[Ffmpeg-devel-irc] ffmpeg-devel.log.20150519

[burek]

[04:16:26 CEST] <cone-879> ffmpeg 03Michael Niedermayer 07master:8ea9334b3154: avcodec/mss2: use < 0 instead of != 0 to check for error of vlc initialization
[04:16:27 CEST] <cone-879> ffmpeg 03Michael Niedermayer 07master:d0ac2f59aa36: avcodec/golomb: Remove disabled and broken code
[06:30:47 CEST] <pross> is anyone bothering with a SILK decoder (Skype SILK V3)?
[07:46:49 CEST] <Compn> pross : there is opus silk decoder in ffmpeg i think
[07:47:00 CEST] <Compn> is that older than the v3 you are talking about ?
[08:00:10 CEST] <Compn> http://linuxfonts.narod.ru/why.linux.is.not.ready.for.the.desktop.current.html
[08:00:12 CEST] <Compn> >slowpoke
[08:00:22 CEST] <Compn> a few ffmpeg topics :P
[08:25:15 CEST] <ubitux> > Mplayer (the most widely used video player in Linux) developers haven't yet merged VAAPI support.
[08:26:21 CEST] <ubitux> seems the guy is stuck 10 years ago
[08:29:07 CEST] <Compn> lol
[08:37:07 CEST] <pross> no mention of mvp. sigh
[08:39:46 CEST] <ubitux> well, mpv is still quite young, so it can be understood he didn't hear of it, but... vlc maybe?
[08:40:18 CEST] <ubitux> and lol @ hw accelerated vp9
[09:18:39 CEST] <wm4> ubitux: "overlooking" vlc and complaining about the lack of vp9 hwaccel... quite obviously this guy hasn't got a clue
[11:33:37 CEST] <cone-748> ffmpeg 03Nicolas George 07release/2.6:1f7343c8145f: lavd/xcbgrab: fix comparison with screen size.
[12:20:11 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07master:d3dcd2881247: avcodec/flacenc: Do not copy unused udata array -> 5x faster calc_rice_params()
[13:32:03 CEST] <nevcairiel> so did libav now re-implement nvenc? it must be different enough to warrant taking over the author line =p
[13:56:17 CEST] <ubitux> lol
[13:59:20 CEST] <j-b> Author line has no legal meaning.
[14:01:15 CEST] <nevcairiel> what does, then? all authorship information is lost, all it says is "based on soem code from some dude"
[14:02:18 CEST] <j-b> Do you add your name to all the files you touch?
[14:03:06 CEST] <nevcairiel> no, but you then at least have git history to see which code can be attributed to me
[14:04:06 CEST] <nevcairiel> but if you take someones patch, mess with it a bit, put yourself in as the author, there is no way to say who wrote what anymore
[14:05:23 CEST] <thardin> maybe plaintext needs per-character metadata
[14:05:44 CEST] <thardin> byte-level xattr, wheee
[14:05:58 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07master:80f9d6e0ab24: avcodec/flacenc: Support calculating rice parameters exactly
[14:05:59 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07master:df5e408d32a0: fate: Add test for -exact_rice_parameters 1
[14:06:01 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07master:7786a91b47e7: avcodec/flacenc: Move udata and sums to FlacSubframe
[14:07:20 CEST] <wm4> thardin: the real solution is assigning a lawyer to each programmer
[14:07:35 CEST] <wm4> and the programmer has to get confirmation from the lawyer for each line of code that is touched
[14:08:15 CEST] <nevcairiel> the real solution is not to waste time on re-implementing every single feature instead of just porting it, and then complaining that there isnt enough time to do all the things =p
[14:10:12 CEST] <j-b> that's a good point
[14:10:16 CEST] <j-b> NIH is strong
[14:10:26 CEST] <j-b> did anyone do a diff?
[14:14:03 CEST] <thardin> heh, rust neophyte in kostya's latest blog entry
[14:17:47 CEST] <ubitux> j-b: not easy, it's split between multiple files
[14:19:28 CEST] <ubitux> you can consider it a third implementation
[14:21:31 CEST] <nevcairiel> btw j-b, you wanted xll samples .. dcadec now has a test suite, you can use its samples to find all the things the native decoder doesn't do
[14:22:03 CEST] <nevcairiel> https://github.com/foo86/dcadec-samples
[14:26:34 CEST] <j-b> ubitux: ...
[14:26:46 CEST] <j-b> sometimes I just want to die.
[14:27:14 CEST] <ubitux> try to kill instead
[14:32:22 CEST] <j-b> maybe
[15:04:08 CEST] <wm4> what got NIHed this time?
[15:04:17 CEST] <wm4> nvenc?
[15:04:33 CEST] <j-b> yes
[15:08:00 CEST] <wm4> cute
[15:28:03 CEST] <iive> again?
[15:29:08 CEST] <BtbN> Are there even that many possible ways to implement an nvenc encoder? The api gives you a quite exact workflow.
[15:31:32 CEST] Action: BBB hugs j-b
[15:31:34 CEST] <BBB> poor you
[15:32:44 CEST] <BBB> is there a date for vdd yet? Im trying to plan my summer
[15:33:09 CEST] <j-b> BBB: 19/20 sept?
[15:33:16 CEST] <BBB> cool ty
[15:34:49 CEST] <kierank> is there a location ?
[15:35:02 CEST] <j-b> hawaii?
[15:35:12 CEST] <Rodeo> sure
[15:35:33 CEST] <Rodeo> but then it may be cheaper to charter a VideoLAN jet
[15:35:53 CEST] <Rodeo> we all rendezvous in Paris and take the VideoLAN jet for Hawaii
[15:36:11 CEST] <j-b> kierank: any wish?
[15:36:19 CEST] <av500> paris
[15:36:22 CEST] <kierank> london
[15:36:26 CEST] <Rodeo> are there castles in Hawaii?
[15:36:29 CEST] <Rodeo> LOL
[15:36:32 CEST] <av500> Rodeo: bouncy ones
[15:36:41 CEST] <Rodeo> ÖK, here goes: Geneva
[15:36:45 CEST] <Rodeo> Zürich
[15:36:47 CEST] <Rodeo> Lausanne
[15:36:49 CEST] <av500> too expensive
[15:36:52 CEST] <av500> allof them
[15:37:03 CEST] <BBB> get multicoreware to pay for it
[15:37:10 CEST] <Rodeo> Port au Prince
[15:38:26 CEST] <Rodeo> Japan
[15:38:43 CEST] <j-b> kierank: too expensive.
[15:38:47 CEST] <j-b> Rodeo: Geneva is doable
[15:38:49 CEST] <Rodeo> wm4: pardon my ignorance, what does "NIHed" mean?
[15:39:00 CEST] <Rodeo> Geneva is expensive too though
[15:39:13 CEST] <wm4> Not Invented Here
[15:39:36 CEST] <wm4> i.e. writing your own stuff instead of using others' because you think others' suck
[15:40:15 CEST] <Rodeo> ah
[15:40:31 CEST] <Rodeo> so there's a new patch floating around?
[15:40:42 CEST] <wm4> yes, on libav-devel
[15:41:10 CEST] <Rodeo> ah OK
[15:41:23 CEST] <Rodeo> I should really check it more often
[15:41:46 CEST] <Rodeo> regardless, I imagine the licensing issue remain the same?
[15:42:16 CEST] <j-b> yes
[15:42:17 CEST] <BtbN> But they did not use the unlicensed libnvenc stuff?
[15:59:00 CEST] <BBB> I think the main objection here is that youll now have two impelmentations
[15:59:00 CEST] <BBB> and typically ffmpeg will merge it so itll have both
[15:59:00 CEST] <BBB> and neither is very good
[15:59:01 CEST] <BBB> (see proresdec/enc)
[16:00:03 CEST] <Compn> and then vlc and j-b will have to support one or the other (or both)
[16:00:03 CEST] <Compn> with different api because fuck api :D
[16:09:44 CEST] <j-b> Compn: or none, because it's not GPL
[16:26:24 CEST] <Compn> hmm
[16:26:24 CEST] <Compn> adobe hds why u hate me
[16:49:23 CEST] <nevcairiel> BBB: i think micheal should not merge the libav re-implementation
[16:49:53 CEST] <BBB> good luck enforcing that
[16:50:05 CEST] <BBB> (I have no opinion)
[16:51:27 CEST] <Compn> nevcairiel : rfc it on mailing list
[16:51:37 CEST] <Compn> we can hash it out there
[17:48:53 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07master:710abaf2c5d8: avutil/pixfmt: Improve ascii art showing the meaning of the AVChromaLocation values
[19:09:09 CEST] <relaxed> Daemon404: ffmpeg now lists the pixel formats for libx265 with high bit depth enabled but when encoding, ffmpeg's output says it's using yuv420p. Then if I run "ffmpeg -i" on the output it lists the correct pixel format. Also, why does it list 8bit pixel formats too? Please see http://johnvansickle.com/ffmpeg-libx265.txt
[19:11:05 CEST] <nevcairiel> afaik, x265 supports 8-bit input even when encoding 10-bit, it converts it internally
[19:17:47 CEST] <nevcairiel> which should be good for you, its faster that way :)
[19:19:44 CEST] <relaxed> Good to know :)
[19:20:07 CEST] <relaxed> but if someone is looking at ffmpeg's output they may get confused
[19:21:29 CEST] <relaxed> it does say "Main 10 profile"
[19:54:56 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07master:5dfcb4f74de5: avcodec/flacenc: Simplify md5 calculation code by using AV_WL24()
[19:54:57 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07master:215410860f33: avcodec/flacenc: Mark pointers/array arguments const which are not changed
[20:34:07 CEST] <cone-748> ffmpeg 03Vittorio Giovara 07release/2.4:7244cefd6e6b: libvpx: Fix mixed use of av_malloc() and av_reallocp()
[20:34:08 CEST] <cone-748> ffmpeg 03Vittorio Giovara 07release/2.4:78a3a4580c5a: mp3: Properly use AVCodecContext API
[20:34:09 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:d61022d55f1a: Merge commit '7244cefd6e6ba7258cb022dfd7a284099d88a3e8' into release/2.4
[20:34:10 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:a7dedd8ea533: avformat/mp3dec: properly allocate dummy AVCodecContext
[20:34:11 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:86be9cda97d2: avformat/mp3dec: Check for avcodec_alloc_context3() failure
[20:34:12 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:d7888ff644f5: Merge commit '78a3a4580c5a547af4ae8682c662ea3a4699a599' into release/2.4
[20:47:48 CEST] <cone-748> ffmpeg 03Andreas Cadhalpun 07release/2.4:9f6c36d961d2: roqvideoenc: set enc->avctx in roq_encode_init
[20:47:49 CEST] <cone-748> ffmpeg 03Luca Barbato 07release/2.4:2af720fe5f04: x86: Put COPY3_IF_LT under HAVE_6REGS
[20:47:50 CEST] <cone-748> ffmpeg 03Vittorio Giovara 07release/2.4:8ae4d4e11762: mov: Fix little endian audio detection
[20:47:51 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:e6f17337cf20: Merge commit '9f6c36d961d27283808310e3ca1d8390b55fce9b' into release/2.4
[20:47:52 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:b8d3c3ea8604: Merge commit '2af720fe5f0418612a8fc26b0147a0e10414fcbe' into release/2.4
[20:47:53 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:60d0d68a8d6a: Merge commit '8ae4d4e117626313e0b7df746e82de84d00d160a' into release/2.4
[21:06:48 CEST] <cone-748> ffmpeg 03Andrey Utkin 07release/2.4:4415d0f3bbae: rtpenc_jpeg: Handle case of picture dimensions not dividing by 8
[21:06:49 CEST] <cone-748> ffmpeg 03Martin Storsjö 07release/2.4:f77c9d71615e: rtsp: Make sure we don't write too many transport entries into a fixed-size array
[21:06:50 CEST] <cone-748> ffmpeg 03John Högberg 07release/2.4:844201e35fe5: mpegts: Update the PSI/SI table only if the version change
[21:06:51 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:d5f2302aca4b: Merge commit '4415d0f3bbaeb287327ef101ae98d727a69d9af1' into release/2.4
[21:06:52 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:efb28e3c510b: Merge commit 'f77c9d71615e17414aacbb1720693b800a5a32d3' into release/2.4
[21:06:53 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:d45eef081469: Merge commit '844201e35fe575710be8218d45828df49b77f205' into release/2.4
[21:06:54 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:74b856e58bd9: avformat/mpegts: reset last_version on seeking
[21:06:55 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:52242a387b36: avformat/mpegts: Also parse the FMC descriptor if the codec has not been identified yet
[21:06:56 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:3d296c0ec1d5: avformat/mpegts: Factorize version checking code out
[21:06:57 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:5b1befb074cf: avformat/mpegts: reset last_ver on corrupted packets
[21:06:58 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:8f3787d068c1: avformat/mpegts: Detect changes in packet through CRC instead of just the 5bit version
[21:17:15 CEST] <cone-748> ffmpeg 03Luca Barbato 07release/2.4:744d813bcf52: avresample: Reallocate the internal buffer to the correct size
[21:17:16 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:386e80610de2: mux: Do not leave stale side data pointers in ff_interleave_add_packet()
[21:17:17 CEST] <cone-748> ffmpeg 03Andreas Cadhalpun 07release/2.4:ac1660509ecf: ape: Support _0000 files with nblock smaller than 64
[21:17:18 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:39cb08cdc20a: Merge commit '744d813bcf527481f2217428fa08bfee8642935b' into release/2.4
[21:17:19 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:8624b4927655: Merge commit '386e80610de282c92ad5897683ccaf2675766ac5' into release/2.4
[21:17:20 CEST] <cone-748> ffmpeg 03Andreas Cadhalpun 07release/2.4:4e4708ad8093: apedec: prevent out of array writes in decode_array_0000
[21:17:21 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:3cc8285439dd: Merge commit 'ac1660509ecfbeca7b63eb5ab8360011180e705b' into release/2.4
[21:36:43 CEST] <cone-748> ffmpeg 03Andreas Cadhalpun 07release/2.4:cb5324200ccd: alac: Reject rice_limit 0 if compression is used
[21:36:44 CEST] <cone-748> ffmpeg 03Andreas Cadhalpun 07release/2.4:21b21aed797b: nut: Check chapter creation in decode_info_header
[21:36:45 CEST] <cone-748> ffmpeg 03Andreas Cadhalpun 07release/2.4:a55a70644872: avi: Validate sample_size
[21:36:46 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:2070149cbbf8: Merge commit 'cb5324200ccdc693dd5b28dcd7d4b722fad83ea2' into release/2.4
[21:36:47 CEST] <cone-748> ffmpeg 03Andreas Cadhalpun 07release/2.4:63afe5b9140b: nutdec: check chapter creation in decode_info_header
[21:36:48 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:95b42188d619: Merge commit '21b21aed797b5e636adcf2df811f96a95f208930' into release/2.4
[21:36:49 CEST] <cone-748> ffmpeg 03Andreas Cadhalpun 07release/2.4:df0003030a97: avidec: avoid infinite loop due to negative ast->sample_size
[21:36:50 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:9f4979b24c9a: avformat/avidec: print a warning for negative sample_size
[21:36:51 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:9005075f397b: Merge commit 'a55a70644872027fdf76a75edf12a09c9008880f' into release/2.4
[21:48:23 CEST] <sh4rm4^bnc> is it possible to use ffserver with huffyuv codec ?
[22:00:56 CEST] <cone-748> ffmpeg 03wm4 07release/2.4:0f50c53cfb95: png: Set the color range as full range
[22:00:57 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:1ab5f63941fd: Merge commit '0f50c53cfb959162f2bccc1a2c2e066d35723595' into release/2.4
[22:00:58 CEST] <cone-748> ffmpeg 03Andreas Cadhalpun 07release/2.4:91aa6d8a8b9b: nutdec: fix memleaks on error in nut_read_header
[22:09:17 CEST] <cone-748> ffmpeg 03Luca Barbato 07release/2.4:1f64b018cbec: nut: Make sure to clean up on read_header failure
[22:09:18 CEST] <cone-748> ffmpeg 03Shiina Hideaki 07release/2.4:5549f693d218: mjpegenc: Fix JFIF header byte ordering
[22:09:19 CEST] <cone-748> ffmpeg 03Mikulas Patocka 07release/2.4:b37bfbfbe539: configure: Disable i686 for i586 and lower CPUs
[22:09:20 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:fb2d1b1b136b: Merge commit '1f64b018cbec018fa66a4a20f79958d9707913de' into release/2.4
[22:09:21 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:dcc800924925: Merge commit '5549f693d2181b3211427f65e48eaa2f4fc5a402' into release/2.4
[22:09:22 CEST] <cone-748> ffmpeg 03Michael Niedermayer 07release/2.4:93ceae749b75: Merge commit 'b37bfbfbe53917820d1f97312fa0b2e8c7a15217' into release/2.4
[00:00:00 CEST] --- Wed May 20 2015