Chromium Adds Support For Animated PNGs

[TheBlackCat]

In what way is that relevant?

It is relevant because in practice it means you can't implement the spec without paying a single organization for permission to do so.

On the other hand, anyone can get the VP8 or VP9 spec for free and implement it without having to go through a single gatekeeper (or any gatekeeper at all) to do so.

I don't think a definition of "proprietary" is very useful when it says that having a free, publicly-available spec anyone can implement without any permission at all is "proprietary" while having to pay a gatekeeper for permission to even see, not to mention implement, a spec is "open".

[Gusar]

If you have to pay a particular entity for the information needed to implement the format, then yes it is most certainly proprietary.

That particular entity is a standards organization. They did not create the format, it was created by several entities working together and the format is not proprietary to any of them, it was a joint collaboration.

It is relevant because in practice it means you can't implement the spec without paying a single organization for permission to do so.

float's question was about redistributing the spec itself, not about implementing it.

And you're not actually paying for permission, that's a different thing related to patents. Permission to use the patented tech is handled by MPEG LA, which is a *different* entity from the ISO/IEC MPEG standards organization that charges a fee for access to the spec.

On the other hand, anyone can get the VP8 or VP9 spec for free and implement it without having to go through a single gatekeeper (or any gatekeeper at all) to do so.

There is a spec now, but in the beginning there wasn't any. The only way to understand the format was to read a particular entity's implementation. And it's that implementation which defines the format, including its bugs (the spec has to accommodate those bugs which need to be replicated in every independent implementation).

I don't think a definition of "proprietary" is very useful when it says that having a free, publicly-available spec anyone can implement without any permission at all is "proprietary" while having to pay a gatekeeper for permission to even see, not to mention implement, a spec is "open".

While the spec is free, the format is still proprietary. Those aren't exclusive terms.

BTW, I just learned something very interesting... Both h264 and h265 were a joint effort by two standards organizations - ITU-T and ISO/IEC MPEG. I knew that before. But what's new to me is that while MPEG charges for the spec, ITU *doesn't*. So the h264 and h265 specs are actually free! Download the h265 spec here. Now what?

[TheBlackCat]

That particular entity is a standards organization. They did not create the format, it was created by several entities working together and the format is not proprietary to any of them, it was a joint collaboration.

You have to pay someone to even see the spec, so it is proprietary to them since they own the rights to distribute it. It is certainly a standard, but it isn't open in any useful sense of the word since you can even look at it without paying to do so.

float's question was about redistributing the spec itself, not about implementing it.

You can't implement a spec if you can't see it, so the two are connected.

And you're not actually paying for permission, that's a different thing related to patents. Permission to use the patented tech is handled by MPEG LA, which is a *different* entity from the ISO/IEC MPEG standards organization that charges a fee for access to the spec.

You are paying for permission to read the spec, which again is a hard requirement for implementing the format.

There is a spec now, but in the beginning there wasn't any. The only way to understand the format was to read a particular entity's implementation. And it's that implementation which defines the format, including its bugs (the spec has to accommodate those bugs which need to be replicated in every independent implementation).

I am talking about now.

But what's new to me is that while MPEG charges for the spec, ITU *doesn't*. So the h264 and h265 specs are actually free! Download the h265 spec here. Now what?

Look at the date. It was released for free today. When I posted yesterday you needed to pay to download it. They could put it behind a paywall at any time, and the next revision will probably be behind a paywall for some unknown amount of time as well.

[Gusar]

You have to pay someone to even see the spec, so it is proprietary to them since they own the rights to distribute it. It is certainly a standard, but it isn't open in any useful sense of the word since you can even look at it without paying to do so.

The format is open because it was developed in the open by anyone who wished to participate. Contrast that to VP9, whose creation was not open to participation to anyone outside Google. It can't be that hard to acknowledge this fundamental difference.

Also, there is a reason we have the "free-as-in-freedom" and "free-as-in-beer" phrases. Because the involvement of money is a separate matter from having the freedom to do something. While the spec is not available freely (an in beer), it is very much available freely (as in freedom) to anyone, anytime.

You can't implement a spec if you can't see it, so the two are connected.

Again, float's post said *nothing* about implementation, but about redistribution of the spec. Entirely different thing, not related to our discussion at all.

I am talking about now..

And I am saying that the availability of the spec now does not change the proprietary nature of the format, because the spec merely describes a proprietary implementation, including said implementation's bugs. (proprietary in the sense of single-vendor; not in sense of closed-source, which is what many people here associate with the word proprietary).

Look at the date. It was released for free today. When I posted yesterday you needed to pay to download it.

Even from here https://www.itu.int/rec/T-REC-H.265 there was no link to a free version?

They could put it behind a paywall at any time, and the next revision will probably be behind a paywall for some unknown amount of time as well.

So it's free only for certain periods of time? I don't really see the point in that, but ok, it it's really true...

[pal666]

h264 and h265 were created in a *public* process that was *open* to everyone. The result of that process was a *spec* than anyone can use to write a compliant implementation.

bullshit. nobody could ship implementation due to patents. it was created in broken process which encumbers results with patents. so process is broken and spec is proprietary

Can't get more open than that.

bullshit again. aomedia was created with one goal: to get more open than that

The patent encumbrance is an entirely different thing. It has a big effect, but what it does not do is make the formats proprietary.

that is in gusar's world. in real world i quoted wikipedia

does not compute.

that is your defect

[Gusar]

bullshit. nobody could ship implementation due to patents.

Nobody like ffmpeg or x264, to provide two prominent examples? They're getting shipped. Under an open source license even. And none of these developers paid for any patent. So yes, you can very well ship implementations. Patent licensing fees may apply to recipients of those implementations, but that doesn't change the fact that those implementations are getting shipped.

it was created in broken process which encumbers results with patents. so process is broken and spec is proprietary

As I already said, patent encumbrance is a *separate* matter. The only thing that's broken is that the patent licensing situation isn't determined beforehand, something like "all participants agree that licensing will be handled from a single patent pool". But the process of creating the spec itself is not broken. The process is how it *should* be done. It's how it is being done even in aomedia. Except with aomedia, the agreement is "all participants provide their patented tech royalty-free".

bullshit again. aomedia was created with one goal: to get more open than that

No. aomedia was created to provide a royalty-free format. Again, different matter from openness of the format. In contrast to VP8/VP9 though, AV1 will not be proprietary. Because it is a joint collaboration, open to participation from anyone and there will be a spec from the very beginning.

that is in gusar's world. in real world i quoted wikipedia

Ah, yes. One word in a wikipedia article, that is contradicted in the very same sentence said word is written. That and only that defines "real world". Right...

that is your defect

Nope, that is you being pwned.

[TheBlackCat]

The format is open because it was developed in the open by anyone who wished to participate.

And my point is that development is only one part. It is much more important whether people can freely implement the spec. If there are gatekeepers that restrict access to the spec, then in practice it is not open.

Also, there is a reason we have the "free-as-in-freedom" and "free-as-in-beer" phrases. Because the involvement of money is a separate matter from having the freedom to do something. While the spec is not available freely (an in beer), it is very much available freely (as in freedom) to anyone, anytime.

Actually, no, it was only open to people who applied to be and were accepted as members of their organization.

Again, float's post said *nothing* about implementation, but about redistribution of the spec. Entirely different thing, not related to our discussion at all.

You seriously can't be this obtuse. You can't implement a spec you can't see. One of the main points of "free-as-in-freedom" is that you can redistribute it That prevents one or two gatekeepers from controlling who can and cannot implement the spec. Who, if anyone, controls who can and cannot implement a spec

Even from here https://www.itu.int/rec/T-REC-H.265 there was no link to a free version?

Correct. You can see from the data, it was released 3/16/2017. Before that, only the paid version was available. And not just paid, you had to be a paid member of their organization to access it.

So it's free only for certain periods of time? I don't really see the point in that, but ok, it it's really true...

That is exactly what I am trying to say: it doesn't matter who contributed to the spec, if your ability to implement it exists solely at the whims of some organization that in practice it is not open.

[Gusar]

And my point is that development is only one part. It is much more important whether people can freely implement the spec. If there are gatekeepers that restrict access to the spec, then in practice it is not open.

People *can* freely implement the spec. I could. Well, I lack the coding skills for something as complex as a video format, but if I had the skill, I could get the spec today and start coding. So could you. Or any other participant of this forum. So if that's the criteria, well, h264/h265 fulfill it.

You're talking as if the gatekeeper is highly selective in who they let through. You'd have a point if that was the case. But the gatekeeper is not selective at all. There is a paywall, but no one will be turned back when they show up with the money.

Actually, no, it was only open to people who applied to be and were accepted as members of their organization.

But there were no restrictions as to who could apply. If you had tech to offer, you were at least considered. Not so with VP8/VP9, where there wasn't even the possibility to apply.

You seriously can't be this obtuse. You can't implement a spec you can't see.

You are the obtuse one. Let's see if this will work: For redistribution to happen you must have something to redistribute. So in order to be in the position to redistribute the spec, you must already have it.

float's question has nothing to do with your or anyone else's ability to implement the spec, he merely asked if you go to jail if you redistribute it. And the answer is - jail is highly unlikely, you'll probably be hit with a cease and desist letter and then be sued in a civil court if you don't comply with the letter, that is if you fail to stop your redistribution after receiving the cease and desist. Regarding float's question, that's all the discussion that can be had. Whatever other thing you're trying to bring into it, is an entirely different discussion.

One of the main points of "free-as-in-freedom" is that you can redistribute it That prevents one or two gatekeepers from controlling who can and cannot implement the spec. Who, if anyone, controls who can and cannot implement a spec

You don't have the freedom to redistribute the spec, that is true. But that doesn't take away other's freedom to implement it. Because they can get the spec the same way you did. The other's ability to implement the spec does not hinge on your ability to redistribute it.

That is exactly what I am trying to say: it doesn't matter who contributed to the spec, if your ability to implement it exists solely at the whims of some organization that in practice it is not open.

Ok, it seems ITU has whims then. But MPEG doesn't. Go here, pay the fee, the spec is yours. Whim free. So if your ability to implement it is tied to whims, but there are no whims, then in practice it is open. That's your own logic at play here.

[chithanh]

They submitted for inclusion as part of the PNG standard but it was rejected due largely to philosophical differences about compatibility with existing PNGs. If that is a "disregard for open standards", what would a "regard for open standards" look like? Just giving up on having a backwards-compatible animated image format?

That is not the point. I have no problems if Mozilla creates APNG (based on PNG or whatever) and uses their own MIME type and file signature for it. It would just be another proprietary format that you can support or ignore if you wish. I would not care whether Chrome supports it or not.

But Mozilla instead chose the embrace-and-extend approach of hijacking PNG, using the same MIME type and file signature, in a blatant disregard for open standards. This is why I think very poorly of them now.

That's what happens when the format is considered stable. Regardless of control. Nobody controls/maintains GIF now, but all its problems are now "enshrined forever" because it's so old. You can't just go and make serious changes in there without breaking compatibility with existing decoders/encoders/old gif files. Not because someone who "controls" it will stop you.

That the libvpx bug is now forever part of VP9 is a direct result of Google's closed process to implement libvpx and then declare it the reference implementation of their standard.
If there had been multiple entities working on independent implementations (as is e.g. W3C rule before a draft standard is promoted to recommendation) then a bug in any single implementation them would not become part of the standard. Also if there had been fair participation rules, then Google would not be the one making the decision.

both h264 and h265 are proprietary due to patents
you've got it backwards. here is quote from wikipedia

If you have to pay a particular entity for the information needed to implement the format, then yes it is most certainly proprietary.

Whether the standard is patent encumbered is inconsequential to the question whether it is open. The openness is defined by the participatory process of creating and developing the standard.

You have to pay someone to even see the spec, so it is proprietary to them since they own the rights to distribute it. It is certainly a standard, but it isn't open in any useful sense of the word since you can even look at it without paying to do so.

Lots of standards bodies make the specs only available to paying members for free, and charge a fee for any outsiders. VESA DisplayPort is one such example. It is however still an open standard, because VESA has fair participation rules. This means any individual or organization can become VESA member and has a say in the standard.

[pal666]

Whether the standard is patent encumbered is inconsequential to the question whether it is open. The openness is defined by the participatory process of creating and developing the standard.

so now you share your fantasy world with gusar. it does not change reality