[Ffmpeg-devel-irc] ffmpeg-devel.log.20141030

[burek]

[00:56] <cone-751> ffmpeg.git 03Carl Eugen Hoyos 07master:23ec8db8a074: Autodetect libxcb.
[00:56] <cone-751> ffmpeg.git 03Carl Eugen Hoyos 07master:f05855414ed4: lavc/utils: Make pix_fmt desc pointer const.
[01:51] <cone-751> ffmpeg.git 03Michael Niedermayer 07master:32e666c354e4: avcodec/g2meet: check tile dimensions to avoid integer overflow
[02:24] <cone-751> ffmpeg.git 03Martin Storsjö 07master:573b1de2d7f1: movenc: Don't use track_id to decide which track is the first in a moof
[02:24] <cone-751> ffmpeg.git 03Michael Niedermayer 07master:647405d1110c: Merge commit '573b1de2d7f1db71030f91ecdded7d0bc071f6b6'
[02:36] <cone-751> ffmpeg.git 03Martin Storsjö 07master:b2b79eca6fae: movenc: Don't check the custom IO flag when using faststart
[02:36] <cone-751> ffmpeg.git 03Michael Niedermayer 07master:39cce77faf15: Merge commit 'b2b79eca6fae2466a53c5daa163a37e2474364fc'
[02:44] <cone-751> ffmpeg.git 03Martin Storsjö 07master:1e0b81abe86d: movenc: Use a local variable consistently
[02:44] <cone-751> ffmpeg.git 03Michael Niedermayer 07master:ad600e10da6b: Merge commit '1e0b81abe86dc09dd34d60d57f92de5f12d65818'
[03:17] <cone-751> ffmpeg.git 03Martin Storsjö 07master:a9911a1b2df3: doc: Document the omit_tfhd_offset flag to the mp4 muxer
[03:17] <cone-751> ffmpeg.git 03Martin Storsjö 07master:600d5ee6b12b: movenc: Signal iso6 in compatible_brands when using tfdt
[03:17] <cone-751> ffmpeg.git 03Michael Niedermayer 07master:be787bccc00c: Merge commit 'a9911a1b2df37ca808e06dc30627215f4ec9a9e4'
[03:17] <cone-751> ffmpeg.git 03Michael Niedermayer 07master:7f24e1e1a2a9: Merge commit '600d5ee6b12bad144756b0772319bb04796bc528'
[06:31] <jamrial> xcbgrab broke a couple fate slots
[10:04] <cone-584> ffmpeg.git 03Carl Eugen Hoyos 07master:0e89d006d87a: Fix disabling libxcb-shm and libxcb-xfixes.
[10:08] <cone-584> ffmpeg.git 03Carl Eugen Hoyos 07master:8c0ae9015a2b: Fix xcbgrab build if shm is not available.
[10:27] <cone-584> ffmpeg.git 03Carl Eugen Hoyos 07master:792f0f20456c: Fix compilation of xcbgrab if xcb/shape.h is not available.
[12:37] <cone-584> ffmpeg.git 03Tobias Rapp 07master:0c9f0da0f765: ffprobe.xsd: add packets_and_frames element definition
[13:01] <saste> michaelni, reynaldo ping
[13:28] <michaelni> saste, ?
[13:28] <saste> michaelni, about OPW qualification ranking, when is it due?
[13:28] <saste> October 31 or what?
[13:29] <saste> I seem to remember definitive ranking was due in date November 7 or something
[13:29] <saste> also, in the OPW "View applications" page what's the "rank" thing?
[13:29] <michaelni> oct 31 for slots for slots from general funding, something later if you/we pay it
[13:30] <michaelni> https://wiki.gnome.org/OutreachProgramForWomen/Admin/ApplicationReview
[13:31] <saste> michaelni, can you quantify "something later"?
[13:33] <saste> michaelni, is it OK to assume Friday 7 Nov?
[13:34] <saste> reynaldo, ^^
[13:34] <michaelni> i see a mail that says nov7 
[13:35] <saste> michaelni, confirmed then
[13:35] <saste> michaelni, I'm writing a summary mail on ffmpeg-opw
[13:36] <saste> another thing, the rank thing is left to organizations, so i suggest to leave it blank, and let the ffmpeg admins to fill it once we defined our internal ranking
[13:36] <saste> so mentors won't have to wonder about how to fill it
[13:50] <wm4> are good english skills part of the OPW requirements?
[13:56] <saste> wm4, why?
[13:57] <wm4> saste: because they make communication really hard
[13:58] <BtbN> good english skills make communication hard?
[14:00] <michaelni> good english is mandatory, yes
[14:01] <michaelni> thats unless the mentor and student know a common language in which they can communicate very well
[14:01] <saste> michaelni, good communication skills, English is a factor but that's not the only one
[14:02] <michaelni> no but i see one or 2 applicants i have great difficulty understanding/communicating with due to their english skill
[14:03] <michaelni> and if basic communication doesnt work theres IMO not much hope for the project
[14:04] <BtbN> btw., https://github.com/01org/libyami is very interesting
[14:04] <BtbN> It's from the intel vaapi guys. Aparently they think their own api is so horrible, they made an official wrapper for it
[14:10] <nevcairiel> this is much more than a wrapper around vaapi
[14:14] <wm4> lol
[14:15] <wm4> more like Intel thought gstreamer is too bad
[14:24] <Daemon404> moar frameworks
[14:25] <kierank> intel have been emailing me trying to sell expensive frameworks
[14:26] <kierank> https://software.intel.com/sites/default/files/managed/cb/77/Intel_Media_Server_Studio_PB%209-2-2014.pdf
[14:36] <michaelni> btw saste, you might want to help arwa debug that segfault, she seems to have a problem with it somehow
[14:37] <kierank> __gb__: intel media server studio is just a wrapper around libva, right?
[14:41] <michaelni> saste, that is iam thinking she needs help there as there a multiple iterations of the patch failing in apparently the same/similar way
[14:48] <cone-584> ffmpeg.git 03Michael Niedermayer 07master:db442c873631: avdevice/xcbgrab: Fix offset support
[15:10] <anshul_mahe> I was looking at http://wiki.multimedia.cx/index.php?title=FFmpeg_Summer_Of_Code_2008#Snow_Job 
[15:11] <anshul_mahe> I know its quite old, but I was I interested in snow codec and would like to work in it
[15:12] <anshul_mahe> can any one tell me what part is already complete and what is still left
[15:12] <anshul_mahe> like is there any document for  "Document the whole Snow algorithm "
[15:12] <av500> anshul_mahe: the task is to document it
[15:12] <av500> afterwards, you have a document to read :)
[15:15] <Compn> https://github.com/kasbah/gsoc/blob/master/ffmpeg/doc/snow.txt
[15:15] <Compn> not sure if that has any other info from official snow.txt document
[15:16] <anshul_mahe> thanks
[15:18] <Daemon404> lolsnow
[15:31] <reynaldo> saste: Whats michaelni said ^
[15:31] <reynaldo> we are targeting Oct 31 for having pretty muche veryone ranked
[15:32] <wm4> why the fuck is snow still dragged along
[15:33] <wm4> by now it's a definitive failure
[15:33] <wm4> it's like GNU Hurd
[15:33] <wm4> just as video codec
[15:33] <wm4> yes there are people actively working on Hurd
[15:33] <wm4> and they're getting revolutionary things done like basic audio playback (probably only for the driver writer's hw of course)
[15:36] <anshul_mahe> wm4: actually I was doing this as part of MS programme, I wanted to work on some lossless video codec.
[15:37] <anshul_mahe> Do you want me to suggest different codec other then snow
[15:39] <anshul_mahe>  do you have any other suggestion for lossless video codec where it is good exposure to algorithms.
[15:39] <kierank> ffv1
[15:39] <kierank> j2k
[15:39] <kierank> etc
[15:40] <av500> tldr?
[15:40] <av500> er, tl3dr
[15:40] <anshul_mahe> ok, I will drop the idea snow then and look for ffv1
[16:11] <__gb__> kierank, there are ways to directly use the VA driver that ships with Intel Media Server Studio, or MediaSDK, or whatever is its name nowadays
[16:11] <__gb__> however, some encoder features depend on additional MediaSDK logics
[16:15] <wm4> i.e. undocumented API artifacts?
[16:18] <__gb__> no, finer tuned encoder logics. The API is here: http://cgit.freedesktop.org/vaapi/libva/tree/va/vendor/intel?h=staging but MediaSDK will have encoding modes that won't make you hit your head :)
[16:19] <__gb__> more detailed documentation of the various bits there live in the gpu docs
[16:25] <kierank> __gb__: I'm guessing the audio stuff is just software then
[16:28] <__gb__> yes, indeed
[16:28] <__gb__> you wanted to offload stuff to the tensilica chips? :)
[16:29] <kierank> just wanted to test out the aac encoder
[16:29] <kierank> I am happier though that all the stuff is in libva
[16:30] <kierank> since i care not for the higher level stuff
[17:23] <cone-584> ffmpeg.git 03Michael Niedermayer 07master:10e32618acce: avcodec/cook: check that the subpacket sizes fit in block_align
[18:38] <cone-584> ffmpeg.git 03Michael Niedermayer 07master:e91ba2efa949: avcodec/svq1dec: zero terminate embedded message before printing
[19:43] <cone-584> ffmpeg.git 03Michael Niedermayer 07master:47f1596ecef3: avcodec/dxtory: use unaligned copy for unaligned addresses
[21:09] <arwa> I am getting this error when I am running fate test " reference file './tests/ref/fate/filter-2xbr' not found", what reference file do I need to add?
[21:11] <ubitux> add GEN=1
[21:11] <ubitux> for the first run
[21:11] <ubitux> make fate-filter-2xbr GEN=1
[21:12] <arwa> okay
[21:12] <arwa> thanks
[21:13] <arwa> It is still giving the same error :/
[21:14] <arwa> now, it worked :D
[21:14] <arwa> thanks
[21:21] <ubitux> arwa: so did you compare the output with the reference?
[21:26] <arwa> i got a reference file, but how do I compare it with the output?
[21:27] <ubitux> no i mean
[21:27] <ubitux> comparing with the reference implementation
[21:27] <ubitux> not talking about fate here
[21:27] <ubitux> like, are you sure you output the exact same thing?
[21:29] <arwa> umm..I dont know actually...! How do 
[21:29] <arwa> I cross check it?
[21:29] <ubitux> well, find the reference implementation and use it
[21:29] <arwa> okay
[21:30] <arwa> but the current implementations work with opengl
[21:31] <ubitux> dunno, make a screenshot
[21:31] <arwa> okay
[00:00] --- Fri Oct 31 2014