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
rvalue has quit [Ping timeout: 264 seconds]
rvalue has joined #ffmpeg-devel
jarthur has quit [Ping timeout: 240 seconds]
jarthur_ has joined #ffmpeg-devel
dellas has joined #ffmpeg-devel
dellas has quit [Remote host closed the connection]
derpydoo has quit [Ping timeout: 260 seconds]
thilo has quit [Ping timeout: 260 seconds]
thilo has joined #ffmpeg-devel
thilo has quit [Changing host]
thilo has joined #ffmpeg-devel
navi has quit [Quit: WeeChat 4.0.4]
cone-342 has quit [Quit: transmission timeout]
AbleBacon_ has joined #ffmpeg-devel
lexano has quit [Ping timeout: 260 seconds]
AbleBacon has quit [Ping timeout: 264 seconds]
Marth64 has quit [Read error: Connection reset by peer]
compn has quit [Read error: Connection reset by peer]
Marth64 has joined #ffmpeg-devel
compn has joined #ffmpeg-devel
georgereynolds8 has quit [Quit: Ping timeout (120 seconds)]
georgereynolds8 has joined #ffmpeg-devel
lemourin has quit [Read error: Connection reset by peer]
lemourin8 has joined #ffmpeg-devel
lemourin8 is now known as lemourin
reynaldo has joined #ffmpeg-devel
jamrial has quit []
kasper93 has joined #ffmpeg-devel
Kei_N has joined #ffmpeg-devel
Martchus has joined #ffmpeg-devel
Martchus_ has quit [Ping timeout: 246 seconds]
blb has quit [Quit: brb]
blb has joined #ffmpeg-devel
AbleBacon_ has quit [Read error: Connection reset by peer]
haihao_ has quit [Remote host closed the connection]
haihao_ has joined #ffmpeg-devel
averne has quit [Ping timeout: 264 seconds]
Marth64[m] has joined #ffmpeg-devel
Marth64 has quit [Ping timeout: 260 seconds]
averne has joined #ffmpeg-devel
Krowl has joined #ffmpeg-devel
mkver has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
Krowl has joined #ffmpeg-devel
ngaullier has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
Krowl has joined #ffmpeg-devel
jarthur_ has quit [Quit: jarthur_]
Marth64[m] has quit [Ping timeout: 246 seconds]
Teukka has quit [Read error: Connection reset by peer]
Teukka has joined #ffmpeg-devel
Teukka has quit [Changing host]
Teukka has joined #ffmpeg-devel
<wbs> frankplow: thanks, all my fate instances are back to green now
dellas has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
elastic_dog has quit [Ping timeout: 240 seconds]
elastic_dog has joined #ffmpeg-devel
cone-122 has joined #ffmpeg-devel
<cone-122> ffmpeg Anton Khirnov master:0291b6f824c5: lavfi/vsrc_ddagrab: add an option to avoid duplicating frames
quietvoid has quit [Ping timeout: 252 seconds]
Krowl has joined #ffmpeg-devel
quietvoid has joined #ffmpeg-devel
dellas has quit [Read error: Connection reset by peer]
dellas has joined #ffmpeg-devel
<cone-122> ffmpeg Anton Khirnov master:3b84140a1bb5: fftools/ffmpeg_dec: split Decoder into a private and public part
<cone-122> ffmpeg Anton Khirnov master:b43d4a069258: fftools/ffmpeg_dec: export subtitle_header in Decoder
<cone-122> ffmpeg Anton Khirnov master:d47bb91f8bc6: fftools/ffmpeg_filter: consolidate decoder/filter type checks
<cone-122> ffmpeg Anton Khirnov master:f2a2094c477d: fftools/ffmpeg: make decoding AVCodecContext private to the decoder
<cone-122> ffmpeg Anton Khirnov master:7ecd45587fbc: fftools/ffmpeg_dec: add an AVClass to Decoder
<cone-122> ffmpeg Anton Khirnov master:5b0e4f945ead: fftools/ffmpeg_dec: pass decoder options as an argument to dec_open()
<cone-122> ffmpeg Anton Khirnov master:9be3f80527e2: fftools/ffmpeg_dec: move decoding counters from InputStream to Decoder
<cone-122> ffmpeg Anton Khirnov master:27de2aacd15c: fftools/ffmpeg_dec: drop useless and racy code
<cone-122> ffmpeg Anton Khirnov master:217b8757db62: fftools/ffmpeg_dec: drop a useless log message
<cone-122> ffmpeg Anton Khirnov master:b2a0a2e6f531: fftools/ffmpeg: move decoder existence check to a more appropriate place
<cone-122> ffmpeg Anton Khirnov master:924a6f3cc778: fftools/ffmpeg_dec: override video SAR with AVCodecParameters value
<cone-122> ffmpeg Anton Khirnov master:474ca6c71e84: fftools/ffmpeg_dec: stop accesing InputStream.fix_sub_duration
<cone-122> ffmpeg Anton Khirnov master:a938f47916f1: fftools/ffmpeg: refactor disabling decoder threading for attached pictures
<cone-122> ffmpeg Anton Khirnov master:ebb8a58c8f24: fftools/ffmpeg_dec: replace InputFile.format_nots with a decoder flag
<cone-122> ffmpeg Anton Khirnov master:a3a9c4ae66e4: fftools/ffmpeg: move hwaccel_retrieve_data() from ffmpeg_hw to ffmpeg_dec
<cone-122> ffmpeg Anton Khirnov master:4bdffec8144c: fftools/ffmpeg_dec: pass hwaccel options to the decoder in a separate struct
<cone-122> ffmpeg Anton Khirnov master:e0a6cb07b21f: fftools/ffmpeg_dec: move flags to DecoderOpts
<cone-122> ffmpeg Anton Khirnov master:052c83e356e8: fftools/ffmpeg_dec: pass forced/estimated framerate though DecoderOpts
<cone-122> ffmpeg Anton Khirnov master:b27de18e4f66: fftools/ffmpeg_dec: move setting compute_edt to demuxer
<cone-122> ffmpeg Anton Khirnov master:970281766200: fftools/ffmpeg_dec: pass input timebase through DecoderOpts
<cone-122> ffmpeg Anton Khirnov master:9ba4bc87e6c0: fftools/ffmpeg_dec: pass top_field_first through DecoderOpts
<cone-122> ffmpeg Anton Khirnov master:097f9ddbe940: fftools/ffmpeg_dec: pass decoder name through DecoderOpts
<cone-122> ffmpeg Anton Khirnov master:d5f9ef766b4e: fftools/ffmpeg_dec: eliminate InputStream use in hw_device_setup_for_decode()
<cone-122> ffmpeg Anton Khirnov master:b3d1916ba030: fftools/ffmpeg_dec: pass AVCodec through DecoderOpts
<cone-122> ffmpeg Anton Khirnov master:1b2c539a0f4b: fftools/ffmpeg_dec: pass AVCodecParameters through DecoderOpts
<cone-122> ffmpeg Anton Khirnov master:a0ca99ccb7a4: fftools/ffmpeg_dec: remove unnecessary InputStream arguments
<cone-122> ffmpeg Anton Khirnov master:fe3be6f78fe3: fftools/ffmpeg_dec: stop passing InputStream to dec_open()
<cone-122> ffmpeg Anton Khirnov master:0d00e2e2f7fc: fftools/ffmpeg_dec: eliminate all remaining InputStream uses
<cone-122> ffmpeg Anton Khirnov master:01c71b78eb3c: fftools/ffmpeg: make InputStream.decoding_needed private to demuxer
<cone-122> ffmpeg Anton Khirnov master:f4b76a1d6a93: fftools/ffmpeg: make InputStream.decoder_opts private to demuxer
<cone-122> ffmpeg Anton Khirnov master:093be1fb060b: fftools/ffmpeg: cosmetics, vertically align Input{File,Stream}
quietvoid has quit [Ping timeout: 256 seconds]
quietvoid has joined #ffmpeg-devel
ccawley2011 has joined #ffmpeg-devel
quietvoid has quit [Ping timeout: 260 seconds]
quietvoid has joined #ffmpeg-devel
Kei_N has quit [Ping timeout: 264 seconds]
Kei_N has joined #ffmpeg-devel
tufei__ has quit [Remote host closed the connection]
tufei__ has joined #ffmpeg-devel
quietvoid has quit [Ping timeout: 240 seconds]
quietvoid has joined #ffmpeg-devel
<cone-122> ffmpeg Zhao Zhili master:bc944168db90: avcodec/mpeg4videodec: Remove write-only variable
dellas has quit [Ping timeout: 276 seconds]
jamrial has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
tufei_ has joined #ffmpeg-devel
tufei_ has quit [Remote host closed the connection]
tufei__ has quit [Remote host closed the connection]
Krowl has joined #ffmpeg-devel
dellas has joined #ffmpeg-devel
<cone-122> ffmpeg Frank Plowman master:36a986d9a193: lavc/vvc: Add check to num_multi_layer_olss
epony has quit [Remote host closed the connection]
navi has joined #ffmpeg-devel
epony has joined #ffmpeg-devel
epony has quit [Remote host closed the connection]
dellas has quit [Remote host closed the connection]
lexano has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
Krowl has joined #ffmpeg-devel
thardin has quit [Ping timeout: 252 seconds]
dellas has joined #ffmpeg-devel
jdek has quit [Ping timeout: 268 seconds]
staceee has quit [Ping timeout: 268 seconds]
Hobbyboy has quit [Ping timeout: 268 seconds]
Manouchehri has quit [Ping timeout: 268 seconds]
Hobbyboy has joined #ffmpeg-devel
c_14 has quit [Ping timeout: 268 seconds]
graphitemaster has quit [Ping timeout: 268 seconds]
termos has quit [Ping timeout: 268 seconds]
staceee has joined #ffmpeg-devel
termos has joined #ffmpeg-devel
Manouchehri has joined #ffmpeg-devel
graphitemaster has joined #ffmpeg-devel
Marth64[m] has joined #ffmpeg-devel
jdek has joined #ffmpeg-devel
epony has joined #ffmpeg-devel
c_14 has joined #ffmpeg-devel
ccawley2011 has quit [Ping timeout: 260 seconds]
ccawley2011 has joined #ffmpeg-devel
thardin has joined #ffmpeg-devel
dellas has quit [Remote host closed the connection]
haihao_ has quit [Ping timeout: 256 seconds]
haihao_ has joined #ffmpeg-devel
Krowl has quit [Quit: Krowl]
Krowl has joined #ffmpeg-devel
<another|> Just finished reading/skimming the STF thread. I feel empty inside.
<Marth64[m]> brutal
<Daemon404> this last year has been very nostalgic
<Daemon404> this is not a good thing
<elenril> another|: next time you can just hire a domme to whip you
* Daemon404 vanishes back into the ether
* elenril throws some peanuts after Daemon404
<elenril> you forgot this
<Daemon404> free food!
<jamrial> we somehow found a way to bikeshed about being given $200k
<elenril> fun fact: local tits like peanuts more than sunflower seeds
<JEEB> it's all semechki
<elenril> JEEB: peanuts are not, at least where i come from
<JEEB> well yea, not in the human snack sense :D
<wbs> "semechki"?
derpydoo has joined #ffmpeg-devel
<JEEB> paahdetut siemenet in finnish
<wbs> right
<elenril> jamrial: ever heard of the resource curse?
<wbs> elenril: thank you :-)
<Traneptora> reading stf thread and ng showed up
<Traneptora> kinda glad I decided to stay out of it
<Marth64[m]> my biggest take away was the word "topical". I like how that was phrased in the context from someone trying to de-escalate
<Marth64[m]> That was a good way to say it
<mkver> Following the STF thread seems to be a full time job.
<Traneptora> speaking of the thread tho I do thing some CoC are happening
<Traneptora> in one thread ng responded to a technical objection with a hithiker's guide to the galaxy quote
<Traneptora> and then when kierank couldn't figure out what he meant, he told him to go back to elementary school and learn to read
<Daemon404> i wrote this yesterday elsewhere but i think it accurate describes this and many other similar threads with e.g. ng:
<Daemon404> > the problem with these threads is the two outcomes are:
<Daemon404> > * stop replying, they win
<Daemon404> > * keep replying and getting gaslit until you burn out or get in trouble, they win
<Daemon404> current status: option 2.
<elenril> mkver: perhaps following the thread could be one of the proposed projects
<Daemon404> coc enforcement is not currently set up to handle this (pattens of behavior, gaslighting, etc.), just Man Said Mean Wordd specific email
<Daemon404> i see CC as a bit of a lame duck
* Daemon404 runs
<elenril> Daemon404: i ignored his complaints and he's endlessly rustled about it now
<elenril> i count that as a win
<Marth64[m]> he was polite to me in the only 2 interactions we had lol
<Daemon404> silence is complicity
<elenril> Marth64[m]: try disagreeing with him
<Marth64[m]> elenril: I did (albeit in the form of a long ramble), he never responded
<elenril> mkver: btw a lot of your cleanup work could fall under what STF wants to fund, if you care about income
<Traneptora> it's gotten especially bad recently. it's always bad, but at least he's always made an attempt
<Traneptora> but quoting fantasy novels and then telling people to learn 2 read is a new low
<Daemon404> how has ng or paul not been banned?
<Daemon404> paul literally only trolls.
<Daemon404> literally.
<Traneptora> paul contributes audio code. or at least he used to
<Daemon404> thats another problem
<Daemon404> he doest now, however
<Daemon404> because someone writes good code doesnt mean they can be a jerk *constantly*
<Daemon404> with no consequences
<thardin> I'd be tempted to say code is completely irrelevant
<Traneptora> mkver: re changing size_t -> uint32_t
<Traneptora> sizeof(JxlPixelFormat) is part of the libjxl public ABI
<Traneptora> that's why we're allowed to place the struct on the stack
<Traneptora> if they were to change it from size_t -> uint32_t, there'd be ABI breakages that go beyond just possible accidental overflow
<thardin> tests and docs and dealing with formalia and so on may well be much more beneficial to the project than yet another filter
<mkver> And? It just means that they may only change it during a major bump. But they are allowed to change it and if they did, your proposed code might not work.
<Traneptora> if they break their ABI the code will already break
<Traneptora> because we place JxlPixelformat on the stack
<Traneptora> what I'm wondering is why we have to plan around an unlikely hypothetical ABI break in an external library on a major version bump
<Daemon404> libjxl has no exactly been the most stable
<Traneptora> sure, but I still don't see why we need to plan around a specific hypothetical unlikely ABI break
<kierank> 15:12:05 <Daemon404> because someone writes good code doesnt mean they can be a jerk *constantly*
<kierank> 15:12:12 <Daemon404> with no consequences
<kierank> i mean that's basically the project mission statement
<Daemon404> for over a decade yes
<Marth64[m]> (wondering as an outsider) are the in-person events as chaotic as the emails lol
<Marth64[m]> hopefully better
<Daemon404> Marth64[m], no
<Daemon404> the chaotic people refuse to meet irl
<Daemon404> because of Reasons
<Marth64[m]> understood
<Daemon404> its als why youll read a lot of Well I Wasnt There And Dont Agree later
<Traneptora> also worth mentioning that they *could*
<Daemon404> yes
<Traneptora> paul lives in europe, he could go to germany for a meeting
<Daemon404> the reasons are not monetary
cone-122 has quit [Quit: transmission timeout]
<Traneptora> ye, especially cause videolan would pay for his train ticket and hotel
<kierank> 15:25:49 <Daemon404> Marth64[m], no
<kierank> wrong
<kierank> you've missed the shouting ones
<Daemon404> i meant now
<kierank> don't need to mention names
<Daemon404> not 5 years ago
<kierank> same
<kierank> don't need to mention names
<Daemon404> noo shouting last year
<kierank> fosdem there was
<Traneptora> it's a bit more complicated for someone like me cause I live in North America so going to europe is actually a real hassle even if it were to be paid for
<Traneptora> but if you live a train ticket away it's a bit different
<elenril> Traneptora: but it's a cultural experience
<elenril> you get to eat the famous brussels cardboard
<Daemon404> everyone has a right to not want to meet irl. i just hate it being used as a spanner to throw into arguments.
<Lynne> kierank: not telling names, but I do not like what you're implying nor what you think you are remembering
<Traneptora> oh I'd love to get a paid trip to europe but I also can't just take a week off work
<mkver> Traneptora: "if they break their ABI the code will already break": When they break their ABI and you recompile lavc against the new version, then the code is supposed to work. Your code would not work under these circumstances.
<Traneptora> ah you mean if it were an ABI but not an API break
<mkver> Yes.
<Traneptora> in either case, I dropped an updated patch in reply to james's email
<Traneptora> changed - jxl_fmt.align * foo to + frame->linesize[0] * foo
<Daemon404> IME putting struct sizes as ABI guarantees always backfires
<Traneptora> for utility metadata structs it seems fine to me
<mkver> Daemon404: This has nothing to do with this.
<Traneptora> the alternative is every metadata struct has to be heap alloced which is a pain
<Daemon404> mkver, no? im referring to libjxl saying it is part of abi guarantees.
<Traneptora> that's just always the case when you put structs on the stack
Krowl has quit [Read error: Connection reset by peer]
<Daemon404> and many libs dont consider that valid ;)
<mkver> Daemon404: ABI guarantees only apply for the same major version. My hypothetical scenario involves a major bump.
<Traneptora> ye, mkver's point is that okay so they change size_t -> uint32_t and break their ABI, and my previous code compiles against the new API so it would break silently
<Daemon404> no i get that.
<Daemon404> my point was providing that guarantee *within the same major version* leads to extra pain on major version bumps
<Daemon404> as people will have put it on the stack
<Daemon404> easy to get bitten
<Traneptora> the alternative is to do what ffmpeg does which is have struct foo *random_struct_new(); for every utility struct
<Daemon404> so better to never guarantee it at all
<Traneptora> which makes sense for major things like AVPacket or AVFrame
<Daemon404> i think it makes sense for more or less eveything
<Daemon404> having been bittn over and over.
<Marth64[m]> (a bit removed, but thought to throw it out there in case it helps) can you #if on a major version something like this? https://github.com/videolan/vlc/blob/master/modules/access/dvdread.c#L214
<Marth64[m]> probably hack or not applicable
<Daemon404> anyway, this is just be waxing about API/ABI design
<Daemon404> s/be/me/.
<Daemon404> do carry on
<Traneptora> Marth64[m]: yea those are for API breaks as well
<Traneptora> if a struct is stack-allocated, then the size of the struct ends up mattering
<Daemon404> not just size, but layout too
<Daemon404> and alignment
<Traneptora> ye, all of the above
<Traneptora> well, layout is still determined if you access struct members directly, right?
<Traneptora> even if it's a pointer to a struct, doing struct foo *bar = struct_few_new(); and then bar->baz still is subject to layout problems
<Traneptora> the only way to solve that is to make the structure entirely opaque
<Daemon404> yes, thats why you cant just reorder
<Daemon404> asm isnt environmentally friendly because you can make your intel processor downclock by using the wrong instruction set
HarshK23 has joined #ffmpeg-devel
<mkver> Daemon404: In the scenario outlined here the pain does not result from people putting something on the stack. The issue would be the same if libjxl provided an allocator.
dellas has joined #ffmpeg-devel
Krowl has joined #ffmpeg-devel
kurosu has joined #ffmpeg-devel
<BBB> Traneptora: fwiw, I live in north america and videolan pays for my travel to vdd. afaik they pay for all active contributors to ffmpeg, they explicitly mention that in the announcement emails
<Traneptora> I know, they would, but it's also a pain
<Traneptora> I can't just take a week off of work to go to europe
<kasper93> vdd in las vegas when?
<BBB> I know. and kids, etc.
<kasper93> :^)
<BBB> why vegas. if we go, might just as well go to hawaii
<elenril> vdd will never be in US
<BBB> I know :(
<elenril> (which is a good thing)
<BBB> you're missing out on free and brave
<BBB> imagine watching a football game and the anthem plays. I mean the other football, not the one played with the foot
<elenril> my friend from physics seems permanently traumatised by his trip to baltimore
<BBB> well, baltimore
<BBB> could've gone to new york, vegas, hawaii, boston, dc, miami, houston. but he went to baltimore
<elenril> he was invited specifically there
<kasper93> I know baltimore only from the wire
<BBB> if I invite you to the homeless encampment in san fransisco, would you go?
<elenril> I'd rather not go to US at all
<elenril> I'd have to buy a burner laptop, phone, leave my hw key at home, etc.
<BBB> that's how corpfolk here talk about china
<elenril> I'd rather not go to china either
<BBB> you're starting to sound like paul
<elenril> clearly you're a poor fflabs employee
<BBB> if you change your mind, I'll happily take you for a beer or another beverage of your choice if/when you visit NYC
<Daemon404> china is a big no-go.
<Daemon404> also baltimor == the wire
<elenril> i might compromise on montreal
<Daemon404> should have it at pauls house
<elenril> good idea
<Marth64[m]> montreal sounds nice
<kierank> •elenril> my friend from physics seems permanently traumatised by his trip to baltimore
<kierank> my trip was fine
<kierank> 16:40:50 <•BBB> why vegas. if we go, might just as well go to hawaii
<kierank> correct
<elenril> might as well go to NZ then
<Marth64[m]> whatever you do, do not leave Canada with citrus (oranges/limes/etc)
<Marth64[m]> big agriculture hassle at the border
<elenril> will they apologise a lot before beheading you?
<Marth64[m]> LOL no but polite the whole time
<jamrial> i wouldn't enter or leave any airport with non sealed package food
<kasper93> yeah, my friend had a sandwich with cheese iirc, dogs pretty quickly found his backpack
APic has quit [Ping timeout: 252 seconds]
<courmisch> Montreal was already voted for VDD'2023 and didn't happen, so don't hold your breath
<courmisch> if we want to go to Americas, we should do VDD in SXM
<Marth64[m]> SXM?
<courmisch> European Carribeans
<Marth64[m]> ahh
<Marth64[m]> flights are $$$$
<courmisch> flights from Schiphol don't see too bad
durandal_1707 has joined #ffmpeg-devel
<courmisch> there's also PTP and FDF
APic has joined #ffmpeg-devel
<Daemon404> basically for vdd to be somewhere, there needs to be someone local-ish to help with venues etc
<Daemon404> ideally provide a venue ;)
<courmisch> outside Europe, it's going to be SEO or (again) TYO, I think
<Daemon404> i find seoul unlikely due to noone local
<durandal_1707> i'm not going to FOSDEM or any LibAV meeting because I'm not respected in FFlabs circles and LibAV dev community
<courmisch> on the bright side, if you don't go to FOSDEM, you can celebrate Candlemas
* courmisch needs to remember to buy milk and flour
<JEEB> I for one just see that people can have internet personas and IRL ones, so unless someone goes over and above to also be an asshat IRL, then you should be fine
<durandal_1707> "asshat" whetever this means, is very derogattory
<courmisch> and on Saturday, I can attend FOSDCJM, the Free and Open-Source Courmisches and JEEBs Meeting
<courmisch> a very exclusive invitation-only elite event if I say
<BBB> durandal_1707: I think from the responses to your (online) talk at vdd, you know you are respected here
<durandal_1707> courmisch: agenda on such elite meeting?
<BBB> durandal_1707: but if online presence is what you prefer, then I look forward to your next online talk for vdd'24
<durandal_1707> BBB: elenril said no
<BBB> maybe we should overrule elenril
<durandal_1707> just merge SDR already
<courmisch> over my undead body
<BBB> sdr as in not-hdr?
<durandal_1707> Software Defined Ratio
<BBB> right, because ratios are really important in multimedia processing
<BBB> ...?
<durandal_1707> only if they fit into AVRational
<courmisch> I vote for VDD in LYR
<durandal_1707> LYR?
<courmisch> We can arrange to "loose" a few people to the hungry polar bears
<Marth64[m]> I would attend a virtual VDD with polar bears
<courmisch> lose*
<durandal_1707> i will make lobotomy on polar bears, they will behave like pets
<durandal_1707> with neuralink
<courmisch> most genetically promising devs will be preserved in the seed vaults for generations
<durandal_1707> how would you evaluate most genetically promising devs?
<courmisch> the scientifically-proven wet finger method
<durandal_1707> which finger?
<Marth64[m]> (1) overall balanced technical skill (2) soft skill (3) can you apply common sense to problem solving,
<courmisch> right index, of course. No unorthodox variations
<durandal_1707> what if person do not have such finger? or have multiple hands?
<courmisch> I think you misunderstand the method
<courmisch> the person doing the evaluation uses their right index, not the victi^Wperson being evaluated
<durandal_1707> whatever
<courmisch> if you insist, we can fallback to basic Stetson-Harrison method
<durandal_1707> if you organize on FOSDEM/VDD poker sessions in flesh, i'm in.
<Marth64[m]> nice
<durandal_1707> 10k spent on SDR merge
mkver has quit [Quit: Leaving]
mkver has joined #ffmpeg-devel
<durandal_1707> BBB: i'm only respected (sort of) by users of my contributions in this project
<BBB> I respect your work
<Marth64[m]> your SCC muxer has saved me
<Marth64[m]> evil closed captions with evil roll up characters that needed to be manually cleaned
<JEEB> durandal_1707: why did you think I somehow meant you under that? there is no proof for me that you would behave like that IRL.
<JEEB> I was specifically making the point that I will do contact with you like with any other person IRL
<durandal_1707> that is like writting troll instead of that
rvalue has quit [Ping timeout: 256 seconds]
<JEEB> so far nobody has been an asshat IRL in the circles I've been :P people I may have issues with on the ML or otherwise, but I can discuss things with them just fine.
<JEEB> thus there's no reason for you to interpret anything I said like I meant anything negative regarding you.
<durandal_1707> ok, my language interpreter is buggy, but my experience of various actions or non-actions is fact.
<durandal_1707> BBB: fine, but I had almost same voting score as NG
<BBB> voting score?
<durandal_1707> that was few years ago
averne_ has joined #ffmpeg-devel
<BBB> I don't know what you mean by voting score. but if poker is what it takes, I'll learn poker and play at vdd'24
averne has quit [Ping timeout: 252 seconds]
averne_ is now known as averne
* JEEB just got into mahjong during the past year
<JEEB> the pieces feel good in hand :3
<klaxa> oh no
<klaxa> another soul lost
rvalue has joined #ffmpeg-devel
<klaxa> enjoy your accelerated aging :(
<JEEB> I blame other finnish programmers who took me to the dark side alleys of Oosaka
<klaxa> damn finns
<durandal_1707> BBB: elections for previous TC
<durandal_1707> Marth64[m]: some entities asked me why I ever wrote SCC muxer (similar happeded with bunch of other contributions)
<JEEB> because you cared enough about it? :D
<JEEB> that's the most common reason for OSS, after all
<JEEB> either you or someone close enough had a need -> stuff happens
<Marth64[m]> durandal_1707: for poor souls like me that can easily get paranoid and hyper over bad quality subtitles and can't afford the $12000+ decoder that is a relic from the 90s
<Marth64[m]> so for that, :tip of the hat:
<Marth64[m]> at least with your work I can do the transformations I need then render it to assa
<Marth64[m]> which ends up looking really good
AbleBacon has joined #ffmpeg-devel
<durandal_1707> JEEB: it was not paid like SDR is paid, or multithreading refactoring in ffmpeg.c is paid, even if I get paid it is like for monkey coder in prison.
<Daemon404> sdr was not paid
<durandal_1707> not sure, looks very complex
<kierank> I paid for sdr
<kierank> we all did
<BBB> courmisch: can you answer a dav1d/riscv MR comment? https://code.videolan.org/videolan/dav1d/-/merge_requests/1463#note_424495
<durandal_1707> where AV1 cameras are used?
kurosu has quit [Quit: Connection closed for inactivity]
paulk-bis has quit [Quit: WeeChat 3.0]
paulk has joined #ffmpeg-devel
<durandal_1707> I am sorry, but I could not find any information on cameras that natively support the AV1 codec. AV1 is a relatively new video codec that is designed to offer internet streaming efficiency upgrades without affecting quality 1. It is the successor to the HEVC (H.265) format that is currently used for 4K HDR video on platforms such as Prime Video, Apple TV+, Disney Plus and Netflix 1.
Krowl has quit [Read error: Connection reset by peer]
<jamrial> lol
<durandal_1707> Netflix has been using the AV1 codec but only for select Android devices, TVs and the PlayStation 4 Pro 2. Other platforms that use AV1 include WebEx 3.
<jamrial> any new enough cellphone supports av1 decoding, and probably encoding, but are unlikely to take photos in avif anyway
* durandal_1707 goes back to mjpeg cameras
<durandal_1707> nitter is dead, cant browse X any more
<BBB> it's ok, you don't need your twitterfriends, you've got us instead!
<durandal_1707> define 'us'
<jamrial> us
<durandal_1707> ok, don't mention my name again here
<durandal_1707> instead just use : 'he who should not be named'
<jamrial> you don't have a twitter account to browse it?
<durandal_1707> no twitter/X, no dumbbook, no facescam, no instagram, no tiktok, no whatsapp, no telegram, no whatever
<durandal_1707> also deleted linkedin
<durandal_1707> all 'social' removed
<durandal_1707> need also to remove irssi
<jamrial> but why? for the purpose of browsing around, you can create one with some random username
<durandal_1707> and chromium/gmail
<durandal_1707> 2FA
<Marth64[m]> its pure freedom
<Marth64[m]> i am in the same boat except linkedin
<Marth64[m]> freedom from garbage ad fueled content or highlight reels and drama
<Traneptora> how does tiff interact with Exif?
averne has quit [Ping timeout: 276 seconds]
<Traneptora> i.e. tiff just has tags. how does tiffdec know if they're tiff-specific or if they're exif tags
<durandal_1707> iirc tiff can have exif injected somehow
<durandal_1707> tiff is worse than mxf
<JEEB> yea, tiff has been extended so many times :D
<Marth64[m]> check GIMP maybe for ideas how its used in practice?
ngaullier has quit [Ping timeout: 256 seconds]
averne has joined #ffmpeg-devel
cone-099 has joined #ffmpeg-devel
<cone-099> ffmpeg Andreas Rheinhardt master:7252e4f8eeb5: avcodec/aac_defines: Remove unused AAC_RENAME_32
<mindfreeze> jamrial: latest pixel do have hw for encoding as per google:), so at somepoint they could do hw encodes
<jamrial> nice
sudden has quit [Ping timeout: 260 seconds]
<kierank> durandal_1707: irc is social media
<elenril> shocking truth!
sudden has joined #ffmpeg-devel
compn has quit [Read error: Connection reset by peer]
compn has joined #ffmpeg-devel
<Marth64[m]> IRC in its prime was the best
<Marth64[m]> i remember when colors were new ;-;
<Marth64[m]> probably clients dont even support IRC colors and rich text anymore
<cone-099> ffmpeg James Almer master:9949c1dd7834: avformat/avformat: fix group index range check in match_stream_specifier()
derpydoo has quit [Quit: derpydoo]
___nick___ has joined #ffmpeg-devel
dellas has quit [Remote host closed the connection]
___nick___ has quit [Ping timeout: 256 seconds]
Marth64[m] has quit [Ping timeout: 255 seconds]
Marth64[m] has joined #ffmpeg-devel
<Traneptora> before I send anything to the ML, if anyone wants to take a look at my Exif overhaul: https://github.com/Traneptora/FFmpeg/tree/exif-overhaul
<Lynne> we don't allow pointers in side data, the data must be flat
<Lynne> nihing a dictionary using flat data will be painful :(
<elenril> json!
* elenril runs
<Lynne> I'm an expert on this, after researching options for binary metadata in avtransport
<elenril> ihappentobeanexpertonthissubject.jpg
<Lynne> in the end I went with CBOR, it's a rushed ietf standard that's a bit too generic
MrZeus has joined #ffmpeg-devel
<Lynne> but feel free to move ebml to lavu and just represent the data using ebml :)
<elenril> not X.690? lame
<durandal_1707> libavjson
<elenril> [trolling intensifies]
<durandal_1707> why libavio ?
<elenril> so that AVIO stuff can be used outside of lavf
<durandal_1707> merge all libs into libffmpeg
<elenril> sigh
<elenril> get new material
<durandal_1707> well, how would you support strange cases, like actimagine or dv or dolby whatever
<durandal_1707> where demuxer is also decoder or whatever
<elenril> everything is a bitstream filter
<Lynne> I'm not a fan of the avio design in general, is it just going too be a direct library move for just the lavf/avio.h functions?
<elenril> for the most part that was the idea, though the API could use some cleanup
<Traneptora> Lynne: if we don't allow a pointer in sidedata, how am I supposed to attach an AVDictionary as sidedata
<Traneptora> or equivalent
<durandal_1707> ffmpeg is binary cancer
<durandal_1707> its got also on Moon?
Marth64 has joined #ffmpeg-devel
<Traneptora> one option that's I'm considering is having the unparsed Exif tags be the side data
<Traneptora> like just the whole buffer
<Traneptora> but then you need to parse it down the line
<durandal_1707> did ffmpeg got on Moon? or I'm hallucinating, and j-b sent only VLC?
<elenril> libavexif
<elenril> durandal_1707: did you eat some weird-colored mushrooms again?
<durandal_1707> i remmember some VLC and Moon post on linkedin?
<Marth64> ffmpeg no longer on the moon
<elenril> rad
<durandal_1707> what is rad?
<Marth64> IRC colors 8)
<pross> hi durandal_1707
<jamrial> Lynne: avio.h and all protocols, afaik
<durandal_1707> hi pross, have a nice day and trip!
<Daemon404> ffmpeg didnt go to the moon
<Daemon404> it went to mars
<durandal_1707> :(
<durandal_1707> mars !== Mars
<durandal_1707> moon !== Moon
<mkver> jamrial: leb |= bits << (i * 7); can be UB if i == 4.
<Lynne> Traneptora: was exif attached as stream sidedata, or frame sidedata before?
<Traneptora> it was attached as AVFrame.metadata
<Traneptora> not as sidedata for either
<durandal_1707> so it crashed?
<jamrial> mkver: it's all unsigned, so no. but i'll fix it anyway
<mkver> You're right. No need to fix anything then.
<durandal_1707> switch to Rust already
<Marth64[m]> isn't the horror story about rust that a lot of the community packages are garage projects a la npm
<Marth64[m]> riddled with vulnerabilities and endless dependency chains
<Marth64[m]> or am i mixing this up
<durandal_1707> yea, i wonder is it social problem or what?
<jamrial> definitely lots of dependencies. that you need to compile for every project
<durandal_1707> that is by design
<Marth64[m]> its bad programmers remaking solutions from 1980s over and over again and putting on github with cute graphics only to be abandoned after 2 years
<durandal_1707> i wondered about abandonware
<Marth64[m]> ok that was harsh, sorry javascript has scarred me for life
<durandal_1707> extreme/radical fragmentations hurts back
<elenril> Marth64[m]: you're not exactly wrong though
<elenril> I've seen so many "$X, but in Rust" projects
<durandal_1707> well, i have no problem with NIH
<durandal_1707> but i Have problem with abandonware
<Marth64[m]> elenril: yes exactly haha that is alwyas the title too
<elenril> that get to the cute toy stage, and stay there
haihao_ has quit [Ping timeout: 268 seconds]
haihao_ has joined #ffmpeg-devel
<Marth64[m]> java is not much better
<Marth64[m]> but at least ancient councils of java elders review feature extension APIs and they become stable for long time
<durandal_1707> future of programming is AI
<elenril> but who uses java
<durandal_1707> AI will code everything once it learn how to code from human examples.
<Marth64[m]> i speak java, too much of it ;-;
<Marth64[m]> ai will destroy us all
<durandal_1707> do current AI copilots know how to debug code?
<Marth64[m]> ehhhhh its like glorified documentation search more than debugging
<Marth64[m]> like the typical inspection tools in the common IDEs will tell you what's wrong then you can chat with a gimmicky text prompt and it will gpt search things for you
<Marth64[m]> but it's great at whipping up prototype code for sure
<Marth64[m]> glorified StackOverflow reader but thats not necessarily a bad thing
<Marth64[m]> i should ask it to write a demuxer
dellas has joined #ffmpeg-devel
Marth64 has quit [Killed (NickServ (GHOST command used by Marth64[m]))]
Marth64[m] is now known as Marth64
<another|> j-b: oh god. more money…
<Traneptora> Marth64: remember this https://meta.stackoverflow.com/questions/425162/
<Marth64> Traneptora: just seeing it now for the first time, honestly I would be furious :X
<Marth64> dear AI please let me compose my own thoughts
<Marth64> even their intro post to it looks like chatgpt
<Marth64> speaking of AI what happened to Duke Abandonman or whatever the spammer
<Marth64> is that bot still spamming the ml I blocked it lol
<Marth64> Duke Abbaddon!
<jamrial> he seems to have stopped. i don't see his email on my spam folder anymore
AbleBacon has quit [Ping timeout: 252 seconds]
<Marth64> glad to hear
<elenril> 2024-01-30T23:13:00.332222+01:00 mail.info mail0.khirnov.net postfix/qmgr[20250]: 0BE972404E5: from=<duke.abbaddon@gmail.com>, size=65989, nrcpt=1 (queue active)
<jamrial> guess not
<elenril> he has a dedicated blacklist entry in my SA config
HarshK23 has quit [Quit: Connection closed for inactivity]
<durandal_1707> I use Duke useful mails for new filter ideas
<Marth64> according to google image search duke is a fantasy mecha-demon creature which honestly totally fits
<Daemon404> is he emailing the list or direct?
<Daemon404> i think he emailed me direct before
<Traneptora> I have no emails from them so idk
<elenril> Daemon404: this is directly to me
<Daemon404> from what i recall they were uh
<Marth64> i deleted them all and added similar block rules so cant remember tbh
<Daemon404> how do you say
<Daemon404> severely mentally ill.
<elenril> could be a bot
<durandal_1707> AI is ill
<Daemon404> elenril, iirc i looked into it and it was just sad
<durandal_1707> needs skilled devs to improve it
<Marth64> i regret any joking if that was the case, always thought he was a bot
<elenril> Daemon404: I can't imagine such a strong case of schizophrenia could operate two email accounts
<durandal_1707> AI do not have brain
<Daemon404> youd be surprise
<Daemon404> d
<elenril> I guess
<jamrial> weren't duke's emails markov chain generated?
<elenril> maybe it's merely pretending
<Daemon404> this prompted me to check my spam folder
<durandal_1707> dropbox is banned here
<durandal_1707> folder is term used only for M$ Windows operating systems
<Traneptora> in the context of emails being sorted it means something other than directories
<Marth64> dir vs ls fight
<elenril> Daemon404: so he knows you
<aaabbb> Marth64: echo * ;)
* Marth64 crash
<Traneptora> Marth64: /usr/bin/dir is owned by coreutils 9.4-3
<durandal_1707> Duke is 'they'
<Daemon404> duke is nukem
<Traneptora> dukem
<durandal_1707> is Plan 9 still developed?
<elenril> planning to switch?
<Marth64> IIRC no but some concepts live on
<Marth64> Traneptora: TIL it's even on my machine now wow
<Traneptora> plan9 people like statically linking everything
<durandal_1707> elenril: yes, this browser == OS is killing me softly
<Daemon404> pauls os is libreoffice
<elenril> durandal_1707: try elinks
<durandal_1707> elenril: already user, but still no JS support, so only use it for limited number of web pages
<elenril> try qutebrowser then
<durandal_1707> can not run it inside tmux
<elenril> sure you can, just need TermX11
<durandal_1707> i wait for DRM powered browser so can use only X11/Xorg for games
<Marth64> horrifying prospect
<elenril> I'm sure you could grab xdummy frames and encode them with caca
<Marth64> thank you whoever shelved this project
<elenril> they'll just quietly revive it in a year
<durandal_1707> elenril: you mean i run dummy X server and proxy it to caca ? but that would be marginally useful...
<aaabbb> fork x11 into librex11
<llyyr> contribute a drm backend to firefox
ccawley2011 has quit [Read error: Connection reset by peer]
<Marth64> i would quit the internet at that point, firefox is all i can tolerate
<durandal_1707> i imagine that would be very hard
<durandal_1707> Marth64: are you against drm backend in firefox?
<Marth64> durandal_1707: yes I don't want it, it's ok. i am also crazy person who hates the modern internet. i'll stick to a firewalled isolated device if I need to watch actual DRM content
<Marth64> i remember when it came out as firebird, it was the coolest idea and project. to change it now and mix DRM in there, would truly be the death of the internet for me
<durandal_1707> DRM != DRM content
<Marth64> ohhhhh, like the renderer
<Marth64> ?
<Marth64> kernel integration for graphics?
<durandal_1707> drm (Direct Rendering Manager) Video output driver using Kernel Mode Setting / Direct Rendering Manager
<Marth64> LOL, I read your message wrong
<Marth64> yes, I am all for that please :D
<Marth64> and with that thought come horrors of trying to install nvidia drivers in 2015 with dkms
<Marth64> truly awful times
<Marth64> pretty good now
<durandal_1707> that is illusion
<durandal_1707> it always becomes more bad than before
<Marth64> the stress of those reboots are like roller coaster ride
<Marth64> will it come back...will it default to archiac VGA port...will it break all your package manager dependencies...the horror
<durandal_1707> do not buy/promote such products
cone-099 has quit [Quit: transmission timeout]
<Marth64> what GPU line do you like?
<durandal_1707> none
<llyyr> gpus are made by satan
<durandal_1707> exp closed ones
<durandal_1707> with bad drivers made by companies that maintain it badly
<Marth64> bad drivers is very true
<Marth64> do you remember ATI drivers early 2000s
<durandal_1707> hw is not much better
<Marth64> save the money and get better CPU?
<durandal_1707> i used 3dfx in 98
<durandal_1707> CPU is closed too
<Marth64> yes sadly
<durandal_1707> and some powerful companies are doing shady stuff
<Traneptora> plz DRM is digital radio mondiale
<Marth64> i don't even know what phone to buy nowadays, they all are adware and i don't have time for root shenanigans
<durandal_1707> if they found you root-ed it you get into jail
<Marth64> DRM the radio standard I remember sounding good
<Traneptora> no idea what it is just know that it comes up when I google DRM and it makes me mad
<Marth64> durandal_1707: some manufacturers give you the bootlaoder keys
<durandal_1707> Marth64: what is that?
<Traneptora> I'm a pleb who has a budget samsung phone because
<Traneptora> well
<Traneptora> I don't want to deal with stuff
<Marth64> durandal_17: like Google Pixel you are free to install your own firmware, which is really nice https://developers.google.com/android/images
<durandal_1707> ah, and what you could install on it?
<Marth64> Traneptora: I can't blame you, mines a couple years old and had no motivation to touch it besides its getting old
<Marth64> durandal_1707: basically you can build android yourself, or get popular builds like LineageOs and install it on the device. so you wipe away factory bloat OS and install this. it's aallowed. they will block you access to heightened security things like payments
<Marth64> I can't blame them to be honest for that
<durandal_1707> everything is going digital
<durandal_1707> how will it handle analog stuff?
<Marth64> they dropped amps long time ago
<Marth64> amps/tdma
mkver has quit [Ping timeout: 252 seconds]
<Lynne> Traneptora: it's insane, but I think you can use avoption for exif
<Lynne> avoption is a public struct
<Lynne> and it supports reading, and it supports multiple types for each option
<durandal_1707> except arrays
<Lynne> or yeah, json, we have utils for reading/writing json I think
<Lynne> or xml
<Marth64> hypothetically if we have tools for reading json. would it make sense to have a loudnorm option to read the first pass parameters from a json string
<Marth64> it currently writes a json, but cant read one. which means trying to do 2-pass you have to interpret and remap from json to loudnorm keys again
<durandal_1707> loudnorm filter is huge slow hack