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
HarshK23 has quit [Quit: Connection closed for inactivity]
rvalue has quit [Ping timeout: 246 seconds]
rvalue has joined #ffmpeg-devel
<Traneptora> iirc the issue with declaration-after-statement was something like
<Traneptora> if (some_func() < 0) goto end; void *a = foo(); end: free(a);
<Traneptora> we make use of forward gotos for error handling, which work poorly with late declarations
dellas has joined #ffmpeg-devel
<Traneptora> if you declare at the beginning it makes it more obvious that you didn't initialize it to NULL or someth
kasper93 has joined #ffmpeg-devel
<Lynne> the compiler will complain
<Lynne> and you generally define what you need to stay for the lifetime of the function after declaration, before any other code
<Marth64> The only real annoyance I have with it is the scrolling in complex/packed functions. but on the other hand, it's sorta cleaner
user1453 has quit [Ping timeout: 264 seconds]
dellas has quit [Remote host closed the connection]
Marth64 has quit [Remote host closed the connection]
kurosu has joined #ffmpeg-devel
crmur__ is now known as orthoplex64
lexano has quit [Ping timeout: 252 seconds]
qeed has joined #ffmpeg-devel
navi has quit [Quit: WeeChat 4.1.2]
rvalue has quit [Ping timeout: 272 seconds]
omegatron has quit [Quit: Power is a curious thing. It can be contained, hidden, locked away, and yet it always breaks free.]
lemourin has quit [Read error: Connection reset by peer]
lemourin has joined #ffmpeg-devel
Wenbin_Chen_ has joined #ffmpeg-devel
HarshK23 has joined #ffmpeg-devel
Wenbin_Chen has quit [Read error: Connection reset by peer]
markh has joined #ffmpeg-devel
haihao has quit [Ping timeout: 252 seconds]
haihao has joined #ffmpeg-devel
epony has quit [Remote host closed the connection]
epony has joined #ffmpeg-devel
jamrial has quit []
Marth64 has joined #ffmpeg-devel
rvalue has joined #ffmpeg-devel
kurosu has quit [Quit: Connection closed for inactivity]
Marth64 has quit [Ping timeout: 272 seconds]
Martchus_ has quit [Ping timeout: 264 seconds]
Martchus has joined #ffmpeg-devel
wyatt8740 has joined #ffmpeg-devel
wyatt8750 has quit [Ping timeout: 256 seconds]
derpydoo has quit [Ping timeout: 240 seconds]
Marth64 has joined #ffmpeg-devel
epony has quit [Remote host closed the connection]
wyatt8750 has joined #ffmpeg-devel
wyatt8740 has quit [Ping timeout: 272 seconds]
Marth64 has quit [Remote host closed the connection]
Marth64 has joined #ffmpeg-devel
AbleBacon has quit [Read error: Connection reset by peer]
haihao has quit [Ping timeout: 260 seconds]
haihao has joined #ffmpeg-devel
<Marth64> sending one last version v8 of the dvd patch, hopefully for real this time, truly done with this horror...but it has fixes for chapter selection edge cases
<Marth64> it works pretty good though
<Marth64> i am done now with dvd i promise
<Marth64> this obsession must die
<Marth64> the alternatives to this demuxer either (1) emit corrupt frames, (2) are shady tools, (3) don't process nav packets or metadata and produce flawed output, (4) ignore some streams altogether or worse (5) all of the above
ngaullier has joined #ffmpeg-devel
<JEEB> :)
Krowl has joined #ffmpeg-devel
Marth64 has quit [Ping timeout: 260 seconds]
Marth64 has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
mkver has joined #ffmpeg-devel
<elenril> poor jonatas, people keep misspelling his name
qeed has quit [Remote host closed the connection]
thilo has joined #ffmpeg-devel
thilo has joined #ffmpeg-devel
thilo has quit [Changing host]
Krowl has joined #ffmpeg-devel
thilo has quit [Client Quit]
thilo has joined #ffmpeg-devel
thilo has quit [Changing host]
thilo has joined #ffmpeg-devel
epony has joined #ffmpeg-devel
markh has quit [Ping timeout: 260 seconds]
ccawley2011 has joined #ffmpeg-devel
cone-619 has joined #ffmpeg-devel
<cone-619> ffmpeg Cosmin Stejerean master:8264f3612c2c: .mailmap: update my mailmap entry
j45_ has joined #ffmpeg-devel
j45 has quit [Quit: ZNC 1.8.2 - https://znc.in]
kurosu has joined #ffmpeg-devel
kasper93_ has joined #ffmpeg-devel
kasper93 has quit [Ping timeout: 264 seconds]
<cone-619> ffmpeg Niklas Haas master:ce81237d6382: avfilter: pass link YUV colorspace to ff_draw_init2
<cone-619> ffmpeg Niklas Haas master:aa1a194a7e9e: avfilter/vsrc_testsrc: switch to YUV colorspace negotiation API
user1453 has joined #ffmpeg-devel
markh has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
thilo has quit [Quit: WeeChat 2.8]
thilo has joined #ffmpeg-devel
thilo has quit [Changing host]
thilo has joined #ffmpeg-devel
Krowl has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
thilo has quit [Ping timeout: 272 seconds]
thilo has joined #ffmpeg-devel
thilo has joined #ffmpeg-devel
thilo has quit [Changing host]
lexano has joined #ffmpeg-devel
Krowl has joined #ffmpeg-devel
<cone-619> ffmpeg Frank Plowman master:a42f884cd2fa: lavc/vvc: Fix slice_idx out-of-bounds memset
novaphoenix has quit [Quit: i quit]
novaphoenix has joined #ffmpeg-devel
Marth64 has quit [Ping timeout: 272 seconds]
kekePower has quit [Quit: The Lounge - https://thelounge.chat]
kekePower has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
<BBB> if we're going to deprecate sdl/gl output, maybe we should document this somewhere and provide helpful examples of how to go about getting the same type of functionality instead?
<BBB> in a way so that people googling it can figure out in a few seconds what to do instead
navi has joined #ffmpeg-devel
dellas has joined #ffmpeg-devel
<haasn> just teach it to chatgpt
<bencoh> is there some kind of replacement even?
<kierank> Gramner: how accurate is the perf hot cpu instruction measurement?
<kierank> it shows weird stuff as being the hottest
<kierank> simple stack moves: mov QWORD PTR [rsp+0x130],0x0
jamrial has joined #ffmpeg-devel
Krowl has joined #ffmpeg-devel
<cone-619> ffmpeg James Almer master:48f4a29bae46: avcodec/libdav1d: print libdav1d version in verbose level
<cone-619> ffmpeg James Almer master:e7a9dd03abcb: avcodec/libaomdec: print libaomdec version in verbose level
<haasn> Is there a way to `zfs allow` setting *all* properties on a dataset?
<Gramner> kierank: cpus are superscalar. such measurements do not have instruction-level accuracy
<Lynne> yeah, the kernel builds that data from seeing where the function was at the time the process was interrupted afaik
<haasn> oh, wrong channel
<Gramner> you can roughly determine in which functions, or in which loops, a lot of time is being spent. that's about it
<Lynne> though if you find an in-order x86 CPU, or use a low-power ARM core, they're a lot more reliable
<BBB> stack can also be slow if it was uninitialized
<BBB> so it's not like it's definitely not the stack making it slow
<Gramner> more specifically cpus do not really have the concept of "uninitialized", bits are bits. accessing memory not in cache can be slow though
<JEEB> bencoh: if you just need playback, ffplay now has libplacebo integration and one can utilize it for video rendering in general. the sdl output I tested during VDD and at least with random yuv420p you couldn't get even a window
<Gramner> also keep in mind that even a (regular) memory store needs to first load the cache line into l1 cache before it can write to it, so cache misses applies to store-only workloads too
<cone-619> ffmpeg Frank Plowman master:5076fa30ab6f: lavc/vvc: Validate alf_list indexes
dellas has quit [Remote host closed the connection]
dellas has joined #ffmpeg-devel
<JEEB> s/during VDD/during FOSDEM/
<JEEB> noticed I had done that mistake :D
jamrial has quit [Read error: Connection reset by peer]
jamrial_ has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
kasper93_ is now known as kasper93
dellas has quit [Remote host closed the connection]
dellas has joined #ffmpeg-devel
Krowl has joined #ffmpeg-devel
<haasn> nice, got swfmt negotiation working
<haasn> -i lena.pnm -vf hwupload,hwdownload -> comes out as rgb0
<haasn> after auto-inserted swscale rgb24->rgb0
<JEEB> fun
dellas26 has joined #ffmpeg-devel
dellas has quit [Read error: Connection reset by peer]
<elenril> haasn: nice
* elenril heading to CRL
<courmisch> eww
<courmisch> eww if you are heading to not-Brussels airfield and eww if you have gone so mad as think of yourself as a certificate to be revoked
* elenril stabs courmisch
<courmisch> that's a CoC violation
<elenril> is it?
<courmisch> also a violation of French penal code
<elenril> don't think it covers virtual stabs
Marth64 has joined #ffmpeg-devel
<courmisch> elenril: up to 1 year in government-paid 24h residence and 15000ā‚¬ in fine for threats of violence
<courmisch> according to said code
<Traneptora> but he didn't threaten to stab you, he stabbed you
<Traneptora> virtually
<courmisch> present tense can be used to express intent
<elenril> the CC would like to remind you that elenril will never threaten to stab you and, in fact, cannot speak
* Traneptora stabs courmisch
<elenril> in the even elenril does speak, the CC asks you to disregard its advice
<courmisch> online harrassment is also illegal under the penal code, AFAIK
<Traneptora> oh no I just threatened to stab courmisch everyone did you hear, please extradite me on an all-expenses-paid trip to europe
<jdek> elenril is a LLM
<courmisch> there is a bit of controversy now that France paid a private jet to extradite a dude from Bosnia
<Traneptora> I mean if you're going to extradite someone you can't charge them
<Traneptora> shrug
<Traneptora> trespassers will get a FREE ride in a police cart
<courmisch> jdek: that sounds reductive
<Traneptora> brb campaigning for rights for LLMs to be recognized as people
* elenril stabs jdek
<elenril> I don't hallucinate nearly as much
* another| patches courmisch up
<another|> threat of healing illegal too?
<Traneptora> >patches courmisch up
<Traneptora> nah, patches welcome
epony has quit [Remote host closed the connection]
<jdek> maybe Iā€™m a LLM
<courmisch> strictly speaking, any verbal human includes an LLM
<jdek> courmisch: does french law include anything for being stabbed by LLMs
<courmisch> IANAL
<Traneptora> I use a shielded neural net inside the head to delegate tasks to various parts of my existential operation
<Traneptora> various networks promote synergy allowing my 24/7 uptime
<courmisch> shielded? are you permanently under substance to prevent "learning" from external inputs?
<Traneptora> ah, no, it's shielded from physical access that may cause external trauma
<Traneptora> physical access to the machine would be problematic so one may only interface with my neural network via the usual communication protocols
<courmisch> ah, collagen
<Traneptora> a skull was what I was thinking
<courmisch> that's 95% collagen
<Traneptora> IANAbiologist
epony has joined #ffmpeg-devel
<courmisch> elenril: business people are supposed to fly to/from BRU. CRL is for tourists
cone-619 has quit [Quit: transmission timeout]
<Marth64> hello world
<courmisch> sorry, this is the zoo. For the world, please check with your local TA
<courmisch> ah no sorry, the lady with the syringe says it's called asylum
Krowl has quit [Ping timeout: 252 seconds]
<Marth64> hahaha
Krowl has joined #ffmpeg-devel
AbleBacon has joined #ffmpeg-devel
derpydoo has joined #ffmpeg-devel
<elenril> courmisch: do I look like a business person?
<elenril> i presented in a t-shirt with pedobear on it, not that anyone noticed
* elenril @CRL
<elenril> departure is not as disgusting as arrival
rvalue has quit [Ping timeout: 272 seconds]
Marth64 has quit [Remote host closed the connection]
<rodeo> wait what? you had a VideoLAN hoodie?
<rodeo> or was it another talk?
<elenril> i had a videolan hoodie, yes
<microchip_> pedo!
<rodeo> I certainly did not notice a pedobear
<courmisch> elenril: Silly con valley style business person, yes
<rodeo> I guess Elon might have
Krowl has quit [Read error: Connection reset by peer]
<courmisch> is that altright for Pikachu?
<courmisch> elenril: also, clothing doesn't make a monk
rvalue has joined #ffmpeg-devel
aljazmc has joined #ffmpeg-devel
<Lynne> neat, even on 32-bit systems, clang supports 128bit ints
user1453 has quit [Ping timeout: 268 seconds]
<courmisch> that can't be very effective
<Lynne> no, but it's probably better performance than doing emulated 64-bit fraction reduction using an iterative algorithm
<rodeo> I think your need to speak French to understand that expression courmisch
<rodeo> the literal translation is not known, as far as I know
<Traneptora> Lynne: what method is there for gcd beyond an interative algorithm
<Traneptora> iterative*
<Traneptora> iirc euclid's method is generally the fastest we got
<courmisch> rodeo: the English idiom is close enough IMO
<courmisch> ("clothes do not make the man")
<courmisch> and he was speaking at a French language institution
<Lynne> Traneptora: libavutil/mathematics.c, #if 1 ... the #else path
<courmisch> I've wanted to use __int128 for timestamp rate conversion in VLC
<Lynne> it just does a straighforward multiplication at arbitrary integer
<courmisch> but it's nonstandard
<courmisch> we can't have nice things in C
<courmisch> Rust has native i128
<Lynne> c23 has BitInt(128)
<Lynne> I'm using it, it works
<Lynne> needs gcc-14, or just #if defined(__GNUC__) && !defined(__clang__) typedef __int128 rational_int; #endif
<courmisch> the reason why C23 has no uint128_t is so dumb
<courmisch> "oh, but then we need to change the ABI of intmax_t"
<Lynne> it is, but nothing's stopping you from deffing it and using it wherever, even in printfs via %w28
<courmisch> that's not portable
<courmisch> even I think it's a bit early to require C23
<Lynne> neither are 128bit ints in general? support for them is optional after all
<Lynne> gcc doesn't support 128bit on 32-bit systems for now
<courmisch> > 32-bit systems
<Lynne> by the way, stdbit.h support finally got merged into glibc(!)
<Lynne> looking forward to using it in <checks calendar> 8 new moons from now
<courmisch> don't care; I already merged my replacement stdbit.h in VLC
<Lynne> nice, link?
<courmisch> but I'm too lazy for stdckdint.h
<Lynne> that one's been supported for years everywhere, iirc
<Lynne> or maybe not, I've added a fallback
<courmisch> unfortunately, not. The syntax is different from the GCC intrinsic
<Lynne> but the has_header directive makes fallbacks so painless
<courmisch> the parameter order is inverted
<courmisch> and the names are different (fortunately)
<Lynne> the only component with support for c23 but no release yet is meson :(
<Lynne> you still have to use c2x
<jamrial_> the fact c23 has to use stdc_ prefixes because they took too long to standarize this is amusing
<courmisch> more like they lacked imagination
<courmisch> they were able to come up with ckd_ for the checked arithmetic
epony has quit [Remote host closed the connection]
TheSashm_ has joined #ffmpeg-devel
TheSashm_ has quit [Remote host closed the connection]
TheSashmo has quit [Read error: Connection reset by peer]
TheSashmo has joined #ffmpeg-devel
ccawley2011 has quit [Read error: Connection reset by peer]
user1453 has joined #ffmpeg-devel
<Traneptora> shoulda been xkcd_
dellas26 has quit [Remote host closed the connection]
dellas has joined #ffmpeg-devel
jamrial_ has quit [Read error: Connection reset by peer]
jamrial has joined #ffmpeg-devel
ngaullier has quit [Ping timeout: 276 seconds]
aljazmc has quit [Remote host closed the connection]
aljazmc has joined #ffmpeg-devel
Krowl has joined #ffmpeg-devel
mkver has quit [Ping timeout: 255 seconds]
j45_ is now known as j45
j45 has quit [Changing host]
j45 has joined #ffmpeg-devel
Krowl has quit [Ping timeout: 276 seconds]
Krowl has joined #ffmpeg-devel
ccawley2011 has joined #ffmpeg-devel
Coinflipper has quit [Quit: ā€‹]
Coinflipper has joined #ffmpeg-devel
aljazmc has quit [Remote host closed the connection]
aljazmc has joined #ffmpeg-devel
Workl has joined #ffmpeg-devel
Workl has quit [Read error: Connection reset by peer]
Krowl has quit [Ping timeout: 260 seconds]
s55 has quit [Quit: ZNC 1.8.2 - https://znc.in]
s55 has joined #ffmpeg-devel
mkver has joined #ffmpeg-devel
dellas has quit [Remote host closed the connection]
s55 has quit [Quit: ZNC 1.8.2 - https://znc.in]
s55 has joined #ffmpeg-devel
aljazmc has quit [Remote host closed the connection]
aljazmc has joined #ffmpeg-devel
aljazmc has quit [Client Quit]
HarshK23 has quit [Quit: Connection closed for inactivity]
ccawley2011 has quit [Read error: Connection reset by peer]
<Daemon404> oh good a third thread
<BBB> yay
dellas has joined #ffmpeg-devel
dellas has quit [Remote host closed the connection]
<Lynne> I've been wondering, but what do av_malloc_attrib/av_alloc_size do?
<Lynne> as well as av_printf_format()
<Lynne> why do compilers need to know which arguments do what?
<kierank> Daemon404: I swear to god this thread is saying exactly what I said last week, it's like slowly the penny drops over the week
<BBB> an obvious workaround would be to not use ffmpeg to apply for funding
<BBB> "I, Michael", or "We, fflabs" and then "apply for STF funding and will use the money to pay people to do maintenance work on ffmpeg
<BBB> "I" or "we" "will do X, Y and Z as decribed in fflabs.org/doc or michaelni.org/doc
<BBB> this could also be videolabs, kierancorp, bbbcorp or anything else
<BBB> then you don't have this issue
<BBB> I'm fundamentally confused why he wants to do this through ffmpeg resources. why?
<kierank> perception of transparency
<BBB> I don't think the transparency is very real if any comment makes me a truthdenier or whatever
<BBB> I've made suuggestions - and you too - about issues and potential solutions, and I don't feel that this input is much appreciated
<BBB> but maybe I feel wrong...
<mkver> Lynne: av_alloc_size is supposed to give the compiler information about the size of allocations, so that it can warn you about stuff like char *ptr = av_malloc(1); ptr[5] = 0;. av_malloc_attrib is used to inform the compiler about aliasing (or rather: the lack thereof). It is also a hint to the compiler that the call likely succeeds.
<Lynne> ah, I guess that's the same for printf
<mkver> av_printf_format is the reason that compilers warn about mismatches between the format string and the supplied arguments.
<pross> is the STF funding stuff intended to fund new FFmpeg developers, or compenstate existing ones?
<Daemon404> i dont think that is defined
<BBB> pross: in its current form, compensate existing, I think michael wants to use some of it for work he does
<Daemon404> i dont think he stf itself has any such rules
<kierank> never ask a barber if you need a haircut
<Daemon404> right, shouldnt have read the third thread
* Daemon404 closes email
<kierank> (obviously michael should be compensated for his work as it's very valuable)
<kierank> but the way this whole process has happened has showed this community cannot organise itself let alone manage money
<Daemon404> a coworker said similar.
<Daemon404> yes, people are watching.
<kierank> I met the coreboot bios people for the first time
<kierank> and told them how nice their community is. they have a mentor to help first time patches like mine, they have weekly meetings instead of ignoring problems.
<Daemon404> their community may be made up of societally functioning human adults
<Daemon404> dont have those here
<kierank> Daemon404: we have people who are paranoid about the government yet at the same time want their money
<kierank> (people are obviously entitled to their own personal views, but the level of mental gymnastics sometimes)
<BBB> :)
dellas has joined #ffmpeg-devel
cone-731 has joined #ffmpeg-devel
<cone-731> ffmpeg Marth64 master:97fb84d74e90: doc/formats: clarify meaning of igndts as per definition in avformat.h