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

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

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

#rockbox log for 2013-01-30

00:10:48Wally [Saint] ha
00:10:53WallyUsing it on my OpenPandora
00:11:14[Saint]The Cowon D2 manual should be mostly correct for this
00:11:19Wallyyeah I didn't see the dtabase feature
00:11:26Wallyhowever its catalogued everything -_-
00:11:27[Saint]...or, any other touchscreen target's manual.
00:12:36[Saint] <−− should be mostly correct (except for references to physical buttons)
00:13:12[Saint](and missing/partially implemented features)
00:15:10 Quit kevku (Ping timeout: 264 seconds)
00:22:35 Join dfkt_ [0] (~dfkt@unaffiliated/dfkt)
00:23:39 Quit dfkt (Disconnected by services)
00:23:42 Nick dfkt_ is now known as dfkt (~dfkt@unaffiliated/dfkt)
00:24:53 Part Wally ("Textual IRC Client:")
00:25:07 Quit prof_wolfff (Ping timeout: 272 seconds)
00:31:33 Quit ruskie (Read error: Connection reset by peer)
00:33:03 Join ruskie [0] (ruskie@sourcemage/mage/ruskie)
00:39:30 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
00:46:42 Quit Sanguinius (Quit: KVIrc 4.2.0 Equilibrium
00:49:35 Quit jhMikeS (Ping timeout: 245 seconds)
00:49:44JdGordonkugel: nope
00:49:50JdGordonprobably to keep the api size down
00:51:05 Quit foolsh (Ping timeout: 272 seconds)
00:56:18 Quit ender` (Quit: You're not drunk if you can lie on the floor without holding on.)
01:04:08 Join foolsh [0] (
01:14:59 Quit pamaury (Ping timeout: 245 seconds)
01:17:17 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
01:23:52 Quit bluebrother (Disconnected by services)
01:23:57 Join bluebrother^ [0] (~dom@rockbox/developer/bluebrother)
01:24:32 Join pamaury_ [0] (~quassel@rockbox/developer/pamaury)
01:24:48 Quit pamaury (Ping timeout: 256 seconds)
01:25:07 Quit fs-bluebot (Ping timeout: 256 seconds)
01:26:32 Join fs-bluebot [0] (
01:31:18 Join SuperBrainAK [0] (
01:37:57 Quit foolsh (Ping timeout: 272 seconds)
01:43:55 Quit lebellium (Quit: ChatZilla 0.9.89 [Firefox 19.0/20130123083802])
01:50:48 Quit Wardo (Read error: Connection reset by peer)
01:50:49 Join foolsh [0] (
01:55:37 Join Strife89 [0] (~Strife89@
01:58:49***Saving seen data "./dancer.seen"
02:02:12 Join kadoban [0] (
02:03:16 Quit Strife89 (Ping timeout: 272 seconds)
02:05:58 Quit pamaury_ (Ping timeout: 276 seconds)
03:58:52***Saving seen data "./dancer.seen"
04:13:30 Join TheSphinX^ [0] (
04:15:25 Quit TheSphinX_ (Read error: Operation timed out)
04:15:51 Quit XavierGr (Ping timeout: 245 seconds)
04:18:40 Join amiconn_ [0] (amiconn@rockbox/developer/amiconn)
04:18:40 Quit amiconn (Disconnected by services)
04:18:41 Quit pixelma (Disconnected by services)
04:18:41 Join pixelma_ [0] (pixelma@rockbox/staff/pixelma)
04:18:43 Nick pixelma_ is now known as pixelma (pixelma@rockbox/staff/pixelma)
04:18:44 Nick amiconn_ is now known as amiconn (amiconn@rockbox/developer/amiconn)
04:20:57 Join foo|sh [0] (
04:21:22 Quit foolsh (Ping timeout: 272 seconds)
04:26:57 Quit foo|sh (Quit: Leaving)
04:27:09 Join foolsh [0] (
05:13:31 Quit TheSeven (Read error: Operation timed out)
05:15:15 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
05:21:15 Quit scorche (Ping timeout: 245 seconds)
05:32:08 Join TheSphinX_ [0] (
05:32:55[Saint]Heh..I somehow managed to break my build in a hilarious way.
05:33:19[Saint]"Shortcuts" lets me bring up the context menu on it and add it to the shortcuts menu ;)
05:33:53[Saint]one of the first tests I ever made was to see if the shortcuts menu was silly enough to let me do this, and, it isnt...normally.
05:33:59[Saint]But apparently I broke it.
05:34:04 Join scorche [0] (~scorche@rockbox/administrator/scorche)
05:35:17 Quit TheSphinX^ (Ping timeout: 276 seconds)
05:48:46 Quit foolsh (Ping timeout: 272 seconds)
05:48:59 Join webguest51 [0] (
05:50:11 Quit webguest51 (Client Quit)
05:58:54***Saving seen data "./dancer.seen"
06:02:07 Join foolsh [0] (
06:38:53 Quit SuperBrainAK (Quit: pbly going to sleep /_\)
06:43:41 Join TheSphinX^ [0] (
06:46:34 Quit TheSphinX_ (Ping timeout: 252 seconds)
07:01:13 Join TheSphinX_ [0] (
07:02:49foolsh[Saint], you were right black is the first thing put on the chopping block. How does the syntax go again? g#390
07:02:52fs-bluebotGerrit review #390 at : Additional colors for lamp.rock by Hayden Pearce (changes/90/390/1)
07:03:46 Quit TheSphinX^ (Ping timeout: 245 seconds)
07:04:13foolshIs there much difference from white dimmed almost all the way down and black anyways?
07:05:13 Join |akaWolf| [0] (~akaWolf@
07:06:04[Saint]a fair bit, the next dimmest candidate is red (closest in the spectrum to black {hence why the military used(/s?) red light for night op map reading.
07:06:49[Saint]I'm more than happy to axe it if it causes to much of a fit.
07:10:23foolshI suppose you must give them something to axe or some feel they're not doing their job, aye?
07:11:05[Saint]I'll just find something I don't like in another task as revenge :P
07:44:56 Join enriched [0] (~quassel@
07:48:22 Quit foolsh (Quit: Leaving)
07:57:09 Join melmothX [0] (~melmoth@unaffiliated/melmothx)
07:58:57***Saving seen data "./dancer.seen"
07:59:08 Quit Rower (Quit: Hmmm...)
08:02:09 Join foolsh [0] (
08:04:24 Quit Raptors (Read error: Connection reset by peer)
08:04:30 Join Raptors_ [0] (
08:04:59 Quit Raptors_ (Client Quit)
08:05:07 Join Raptors_ [0] (
08:05:08 Quit Raptors_ (Client Quit)
08:05:28 Join Raptors [0] (
08:07:39 Join esperegu [0] (
08:09:48 Join LinusN [0] (
08:20:32 Join wodz [0] (
08:21:49 Join ender` [0] (
08:22:02wodz[Saint]: its not that I care much about 'black' in lamp. It just looks weird to have lamp in 'black'. Red has its merit (for example to inspect stereoscopic r/g images to filter one frame out)
08:23:13 Join mortalis [0] (
08:23:26wodz[Saint]: But to be honest - if you want color selection why not code color picker and satisfy everyone?
08:23:39[Saint]I just dragged that out of me personal tree, I can easily remove it. I use it, for example, if I need enough light to see by but don't want to disturb my partner.
08:24:07[Saint]That's "step 2", I am thinking of adding a color picker, but even if I do, I think it sane to have a set of defaults.
08:24:29wodzIf you have usecase for 'black' thats fair enough
08:25:11wodz[Saint]: true, I would expect a (small) set of default colors + one 'custom'
08:25:36[Saint]I'm happy to axe black if it is a source of contention.
08:25:54wodzthe easier path for color picker would be config file where you could put rgb value for custom slot
08:25:59[Saint]I didn't think to remove it before I pushed to gerrit.
08:26:23wodz[Saint]: I wouldn't object for 'black' if it has some merit
08:26:42[Saint]I was thinking about re-using the color picker for foreground/background color.
08:27:03[Saint]iirc, there's a color picker that spits out a hex value used somewhere in the core.
08:27:46[Saint]I _think_ it's for foreground/background colors, but, it may be for filetype colors now I think of it.
08:27:51wodzreally? I never hit it then
08:29:17[Saint]Settings - Theme Settings - Colors
08:29:43[Saint]I was right the first time (damn second guessing myself), it is used for foreground/background color.
08:30:09[Saint]there's three sliders (for R, G, and B), and a preview window.
08:30:29[Saint]Shouldn't be too difficult to kludge into the lamp plugin.
08:31:14wodzyeah, I see
08:31:33wodzso you probably need to export this function in plugin api and you are set
08:31:42*[Saint] nods
08:34:50 Join Zagor [0] (
08:34:51 Quit Zagor (Changing host)
08:34:51 Join Zagor [242] (~bjst@rockbox/developer/Zagor)
08:37:31wodzZagor: any more thought about random build errors?
08:38:13wodz[Saint]: setc_color() is even exported in plugin api already
08:38:34[Saint]Ah, helpful.
08:38:48Zagorwodz: I still think it's a dependency failure. I just don't know exactly how.
08:39:17wodz[Saint]: it is used in text_editor and lrcplayer
08:39:46[Saint], the Recorder outright shat its pants last round.
08:40:08[Saint]oh, no, derp...for the last N rounds.
08:40:10Zagorlast round? it's been doing it for a long time.
08:40:35Zagorour missing irc reports make a big difference
08:42:01Zagorit's a bit weird that n01-roolku is so heavily over-represented
08:42:46wodzZagor: when we speak about announcements: + irker could be drop in replacement (or rewrite hooks and use simply irker)
08:43:00 Join kevku [0] (~kevku@2a01:d0:ffff:34a::8:3)
08:43:07 Quit enriched (Remote host closed the connection)
08:43:10Zagorit's practically only n01-roolku getting it, and only occasionaly storebror-daniel
08:43:30wodzZagor: I guess its because when it starts to get failing builds it appears as super fast machine
08:43:31 Join enriched [0] (~quassel@
08:43:45wodzZagor, n05-roolku pop from time to time
08:43:48Zagorwodz: right, but n01 isn't even the fastest.
08:44:10wodzI saw also a few other random machines failing this way
08:44:23Zagorwodz: yes, some others are seen occasionally but n01 is the vast majority.
08:44:35[Saint]rigaud-danya, microserver-jdgordon, hex-gevaerts, etc.
08:45:11ZagorI'm not saying there is anything wrong with n01, just that knowing why it hits the problem so frequently could give a clue.
08:45:21*[Saint] nods
08:45:49wodzit would be nice to inspect failing build (and logs for it)
08:45:54Zagor[Saint]: I am talking about the libtlsf.a error, not the recorder issue
08:46:19Zagorwodz: how do you mean? we have all the logs.
08:46:28wodznot the verbose ones
08:47:00ZagorI don't think seeing all commands would help much. we're already seeing the order.
08:48:04wodzbut the order seems to be correct
08:48:31[Saint]it seems to be completely random selection as to whether a given build wants to show me the build log or not.
08:51:04Zagor[Saint]: hmm, looks like I'm cleaning old logs too soon
08:51:08wodzthe only way it could fail is if `ar' command fails for some reason and if I read libtlsf.make correctly we throw away errors from ar
08:52:17[Saint]Zagor: aha, great (well, not really.) I was just about to start blaming Chrome :)
08:52:41[Saint]Almost fired up FF, that was lucky. ;p
09:00:14 Join einhirn [0] (
09:01:32Zagorwodz: I wonder why we added that redirect. trying without now and there is no output
09:01:57wodzmaybe gevaerts remembers
09:05:15Zagoroh right, we don't even redirect stderr. only stdout. errors would show anyway.
09:07:30 Join petur [0] (~petur@rockbox/developer/petur)
09:10:16wodzlibtlsf is linked way before this error pops up
09:19:40 Quit wodz (Quit: Leaving)
09:48:30 Join lebellium [0] (
09:54:38 Join wodz [0] (
09:55:56 Join bebna [0] (~a.fasold@
09:59:01***Saving seen data "./dancer.seen"
10:00:31 Quit petur (Quit: *plop*)
10:02:30 Join XavierGr [0] (~xavier@rockbox/staff/XavierGr)
10:03:22 Join petur [0] (~petur@rockbox/developer/petur)
10:04:52 Quit petur (Read error: Connection reset by peer)
10:04:59 Quit Unhelpful (Remote host closed the connection)
10:05:09 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
10:06:10 Join petur [0] (~petur@rockbox/developer/petur)
10:14:17 Join Kaii [0] (6e4d951a@gateway/web/freenode/ip.
10:16:04KaiiHi every, Does anyone face this problem? Once I browse by using Database and select any song, the player will play another song instead.... to be precise, it will be the next 3 songs after the selected one
10:16:44KaiiThis problem occur only with Asian songs, Thai Japanese, Korean
10:16:59KaiiMaybe something to do with file name?
10:21:02 Quit petur (Quit: *plop*)
10:21:15[Saint]Is it *always* exactly three songs ahead of the selected track?
10:21:49[Saint]...that is odd.
10:24:32KaiiYes, always 3 songs (as shown on the list generated by Database), after the selected one :)
10:25:47[Saint]It would help if you reported the bug here:
10:26:04Kaiinoted, will do Thank you
10:27:01 Quit Kaii (Quit: Page closed)
10:41:21 Quit foolsh (Ping timeout: 272 seconds)
10:42:42 Quit mortalis (Read error: Operation timed out)
10:50:15 Join mortalis [0] (
10:54:27 Join petur [0] (~petur@rockbox/developer/petur)
11:04:02 Join Ward [0] (
11:04:25 Nick Ward is now known as Guest78302 (
11:07:17 Quit kadoban (Ping timeout: 264 seconds)
11:17:58 Join foolsh [0] (
11:30:15 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
11:59:04***Saving seen data "./dancer.seen"
12:10:59 Quit foolsh (Quit: Leaving)
12:29:25 Join foolsh [0] (
12:33:21mortaliswodz: one user complains about ata error -2. He says that error doesn't occur after hard reset.
12:34:19mortalisif i read sd spec correctly 1 sec is minimal timeout, shouldn't increase it in this case?
12:35:59wodzfeel free to fix
13:19:39 Quit petur (Quit: *plop*)
13:25:23 Quit foolsh (Ping timeout: 272 seconds)
13:37:35 Join foolsh [0] (
13:44:33gevaerts[Saint]: that's alphabetic sort!
13:59:07***Saving seen data "./dancer.seen"
14:01:27 Quit esperegu (Ping timeout: 244 seconds)
14:04:08 Join esperegu [0] (
14:10:25 Join amayer_ [0] (
14:29:48 Quit wodz (Quit: Leaving)
14:37:23 Quit the-kyle (Ping timeout: 264 seconds)
14:38:01 Join the-kyle [0] (~kyle@
14:39:13dv_hello. who is the person I should speak to about Opus optimizations for ARM?
14:45:18amayer_dv_: i believe its bertrick. but i could be wrong
14:45:29gevaertsLots of people, really
14:45:48amayer_gevaerts: true.
14:45:50gevaertsOptimisation tends to be a collaborative effort I'd say
14:45:56gevaertsAlso, he has no c in his name :)
14:46:49amayer_oh... he isnt signed in so i gave my best shot at his name
14:48:27amayer_dv_: gevaerts is right tho. if you want to start a discussion about it just start asking questions or throwing ideas out there. usually someone will respond
14:48:29dv_I would try my hands on arm NEON optimizations, but I am unfamiliar with it, dont have the budget to do it at work, and perhaps somebody is already doing just that
14:48:46dv_specifically, optimizations of the FFT and MDCT code
14:49:12dv_I run opus on a cortex-A8 though, which I guess is not the typical target platform for rockbox, right?
14:49:36gevaertsWell, RaaA runs on those
14:50:03 Join redhot [0] (~kvirc@
14:50:13dv_I use opus for internal audio transmissions between devices, and is currently one significant performance sink
14:51:27gevaertsThere was some optimisation talk mentioning neon at
14:53:50 Join TheSphinX^ [0] (
14:54:36 Join petur [0] (~petur@rockbox/developer/petur)
14:56:09 Quit TheSphinX_ (Ping timeout: 240 seconds)
15:04:41 Quit esperegu (Ping timeout: 276 seconds)
15:04:58 Join esperegu [0] (
15:24:47 Join Rower [0] (
15:27:28 Quit the-kyle (Ping timeout: 255 seconds)
15:34:29 Join lorenzo92 [0] (
15:35:30lorenzo92anyone knows about the possibility to read/write (at least write) registers of a codec registered in ALSA? I know there are the proper functions but do you know whether it's possible to access them from user-space?
15:35:38lorenzo92kugel: maybe you know something...
15:36:03 Quit mortalis (Quit: Leaving)
15:36:40*lorenzo92 needs to access register from R1's userspace, but doesn't find a way, just always a "symbol not found" while loading my module...
15:38:53 Join TheSphinX_ [0] (
15:38:55pamaurylorenzo92: if the driver doesn't export the functions I'm not sure you can do it
15:39:35 Quit TheSphinX^ (Ping timeout: 245 seconds)
15:39:45pamauryif the underlying driver (i2c ?) is available from user space, you need to interface with the codec directly perhaps, bypassing alsa driver
15:41:06lorenzo92pamaury: yes i2c is available, indeed I've tought about this but before I should kill the kernel i2c driver for it, isn't it?
15:41:40pamauryhum, the audio driver you mean ? if you the i2c driver how can you access i2c ? :)
15:42:10lorenzo92uhm audio is sent via i2s, while control via i2c
15:42:16lorenzo92I just need control
15:42:23 Quit TheSphinX_ (Read error: Operation timed out)
15:42:54lorenzo92i2c driver for it = wm1808 reg read/write
15:43:44lorenzo92indeed I need to explore this possibility since it's the only way - apart if someone tells me it's possible to use ALSA features...
15:43:47pamauryoh so the i2c+wm18080 reg read/write are in the same driver ?
15:43:53 Join TheSphinX^ [0] (~briehl@
15:44:03lorenzo92pamaury: nono sorry you misunderstood
15:44:24lorenzo92we have separate i2c driver + wm1808 driver (which is attacched to i2c)
15:44:56pamauryand is it the same driver controlling i2c and i2s for the codec ?
15:45:01lorenzo92so you think it's not needed to detach i2c internal driver FOR wm1808? I can simply write my own?
15:45:07lorenzo92pamaury: no it isn't
15:45:18lorenzo92imx37 has a separate interfaec for that
15:45:43pamaurythe problem is that linux will not let you kill a needed driver I think
15:46:10pamauryif the alsa drivers relies on both the i2s and i2c driver, you can't remove any of them without removing the alsa driver
15:46:34lorenzo92pamaury: indeed, seems logical... :(
15:46:47 Join the-kyle [0] (
15:47:03lorenzo92not a good situation, at least with R0 I had the possibility to do that on as35434
15:47:20pamauryI guess you don't have the sources ?
15:47:23 Join TheSphinX_ [0] (
15:47:34lorenzo92yes I have them
15:47:56lorenzo92of course I tried to poke with extern and such, but without any luck
15:47:57pamauryah, then have a look at them, perhaps there is a way to send ioctl to the drivers
15:48:11 Part redhot ("Once you know what it is you want to be true, instinct is a very useful device for enabling you to know that it is")
15:48:15lorenzo92of course I did
15:48:26 Part LinusN
15:48:56 Join LinusN [0] (
15:49:17 Quit TheSphinX^ (Ping timeout: 245 seconds)
15:49:30pamauryand what is the outcome ? You can't access a function if it's not explicitely exported from the driver. But perhaps the driver implementes ioctl to read/write the registers. I don't know how that work really.
15:50:01 Quit Guest78302 (Read error: Connection reset by peer)
15:51:19 Part LinusN
15:51:39 Join LinusN [0] (
15:51:53 Join TheSphinX^ [0] (
15:52:37 Quit Zagor (Quit: Clint excited)
15:54:04lorenzo92pamaury: yes in fact there are not ioctls unfortunately, I guess I need to exploit alsa in some way
15:54:11 Quit TheSphinX_ (Ping timeout: 248 seconds)
15:55:35pamauryYou can use alsa to tweak standard settings, but you can't tweal what alsa doesn't export. I doubt the alsa driver exports many things, I'm downloading the code to see, I'm curious
15:55:54 Join TheSphinX_ [0] (
15:56:57pamauryhow does that work on the r0 ?
15:57:05 Join homer21 [0] (~homer21@
15:57:34lorenzo92pamaury: I directly use rockbox driver ... I wrote a wrapper to the afe.ko module, which exports read/write registers :)
15:58:03pamauryyou might have to do the same then
15:58:37lorenzo92yes, I was doing that, but unfortunately this time no afe.ko-like module, so I was writing my own but now I'm stuck
15:58:56lorenzo92module is already-ready but I'm missing the function to write reg
15:58:59 Quit TheSphinX^ (Ping timeout: 264 seconds)
15:59:09***Saving seen data "./dancer.seen"
15:59:28lorenzo92well to be precise I have a lot of methods which write registers but NONE of them is exported pfff
16:00:00lorenzo92pamaury: everything is inside /sound
16:00:15lorenzo92ah btw
16:00:38lorenzo92this time samsung did a shotty package, because symlinks are broken
16:01:18lorenzo92but fortunately I have the old package of ~a year ago, possibly this works out of box, I will try that because otherwise atm I wasn't able to compile my own kernel
16:05:26pamaurythat's a wm1808 right ? the driver is in sound/soc/ ?
16:05:43lorenzo92just do a search for wm1808 and you get everything
16:06:16lorenzo92pamaury: as last resort I would need to recompile the whole kernel, but I hope not!
16:08:49pamauryI think you always have the option to send ioctl to the i2c driver, and hope that the alsa driver doesn't override them :-/
16:09:37lorenzo92pamaury: ah! okay, yes it WON't, tested on R0 ... if that's possible we are lucky :D
16:09:59pamaurywell, except if the i2c driver doesn't any ioctl for this of course
16:10:02lorenzo92because alsa plays only the role of PCM "manager" in our case, mixer is left alone
16:10:16lorenzo92the i2c driver *should* be standard
16:10:31pamauryI have no idea if it exports ioctl, I'll check
16:10:33lorenzo92the cpu code was written by freescale (fortunately ^^)
16:10:35pamauryindeed the wm1808 i2c driver doesn't have an could also modify it and implement them
16:10:53pamaurybut that requires recompiling a driver
16:11:01pamauryso let's check freescale i2c driver...
16:11:02lorenzo92but this involves a new kernel, little impractical with rockbox
16:11:25pamaurywhy a new kernel ? only a new module ? Or is it not compiled as a module ?
16:12:27pamauryit's imx37 ?
16:15:27lorenzo92unfortunately it is not a module
16:16:10pamaurywhere is this damn freescale i2c driver ?!
16:16:15 Join exouy [0] (
16:16:20lorenzo92no idea :D
16:16:37lorenzo92i've seen standard i2c interface for linux in the net and it is very complete
16:17:30lorenzo92pamaury: drivers/i2c/busses
16:17:38 Quit TheSphinX_ (Read error: Operation timed out)
16:17:49pamauryI can't see it, am I blind ?
16:19:01lorenzo92btw, there are standard devices i.e. i2c-0 - i2c-1
16:19:06lorenzo92so I'm quite sure it will work
16:20:50pamaurythen you'll probably have to use this
16:21:10 Quit exouy (Quit: CGI:IRC (Ping timeout))
16:22:17lorenzo92in this case it would be even better, since i move everything in rockbox, thing that I quite like....
16:23:27lorenzo92because thinking about a future RaaA like the YP-R series, it would be quite nice to have some standardization as we were tallking yesterday
16:24:04pamauryyou just have to study the wm1808 driver to see how the read/writes are encoded but that's even simpler. Yeah indeed
16:25:31 Join TheSphinX^ [0] (
16:30:58 Join jhMikeS [0] (~jethead71@
16:30:59 Quit jhMikeS (Changing host)
16:30:59 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
16:32:48 Join y4n [0] (~y4n@unaffiliated/y4ndexx)
16:33:02 Quit lorenzo92 (Quit: ChatZilla 0.9.89 [Firefox 18.0.1/20130117041235])
16:35:55 Join TheSphinX_ [0] (
16:38:15 Quit TheSphinX^ (Ping timeout: 252 seconds)
16:46:43 Quit hype (Quit: Computer has gone to sleep.)
16:50:58 Join dfkt [0] (dfkt@unaffiliated/dfkt)
16:56:00 Join hype [0] (~hype@
16:59:09 Quit hype (Client Quit)
17:11:42 Quit esperegu (Remote host closed the connection)
17:17:29 Quit kevku (Ping timeout: 245 seconds)
17:24:16 Quit XavierGr (Disconnected by services)
17:24:18 Join XavierGr [0] (~xavier@rockbox/staff/XavierGr)
17:38:33 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
17:42:13 Join pretty_function [0] (~sigBART@
17:59:11***Saving seen data "./dancer.seen"
18:09:31 Quit pretty_function (Ping timeout: 252 seconds)
18:11:58 Join pretty_function [0] (~sigBART@
18:12:34 Join kevku [0] (~kevku@2001:470:27:773:0:feed:c0f:fee)
18:16:42 Quit pretty_function (Ping timeout: 260 seconds)
18:22:39 Join saratoga [0] (123e1f35@gateway/web/freenode/ip.
18:23:06saratogadv_: I'm not working on NEON specifically, but i'm very interested in it for Opus
18:24:10saratogaderf and I discussed it recently:
18:26:34 Join pretty_function [0] (~sigBART@
18:28:06dv_ah so you are also eyeing the MDCT and FFT implementations
18:28:30dv_one big concern of mine is that floating point with the A8 is slow, but NEON is useless for integer on the A8
18:28:58saratogadv_: doesn't NEON have packed integer multiplies on the A*?
18:29:06saratogaI haven't looked at it much
18:34:59 Quit pretty_function (Ping timeout: 264 seconds)
18:35:39 Join advcomp2019_ [0] (
18:35:39 Quit advcomp2019_ (Changing host)
18:35:39 Join advcomp2019_ [0] (~advcomp20@unaffiliated/advcomp2019)
18:36:38 Join pretty_function [0] (~sigBART@
18:38:42 Quit advcomp2019 (Ping timeout: 248 seconds)
18:39:43 Join kadoban [0] (
18:45:46 Join ender1 [0] (
18:46:15 Quit ender| (Read error: Operation timed out)
18:47:45 Quit foolsh (Ping timeout: 272 seconds)
18:48:09 Quit ender` (Ping timeout: 240 seconds)
18:48:21 Quit pretty_function (Ping timeout: 256 seconds)
18:48:42 Join lebellium_ [0] (
18:51:01 Quit lebellium_ (Read error: Connection reset by peer)
18:51:46 Quit lebellium (Ping timeout: 255 seconds)
18:52:30 Quit pamaury (Ping timeout: 246 seconds)
18:52:59 Join lebellium [0] (
18:54:54 Join foolsh [0] (
19:00:06 Join pretty_function [0] (~sigBART@
19:02:16 Join ender| [0] (krneki@2a01:260:4094:1:42:42:42:42)
19:04:48 Part LinusN
19:08:59 Quit homer21 ()
19:09:18 Quit Zambezi (Remote host closed the connection)
19:13:32 Nick amee2stolen is now known as amee2woof (
19:19:05 Join hype [0] (~hype@
19:20:36 Quit pretty_function (Ping timeout: 252 seconds)
19:21:11 Join LinusN [0] (
19:22:11 Quit kadoban (Ping timeout: 248 seconds)
19:22:15 Join pretty_function [0] (~sigBART@
19:37:18 Quit foolsh (Quit: Leaving)
19:46:04 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
19:55:11 Join Strife89 [0] (~Strife89@2602:306:250e:61d9:225:d3ff:fed6:15a)
19:59:13***Saving seen data "./dancer.seen"
20:10:26 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
20:15:38 Quit jhMikeS (Ping timeout: 256 seconds)
20:19:36 Quit saratoga (Ping timeout: 245 seconds)
20:23:52 Quit pretty_function (Remote host closed the connection)
20:28:29 Quit melmothX (Quit: #)
20:29:23 Quit Strife89 (Ping timeout: 276 seconds)
20:36:15 Join SuperBrainAK [0] (
20:38:46 Quit mc2739 (Ping timeout: 276 seconds)
20:40:11 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
20:48:56 Join IncognitoMan [0] ( a sansa clip zip in the mail today and wanted to play some sid's
20:49:58 Quit kevku (Ping timeout: 264 seconds)
20:50:27IncognitoMandoesnt seem to work...
20:50:58 Quit bebna (Quit: Leaving.)
20:51:21dfktIncognitoMan, that sid works fine here
20:51:36IncognitoMannot working for me on the latest build
20:51:41IncognitoManjust glitches
20:52:43dfktworking on build 1fbdc28
20:53:12IncognitoManthen what has changed between builds?
20:53:37 Quit |akaWolf| (Ping timeout: 256 seconds)
20:53:41IncognitoManim guessing 1fbdc28 is when they pushed the patches for sudoku?
20:53:53IncognitoMan( dont remember anymore )
20:55:07dfkt(guess that build is kinda old, from december - a lot could have happened in the meantime)
20:56:53IncognitoManheres how its suppose to sound
20:57:01IncognitoManits as if timing is off
21:05:51 Quit XavierGr (Ping timeout: 246 seconds)
21:06:45 Join ender` [0] (
21:06:54 Quit ender| (Read error: Operation timed out)
21:09:00 Quit ender1 (Ping timeout: 246 seconds)
21:14:26 Join jhMikeS [0] (~jethead71@
21:14:26 Quit jhMikeS (Changing host)
21:14:26 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
21:22:54 Join ender| [0] (whatever@2a01:260:4094:1:42:42:42:42)
21:22:55 Join lebellium_ [0] (
21:23:40 Quit lebellium (Ping timeout: 256 seconds)
21:23:50 Nick lebellium_ is now known as lebellium (
21:25:53 Quit petur (Quit: Leaving)
21:26:36 Join petur [0] (~petur@rockbox/developer/petur)
21:30:45IncognitoMan....anyone with latest build have issues playing that sid?
21:34:12pamauryperhaps bertrik can test it
21:39:24 Quit y4n (Quit: coob ov vood?)
21:44:35pamaurymt: I've begun some analysis of the X-Fi Style firmware, it is indeed extremely similar to the other firmwares I've seen
21:44:45 Join saratoga [0] (123e1f35@gateway/web/freenode/ip.
21:44:46pamauryI've already identified many entry points, it should be ready soon
21:45:14saratogadv_: i don't see the relevance of that link?
21:45:46 Join Wardo [0] (
21:47:51saratogathat just looks like some really not well thought out assembly code :)
21:48:06saratogaif you write bad assembly, gcc will beat you everytime
21:48:14[Saint]it seemed to me like "yes, it's slower on NEON, if you're doing it wrong" ;)
21:54:24 Quit petur (Quit: Leaving)
21:55:32IncognitoManbertrik: if you get around to it can you test this sid on the latest build?
21:55:53 Join kevku [0] (~kevku@2001:470:27:773:0:feed:c0f:fee)
21:56:07 Join lebellium_ [0] (
21:57:38 Quit lebellium (Ping timeout: 248 seconds)
21:57:47 Nick lebellium_ is now known as lebellium (
21:59:14***Saving seen data "./dancer.seen"
22:13:43bertrikIncognitoMan: ok, I'll have a try
22:15:00IncognitoManbertrik: as a reference...heres how it *should* sound
22:18:34bertrikI hear some similarity ☺ and at least it doesn't crash
22:19:01bertrikdoesn't sound right
22:19:41bertrikdoes it sound OK on the simulator?
22:20:03IncognitoMandont know about the simulator
22:20:33IncognitoManbut on the actual hardware it sounds quite wrong lol
22:21:01IncognitoManthen again does the zip have a simulator?
22:22:24IncognitoManguess i should setup my server for being a build machine then
22:22:55IncognitoManpamaury: when you tried that sid...what device did you try it on
22:23:20pamauryIncognitoMan: I didn't try it, wait a second, I'll try on the fuze+
22:23:52IncognitoMansorry...mean dfkt
22:23:54pamauryoh wait, I can't, my internet connection is way to slow here
22:24:04IncognitoManpamaury: its a 6kb file
22:24:14pamauryah ok, let me try ^^
22:25:03IncognitoMandfkt: what device did you try the sid on if i may ask
22:26:39bertrikit sounds equally bad on the sim as it does on the actual clip zi
22:27:19bertrikwe could try it on a few other sims, if it still sounds bad, it's probably some problem with the .sid player
22:27:21IncognitoManhuh...then i wonder what dfkt ran it on
22:27:33IncognitoManas he says it played it just fine
22:27:45IncognitoMan( on an older build )
22:27:56IncognitoMan14:51 < dfkt> IncognitoMan, that sid works fine here
22:28:05IncognitoMan14:52 < dfkt> working on build 1fbdc28
22:30:26pamauryIncognitoMan: I can't hear the original, not enough bandwidth for youtube but it definitely doesn't sound the same with vlc and with my fuze+
22:30:43pamauryactually it sounds strange on the fuze+
22:34:15bertriklooks like the core of sid.c codec hasn't seen changes in a long time
22:35:15bertrikat least we haven't tracked upstream fixes for a while
22:35:26IncognitoManhmm :/
22:40:43 Part LinusN
22:43:00 Quit Belzebub (Ping timeout: 252 seconds)
22:47:14 Join Belzebub [0] (
22:53:46bertriklooks like the rockbox sid codec is just one file, while I see multiple files in a repositorty like this
22:54:04 Quit SuperBrainAK (Ping timeout: 264 seconds)
22:56:37bertrikhm, or is that a different tinysid
23:04:52bertrikbah, can't find the tinysid source code
23:07:58bertrikany french speaking people: does this translation look ok?
23:09:52gevaertsThe last one looks wrong
23:11:01 Quit amayer_ (Quit: amayer_)
23:13:05IncognitoMana wild gevaerts appears
23:20:43mtpamaury: That's great. I'm a bit too busy currently, but will have time in the weekend.
23:21:31pamaurybertrik: wait a sec
23:21:56mtI think the port will not need much work, relatively.
23:26:07pamauryI agree
23:50:16pamauryme: the x-fi style has tvout support apparently
23:51:08pamaurylcd seems to be a random smart lcd, perhaps a known one, you'll have to extract the init sequences but i've identified them for you. The code is bit complicated because it handles both tv out and lcd, but I'm sure you'll figure out.
23:54:08 Quit TheSphinX_ (Ping timeout: 260 seconds)
23:56:16 Join TheSphinX^ [0] (
23:56:23pamaurybacklight done with pwm, channel 4, running at unusual frequency of ~2.3MHz

Previous day | Next day