dav2d is the fastest AV2 decoder on all platforms :)
Targeted to be small, portable and very fast.
If you're out of the loop like me:
AV2 is the next-generation video coding specification from the Alliance for Open Media (AOMedia). Building on the foundation of AV1, AV2 is engineered to provide superior compression efficiency, enabling high-quality video delivery at significantly lower bitrates. It is optimized for the evolving demands of streaming, broadcasting, and real-time video conferencing.
They've done the same thing with AV1, and I can't see that having prevented adoption, nor can I imagine Sisvel wanting to poke the bear that is AOMedia unless they're certain their case is absolutely watertight.
I see zero public evidence that they've filed any lawsuits against the members of AOM in any jurisdiction. I'm sure there's been a lot of threatening letters sent...
Same, which is what makes it seem to me that that case is absolutely not watertight. Those patents are probably all about esoteric minutiae (to be fair, that's because that's what it takes to make a better video codec these days) and everything and anything that can seemingly be connected to AV2 (or AV1 for that matter), many of which have only gotten a patent because the person approving it only barely understands what it's saying.
Sisvel is a patent troll. Take a look at the combined list of all companies that are the AOM and tell me with a straight face that all of their corporate in house counsel specializing in intellectual property law are wrong.
I don't know this stuff super well but I imagine it's not necessarily about the lawyers being right or wrong so much as what they can convince people of. The ideal scenario for the patent troll is they can intimidate you into licensing with them. Another good outcome for them (though more costly) is they can convince some non-expert in court. In either case the big players behind the codec can defend themselves but a small one just picking it up downstream as OSS can't.
Not on topic, but wow the internet has very quickly devolved into: click -> "making sure you're not a bot", click -> "making sure you're a human", click -> "COOKIES COOKIES COOKIES", click -> "cloudflare something something"
Maybe I’m naive about this, but I didn’t expect AI scrapers to be that big of a load? I mean, it’s not that they need to scrape the same at 1000+ QPS, and even then I wouldn’t expect them to download all media and images either?
What am I missing that explains the gap between this and “constant DDoS” of the site?
I highly doubt there is no other technically feasible option to block the AI bots.
You end up blocking not just bots, but many humans too. When I clicked on the link and the bot block came up, I just clicked back.
I think HN posts should have warnings when the site blocks you from seeing it until you somehow, maybe, prove you are human.
I'm sure there are many solutions for many problems, but expecting a small Foss development team to know or implement them all is rather unreasonable.
I think the world gains more if the VLAN team focuses on their amazing, free contribution to the world, than if they spend the same time trying to figure out how to save you two clicks.
We all hate that this is happening, but you don't need to attack everyone that is unfortunately caught up in it.
> I highly doubt there is no other technically feasible option to block the AI bots.
If you have discovered such an option, you could get very wealthy: minimizing friction for humans in e-commerce is valuable. If you're a drive-by critic not vested in the project, then yours is an instance of talk being cheap.
Its pretty explicitly not a tragedy of the commons. Its a tragedy of the ruling class abusing the resources of the 'commons' to extract value. There is nothing 'commons' about trillion dollar companies extracting all available value from the labor of the working class. That's just the tragedy that'll bring around the death of society, the same tragedy that brings all other tragedys
While a lot of the plebs do suck, a pleb who sucks causes way less problems than a big corp that sucks simply by virtue of not having too much resources.
> its citizens that act selfishly and in bad faith will slowly make it unusable
It's rarely been the citizens that have been the problem, but the governments and companies that seek the use the network connection for their overwhelming benefit.
Glorious. Really looking forward to seeing how much better than AV1 it actually turns out to be. It's a shame it'll take a while before we'll have a decent encoder (it took an annoyingly long time until SVT-AV1 was usable).
Not just C, dav1d and dav2d are actually mostly written in ASM! Then there's a bit of C as the glue or for functions that don't have optimized ASM yet.
Since dav2d is newer it has a higher fraction of C, but not enough for it to be the main language in the codebase :)
That codecs should be written in safer languages given that they usually process untrusted files. There have been a number of serious hacks from file parsing bugs due to them being written in unsafe languages.
There's literally a DSL designed for this purpose (Wuffs) so it would be interesting to hear why they didn't use it.
We must not continue to develop media codecs in memory unsafe languages. Small, auditable sections can opt-out perhaps, but choosing default-unsafe for this type of software is close to professional negligence.
Cryptography and video codecs are notable exceptions, they put a lot of effort to making the code provably memory safe: no recursion, limited use of stack variables, no dynamic allocations, etc. As a result, memory safe languages bring nothing but trouble by making it non deterministic, that’s especially true for crypto where compiler “optimisations” guarantee you side channels attacks.
Of the 3 software AV1 encoders, the only one that is fully dead is the Rust encoder (rav1e). If people truly wanted memory safe encoders/decoders, they would fund and develop them.
Project description:
If you're out of the loop like me: - from https://av2.aomedia.org/looks at if AV2 is dead in the water
https://www.sisvel.com/insights/av2-is-coming-sisvel-is-prep...
yep
This is a thinly veiled extortion racket and any competent system would fine them into bankruptcy.
They've done the same thing with AV1, and I can't see that having prevented adoption, nor can I imagine Sisvel wanting to poke the bear that is AOMedia unless they're certain their case is absolutely watertight.
I see zero public evidence that they've filed any lawsuits against the members of AOM in any jurisdiction. I'm sure there's been a lot of threatening letters sent...
Same, which is what makes it seem to me that that case is absolutely not watertight. Those patents are probably all about esoteric minutiae (to be fair, that's because that's what it takes to make a better video codec these days) and everything and anything that can seemingly be connected to AV2 (or AV1 for that matter), many of which have only gotten a patent because the person approving it only barely understands what it's saying.
Sisvel is a patent troll. Take a look at the combined list of all companies that are the AOM and tell me with a straight face that all of their corporate in house counsel specializing in intellectual property law are wrong.
I don't know this stuff super well but I imagine it's not necessarily about the lawyers being right or wrong so much as what they can convince people of. The ideal scenario for the patent troll is they can intimidate you into licensing with them. Another good outcome for them (though more costly) is they can convince some non-expert in court. In either case the big players behind the codec can defend themselves but a small one just picking it up downstream as OSS can't.
I don't doubt for a minute that they are going to attempt to intimidate companies using av1 which are much smaller than the AOM founders.
Not on topic, but wow the internet has very quickly devolved into: click -> "making sure you're not a bot", click -> "making sure you're a human", click -> "COOKIES COOKIES COOKIES", click -> "cloudflare something something"
We had to set it up on the parts of VideoLAN infra so the service would remain usable.
Otherwise it was under a constant DDoS by the AI bots.
Maybe I’m naive about this, but I didn’t expect AI scrapers to be that big of a load? I mean, it’s not that they need to scrape the same at 1000+ QPS, and even then I wouldn’t expect them to download all media and images either?
What am I missing that explains the gap between this and “constant DDoS” of the site?
I highly doubt there is no other technically feasible option to block the AI bots. You end up blocking not just bots, but many humans too. When I clicked on the link and the bot block came up, I just clicked back. I think HN posts should have warnings when the site blocks you from seeing it until you somehow, maybe, prove you are human.
I'm sure there are many solutions for many problems, but expecting a small Foss development team to know or implement them all is rather unreasonable.
I think the world gains more if the VLAN team focuses on their amazing, free contribution to the world, than if they spend the same time trying to figure out how to save you two clicks.
We all hate that this is happening, but you don't need to attack everyone that is unfortunately caught up in it.
> I highly doubt there is no other technically feasible option to block the AI bots.
If you have discovered such an option, you could get very wealthy: minimizing friction for humans in e-commerce is valuable. If you're a drive-by critic not vested in the project, then yours is an instance of talk being cheap.
I'm all ears on how we can fix it otherwise.
Keep in mind that those kinds of services: - should not be MITMed by CDNs - are generally ran by volunteers with zero budget, money and time-wise
The internet is such a Tragedy of the Commons… its citizens that act selfishly and in bad faith will slowly make it unusable.
No, it is because citizen allow treating them like this.
Its pretty explicitly not a tragedy of the commons. Its a tragedy of the ruling class abusing the resources of the 'commons' to extract value. There is nothing 'commons' about trillion dollar companies extracting all available value from the labor of the working class. That's just the tragedy that'll bring around the death of society, the same tragedy that brings all other tragedys
The commons in question is the internet itself.
Thank you for describing the tragedy of the commons
The commons were never unregulated. This is a tragedy of enclosure.
https://en.wikipedia.org/wiki/Enclosure
There’s definitely lots of problems with the ruling class and wealth disparity. Perhaps the defining problems of our current age.
That being said, so many of the plebs suck. Like 2% will ruin everything for everyone.
While a lot of the plebs do suck, a pleb who sucks causes way less problems than a big corp that sucks simply by virtue of not having too much resources.
> its citizens that act selfishly and in bad faith will slowly make it unusable
It's rarely been the citizens that have been the problem, but the governments and companies that seek the use the network connection for their overwhelming benefit.
wat. The protections in place that the OP is talking about are almost entirely due to (not government and company) bad actors.
Nearly every single website I'm not logged into these days want me to "confirm I'm not a bot".
it is incredibly annoying but what can you do? AI scrapers ruined the web.
Wow I’m glad it’s not just me. I thought my IP block had gotten caught up in some known spamming or something.
I get exactly none of that. Is your adblocker still working?
renders your gigabit connection pointless
Glorious. Really looking forward to seeing how much better than AV1 it actually turns out to be. It's a shame it'll take a while before we'll have a decent encoder (it took an annoyingly long time until SVT-AV1 was usable).
>video decoder implementation
>look inside
>it's C
Not just C, dav1d and dav2d are actually mostly written in ASM! Then there's a bit of C as the glue or for functions that don't have optimized ASM yet.
Since dav2d is newer it has a higher fraction of C, but not enough for it to be the main language in the codebase :)
What are you even implying?
I think they mean that video decoders and encoders tend to have custom assembly code for speedup.
That codecs should be written in safer languages given that they usually process untrusted files. There have been a number of serious hacks from file parsing bugs due to them being written in unsafe languages.
There's literally a DSL designed for this purpose (Wuffs) so it would be interesting to hear why they didn't use it.
muh rust muh safety muh Safe Code
I wonder if the author is a Dave2D fan?
https://www.youtube.com/@Dave2D
I think it's an increment on this: https://www.videolan.org/projects/dav1d.html
They’re more of a D4vd fan
I would even remove the C code and lower the usage of the assembler pre-processor to a basic C pre-processor.
Happy, AV2 decoding already here.
:)
We must not continue to develop media codecs in memory unsafe languages. Small, auditable sections can opt-out perhaps, but choosing default-unsafe for this type of software is close to professional negligence.
Cryptography and video codecs are notable exceptions, they put a lot of effort to making the code provably memory safe: no recursion, limited use of stack variables, no dynamic allocations, etc. As a result, memory safe languages bring nothing but trouble by making it non deterministic, that’s especially true for crypto where compiler “optimisations” guarantee you side channels attacks.
Of the 3 software AV1 encoders, the only one that is fully dead is the Rust encoder (rav1e). If people truly wanted memory safe encoders/decoders, they would fund and develop them.
Fully dead in what sense? Seems like it still has active development to me.
It hasn't had any proper quality/speed improvements in years. Only thing that has changed is updating deps and some bug fixes.
> If people truly wanted memory safe encoders/decoders
Really? How many codecs have your neighbors contributed money for the development of, just curious.
Given Netflix's involvement with SV1-AV1, (not even that) indirectly, at least 1.