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 6.1.1 has been released! | Please read ffmpeg.org/developer.html#Code-of-conduct
philipl has quit [Quit: leaving]
user23 has quit [Remote host closed the connection]
user23 has joined #ffmpeg-devel
philipl has joined #ffmpeg-devel
iive has quit [Quit: They came for me...]
blb has quit [Ping timeout: 255 seconds]
blb has joined #ffmpeg-devel
Gramner has joined #ffmpeg-devel
hamzah has joined #ffmpeg-devel
mkver has joined #ffmpeg-devel
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg-devel
dykai has joined #ffmpeg-devel
thilo has quit [Ping timeout: 256 seconds]
jarthur has quit [Quit: jarthur]
thilo has joined #ffmpeg-devel
jarthur has joined #ffmpeg-devel
marcj has joined #ffmpeg-devel
dykai has quit [Ping timeout: 264 seconds]
cone-232 has joined #ffmpeg-devel
<cone-232>
ffmpeg Andreas Rheinhardt master:dfb9d8a5a2f5: avformat/avio: Make avio_print_string_array() accept const pointers
<cone-232>
ffmpeg Andreas Rheinhardt master:1e98cc47879f: avcodec/bsf/mp3_header_decompress: Remove BSF
<cone-232>
ffmpeg Andreas Rheinhardt master:b295aafb0820: swresample/swresample: Constify swr_convert()
<cone-232>
ffmpeg Andreas Rheinhardt master:58e3ef7f5468: avutil/timestamp: Constify av_ts_make_time_string()
<cone-232>
ffmpeg Andreas Rheinhardt master:6c693da690d2: fftools/ffmpeg_demux, sync_queue: Constify a bit
<cone-232>
ffmpeg Andreas Rheinhardt master:88b3b09afa5e: avcodec/aacenc: Move initializing DSP out of aacenc.c
<cone-232>
ffmpeg Andreas Rheinhardt master:4460cb485be4: avcodec/svq1enc: Move initializing DSP out of svq1enc.c
<cone-232>
ffmpeg Andreas Rheinhardt master:33a3a23b791d: avcodec/vvc/vvc_intra: Move utils out of vvc_intra.c
<cone-232>
ffmpeg Andreas Rheinhardt master:540d8e77404a: avcodec/vvc/vvc_mvs: Add proper header include
<cone-232>
ffmpeg Andreas Rheinhardt master:cfe21a66e20f: checkasm/vvc_mc: Don't use declare_func_emms
<cone-232>
ffmpeg Andreas Rheinhardt master:0f944c8b2771: tests/checkasm: Improve included headers
<cone-232>
ffmpeg Andreas Rheinhardt master:d25b3497f2cd: avcodec/x86/rv40dsp, simple_idct: Remove remnants of MMX
dykai has joined #ffmpeg-devel
mkver has quit [Ping timeout: 252 seconds]
jamrial has quit []
AbleBacon has quit [Read error: Connection reset by peer]
Sean_McG has quit [Quit: leaving]
dykai has quit [Ping timeout: 260 seconds]
elastic_dog has quit [Ping timeout: 255 seconds]
user23 has quit [Remote host closed the connection]
<cone-232>
ffmpeg Andreas Rheinhardt master:45c612702f6b: configure: Add missing vvc_decoder->cbs_h266 dependency
<cone-232>
ffmpeg Michael Niedermayer master:b95c0d93fff0: avcodec/vvc/vvcdec: allow vvc_decode_close() on failed open
hamzah has quit [Ping timeout: 250 seconds]
<elenril>
><@JEEB> and defining a variable and initializing it to the final value should generally be the best way
<elenril>
I disagree
<elenril>
I believe it's bad practice to mix constant/trivial initialization with complex operations like memory allocation
Poorvagaikar2003 has joined #ffmpeg-devel
Poorvagaikar2003 has quit [Ping timeout: 260 seconds]
Poorvagaikar2003 has joined #ffmpeg-devel
dykai has joined #ffmpeg-devel
TheSashm_ has joined #ffmpeg-devel
TheSashmo has quit [Ping timeout: 264 seconds]
lexano has quit [Ping timeout: 264 seconds]
MisterMinister has quit [Ping timeout: 272 seconds]
lexano has joined #ffmpeg-devel
dykai has quit [Ping timeout: 268 seconds]
<cone-232>
ffmpeg Marth64 master:a1304272c3a0: libavformat/dvdvideo: add DVD-Video demuxer, powered by libdvdread and libdvdnav
Poorvagaikar2003 has quit [Ping timeout: 268 seconds]
dionisis has joined #ffmpeg-devel
hankin has joined #ffmpeg-devel
hankin has quit [Quit: Client closed]
Poorvagaikar2003 has joined #ffmpeg-devel
Poorvagaikar2003 has quit [Remote host closed the connection]
Poorvagaikar2003 has joined #ffmpeg-devel
Poorvagaikar2003 has quit [Remote host closed the connection]
Poorvagaikar2003 has joined #ffmpeg-devel
Poorvagaikar2003 has quit [Ping timeout: 268 seconds]
Poorvagaikar2003 has joined #ffmpeg-devel
elastic_dog has quit [Ping timeout: 268 seconds]
cone-232 has quit [Quit: transmission timeout]
elastic_dog has joined #ffmpeg-devel
Poorvagaikar2003 has quit [Ping timeout: 260 seconds]
Krowl has joined #ffmpeg-devel
dykai has joined #ffmpeg-devel
dykai has quit [Ping timeout: 260 seconds]
jamrial has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
mkver has joined #ffmpeg-devel
Poorvagaikar2003 has joined #ffmpeg-devel
Poorvagaikar2003 has quit [Ping timeout: 240 seconds]
dre-droid has joined #ffmpeg-devel
<mkver>
jamrial: Please wait with applying that flv set; at least the last patch.
<jamrial>
ok
<haasn>
elenril: what about if you have scoped destructors
stevenliu has quit [Remote host closed the connection]
stevenliu has joined #ffmpeg-devel
AbleBacon has joined #ffmpeg-devel
user23 has joined #ffmpeg-devel
<mkver>
elenril: Have you actually tested the dvd demuxer?
<mkver>
elenril: Screw that question.
dre-droid has left #ffmpeg-devel [#ffmpeg-devel]
dre-droid has joined #ffmpeg-devel
dre-droid has quit [Quit: Leaving...]
DuckInspector has joined #ffmpeg-devel
<kierank>
Can we ban balling
<kierank>
From trac
Poorvagaikar2003 has joined #ffmpeg-devel
<kierank>
I emailed cc
<JEEB>
he was already banned from IRC and ML so I wouldn't consider trac anything extra
<haasn>
kierank: +1
<JEEB>
so IMO no CC required for that since he's already supposed to be banned in our communications
<courmisch>
he has been unbanned by elenril
<courmisch>
from IRC
<courmisch>
or more accurately, he unbanned Mordor IP ranges
<kierank>
The reason it needs to go to CC is because of the controller of trac.ffmpeg.org
<JEEB>
right
AbleBacon_ has joined #ffmpeg-devel
AbleBacon has quit [Ping timeout: 256 seconds]
<elenril>
mkver: I did
<elenril>
not very extensively, admittedly
<elenril>
haasn: ?
user23 has quit [Ping timeout: 250 seconds]
<j-b>
good morning
<haasn>
elenril: int *tmp __attribute__ ((__cleanup__(free)) = malloc(1000);
<j-b>
libdvdread is NOT LGPL
<elenril>
seems non-portable
<haasn>
it definitely isn't
<haasn>
s/free/av_freep/
<elenril>
but that's beside the point
<elenril>
the main issue i have with it is that it needs to be checked
<elenril>
and you're separating the check from the operation itself
<haasn>
fair
<elenril>
and making it less clear that a "heavy" operation is being performed, by mixing it with other declarations
<jamrial>
haasn: that kind of "on return" behavior is something afaik c++ has that i wish c had
<JEEB>
elenril: just to explain my thought with these things is: 1) I like early exit, so if there are any checks that you need to do with input values, you do those first, and kick the logic out of that function 2) thus in the best case you have those checks at the ver beginning of the function 3) the C options we use don't let the declarations be done after such a check 4) if you add declaration at the
<JEEB>
beginning with a safe "failure" value, then someone might come and note that the value is not utilized and make it uninitialized (zomg!) 5) then I learned that you can just start a new context block to declare stuff there
<JEEB>
but anyways, this is not a big thing. I already moved some declarations on top with comments from jamrial. I just want you to understand the flow of thought with those
AbleBacon_ is now known as AbleBacon
<haasn>
elenril: "by mixing it with other declarations" -> this is only an issue if you're married to -Wdeclaration-after-statement
<JEEB>
that this is not just random adding of { }, it's trying to enable declarations after some logic has already been done.
<JEEB>
most often after some sort of early exit
<haasn>
or if you need a `cleanup:` label on exit, which attribute __cleanup__ solves
DuckInspector has quit [Remote host closed the connection]
Poorvagaikar2003 has quit [Remote host closed the connection]
<haasn>
maybe we will get them as part of the language standard in a decade
<elenril>
JEEB: I understand why you do it, but I still wish you didn't
<JEEB>
alright, just wanted to make sure that you understand that there's a logic behind it
<JEEB>
because with something like ret = AVERROR_BUG or ptr = NULL, you never know when someone decides to do the compiler optimization for it :)
<JEEB>
(and in the worst case then someone else changes the latter code so the uninitialized value is read)
user23 has joined #ffmpeg-devel
zhupengfei08 has joined #ffmpeg-devel
zhupengfei08 has quit [Quit: Client closed]
DuckInspector has joined #ffmpeg-devel
<DuckInspector>
Hi! I'm new to contributing to ffmpeg and wanted to work on a ticket to get more comfortable with the codebase. Is there a specific one I should get started on? I apologise if this is not the place to ask
<JEEB>
not sure how many specific known simple things we have
<microchip_>
work on adding cover art support for ogg
<mkver>
One could also do something similar in dump.c, but the strings printed by it do not completely coincide with the ones from av_disposition_to_string() and would need to be postprocessed ('_'->' ' and a leading "non").
Krowl has quit [Read error: Connection reset by peer]
jarthur has joined #ffmpeg-devel
Krowl has joined #ffmpeg-devel
jamrial has quit [Ping timeout: 240 seconds]
Krowl has quit [Read error: Connection reset by peer]
<mkver>
Lynne: "const crime crimes = (crime){ 11 }": We don't use this (and when we do, it's a bug, because MSVC doesn't like it; see a25a60d763061da695c3c563dcafbf392aea5579).
hamzah has joined #ffmpeg-devel
<Lynne>
really? we have a massive codebase, there's bound to be lots of this in old code
<mkver>
Definitely not in code that can be compiled with MSVC.
<mkver>
And I also don't see it elsewhere; while we have lots of code, we do not have an unbounded number of idioms.
hankin has quit [Quit: Client closed]
j45_ has joined #ffmpeg-devel
j45 has quit [Ping timeout: 260 seconds]
j45_ is now known as j45
j45 has quit [Changing host]
j45 has joined #ffmpeg-devel
<nevcairiel>
I think the error was about using variables in that construct, with constants it would've worked, but regardless its apparently not expected to work
jamrial has joined #ffmpeg-devel
<nevcairiel>
although the macro is only used with constant values .. guess not. I see that article above says that structs are not considered constant in some situations like these
<mkver>
The commit is exactly about the stuff mentioned by thephd.dev: (VLCElem){.sym = (a), .len = (b) } creates an unnamed object which not a constant expression even if all initializers are const expressions.