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 2011-04-12

00:02:50bluebrotheryay, now it looks really good :)
00:03:11 Quit Barahir_ (Read error: Operation timed out)
00:03:19 Quit domonoky (Read error: Connection reset by peer)
00:04:34bertrikshould we write "customize" or "customise" in the user manual?
00:04:43***Saving seen data "./dancer.seen"
00:04:43 Join Barahir [0] (
00:04:44 Quit petur (Quit: Leaving)
00:04:58AlexPbertrik: Customise for me
00:05:06AlexPBut this will be another fun argument
00:05:17bertrikdo we have a tag for that?
00:08:47bertriknever mind, it was a stupid attempt at a joke, to have a tag to switch between -ize and -ise
00:08:54AlexPah :)
00:09:07AlexPAnyway, I'm off to bed, can argue later :)
00:09:52bluebrotherbertrik: changed :)
00:09:56CIA-87New commit by bluebrother (r29705): Allow URLs to break at hyphens and allow more space at breaking characters. ...
00:10:04 Quit funman (Read error: Operation timed out)
00:13:44 Quit bieber (Ping timeout: 264 seconds)
00:14:02 Join bieber [0] (
00:14:31CIA-87r29705 build result: All green
00:14:36 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
00:21:56 Quit esperegu (Read error: Connection reset by peer)
00:24:09 Quit ender` (Quit: What if there were no hypothetical questions?)
00:24:21 Quit Judas_PhD (Quit: This is a quitting message)
00:25:54 Join ender| [0] (
00:31:16 Join sideral [0] (~sideral@
00:31:16 Quit sideral (Changing host)
00:31:16 Join sideral [0] (~sideral@rockbox/developer/sideral)
00:31:16 Quit pamaury (Remote host closed the connection)
00:35:42 Quit amee2k (Read error: Operation timed out)
00:37:23 Join amee2k [0] (
00:37:27 Join scromple [0] (
00:38:36 Quit [Saint] (Ping timeout: 246 seconds)
00:40:53 Quit efyx (Remote host closed the connection)
00:42:17CIA-87New commit by bertrik (r29706): Fix spelling of "appearance" in the user manual.
00:43:53 Join [Saint] [0] (
00:44:14 Quit t0rc (Quit: Give someone code, help them with one project. Teach someone to code, help them rule the world.)
00:45:51 Quit sideral (Quit: Leaving.)
00:45:51 Quit n1s (Read error: Connection reset by peer)
00:45:59CIA-87r29706 build result: All green
00:46:25 Quit Keripo (Quit: Leaving.)
00:46:56 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
00:50:17 Join MethoS- [0] (~clemens@
00:50:27 Quit bertrik (Quit: :tiuQ)
00:51:59 Quit kevku (Quit: KVIrc 4.0.4 Insomnia
00:53:08 Quit ChickeNES (Quit: Computer has gone to sleep.)
00:53:31 Quit krazykit (Read error: No route to host)
00:53:46 Quit bieber (Ping timeout: 276 seconds)
00:54:05 Join bieber [0] (
00:55:29 Join krazykit [0] (
01:05:41 Join lennyk [0] (
01:10:47 Join webguest00 [0] (
01:11:23webguest00Hello my sansa clip 2gig v01.01.35A doesent show up in the selection menue for rockbox
01:13:36 Quit robin0800 (Ping timeout: 246 seconds)
01:13:47webguest00hello? anyone there?
01:13:53 Quit krazykit (Quit: awe yeeeeeee)
01:14:12 Join bluebroth3r [0] (
01:14:16 Quit bluebroth3r (Changing host)
01:14:16 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
01:14:16 Quit bluebrother (Disconnected by services)
01:14:30webguest00Can some one help meplease?
01:16:46 Join krazykit [0] (~krazykit@
01:17:16webguest00hello!!! I would like some assistence please
01:22:09 Join kevin__ [0] (
01:22:38kevin__Hello can i get some assistance please?
01:23:39kevin__I need some help with rockbox that isnt covered in the faq
01:27:04krazykityou can if you ask a question
01:28:02kevin__Finnally. okay so my sansa 2gig v1, dosent show up in the list of drives i can install rockbox to
01:28:41krazykityour device is probably in MTP mode. it needs to be in MSC mode
01:29:00kevin__How do I do that?
01:30:07krazykitthat's covered in the installation section of the manual
01:30:13krazykiti don't recall off-hand
01:32:48 Quit bieber (Ping timeout: 248 seconds)
01:33:30 Join bieber [0] (~quassel@
01:40:49 Join vlam [0] (
01:40:55vlamCan someone help me out here
01:42:23vlamI'm trying to uninstall RockBox
01:42:29vlamSpecifically t he bootloader
01:43:45 Part vlam
01:44:44kevin__Hey thanks guys
01:47:25 Quit leavittx (Ping timeout: 240 seconds)
01:50:50kevin__I don't think it's on there anymore anyways
01:52:43 Join Keripo [0] (
01:53:18 Quit kevin__ (Quit: Leaving)
01:55:24 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
02:03:53 Quit T44 (Quit: Leaving)
02:04:03 Quit MethoS- (Remote host closed the connection)
02:04:14 Join G4Oblivion [0] (
02:04:38 Join Topy44 [0] (
02:04:44***Saving seen data "./dancer.seen"
02:06:55 Quit G4Oblivion (Client Quit)
02:08:20 Quit webguest00 (Quit: CGI:IRC)
03:08:03 Quit lennyk (Quit: lennyk)
03:21:16 Quit amiconn (Disconnected by services)
03:21:17 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
03:21:34 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
03:22:17 Quit pixelma (Disconnected by services)
03:22:20 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
03:22:22 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
03:23:03 Quit ps-auxw (Ping timeout: 260 seconds)
03:23:20 Join ps-auxw [0] (~arneb@2001:470:c807:0:1532:4e5f:2ad3:4123)
03:24:41 Quit mystica555 (Remote host closed the connection)
03:27:39 Join michaeltwofish [0] (
03:29:02 Join Judas_PhD [0] (
03:33:33 Join mikelietz [0] (
03:33:38 Part mikelietz
03:38:50 Quit michaeltwofish (Quit: leaving)
03:54:45 Join ChickeNES [0] (~ChickeNES@
03:56:03 Quit Keripo (Quit: Leaving.)
03:57:06 Join CaptainKwel [0] (
03:59:30 Join Keripo [0] (
04:00:32 Join T44 [0] (
04:03:53 Quit Topy44 (Ping timeout: 240 seconds)
04:04:46***Saving seen data "./dancer.seen"
04:08:53 Join kugel_ [0] (~kugel@rockbox/developer/kugel)
04:11:03 Join mystica555 [0] (
04:11:53 Quit kugel (Ping timeout: 246 seconds)
04:14:18 Quit Judas_PhD (Quit: This is a quitting message)
04:16:43 Quit FoH (Ping timeout: 246 seconds)
04:17:15 Join FoH [0] (
04:20:48 Quit DerPapst (Quit: Leaving.)
04:29:10 Quit BHSPitMonkey (Remote host closed the connection)
04:29:38 Join DX3 [0] (~Dre@
04:32:32 Quit Dreamxtreme (Ping timeout: 246 seconds)
04:33:51 Quit CaptainKwel (Quit: Ex-Chat)
04:38:57 Join byteframe [0] (~byteframe@unaffiliated/byteframe)
04:39:08byteframerockbox on my new sansa clip is secksee howt.
04:41:02 Part byteframe ("Leaving.")
04:45:23 Join wtachi [0] (
04:47:29 Quit T44 (Ping timeout: 240 seconds)
04:48:57 Quit pixelma (Disconnected by services)
04:48:59 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:49:01 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:49:58 Quit amiconn (Disconnected by services)
04:49:59 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:50:02 Join Topy44 [0] (
04:50:12 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
05:00:33 Quit TheSeven (Ping timeout: 260 seconds)
05:01:01 Quit bieber (Ping timeout: 276 seconds)
05:04:05 Join bieber [0] (
05:04:16 Quit [Saint] (Quit: I'm only going to Heaven if it feels like Hell, I'm only going to Heaven if it tastes like caramel...)
05:04:56 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
05:08:09 Quit bieber (Ping timeout: 246 seconds)
05:08:30 Join bieber [0] (
05:09:17 Join linuxguy3 [0] (
05:09:52 Quit JdGordon| (Quit: leaving)
05:13:25 Join Rob2222 [0] (
05:15:51 Quit bieber (Ping timeout: 246 seconds)
05:16:11 Join bieber [0] (
05:16:39 Quit Darkknight512 (Quit: CGI:IRC)
05:17:19 Quit Rob2223 (Ping timeout: 246 seconds)
05:28:32 Join [Saint] [0] (
05:32:13 Quit [Saint] (Client Quit)
05:39:25 Quit Horscht (Quit: Verlassend)
05:40:05 Join byteframe [0] (~byteframe@unaffiliated/byteframe)
05:40:36byteframemy sansa clip+ boots into original firmware upon usb removeal. if I reboot, rockbox loads. is there something I can do?
05:42:43DrDnarSure, you can listen to music.
05:43:04DrDnarIs there something you think is wrong?
05:43:43byteframeI quess. If It could, shouldn't it boot into rockbox?
05:44:09wtachibyteframe: Rockbox doesn't have 100% complete USB support for the clips
05:44:23wtachiit has to use the OF
05:44:32byteframewtachi, I read, but I read/write from a linux host, so should I be ok?
05:45:10DrDnarI have a Sansa e200. It doesn't need to reboot after using USB, but that's because Rockbox's USB support is completed for that model.
05:45:27byteframeWhat functionality is it missing?
05:45:56wtachibyteframe: IIRC, it sometimes connects to Windows at a low speed
05:46:15wtachiyou can compile your own version with USB enabled, and try it
05:46:25byteframeThat's fine. I don't use windows.
05:46:27DrDnarHuh, sounds like a descriptors problem.
05:46:40wtachiit's been a known problem for a while
05:46:40byteframeSo far I've only had it crash twice (I think both going 'back' while playing with the database).
05:46:57wtachiyeah, the database is flaky because the developers don't use it
05:47:06byteframeI can deal with the reboot after no usb, are there anyother tips?
05:47:20DrDnarThe host operating system shouldn't matter.
05:47:59byteframeIs the database flaky? I easily do the files (or just get my playlists on there) and avoid all that battery drain and such.
05:48:00DrDnarRockbox probably just simulates a generic USB MSD.
05:48:30byteframeDrDnar, I had to change it from MTP to MSC (UMS?) after opening the box.
05:48:35wtachibyteframe: it's up to you, if you're having problems
05:48:43wtachibyteframe: that's in the OF; Rockbox is MSC only
05:48:57byteframeDoes MSC == UMS?
05:49:25byteframeok yes. thanks.
05:52:51byteframeCan I hide something (Database or FM Radio) from the main menu?
05:54:04wtachibyteframe: not without modifying the code yourself
06:04:49***Saving seen data "./dancer.seen"
06:08:22 Quit Keripo (Quit: Leaving.)
06:32:00 Quit krazykit (Ping timeout: 276 seconds)
07:07:42 Quit fyre^OS (Ping timeout: 260 seconds)
07:10:15 Join fyrestorm [0] (
07:19:44 Join Keripo [0] (
07:26:10 Quit fyrestorm (Ping timeout: 240 seconds)
07:31:04 Join Judas_PhD [0] (
07:32:32 Join DerPapst [0] (
07:35:22 Join Buschel [0] (
07:38:44 Quit factor (Read error: Connection reset by peer)
07:44:16 Join Keripo1 [0] (
07:48:03 Quit Keripo (Ping timeout: 276 seconds)
07:51:09 Quit TheSeven (Read error: Connection reset by peer)
07:52:10 Quit bieber (Ping timeout: 240 seconds)
07:52:29 Join krazykit [0] (
07:53:23 Join bieber [0] (
07:56:13 Join [Saint] [0] (~St.]
07:56:42 Join factor [0] (~factor@
08:04:52***Saving seen data "./dancer.seen"
08:09:35 Join ender` [0] (
08:10:43 Quit ender` (Read error: Connection reset by peer)
08:10:57 Join ender` [0] (
08:11:51 Quit Buschel (Quit: ChatZilla [Firefox 3.6.16/20110319135224])
08:15:58 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
08:18:46 Join WhoisMaple [0] (
08:20:51WhoisMapleCan someone help me quick, i have to leave for work. My while playing screen is stuck as the failsafe. No matter which theme i choose it remains in the failsafe wps. I am running a sansa fuzev1 latest stable build. Any way to remedy this w/o a full wipe and install?
08:24:08saratogausually means the theme you selected didn't parse correctly
08:24:37saratogathe default should always work unless you went into the rockbox folder and deleted it
08:25:14 Quit DerPapst (Quit: Leaving.)
08:25:16 Quit wtachi (Quit: Quit: Quit: Quit: Quit:)
08:25:29WhoisMapleAll of my themes do the same thing even the default.
08:26:13saratogasomething wrong with your install then
08:26:24saratogacheck the disk for errors, etc
08:27:01WhoisMapleit was literally working fine a minute ago :( jsut my luck
08:29:48 Quit Judas_PhD (Quit: This is a quitting message)
08:32:53WhoisMapleSo i got the default theme working now, How would i go about getting the other themes to work? delete and reinstall em?
08:37:21 Join efyx [0] (
08:39:14 Join mudd1 [0] (
08:39:20 Join B4gder [0] (
08:39:20 Quit B4gder (Changing host)
08:39:20 Join B4gder [0] (~danielx@rockbox/developer/bagder)
08:41:49 Quit bieber (Ping timeout: 276 seconds)
08:42:07 Join bieber [0] (
08:45:17 Quit WhoisMaple (Quit: Living is easy with eyes closed)
08:47:15 Join leavittx [0] (~leavittx@
08:50:39 Join DerPapst [0] (
08:51:06 Quit DerPapst (Client Quit)
08:51:24 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
08:54:34 Join Judas_PhD [0] (
08:55:33 Join DerPapst [0] (
08:57:29 Quit DerPapst (Client Quit)
09:00:03 Join LinusN [0] (~linus@rockbox/developer/LinusN)
09:10:01 Join T44 [0] (
09:12:27 Quit Topy44 (Ping timeout: 246 seconds)
09:20:38 Part scromple ("Leaving")
09:27:19 Quit leavittx (Ping timeout: 276 seconds)
09:30:05 Quit sasquatch (Quit: WeeChat 0.3.2)
09:30:29 Join sasquatch [0] (
09:31:51 Quit linuxstb (Read error: Connection reset by peer)
09:33:10 Quit bertrik (Ping timeout: 276 seconds)
09:36:45 Join swilde [0] (
09:39:19 Join linuxstb [0] (~linuxstb@rockbox/developer/linuxstb)
09:46:14 Join einhirn [0] (
09:51:13 Quit Kuitsi (Remote host closed the connection)
09:56:27 Quit Keripo1 (Read error: Connection reset by peer)
09:59:30 Join Keripo [0] (
10:04:54***Saving seen data "./dancer.seen"
10:11:37 Nick kugel_ is now known as kugel (~kugel@rockbox/developer/kugel)
10:12:10 Quit sasquatch (Ping timeout: 240 seconds)
10:12:44 Join sasquatch [0] (
10:16:20 Quit otih (Remote host closed the connection)
10:19:16 Join mt [0] (~mtee@
10:19:18 Quit mt (Changing host)
10:19:19 Join mt [0] (~mtee@rockbox/developer/mt)
10:37:52 Join pamaury [0] (81680b01@rockbox/developer/pamaury)
11:07:07 Quit liar (Read error: Connection reset by peer)
11:07:59 Join liar [0] (
11:15:50 Join Hindu [0] (~Hindu@
11:19:06 Quit bieber (Ping timeout: 260 seconds)
11:19:26 Join bieber [0] (
11:32:14 Join DerPapst [0] (
11:32:27HinduWhat about our GSOC proposals ??
11:33:46AlexPHindu: We are evaluating them
11:34:10AlexPIf we need more detail/interviews/whatever, we'll ask - until then patience is the key
11:36:53HinduWell , today is the fourth day after proposals....:)
11:37:26gevaertsAh, nice of you to keep track of the numbers for us!
11:38:07AlexPHindu: We know
11:38:29AlexPHindu: We all also have real lives, and there are many developers we need to co-ordinate
11:38:30HinduWell , can you give me a rough idea about the no. of proposals to Rockbox ? Mentors can tell that I suppose
11:38:42AlexPHindu: Sorry, not for the moment, no
11:39:20tguinotWait & see ;]
11:39:23HinduSo ?
11:39:45 Join honkytonkwillie [0] (
11:39:52 Quit mudd1 (Ping timeout: 240 seconds)
11:40:09AlexPHindu: You aren't going to achieve anything positive by continually asking
11:40:25AlexPWe've said that we are evaluating and that we will ask if we need anything
11:40:32AlexPUntil then please just be patient
11:40:33HinduSo what should I do?
11:40:58tguinotWait, begin to read doc, etc...
11:41:14AlexPGet involved in other areas, read the source etc.
11:41:24AlexPAlthough really that should have been done before now :)
11:41:48HinduAllright !!!
11:43:21Hinduyou guys only mention on the website that we should be interacting with the mentors , Be in touch , So I just asked about the applications......
11:44:23AlexPSure, but interacting doesn't mean keeping asking the same question when you have been told we will ask if needed and to be patient
11:44:45AlexPinteracting means getting involved in the community, familiarising yourself with the code, things like that
11:44:55HinduAllright .... :)
11:46:16 Quit honkytonkwillie (Quit: Leaving.)
11:47:09tguinotHindu: I think you can start to work, even if you are not selected, It will help the project
11:48:02 Part Zagor ("Leaving")
11:48:08 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
11:52:53 Join Hindu_Away [0] (~Hindu_Awa@
11:54:09 Quit ChickeNES (Read error: Connection reset by peer)
11:54:29 Join ChickeNES [0] (~ChickeNES@
11:54:39 Quit Hindu (Ping timeout: 240 seconds)
11:55:03 Nick Hindu_Away is now known as Hindu (~Hindu_Awa@
12:00:15 Quit bluefoxx (Ping timeout: 260 seconds)
12:04:32 Join bluefoxx [0] (
12:04:58***Saving seen data "./dancer.seen"
12:07:39 Join wodz|work\ [0] (
12:07:57 Join wodz|work [0] (
12:08:03 Quit wodz|work\ (Client Quit)
12:09:17wodz|workabout GSOC proposals - only thouse who signed up as mentor read it or any rb developer can help evaluating?
12:10:12gevaertsYou need to be signed up as a mentor to see them
12:10:25gevaertsBut any developer can sign up as a mentor
12:10:58wodz|workok, maybe next time
12:11:22gevaertsYou can still sign up
12:12:43wodz|workgevaerts: sorry for my ignorance - what is the procedure?
12:13:30gevaertsGo to and look for the "Mentors: Apply now" bit
12:13:52gevaertsDon't ask me more details, I signed up before the new user interface was launched :)
12:20:24 Quit Hindu (Ping timeout: 252 seconds)
12:26:25 Quit wodz|work (Quit: CGI:IRC)
12:37:12 Join dfkt [0] (~dfkt@unaffiliated/dfkt)
12:51:43 Join wodz|work [0] (
12:56:36 Quit Keripo (Ping timeout: 276 seconds)
13:01:20 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
13:10:48B4gderwodz|work: #rockbox-gsoc is the channel for the gsoc talk
13:19:02 Quit wodz|work (Quit: CGI:IRC)
13:20:38 Quit krazykit (Ping timeout: 258 seconds)
13:22:22 Join krazykit [0] (
13:32:08 Quit amee2k (Ping timeout: 258 seconds)
13:32:34 Join amee2k [0] (
13:57:02 Join MethoS- [0] (~clemens@
14:01:53 Quit sinthetek (Read error: Operation timed out)
14:03:15 Quit jhMikeS (Ping timeout: 248 seconds)
14:05:02***Saving seen data "./dancer.seen"
14:12:06 Join smk [0] (~smk@
14:16:27 Quit smk (Quit: Leaving)
14:19:35 Join smk [0] (~smk@
14:24:44 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
14:31:17 Quit Xerion (Ping timeout: 246 seconds)
14:42:01smkhi. a very fundamental question. why did rockbox choose to have it's own kernel? why not run it as an app on existing kernel of the device?. is it because running as app would make the apps device dependent?
14:42:24B4gdersmk: because the targets we started rockbox on were far too small to run linux
14:43:01kugelmore fun, too :)
14:44:15smkyeah ofcourse. I meant why not use the kernel that must be already running on a device. Currently, rockbox replaces the original kernel with it's own.
14:44:46smkdid the earlier targets not have an OS on them?
14:44:54Zagormost devices don't
14:45:18TorneDo *any* of the devices we run on, other than the ones we run as an app on, have any kind of OS?
14:45:22Zagoronly the very latest like ipod touch and android devices have an operating system that runs apps
14:46:05B4gdersome of them run some RTOS
14:46:26B4gderbut we couldn't "run apps" on those anyway
14:46:59Tornesmk: most embedded devices don't *have* a kernel. They run a single binary on the bare metal, generally
14:47:00smkyeah. i thought every DAP must require resource management in some way. so all of them might be having OSes. i assumed so.
14:47:13B4gdersmk: malloc doesn't require an OS...
14:47:28gevaertsDefine "OS" :)
14:47:30TorneRockbox is also a single binary running on bare metal; our "kernel" is a logical component in the source, not a seperate actual thing that gets built
14:49:20smki would define OS as a resource manager. rockbox has thread to schedule and service. so i figured the firmware that originally comes with DAPs must be having a similar architecture, hence the need of a kernel
14:49:35TorneIf you define OS that way then yes, probably all of them ahve an OS
14:49:51B4gderbut they're all pretty custom and proprietary
14:49:51Tornebut the OS and the "application" are invariably combined into a single thing
14:49:53pamaurythe main point here is that nearly all devices have a single binary blob which contains everything (all the code)
14:50:04TorneThere's no actual separation in the produced firmware binary
14:50:31Tornethere is no requirement that there be any separate components/layers, or that it be possible to replace any part of it
14:50:42gevaertssmk: trying to link to the existing firmware would be (a) legally very dubious, (b) technically insane, and (c) no fun at all
14:51:42 Join TheLemonMan [0] (
14:51:52pamauryyeah, I had to do it once to isolate code and it was a nightmare, so (b) is a real stopper
14:52:04smkhmm okay. i agree with (c). I figured writing device specific app would be easier than writing device specific core or kernel (whatever it is called)..
14:52:37smknow i get it though.
14:52:59TorneNo, that would be about a million times harder
14:53:17TorneThe original firmwares of these devices are much more different than the actual hardware is ;)
14:53:50Tornethe hardware/software boundary is at least a clean point of separation where you are supposed to be able to replace things
14:53:55Torneand hardware is often documented
14:54:10Tornethe original firmwares have no such separations, and *never* any documentation
14:55:19smkright. when i learnt that rockbox has it's own kernel, i imagined a layered architecture like with any system..
14:55:30TorneIt's layered logically, yes
14:56:02Tornethat doesn't mean they're not all linked into one big blob at build time :)
14:57:00 Quit B4gder (Quit: Konversation terminated!)
14:57:05smkyeah that's true.
14:58:26smkso all apps,core , plugins come bundled. But since there is a logical layer, is it possible to load a rockbox-app onto the device after it has been installed on it?
14:58:38smklike we do for android
14:59:28Tornenot sure what you mean.
14:59:36Torne"core" is the app and the firmware layer, linked together
15:00:12Tornethe rockbox-as-app ports have a core which contains a lot of the same app code, but only a rather small set of firmware code which just stubs onto the underlying OS's interfaces
15:01:20pamauryon all devices, there is a binary whichs contains the "firmware" and "apps" code; all the rest (plugins and codecs) are loaded on demand and are in other files
15:01:33smkwhat i mean is can anyone write a rockbox-app (say a stopwatch with a stylish interface) , and install it on a device which already has the rockbox core?
15:02:25smkpamaury: yeah thanks. that's what i wanted to know :)
15:02:36pamauryyou can write it as a plugin and then run it from rockbox
15:02:40Torneyou're misusing "app" here
15:02:58Torneour plugins behave kind of like applications but they are not really distributable in any other way, and they do not replace the rockbox app code
15:03:16TorneDistributing them seperately from the core generally doesn't work, though in an ideal world it maybe would.
15:04:16Torneso yes, you can write a plugin which does that kind of thing, but you still have to compile the plugin seprately for every rockbox target
15:04:19Torneplugins are not portable
15:04:36Torneand even within the same target it will only work within a fairly narrow range of builds where we haven't changed anything in the plugin API
15:04:47Torneso, in practise the way to do this is contribute the code of the plugin to rockbox :)
15:04:53Tornethen it gets built witht eh rest, and included in the build.
15:05:17smkright. got the concept..
15:05:40Tornethe plugins are, as far as the compiler and linker are concerned, also part of the same binary
15:05:48Torneit just happens that they're a part that's kept in a separate file
15:10:25smkB4gder: malloc is a request to the OS for more memory. on receiving such a request , the OS returns a handle to the appropriate chunk, to the process that requested memory.
15:10:41smkso i think OS is required . i might be wrong though :P
15:10:59gevaertsmalloc() is usually not implemented in the kernel :)
15:11:19Torneno; sbrk() is the standard unixy name for the syscall that actually gets memory from the kernel
15:11:23*gevaerts is being pedantic
15:11:33Tornemalloc is a layer above that generally
15:12:52smktrue. but either ways, OS intervention is necessary.
15:17:52Tornemalloc just needs a big chunk of memory to carve up to satisfy requests.
15:18:19Torneon an OS with process isolation it generally starts with a very small piece, then calls sbrk as needed to ask the OS for more
15:18:35Torneon an embedded system the malloc heap might just be "all of the memory that's left that we aren't using for anything else"
15:18:42Torneso there's no need to ahve any "os" involvement
15:18:54Tornewhen there's no more room in the heap you just fail
15:19:00gevaertsAnd threads don't count either!
15:19:27Torneyah, you can do threading without OS help as well ;)
15:19:36gevaertsWell, we do in RaaAoA :)
15:19:37smkoh. so the allotment is already there. malloc just picks from the quota given to it.
15:19:48Tornemalloc works however you implement it to work
15:19:51Torneit's just a library function
15:20:11gevaerts(which we don't have)
15:20:47smkgevaerts: yeah, working on that ;)
15:21:29Tornehopefully a nice one :)
15:22:26 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
15:23:00smknot sure about threads though. This discussion asks me to revisit my OS concepts, which i should do :)
15:23:33Torneit comes down to "an OS is not a well defined thing" :)
15:23:44Torneit's whatever the person who said it means ;)
15:24:03smktrue. the boundary is blur.
15:27:20 Join mem_ [0] (
15:27:42pamauryTheLemonMan: I just had a look at your OpenBLTC repo on github, the protocol.txt *might* be incomplete: you should say the payload size is LE the second time. Also, I'm not sure about the transfer size, since you seem to send 0x40+1 per transfer and I send 1024+1
15:32:51 Quit rasher (Ping timeout: 248 seconds)
15:35:16TheLemonManpamaury: yeah i forgot to update the header explanation :). My usb traces were all using 0x41 chunks and yours too iirc
15:38:09 Join rasher [0] (~rasher@rockbox/developer/rasher)
15:41:36pamauryTheLemonMan: I doubt it, mine were sending 1025 bytes each time, I'm pretty sure
15:42:45TheLemonManpamaury: here's mine, As you can see it's using 0x41 chunks
15:42:48pamaurydo you know what was the maximum size of a ep0 transfer on your device
15:42:59TheLemonMan0x40 is the max size
15:43:32pamauryhum, I should check with mine, I pasted the trace, need to find the link on my history
15:47:32 Quit smk (Quit: Leaving)
15:48:56pamaurymine too has a 64 bytes ep0 control size (the maximum), so the maximum transfer size might be independent of the maximum packet size
15:58:16 Part LinusN
15:58:32TheLemonManhave you tried sending something to your player with my tool ?
15:59:32pamauryI'll try when I'll be home if you want
15:59:39pamauryhave you tried with mine ?
16:01:40TheLemonMandid you send me the code ?
16:03:42 Quit antil33t (Read error: Connection reset by peer)
16:03:52pamaurysend you the link: pamaury/pa-tools/blob/master/fuze%2B/recovery_hid/recovery_hid.c">
16:03:52 Join antil33t [0] (
16:05:05***Saving seen data "./dancer.seen"
16:08:48TheLemonManpamaury: fails at transfer step 0 with 1024, at step 2016 with 64
16:10:33pamauryok, so you device doesn't seem to support such a size
16:10:33TheLemonManerr, it's working with 64 but doesnt like the last interrupt transfer
16:10:54pamaurymaybe they have a slightly different protocol
16:11:22TheLemonManmaybe depends on the fuse config
16:16:15TheLemonManbrb in a few hours
16:16:24 Quit TheLemonMan (Quit: Destructor called)
16:16:37 Join pepsi_ [0] (~pepsi@unaffiliated/jbutera)
16:17:04 Quit JackWinter (Read error: Connection reset by peer)
16:19:58 Quit Barahir (Ping timeout: 252 seconds)
16:21:36 Quit pepsi_ (Remote host closed the connection)
16:21:37 Join Barahir [0] (
16:22:01 Join n1s [0] (~quassel@rockbox/developer/n1s)
16:27:40 Join JackWinter [0] (
16:32:10 Join pepsi_ [0] (~pepsi@unaffiliated/jbutera)
16:45:39 Quit pepsi_ (Ping timeout: 240 seconds)
16:46:35 Quit pamaury (Quit: Page closed)
16:47:31 Join pepsi_ [0] (~pepsi@unaffiliated/jbutera)
16:51:51 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
17:01:48 Join LambdaCalculus37 [0] (~LambdaCal@rockbox/staff/LambdaCalculus37)
17:12:21 Part Zagor
17:13:00 Join L-Strife89 [0] (~Strife89@
17:18:12 Join leavittx [0] (~leavittx@
17:25:18 Join Xerion [0] (
17:25:19 Join Biont [0] (
17:28:08 Join mishra [0] (
17:28:44mishrahello everyone
17:29:00mishrai read that #rockbox-gsoc is the channel for the gsoc talk
17:29:10mishrabut its invite only
17:29:20BiontHell. A friend of mine just brought her rockboxed clipv2 which has been working flawlessly for well over a year now. Now it refuses to play anything, and all folders appear empty. However, the stock firmware works just fine and all files are in place. We got an error message once, but I can't reproduce it yet:
17:29:58Biont+o ^^ Well, everything worked fine until yesterday. any idea? I already did an update, of course
17:31:20gevaertsmishra: it's the channel for gsoc mentor talk to be specific
17:32:35mishraok thanks
17:32:58mishrawho should I catch to discuss about AAC ?
17:36:49 Quit pepsi_ (Ping timeout: 246 seconds)
17:36:50evilnick_BBiont: What is Show Files set to?
17:37:01evilnick_BSounds like it's accidentally been changed to Playlists
17:39:31 Join pamaury [0] (
17:39:31 Quit pamaury (Changing host)
17:39:31 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
17:40:56Biontah okay. that fixed the file browser :) But it still does not play
17:43:58 Join japc [0] (~japc@
17:44:07saratogamishra: i can talk to you about AAC, also probably Buschel if hes on since hes worked on it more then me
17:44:43 Quit TheSeven (Ping timeout: 276 seconds)
17:44:58Biontfast forward also works, but no sound/playback
17:48:46mishraOk, I have read about the basics of AAC now, so I want to ask if their are already some problem areas or bottlenecks identified in the current implementation?
17:49:58saratogaSBR is still pretty slow
17:50:10saratogaand the PS code is a mess as well, i don't think anyone has ever looked at it
17:50:36saratogathe code AAC-LC decoder is pretty well optimized, although not nearly so fast as WMA/Vorbis
17:51:04saratogaBuschel could tell you about SBR since he committed a lot of improvements over the summer
17:53:35 Join JesusFreak316 [0] (
17:55:01mishraOk will catch him sometime. Is there some documentation of the current implementation ?
17:56:33saratogaof what?
17:58:50 Join pepsi [0] (
17:58:50 Quit pepsi (Changing host)
17:58:50 Join pepsi [0] (~pepsi@unaffiliated/jbutera)
17:59:43mishraof the flow, like a block diagram etc
18:00:15 Join xmatrix [0] (xmatrix@
18:00:30saratogaah no, but generally people just generate those if they want them
18:01:10saratogaalthough maybe putting something up for libfaad would make sense since its such a fucking mess
18:01:35 Join skapazzo [0] (~skapazzo@
18:02:22xmatrixAnyone else using rockbox on the iPod classic and want to share their experiences? So far I've been using it for a few days and noticed that charging time takes roughly 6 hours (for 94% which it caps at for me) and battery life lasts roughly 9 hours.
18:03:15mishraI see, how should one approach to optimize something, lets say SBR, should we profile the code with say gprof and try to optimize the most used functions ?
18:03:44saratogayou won't be able to run gprof on rockbox, but it is possible to profile rockbox code, although its not always easy to do
18:04:09saratogastripwax has profiled a lot of the codecs, usually takes a lot of fiddling with the code and inlining to get meaningful results
18:04:30saratogai usually just skim the code, look for functions that are likely to be slow, and then comment them and benchmark without them
18:04:33 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
18:04:52saratogamost codecs aren't that large, so it doesn't take all that long
18:05:07***Saving seen data "./dancer.seen"
18:05:21saratogai posted a couple measurements here last summer:
18:05:39saratogadone with the 64kaache test file
18:09:55 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
18:11:54 Quit dionoea (Read error: Operation timed out)
18:11:59 Join dionoea [0] (
18:18:37mishraNice, I will try to put test_codec on my player and run some tests to start with. Thanks.
18:22:13evilnick_BBiont: What codec are the files that don't play? Is there anything unusual about the files? Large comment tags?
18:25:45 Quit L-Strife89 (Ping timeout: 258 seconds)
18:26:13BiontI have yet to find a file that will play at all. It really seems to affect playback as a whole. The files are mostly mp3 and maybe a couple of wma it seems
18:26:34saratogacheck the disk for errors
18:26:53 Join Sudos|i1520 [0] (~FLOORISLA@
18:28:49 Quit pepsi (Ping timeout: 258 seconds)
18:30:08 Quit Sudos|i1520 (Client Quit)
18:30:53 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
18:30:54BiontShould I rid the player files from comment tags just to make sure?
18:31:39 Quit krazykit (Quit: awe yeeeeeee)
18:35:15 Join u42p [0] (
18:35:21 Quit LambdaCalculus37 (Quit: switching to web)
18:35:39 Join LambdaCalculus37 [0] (~3f74f70d@rockbox/staff/LambdaCalculus37)
18:37:20 Join xmatrix_ [0] (xmatrix@
18:41:06 Quit xmatrix (Ping timeout: 248 seconds)
18:42:23 Quit xmatrix_ (Ping timeout: 276 seconds)
18:48:56 Quit user890104 ()
18:51:32Biontany specific utilities you want me to use? I just used the chkdsk from the explorer and it reported no errors.
18:51:32 Join pepsi_ [0] (~pepsi@unaffiliated/jbutera)
18:51:44Biontgoing to erase the comment tags now
18:53:24 Join fyrestorm [0] (
18:53:34 Join xmatrix [0] (xmatrix@
18:53:57 Quit Xerion (Read error: Operation timed out)
18:54:38 Quit u42p (Quit: Leaving)
18:54:42 Join bertrik [0] (
18:54:42 Quit bertrik (Changing host)
18:54:42 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
18:55:00 Join Xerion [0] (
18:56:35 Join u42p [0] (
18:57:50 Quit xmatrix (Ping timeout: 240 seconds)
18:58:55 Join kevku [0] (~kevku@2001:470:28:773:babe:feed:dead:bee)
19:00:31 Join wtachi [0] (
19:02:04 Join xmatrix [0] (xmatrix@
19:08:04 Quit t0rc (Quit: Give someone code, help them with one project. Teach someone to code, help them rule the world.)
19:11:17 Join L-Strife89 [0] (~Strife89@
19:13:23 Nick DX3 is now known as Dreamxtreme (~Dre@
19:15:32 Quit Biont (Quit: CGI:IRC)
19:16:07 Join Biont [0] (
19:16:12 Join ender [0] (
19:16:30 Quit ender` (Read error: Connection reset by peer)
19:16:30 Nick ender is now known as ender` (
19:21:52 Join TheLemonMan [0] (
19:23:10 Quit DerPapst (Ping timeout: 246 seconds)
19:24:40 Join DerPapst [0] (
19:26:35 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
19:27:06TheLemonManpamaury: youre working on a 3780, right ?
19:30:24 Quit LambdaCalculus37 (Quit: back to work!)
19:30:29 Quit tmzt (Ping timeout: 276 seconds)
19:35:35 Quit factor (Read error: Connection reset by peer)
19:36:20 Join tmzt [0] (~tmzt@
19:38:47 Join factor [0] (~factor@
19:41:01 Join Keripo [0] (
19:41:12 Join Horscht [0] (
19:41:12 Quit Horscht (Changing host)
19:41:12 Join Horscht [0] (~Horscht@xbmc/user/horscht)
19:44:51 Part mem_
19:54:24 Join ChickeNE_ [0] (~ChickeNES@
19:54:43 Quit japc (Ping timeout: 276 seconds)
19:56:07 Quit ChickeNES (Ping timeout: 246 seconds)
19:57:20 Join msh_lv [0] (
20:00:23 Quit xmatrix (Ping timeout: 276 seconds)
20:01:28 Quit dionoea (Changing host)
20:01:28 Join dionoea [0] (~dionoea@videolan/developer/dionoea)
20:02:55 Quit L-Strife89 (Quit: Restart.)
20:04:55 Quit msh_lv (Quit: reb00t)
20:05:08***Saving seen data "./dancer.seen"
20:06:09 Join L-Strife89 [0] (
20:08:55 Quit swilde (Remote host closed the connection)
20:16:37 Join msh_lv [0] (
20:17:48 Quit [Saint] (Quit: I'm only going to heaven, if it feels like Hell. I'm only going to heaven, if it tastes like caramel...)
20:22:14 Quit rasher (Ping timeout: 240 seconds)
20:23:53 Join [Saint] [0] (~St.]
20:25:23 Quit Keripo (Quit: Leaving.)
20:26:25 Join dfkt [0] (
20:26:25 Quit dfkt (Changing host)
20:26:25 Join dfkt [0] (~dfkt@unaffiliated/dfkt)
20:26:30 Join Keripo [0] (
20:31:38 Nick msh_lv is now known as mshathlonxp (
20:37:26 Join Buschel [0] (
20:37:53BiontAlright, after checking for errors on the clip and deleting all comment tags, I still get no playback and an error message like this:
20:38:16BiontIt also said "Codec failure" several times now. I didn't see that before
20:39:31Biontregular mp3, btw.
20:39:38Biontany ideas?
20:39:52 Join rasher [0] (~rasher@rockbox/developer/rasher)
20:40:42gevaertsBiont: did you check the filesystem?
20:40:53gevaertsAh, I see you did...
20:41:13gevaertsWhich rockbox version are you running?
20:41:48BiontI updated to the latest version before joining here to see if the problem persists
20:42:15 Join user890104 [0] (
20:42:16gevaertsSo r29706?
20:42:19Bionti.e. recent build
20:43:26Buschelgevaerts: do we have the map-files of those builds?
20:43:34gevaertsI don't think so
20:43:41Biontshe came here after running her rockboxed clip for more than a year so I figured an update would be the first thing to try, but it didn't help.
20:43:51 Join Darkknight512 [0] (
20:44:47Biontjust to stress again: the stock firmware runs fine
20:45:20saratogacould be an incomplete update
20:45:30saratogaif you have old codecs with a new binary that won't work
20:46:19Buschel"codec failure" could also be a sign of incomplete update...
20:46:26Buschelsaratoga: you were faster :)
20:46:55[Saint]kugel: were right about KEYTOOL
20:47:29BiontHow do I update properly if I missed something?
20:47:56 Quit L-Strife89 (Quit: Restart.)
20:48:46 Part u42p ("Leaving")
20:49:49[Saint]Biont: to update "properly"...just use RbUtil
20:51:04[Saint]Using Rockbox Utility guarantees that everything ends up where it's supposed to be.
20:51:26Biontokay. Fancy. I'll do that
20:56:28BiontFeels good to finally use this program at least once :)
20:59:32Bionthm...extraction froze at 91%
21:00:07 Quit JesusFreak316 (Ping timeout: 246 seconds)
21:01:03 Join promyloph [0] (~promy@unaffiliated/promyloph)
21:02:20Bionttrying again :D
21:03:03 Join L-Strife89 [0] (
21:05:09 Quit saratoga (Ping timeout: 252 seconds)
21:07:37Buschelhmm, I just built clipv2 r29706. 0x3004EA6C is the address of strcasecmp
21:08:35 Join sinthetek [0] (
21:08:35 Quit sinthetek (Changing host)
21:08:35 Join sinthetek [0] (~sinthetek@unaffiliated/sinthetek)
21:08:45 Join mudd1 [0] (
21:21:58BiontI don't know what just happened. Rbutil froze again after I tried to do a full reinstall...I feared I would brick it or something, but I unplugged and restarted it it works o_O
21:22:50 Quit pepsi_ (Ping timeout: 260 seconds)
21:23:22Biontis there any way to check if everything is in place and up to date? I don't feel too comfortable about giving it back just now
21:31:53evilnick_BBiont: Which player?
21:32:06Biontsansa clipv2
21:34:01*evilnick_B isn't totally sure, but on the players that he's used, the bootloader version can be found when powering on the player
21:34:23 Join ptrkmj [0] (b24923a2@gateway/web/freenode/ip.
21:34:39evilnick_BAnd for the main rockbox build, check System > Rockbox Info
21:36:24BiontBoot 2.0 / r29706
21:36:59 Join jhMikeS [0] (
21:36:59 Quit jhMikeS (Changing host)
21:36:59 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
21:38:13ptrkmjHi. I have a strange problem with my 5.5G iPod. Hard reset function (pressing Menu + Select button) is not working. Click wheel works fine. Has anyone experienced anything similar?
21:38:46BiontNot sure about the bootloader, but it said 29706 before and I was apparently missing codecs. 2 frozen and aborted installations later, it's working again. Can I really trust this?
21:40:20evilnick_BBiont: It's a difficult question to answer as we don't know what you've done to get to this stage.
21:41:00[Saint]ptrkmj: If that button combination is not working, then it is not a Rockbox issue.
21:41:17[Saint]that combination is handled by hardware, if it does not work, your player is broken.
21:41:19Torneptrkmj: turn the hold switch on and off first
21:41:21evilnick_Bi.e. you could extract the contents of a new build but tell the compression program to not overwrite files so you'd be left with a partial old build and a partial new build.
21:41:23ptrkmjWell, I know. I haven't even installed it :)
21:41:23Tornethen hold it for longer
21:42:28ptrkmjTorne: I tried everything from this page including that
21:42:56gevaertsptrkmj: is the ipod flat on a table?
21:43:07gevaertsThat helps a lot
21:43:09Tornewell so far basically everybody who has ever asked that question has in fact just been doing it wrong. maybe one or two people have turned out to have broken hardware.
21:43:12 Join saratoga [0] (9803c6dd@gateway/web/freenode/ip.
21:43:18TorneIt's never been anything else that I'm aware of :)
21:43:35 Quit saratoga (Changing host)
21:43:35 Join saratoga [0] (9803c6dd@rockbox/developer/saratoga)
21:43:47TorneMake sure you're holding just those two buttons, that you don't let go of either of them even a little, and you hold it for, like, 30+ seconds before giving up
21:43:53[Saint]that's correct. If it's not working, either you're messing it up or your player is busted.
21:47:19 Quit L-Strife89 (Quit: Dorm-bound.)
21:47:28ptrkmjI doubt I'm doing it wrong (I tried so many times). 6-7 seconds used to work. I'll try few more times, though. My iPod is working just fine (it's not 'frozen' at the moment).
21:48:28 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
21:49:06[Saint]ptrkmj: It's perfectly capable of "working fine" and still being broken.
21:49:10 Join krazykit [0] (
21:49:14[Saint]this is a hardware issue, not software.
21:49:23*gevaerts still thinks it's a finger issue
21:49:36 Quit TheLemonMan (Quit: Destructor called)
21:49:37[Saint]gevaerts: not if he used to be able to do it.
21:49:54[Saint]I'd like to assume people don't just forget how to use their iPods.
21:50:03[Saint]however flawed that may be ;)
21:50:14ptrkmjSaint:: Any particular hardware issue? My click wheel works fine...
21:50:50[Saint]ptrkmj: it's apparently not capable of registering that particular button combination anymore.
21:50:53BiontWell, I checked all overwriting options during the manual update. Don't know what RbUtil did...or what I did to crash it. It ran as administrator on a freshly installed Win7. Guess I'll just give the player back to my friend and tell her to keep an eye on it in case there are strange things happening
21:50:59[Saint]which kinda sucks, as, it's important.
21:52:27BiontThank you for your time and help then. I'll report back if the initial problem turns up again. According to my friend it literally stopped playing anything over night. she didn't put on new music or anything
21:52:41[Saint]I would suggest trying to enter the Apple debug mode (select+previous) and doing a button check in there.
21:52:45evilnick_BOh,. hang on...
21:52:47[Saint]ptrkmj: ^^
21:52:52evilnick_BBiont: Hang on a second
21:53:03ptrkmjCan it be connected with removing a ground strap? (The metal plate between click wheel and mainboard)
21:53:10 Quit benedikt93 (Quit: Hofstadter's Law: It always takes longer than you expect, even when you take into account Hofstadter's Law.)
21:53:11evilnick_BBiont: Can you have a look at the manual for Party Mode? See if that sounds familiar?
21:53:28[Saint]ptrkmj: Err...what?
21:53:42evilnick_BBiont: There's a chance that your friend managed to both change the File View setting AND enabled Party Mode.
21:53:51ptrkmjSaint: Have you ever disassembled iPod? :)
21:54:08[Saint]Yes, I have.
21:54:12evilnick_BSo File View made sure the music couldn't be seen, and Party Mode means that it can't be played
21:54:27evilnick_Bs/can't be played/can't be played the usual ways
21:55:06[Saint]ptrkmj: If you specifically removed insulation, I think you found your answer...
21:55:22[Saint]surprise surprise, it's there for a reason.
21:55:47ptrkmjThere's a metal grounding plate situated between click wheel and logic board. My new board (bought on eBay) lacks it.
21:56:39[Saint]I'd say it's likely that is it.
21:56:45[Saint]It's definitely a HW issue.
21:57:11[Saint]if you're 100% that you're doing the reset combination correctly, it *must* be a HW issue.
21:57:29Bionthm...that does sound interesting
21:57:31saratogaBuschel: did you have any specific thoughts about what could be improved for AAC or AAC-HE for GSOC?
21:59:38ptrkmjSaint: I think I'll make a topic on forum. Someone might verify this. Thank you for your help.
22:00:05Buschelsaratoga: I saw your comment and also think AAC-HE should be the focus. starting with profiling it
22:00:29saratogado you know if stripwax ever managed to profile AAC-HE?
22:00:44Buschelno, I do not knwo
22:01:03Biontwell, I can't tell for sure now, but I'm sure pause/stop did least the icon changed.also fast forward seemed to work on paper. not sure about skipping
22:01:14Buschelsaratoga: the right samples will be needed... just to ensure PS is measured right...
22:01:38saratogayes I should make a PS sample
22:01:43 Join sideral [0] (~sideral@rockbox/developer/sideral)
22:02:18 Join japc [0] (
22:05:12***Saving seen data "./dancer.seen"
22:05:38BuschelI would be interested how much CPU all the pre/post rotations in SBR cost.
22:06:01Buschelsame for the hf-stuff...
22:06:03 Join funman [0] (~fun@rockbox/developer/funman)
22:09:26saratogai apparently measured hf_adjustment at 14MHz and hf_generation at 20MHz last summer on AMSv2
22:09:32 Join DX3 [0] (~Dre@
22:10:46*[Saint] waves in funman's general direction.
22:10:56saratogaon the upside both those functions look like they could be massively accelerated on armv5e without too much trouble
22:11:12 Quit japc (Ping timeout: 276 seconds)
22:12:05saratogaASM might help a lot for armv4 too depending on how well gcc handles them
22:12:33 Quit Dreamxtreme (Ping timeout: 252 seconds)
22:22:03saratogaBagder, Zagor: could one of you upload this file to ?
22:22:35saratogaand then further rename "64kaache.m4a" to "nero_he_64.m4a" so that its consistent?
22:24:45CIA-87New commit by saratoga (r29707): Update parse_testcodec to parse AAC-HE files with parametric stereo.
22:25:04saratogahope thats right (i don't know ruby but it looks like perl so i made an educated guess)
22:25:59 Quit Biont (Quit: CGI:IRC (EOF))
22:26:46saratoganot sure if 64k was a sensible bitrate for HEv2, but I think its easier to compare if HE and HEv2 have similar bitrates
22:28:27Buschelshould be fine imho −− I have a real life hev2 @56kbps
22:28:57CIA-87r29707 build result: All green
22:30:40saratogathe ps functions mostly look like filterbanks using small constant tables
22:30:57saratogacould pretty easily be made 2-3x faster on armv5e just by using the DSP instructions
22:31:10saratogaalthough i don't really know how much time is really spent in those functions
22:32:06Buschelon my nano 2g there is not diference between he and hev2, on PP5022 the difference is ~16 MHz
22:32:22Buschel(from a very small test base)
22:32:22saratogawow thats surprising
22:32:32saratogai suppose lack of IRAM could be an issue for PS
22:32:33Buschel2 he and 2 hev2
22:32:43saratogaare the bitrates the same?
22:33:40Buschelboth he ~64kbps, hev2 are 56 and 48 kbps
22:33:53saratogaso maybe that helps
22:34:02saratogaactually i'm not really sure how the LC part works with PS
22:34:06saratogaperhaps the MDCT is faster?
22:34:35saratogaAn HE-AAC v1 decoder will only give mono sound when decoding an HE-AAC v2 bitstream.
22:34:44saratogaah so the MDCT is likely 2x as fast for PS
22:34:51saratogaso thats probably a quite large savings
22:36:39saratogaactually looking at these filters, i bet gcc does quite a poor job of using LDM, so that may be why your ARM9 devices does so much better then ARM7, almost 2x the load store throughput
22:37:57Buschelyou're looking at the channel_filter stuff?
22:38:26 Quit tmzt (Ping timeout: 248 seconds)
22:39:18saratogabut hybrid_analysis, hybrid_synthesis and ps_decorrelate also look bad
22:39:29Buschelnot only ldm, but multiply-add could be used a lot
22:39:43mishramay i ask what is LDM ?
22:39:56Buschelload multiple
22:40:43saratogaon arm7 theres a unified cache 32 bits wide, so each load or store instruction must stall for one cycle
22:41:02saratogaso doing multiple transfers greatly accelerates memory accesses
22:41:18mishragot it
22:41:21saratogaarm9 corrects this by using separate D and I cache so that data can load in parallel with instructions
22:41:50saratogaso LDM is no faster (until arm11 where the cache becomes 64 bit wide and thus ldm is faster again)
22:43:22mishrawhich CPU does Ipod Nano 2g has ?
22:43:22saratogathat page breaks down all the arm cpus and players
22:43:22Buschelwith sloooow memory
22:43:27mishrabecause I have that device.
22:43:31 Join gbl08ma_ [0] (~gbl08ma_@
22:44:50 Join tmzt [0] (
22:45:17saratogaarm9 isn't bad to work with since generally doing things right on it will help arm7 a lot as well, provided you try to use ldm, stm
22:45:35saratogaalthough you won't be able to try the really fun stuff using the dsp operations added on the arm9e cores and later
22:45:56 Quit ptrkmj (Quit: Page closed)
22:47:10mishraI was under the impression that we will go for cpu independent optimization
22:47:38mishrasince it will improve performance on all devices
22:48:42gbl08ma_hello everyone. is it just me, or the option "fade on stop/pause" isn't fading anything? I tried both on and off I can't listen to a difference.
22:49:04saratogaalgorithmic optimizations are of course best, but in practice they're often extremely difficult
22:51:18mishraThis reminds me of one question, the libfaad implementation and ffmpeg implementation implement the same decoding logic or not ?
22:51:37saratogathey implement the same specification, but they're essentially independent of one another
22:52:01saratogathe ffmpeg people made a number of different algorithmic choices in SBR for example
22:52:23mishraok that was what i was looking for
22:52:45mishraso maybe we can try some of their code to see if we gain in performance
22:52:51saratogahave you compared their implementations?
22:52:59mishranot yet
22:53:20 Quit promyloph (Quit: WeeChat 0.3.3)
22:54:18mishrafirst i am reading libfaad
22:54:45saratogareading ffmpeg might be easier
22:55:00saratogaalthough fortunately they tend to use similar function names
22:55:36saratogaalthough comparing similarly named functions (e.g. hybrid_analysis) is quite interesting, as they're implemented very differently
22:57:26mishrainteresting, if it can be proved that ffmpeg is better implemented, then can we go for porting it on RB ?
22:58:26funmanbtw, i read about the aac encoder that google open sourced for android
22:58:31funmanno idea about its performance though
22:58:51saratogai would probably try to go algorithm by algorithm and pick the best available one
22:59:09saratogaalthough i have no idea how difficult it would be to bolt on say the entire PS code onto libfaad
22:59:20saratogaprobably fairly so, given that its not fixed point in ffmpeg
23:00:37funmanthis is a repo created by a FFmpeg dev
23:00:42mishraoh, didnt know that ffmpeg has floating point
23:00:46funmanApache License v2
23:01:26saratogafunman: do they have a decoder too?
23:01:41saratogamishra: we're pretty much alone as far as open source fixed point anything
23:01:50funmaniirc yes
23:02:44funmanthe encoder has a bit of armv5e asm (and armv7 too)
23:04:14merbanansaratoga: we have a fixed point fft niw
23:04:19 Quit kevku (Ping timeout: 260 seconds)
23:04:22saratogayes i saw
23:04:23merbananused in the ac3 encoder
23:04:25saratogaits nearly identical to ours
23:05:06saratogaalthough we have added the questionable memory optimization from tremor's mdct that saves a lot of IRAM but hurts cache locality
23:07:35saratogathat android decoder looks quite interesting, very well documented code
23:07:37bluebroth3rhmm, why does Rockbox leave around on Android after extracting it?
23:07:43saratogaeven references for some of the algorithms
23:08:08 Quit mshathlonxp (Quit: Leaving)
23:09:07bluebroth3rremoving it makes the install about 2.5MB smaller, and it still works (unsurprisingly since it's already extracted)
23:09:59mishrayes google code documentation is awesome, used their vp8 code, each and every algo was described.
23:10:47saratogaha the mp3 decoder uses a similar approach that i am implementing
23:11:17merbananmishra: well not in the beginning
23:11:55merbananafter the ffmpeg decoder was done and all the missing pieces where found it got alot better
23:12:08mishraI took it around December 2010
23:12:13saratogayes vp8 seems to have been somewhat of a mess
23:12:29saratogabut thats not really our problem
23:14:23merbanansaratoga: the android decoder is a maze of files written by Packet Video
23:15:11saratogayes i'm reading through the aac and mp3 decoders now
23:15:18saratoganothing too surprising so far
23:15:38saratogaalthough its hard to imagine they're worse then libfaad
23:15:58 Join wodz [0] (
23:16:49merbananwould be interesting to see how a libavcodec aac fixed point decoder port would fair against them
23:18:08saratoganot interesting enough for me to want to attempt
23:18:28saratogai'd rather just take the good parts of ffmpeg and combine them with the acceptable parts of other fixed point codecs
23:19:30saratogano sense going through all the trouble of tracing down fixed point overflows in code thats not that performance sensitive
23:20:06saratogaalthough porting that android decoder to rockbox and then swapping out its MDCT for ours might be pretty interesting
23:20:19saratogarelative to the degree of difficulty involved
23:20:38 Quit leavittx (Ping timeout: 240 seconds)
23:20:50merbananthe files are *.cpp, but the code looked like c
23:20:50 Join petur [0] (~petur@rockbox/developer/petur)
23:20:51 Quit bieber (Ping timeout: 246 seconds)
23:21:05merbananthat might make someone cry for a few days
23:21:34mishrabut android targets have more powerful cpus, so maybe it will not benefit us straight away
23:21:35saratogai don't see any actual cpp in it
23:22:03saratogaand very little assembly unfortunately
23:22:28merbanananyway license way fsf thinks that gplv3 and apache 2.0 are fine
23:23:16 Join bieber [0] (
23:26:21saratogafree software foundation
23:26:35wodzI erased whole bootloader on my rockchip just to find out that although it is not on the flash it is still operating from ram somehow
23:30:13 Quit n1s (Remote host closed the connection)
23:35:14 Quit gbl08ma_ (Ping timeout: 260 seconds)
23:37:02 Nick sheepshank is now known as alexbobp (
23:45:03 Quit Buschel (Quit: ChatZilla [Firefox 3.6.16/20110319135224])
23:51:33 Quit JackWinter (Read error: Connection reset by peer)
23:52:26 Quit Keripo (Quit: Leaving.)
23:53:00 Join Xerion_ [0] (
23:53:18 Join JackWinter [0] (
23:54:13wodzha, there is spare copy of the bootloader on the flash :-) Let's kill this as well
23:55:06 Quit Xerion (Ping timeout: 246 seconds)
23:55:06 Nick Xerion_ is now known as Xerion (

Previous day | Next day