
14:09:56wodzamiconn, alive?
14:12:09***Saving seen data "./dancer.seen"
14:13:52 Quit jhMikeS (Read error: Connection reset by peer)
14:14:14kisak_funman: thanks, I've made the changes that make sense to me
14:14:25 Join jhMikeS [0] (
14:14:26 Quit jhMikeS (Changing host)
14:14:26 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
14:20:27CIA-5New commit by 03mt (r25764): Modify the ffmpeg source files in apps/codecs/libwmapro in order to compile the codec standalone.
14:20:50 Quit JohannesSM64 (Quit: WeeChat 0.3.2-dev)
14:23:09 Join JohannesSM64 [0] (
14:24:46 Quit mapi (Ping timeout: 276 seconds)
14:24:57 Join mischasworld [0] (
14:28:44 Part rajendrau
14:33:10 Quit mt (Ping timeout: 240 seconds)
14:34:14 Quit B4gder (Quit: It is time to say moo)
14:34:22 Quit robin0800 (Remote host closed the connection)
14:34:28TorneThe tracker reported version field really needs to be updated to include 3.5...
14:34:31Tornewho has permission to do that?
14:34:46Torne(due in version and other fields might want some love too)
14:35:58wodzTorne: what is the status of ipod charging? There is experimental patch which works for a few people but is hard to find
14:36:30 Quit xavieran (Quit: ➤➤➤➤➤➤ UniCode shall rein forever! ➤➤➤➤➤➤)
14:37:10 Join xavieran [0] (
14:37:47Tornewodz: the status is i have a plan to rework charging in general to handle USB in a vaguely correct way
14:38:05Tornei got some way with this but hit a bit of a problem and haven't put in the time to work out a better way
14:38:33 Quit xavieran (Client Quit)
14:38:59TorneFS #8802 needs to be tested on more models with a multimeter handy, also
14:39:02 Join xavieran [0] (
14:39:24Torneto make sure that the GPIO it's toggling is definately the current limit :)
14:39:32funmanTorne: only handy? don't you need to *use* the multimeter? :)
14:39:53Tornei might try and do this at devcon if there are other ipods around
14:40:17Tornewodz: anyway, yeah. i want to fix how we handle USB charging in general first
14:40:37Tornebecause right now we ignore the actual state of the usb connection pretty much, and draw whatever power we feel like
14:40:56soapTorne, what models and what needs physically tested?
14:41:13Tornesoap: er, whatever doesn't say it's been tested on FS #8802. I have only personally tried ipodvideo.
14:41:23 Join mapi [0] (
14:42:04Tornerun down your battery some, apply patch from 8802, then measure USB Vbus current draw with the usb charging setting enabled/disabled
14:42:10Torneshould be ~500 and ~100 mA respectively
14:42:26soapoh? That's all?
14:42:38Tornei don't want to commit the patch as-is regardless, though
14:42:51Tornebecause all it does is bring the ipod to the same dodgy state as several existing targets, which draw power when they are not allowed to.
14:42:53CIA-5New commit by 03wodz (r25765): HD200 - remove moving average filtering of ADC readouts
14:43:11Tornebut the patch is perfectly fine to test and make sure we have the right GPIOs on each model
14:43:11soapYea, I'm clear on that. Wasn't clear what multimeter testing was implicit in the #8802 discussion, though.
14:43:37Torneyeah, that's it. the code from 8802 is only concerned with switching the 100/500 limit
14:44:02Torneit's possible that on some models we could also set it to 0 or unlimited, depending what other charger chip lines are hooked up, but i don't know :)
14:44:37Torne(setting to unlimited, if possible, would be useful in concert with positive charger detection, i guess; setting it to 0 has possible use cases but also the huge pitfall of not being able to charge a dead battery) :)
14:57:26 Quit nima (Ping timeout: 246 seconds)
14:58:41 Quit anewuser (Quit: What do you know...THE WORLD'S first NTRQ (that's for NES/FAMICOM) tracking compo. Have powerpak? Try it out! Otherwise ROM IMAGE.)
14:58:43 Quit bertrik (Read error: Connection reset by peer)
14:59:02kisak_I've gone and bought a Sanza Fuze, here's to hoping it's a V1
14:59:05 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
14:59:11 Join nima [0] (
15:07:31 Quit shai (Read error: Connection reset by peer)
15:07:49 Join shai [0] (
15:09:49 Join evilnick_B [0] (~0c140464@rockbox/staff/evilnick)
15:16:18 Part b0hoon ("Back to work.")
15:23:57 Quit funman (Quit: free(random());)
15:28:57 Join shaggy-h [0] (
15:34:12 Quit esperegu (Remote host closed the connection)
15:35:16 Quit FlynDice (Remote host closed the connection)
15:38:42 Join FlynDice [0] (
15:39:49 Quit mc2739 (Ping timeout: 258 seconds)
15:45:22 Join CGL [0] (~CGL@
15:49:53 Quit mitk (Quit: Leaving)
15:55:33 Join adnyxo [0] (
15:55:50 Quit mischasworld (Ping timeout: 246 seconds)
15:55:58*Torne prods people with ipods other than the video to test his bootloader test build:
15:56:19 Quit japc (Quit: Ex-Chat)
15:59:02CIA-5New commit by 03wodz (r25766): HD200 - small fix in backlight code
15:59:23linuxstbTorne: It would be nice to make the nano2g bootloader consistent (i.e. enter disk mode on usb detect after failing to start rockbox)
16:00:43Tornelinuxstb: well, i can do that, but someone with a 2g would have to test it
16:01:09linuxstbI should be able to try it (tonight or over the weekend)
16:01:32Torneok, i'll poke it in a bit.
16:02:16linuxstbI assume you just need a working usb detect? IIRC, there's no button handling (apart from the hold switch) in the Nano2G bootloader.
16:03:05 Nick fxb__ is now known as fxb (
16:05:16S_a_i_n_tif you could fix being able to use Menu to boot the OF for Nano2g, it'd be great.
16:05:34S_a_i_n_tHaving to use hold ucks...
16:05:50S_a_i_n_tTorne: ^
16:05:51linuxstbDon't use the OF then...;)
16:06:32S_a_i_n_tlinuxstb: Hardy har har... ;P
16:09:08S_a_i_n_tTheSeven was going to look into it IIRC, but got himself tangled up in the mess that is Nano2g FTL
16:12:12***Saving seen data "./dancer.seen"
16:13:53CIA-5New commit by 03wodz (r25767): HD200 - implement blit functions
16:14:45 Quit wodz (Quit: Leaving)
16:24:49*S_a_i_n_t scrolls up a tad...I will gladly test the Nano2G bootloader. I'm not having any issues with the test-build + new bootloader on my Nano1Gs, all works as intended.
16:24:57S_a_i_n_tTorne: ^
16:26:45 Join liar [0] (
16:28:48 Quit S_a_i_n_t ()
16:29:09 Join S_a_i_n_t [0] (S_a_i_n_t@
16:43:47 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
16:47:26 Quit liar (Ping timeout: 252 seconds)
16:48:10 Join liar [0] (
16:53:37 Quit adnyxo (Ping timeout: 260 seconds)
16:53:52TorneS_a_i_n_t: Unfortunately the lack of button handling in the nano 2g bootloader appears to be because it just says "todo" instead of reading hte keypad :)
16:54:07TorneS_a_i_n_t: so I think you are stuck with using the hold switch until someone who knows more about nano2g looks :)
16:54:15TorneS_a_i_n_t: i'll see if i can get it to boot disk mode though
16:58:45S_a_i_n_tTorne: Thanks. I'll poke TheSeven when I see him next and see if he can do something about it. The hold switch on the Nano2Gs is a cheap plastic piece of shit, and if you've ever opened the case of your 2G there's a good chance that 1 in 5 times or so it decides not to work.
16:59:14S_a_i_n_tThe hold switch is nowhere near as sturdy as the one on the Nano1g, or any other iPod for that matter.
17:00:36 Quit petur (Quit: connection reset by beer)
17:00:58S_a_i_n_tIIRC, TheSeven said it would be trivial to implement checking if menu is pressed (like the other iPods do) on boot, but he had his head pretty deeply stuck into the Nano2G FTL at the time, and then he got stuck into iLoader...and now has no personal need for this feature I guess ;)
17:01:12Tornewell, ok, i'll have a quick look ;)
17:01:56S_a_i_n_tI didn't mean do it yourself lol, I'll poke TheSeven when I see him next and try to do some sweet talkin' there ;)
17:02:14S_a_i_n_tI'l flutter my eyelids :P
17:02:17 Join Strife89|PalmTX [0] (
17:06:14Tornehm, usb is not 100% trivial on nano2g
17:06:21Torneit doesn't include the usb driver in the bootloader :)
17:06:43S_a_i_n_t*sad trombone*
17:06:54Tornewell, I could add it ;)
17:06:57S_a_i_n_tThere had to be *something* ;P *sigh*
17:06:59 Join archivator [0] (
17:07:00Torneor just use charger_detected..
17:07:07Tornewhich is all usb_detect does atm on nano2g
17:07:33S_a_i_n_tIIRC, I noticed it was broken...and you fixed that up.
17:07:41S_a_i_n_t*charger detect that is.
17:08:06*S_a_i_n_t was pretty sure you did...
17:08:15Tornethat wasn't the 2g ;)
17:08:35S_a_i_n_tAhhhh, right, sorry. 1g, you're quite correct.
17:08:53Torneundefined reference to udelay? what? :)
17:09:02Tornethe nano2g bootloader doesn't know how to sleep?
17:10:04S_a_i_n_tI don't think the bootloader has had as much attention as it should have done really, probably due to TheSeven using iLoader instead ;)
17:10:26TorneS_a_i_n_t: try
17:10:34Torneshoudl reboot on cable insert
17:11:00Tornenot sure if i want to commit that, though (it's just using charger_inserted() which means if the definitoin of usb_detect() gets altered to be a real USB detection it won't change the bootloader
17:11:09S_a_i_n_tis that if rockbox.ipod is not present?
17:11:21Torneyeah, that's the easiest way to provoke it ;)
17:11:24Torneit's any fatal error
17:11:27Tornejust rename rockbox.ipod
17:12:49Tornei wonder if TheSeven has some reason not to include the usb driver, other than it's not needed
17:12:56Tornethe other ipods included it even though it wasn't needed ;)
17:13:08Tornei suspect he was just being consciencious
17:14:38Tornewell, including the usb driver works :)
17:14:58Tornethat migh tbe better.
17:14:58 Join mischasworld [0] (
17:15:02Torneit's only 2k
17:18:29S_a_i_n_tTorne: ???
17:18:47TorneOh, that binary isn't encrypted
17:18:49TorneI don't have a nano2g
17:18:55Torneyou will have to encrypt it on yours first
17:19:03*S_a_i_n_t slaps his forhead...
17:19:14Tornethat's what the 'x' means
17:19:54archivatorIs there an elegant way to link codeclib's tlsf in a plugin?
17:20:21S_a_i_n_tTheSeven has some Python lib that can decrypt/encrypt Nano2g stuff IIRC
17:20:37Torneer, i thought you needed the (unknown) keys to do that
17:20:41Tornewhich is why we do it on device
17:20:47Tornei really don't know, anyway
17:21:21liarS_a_i_n_t: IIRC linuxstb wrote a rockbox plugin to encrypt firmware/bootloader files on nn2g
17:21:22Tornedoes anyone have an opinion on having the usb driver (just the target chip driver, not the usb stack) built into the bootloader in order to use its usb_detect, btw?
17:21:41TorneThe PP ipods already had this, but probably by accident rather than design, and I'm now using it..
17:26:41S_a_i_n_tTorne: It doesn't reboot on cable insert.
17:27:00S_a_i_n_tJust gives the same message on failure it ususally does... :'(
17:27:07Torneer, it shouldn't be the same message
17:27:19TorneShould say "Insert USB cable, or" at the start
17:27:29Torneif it doesn't say that it's not the one i just built :)
17:27:36S_a_i_n_tHold menu+select the menu+play blah blah blah...
17:27:49Torneer, then you didn't install it right
17:27:55 Quit DerPapst (Quit: Leaving.)
17:27:56Tornei changed the message in line with the others
17:28:05S_a_i_n_tErr, *does* say insert cable
17:28:11S_a_i_n_tI missed that part.
17:29:05S_a_i_n_tInsert USB cable, or hold MENU+SELECT (blah blah blah)
17:29:10Tornewhat happens if you turn the hold switch
17:29:21Tornedoes it say "Hold switch on!" until you turn it off again?
17:29:37Tornebecause that is the same loop i am checking charger_inserted() in :)
17:31:23S_a_i_n_tI've just tried it enough times in a row now to cancell out fluke/user error and still, no go.
17:31:36S_a_i_n_tIt doesn't seem to have any idea the cable is inserted.
17:31:37Torneit's not even the cable insert event it detects, it's literally polling
17:31:44S_a_i_n_tOh well, it was worth a try ;)
17:32:33Tornecharger_inserted is from the powermanagement thread ;)
17:32:40Tornewhich doesn't run in bootloader builds
17:32:43saratogaarchivator: i wouldn't use malloc in a new codec unless you must have it
17:34:44archivatorsaratoga: I want to use it in a plugin, actually.
17:35:35TorneS_a_i_n_t: uploaded a new one that should work, but is even *more* horrible ;)
17:35:41Torne(calls power_input_status() directly)
17:36:02S_a_i_n_tsame link?
17:36:30 Quit mischasworld (Remote host closed the connection)
17:37:50Tornetbh I think you should just poke TheSaint though, I have no idea if this is really a sensible long term solution
17:37:56Torneit doesn't look like nano2g has very good usb detection at all atm
17:38:09Torneand its current usb_detect() just won't work in the bootloader.
17:40:06linuxstbTorne: I thought the older ipods had a simple usb detect now as well?
17:40:21saratogaarchivator: yes sorry I meant plugin
17:40:29S_a_i_n_tTorne: That last one won't boot at all...just gets stuck in the apple screen.
17:40:46S_a_i_n_tI'm also gonna be annoying and head to bed, I'm pretty damn tired.
17:41:11TorneS_a_i_n_t: there's no logical reason for it to do that ;0
17:41:19TorneS_a_i_n_t: the diff is tiny
17:41:43Tornelinuxstb: the PP ipods all detect usb from GPIOs, but I'm relying on the implemention of that that's in usb_detect in the PP-model-specific USB driver
17:41:51Tornewhich is included in the bootloader anyway even though it's otherwise not used
17:42:49archivatorsaratoga: I don't currently have a choice. Will try to remove the malloc-ness later on.
17:43:34archivatorHow do I expose the rb variable to files outside the main plugin file?
17:48:23linuxstbarchivator: I think just "'#include plugin.h"
17:48:42linuxstbThat declares "extern const struct plugin_api *rb;"
17:49:16Torneindeed, you need plugin.h to know what struct rb *contains* anyway
17:50:41 Quit ender` (Quit: Shortly after Dr. Samuel Johnson had published his Dictonary of the English Language (in 1755), a splendid British lady approached him and congratulated him on omitting all improper and coarse words from his dictionary."Ah," replied Dr. Johnson, "so yo)
17:50:41S_a_i_n_tTorne: I tried it again (sometimes crypt_firmware can choke...), and it works to boot into diskmode on cable insert, takes you to .rockbox when the host detects it, but then I get "Undefined instruction at 0803FB4C when it tries to boot back into RB :/
17:51:19Tornewell, it's just doing what the nano2g code does anyway if you undefine HAVE_ROCKBOX_USB
17:51:19archivatorThanks, that did it. (for the record, it was a call to isupper that errored out with "rb undefined")
17:51:51S_a_i_n_tOnly was to recover from that apparently is to use ipodpatcher to install a known-good bootloader and hard-restart.
17:52:24S_a_i_n_t*do'h! *way
17:53:29S_a_i_n_tCould all this dodgyness with USB on the Nano2g be why USB transfer is so dog slow, and why HID fails in 9/10 cases?
17:54:32*Torne has no idea.
17:54:37Torneanyway it sounds like no, i can't do this :)
17:58:59S_a_i_n_tTorne: Well, many thanks for trying anyway...I appreciate it. I'll keep a copy of that bootloader that got as far as it did, and see if TheSeven has any ideas.
17:59:16S_a_i_n_tThat man can do some crazy stuff when he sets his mind to it.
17:59:45S_a_i_n_tMy major peev with the Nano2g bootloader is the lack of MENU to boot OF.
18:05:16 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
18:12:14***Saving seen data "./dancer.seen"
18:18:08linuxstbTorne: If you give up on the Nano2G, can you post a patch somewhere? I may try and dig into it.
18:18:27Torneit's miniscule..
18:19:22Tornethe reboot code is copypasted from usb.c, and then i'm just testing power_input_status directly to decide if the cable is attached
18:19:27S_a_i_n_tGo linuxstb, GO!
18:20:47linuxstbS_a_i_n_t: I'm still working, but maybe later this evening or over the weekend.
18:20:48*S_a_i_n_t would *really* like to see this happen, even if the sole reason is just to keep things consistent between the iPod targets.
18:21:20S_a_i_n_tlinuxstb: No rush man ;)
18:21:38linuxstbS_a_i_n_t: You just said "Go GO!" ;)
18:21:59 Join Strife89 [0] (
18:22:13S_a_i_n_tOdd mood strikes again...
18:24:10 Part Strife89|PalmTX
18:26:36archivator.... and I managed to compile flite for rockbox (albeit, with tons of implicit declarations for things like fprintf, etc.). Next step: actually use it.
18:27:44domonokyarchivator: sounds good. :-)
18:31:27 Join DataGhost [0] (~dataghost@unaffiliated/dataghost)
18:32:58 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
18:34:00 Quit S_a_i_n_t ()
18:34:18 Join S_a_i_n_t [0] (S_a_i_n_t@
18:36:58archivatortlsf segfaults :(
18:38:39 Join Tux2 [0] (~Tux2@
18:39:34archivatornevermind me, I'm just stupid.
18:39:49gevaertsoh, good to know :)
18:46:18archivatorRight, so the basic initialization process works well. The voice registers itself fine, just need to figure out how to get synthesis to work before I move on to translating everything into fixed-point.
18:46:37 Join kugel [0] (~kugel@rockbox/developer/kugel)
18:48:19 Join adnyxo [0] (
18:50:04 Quit liar (Ping timeout: 252 seconds)
18:52:26 Join powell14ski_ [0] (
18:54:16archivatorShould I use pcm_play_data to play PCM data from a plugin? :)
18:54:49linuxstbarchivator: There's nothing better...
18:55:10archivatorlinuxstb: just asking, with the pcm* functions, you never know :)
18:55:25archivatoror at least, I never know :)
18:55:49linuxstbI guess in theory you could use the playback engine's pcm/dsp code, but I'm not sure any plugins do so.
18:57:37kugeldoes anyone know where the sim gets its fopen/fclose/fwrite declarations from?
18:58:36linuxstbShouldn't it be stdio.h?
18:58:48kugelit's not in our stdio.h
18:59:06linuxstbI would expect the sim to use the system stdio.h.
18:59:14kugelbut the sim uses it. our stdio.h is preferred as per include path order
18:59:23kugelit doesn't
19:00:06kugelI can make it use the systems one, but then there are incompatibilties in ctype.h (so that ctype.c doesn't build)
19:01:07kugelat some point I'd like to not use any of our libc replacements but completely use the systems one, at least for raaa. I could understand that it's desirable to use the code we use on the targets in the simulators
19:05:13pamaurykugel: I don't remember but I'm pretty sure it uses some system headers in the uisimulator/ part
19:05:33kugelok guys, I poked a bit around with the idea of moving sdl stuff into the target tree
19:05:43kugelpamaury: grepped it without results
19:06:14pamaurywait a minute if I can find it quickly, otherwise you'll have to investigate :/
19:07:33kugeluisimulator/sdl/ contains the very specific sdl code (drawing, threads, kernel, timers, audio). that's what I'd like to move into the target tree
19:08:30pamaurykugel: answer, it doesn't use fopen, ... in uisimulator/io.c, it uses _wopen, close, read and write
19:08:56kugelpamaury: sound.c uses f* functions
19:09:07kugeluisimulator/common contains stubs and utility functions specific to simulating a target. this is what I'd like to keep in uisimulator.
19:09:33pamauryhum, you say that sound.c uses f* functions ?
19:09:37 Quit lpereira (Quit: Leaving.)
19:10:00kugelpamaury: exactly what I said
19:10:11 Join DerPapst [0] (
19:10:23pamaurytrue, only for debug purposes it seems
19:10:45kugelthe purpose doesn't matter. I can't find where it gets the prototypes for it
19:12:01kugelI pused a "sample commit" too kugel-rb.git/shortlog/refs/heads/sim-target-tree">
19:12:14kugelit moves the pcm/audio part into the target tree
19:12:22pamaurylooking at the Makefile in uisimulator/common, it will get them from the system
19:12:29archivatorHa! Synthesis works as well.
19:13:49kugelpamaury: wtf, Makefiles are supposed to be inexistent in our tree. at least I thought so
19:14:19pamaurywell for the simulator it's quite specific, I don't know if it's really used however
19:14:44kugelwhich line is that?
19:16:32 Join Strife89|PalmTX [0] (
19:16:45 Quit Strife89 (Quit: Rebooting.)
19:16:56pamauryyou might have a look in uisimulator/uisimulator.make, line 10 and uisimulator/common/Makefile, line 26
19:17:31pamaurythe Makefile is uisimulator has a licence header and is not generated so I guess it's the an exception to the Makfile rule you stated
19:19:11kugelpamaury: I saw those INCLUDE= lines, but I don't see where it picks the systems' include dir
19:19:29 Join archivator_ [0] (
19:19:43pamauryit's always included by default, except if you add a -nostdinc or something like that I think
19:19:44 Quit archivator (Disconnected by services)
19:19:53 Nick archivator_ is now known as archivator (
19:20:50kugelah, it doesn't add firmware/include to INCLUDE
19:21:04 Join liar [0] (
19:21:34pamauryand furthermore, the Makefile generated for the sim build doesn't have the -nostdlib switch so it links against the system libc
19:21:53kugelright, that's why \n works in the sim
19:22:59kugel(in snprintf for example)
19:23:24pamaurytrue, it's not implemented in rockbox
19:24:37kugelsoo, would anyone like to comment on what I said about the target tree thing?
19:26:37*pamaury looks at the code
19:29:33 Join Boldfilter [0] (
19:29:43pamaurykugel: moving sdl audio code to firmware/ and removing the table in sound.h makes sense to me, I already noticed it was awkward
19:31:15 Join Luca_S [0] (
19:31:19pamaurybut how do you handle sdl inclusion ? Shouln't it change the Makefile ?
19:31:32kugelI don't understand
19:32:14kugelconfigure just adds the sdl-config −−libs/−−includes output to the some environment variables
19:32:52pamauryah ok then
19:33:34kugelI don't handle the system includes yet, which is why I spotted the problem of fopen&friends
19:34:16pamauryhandling system includes in firmware/ will be tricky
19:34:42kugelgevaerts: I'm starting my technical discussion, would you like to attend? :)
19:35:22gevaertskugel: if enough people are involved, can I limit myself to reviewing patches? :)
19:35:22kugelpamaury: it's not that tricky actually, for instance firmware/common/ctypes.c is incompatible with my system's ctype.h
19:35:27*pamaury will soon leave anyway
19:35:59*kugel adds a 'but' before for
19:36:29pamaurydo/will you really need system functions in firmware/ ?
19:37:02 Join phanboy4 [0] (
19:37:31kugelnot for raaa, but it might be desirable for the sims
19:37:52kugellinuxstb: you were skeptical about it IIUC?
19:38:18kugelgevaerts: you can start by reviewing kugel-rb.git/commit/925feaf1dd84d069495f847dcc47d1e75d6fb5ef"> :)
19:38:56 Quit n17ikh (Ping timeout: 276 seconds)
19:39:06pamauryanyway your commit seems ok me if you just moved code
19:39:33kugelthe question is whether moving sdl stuff into the target tree is wanted or not
19:39:46pamauryyes that's the question
19:40:02kugelin my interview it sounded like some people might not want it and that I should open a discussion for it before doing too much in that direction
19:40:49 Quit phanboy4 (Client Quit)
19:40:59gevaertsyes, that decision is the first hurdle
19:41:09kugelmy opinion is that using sdl as backend really counts as a port. it's not an actual device but it gives us what an device would give us too
19:41:13 Quit Kitar|st (Read error: Connection reset by peer)
19:41:15 Join Kitar|st [0] (
19:41:31pamaurymy opinion is that it's ok, I don't see why it should not be the case if it's stays completely hidden in .c file. I mean, the current code works with hardware, one can consider SDL has hardware
19:42:16kugeland as RaaA is supposed to be an actual port too, I think it should be handled the same as classical ports
19:42:25gevaertsIn my opinion, it's a "hardware" (or envoronment-specific) driver, so if belongs in the target tree, at least for RaaA
19:42:48 Join n17ikh [0] (
19:43:03pamauryyes, I agree
19:43:36gevaertsFor the sim it can be argued that this isn't relevant, but I don't think we want duplicated code, and I don't think there will be SDL bits that are needed for RaaA but nor for the sim
19:43:47kugelwe could consider whether firmware/target is appropriate for the RaaA target tree or maybe something like firmware/app would be better, but the general idea is the same
19:44:07gevaertsIt's firmware/target, not firmware/hardware
19:44:38*pamaury leaves
19:44:47kugelthen we could consider firmware/target/app/sdl :)
19:45:23kugelI doubt that sdl will get any subdirs, or an andriod port, so it might be useful to separate app and hardware ports in a way
19:45:35gevaertsOr firmware/target/hosted/sdl ?
19:45:48kugelbut that's all minor, the general idea of a target tree system for RaaA stays the same, and I like that idea
19:45:51gevaerts"app" might annoy the "The sim is different!" crowd
19:46:00 Join Strife89 [0] (
19:46:06 Quit Kitar|st (Ping timeout: 264 seconds)
19:49:26*archivator fails miserably at trying to convince the sim that his audio is sampled at 8kHz.
19:49:46archivatorNo matter how I call pcm_set_frequency and pcm_apply_settings, I *cannot* change the sample rate.
19:49:59archivatorOh, well, at least I can write it to a file and then use aplay..
19:50:59gevaertsarchivator: maybe the sim doesn't handle that situation well?
19:51:19 Join Kitar|st [0] (Kitar_st@
19:51:29gevaertsDoes it play 8kHz audio files properly?
19:51:51kugelarchivator: the sim hardcodes 44.1kHz
19:52:09archivatorThat explains *everything*! :)
19:52:11kugelsee pcm_dma_apply_settings_nolock()
19:52:35kugelah wait, maybe not
19:53:05kugeloh no it doesn't
19:53:57archivatorkugel: oh, well, I'll figure it out tomorrow, then, have to go now.
19:57:54 Quit hebz0rl (Read error: Connection reset by peer)
20:02:26 Join jgarvey [0] (
20:02:33 Join panni_ [0] (
20:06:44 Quit kugel (Quit: exit(0);)
20:06:54 Join kugel [0] (~kugel@rockbox/developer/kugel)
20:12:12*kugel sees Wincent Balin has appeared on the wiki
20:12:18***Saving seen data "./dancer.seen"
20:13:01 Join Sjano [0] (
20:13:50SjanoIs there any way to use .cue files in rockbox?
20:14:46SjanoSo that we could rip cd's into one large file instead of one for each song?
20:14:49 Join mischasworld [0] (
20:15:51kugelSjano: yes
20:19:20 Quit krazykit (Read error: Operation timed out)
20:21:15SjanoHow? I cant seems to get it to work. :S
20:21:40SjanoI have to leave now, but i'll read it when i come back ^^
20:21:57 Join merbanan [0] (
20:23:56kugelSjano: the manual should describe it
20:31:28 Join krazykit [0] (
20:34:37 Join phanboy4 [0] (
20:40:03 Nick YPSY is now known as Ypsy (
20:40:42kugelso, I would still appreciate comments on my "issue"
20:41:48 Quit komputes (Remote host closed the connection)
20:42:34 Quit panni_ (Quit: ( :: NoNameScript 3.81 :: ))
20:42:47 Join panni_ [0] (
20:49:03 Join MethoS- [0] (~clemens@
20:50:50 Quit Strife89 (Quit: Reboot to test.)
20:52:51 Nick Ypsy is now known as YPSY (
20:53:53 Join halmi [0] (
20:55:09 Quit halmi (Client Quit)
20:55:25 Join halmi [0] (
20:56:40 Join Strife89 [0] (
20:58:05 Quit jgarvey (Quit: Leaving)
20:58:31 Quit Zarggg (Quit: Zarggg)
20:59:56 Join Xerion_ [0] (
21:00:14kugelgevaerts: I think firmware/target/hosted would be a bit better, since some not-so-much plattform dependent files could be under hosted/
21:01:25gevaertskugel: same as the other target trees basically. firmware/target/hosted for all non-raw-hardware stuff, and then firmware/target/hosted/sdl, firmware/target/hosted/android, firmware/target/hosted/toaster,...
21:02:04 Quit Xerion (Ping timeout: 245 seconds)
21:02:04 Nick Xerion_ is now known as Xerion (
21:04:02kugelwe use newlib ctype, no wonder it's incompatible
21:09:52kugelwe probably don't use most if common/ in the sim even though it's compiled :/
21:10:49archivatorIf there are no objections to FS #11219 and FS #11229, I don't mind seeing them committed :)
21:13:51kugelI had this idea before. could we split up firmware/common and firmware/include so that standard c library functions are isolated in a folder?
21:15:07kugelthe mixup creates if you want to use the systems library but still need rb specific stuff from there (a problem I also noticed with uchida's database tool patch)
21:15:48kugelcreates problems*
21:16:13gevaertsThat might be a good idea
21:17:26kugelI just don't know what belongs to the c library and what not :P but a few hours of research would solve that
21:18:13linuxstbIt looks like _most_ of that stuff is the C library, but a few other things have sneaked in.
21:18:25 Quit FlynDice (Remote host closed the connection)
21:19:01 Quit Xerion (Ping timeout: 276 seconds)
21:19:35kugelI would propose to move c library stuff into a firmware/libc, and leave the corresponding headers in include; and leave the other files in common but move the corresponding headers into export
21:20:31 Join FlynDice [0] (
21:20:33kugelis that ok?
21:21:24kugelit might actually be required for RaaA
21:21:26linuxstbWhat about having a firmware/libc/include ? That would be more obvious than a firmware/include referring to firmware/libc
21:21:59kugeland .c files directly under firmware/libc=
21:22:11 Quit FOAD (Quit: I'll be back)
21:22:13 Join M3DLG [0] (~M3DLG@
21:22:17kugelthat sounds ok also
21:22:25 Join halmi_ [0] (
21:22:57 Join Zargold [0] (
21:23:57 Join Zarggg [0] (~zarggg@2001:0:4137:9e74:0:fbf3:beb1:ba3d)
21:24:51 Quit halmi (Ping timeout: 265 seconds)
21:27:01linuxstbkugel: Have you thought about dircache?
21:27:15kugelin what way?
21:29:04*kugel doesn't remember why he should think about dircache
21:29:21 Quit flydutch (Quit: /* empty */)
21:30:06 Quit M3DLG (Ping timeout: 260 seconds)
21:31:31 Part Zargold
21:32:25kugellinuxstb: can you elaborate?
21:32:32linuxstbkugel: About if/how to handle it in RaaA. Looking at the code in common reminds me that it's intertwined with our file I/O functions. But I guess it works in the sim already?
21:33:01kugelnot really and yes
21:33:45kugelnot all targets have dircache (they don't define it at compile time), so there's a completely dircache free way possible
21:34:24kugeldircache is not something we'd want I think. if there's something worth to cache I'd expect the host to do it
21:35:11 Join FOAD [0] (~dok@
21:35:28linuxstbBut aren't some database features still dependent on dircache being present (or work much better with dircache) ?
21:35:53kugelthe database should work without dircache
21:36:18linuxstb"work", or "work well" ?
21:36:20 Join halmi [0] (
21:36:41kugelhow well it works depends on the platform I guess. I expect it would work just as well if the host caches things
21:37:32kugelbut I think optimizing the database for best possible performance is a bit out of scope, that's something that can be done later IMO
21:38:36 Quit halmi_ (Ping timeout: 240 seconds)
21:39:06linuxstbI'm just saying that you may want to consider keeping the possibility of dircache as an option in RaaA.
21:39:47kugeldircache doesn't make sense if we're a music player only
21:40:29kugelit would cache all files, RaaA wouldn't touch 99% of them
21:40:51archivatorDoes Rockbox have *any* support for the standard tell() and seek() file ops?
21:41:42kugeldircache is something you want if you're an OS, not if you're a media player app
21:42:15kugelbut I don't plan to make it explicitely impossible to use with RaaA so it might just work in the end
21:43:01linuxstbarchivator: lseek should be there IIRC.
21:43:30archivatorlinuxstb: but no tell? That's weird.
21:44:06archivatormy bad, I read up on lseek
21:44:10linuxstbarchivator: Isn't ftell just a stream function?
21:44:56 Quit Llorean (Quit: Leaving.)
21:52:20 Join halmi_ [0] (
21:54:33 Quit halmi (Ping timeout: 246 seconds)
21:58:05 Quit halmi_ (Quit: halmi_)
21:58:21 Quit Luca_S (Quit: CGI:IRC (EOF))
22:00:52 Join Llorean [0] (
22:02:34archivatorRight, for the record: flite compiles fine for the sim but needs some extra work in file-handling and math to compile for ARM.
22:02:52archivatorAlso, do we have an strstr implementation hidden somewhere?
22:03:28gevaertsarchivator: firmware/common/strstr.c ?
22:03:59archivatorgevaerts: gah, I was only looking at plugin.h..
22:04:10archivatorThanks, will have to expose it to the plugins, though.
22:05:16 Join Strife1989 [0] (
22:05:25linuxstbarchivator: Rockbox tries to be POSIX-compliant wherever possible. i.e. if there's a standard function to do something, we'll use it.
22:06:52 Quit Strife89|PalmTX (Ping timeout: 276 seconds)
22:07:36 Quit Strife89 (Ping timeout: 240 seconds)
22:08:52 Quit bluebrother (Disconnected by services)
22:08:55 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
22:12:22***Saving seen data "./dancer.seen"
22:13:27kugelso, timefuncs.c contains libc and non-libc functions. I would propose to move mktime out to a seperate file and move timefuncs somewhere lese
22:14:26archivatorCurrent status of linking flite for ARM: "region PLUGIN_RAM is full", plus a bunch of unresolved references. Not good but I'm leaving it at that for tonight.
22:15:01linuxstbarchivator: How big are you expecting it to be (code+data+BSS) ?
22:15:36archivatorlinuxstb: it's about 3MB on x86_64..
22:15:40Bagderkugel: you mean to better allow an actual libc function to be used instead?
22:16:04 Quit Strife1989 (Ping timeout: 265 seconds)
22:16:20linuxstbkugel: What about trying to make timefuncs more standard? That would help when trying to integrate with a host.
22:16:49kugelBagder: yes, that too
22:17:01Bagderwhat other reason is there+
22:17:12 Join Blue_Dude [0] (~chatzilla@rockbox/developer/Blue-Dude)
22:17:21BagderI mean, why does it matter if libc and non-libc functions are mixed?
22:17:24kugellinuxstb: I just would like to clean it up for now
22:17:48Bagderin rockbox, we don't have a libc...
22:17:54kugelBagder: we talked about moving standard C stuff into a seperate directory
22:18:13Bagderbut why
22:18:15 Join Strife89 [0] (
22:18:33kugelthen have an include dir in that libc folder
22:18:35gevaertsBagder: there's this guy somewhere in Sweden who wrote on his blog that rockbox should be an application!
22:18:36linuxstbBagder: To use the host's libc in RaaA
22:18:43kugelbecause our replacements header class with the system's ones
22:18:53Bagderexactly my point, my kugel said there's another reason too
22:19:01Bagderso I was curious what it was
22:19:15Bagderdarnit I type bad
22:19:16linuxstbBagder: Your kugel?
22:19:24Bagderhaha yeah, oh I suck
22:19:57kugelI think a clear separation is a good idea nevertheless
22:20:01Bagders/, my/, but/
22:20:36Bagderkugel: but if there was no raaa, there is no "separation" - just functions rockbox provides
22:20:55 Quit JohannesSM64 (Ping timeout: 265 seconds)
22:21:26Bagderreally not worth discussing though ;-)
22:22:23 Quit Llorean (Quit: Leaving.)
22:22:28kugelhere's the list I collected now (the files in firmware/common with there according "standard")
22:22:47kugelI need to go now, it would be nice if someone could look over it and confirm/correct
22:23:31 Quit kugel (Remote host closed the connection)
22:25:46Bagderkugel: strcasecmp is rather libc I would say
22:28:18 Join mt [0] (~mtee@rockbox/developer/mt)
22:28:49gevaertsBagder: technically it's not though
22:29:01Bagderyou mean originally?
22:29:15Bagderwhich libc doesn't have that these days?
22:29:34Bagderor you mean it isn't in the C standard?
22:29:42gevaertsthe latter
22:29:57gevaertsAll non-rb things in that list are in
22:30:12Bagderright, if that what the list means then I agree
22:30:29 Quit mischasworld (Ping timeout: 264 seconds)
22:30:45Bagderthe gnuc one is just not in "regular" libcs
22:31:15gevaertsI wouldn't split things too much though
22:31:53Bagderno, there's no point in doing that. We just need to possibly provide a strcasestr() replacement when we run raaa on a system without it in its libc
22:32:02 Quit adnyxo (Ping timeout: 264 seconds)
22:32:51Bagderit seems to exist on android...
22:35:03 Join JohannesSM64 [0] (
22:36:28mtsaratoga, linuxstb : Wouldn't it be better if all the container-specific stuff in apps/codecs/wma.c were factored out to some independent asf parsing library ? (sort of like librm and the different real audio codecs)
22:37:02Blue_DudeNew task at FS #11232: software mixer. It's not done yet but I wanted to take a snapshot of where I'm at to solicit comments. Thanks!
22:37:30 Quit Tux2 (Quit: Leaving)
22:38:24linuxstbmt: Yes, that would make sense. But there was no need when there was only one asf codec.
22:38:45mtYes, I assumed that was the case
22:43:36 Join Necos [0] (
22:46:41Necosi've been looking in the bug tracker for this bug, but haven't found it. if i go into a menu on my e260, and scroll past the "edges" of the database listing, all the text on the screen disappears
22:47:51 Join adnyxo [0] (
22:51:01Necosso, basically, if i scroll up past A or down past Z, everything on screen disappears, but the theme's background is still there
22:52:42mtBlue_Dude: Is this patch to have voice and playback running together ?
22:53:08 Join Llorean [0] (
22:53:25 Quit Llorean (Changing host)
22:53:25 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
22:54:04Blue_DudeYeah, that's the plan. Voice and playback currently plays together fine, but bad things happen when you pause.
22:54:06 Join Llorean1 [0] (
22:54:28 Quit Llorean (Disconnected by services)
22:54:31 Nick Llorean1 is now known as Llorean (
22:54:34 Quit Llorean (Changing host)
22:54:34 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
22:56:23 Quit JohannesSM64 (Quit: WeeChat 0.3.2-dev)
23:00:21 Quit DerPapst (Quit: Leaving.)
23:12:27 Quit Blue_Dude (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
23:16:20 Part Boldfilter
23:18:23 Join M3DLG [0] (
23:24:09 Join mischasworld [0] (
23:36:03 Quit evilnick_B (Quit: Page closed)
23:41:47 Nick fxb is now known as fxb__ (
23:43:40 Quit MethoS- (Remote host closed the connection)
23:50:58 Quit mischasworld (Remote host closed the connection)
23:52:31 Quit Llorean (Quit: Leaving.)
23:56:00 Quit DataGhost (Ping timeout: 246 seconds)

Previous day | Next day