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 7.0.1 has been released! | Please read ffmpeg.org/developer.html#Code-of-conduct
dec05eba has joined #ffmpeg-devel
lemourin has quit [Ping timeout: 256 seconds]
dec05eba has quit [Quit: Leaving]
lemourin has joined #ffmpeg-devel
<michaelni>
BtbN, --global doesnt work with www-data because its home is /var/www and if one places a .gitconfig there it doesnt work from inside apache because i thing it unsets $HOME
<michaelni>
but thats not the problem
<BtbN>
What else is?
<BtbN>
That's the only really disruptive change git made in a while
<michaelni>
ive set safe.directory systemwide and thats what we have with michal.git now
<BtbN>
systemwide "*" would probably also work
<michaelni>
raz said he will look at it tomorrow
<BtbN>
not like that system has any removable disks plugged into it or is multi-user
<BtbN>
safe.directory is a list, not one property. So if you don't want to just set "*", the only other solution is to add the path to every single repo there
<michaelni>
the big mystery is why it doesnt work when its set, i mean the 500 is gone and its a normal http success then
<michaelni>
yes i would set it to a list of every repository
<BtbN>
seems unneccesary to me. A git server is not susceptible to the issue the flag fixes
<michaelni>
maybe, i dont know
<BtbN>
"This vulnerability affects users working on multi-user machines where a malicious actor could create a .git directory in a shared location above a victim’s current working directory."
<BtbN>
That's really not the case on a server where nobody but admins log in into interactive shells
<BtbN>
If after fixing that, there is still another issue, that's likely an entirely unrelated second issue
<michaelni>
ok, ive done a workaround, raz will look into what really was going on
<BtbN>
Just downgraded git again?
<michaelni>
yes
<michaelni>
if the fix is not relevant for us thats simplest
<BtbN>
well, if adding the repository to safe.directories does not fix it, something else is going on as well
<BtbN>
so yeah, for a quick fix absolutely fine
<michaelni>
yes, iam really confused
<BtbN>
the security issue is not relevant here
<michaelni>
maybe ubuntu messed up the backported fixes but i dont want to say it to loud because i dont know at all it could be some really dumb mistake on our / mine side too
<BtbN>
it's odd that gitweb seems to work just fine, but cloning doesn't
<michaelni>
its odd that the git-http thing needs to be the same user as the repo. I dont know how other people set this up but serving git to the public i imagine wouldnt be done frm the user of each repo
<BtbN>
It shouldn't need to. It just needs read access
<BtbN>
(And now this safe.directory thing. The fact that it triggers a 403 shows that it actually is a different user, since that's the condition it checks on)
<michaelni>
yes http-data is in the same group as git and that group should have only read access
<michaelni>
and i imagined that git-http cgi thing would never ever execute a hook because if it did that would just be dumb
<BtbN>
Yeah, the fix for the security issue seems WAY over-board
<BtbN>
like, just disable executing stuff and print a warning if it was attempted
<BtbN>
and not just flat out hard-fail every single repo operation
thilo has quit [Ping timeout: 240 seconds]
thilo has joined #ffmpeg-devel
thilo has quit [Changing host]
thilo has joined #ffmpeg-devel
iive has quit [Quit: They came for me...]
_whitelogger has quit [Remote host closed the connection]
tufei has quit [Remote host closed the connection]
<Lynne>
jamrial: could you add the xhe-aac fate samples I linked in the last commit?
<Lynne>
they should just be added to aac/
<jamrial>
ok
<Lynne>
thanks
<Lynne>
both samples and references are from the official test samples, I only zero-padded the reference output to match what we output
<jamrial>
Lynne: isn't there a fate option to discard samples while doing comparison?
<jamrial>
CMP_SHIFT
<jamrial>
or is that for something else?
tufei has joined #ffmpeg-devel
<Lynne>
no, that's for the start offset
<Lynne>
tiny_psnr does support discarding last N samples, but that's not hooked up
tufei has quit [Ping timeout: 260 seconds]
<jamrial>
may be worth adding it
<Lynne>
hmm, it doesn't do what I think it does
<Lynne>
yeah, it's definitely not a trim parameter
<Lynne>
I think its best we leave it as-is, the reference files are wavs anyway which I converted, a few bytes at the end don't add too much, and when we eventually fix all lossy codec padding issues in mp4, it'll still work
<jamrial>
ok
<jamrial>
all four files uploaded
<Lynne>
thanks
cworley has quit [Quit: bye]
cworley has joined #ffmpeg-devel
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg-devel
jamrial has quit []
Martchus_ has joined #ffmpeg-devel
Martchus has quit [Ping timeout: 264 seconds]
zsoltiv_ has quit [Ping timeout: 260 seconds]
AbleBacon has quit [Read error: Connection reset by peer]
rvalue has quit [Ping timeout: 252 seconds]
DHE has joined #ffmpeg-devel
DeHackEd has quit [Ping timeout: 240 seconds]
kurosu has joined #ffmpeg-devel
rvalue has joined #ffmpeg-devel
arch1t3cht7 has joined #ffmpeg-devel
arch1t3cht has quit [Ping timeout: 256 seconds]
arch1t3cht7 is now known as arch1t3cht
naveen521kk has joined #ffmpeg-devel
naveen521kk has quit [Client Quit]
tr4nq_ has quit [Quit: Lost terminal]
tr4nq_ has joined #ffmpeg-devel
scat117 has quit [Ping timeout: 240 seconds]
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg-devel
Krowl has joined #ffmpeg-devel
ngaullier has joined #ffmpeg-devel
System_Error has quit [Ping timeout: 260 seconds]
System_Error has joined #ffmpeg-devel
ccawley2011 has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
dionisis has quit [Ping timeout: 260 seconds]
Krowl has joined #ffmpeg-devel
dionisis has joined #ffmpeg-devel
natto has quit [Read error: Connection reset by peer]
natto17 has joined #ffmpeg-devel
<thardin>
>Test source failed.
<thardin>
well that's an annoying test
IndecisiveTurtle has joined #ffmpeg-devel
mkver has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
arch1t3cht1 has joined #ffmpeg-devel
arch1t3cht has quit [Ping timeout: 256 seconds]
arch1t3cht1 is now known as arch1t3cht
lexano has joined #ffmpeg-devel
jamrial has joined #ffmpeg-devel
Krowl has joined #ffmpeg-devel
<BBB>
thardin: why would you want to drop msvc support?
Krowl has quit [Read error: Connection reset by peer]
cone-762 has joined #ffmpeg-devel
<cone-762>
ffmpeg Zhao Zhili master:093f5da53474: avfilter/dnn: Don't show backends which are not supported by a filter
<cone-762>
ffmpeg Zhao Zhili master:4f051c746be9: avfilter/dnn: Use dnn_backend_info_list to search for dnn module
<cone-762>
ffmpeg Zhao Zhili master:d3db7bbc0392: avfilter/dnn_backend_tf: Remove one level of indentation
<cone-762>
ffmpeg Zhao Zhili master:ac52cee72eb5: avfilter/dnn_backend_openvino: Fix free context at random place
<cone-762>
ffmpeg Zhao Zhili master:6de951923b2d: avfilter/dnn: Remove a level of dereference
Krowl has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
<JEEB>
Lynne: btw I just noticed for the top channels, the description shortname string is TTL/TTR, while it's "top surround left/right" :D is there some link between those that I don't get or is that a leftover from some earlier version?
Krowl has joined #ffmpeg-devel
<thardin>
BBB: we may already have, considering c99toc89 exists
<BBB>
new versions of msvc support c99
<BBB>
c99toc89 is (was) only needed for super-old versions of msvc
<BBB>
FFmpeg can be built with MSVC 2012 or earlier using a C99-to-C89 conversion utility and wrapper, or with MSVC 2013 or later natively.
<thardin>
for the longest time MS categorically refused to improve their C compiler
<JEEB>
yea, 2013 update 2 or so is when it finally compiled FFmpeg I think?
<JEEB>
`/15
<JEEB>
whoops
<mkver>
BBB: FFmpeg relies on C11 nowadays.
<mkver>
Everything <= 19.27 is unsupported to build FFmpeg with.
<BBB>
didn't you say -Dstatic_assert=_Static_assert can be used as a workaround?
<mkver>
BBB: This was for an early version of the patch; the one that has been applied 0d43adcbef9a intentionally omitted this, because 19.27 is only a preview implementation of C11.
<mkver>
And anyway, MSVC 19.27 would need -D_Static_assert=static_assert
<BBB>
that seems reasonable. so it would be trivial to compile with msvc 19.27 if some poor soul wanted to do so
<mkver>
-Dstatic_assert=_Static_assert would work for michaelni's old fate boxes, but he chose to update to newer toolchains instead.
<jamrial>
until we started using something like _Generic
<BBB>
I believe that wiki page could use some updating perhaps. but my more general point was that msvc *is* supported (for some versions) and I don't think that's a particular reason to drop that
<mkver>
Actually, any C11 compiler should work out of the box. Trying to change that by blacklisting a particular compiler or by only supporting a whitelist of {GCC,Clang} should be instantaneously rejected.
<courmisch>
MSVC is not C11-conformant though is it
<courmisch>
AFAIR, e.g. aligned_alloc() isn't there
<BBB>
shocking, ain't it
<courmisch>
ironic actually
<mkver>
courmisch: True, but C11's aligned_alloc is broken anyway.
<courmisch>
MS pushed to make VLAs optional in C11 to make MSVC compliant the cheesy way
<courmisch>
and in the end they failed
<courmisch>
worksfinehere™
<courmisch>
well, I would argue that C11 and earlier heap is broken, because unlike C23 and Rust, you don't tell free() the alloc size, but that's another problem (that will never be fixed)
<Lynne>
c23 frees the alloc size?
<mkver>
courmisch: IMO not supporting _Alignas and _Atomic is way more significant.
<courmisch>
Lynne: c23 has free_sized()
<courmisch>
ofc it's useless because it still has free() for backward compat
<courmisch>
mkver: no disagreement there
rvalue has quit [Read error: Connection reset by peer]
rvalue has joined #ffmpeg-devel
Arsen has quit [Quit: Quit.]
Arsen has joined #ffmpeg-devel
Arsen is now known as Guest8883
Guest8883 has quit [Client Quit]
<cone-762>
ffmpeg Rémi Denis-Courmont master:add8c4621539: Revert "avformat/file: seek back to initial position for fd protocol"
aaabbb has quit [Remote host closed the connection]
aaabbb has joined #ffmpeg-devel
another is now known as another|
IndecisiveTurtle has joined #ffmpeg-devel
cone-762 has quit [Quit: transmission timeout]
Krowl has quit [Read error: Connection reset by peer]
<courmisch>
hmm, so why is framecrc computing not-a-CRC?
<courmisch>
Adler-32 is not a cyclic redundancy check
<courmisch>
false advertising, tsk tsk
<Lynne>
when I'm in power, passing something off as a CRC would be a crime
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg-devel
kurosu has quit [Quit: Connection closed for inactivity]
Livio has joined #ffmpeg-devel
Livio has quit [Ping timeout: 268 seconds]
AbleBacon has joined #ffmpeg-devel
_whitelogger_ has quit [Ping timeout: 260 seconds]
_whitelogger has joined #ffmpeg-devel
iive has joined #ffmpeg-devel
<Sean_McG>
QEMU ppc FATE node might be down for a bit, I need to integrate it in to libvirt so it survives whenever I need to login/out of my DE. Last time I tried, it wouldn't boot, even though it runs fine manually via qemu-system-ppc
ccawley2011 has quit [Read error: Connection reset by peer]
bencoh has quit [Ping timeout: 268 seconds]
<Sean_McG>
it'd be nicer if I could get my old G4 Mac mini up and running but I haven't had the kind of time it would probably take to do that.
bencoh has joined #ffmpeg-devel
merbanan has quit [Quit: Leaving]
<Lynne>
Sean_McG: if you want something else to get rid of, there's atmega assembly with zero proof it was ever used in libavutil
<Lynne>
I seriously doubt anyone has ever managed to run ffmpeg on an atmega, I found no evidence, and we use far too much memory for an 8-bit micro
<michaelni>
courmisch, historic reasons, i too am annoyed by calling adler-32 a crc ...
<Lynne>
glibc 2.39 still doesn't support free_sized
<Lynne>
hashtag sadface
lexano has quit [Ping timeout: 268 seconds]
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg-devel
<Sean_McG>
Lynne: can do.
<Sean_McG>
even the Wikipedia article mentions that Adler-32's primary interest was speed over accuracy