[Ffmpeg-devel-irc] ffmpeg-devel.log.20180830

[burek]

[06:44:04 CEST] <gagandeepsingh> finally quiz in university over, going to update the patches
[10:01:57 CEST] <cone-385> ffmpeg 03Raphael Graf 07master:a3c6b7ff5903: frei0r: handle string params
[11:48:34 CEST] <durandal_1707> anybody compared av1 with xvc?
[12:05:25 CEST] <cone-385> ffmpeg 03Paul B Mahol 07master:ca079b095499: avcodec/get_bits: add cached bitstream reader
[12:05:26 CEST] <cone-385> ffmpeg 03Paul B Mahol 07master:042fc12344d6: doc/libav-merge: bitstream reader is now merged
[12:05:27 CEST] <cone-385> ffmpeg 03Paul B Mahol 07master:562f00ed0720: avcodec/utvideodec: use cached bitstream reader everywhere except on x86_32
[13:20:52 CEST] <durandal_1707> -300 scpr version 3, now at ~2111, need to be less than 1000
[13:34:46 CEST] <TekniQue> I was looking at this ticket, and I tried out the patch someone submitted, that almost worked, needed one little change to really work properly https://trac.ffmpeg.org/ticket/5419
[13:36:14 CEST] <TekniQue> I'm wondering by what route it's best for me to get patches accepted and merged into the main branch
[13:44:42 CEST] <durandal_1707> TekniQue: http://ffmpeg.org/developer.html#Submitting-patches-1
[15:24:21 CEST] <kierank> January: maybe worth having discussion here
[16:14:50 CEST] <durandal_1707> nothing worth here, feel free to leave
[17:41:31 CEST] <cone-786> ffmpeg 03Michael Niedermayer 07master:6b1b5af0240b: avfilter/vf_frei0r: Remove duplicate }, fix build
[22:05:24 CEST] <j-b> What's the use of using wolfSSL?
[22:06:19 CEST] <BtbN> The hell is WolfSSL?
[22:08:01 CEST] <atomnuker> pull
[22:08:23 CEST] <atomnuker> erm switched focus halfway through
[22:11:34 CEST] <durandal_1707> j-b: what's use of vlc supporting cineform but not sheervideo?
[22:11:48 CEST] <j-b> durandal_1707: no idea.
[22:12:03 CEST] <j-b> durandal_1707: people ask me for cineform all the time. Almost never sheervideo.
[22:15:33 CEST] <durandal_1707> Compn: still waiting for your changes to mplayer repo
[22:21:27 CEST] <durandal_1707> jamrial: can you please explain what i should do? I failed to understand your message.
[22:22:18 CEST] <jamrial> look at how UNCHECKED_BITSTREAM_READER is defined and checked in get_bits.h, then do the same for the new define
[22:23:21 CEST] <durandal_1707> what about using: #if ARCH_X86_32 #define ...
[22:29:10 CEST] <jamrial> durandal_1707: https://pastebin.com/vHWdi9Nz
[22:34:59 CEST] <kurosu> so, do people in here think OPEN_READER & friend should eventually go (like in libav) or there is a *functional* difference
[22:36:16 CEST] <kurosu> from a quick glance, I'd say this is mostly better for the unchecked case, and having some heavily used struct members as variables, which I imagine helps compilers use registers
[22:38:07 CEST] <kurosu> because the opposite would be to then port durandal_1707's cached part to this macro API
[22:43:12 CEST] <durandal_1707> kurosu: someone needs to make benchmarks, lot of that belongs to microoptimization
[22:44:06 CEST] <BBB> holy crap that email does not render correctly in gmail
[22:44:16 CEST] <BBB> its like word vomit all over my screen, what happened
[22:44:26 CEST] <BBB> (the wolfssl one)
[22:44:30 CEST] <BBB> does anyone else have that issue also?
[22:44:48 CEST] <kurosu> durandal_1707, yeah, but that would be ignoring why there is a benefit and how it could be reused
[22:46:29 CEST] <durandal_1707> BBB: mailing list archive attachment is fine
[22:48:27 CEST] <durandal_1707> jamrial: that patch have some issues, i fixed it and will push it under your authorship
[22:58:58 CEST] <cone-733> ffmpeg 03James Almer 07master:9305bdc68f09: avcodec/get_bits: actually make cached reader correctly disabled
[23:11:32 CEST] <thardin> https://www.gamasutra.com/blogs/TrevorDiem/20171018/301247/The_Full_Throttle_Remastered_FMV_Pipeline_Part_1.php  http://gamasutra.com/blogs/TrevorDiem/20180829/321112/The_Full_Throttle_Remastered_FMV_Pipeline_Part_2.php  interesting considering the smush support in ffmpeg
[00:00:00 CEST] --- Fri Aug 31 2018