[Ffmpeg-devel-irc] ffmpeg-devel.log.20190823

[burek]

[00:00:04 CEST] <thardin> how long does getting to/from greenland by boat take?
[00:00:43 CEST] <durandal_1707> small or big boat?
[00:01:09 CEST] <thilo> durandal_1707: whats you travel limit? Front door, your city, your country? expenses can be paid for
[00:01:46 CEST] <durandal_1707> my city, the current location
[00:02:25 CEST] <thilo> is there an airport in that city?
[00:02:44 CEST] <durandal_1707> i just do not like traveling
[00:02:45 CEST] <Compnn> can still make a product off gpl code too
[00:02:49 CEST] <Compnn> no reason to even relicense
[00:03:09 CEST] <durandal_1707> thilo: yes
[00:03:16 CEST] <thardin> Compnn: possibly, but proprietary vendors can hide it behind a cloud
[00:03:32 CEST] <thardin> of course there's always adapting it to clients' needs
[00:04:08 CEST] <thilo> so if it is not Kiev and somehow not the smalles airport, that meeting might happen in your city. mind to reveal which one it is?
[00:05:01 CEST] <durandal_1707> http://www.zagreb-airport.hr/en
[00:07:32 CEST] <thilo> I find direct flights from Berlin, Paris, London, Vienna
[00:07:47 CEST] <thilo> And the civil war is over, isn't it?
[00:07:58 CEST] <durandal_1707> i do not fly
[00:08:00 CEST] <thilo> Zagreb seems like a totally doable location
[00:08:17 CEST] <thilo> We might do the meeting in Zagreb, that's what I'm saying
[00:08:44 CEST] <durandal_1707> you do not have resources
[00:09:10 CEST] <thilo> I'll have someone pay for it
[00:15:06 CEST] <durandal_1707> ok then, maybe ask other people for opinions
[00:15:31 CEST] <thilo> good, I'll send a proposal to the ML tomorrow
[00:16:02 CEST] <TD-Linux> durandal_1707, prores is a pretty simple format. he can't be doing anything terribly complicated
[00:16:05 CEST] <kierank> durandal_1707: there is not much point unless michaelni_ comes
[00:16:10 CEST] <kierank> and for years we wanted michaelni_ to come to vdd
[00:16:13 CEST] <kierank> and it never happened
[00:16:17 CEST] <kierank> so i think it's pointless
[00:16:20 CEST] <thilo> michaelni_ will never come anywhere
[00:16:50 CEST] <thilo> so FFmpeg devs are checkmate by that?
[00:17:05 CEST] <durandal_1707> so organize meeting in his place/room
[00:17:15 CEST] <thilo> not gonna happen either
[00:17:35 CEST] <thilo> are we spinning about michaelni's no-meeting policy again?
[00:17:48 CEST] <kierank> well it seems paul is unhappy with the fuzzing patches which are all from michaelni_ 
[00:17:56 CEST] <thilo> lets just accept that, nothing will change it
[00:18:06 CEST] <durandal_1707> since when is that policy?
[00:18:21 CEST] <thilo> well many people are unhappy about many things, which is why we want to meet, right?
[00:18:49 CEST] <thilo> I don't know, its not my policy but michaelni's
[00:18:50 CEST] <durandal_1707> we can meet remotely also
[00:18:59 CEST] <Compnn> thilo, spinning spinning
[00:19:11 CEST] <Compnn> who is unhappy ?
[00:19:16 CEST] <thilo> Compnn: dizzy already
[00:19:36 CEST] <Compnn> that was other project idea, to meet in person
[00:19:44 CEST] <Compnn> meet in person didnt really help that project in vdd either
[00:20:09 CEST] Action: Compnn meets everyday in irc and email :p
[00:20:36 CEST] <Compnn> "if we meet in person we can all be friends" is a nice ideal
[00:21:11 CEST] <durandal_1707> Compnn can only meet himself?!
[00:21:21 CEST] <Compnn> i was at a few vdd durandal_1707 
[00:21:26 CEST] <Compnn> you can always drive there :P
[00:21:35 CEST] <Compnn> or take the bus, train, etc
[00:21:50 CEST] <durandal_1707> through oceans
[00:22:02 CEST] <thilo> driving through croatia? even i wouldn't do that
[00:22:14 CEST] <Compnn> i think i am on an island further away from you durandal_1707 :)
[00:23:02 CEST] <thilo> we will see how many people would like an actual meeting. if more people really prefer remote, than so be it. in my eyes, we are long overdue for an annual meeting
[00:23:14 CEST] <Compnn> where is next vdd? :P
[00:23:21 CEST] <durandal_1707> tokyo
[00:23:30 CEST] <Compnn> i could actually do tokyo
[00:24:18 CEST] <Compnn> bleh still 10 h flight
[00:24:36 CEST] <durandal_1707> which vlc devs live in japan?
[00:32:53 CEST] <Compnn> not many i'd guess
[00:40:48 CEST] <thardin> https://multun.net/obscure-c-features.html
[02:00:06 CEST] <cone-438> ffmpeg 03Jarek Samic 07master:d3cd33ab1b23: lavfi: add utilities to reduce OpenCL boilerplate code
[02:00:06 CEST] <cone-438> ffmpeg 03Jarek Samic 07master:5b5746b1e0d2: lavfi: modify avfilter_get_matrix to support separate scale factors
[02:00:06 CEST] <cone-438> ffmpeg 03Jarek Samic 07master:b29c7bcbf6bc: lavfi: add deshake_opencl filter
[02:41:56 CEST] <kierank> durandal_1707: none
[05:12:20 CEST] <Compnn> durandal_1707, pretty much i am the only person who wants a binary codec loader in ffmpeg :D
[05:12:33 CEST] <Compnn> well probably youtube would use it too
[08:37:15 CEST] <JEEB> ok, so we already have an AVPacketList
[08:39:43 CEST] <JEEB> but I remember we recently got some general list structure from which one could just push and pull from
[08:41:53 CEST] <JEEB> ok, I guess AVPacketList works for me
[12:24:36 CEST] <kurosu> I suspect I know what to do about the prores encoders (parts of the quantization are a bit wtfy) but I'd instead try to nerd-snipe Daemon404 as he is getting the experience needed atm for other reasons
[12:30:02 CEST] <durandal_1707> patch welcome
[12:31:46 CEST] <kurosu> My point
[12:36:22 CEST] <durandal_1707> why quantization is wtfy?
[12:37:48 CEST] <durandal_1707> whe should kill 2 encoders of prored by supreme one written by elvis
[12:46:36 CEST] <kurosu> 1) eob 2) dead zone 3) lagrangian 4) RDO
[12:48:53 CEST] <durandal_1707> please explain in details each of these points
[13:00:52 CEST] <kurosu> Not going to spend that much time over this, thus why I mentioned Derek
[14:40:41 CEST] <cone-124> ffmpeg 03Michael Niedermayer 07master:ac77c5492f5f: tools/target_dec_fuzzer: Do not increase max_pixels
[14:40:42 CEST] <cone-124> ffmpeg 03Michael Niedermayer 07master:c9fcf881e69e: avcodec/idcinvideo: Add 320x240 default maximum resolution
[14:40:43 CEST] <cone-124> ffmpeg 03Michael Niedermayer 07master:48b86dd8a6bf: avcodec/aacdec: Add FF_CODEC_CAP_INIT_CLEANUP
[15:32:39 CEST] <JEEB> wbs: first "d'uh, I'm dumb." the whole moov_written thing 8)
[15:35:22 CEST] <JEEB> but at least for basic fragmentation we seem to be rolling http://up-cat.net/p/e605863c
[15:46:09 CEST] <JEEB> wbs: also I am already seeing how many time-based things this would be breaking since the entries are no longer done as they get buffered first and only written when flushed xD
[16:10:24 CEST] <cone-124> ffmpeg 03James Almer 07master:92c32b0f0c3e: tools/target_dec_fuzzer: use refcounted packets
[16:10:44 CEST] <jamrial> durandal_1707: ^
[16:20:12 CEST] <thardin> \o/
[17:28:56 CEST] <cone-124> ffmpeg 03Jun Li 07master:0b34cdf424d6: avcodec/h264_parse: retry decoding SPS with complete NAL
[19:43:18 CEST] <tmm1> shared code between avcodec and avfilter has to go into avutil?
[19:46:26 CEST] <nevcairiel> could be in avcodec as well if it fits there
[19:53:02 CEST] <tmm1> oh ok, so avfilter dependency on avcodec is ok
[19:53:13 CEST] <tmm1> i want to add a vf_scale_{omx,v4l2}
[19:54:46 CEST] <nevcairiel> ideally we abstract hardware access through the hwcontext though
[20:11:24 CEST] <cone-124> ffmpeg 03elliottk 07master:711c59bc571c: Change libaom default to crf=32.
[20:14:58 CEST] <kierank> durandal_1707: I need to restart my threaded fuzzing
[21:42:58 CEST] <durandal_1707> kierank: ask google to do it
[22:06:55 CEST] <thardin> can we fuzz google?
[22:32:10 CEST] <cone-124> ffmpeg 03leozhang 07master:b2bb09bcc330: avformat/flvdec: delete unused code
[22:32:12 CEST] <cone-124> ffmpeg 03Michael Niedermayer 07master:965e766e4892: avcodec/rl2: set dimensions
[22:32:12 CEST] <cone-124> ffmpeg 03Michael Niedermayer 07master:4d4734bdc881: avcodec/lcldec: Check mthread_inlen instead of cliping
[22:32:13 CEST] <cone-124> ffmpeg 03Michael Niedermayer 07master:77abf3145344: avformat/mpsubdec: Check pts / duration before cast
[22:32:14 CEST] <cone-124> ffmpeg 03Michael Niedermayer 07master:1a0f106232f5: avcodec/pnm: Check magic bytes directly without pnm_get()
[22:32:15 CEST] <cone-124> ffmpeg 03Michael Niedermayer 07master:68f30567df56: avcodec/pnm: skip reading trailing bytes in get_pnm()
[22:32:16 CEST] <cone-124> ffmpeg 03Michael Niedermayer 07master:8a24d2cc304c: avcodec/parser: Optimize ff_combine_frame() with massivly negative next
[22:32:17 CEST] <cone-124> ffmpeg 03Michael Niedermayer 07master:34f6d1a7d785: avcodec/pnm_parser: Use memmove() to handle "overread"
[22:32:18 CEST] <cone-124> ffmpeg 03Michael Niedermayer 07master:5231e89eb9ee: avcodec/pictordec: Optimize picmemset() for single plane full lines
[22:32:19 CEST] <cone-124> ffmpeg 03Michael Niedermayer 07master:1c9a56b13965: avcodec/smacker: Check that not all tables are skiped
[22:32:20 CEST] <cone-124> ffmpeg 03Michael Niedermayer 07master:ae3d6a337ad2: avcodec/alac: Fix multiple integer overflows in lpc_prediction()
[23:13:24 CEST] <cone-124> ffmpeg 03Marton Balint 07master:6f499ae3904b: avformat/mpegtsenc: rename pcr_period variable to pcr_period_ms
[23:13:25 CEST] <cone-124> ffmpeg 03Marton Balint 07master:8bf732cda69b: avformat/mpegtsenc: add support for setting PCR interval for VBR streams
[23:13:26 CEST] <cone-124> ffmpeg 03Marton Balint 07master:5c119bf189c3: avformat/mpegtsenc: move some code around and simplify a bit
[23:13:27 CEST] <cone-124> ffmpeg 03Marton Balint 07master:2fb550893f75: avformat/mpegtsenc: get rid of packet counting for sdt/pat/pmt
[00:00:00 CEST] --- Sat Aug 24 2019