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 2010-04-26

00:00:40kugelgood, that's the point of the release action, do allow for mapping select and long select to different actions
00:01:06kugelnot a huge fan of that, but it's ok I guess
00:05:07pixelmaI like it better than button combos in most cases (and on some targets button combos aren't really possible, so long and short button presses add more possibilities there)
00:09:56kugelamiconn: yea, I see that point is valid
00:11:15kugelpixelma: ok, so I guess it can go in soonish
00:14:23pixelmaI'll also test on my other two targets (already put the build on them) but the Ondio was the one I have with the least amount of buttons... and I currently use it quite a bit for other reasons anyways
00:14:57kugelthat would be nice
00:14:57 Quit Luca_S (Quit: CGI:IRC (EOF))
00:18:33 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
00:19:36 Quit stripwax (Read error: Connection reset by peer)
00:22:35 Quit Topy (Ping timeout: 276 seconds)
00:23:14 Quit domonoky (Read error: Connection reset by peer)
00:24:04 Join Cornwall [0] (~cornwall@nateres205.tel.miami.edu)
00:24:23CornwallHi, just spreading some love. This program is the greatest thing that happened to my ipod :)
00:25:59 Quit ender` (Quit: Do not meddle in the internals of kernels, for they are subtle and quick to panic.)
00:26:48 Join Topy [0] (~topy@my.fastsh.it)
00:27:32 Part Cornwall ("Leaving.")
00:52:49 Join anewuser [0] (anewuser@unaffiliated/anewuser)
00:55:00CIA-5New commit by 03kugel (r25719): Remove lots of simulator-speficic #ifdef'ing in sound.c to allow for DSP effects in the sims, e.g. for bass and treble (targets that do those effects ...
01:00
01:02:23CIA-5New commit by 03kugel (r25720): Fix yellows
01:04:44 Quit DerPapst (Quit: Leaving.)
01:05:54*kugel needs sim dsp for the alsa driver, for volume changing :)
01:11:51*kugel wonders if/when 3.6 happens
01:13:18 Quit jordan` (Ping timeout: 245 seconds)
01:13:39***Saving seen data "./dancer.seen"
01:14:36saratogacouple weeks now I guess
01:15:24 Join ChrisJJ [0] (~560ee12f@giant.haxx.se)
01:15:42 Quit ChrisJJ (Client Quit)
01:15:52 Join ChrisJJ [0] (~560ee12f@giant.haxx.se)
01:16:52ChrisJJHi. Please add me to the WikiUsersGroup. Thanks.
01:17:33linuxstbChrisJJ: The registration page asked you to use your full name as your wiki name...
01:19:37 Quit ChrisJJ (Client Quit)
01:22:20 Join ChrisJordan [0] (~560ee12f@giant.haxx.se)
01:23:17ChrisJordanOops, sorry - the form allowed me to depart from the default, but I now see this must have been a disobediance test!
01:23:28ChrisJordanNow done. Please add ChrisJordan
01:24:10 Quit scorche (Ping timeout: 260 seconds)
01:24:21 Join shai_ [0] (~Shai@l192-117-110-233.cable.actcom.net.il)
01:24:34amiconnkugel: dsp for volume changing sounds nasty
01:24:51amiconnYou'll lose a lot of dynamic range at low volume
01:25:40kugelpossibly, but I don't see a way to do it without with alsa-lib
01:25:47saratogadon't the MIPS devices already do that though?
01:26:22 Quit Topy44 (Quit: Leaving)
01:26:49 Join elcan [0] (user36@pr0.us)
01:27:04kugelsaratoga: yep
01:27:12linuxstbkugel: If alsa-lib doesn't give you access to hardware volume control, it sounds like you shouldn't be using alsa-lib...
01:27:52kugelI don't necessarily want hardware access, in an app environment the rockbox volume shouldn't affect other apps
01:27:54 Quit shai (Ping timeout: 260 seconds)
01:29:08 Quit archivator (Quit: Leaving)
01:29:18kugelthough, I expect that you're most likely going leaving leave rockbox at 100% and use the hardware controls (via buttons) to control the volume
01:29:31kugels/leaving//
01:31:51ChrisJordanAnd please delete the Wiki user "ChrisJJ" to avoid confusion. Thanks.
01:32:34 Join BHSPitMini [0] (~BHSPitMon@adsl-66-140-45-134.dsl.rcsntx.swbell.net)
01:32:52LloreanWhat's wrong with DSP for volume change in an RBaaA evironment?
01:32:54saratogakugel: i don't know much about how sound libraries work, but I think most let you specify the volume you want to the OS mixer, which then adjusts the amplifier output power in conjunction with digital gain adjustments
01:33:28saratogaLlorean: nothing wrong exactly, but its kind of a weird way to do things, most apps use system calls and let the OS figure out the best way to do it
01:33:58Lloreansaratoga: I know that some of my Windows apps will change volume within the app if I use volume keys, but also allow me the option to disable this so that the volume keys adjust the host OS volume instead.
01:34:04 Join CaptainKewl [0] (~jason@207-237-106-60.c3-0.nyr-ubr1.nyr.ny.cable.rcn.com)
01:34:18LloreanThis is, more or less, the behaviour I'd expect from RBaaA (and I assume "rockbox volume" would be dsp?)
01:34:20linuxstbChrisJordan: I've given ChrisJordan write access, but it needs a wiki admin to delete users. ChrisJJ will get deleted during the next cleanup...
01:34:25saratogaLlorean: yeah DS lets you do that, if you use KS you can only adjust the global volume
01:35:13LloreanI think the option to adjust rockbox independently of the host is essential for environments where Rockbox gets mixed into something else (so, for example, you hear the incoming message noise)
01:35:51kugelsaratoga: DS/KS?
01:35:57saratogai guess for now it doesn't really matter, this can be made target specific later on (and i doubt most mobile devices will do it the windows or linux way anyway)
01:36:05saratogakugel: direct sound, kernel streaming
01:36:14saratogatwo apis on windows i'm vaguely aware of having used
01:36:50saratogakernel streaming basically just copies the sound to the output buffer without mixing, DS uses the OS mixer in some way (depends on the windows flavor)
01:38:36kugelsaratoga: what you're saying might be possible with alsa-lib, I haven't looked throughoutly at it, mostly only at the pcm part to get working sound output
01:39:22 Quit mirak (Quit: Ex-Chat)
01:39:33saratogai don't know anything about linux sound, but I would be amazed if windows was the only OS with a mixer that could adjust each stream's volume
01:39:55kugelthere's also some mixer part, but I had the feeling it's more about the output mixers (not sure if the term is correct, I mean that what you see when running alsamixer). I might have another look thoughj
01:40:39kugelwell, gnome has some applet that lets you change the volume per application outside of the application I guess there's a way I haven't found yet
01:41:34kugelotoh, I looked at alsaplayer and it also fixes up the volume of the stream before it's fed into the alsa-lib pcm functions in the audio callback
01:42:38kugeland changing the volume in ario (an mpd client) affects the global volume :/
01:43:18saratogahasn't alsa changed a lot in the last few years, maybe some software hasn't been updated?
01:44:03kugelpossibly
01:44:59kugelmaybe looking at how SDL does it could be interesting
01:45:07saratogakugel: wikipedia says pulseaudio handles per application volume control rather then alsa
01:45:41pixelmathan?
01:46:43saratogahttp://en.wikipedia.org/wiki/PulseAudio#Alternatives
01:47:25saratogamy very quick reading of that is that alsa is a kernel module, while pulseaudio is a complementary userspace layer which provides additional mixing and per application volume control
01:47:48saratogaso it sounds like if you use alsa directly you're doing something more like kernel streaming in windows, and get to handle volume control yourself
01:48:30kugelalsa has both kernel support (e.g. as a module) and a user space library
01:48:58kugelthe german article says that alsa apps automagically redirect sound to pulseaudio, which in turn uses alsa again to feed the hardware directly
01:50:56saratogamy completely uninformed take is that its probably worth looking at how some embedded linux or android or osx device handles this, and seeing if its even comparable to alsa on the PC
01:51:37saratogaif its not, just use software volume control or change the system volume for now and worry about how to do it properly later
01:52:03kugelfrom my understanding is that alsa is what generally you use on linux directly or indirectly, OSS is deprecated also
01:52:41kugelthat'd also include embedded systems
01:53:03kugelI can imagine that embedded systems leave out pulseaudio, but probably not alsa
01:53:22kugelI'm not sure what android does but you need to go through the java thing anyway for sound output (which I'd expect to use alsa behind the scenes)
01:55:01kugelI haven't really heard of an alsa alternative, expect SDL
01:55:52 Join scorche [0] (~scorche@rockbox/administrator/scorche)
01:56:13kugelemdebian definitely comes with alsa as my test app ran without changes, I didn't even need to install extra packages
02:00
02:01:06 Quit CaptainKewl (Ping timeout: 240 seconds)
02:02:09 Join krabador [0] (~darkham@host165-5-dynamic.55-79-r.retail.telecomitalia.it)
02:04:09 Quit linuxstb (Read error: No route to host)
02:05:58 Join linuxstb [0] (~linuxstb@rockbox/developer/linuxstb)
02:06:09 Quit kugel (Remote host closed the connection)
02:07:56 Join kramer3d [0] (~kramer@unaffiliated/kramer3d)
02:09:06 Quit efyx (Remote host closed the connection)
02:19:45 Quit liar (Ping timeout: 240 seconds)
02:22:31 Join mikroflops_ [0] (~yogurt@90-227-45-110-no112.tbcn.telia.com)
02:23:41ChrisJordanThanks linuxstb!
02:23:46 Quit ChrisJordan (Quit: CGI:IRC)
02:26:18 Quit mikroflops (Ping timeout: 258 seconds)
02:29:24 Join mikroflops [0] (~yogurt@90-227-45-110-no112.tbcn.telia.com)
02:32:58 Quit mikroflops_ (Ping timeout: 246 seconds)
02:36:51 Quit Unhelpful (Remote host closed the connection)
02:40:33 Nick fxb is now known as fxb__ (~felixbrun@h1252615.stratoserver.net)
02:41:35 Join mikroflops_ [0] (~yogurt@90-227-45-110-no112.tbcn.telia.com)
02:43:31 Join Unhelpful [0] (~quassel@71.173.205.32)
02:43:31 Quit Unhelpful (Changing host)
02:43:31 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
02:44:42 Join komputes [0] (~komputes@ubuntu/member/komputes)
02:45:13 Quit mikroflops (Ping timeout: 246 seconds)
02:48:47 Nick fxb__ is now known as fxb (~felixbrun@h1252615.stratoserver.net)
02:50:08 Quit MethoS- (Remote host closed the connection)
02:50:17 Quit krabador (Read error: Connection reset by peer)
02:51:02 Quit Spaceghost (Read error: Connection reset by peer)
03:00
03:01:18 Join xpansive [0] (~8ea592bb@giant.haxx.se)
03:01:38 Quit xpansive (Client Quit)
03:13:42***Saving seen data "./dancer.seen"
03:14:22 Join RandomInsano [0] (~Edwin@wnpgmb0806w-ad07-62-231.dynamic.mts.net)
03:20:42RandomInsanoI need to figure out how to get code running on a Telechips 8200 cpu. I know I need to prepend a header, but how do I get started? There is a LOT of code for the Cowon D2 that I've been sifting though, but it's hard to filter out the basic bootloaderish stuff. I'm simply trying to make a 'hello world' of sorts for now. Blinking the backlight is the goal, and it should just take either GPIOF or GPIOB.
03:22:04RandomInsanoOr... maybe it's under the bootloader directory?
03:30:08saratogacan't you just do what the D2 does?
03:33:08RandomInsanoThere is a LOT of code for what the D2 does.
03:33:15RandomInsanoIt's hard for me to figure out the basic steps
03:33:48RandomInsanoI also don't know the whole codebase, how everything compiles together...
03:34:37 Join JdGordon_ [0] (~jd@175.35.161.102)
03:34:37 Quit JdGordon_ (Changing host)
03:34:37 Join JdGordon_ [0] (~jd@rockbox/developer/JdGordon)
03:35:27RandomInsanoI'm still looking though
03:35:39saratogahave you compiled a D2 build and watched what it builds?
03:35:45saratogaD2 bootloader I mean
03:35:59RandomInsanoCan I just build a bootloader?
03:36:02saratogathere should only be a few files involved, most of which are drivers you don't need
03:37:17 Quit ps-auxw (Read error: Operation timed out)
03:37:28RandomInsanoGood point. I've read through 'trunk/bootloader/telechips.c' but not sure if I can just build that with the build system
03:38:13 Join Rob2223 [0] (~Miranda@p4FDCBC10.dip.t-dialin.net)
03:38:13saratogathe D2 install directions explain how to prepare a D2 binary for use with tcctool
03:38:19saratogamaybe the process is similar for your player
03:38:25 Quit pixelma (Disconnected by services)
03:38:27 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
03:38:33RandomInsanoMaybe it is. I'll start there. Thanks
03:38:48 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
03:39:11 Join ps-auxw [0] (~arneb@2001:470:c807:0:1532:4e5f:2ad3:4123)
03:39:59saratogamy guess would be you want to compile a stripped down version of the D2 bootloader, and then try loading that with tcctool like on the D2
03:40:00 Nick bgs100 is now known as bgs000 (57o9@unaffiliated/bgs100)
03:40:33saratogasince the lcd drivers probably aren't exactly the same as teh D2 the screen probably won't work, but hopefully you have the GPIO for the backlight or something similar
03:41:01 Quit saratoga (Changing host)
03:41:01 Join saratoga [0] (~9803c6dd@rockbox/developer/saratoga)
03:41:05 Quit Rob2222 (Ping timeout: 240 seconds)
03:41:26 Quit JdGordon_ (Ping timeout: 258 seconds)
03:41:36RandomInsanoI can enable them myself. There are four PWM channels on the TCC8200. They map to a few GPIO pins.
03:42:05RandomInsanoSince I have backlight brightness control in the factory firmware I assume they use one of the PWM channels on the chip.
03:48:05 Quit adnyxo (Quit: Leaving)
03:48:30 Join n1s [0] (~n1s@rockbox/developer/n1s)
03:49:00 Join JdGordon_ [0] (~jd@rockbox/developer/JdGordon)
03:53:42 Quit JdGordon_ (Ping timeout: 260 seconds)
04:00
04:04:40 Quit tchan (Quit: WeeChat 0.3.1.1)
04:06:47 Quit anewuser (Quit: http://xrl.us/NitroQueer What do you know...THE WORLD'S first NTRQ (that's for NES/FAMICOM) tracking compo. Have powerpak? Try it out! Otherwise ROM IMAGE.)
04:09:56 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
04:13:52 Quit Barahir_ (Ping timeout: 246 seconds)
04:15:24 Quit komputes (Ping timeout: 276 seconds)
04:15:48 Join Barahir [0] (~jonathan@gssn-5f756cce.pool.mediaWays.net)
04:27:43 Quit Unhelpful (Remote host closed the connection)
04:28:16 Join Rondom [0] (~quassel@dslb-084-057-141-252.pools.arcor-ip.net)
04:28:50 Quit Rondom_ (Ping timeout: 260 seconds)
04:31:12 Quit emrecelikten (Quit: Leaving.)
04:35:49 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
04:36:58 Quit rvvs89 (Ping timeout: 248 seconds)
04:39:37 Quit panni_ (Read error: Connection reset by peer)
04:45:45 Quit bzed (Read error: Connection reset by peer)
04:45:49 Join bzed [0] (~bzed@devel.recluse.de)
04:46:33 Nick fxb is now known as fxb__ (~felixbrun@h1252615.stratoserver.net)
04:47:35 Quit n1s (Ping timeout: 276 seconds)
04:50:53 Join n1s [0] (~n1s@rockbox/developer/n1s)
04:52:02 Join SirFunk [0] (~Sir@97-92-38-108.dhcp.aldl.mi.charter.com)
04:54:07 Quit n1s (Client Quit)
04:55:33 Join Rob2222 [0] (~Miranda@p4FDCAA68.dip.t-dialin.net)
04:55:47 Quit TheSeven (Ping timeout: 276 seconds)
04:59:23 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:59:43 Quit Rob2223 (Ping timeout: 265 seconds)
05:00
05:10:03 Join saratogalab [0] (~9803c20d@gateway/web/freenode/x-lbsuhyczsnsfvlqa)
05:13:44***Saving seen data "./dancer.seen"
05:32:06 Quit Horschti (Quit: Verlassend)
05:58:52 Quit ollebe (Quit: Leaving)
06:00
06:07:56 Quit esperegu (Ping timeout: 276 seconds)
06:08:16 Join lifeless [0] (~lifeless@188.16.66.110)
06:11:24 Quit bluebrother (Disconnected by services)
06:11:27 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
06:11:29 Quit BHSPitMini (Ping timeout: 260 seconds)
06:21:51 Join XPANSiVE [0] (~XPANSiVE@142.165.146.187)
06:25:05 Part RandomInsano
06:32:24 Quit XPANSiVE (Ping timeout: 240 seconds)
06:48:59 Join LinusN [0] (~linus@rockbox/developer/LinusN)
07:00
07:05:21 Nick shai_ is now known as shai (~Shai@l192-117-110-233.cable.actcom.net.il)
07:13:48***Saving seen data "./dancer.seen"
07:19:57 Quit CGL (Quit: Saliendo)
07:22:46 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
07:27:23 Join esperegu [0] (~quassel@145.116.15.244)
07:35:16 Join esperegu_ [0] (~quassel@145.116.15.244)
07:35:55 Quit esperegu (Ping timeout: 276 seconds)
07:45:40 Join esperegu [0] (~quassel@145.116.15.244)
07:46:07 Quit esperegu_ (Ping timeout: 258 seconds)
07:53:29 Quit esperegu (Ping timeout: 252 seconds)
07:55:49 Quit kramer3d (Quit: Leaving)
07:57:40 Join Xerion_ [0] (~xerion@84.25.7.202)
07:58:51 Join jordan` [0] (~jordan@jem75-13-78-235-252-137.fbx.proxad.net)
07:59:08 Quit Xerion (Ping timeout: 264 seconds)
07:59:08 Nick Xerion_ is now known as Xerion (~xerion@84.25.7.202)
08:00
08:04:24 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
08:07:05 Join esperegu [0] (~quassel@145.116.15.244)
08:12:21 Join TheSphinX^ [0] (~cold@p54A5DAED.dip.t-dialin.net)
08:13:13 Part Llorean
08:14:13 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
08:23:39 Join kugel [0] (~kugel@rockbox/developer/kugel)
08:33:16 Nick jnss is now known as ANNA (janes@gateway/shell/sign.io/x-vwlduembtdtjdmud)
08:53:37 Join flydutch [0] (~flydutch@host24-146-dynamic.15-87-r.retail.telecomitalia.it)
08:54:25 Join n1s [0] (~n1s@rockbox/developer/n1s)
08:57:16 Quit saratogalab (Quit: Page closed)
09:00
09:07:20 Join petur [0] (~petur@rockbox/developer/petur)
09:12:58kugelsdl also changes the volume in a dsp-like fashion
09:13:50***Saving seen data "./dancer.seen"
09:13:57 Join ender` [0] (krneki@foo.eternallybored.org)
09:16:12 Join mischasworld [0] (~quassel@f051115178.adsl.alicedsl.de)
09:18:22 Join mikroflops [0] (~yogurt@90-227-45-110-no112.tbcn.telia.com)
09:22:11 Quit mikroflops_ (Ping timeout: 246 seconds)
09:25:44 Join DerPapst [0] (~Alexander@dslb-088-069-156-106.pools.arcor-ip.net)
09:26:46S_a_i_n_tWhy is it that the Cabbie themes don't seem to use viewports much?
09:27:08S_a_i_n_tI mean, they all seem to draw in the 'default viewport' when it comes it icons etc.
09:27:22S_a_i_n_tIsn't Cabbie supposed to be the "demo" theme?
09:27:31 Join liar [0] (~liar@213162066160.public.t-mobile.at)
09:29:34S_a_i_n_tIMO the Cabbie themes should also be fairly heavily commented, to give users a better idea of which part of the code does what. I'd be happy to put a patch together, but it might not be something I can do immediately.
09:30:00S_a_i_n_ts/might not/probably won't/
09:31:20S_a_i_n_tAs well as the 'default' theme, it is also supposed to serve as an example...is it not?
09:37:07 Join lpereira [0] (~lucien@did75-8-82-226-27-213.fbx.proxad.net)
09:39:24 Quit kugel (Ping timeout: 276 seconds)
09:40:36pixelmayou are allowed to draw (images) in the default viewport. I guess the reason is just that cabbiev2 was designed before viewports or when viewports, especially conditional viewports, were still fairly new. And since they "just work" no-one bothered to clean them up yet.
09:42:21S_a_i_n_tYou *can* draw images in the default viewport...but it kinda teaches bad practice.
09:42:54S_a_i_n_tAnd yes, I think you're right. They "just work", so I doubt anyone's bothered keeping the code up to date.
09:43:43S_a_i_n_tAt the very least, I want to put a patch together that comments the .wps files explaining which lines/segments do what.
09:44:26S_a_i_n_tI doubt there would be any objections to me doing that, would thee?
09:44:30S_a_i_n_t*there
09:44:52pixelmaif you like working on it go ahead. :) And while you are at it, I'd like every port to show volume in dB instead of the icon while adjusting - it was added to the Clip port but not others (an colour targets I imagine that in yellow font too)
09:45:36 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
09:46:26 Quit esperegu (Read error: Connection reset by peer)
09:46:28pixelmanon-existence of viewports makes this task a bit harder because for this you will need viewports
09:48:12S_a_i_n_tYes, I guess I'm going to have to convert all the (applicable) cabbies to use viewports, I believe JdGordon talked about doing this (or the need to do this) quite a long time ago...
09:49:00S_a_i_n_tI don't necessarily *like* doing things like this (I'd rather work on my own code/themes), but it is something that needs to be/should be done.
09:49:24 Join esperegu [0] (~quassel@145.116.15.244)
09:49:32S_a_i_n_tAnd I'm sure the new themers will thank me for it.
09:49:59S_a_i_n_ts/thank me/be thankful/
09:50:51linuxstbS_a_i_n_t: I'm not sure I agree that Cabbie should be a _demo_ theme. It's the _default_ theme. Maybe instead of heavily commenting cabbie, it would be more useful to do a new theme, designed to be a demo, using as many WPS features as you can.
09:54:13S_a_i_n_tlinuxstb: When the other shipped themes are taken out of the release builds (which *will* happen eventually, that's the reason they were all added to the themesite), then Cabbie will have to serve the purpose of being both the default *and* the demo theme. The idea I believe is to update the code of Cabbie as heavily as possible without changing its appearence.
09:54:33pixelmait would be nice if the existing code of cabbie was a bit commented though
09:54:35S_a_i_n_tIt is just a bit of a daunting task, which is why I haven't started doing this already...
09:55:51pixelmalinuxstb: I think S_a_i_n_t is not talking about "demo" in the sense of "show off" just showing and explaining some basics
09:56:03linuxstbS_a_i_n_t: That logic doesn't follow... The demo theme doesn't have to be shipped with Rockbox. In fact, I would doubt anyone would want to use it, as it would have too many features. But of course, commenting cabbie would also be a good thing.
09:56:41S_a_i_n_tCommenting *and* converting to viewports ;)
09:57:02 Join rvvs89 [0] (ivo@bright-snat.ucc.asn.au)
09:57:04GodEaterI think no-one has been near this because of the amount of work it took to get CabbieV2 stable on all targets
09:57:12GodEaterit took *months* of work by a lot of people
09:57:42S_a_i_n_tGodEater: Yep...and I'm *slowly* getting through it all ;)
09:58:10S_a_i_n_tAnother task I have at hand is making iconsets for the largescreen players.
09:58:12GodEaterI happen to agree with linuxstb too - I don't want a default theme which uses every single feature of WPS
09:58:23S_a_i_n_tSo that the icns don't look all tiny and weird.
09:58:34pixelmaif someone else works on it, it might give me incentive to finally make the c200 port use a 10-pixels tall font, using the screen estate better - one more line possible which can finally show playing time info
09:59:04GodEaterwhat size does it use now?
09:59:15pixelma11-pixels
09:59:38pixelma> 7 lines on the 132x80 display
09:59:46GodEaterS_a_i_n_t: I'd be happy to see someone start a new default theme effort too - if you think some new WPS features would be worth using in it.
09:59:48S_a_i_n_tGodEater: My intention is not to change the appearence of Cabbie *at all*, just to update the code, and decrease the total size of the theme by using progressbars for the battery and volume and kicking the bitmap strips out.
10:00
10:00:22S_a_i_n_tA new default theme will probably start a huge flamewar ;)
10:00:51pixelmaas I said earlier to linuxstb, it seems you are talking about different meanings of "demo"
10:01:07 Quit rvvs89 (Changing host)
10:01:07 Join rvvs89 [0] (ivo@pdpc/supporter/base/rvvs89)
10:01:52S_a_i_n_tpixelma: Yes, well...kinda. I don't want to pack it with every feature, but I want the code to be current at the very least.
10:01:56linuxstbpixelma: My understanding of "demo" is a theme which uses as many features as possible, to demonstrate to a theme author how to do things.
10:03:03S_a_i_n_tI guess what I meant by "demo" is more to do with comments that reflect the code. I suppose "learning theme" would be more appropriate?
10:03:38linuxstbThe default theme should be more like "showing off" - i.e. you've just installed Rockbox, this is what it can do.
10:04:06S_a_i_n_tAnd, although it isn't *that* important, I would like to get rid of the gratuitous use of bitmap strips wherever possible.
10:04:41S_a_i_n_tHehehe....Cabbie is hardly "showing off" ;)
10:04:53S_a_i_n_tPerhaps it used to be, but it certainly isn;t anymore :P
10:05:13 Join wodz [0] (~wodz@chello087206240004.chello.pl)
10:06:29wodzI would like to ask for advice. WM8750 is very flexible codec and I am thinking how to setup it for recording and monitoring. Generally I have 3 options:
10:06:58wodz1) Setup monitoring as a analog bypass
10:06:59S_a_i_n_tI *believe* that getting rid of that battery and volume bitmap strips can reduce the total size of the theme by ~20/25% or so.
10:07:36wodz2) Setup monitoring as analog bypass after input amplifier stage (PGA)
10:07:52wodz3) Setup full digital monitoring
10:09:04wodzoption 1 will allow to turn off most of the subsystems of codec
10:09:39wodzoption 2 is intermediate
10:09:59wodzoption 3 will allow to use digital filtering features of the codec
10:10:22wodzbut 3 will use the most energy
10:11:01linuxstbwodz: I think the important thing is that monitoring should be as close to an actual recording as possible. So probably 3... But I'm not sure what other targets do.
10:11:21*linuxstb pings petur as resident recording guru...
10:11:34*petur wakes up
10:11:48wodzlinuxstb: I am thinking of setting up option 1 for FM listening and option 3 for all others monitoring
10:12:59peturI think the irivers do (1)
10:13:33wodzpetur: Isn't it because hardware limits monitorin to analog domain in irivers?
10:13:38peturbut I haven't looked at that code for a looong time
10:13:47peturyes
10:13:51 Join B4gder [0] (~daniel@static-213.88.151.242.addr.tdcsong.se)
10:13:55 Quit B4gder (Changing host)
10:13:55 Join B4gder [0] (~daniel@rockbox/developer/bagder)
10:14:24peturFor normal radio listening, doing it all analog is the easiest with best quality
10:14:46wodzpetur: You mean in general or on irivers?>
10:15:01peturif you want to monitor recording, maybe doing it all digital gives the safest option
10:15:43peturin general... If you're only listening to radio (not recording), why would you go digital?
10:15:58wodzpetur: digital filtering?
10:16:07peturhmmm right
10:16:30petur and probably also mixing in voice navigation
10:16:50wodzhmm interesting
10:17:09wodzhow is voicing implemented?
10:17:11peturalthough the audio chip can probably mix too
10:17:28peturright now it plays an mp3
10:17:41petur(TTS is a GSoC project)
10:18:24wodzaa so I can't just power down DAC and do analog bypass of FM signal right?
10:19:39pixelmait doesn't play an mp3 on swcodec players... it's a special speex file
10:19:53wodzthat doesn
10:19:54*petur stands corrected
10:20:13wodzt matter - it feeds DAC with data to play
10:20:44peturI can't remember how this works on iriver...
10:21:17wodzanyway to keep voicing working I have to have DAC powered up all the time - am I correct?
10:24:53peturI don't know if it makes sense to only power it on if required to play. What powersavings do you gain?
10:27:15S_a_i_n_thttp://imgur.com/DIIrR.png <== A new "Cabbie" I'm working on for iPod Nano 1/2g, comments welcome.
10:27:29 Quit n1s (Ping timeout: 240 seconds)
10:27:37S_a_i_n_tNo album/track info as yet. Only been working on it for like 30 minutes.
10:27:58 Join n1s [0] (~n1s@rockbox/developer/n1s)
10:28:26 Quit rasher (Quit: restarting tmux)
10:28:47 Join rasher [0] (~jonas@0x5550f5a3.adsl.cybercity.dk)
10:28:53 Quit rasher (Changing host)
10:28:53 Join rasher [0] (~jonas@rockbox/developer/rasher)
10:29:10peturthe reflection wastes so much vertical space...
10:29:44wodzfrom datasheet for codec - paying to 16 ohm headphone (so DAC up output stage up) 34.5 mW, line-in to 16 ohm headphones (so basically analog bypass from analog input source) 8 mW, recording and playback at the same time 80 mW
10:30:09wodzs/paying/playing/
10:30:48S_a_i_n_tpetur: yes, it does...it's a port of "Cabbie 3.0"
10:31:00JdGordonS_a_i_n_t: is txt going to go under the icons?
10:31:10 Quit TheSeven (Ping timeout: 246 seconds)
10:31:57S_a_i_n_tJdGordon: *possibly*, but probably not. I'll probably end up dropping the icons and the playbar down by ~3/4px
10:32:22S_a_i_n_t3 or 4, not the quarters ;P
10:32:25JdGordonand to the bottom of the display
10:32:55S_a_i_n_tI don't parse that...you mean drop the reflection altogether?
10:33:14JdGordonyes
10:33:22JdGordonI agree that they take up too much space
10:33:28JdGordonor appear to in the screenshot
10:33:46 Quit rasher (Quit: Changing server)
10:33:48S_a_i_n_tBlargh...I wanted to keep the "Cabbie 3.0" feel.
10:34:08S_a_i_n_tI may go for a slightly thinner (2px or so) playbar also.
10:34:18 Quit Farthen (Ping timeout: 276 seconds)
10:34:27S_a_i_n_tIt *may* seem more proportionate when there is text on the screenshot also
10:34:57 Quit bluebroth3r (Ping timeout: 276 seconds)
10:34:59 Join rasher [0] (~jonas@0x5550f5a3.adsl.cybercity.dk)
10:35:00 Quit rasher (Changing host)
10:35:00 Join rasher [0] (~jonas@rockbox/developer/rasher)
10:35:36 Quit bug2000 (Ping timeout: 276 seconds)
10:35:47 Quit esperegu (Read error: Connection reset by peer)
10:35:47 Join Farthen [0] (~Farthen@static.225.178.40.188.clients.your-server.de)
10:36:12S_a_i_n_tPresently, there is 83px above the playbar...which allows for 5 or 6 lines of 12~15px font.
10:36:20 Join bluebrother [0] (~dom@f053154190.adsl.alicedsl.de)
10:36:20 Quit bluebrother (Changing host)
10:36:20 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
10:36:40 Join esperegu [0] (~quassel@145.116.15.244)
10:36:53S_a_i_n_tI *may* be able to stretch that out to 90px and still keep *some* of the reflection.
10:38:35S_a_i_n_tOr I could put aome text under the icons, on top of the reflections and keep the current positioning. Though, I'm not sure what text I would display down there.
10:38:44 Quit mischasworld (Remote host closed the connection)
10:39:02S_a_i_n_ts/aome/some/
10:39:23 Join rasher_ [0] (~rasher@0x5550f5a3.adsl.cybercity.dk)
10:39:23 Quit rasher_ (Changing host)
10:39:23 Join rasher_ [0] (~rasher@rockbox/developer/rasher)
10:41:08 Quit rasher (Quit: leaving)
10:41:09 Nick rasher_ is now known as rasher (~rasher@rockbox/developer/rasher)
10:43:26 Join DerPapst1 [0] (~Alexander@dslb-088-069-156-106.pools.arcor-ip.net)
10:44:28 Join efyx [0] (~efyx@lap34-1-82-225-185-146.fbx.proxad.net)
10:44:46 Join DerPapst2 [0] (~Alexander@dslb-088-069-156-106.pools.arcor-ip.net)
10:45:06S_a_i_n_tJdGordon: petur: <== Is this a bit better?
10:45:11 Quit DerPapst (Ping timeout: 240 seconds)
10:45:27S_a_i_n_toooops, need the link too http://imgur.com/9tRum.png
10:45:43peturyes
10:45:46JdGordonyeah, that looks better
10:46:01JdGordonshould give you enough room for text between the bar and icons
10:46:10S_a_i_n_tThis makes space for current time/time left between the playbar and the icons.
10:46:18S_a_i_n_thahahah, snap!
10:46:26S_a_i_n_tyes, 12px font will fit there.
10:47:23S_a_i_n_tAnd I still get to keep (a little bit of) the reflection then \0/
10:47:32 Quit DerPapst1 (Ping timeout: 245 seconds)
10:48:20 Join bug2000 [0] (~bug@unaffiliated/bug2000)
10:49:11 Quit mikroflops (Ping timeout: 240 seconds)
10:51:26 Join mikroflops [0] (~yogurt@90-227-45-110-no112.tbcn.telia.com)
10:53:03wodzwhat is exactly datapath when voicing in recording mode?
10:54:34JdGordonthe only way you'll get an exact answer is by staring at the code for a few hours...
10:55:15wodzI love such answers :-)
10:57:15JdGordonsorry :)
10:58:11 Quit mikroflops (Ping timeout: 240 seconds)
10:58:44 Join mikroflops [0] (~yogurt@90-227-45-110-no112.tbcn.telia.com)
10:59:34 Quit wodz (Quit: Leaving)
10:59:40 Quit DerPapst2 (Read error: Connection reset by peer)
11:00
11:10:24*amiconn wonders whether he should change his Elio RE approach once more and try to hack together (preliminary) pp5020 sxupport for skyeye
11:13:52***Saving seen data "./dancer.seen"
11:15:52linuxstbamiconn: What have you been doing so far? Have you tried the Sansa emulator?
11:16:11pamaurygevaerts: ping
11:16:50 Quit phanboy4 (Quit: Leaving)
11:22:45 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:34:56 Quit BHSPitMonkey (Remote host closed the connection)
11:34:59 Join DerPapst [0] (~Alexander@p5099d40e.dip0.t-ipconnect.de)
11:38:19 Join wodz [0] (~wodz@skatol.ch.pw.edu.pl)
11:42:30JdGordondoes anyone care about the rm skin patch?
11:43:15 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
11:45:44 Quit TheSeven (Read error: Connection reset by peer)
11:45:48 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
11:48:18JdGordonfm/radio skin patch
11:49:18JdGordonI'm starting to think that there is little point waiting for the skin buffer resizing patch so I'm going to make the buffers large enough to fit (IIRC mutlifont doesnt work on some targets because the buffers are too small anyway) and commit the fm patch
11:49:43JdGordoni.e in the next few days
11:49:58 Quit TheSeven (Disconnected by services)
11:50:00 Join The_Seven [0] (~TheSeven@rockbox/developer/TheSeven)
11:51:12 Quit Zagor (Remote host closed the connection)
11:53:42JdGordonalso on that topic I think an option for how much buffer to steal is a good idea
11:54:02 Join b0hoon [0] (~quassel@62.87.184.82)
11:54:25 Quit The_Seven (Ping timeout: 246 seconds)
11:57:34JdGordonwe can then add some logic (2 lines of code) to splash when a theme doesnt load because it needs more buffer
11:57:40JdGordonshould keep everyone happy
11:58:20TorneJdGordon: had any bright ideas re font loading yet? :)
11:58:25 Join mikroflops_ [0] (~yogurt@90-227-45-110-no112.tbcn.telia.com)
11:59:06JdGordonnope, havnt thought about it at all
11:59:43JdGordonwe could maybe get the skins (ther than the base skin) loading on demand instead of on boot to speed it up
11:59:57JdGordonexcpet the bad font is needed there so maybe that wont affect anything
12:00
12:01:02Tornethere's several "obvious" solutions but thye all suck for one reason or another :)
12:02:03Tornejust dumping the whole buffer as a file might be best but it's a little fiddly
12:02:10 Quit mikroflops (Ping timeout: 240 seconds)
12:02:42Tornethere's a few pointers to fix, you need a file per font, you may have to have more than one per font if you load the font into different size buffers at different times (one theme uses it as main font, one as a skin font?)
12:03:52amiconnlinuxstb: I've tried skyeye with another (existing) SoC option, occasionally fixing up registers when the I/O area is being read
12:04:06amiconnBut it's time consuming to step through all the time
12:04:07 Quit esperegu (Read error: Connection reset by peer)
12:04:56amiconnI didn't try the sansa emulator
12:05:11 Join esperegu [0] (~quassel@145.116.15.244)
12:05:31linuxstbamiconn: Yes, I can imagine it's a slow process. But implementing what we know of the PP502x for skyeye would sound like a generally useful thing to do.
12:05:46wodzI would like to commit mpio hd200 port finally. Could I ask for look at #11137?
12:06:02wodzAny comments welcome
12:06:55linuxstbwodz: Are you intending on committing the whole patch at once?
12:07:12wodzlinuxstb: no
12:07:58 Join watto [0] (~watto@193.203.81.165)
12:07:58wodzI would like to split it at least for new files and existing files
12:08:26wodzbut maybe each changed existing file should be separate commit
12:08:39wodzwhat do You think?
12:10:45n1si think new files and changed files in two commits is fine
12:11:06 Quit antil33t (Ping timeout: 264 seconds)
12:11:13 Join antil33t [0] (~Mudkips@203-184-54-232.callplus.net.nz)
12:11:16linuxstbwodz: Why model number 170 in the config file? 69 looks to be the next free one. (I don't know why the lyre and mini2440 chose 130 and 131)
12:11:51wodzit was just arbitrary choice
12:12:27*linuxstb also wonders why the mini2440 has target_id 99
12:12:31 Join M3DLG [0] (~M3DLG@212.183.140.55)
12:13:12wodzbut target_id is 69 for MPIO
12:14:33linuxstbwodz: As you say, the choice is arbitrary, but I think we may as well just use the next free numbers.
12:14:38wodzshould target_id == MODEL_NUMBER ?
12:15:18linuxstbTo be honest I'm not sure why we have two... But I think it makes sense to keep them the same if you can.
12:16:21wodzhmm it is big mess
12:16:53linuxstbI think target_id is used in voice files (and maybe lang files). MODEL_NUMBER is used for the checksum in the "rockbox.iriver" etc file headers.
12:18:24wodziaudio M3: MODEL_NUMBER==25, target_id==37, number in menu 34
12:18:36wodzso I don't see any pattern here
12:19:13b0hoonfrom what i understood the targed_id should be the next free number but the "target platform number" is reserved and grouped by a vendor name
12:19:49wodzb0hoon: I understood it the same
12:20:04b0hoonso for example sandisk has numbers from 50 to lets say 70
12:20:52 Join kugel [0] (~kugel@rockbox/developer/kugel)
12:21:00b0hoonbut don't know for sure, ive assumed it from my obserwations
12:21:26linuxstbThe menu number is something that can change easily - it's just used iin the configure menu (and the target name can also be used in that menu, and should be used by scripts running configure)
12:22:28b0hoonso the packard bell got 160 because it was the next free interval
12:22:49wodzso is 170 ok as menu number for mpio?
12:23:03linuxstbDon't worry about the menu number... That's intentionally different to the internal target_id and MODEL_NUMBER values.
12:23:33wodztarget_id is next free so the value of concern is MODEL_NUMBER
12:23:36b0hooni think: yes, but i dont know what the other developers think about it
12:25:23linuxstbwodz: I would say use the same as target_id if available. It would seem sensible to make them the same wherever possible.
12:26:03wodzI'll check if 69 is free than
12:27:13wodzseems free
12:28:31wodzsome other comments?
12:33:10 Quit liar (Ping timeout: 240 seconds)
12:34:13linuxstbwodz: I'm just looking at your crt0.S - you've put a #ifdef MPIO_HD200 inside a #ifndef IAUDOI_M3. Couldn't those be done differently?:
12:39:53wodzlinuxstb: it can for sure
12:40:21wodzbut why this is wrong?
12:40:21n1si think merging your bytsaw changes into ata-as-coldfire would be better to avoid duplication
12:40:35n1sbyteswap even
12:40:47linuxstbwodz: It's not wrong (logically), it just struck me as an awkward way to do the #ifs
12:41:20wodzlinuxstb: current ifndef IAUDIO_M3 is not any better
12:41:55linuxstbwodz: I think just change the #else to #elif !defined(IAUDIO_M3) (and then remove the outer #ifndef IAUDIO_M3)
12:42:03linuxstbIIUC, it's because the M3 doesn't have a main LCD.
12:42:57wodzlinuxstb: ok
12:45:09wodzn1s: I would like to postpone the merge. First I would like to have optimized writes working
12:45:29 Join liar [0] (~liar@213162066168.public.t-mobile.at)
12:45:35n1sah, ok
12:45:39linuxstbwodz: Is your app.lds different to the other two Coldfire versions? I'm not saying you should do it, but those two look like they could be merged.
12:47:25wodzit is the same for all 5249 based daps it differs for 5250 based
12:47:39n1swodz: the test_disk.c diff contains +#define TEST_SIZE (20*1024*1024) // hacked other than that i think the test_disk diff can be committed separately (assuming it fixes warnings)
12:47:45*linuxstb wonders about http://pastebin.com/qbXPjse5 in iaudio/app.lds
12:48:22n1slinuxstb: seems useful! :)
12:48:28*linuxstb blames buschel for that gem ;)
12:48:32 Quit M3DLG (Ping timeout: 258 seconds)
12:49:02wodzlinuxstb: I just saw that - it is plain wrong :-)
12:49:24linuxstbwodz: Well, it's logically correct...
12:49:36n1syeah, the x5 has more iram
12:49:43wodzn1s: oops You are right
12:52:32wodzn1s: I can't remember now why I changed %d to %ld - I'll revert this
12:55:41 Join JohannesSM64 [0] (~johannes@cm-84.215.116.196.getinternet.no)
12:56:10n1swodz: the void bootmenu(void) funtion is missing indentation
12:56:26linuxstbbootmenu? ;)
13:00
13:05:34amiconnlinuxstb: Skyeye doesn't support multicore yet though; I would limit my implementation to CPU, and only those i/o modules actually accessed in the elio rom
13:08:43*amiconn first needs to find out how to build skyeye
13:13:53***Saving seen data "./dancer.seen"
13:16:52 Join robin0800 [0] (~robin0800@cpc2-brig8-0-0-cust964.brig.cable.ntl.com)
13:19:06linuxstbwodz: I thought when we discussed the idea of a boot menu, you concluded that it wasn't needed? Or am I mis-remembering?
13:20:39 Nick YPSY is now known as Ypsy (~ypsy@geekpadawan.de)
13:23:01wodzlinuxstb: As far as I remember we didn't come to any final conclusion
13:25:06wodzanyway at this stage of port I don't think that this does matter much
13:25:35n1swe traditionally don't have them
13:26:00n1sbut i don't feel stronlgy about it as long as it doens't popup on every boot
13:26:09linuxstbwodz: The problem is that once things like that go in, it's hard to remove them...
13:26:44linuxstbI'm also not against boot menus in principle, but I think there should be a way for a user to disable it and go immediately to Rockbox without any significant delay.
13:27:10linuxstbIf I understand your menu, there's a 10-second timeout?
13:27:57wodzyes
13:29:00n1samiconn: did you see linuxstb's pastebin, seems all iauduio
13:29:11n1ss use 96kB iram now...
13:29:37wodzBootloader need rework anyway as it can corrupt filesystem in bootloader USB mode
13:30:41wodzI am going to work it out after adding support for FM, LINEIN, MIC and recording
13:32:21 Quit TillW (Quit: This now concludes our broadcast day.)
13:33:09amiconnn1s: That's correct. We used to have an even split of iram between core and plugins/codecs, but Buschel changed that in favor of plugins/codecs in r25057
13:33:11n1sam i correct thinking that only the x5 has the 5250 with 128kB
13:33:28amiconnThis ifdef looks like an oversight in that commit
13:33:42n1sah, right, my internal hex parser is off it seems, thanks
13:33:44amiconnn1s: No, both the M5 and the X5 have the 5250. The M3 has 5249
13:34:15n1sok, so just a bit of useless linkerscript, no biggie :)
13:36:53*n1s fixes
13:38:01CIA-5New commit by 03nls (r25721): remove some useless lines
13:39:40pamaurygevaerts: time to discuss ?
13:39:50gevaertsnot really
13:39:57*gevaerts has a full-time job!
13:40:23 Join archivator [0] (~archivato@stu0279.keble.ox.ac.uk)
13:40:27*pamaury is having a pause !
13:43:11 Quit kugel (Ping timeout: 260 seconds)
13:43:27 Quit n1s (Ping timeout: 240 seconds)
13:47:16 Join hebz0rl [0] (~hebz0rl@dslb-088-067-206-019.pools.arcor-ip.net)
13:57:35 Quit efyx (Remote host closed the connection)
13:59:18 Join efyx [0] (~efyx@lap34-1-82-225-185-146.fbx.proxad.net)
14:00
14:01:08 Join webguest20 [0] (~5b6281f1@giant.haxx.se)
14:01:50b0hoonis there any person who has the Philips HDD6330 or 6320?
14:02:41 Quit webguest20 (Client Quit)
14:05:18 Join adnyxo [0] (~aaron@adsl-065-013-002-216.sip.asm.bellsouth.net)
14:08:15JdGordonTorne: I really dont think the skin fonts need to bother with loading from a cache file. I would have no problem with the 60k setting font loading from a disk copy and the others using the first 10k of glyphs and then let the cache sort itself out
14:09:17TorneJdGordon: heh..
14:09:27JdGordon?
14:12:31 Quit scorche (Read error: Connection reset by peer)
14:12:55 Join scorche [0] (~scorche@rockbox/administrator/scorche)
14:13:16Tornewell, that might kinda work, but the obvious way to do it will leave ASCII characters as the first to get dropped from the cache
14:14:12JdGordonhows this for a fun idea... have something read every tag on your DAP and keep a count of how often each glyph would be required. Then have it write the stats out to a txt file, then fix bdf2fnt to be able to write the glyphs on disk in an order by those stats?
14:14:21Tornehaha
14:14:30Tornethat could be.. useful :)
14:14:37JdGordonthen fix the font loading to look for <fontmane>-mine.fnt and use that for fontname
14:14:56gevaertsJdGordon: you need to take the rating into account as well
14:14:57 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
14:15:16JdGordonwhich rating?
14:15:19JdGordonplaycount?
14:15:39gevaertsall of them!
14:16:18JdGordonnot really... if you only have one track with umlaut-a and 100000 without, even if that one track is played alot more than the others that glyph wouldnt rate very highly
14:17:15 Quit Zagor (Remote host closed the connection)
14:17:52JdGordonor... how about load the whole .fnt into the audiobuffer and seek in that instead of on disk?
14:18:03JdGordonif we are doing this before audio is going then that actually would work
14:18:04 Nick fxb__ is now known as fxb (~felixbrun@h1252615.stratoserver.net)
14:18:20JdGordon^ actually a workable fix!
14:19:02Torneindeed...
14:19:10JdGordonunifont is only 2mb so this could almost work on all targets
14:19:28Tornefont_load would need to be able to do it both ways, though, of course
14:19:36JdGordonsure
14:19:42Torneunless you were going to insist on stopping playback to change fonts
14:19:48JdGordonhell no!
14:19:52Torneheh
14:20:06Torneanyway, i was thinking that the caching system itself might be easily improved..
14:20:17Torneprincipally, why do we bother to track single glyphs?
14:20:29Torneunicode is helpfully grouped in a useful way already
14:20:51JdGordonwe have to keep track somehow to know which to dump...
14:21:12Torneyah, but i mean, you could load/unload/index glyphs in whole pages
14:21:22 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
14:21:39JdGordonwont that be very wasteful?
14:21:41Torneload characters 0-127 together as a single block, have them be a single entry in the LRU list
14:21:47TorneNot really, no
14:22:05TorneOther than weird symbols, if you have a single character from a given page you probably have a large majority of them
14:22:22JdGordonthat makes sense
14:22:32Torneand loading a page is not much more costly than loading a glyph
14:22:45gevaertsno, but it takes more memory
14:22:47TorneWell yes
14:23:14TorneBut actually, the LRU list takes a little memory anyway :)
14:23:45JdGordoncan we track how many glyphs are used in each page and then has a seperate list of singles which are usead heavily?
14:23:48*Torne shrugs, it was a thought.. my intuition is that it wouldn't reduce the cache hit rate significantly, but this may not be true for everyone
14:24:25gevaertssure, but the system still works (in theory anyway) if you have a huge font where each glyph takes 50K. A Page-based system needs to be able to load a full page, so we'd have to check what that means for all fonts
14:24:33JdGordonit depends how big each page is...
14:25:11JdGordoncan "page" sizes be chosen arbitrarily on load?
14:25:46Tornewell yeah, the block size is arbitrary
14:26:02Tornei was thinking 128 chars because that's the smallest grouping the unicode spec itself uses
14:26:05gevaertstrue
14:26:24gevaertsYou'd go for 128 glyphs or that maximum that fits, whichever is smaller
14:26:40Tornebut anyway, if we did this, it would dramatically reduce the number of reads to fill the cache, without changing the actual save/load procedure at all
14:27:04Torneit wouldn't be 128 chars for all pages, also, the font doesn't necessarily *have* all 128 of those glyphs
14:27:12JdGordonare the .fnt files layed out for this to work?
14:27:21Tornefew fonts have glyphs for 0-31, for example :)
14:27:40TorneJdGordon: they're just in unicode point order, with a table that says which are present, pretty much
14:27:47Torneyou can easily load a range from that
14:27:57Torneer, which are present at what offset
14:28:16JdGordonso you can read a 16K block and know which fonts you are definitly getting?
14:28:24Tornei think so. lemme check.
14:28:58Torneof course if we were going to do this we could always make .fnt files actually get *stored* like this
14:29:12Tornefile index says where each block is, and the blocks have a smaller header with the offset to each char's start
14:33:13Torneyeah, the font currently stores a table mapping unicode char number to an offset into the file, with -1 for chars that are not present
14:33:34Torneif you assume those are always in order, loading a block is trivial
14:44:14 Join kugel [0] (~kugel@rockbox/developer/kugel)
14:44:26*kugel still thinks dumping the font buffer(s) as is is the easiest and best solution
14:44:28 Join Schmogel [0] (~Miranda@p3EE22D2F.dip0.t-ipconnect.de)
14:46:17Tornekugel: that's not mutually exclusive
14:46:29 Quit liar (Read error: Connection reset by peer)
14:46:33Torneso, write the code for that? :)
14:46:50kugelI don't experience the problem! :)
14:47:00Tornenor do I
14:47:08Tornei use a smaller font normally ;)
14:47:19Tornei only noticed it when i was benchmarking with default settings, because of cabbie ;)
14:55:40 Quit JohannesSM64 (Ping timeout: 276 seconds)
14:59:42wodzI splitted mpio port into 3 patches and uploaded it to FS #11137
14:59:50wodzcould I ask for final glance?
15:00
15:00:10 Quit lifeless (Ping timeout: 260 seconds)
15:01:34kugelwodz: I think mkmpioboot should be under rbutil/
15:04:46wodzhmm
15:04:58wodztools for irivers are in tools/
15:06:20linuxstbHistorically, everything was in tools. But things not required by the build system shouldn't be there any more.
15:06:47kugelthe reason is that rbutil wants the link library versions mk*boot into the binary
15:07:54wodzI see
15:07:54 Part LinusN
15:08:30 Join lifeless [0] (~lifeless@188.17.70.234)
15:09:44 Join JohannesSM64 [0] (~johannes@cm-84.215.75.42.getinternet.no)
15:09:59 Join CGL [0] (~CGL@190.207.195.211)
15:11:01 Join LinusN [0] (~linus@rockbox/developer/LinusN)
15:11:56wodzshould I change somehow mkmpioboot or simply move it?
15:13:08linuxstbI would just move it for now. Integration with rbutil can come later.
15:13:54***Saving seen data "./dancer.seen"
15:15:02 Join M3DLG [0] (~M3DLG@212.183.140.55)
15:20:41linuxstbwodz: The convention (rightly or wrongly) is to use the 4-char string that is stored in the file header as the parameter to the -add option. IIUC, you're still only storing 4 chars in the file, but using a 5-char string as the argument?
15:21:09 Quit M3DLG (Ping timeout: 240 seconds)
15:21:56*linuxstb thinks he was looking at an old version of the patch...
15:22:33linuxstbwodz: Ignore that last sentence. I think you forgot to change modelnum from 170 to 69 in scramble.c though
15:23:30wodzehh, yes I missed scramble.c
15:25:16 Join evilnick [0] (~0c140464@rockbox/staff/evilnick)
15:26:08kugelwodz: I skimmed through it and it looks largely fine to me
15:27:56linuxstbwodz: Do you need the bootloader converted to "rockbox.mpio" ? If mkmpioboot just uses the raw bootloader.bin, then you can do "bootoutput=bootloader.bin" and "boottool=cp"
15:29:52wodzlinuxstb: good point
15:30:18linuxstbAlthough from a safety point of view, it's nicer if the mk*boot tool checks for a file header/checksum.
15:31:05linuxstbSo you could use something like bootloader-mpiohd200.mpio and keep boottool as scramble -add
15:31:20linuxstbBut again, that's something you could tweak later if you wanted.
15:31:47wodzlinuxstb: we are talking about tools/configure right?
15:32:20linuxstbYes, you can add boottool and bootoutput there, to override tool and output
15:33:47wodzcurrent mkmpioboot uses raw binary so I go the first route You proposed and make a record on my TODO list to improve mkmpioboot later
15:35:22 Part LinusN
15:40:35 Join funman [0] (~fun@rockbox/developer/funman)
15:48:06 Join Jaykay [0] (~chatzilla@p5DC578C3.dip.t-dialin.net)
15:49:53 Quit lifeless (Ping timeout: 260 seconds)
15:50:30 Nick ANNA is now known as jnss (janes@gateway/shell/sign.io/x-vwlduembtdtjdmud)
15:50:45 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
15:53:27 Join lifeless [0] (~lifeless@83.219.8.54)
15:58:42 Quit lifeless (Ping timeout: 246 seconds)
16:00
16:00:48funmankugel: RE: fuzev2, adding a delay *before* clearing CCU_IO bit 12 (and after reading the scrollwheel) fixes problems
16:01:10kugelI know, it has been there :)
16:02:07kugelor, I'm sure it was :\
16:02:12funmanyes i removed it
16:02:36funmanget_scrollwheel() calls buttonlight_on() and backlight_on(), that's the only hardware bits it touches
16:03:25kugelI never figured out why a delay in the function prolougue helped so much though
16:03:59funmanperhaps it belongs in the epilogue of another function (lcd comes to my mind)
16:04:57kugelI probably should commit my semi-udelay() for that single purpose, 500 is quite a bummer and it's needed twice there
16:05:08funman500 is not enough
16:05:46 Join xiainx [0] (~xiainx@modemcable195.238-202-24.mc.videotron.ca)
16:05:48 Join MethoS- [0] (~clemens@134.102.106.250)
16:06:25kugelperhaps do it in a different way: read buttons after lcd updates and not during the tick tasks if an update is currently running
16:06:47kugelI imagine you don't need a huge delay if the lcd is not currently updating
16:10:29 Join krazykit` [0] (~kkit@ppp-70-225-135-39.dsl.ipltin.ameritech.net)
16:10:29 Quit krazykit (Read error: Connection reset by peer)
16:10:51 Quit wodz (Quit: Leaving)
16:11:51JdGordonanyone got a <16bit target handy?
16:12:28funmanonly 1bit (clips)
16:12:46JdGordoncan you quickly check how much skin buffer cabboe uses please?
16:13:10jnssregarding the sansa e200 as microsdhc capability was added three years ago, will the device recognize 16gb sdhc cards?
16:13:39funmanjnss: it recognizes all sdhc cards
16:13:54jnssfunman, so no limit to the size of cards
16:14:10funmanno limits in any way
16:14:27jnssgreat, thanks
16:14:35 Join lifeless [0] (~lifeless@90.151.43.12)
16:16:44funmanJdGordon: where is this info? didn't see it in buffering thread (although it might be missing due the clip only having 8 lines of text)
16:16:59JdGordonsystem > rockbox info > last line
16:17:19funman3.8 / 10
16:17:57JdGordonta
16:21:34ranmafunman: Regarding the AUDIOSET2/AUDIOSET3 values, AUDIOSET2 = 6 means Speakeroutput driver set form >= 16 Ohm loads and 50% bias current reduction
16:22:33ranmaWith AUDIOSET3 I set the IBR_HPH to 1 (Headphone output for >= 32 Ohm loads)
16:22:35funmanok, but i've no idea what an ohm load means for headphones, or a 'bias'
16:23:05funmanif the Ohm value is higher, would it mean lower sound output ? (and lower battery usage)
16:23:24ranma'Load' just means what's connected to the amp output terminals
16:23:38ranmaNormal headphones have around 32 Ohms
16:23:45 Quit lifeless (Ping timeout: 245 seconds)
16:24:27ranmaAt least portable ones, bigger expensive high-quality headphones have more like 600 Ohms sometimes IIRC
16:24:58ranmaThe volume is unchanged by the value, I checked that.
16:25:22funmanjust wondering how this affects sound quality/volume/whatever, if we should apply these changes, and also if we should apply them for other targets using as3514
16:25:23ranmaIBR_HPH also points to I_Bias_Reduction I'd say :)
16:25:41ranmaBias current most likely means the AB amp's bias current
16:26:09ranmaThis could affect cross-over distortion and high frequency accuracy I think
16:26:16vedthis value matters for fitting the output of amp with input of speakers, they should match to minimise reflections, at least that would be the case in high frequencies
16:26:32vedand in terms of impedance, not just resistance
16:27:07funmansince we can't change the speakers (headphones) anytime, then it's not worth trying to fit with them?
16:27:15funmans/can't/can/
16:27:53ranmaI think 16 Ohms is rather uncommon and I just changed the setting to what the original firmware was using :)
16:28:19ranmaThe DAC bias current is also reducable, didn't try that one...
16:28:50funmanranma: btw about the GPIO registers returning 0 : it would make sense if pclk is disabled
16:29:32 Join TillW [0] (~Till@h92-net09.simres.netcampus.ca)
16:29:41funmanhm but the direction registers can be read ..
16:29:50ranmaAn another note, I received FlynDice broken Clip+ yesterday
16:29:53funmancan you try to read GPIO_BASE+0x3FC instead ?
16:30:33ranmaHmm, Indeed, I should try that.
16:30:57funmando you want to solder it to a JTAG adaptor?
16:31:41funmanI noticed that the m200v4 and c200v2 (both 7 pins) share the same layout, e200v2 and Clipv1 have a 8 pin layout (i don't remember if I saw pictures of a disassembled Clip+)
16:31:56ranmaYeah. Found the JTAG pads (8 pads on the backside), am going to solder wires to it tomorrow
16:31:57archivatorIs there a non-callback recording API? Also, are there any docs about it? E.g. which thread calls the callback and so forth?
16:32:39funmanarchivator: i don't think there is any docs beside the source code
16:32:51ranmaIt comes up in bootrom-mode and shows up as a 4MB usb-storage, but writes don't seem to do anything
16:33:28funmanwhat can you read from it ?
16:34:07archivatorfunman: that's a pity. Oh, well, I don't need recording in FFT anyway :)
16:34:08ranmaSome repeating pattern
16:35:30funmanarchivator: what's wrong with a callback ?
16:36:23funmanyou could emulate synchronous calls with yield() + atomic operation on some variable modified by the callback
16:36:44ranmaBTW, the clip+ board is sticks to the display using some kind of double-sided adhesive tape, so it doesn't come off easily after removing the screws...
16:36:52ranma-is
16:38:09archivatorfunman: I wan't to keep the current 2-thread model. I was hoping I would just need to add a new entry point to the thread and simply switch between the two. However, I would have to rewrite a bit more than I imagined if I wanted to add a callback.
16:38:38archivatorBesides, I'm not even sure which thread calls the callback..
16:38:46 Quit xiainx (Quit: Good Bye)
16:39:06funmangrep doesn't help ?
16:40:42archivatorfunman: well, the pitch plugin obviously uses recording but reading the TODO, he doesn't seem to be using the callback properly at all.
16:40:50funmanarchivator: it's not called from a thread, it's called from the recording ISR
16:40:56 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
16:42:30 Quit robin0800 (Ping timeout: 245 seconds)
16:42:40archivatorfunman: thanks but I'm not sure I know what to do with that info.. In any case, I'll leave this for another day.
16:42:50 Join robin0800 [0] (~robin0800@cpc2-brig8-0-0-cust964.brig.cable.ntl.com)
16:44:40funmanit means in the callback you can use any threads-related functions like mutexes, or yield(), and that you must return as fast as possible
16:48:55archivatorfunman: alright, that much I understand. Now, how do I use recording in the sim? Any way to trick it to "record" from a file?
16:48:59 Join M3DLG [0] (~M3DLG@212.183.140.17)
16:49:54CIA-5New commit by 03funman (r25722): Fuzev2: fix screen corruption ...
16:51:04funmanoops i forgot to test fps after adding the delay
16:51:52JdGordonranma: how hard is it to get to the jtag pins on the cip+?
16:51:56*JdGordon wants his unbricked
16:52:08funman=> less than 1%
16:52:33Torneranma: also do you know if the clipv2 has similar/identical jtag? :)
16:52:51funmanTorne: do you have pictures of the clipv2 board?
16:52:58 Join ollebe [0] (~olle@root.rot.sgsnet.se)
16:53:08TorneI don't personally, no
16:53:13Tornei've not opened mine
16:53:20Torneor really used it much :)
16:54:15funmanhm i can crash the Clip+ simulator when trying to record :o
16:54:46ranmaJdGordon: I think you have to be careful to not break anything while removing the pcb from the case to get a the pads.
16:54:53ranmaTorne: Dunno
16:55:15 Part b0hoon ("Back to work.")
16:56:50Tornewe dont' seem to have a pic of the clipv2/clip+ pcb
16:57:19ranmaranma/S6001777.JPG">http://uguu.de/~ranma/S6001777.JPG
16:57:20Tornehow easy is it to take apart? :)
16:57:27funmanvery
16:57:42ranmaranma/S6001783.JPG">http://uguu.de/~ranma/S6001783.JPG
16:57:42Torneranma: which ones are the jtag there?
16:57:54ranmaranma/S6001790.JPG">http://uguu.de/~ranma/S6001790.JPG
16:58:21ranmaUpper right near the display cable in S6001783.JPG
16:59:08funmanhalf between the hole and the cable ?
16:59:43Tornei may pop my v2 open later then and scan it :)
17:00
17:00:38funmanarchivator: i thought recording worked fine in the sim, but it crashes for me (perhaps due to my recent install of pulseaudio?)
17:01:06 Quit CGL (Ping timeout: 276 seconds)
17:02:50 Join CGL [0] (~CGL@190.207.195.211)
17:04:41funmankugel: it's weird that we must wait when reading buttons because we aren't touching dbop afaict
17:04:52 Quit MethoS- (Remote host closed the connection)
17:05:47kugelI think gpio c/d are still wired to the lcd somehow as on as3525, just that CGU_IO bit 12 is another way of setting GPIOx_AFSEL
17:05:52kugelb/c*
17:06:46funmananything missing for fs#11172 (read scrollwheel from isr) ?
17:07:49CIA-5New commit by 03funman (r25723): Fuzev2: button_read_dbop() doesn't exist
17:07:54ranmafunman: Yeah, the 2x4 padrows
17:08:11funmankugel: ^
17:09:43archivatorfunman: it doesn't do anything here either. It doesn't crash, though, just doesn't record.
17:10:05kugelfunman: I don't think so
17:10:11 Quit Zagor (Quit: Leaving)
17:10:21kugelI meant to add a user setting for the sensitivity but that can be done later
17:10:35funmanarchivator: ah right, there are only empty stubs
17:11:30archivatorfunman: that's not very helpful :-/ Oh, well, at least I found about recorder/pcm_record.c - looks like a good reference.
17:12:16pamauryWhat do you people think about the way to select between UMS and MTP ? And on top of that, select HID or not. How do you think it should be handled ? I need to implement somthing before being able to commit something and the discussion never ends up with a solution each time I ask :)
17:13:08 Join xiainx [0] (~iain@modemcable195.238-202-24.mc.videotron.ca)
17:13:56***Saving seen data "./dancer.seen"
17:15:22 Quit M3DLG (Ping timeout: 246 seconds)
17:15:26archivatorfunman: it seems like SDL itself does not support recording. Adding that to the sim would mean another dependency (portaudio, perhaps?)
17:18:09linuxstbarchivator: Hmm, that's unfortunate...
17:18:27 Join n1s [0] (~n1s@rockbox/developer/n1s)
17:19:33funmanpamaury: perhaps commit first and discuss after then
17:19:52archivatorlinuxstb: one of the first results google gives is a patch against sdl 1.2.4 but I guess that's ancient now
17:20:17pamauryYou mean, I commit it with the current system ? (I currently have a debug entry, and one need to select this debug entry and then hold select during usb insertion)
17:20:53pamauryarchivator: I *think* 1.3 will support it but the question is when it will be out :)
17:22:31kugelalso if rockbox will run under it, I once tried a snapshot and it didn't
17:22:57kugelthe alsa pcm driver could support recording, I haven't added it though
17:24:10kugelis there no tick_task equivalent in linux? :(
17:24:17 Join emrecelikten [0] (~58f08cca@gateway/web/freenode/x-djbcrnoocvkilkqu)
17:24:31pamaurywhat does tick_task does ?
17:24:41 Join nls [0] (~n1s@113.166.0.45)
17:24:52pamaury*do
17:24:58archivatorIs it reasonable to use global_settings.rec_* in a plugin? I mean, can they be guaranteed to be valid?
17:24:59kugeladd a callback for the period timer interrupt
17:25:33kugelarchivator: should be pretty safe if you do rb->global_settings->rec_*
17:26:04ranmaYou have times in Linux, with tickless kernels it's no longer certain you have a tick
17:26:14ranmas@times@timers@
17:26:30archivatorkugel: that's what I meant, yes. thanks.
17:26:46 Join kugel_ [0] (~kugel@e178067065.adsl.alicedsl.de)
17:27:00 Quit kugel (Disconnected by services)
17:27:04 Nick kugel_ is now known as kugel (~kugel@e178067065.adsl.alicedsl.de)
17:27:10 Quit kugel (Changing host)
17:27:10 Join kugel [0] (~kugel@rockbox/developer/kugel)
17:27:16kugelranma: different names for the same thing basically :)
17:27:25 Quit n1s (Ping timeout: 268 seconds)
17:27:46kugelanyway, I found setitimer, but I find it a bit unfortunate that it works via SIGALRM which interrupts sleep()
17:28:24ranmaAh, I was thinking of kernel space there...
17:28:29pamauryhuh, iirc sigalrm only works with seconds, not under
17:28:55kugelranma: I found nothing in that direction, do you know more?
17:29:01 Quit pamaury (Quit: Quitte)
17:29:09kugel(is trying to think of how to replace the sdl timers with linux' ones)
17:30:13ranmaFor userspace I only know SIGALRM, I don't see why setitimer shouldn't work with sub-second resolution
17:30:25 Quit lpereira (Ping timeout: 245 seconds)
17:30:48B4gderpthreads has some kind of timers, doesn't it?
17:31:01gevaertskugel: timer_create() and timer_settime() and friends?
17:31:16funmankugel: create_timer() ?
17:31:26funmanoops, what gevaerts said
17:31:45kugelranma: setitimer does
17:32:27kugelbut it works by signalling SIGALARM to the process (i.e. you add the callback with sigalarm/sigaction), which breaks sleep()s. I think the same applies for timer_create()
17:32:58 Join Luca_S [0] (~5d3fc54b@giant.haxx.se)
17:33:08 Join mt [0] (~mtee@rockbox/developer/mt)
17:33:22kugelwell, we don't call sleep() anyway so it hopefully won't be a problem
17:34:20kugelthe relating manpages are horrible
17:35:21B4gderand signals and threads don't match very good
17:37:35*FlynDice sees ranma has recieved the package and regains some lost confidence in the US Postal Service... Happy Hacking!!
17:38:29 Join b0hoon [0] (~quassel@xdsl-7253.bielsko.dialog.net.pl)
17:39:18ranmaI didn't sign for the package receival though, I was just lying in the entrance when i came home on sunday :)
17:39:36ranma'it was' of course ^^;
17:40:08 Quit funman (Quit: free(random());)
17:40:14FlynDiceBy the way, that's not me in those pictures and I've never seen those goats before......
17:41:24 Quit nls (*.net *.split)
17:41:24 Quit domonoky (*.net *.split)
17:41:24 Quit scorche (*.net *.split)
17:41:24 Quit bzed (*.net *.split)
17:41:24 Quit Topy (*.net *.split)
17:41:25 Quit markun (*.net *.split)
17:41:25 Quit slck (*.net *.split)
17:41:25 Quit parafin (*.net *.split)
17:41:58 Join scorche [0] (~scorche@rockbox/administrator/scorche)
17:42:27 Join nls [0] (~n1s@113.166.0.45)
17:42:27 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
17:42:27 Join bzed [0] (~bzed@devel.recluse.de)
17:42:27 Join Topy [0] (~topy@my.fastsh.it)
17:42:27 Join markun [0] (~markun@rockbox/developer/markun)
17:42:27 Join slck [0] (Venci@Slackware.SlackPix.Com)
17:42:27 Join parafin [0] (parafin@paraf.in)
17:50:22 Join lifeless [0] (~lifeless@188.18.74.67)
17:56:33 Quit B4gder (Quit: It is time to say moo)
17:57:21 Quit petur (Quit: moo)
17:57:30kugelmeh, they all seem to work with signals
17:59:46 Quit lifeless (Ping timeout: 258 seconds)
17:59:47 Join panni_ [0] (hannes@ip-95-222-52-93.unitymediagroup.de)
18:00
18:00:48 Quit nls (Ping timeout: 276 seconds)
18:01:08 Join pamaury [0] (~c2c7a50a@rockbox/developer/pamaury)
18:01:08 Join lifeless [0] (~lifeless@89.20.121.185)
18:06:39 Quit Topy (Ping timeout: 276 seconds)
18:08:25 Join Topy44 [0] (~topy@my.fastsh.it)
18:10:56 Join Galois [0] (djao@efnet.math.uwaterloo.ca)
18:11:27 Quit Galois (Client Quit)
18:11:41 Join komputes [0] (~komputes@ubuntu/member/komputes)
18:12:27 Quit b0hoon (Quit: No Ping reply in 180 seconds.)
18:12:35 Join Galois [0] (djao@efnet.math.uwaterloo.ca)
18:12:39 Join nls [0] (~n1s@113.166.0.45)
18:12:48 Quit TillW (Remote host closed the connection)
18:12:56 Join b0hoon [0] (~quassel@xdsl-7253.bielsko.dialog.net.pl)
18:20:56 Quit b0hoon (Quit: No Ping reply in 180 seconds.)
18:21:35 Join b0hoon [0] (~quassel@xdsl-7253.bielsko.dialog.net.pl)
18:21:41 Quit n17ikh (Ping timeout: 268 seconds)
18:23:49pixelmapamaury, gevaerts: speaking of USB - would it be possible to probe the host OS if it can deal with USB composite device correctly? Just an idea I had because switching HID on and off ge ts a bit annoying (using my c200 on XP and MacOS 10.4)
18:25:06pamaurypixelma: not easily not. You can probably determine which OS it is by looking at the sequence of descriptors it ask (for example Windows doesn't ask other speed descriptor iirc). But that's not reliable imo.
18:25:08 Join toffe82 [0] (~chatzilla@12.169.218.14)
18:25:48pixelmaa pity
18:25:50 Join n17ikh [0] (~n17ikh@host-69-59-126-212.nctv.com)
18:26:13pamauryBut perhaps gevaerts knows a way. Normally, XP handles composite devices
18:27:29S_a_i_n_texcept when it comes to HID on the Nano2g for some reason...
18:27:30pixelmayes, but MacOS 10.4 doesn't :\
18:27:36*S_a_i_n_t shakes a fist at HID again.
18:28:01 Join Zarggg_ [0] (~zarggg@65-78-69-194.c3-0.eas-ubr6.atw-eas.pa.cable.rcn.com)
18:28:31pamauryHum, perhaps MacOS would work better with interface association descriptors...
18:28:51 Quit Zarggg (Read error: Connection reset by peer)
18:30:27pixelmanewer MacOS can deal with it, although my impression was that controls are not as smooth as they are on XP but I only tried 10.5 once and haven't tested 10.6 at all
18:31:12pixelmabut if I don't turn off HID on 10.4 then I don't get a USB data connection
18:32:10pamauryWhat do you mean ? It probe the devices but acts as if it didn't see the UMS interface ?
18:32:10 Quit Luca_S (Quit: CGI:IRC (EOF))
18:32:58 Quit b0hoon (Quit: No Ping reply in 180 seconds.)
18:33:22 Join b0hoon [0] (~quassel@xdsl-7253.bielsko.dialog.net.pl)
18:33:22 Quit b0hoon (Client Quit)
18:36:04pamauryIf I had a Mac, I could test some things but without it that's quite difficult. And it's your fault it you use a buggy OS with an obsolete version ;)
18:37:01S_a_i_n_tOSX Burn!!! ;)
18:37:42saratogathe clip+ JTAG pins look downright accessible
18:37:47saratoganice of sandisk to give us a break
18:38:11pamaurypixelma: let me think about it, I have a idea in mind, I'm recalling something but I need research to check it if feasable
18:39:09 Quit komputes (Ping timeout: 276 seconds)
18:39:52 Join M3DLG [0] (~M3DLG@212.183.140.33)
18:41:05 Join TillW [0] (~Till@h92-net09.simres.netcampus.ca)
18:43:05 Quit JohannesSM64 (Ping timeout: 258 seconds)
18:46:38pixelmapamaury: if I have HID enabled then I can't access the device, and HID controls usually don't work. There's one weird exception though if I let a program run called "USB prober" HID works at least but I still can't access the device
18:48:38 Quit arbingordon (Quit: `)
18:50:54 Quit robin0800 (Remote host closed the connection)
18:52:12 Join komputes [0] (~komputes@ubuntu/member/komputes)
18:54:42 Join bertrik [0] (~bertrik@ip117-49-211-87.adsl2.static.versatel.nl)
18:55:00 Quit kugel (Read error: Operation timed out)
18:55:17 Quit TillW (Remote host closed the connection)
18:58:38linuxstb(taking the conversation from -community) - shouldn't commits get more "testing" than just "it compiles"... ?
18:59:15saratogaisn't that what our users are for :)
18:59:20 Join JohannesSM64 [0] (~johannes@cm-84.215.116.196.getinternet.no)
19:00
19:00:03*S_a_i_n_t always thought that all commits were supposed to be cross-target tested *before* commit.
19:00:05archivatorlinuxstb: sure but don't forget that not all developers are committers.
19:00:17*domonoky thinks all patches/changes should get tested on at least 1-2 targets before commiting.
19:00:18S_a_i_n_tI'm sure the wiki says this *should* happen...but it never does.
19:00:24linuxstbarchivator: Err, so?
19:00:35 Join TillW [0] (~Till@h92-net09.simres.netcampus.ca)
19:01:00pixelmaS_a_i_n_t: I do, and I help out testing a lot
19:01:17pixelmaI don't see users as testers
19:01:38domonokyexpirience shows that what isnt tested, doesnt work.
19:01:47S_a_i_n_tIt's usually, "commit...RED...crap...commit...FIXRED"
19:01:55archivatorlinuxstb: take me, for example. I wrote the fft plugin, do something rbutil related every once in a while but don't have committer access. Hey, I'm not complaining, I prefer using FS# - keeps me disciplined and moves some of the weight off my shoulders. But every once in a while, I have to test something on a multitude of targets and it takes *long* to do it by hand.
19:03:33bertrikif you test-compile a HWCODEC target + sim, a SWCODEC + sim and a touchscreen target + sim, you have a lot of targets covered
19:03:52nlsS_a_i_n_t: well, testing if something compiles and testing if it works is two different things
19:04:10nlsbertrik: it all depends on what you changed
19:04:24*S_a_i_n_t realises this...hence the statement ;)
19:04:24bertrikof course, but those cover a lot
19:05:15 Quit MagusG (Ping timeout: 276 seconds)
19:05:29S_a_i_n_tarchivator: You wrote the fft plugin?
19:05:38archivatorS_a_i_n_t: I did.
19:06:08archivatorNow you know who to blame for the mess that it is. :)
19:06:28S_a_i_n_tYou've reminded me of something..."bars" on "linear" displays nothing onthe Nano 1/2g
19:06:45S_a_i_n_tPretty sure Llorean mentioned this on some other target also.
19:07:13S_a_i_n_twhereas on dB, it's fine.
19:07:18archivatorS_a_i_n_t: I'll take a look, it's probably a division somewhere.
19:07:51S_a_i_n_tI *could* have linear/dB mixed up...I don't have a target on hand.
19:07:52archivatorS_a_i_n_t: for various definitions of "fine" - it's probably edgy since it just clips instead of properly handling overflow :(
19:08:00S_a_i_n_t*one* of them doesn't work lol ;)
19:08:10S_a_i_n_totherwise, I think it's awesome.
19:10:03 Join Brewtality [0] (~48e1b9c7@giant.haxx.se)
19:10:37archivatorS_a_i_n_t: you know, there is a valid scenario where that could happen with linear. If there's a single spike, it might get zeroed out during the averaging but (I think) it would still influence the scaling. Meaning, everything would display as either 0 or 1 pixel at the bottom. That's obviously wrong but was done to save an iteration through the results (reuse a loop).
19:11:28Brewtalityhey, sorry for jumping in here all noobish, but i had questions about getting Rockbox on my 240 gig iPod 5.5
19:11:44S_a_i_n_tIn that case, would it happen every single time, or just seemingly randomly?
19:11:55S_a_i_n_tAs, I've actually never seen it work...
19:12:16S_a_i_n_tdBs though (I'm fairly confident it's dBs) works fine.
19:12:41S_a_i_n_tI like it a buttload more than the peakmeter.
19:12:45archivatorS_a_i_n_t: should be pretty much random. It might be something to do with the screen size, then. I'll take a look when I'm done with this recording nightmare.
19:12:47linuxstbBrewtality: What's your question? You just need to install the bootloader from here - http://forums.rockbox.org/index.php?topic=24570.0 - plus a Rockbox "current build".
19:13:31S_a_i_n_tarchivator: It is hardly crucial ;) I just thought I'd mention it since it popped into my head right then, and you're here :D
19:13:52archivatorS_a_i_n_t: no worries, thanks for reporting.
19:13:58***Saving seen data "./dancer.seen"
19:14:07S_a_i_n_tI obviously didn;t find it annoying enought to lodge a bug for it, so I obviously kept forgetting about it ;P
19:14:38S_a_i_n_t-1 obviously somewhere in there...
19:14:46Brewtality guess I am just very confused by the steps i need to take. so i should download the bootloader and ipodpatcher from the thread, plus the current build for a iPod 80gig from here: http://build.rockbox.org/
19:15:10linuxstbBrewtality: Yes.
19:15:20 Quit M3DLG (Ping timeout: 268 seconds)
19:15:56linuxstbBrewtality: Is your ipod currently functioning normally with the Apple firmware?
19:16:08Brewtalityyes
19:16:17linuxstbThen you're good to go.
19:20:31Brewtalityso I just extract the latest build to the iPod, run the patcher using the bootloader from that thread, then eject the disk, and i'm done?
19:23:08S_a_i_n_tyep
19:23:34S_a_i_n_t"./ipodpatcher.exe -a bootloader-youripodmodel.ipod"
19:24:13S_a_i_n_tmaking sure that ipodpatcher and the bootloader are in the same dir of course.
19:27:42bluebrotherdoes the h300 bootloader boot into the OF when Rockbox is not found (like the h100 bootloader does)?
19:30:19 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
19:31:05BrewtalityHey, it worked. All loaded up. Thanks a ton guys!
19:31:10 Join jgarvey [0] (~jgarvey@cpe-065-190-066-089.nc.res.rr.com)
19:31:18S_a_i_n_tNo worries, enjoy ;D
19:32:07Brewtalitynow, time to find a theme i like
19:32:36S_a_i_n_tThere is *plenty* on themes for the Video on the themesite l)
19:32:48S_a_i_n_t* ;) rather
19:32:58 Join Luca_S [0] (~5711b744@giant.haxx.se)
19:34:01*bluebrother found it in the manual
19:34:19 Join Kitr88 [0] (~Kitar_st@BSN-143-102-221.dial-up.dsl.siol.net)
19:35:00nlsbluebrother: but can that be trusted? ;)
19:35:27 Quit nls (Quit: Lämnar)
19:36:41 Nick fxb is now known as fxb__ (~felixbrun@h1252615.stratoserver.net)
19:37:32 Quit Kitar|st (Ping timeout: 268 seconds)
19:39:02 Quit Kitr88 (Ping timeout: 276 seconds)
19:47:36 Join Kitar|st [0] (Kitar_st@BSN-210-226-207.dial-up.dsl.siol.net)
19:52:34 Join kugel [0] (~kugel@rockbox/developer/kugel)
19:56:18 Quit jordan` (Ping timeout: 264 seconds)
19:59:48 Quit CGL (Remote host closed the connection)
20:00
20:00:32 Join liar [0] (~liar@clnet-p09-185.ikbnet.co.at)
20:03:22*amiconn thinks the glyph group idea won't work well
20:07:25 Join wodz [0] (~wodz@chello087206240004.chello.pl)
20:09:10 Join Boldfilter [0] (~Boldfilte@adsl-82-101-227.jax.bellsouth.net)
20:12:05 Quit Brewtality (Quit: CGI:IRC (EOF))
20:23:41 Join syrius [0] (~syrius@unaffiliated/syrius)
20:24:23syriushow do I create playlists for rockbox on ubuntu without having to edit the playlist file so it can play on the device?
20:24:39syriuslike I have to delete /media/ROCKBOX/
20:24:56 Quit bug2000 (Ping timeout: 260 seconds)
20:26:25S_a_i_n_tsyrius: Why not just create the playlist on the device?
20:27:19syriusI don't want to do that
20:27:23syriusit takes forever
20:27:40syriusalso it wastes battery power
20:27:43S_a_i_n_tfaster than making it on a PC, transfering it over etc.
20:28:01archivatorHuh, can anyone confirm that fft from svn works on the sansa *200?
20:28:13syriusyou can to name the playlist
20:28:16syriusthat takes awhile
20:28:24syriusunless you don't care what it is called
20:28:24archivatorbasically, anything PP with limited memory.
20:28:44pixelmasyrius: you shouldn't have to delete this part of the path (if I understand your question correctly), I believe Rockbox still tries to find the subfolders. I am not sure with this part of the linux paths, did you try?
20:29:10S_a_i_n_twhatever...since you're going to have to create the playlist by hand, I'd say it's going to be a lot faster on the device.
20:29:32amiconnYes, path stripping also works for ordinary path components (linux and windows), not just drive letters
20:29:48syriusic
20:30:03pixelmaS_a_i_n_t: have you ever tried creating an m3u playlist on the PC?
20:30:16kugelarchivator: it shows the same problems as the fuze
20:30:22syriuswell perhaps last time I tried it had a bug
20:30:25kugeli.e. what S_a_i_n_t reported
20:30:27archivatorkugel: that is?
20:30:30archivatorah.
20:30:31syriusor did something wrong
20:30:45archivatorkugel: does it display anything in any of the modes?
20:30:56kugelit has a hamm/hamming window on the e200 (I haven't found that on the fuze so far). what does that do?
20:31:01archivator'cause I think I know the cause (it wasn't me, I swear!)
20:32:18S_a_i_n_tpixelma: yes, but once you factor in making the playlist, connecting the device, transfer etc...I find it a lot faster/easier to just make the playlist on the device personally.
20:32:19archivatorkugel: two different window functions. Kinda there just for the kicks, since it's not a moving window any way.
20:32:19kugellog scale bars does not work, spectral view linear scale looks strange also
20:32:42kugelI think the amplites are too low in the linear scale
20:32:50archivatorkugel: can you try it with 24613 reverted?
20:33:43archivatorkugel: actually, that might not work.
20:33:54archivatorback to square one, no need for testing.
20:34:34pixelmasyrius: what player do you have?
20:35:09 Quit sevard (Ping timeout: 252 seconds)
20:35:18 Join petur [0] (~petur@rockbox/developer/petur)
20:35:18syriussweet nevermind guys
20:35:21syriusit works now
20:35:29syriusI tried a year ago I think
20:35:49syriuswith rythmbox it didn't work
20:35:56linuxstbAny forum admins (scorche?) know if the length of URL that the forum software truncates is configurable? It's very annoying to see them cut off...
20:35:57 Join sevard [0] (sev@216.164.6.24)
20:35:57syriusnow it works
20:36:16pixelmaI mean your Rockbox device
20:36:31syriusyes on my rockbox device it works now
20:36:34syriusthe playlist
20:36:51syriusI tried last time and it didn't work
20:36:54syriusdunno why
20:36:56syriusnow it works
20:36:57pixelmawhat Rockox player do you have?
20:37:15syriusipod greyscale 4gen
20:37:24syriusI am using hte current release
20:37:27syrius3.5 I think
20:37:54syriusyou will always have the manual install available right?
20:38:05syriuscause I don't want to have to learn how to do it a new way
20:39:40syriusbadass
20:39:43syrius:)
20:39:49saratogaarchivator: did you see we added an FFT to the codeclib?
20:40:18archivatorsaratoga: I did, a long time ago, just haven't worked on the plugin all that time.
20:41:34archivatorsaratoga: as I said the last time we discussed it, the fft is not the bottleneck - the lack of PCM data is. Therefore, unless I find a way to get more data, there's hardly any point in doing the transform faster (won't increase perceivable speed).
20:41:35syriusrockbox rocks
20:41:41wodzam I correct that rockbox checksum added with add method in scramble.c is calculated as a checksum = modelnumber + sum of all unsigned bytes of file?
20:42:09syriushow much should I donate to rockbox if I get paid to install rockbox on people's devices?
20:42:42syriuslike if I get paid 50 USD should I donate 10 USD?
20:43:03 Join M3DLG [0] (~M3DLG@212.183.140.32)
20:43:08syriusI think I'll do that
20:43:09gevaertssyrius: as much as you like. If you get paid to do this, you should probably pay attention to the source distribution clauses in the GPL though
20:43:33syriusI don't sell rockbox
20:44:02syriusI just get paid to install and install an open source media player and show them how to do stuff
20:44:03S_a_i_n_tStill, if you distribute binaries, you need to provide the source if it is asked for.
20:44:12syriusI see
20:44:17syriuso yes
20:44:23domonokysyrius: but you distribute, so make sure you comply with the licence. And donations are not needed, but its nice, then we can drink more beer at devcon :-)
20:44:41syriusdevcon?
20:44:54S_a_i_n_tdevelopers convention
20:44:55syriusdev conference
20:44:56domonokydeveloper meeting..
20:44:57syriusI see
20:45:20syriusare you serious?
20:45:36syriusI only would like to donate if it helps in the development of rockbox
20:45:47S_a_i_n_twell...they don't spend it *all* on beer
20:45:52syriusokay
20:45:54S_a_i_n_t;)
20:46:12gevaertssyrius: the easiest way to work in accordance with the GPL is probably to copy the source archive to the device as well
20:46:22syriusI see
20:46:25gevaertsIf people then delete that, it's up to them
20:46:27S_a_i_n_tAnyway, how would beer not help in the development of Rockbox? :P
20:46:54 Join mikroflops [0] (~yogurt@90-227-45-110-no112.tbcn.telia.com)
20:46:59syriuswell beer can make it harder to concentrate
20:47:20S_a_i_n_tthe source is only what ~35/40MB or so when compressed?
20:47:24syriuswell I would like it that you use it to buy new devices or new hardware to work with rockbox
20:47:38saratogaarchivator: ok couldn't remember if i mentioned that to you
20:48:08syriuswhy can't they just go to the site to use get the sourcecode?
20:48:27syriuswell perhaps I should just put it in a folder called sourcecode
20:48:32S_a_i_n_tbecause *you* need to give it to them, not say "its there, go get it"
20:48:45syriusic
20:49:03syriusokay I certainly don't want to get sued
20:49:05syriusokay
20:49:09gevaertssyrius: have a look at http://www.gnu.org/licenses/old-licenses/gpl-2.0.html section 3. That's the full details
20:49:16syriusyou will be seeing a 10 dollar donation
20:49:33syriusperhaps another one
20:49:39syriusif I get another guy
20:49:51syriuswell we could do it half and half
20:49:53pixelmawe usually get a lot more done than alone during DevCon, it's very motivting to work on something together
20:49:54syriusbut I need $$$
20:49:57gevaertsIf you get paid to install, it probably doesn't count as noncommercial
20:50:03saratogaassuming this is stock rockbox, you could probably just tell them about rockbox.org
20:50:06syriusif the economy was better perhaps it would be half
20:50:14saratogai'm sure we'd rather prefer people see the website then ask you for the source
20:50:20saratoga(assuming you haven't modified it yourself)
20:50:30pixelmamotivating too
20:50:37gevaertssaratoga: that's for noncommercial distribution, I'm really not sure if that applies if you get paid to install it
20:50:40 Join mikroflo1s_ [0] (~yogurt@90-227-45-110-no112.tbcn.telia.com)
20:51:07gevaertssyrius: you don't have to feel bad for every dollar you don't donate. We do this for fun anyway
20:51:11 Quit mikroflops_ (Ping timeout: 276 seconds)
20:51:19syriuswell it isn't just install
20:51:24syriusit is also support
20:51:32archivatorHa, I'm accepted for GSoC. Guess I should start working on flite, then. :)
20:51:33syriusso they know how to use it
20:51:47gevaertssyrius: even better! We don't have to support them then :)
20:51:49S_a_i_n_tarchivator: Mush!
20:51:54*S_a_i_n_t cracks the whip!
20:51:57kugelarchivator: high 5!
20:52:04 Quit liar (Ping timeout: 245 seconds)
20:52:07saratogagevaerts: getting paid to install something doesn't mean you're distributing the software, as I understand it hes just downloading our binary and putting it on the disk, so I think we're still distributing it
20:52:16syriusokay my 10 dollar better go to good use
20:52:18syrius:)
20:52:40 Nick Zarggg_ is now known as Zarggg (~zarggg@65-78-69-194.c3-0.eas-ubr6.atw-eas.pa.cable.rcn.com)
20:52:49saratogaIMO if its our build and our server we are the distributor, although I am not a lawyer
20:53:02syriusokay
20:53:03gevaertssaratoga: I suspect that it depends on how exactly this is done
20:53:04FlynDicedon't worry it'll be good beer I'm sure.
20:53:08syriusI will just put it in a folder
20:53:10syriusleave it that
20:53:21syriusI say that the gpl says that I have to give you source with it also
20:53:25 Join mikroflops_ [0] (~yogurt@90-227-45-110-no112.tbcn.telia.com)
20:53:37syriusand that it is in the sourcecode directory and they can delete that if they don't want it
20:53:44gevaertsthat's fine
20:53:51 Quit flydutch (Quit: /* empty */)
20:54:19 Quit mikroflops (Ping timeout: 240 seconds)
20:54:29 Join liar [0] (~liar@clnet-p09-185.ikbnet.co.at)
20:54:39gevaertssyrius: you can of course offer to delete it for them as well ;)
20:54:46saratogaIMO if you use an official rockbox build, and do not compile it yourself, your are not the distributor regardless of how it comes to be installed, and so GPL compliance is still our responsibility
20:55:28syriusic
20:55:29gevaertssaratoga: that interpretation would make section 3c totally meaningless
20:55:33syriuswell I don't want to get sued
20:55:38syriusfor gpl violation
20:56:12syriusokay making playlist now
20:56:17syriuswell I don't usually make playlist
20:56:23gevaertsDon't worry about getting sued...
20:56:23syriusso I wanted to make sure how to do that
20:56:30syriusso I can show this guy how to
20:56:38syriusgoing to install songbird on his mac
20:56:52syriushe doesn't want itunes
20:57:15 Quit mikroflo1s_ (Ping timeout: 246 seconds)
20:57:34S_a_i_n_tsyrius: With Rockbox...he doesn't *need* iTunes :)
20:57:35syriusbecause I told him if he wants to use it like a normal mass storage device to just put the music in a folder like drag and drop you need to use another player because itunes doesn't do it like that
20:57:45syriusyes of course he doesn't
20:57:49saratogagevaerts: assuming a license is logically consistent makes you a bad lawyer :)
20:58:05syriushe says he prefers to do it that way anyways
21:00
21:01:29Bagderhttp://socghop.appspot.com/gsoc/program/list_projects/google/gsoc2010
21:04:35 Quit Luca_S (Quit: CGI:IRC)
21:04:57linuxstbBagder: Nowhere just showing Rockbox's projects?
21:05:19BagderI didn't spot any, that list doesn't even mention which org that hosts the projects!
21:06:00bertrikWho got accepted for rockbox GSOC?
21:06:23 Quit hebz0rl (Quit: Ex-Chat)
21:07:09linuxstbkugel (Rockbox as an App), mt (WMA Pro and WMA Voice), archivator (TTS in core Rockbox) and bieber (Theme editor)
21:07:14*linuxstb hopes he got that right.... ;)
21:07:40gevaertslinuxstb: those are the ones we rejected! ;)
21:07:44*gevaerts runs
21:07:50*linuxstb runs
21:13:20 Quit Rondom (Disconnected by services)
21:13:37 Join Rondom_ [0] (~quassel@dslb-084-057-192-003.pools.arcor-ip.net)
21:14:01***Saving seen data "./dancer.seen"
21:14:08 Nick evilnick is now known as evilnick_silly_b (~0c140464@rockbox/staff/evilnick)
21:14:40 Quit TheSphinX^ (Quit: XChat)
21:16:56 Join Rondom [0] (~quassel@dslb-088-064-124-136.pools.arcor-ip.net)
21:17:09 Quit DerPapst (Quit: Leaving.)
21:17:12 Quit M3DLG (Ping timeout: 246 seconds)
21:20:36 Quit Rondom_ (Ping timeout: 248 seconds)
21:20:37 Quit shaggy-h (Ping timeout: 240 seconds)
21:20:46kugelgevaerts: hey, you're my mentor
21:20:54*gevaerts looks
21:20:57gevaertsOh, right
21:20:59S_a_i_n_tgevaerts: run!
21:21:02*gevaerts runs
21:21:14bluebrothergevaerts: run faster! :)
21:21:34gevaertskugel: yes, indeed :)
21:22:03kugel:)
21:23:44 Quit Rondom (Ping timeout: 260 seconds)
21:24:26 Quit kugel (Remote host closed the connection)
21:24:38 Join kugel [0] (~kugel@rockbox/developer/kugel)
21:25:08 Join lpereira [0] (~lucien@112.46.70-86.rev.gaoland.net)
21:27:19archivatordomonoky: hello, mentor! :)
21:27:51domonokyhello student :-)
21:28:12S_a_i_n_tHello World.
21:28:28archivatorSerious question: how do I use logf in a plugin?
21:28:36domonokyarchivator: congratulations for getting accepted !
21:28:50archivatorthanks!
21:30:38S_a_i_n_tThanks Ants, thants.
21:31:31 Part watto
21:32:21mtsaratoga: Congratulations ! you get to mentor me again. :P
21:32:49Bagdersome guys never learn! ;-)
21:33:00mthaha
21:33:20domonokyarchivator: i never used logf in a plugin, but i would suspect it works the same as in the core: make a logf build, place the logf enable Macro in your .c file and call rb->logf...
21:34:41S_a_i_n_tWhy is it that after selecting "Advanced - Voice Build" you get rewarded with a confirmation telling you what you already know (that you selected a voice build) instead of it just going "righto, I'll create that makefile for you..."?
21:34:59Tornebecause you can pick more than one advanced option at once
21:35:02gevaertsS_a_i_n_t: you can select more than one advanced option
21:35:06 Quit komputes (Ping timeout: 252 seconds)
21:35:06 Part xiainx ("Good Bye!")
21:35:07Torneyou have to enter a blank one to tell it you're done
21:35:29S_a_i_n_tAha...that makes some more sense now. Thanks.
21:37:48archivatordomonoky: right, I just managed to corrupt the build in the process, so I thought there might be something more to it.
21:38:24 Quit pixelma (Disconnected by services)
21:38:26 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
21:38:35 Quit amiconn (Disconnected by services)
21:38:37 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
21:38:46 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
21:38:48 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
21:39:16saratogamt: woot
21:39:44mt:)
21:39:56 Nick fxb__ is now known as fxb (~felixbrun@h1252615.stratoserver.net)
21:40:49archivatorHm, is the logf build supposed to be really, *really* slow? 'Cause I can't get it to boot at all.
21:41:53 Join mikroflops [0] (~yogurt@90-227-45-110-no112.tbcn.telia.com)
21:41:53Torneturning logf on makes no actual runtime difference at all, in theory
21:41:53Tornesince you have to add LOGF_ENABLE to the individual files..
21:42:47gevaertsarchivator: what target? There's FS #11204...
21:43:13archivatorgevaerts: ipodvideo
21:43:26archivatorit doesn't even get to the logo, freezes with the apple on.
21:43:31gevaertsok, useful to know that it's not just CF then
21:43:37*gevaerts wonders what could be wrong
21:43:51Tornethat's very bizarre.. I use logf builds all the time
21:44:00bluebrotherBagder, linuxstb: list with only Rockbox projects is here: http://socghop.appspot.com/gsoc/org/home/google/gsoc2010/rockbox
21:44:04Tornein fact i have bootchart permanently enabled in my regular build
21:44:11Bagderthanks
21:44:16archivatorthe disk is spinning and accessing something at regular intervals.
21:44:55archivatorsomewhat less than 1sec I think.
21:45:15gevaertshm
21:45:28 Quit mikroflops_ (Read error: Operation timed out)
21:46:36gevaertsI think I've seen that on my gigabeat F once, but I can't remember if that was a logf build
21:47:27archivatorI'm making a new build, ccache disabled. See if that changes anything.
21:47:27gevaertsarchivator: could you try reverting only r25459?
21:48:21archivatorgevaerts: sure, let me just check that it's not ccache interference.
21:49:40kugelthe odds for ccache being the cause are extremely low
21:50:03archivatorkugel: I've had it happen..
21:50:15archivatornot with rockbox, though.
21:50:34 Join komputes [0] (~komputes@ubuntu/member/komputes)
21:50:49archivatorand .. it works.
21:50:58archivatorriddle me that.
21:51:17kugelextremely low is still non-zero :P
21:51:35 Quit JohannesSM64 (Max SendQ exceeded)
21:51:43*gevaerts doesn't like that sort of issue
21:51:52gevaertsarchivator: did you make clean earlier?
21:52:03archivatorgevaerts: veryclean, actually. Several times.
21:52:09archivatorafter each attempt, that is.
21:52:10 Join mirak [0] (~mirak@85-171-108-160.rev.numericable.fr)
21:52:48 Join JohannesSM64 [0] (~johannes@cm-84.215.116.196.getinternet.no)
21:52:59 Quit S_a_i_n_t ()
21:53:08 Join bieber [0] (~quassel@162-78.97-97.tampabay.res.rr.com)
21:53:12 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.0.156)
21:56:08saratogakugel: while you're looking at audio output in the sim, could you keep your eyes open for a way to make the sim run faster then real time
21:56:40saratogadebugging playback and codec issues would be a lot easier if we could make the sim output PCM at 2x, 4x, ... realtime so I don't have to wait half an hour for a deadlock
21:56:47gevaertsbieber: congratulations!
21:57:22kugelsaratoga: 2x should work now, i.e. pitch. timestretch also works
21:59:39saratogakugel: I was thinking more of a way to do it by discarding the audio and just running everything faster, e.g. don't sleep when the pcm buffer fills up, just empty it as fast as we can
22:00
22:00:41kugelisn't the 2x pitch limit arbitary anyway? you could try simply upping it
22:00:57 Quit bieber (Remote host closed the connection)
22:01:07 Nick evilnick_silly_b is now known as evilnick (~0c140464@rockbox/staff/evilnick)
22:01:30saratogadoes that run everything faster? i wonder if its the same as running longer at regular pitch
22:02:40kugelit should be
22:04:42archivatorgevaerts: what should I do to get logf over usb? I have -DUSB_ENABLE_SERIAL and usb_serial was seemingly compiled in. I've also modprobe'd usbserial. Yet, no /dev/ttyUSB devices
22:05:06 Nick bgs000 is now known as bgs100 (57o9@unaffiliated/bgs100)
22:05:24gevaertsarchivator: maybe disable HID
22:05:38gevaertsAlso it you might have to enable it in the debug menu
22:06:28archivatorhmm, no option under Debug, I guess it wasn't compiled in after all.
22:06:41 Quit bertrik (Changing host)
22:06:41 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
22:11:05 Join merbanan [0] (~banan@c-94-255-217-199.cust.bredband2.com)
22:12:12 Quit syrius (Quit: Leaving)
22:13:03bertrikarchivator, wasn't it the cdc_acm module rather than usbserial that you need to modprobe?
22:13:29 Part chaos
22:13:40archivatorbertrik: nah, it's usbserial, I got it working.
22:19:11 Join mikroflops_ [0] (~yogurt@90-227-45-110-no112.tbcn.telia.com)
22:20:58 Quit JohannesSM64 (Quit: WeeChat 0.3.2-dev)
22:22:57 Quit mikroflops (Ping timeout: 246 seconds)
22:24:50 Join ajpm [0] (~ajpm@84.13.161.225)
22:27:27 Quit n17ikh (Ping timeout: 265 seconds)
22:27:54 Quit bmbl (Quit: Bye!)
22:30:15ajpmwhat's the trick for booting to original firmware (on ipod 4g)?
22:31:54archivatorajpm: have hold on while it's booting?
22:32:03 Join n17ikh [0] (~n17ikh@host-69-59-126-212.nctv.com)
22:32:28archivatorajpm: http://download.rockbox.org/daily/manual/rockbox-ipod4g/rockbox-buildch3.html#x5-290003.1.3
22:33:03archivatorplease, do read the manual. You'll find most your answers there.
22:34:58ajpmarchivator, thanks a lot you've saved from having to trawl through it
22:48:48 Quit S_a_i_n_t ()
22:49:15wodzkugel: could You look at mkmpioboot part of the patch at FS #11137 - is it ok for You?
22:50:42 Join mikroflops [0] (~yogurt@90-227-45-110-no112.tbcn.telia.com)
22:52:05archivatorlogf() do not do anything in a normal build, right? Is it safe to leave them in a my plugin or do I need to #ifdef them?
22:52:18kugelwodz: heh, you even prepared it for rbutil. you're awesome :)
22:52:54gevaertsarchivator: that's fine
22:52:55 Join DerPapst [0] (~Alexander@p4FE8F949.dip.t-dialin.net)
22:53:35wodzkugel the bit which bothers me is conversion int2be. It is ok on intel but probably wrong on ppc
22:53:37 Quit Jaykay (Ping timeout: 276 seconds)
22:54:11kugelwodz: I assume it should do nothing on ppc?
22:54:29 Quit mikroflops_ (Ping timeout: 246 seconds)
22:54:35kugelI guess looking at other installers might give a clue
22:54:59 Quit ajpm (Quit: sticks, stones /= bones)
22:59:59archivatorlogf is supposed to be printf-like, correct? %i doesn't seem to do anything...
23:00
23:00:38gevaertsWe probably don't do %i
23:03:11kugelarchivator: we only support a subset of the formatting things, try %d/%u
23:03:50archivatorkugel: I figured as much, %d works. %i is in the specs, though! (synonymous to %d for printf) :)
23:04:12gevaertsarchivator: is it worth even one byte of binsize?
23:04:46archivatorgevaerts: yes, yes, I see your point. It's just non-standard, 'tis all.
23:05:12kugelnoooooo, if we add %i, people start asking for %f! :)
23:06:18archivatorkugel: I was actually thinking about that the other day - since we use fixed-point numbers so much, can't we not add a utility function that's like %f but for fixed-point numbers. Might come in handy :)
23:07:54 Join sudoman [0] (~d8ecfceb@gateway/web/freenode/x-kkkxaityihbjbcoh)
23:08:03gevaerts%i seems to mostly exist for compatibility with scanf
23:08:40*kugel never understood the existence of %d actually
23:13:54 Join ball [0] (~ball@adsl-99-142-84-135.dsl.emhril.sbcglobal.net)
23:14:04***Saving seen data "./dancer.seen"
23:14:47sudomani'm trying to build rbutil for ppc linux. however, i get the error: /usr/bin/lrelease: Command not found
23:15:13sudomani'm on ubuntu lucid
23:15:24domonokysudoman: do you have the Qt Dev package installed ?
23:15:48sudomanlemme check
23:15:52domonokylrelease is part of Qt. (you also need libusb-dev to build rbutil)
23:16:15wodzok I am about to commit HD200 port
23:16:24wodzany objections?
23:16:53domonokywodz: if you are sure, you are not breaking other targets, go for it ! :-)
23:17:21wodzthe only thing I am sure and certain is that I will die
23:17:27archivatorcan someone with a target with a mic test FS #11219. You shouldn't be playing anything when you start the plugin.
23:18:30 Quit evilnick (Quit: Page closed)
23:18:46ballWhere should I look for a manual for Rockbox 3.5.1 for the iPod mini? I get a 404 from the usual page (http://www.rockbox.org/download/byhand.cgi )
23:18:47sudomandomonoky: what if i want to use firewire?
23:20:55domonokysudoman: that doesnt matter, libusb is used to detect targets.
23:21:51archivatorball: that page does appear to link to non-existent files. Try the daily builds - there might be a few features not available in your version but everything else will be applicable.
23:21:53 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
23:22:06 Join arbingordon [0] (~w@unaffiliated/arbingordon)
23:22:11ballarchivator: Thanks
23:22:36ballI hope I have all this music backed up somewhere. I think I'm going to have to newfs the thing.
23:22:39archivatorball: clarification: the daily builds of the manual, not of rockbox.
23:22:49ballarchivator: I understand, thanks.
23:23:05archivatorjust making sure I'm not sending you down the rabbit hole :)
23:23:36ballbrb, phone
23:24:18ballUgh.. I have to go
23:24:24 Quit ball (Quit: leaving)
23:28:09archivatorOkay, I'll just leave this here and go to bed, then - FS #11219 - call for testers with mic- or linein-equipped targets.
23:29:34 Quit archivator (Quit: Leaving)
23:29:54 Quit sudoman (Ping timeout: 252 seconds)
23:31:24 Quit pamaury (Quit: Page closed)
23:31:48linuxstbBah, too late, but 3.5.1 manuals are at http://download.rockbox.org/release/3.5/
23:32:05linuxstb(we never rebuilt 3.5.1 manuals by the look of it)
23:32:27gevaertsindeed not. It seemed a bit pointless
23:33:31linuxstbAlthoughj is that the cause of link breakage?
23:33:55linuxstbLooks like it is... http://www.rockbox.org/download/byhand.cgi
23:34:16linuxstbI guess someone needs to cp/ln on the download server... Bagder ?
23:34:19gevaertsBagder: ping
23:34:50Bagderhm
23:34:58Bagdercan I blame zagor please? B)
23:35:12gevaertsYou can blame whoever you want, but could you fix it? :)
23:35:52BagderI'll check
23:38:43 Quit merbanan (Ping timeout: 245 seconds)
23:40:03CIA-5New commit by 03wodz (r25724): Add MPIO HD200 port - changed files
23:40:19CIA-5New commit by 03wodz (r25725): Add MPIO HD200 port - new files
23:40:28CIA-5New commit by 03wodz (r25726): Add MPIO HD200 port - binary files
23:40:47 Join EsotericWisp [0] (~636125a5@giant.haxx.se)
23:43:36 Quit emrecelikten (Quit: Page closed)
23:45:47 Quit bertrik (Quit: De groeten)
23:46:11 Quit kugel (Ping timeout: 248 seconds)
23:46:42CIA-5New commit by 03dast (r25727): let manuals and voices have separate release numbers
23:47:20BagderI need to delete that repo, I have the wrong user name there :-!
23:49:41wodztime to sleep
23:49:44wodzbye
23:49:52 Quit wodz (Quit: Leaving)
23:52:54 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.0.246)
23:53:03 Nick Ypsy is now known as YPSY (~ypsy@geekpadawan.de)
23:54:47CIA-5New commit by 03bagder (r25728): versioning: separate manual and voices from publicrelease ...
23:56:17Bagdergrrr
23:56:19 Quit w1ll14m (Ping timeout: 260 seconds)
23:57:26EsotericWispHi, Why can I not install the doom files onto my H10 20GB?
23:57:43CIA-5New commit by 03bagder (r25729): release build table: use the correct dirs for manuals and voices too
23:58:13Bagdernow I can crawl back to my cave
23:58:58EsotericWispIt says<"Your Deviceoesn't have a oom plugin. Aborting" Why?

Previous day | Next day