michaelni changed the topic of #ffmpeg-devel to: Welcome to the FFmpeg development channel | Questions about using FFmpeg or developing with libav* libs should be asked in #ffmpeg | This channel is publicly logged | FFmpeg 7.0 has been released! | Please read ffmpeg.org/developer.html#Code-of-conduct
<Daemon404>
was this written pre-sidedata or something?
<Daemon404>
seems somewhat cursed
Livio has quit [Ping timeout: 255 seconds]
Krowl has quit [Read error: Connection reset by peer]
ngaullier has quit [Ping timeout: 256 seconds]
<Daemon404>
i regret reading the http server code
tufei_ has joined #ffmpeg-devel
tufei__ has quit [Ping timeout: 260 seconds]
Krowl has joined #ffmpeg-devel
mkver has quit [Remote host closed the connection]
mkver has joined #ffmpeg-devel
<thardin>
why do we implement http again?
<j-b>
lol
<thardin>
that's my point. but suggesting we should use well-maintained libraries instead of reimplementing stuff is heresy of the highest order
<j-b>
BURN, HERETIC 𐠒
<thardin>
we mustn't have dependencies because ffabrice (pbuh) wrote his own mpeg2 encoder due to there being no decent libre one back in 2000
<j-b>
/s
<Lynne>
Daemon404: same reason png exports the intended gamma as metadata
<Lynne>
too miscellaneous to define side data
<Lynne>
side data is inflexible or bothersome to define for such,
jamrial has quit []
jamrial has joined #ffmpeg-devel
<Daemon404>
Lynne, certainly confusing from the perspective of an API user
<Gramner>
NIH:ing simple features is reasonable as having dependencies comes with it's own set of caveats, but http with all it's quirks seems like an excellent candidate for using some existing widely used and well-tested library
<Lynne>
like curl!
<Lynne>
there was a patch to add support for that years ago IIRC
cone-314 has joined #ffmpeg-devel
<cone-314>
ffmpeg James Almer master:6b6a0fc53df5: avformat/iamf_writer: reject duplicated stream ids in a stream group
Arsen has quit [Quit: Quit.]
philipl has quit [Ping timeout: 255 seconds]
Arsen has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
<JEEB>
Gramner: I mean like 15-20 years ago HTTP was simpler. since then... :D
philipl has joined #ffmpeg-devel
<Lynne>
well, http/2 was a failure
<Lynne>
http/3 was unusable outside of browsers until literally a few months ago when openssl finally got support for it
<thardin>
I wouldn't say HTTP was simple even 20 years ago. it's easy to get something kinda working, and that is also the danger
<JEEB>
stuff like keepalive etc
<cone-314>
ffmpeg arch1t3cht master:5a856ac6e6a6: avcodec/h264dec: Properly mark frames as recovered when draining
<cone-314>
ffmpeg arch1t3cht master:728ffe6ca6d4: avcodec/h264dec: Handle non-recovered frames when draining
<cone-314>
ffmpeg arch1t3cht master:e9e1932c1ca8: avcodec/h264dec: Reindent after the previous commit
System_Error has quit [Remote host closed the connection]
Krowl has joined #ffmpeg-devel
System_Error has joined #ffmpeg-devel
<cone-314>
ffmpeg Mark Thompson release/7.0:0d851a82dd97: lavc/av1: Record reference ordering information for each frame
<cone-314>
ffmpeg Mark Thompson release/7.0:48721a415a05: lavc/vulkan_av1: Use av1dec reference order hint information
<cone-314>
ffmpeg Lynne release/7.0:8dfafe536657: vulkan_av1: add workaround for NVIDIA drivers tested on broken CTS
<Lynne>
ffmpeg 7.0 is still only in experimental in debian, and not packaged by ubuntu yet, at least
<Lynne>
and arch, the channel layout switch broke most users
<JEEB>
yea, most API clients not following new warnings from master would have not adjusted
<jamrial>
Lynne: really? projects haven't migrated to the new channel layout api?
<jamrial>
wonder if they will do like with 4.4 and create yet another ffmpeg package to keep supporting such projects
<jamrial>
they have ffmpeg4.4 basically for vlc stable
<JEEB>
I mean, just like we learned before, there are N projects which just don't want to care about FFmpeg. they just want to focus on their own part of the bargain
<BtbN>
VLC _still_ being stuck on 4.4 is just sad at this point. It's the sole reason a lot of distros have not made the jump to... 5
<jamrial>
this is why people argued the two year deprecation period was pointless
<jamrial>
projects migrate either on the spot to get rid of deprecation warnings, or when they get compilation errors due to missing api
<BtbN>
Maybe we should do "active deprecation"
<BtbN>
i.e. when something gets deprecated, you have to set a define in your project, otherwise it fails to build
<BtbN>
AV_YES_I_WILL_MIGRATE_TO_FFMPEG_7
<JEEB>
BtbN: wasn't that thankfully just for the vaapi module? although with AVChannelLayout being the only layout now, of course that's going to break things with 7.0 :V
<BtbN>
yes, it's vaapi. But distros don't care
<BtbN>
hwaccel is kinda a priority as well
<haasn>
statically built binaries are the future
<haasn>
just pin all your dependencies to ancient versions and you never have to upgrade
<BtbN>
I'm still waiting for the rustpocalypse. Cause that's what they do, and it's such an obvious security nightmare waiting to explode...
<BtbN>
Go already had its first instance. They NIH'd crypto, and it was broken. And to this day, the broken version is still used in various projects
<jamrial>
like, the channel layout migration is more or less s/.channels/.ch_layout.nb_channels/ and s/.channel_layout/.ch_layout.u.mask/ for all the pre new api usecases
<jamrial>
yes, i know it's a bit more than that, but compared to migrating to the new decode/encode api, it's super trivial
Krowl has quit [Read error: Connection reset by peer]
<Lynne>
jamrial: I only switched cyanrip to the new channel layout a few months ago
<Lynne>
it wasn't so bad, as long as you copy+paste code from already ported code, the channel layout struct is daunting
<jamrial>
well, you wanted opaque pointers and names
<Lynne>
guilty
<Lynne>
I regret nothing, though!
<jamrial>
but for 99% of uses, just looking at nb_channels and in case of native layouts u.mask is enough
System_Error has quit [Ping timeout: 260 seconds]
<jamrial>
we have one decoder exporting ambisonic layout and one demuxer exporting custom for now
HarshK23 has quit [Quit: Connection closed for inactivity]
<Lynne>
which decoder was that again? opus?
<jamrial>
yeah
System_Error has joined #ffmpeg-devel
AbleBacon has joined #ffmpeg-devel
HarshK23 has joined #ffmpeg-devel
muiz has joined #ffmpeg-devel
muiz has quit [Changing host]
muiz has joined #ffmpeg-devel
muiz has quit [Client Quit]
muiz has joined #ffmpeg-devel
muiz has quit [Changing host]
muiz has joined #ffmpeg-devel
psykose has quit [Remote host closed the connection]
psykose has joined #ffmpeg-devel
psykose has quit [Remote host closed the connection]
psykose has joined #ffmpeg-devel
psykose has quit [Remote host closed the connection]
psykose has joined #ffmpeg-devel
jarthur has joined #ffmpeg-devel
cone-314 has quit [Quit: transmission timeout]
Krowl has joined #ffmpeg-devel
Sean_McG has quit [Quit: leaving]
Sean_McG has joined #ffmpeg-devel
Gramner has quit [Ping timeout: 256 seconds]
Gramner has joined #ffmpeg-devel
razor_ has quit [Ping timeout: 260 seconds]
razor_ has joined #ffmpeg-devel
cone-763 has joined #ffmpeg-devel
<cone-763>
ffmpeg Andreas Rheinhardt master:639013aafcd9: avformat/lc3: Only allow AV_CODEC_ID_LC3 in muxer
Krowl has quit [Read error: Connection reset by peer]
razor_ has quit [Ping timeout: 240 seconds]
iive has joined #ffmpeg-devel
Sean_McG has quit [Quit: leaving]
Livio has joined #ffmpeg-devel
ngaullier has joined #ffmpeg-devel
microchip__ has joined #ffmpeg-devel
microchip_ has quit [Ping timeout: 256 seconds]
ngaullier has quit [Ping timeout: 260 seconds]
microchip__ is now known as microchip_
microchip_ has quit [Quit: There is no spoon!]
microchip_ has joined #ffmpeg-devel
mkver has quit [Ping timeout: 264 seconds]
darkapex has quit [Remote host closed the connection]
darkapex has joined #ffmpeg-devel
ccawley2011 has quit [Read error: Connection reset by peer]
kurosu has quit [Quit: Connection closed for inactivity]
System_Error has quit [Remote host closed the connection]