5 ms·

Google submits VP8 bitstream to IETF, but not as a standard

15 points by abraham 16y ago

alanh 16y ago
I don't have a link on me, but I remember reading about big differences between the spec and implementation.
- othermaciej 16y ago
  The following blog posts from an x264 developer point out problems with the spec, including discrepancies with the implementation and chunks of cut&pasted C code: http://x264dev.multimedia.cx/archives/377 http://x264dev.multimedia.cx/archives/377 http://x264dev.multimedia.cx/archives/486 http://x264dev.multimedia.cx/archives/486 Unfortunately, the VP8 Internet-Draft seems to be just a plaintext version of the bitstream guide complained about above, including the disclaimer that the implementation takes precedence and the snippets of C code. It would need a lot of work to actually be published as any kind of standard, this perhaps being a reason Google is going the individual submission RFC route.
  - ZeroGravitas 16y ago
    Interestingly the latest draft of the IETF's Opus codec spec actually contains source code, zipped and then base64 encoded: http://www.ietf.org/mail-archive/web/codec/current/msg02022.html http://www.ietf.org/mail-archive/web/codec/current/msg02022....
vu3rdd 16y ago
I "heard" that the reason why this cannot be a standard is that the hardware vendors are already making silicon out of the currently published bitstream spec of VP8 and any discussion and a later change in the bitstream can upset such plans. Hardware support is crucial for the success of any multimedia format.
- robin_reala 16y ago
  3 of the last 4 posts on the WebM blog ( http://blog.webmproject.org/ http://blog.webmproject.org/ ) have been about upcoming hardware support. Sooner rather than later I’m sure. Course, these seem to be mostly discrete parts. For real success WebM needs to make it into SOCs.
- ZeroGravitas 16y ago
  They can easily get a tame standards body to bless an existing implementation without any changes if they want. This is basically ECMA's business model. The question is, if there are already multiple interoperable hardware and software implementations, including a couple of independently produced open source ones then what benefit does a further English language description bring? If I ruled the world then two independent open source implementations plus a spec would be required for "standards". If I had to give up one of those, it would be the the spec that went first.
eklitzke 16y ago
So, is "standard" here being used to differentiate something approved by an official standards body like ANSI/ISO vs. the IETF/RFC system? I can kind of see how that would make sense, since RFCs are just memoranda published by the IETF that happen to carry a lot of technical weight (v.s. the long, formal, arduous, and bureaucratic standardization process bodies like the ISO are known for). It's still weird to think though that by this definition TCP, IP, etc. aren't really standards. The other explanation that I can think of is that this is "not a standard" since this is just a working draft, and not an approved/numbered RFC, but clearly the draft here is the first step along that process.
- robin_reala 16y ago
  Add to that all ‘Web Standards’ – the W3 aren’t an official standards body and as such only make recommendations.

15 points by abraham 16y ago

alanh 16y ago

I don't have a link on me, but I remember reading about big differences between the spec and implementation.

othermaciej 16y ago
The following blog posts from an x264 developer point out problems with the spec, including discrepancies with the implementation and chunks of cut&pasted C code: http://x264dev.multimedia.cx/archives/377 http://x264dev.multimedia.cx/archives/377 http://x264dev.multimedia.cx/archives/486 http://x264dev.multimedia.cx/archives/486 Unfortunately, the VP8 Internet-Draft seems to be just a plaintext version of the bitstream guide complained about above, including the disclaimer that the implementation takes precedence and the snippets of C code. It would need a lot of work to actually be published as any kind of standard, this perhaps being a reason Google is going the individual submission RFC route.
- ZeroGravitas 16y ago
  Interestingly the latest draft of the IETF's Opus codec spec actually contains source code, zipped and then base64 encoded: http://www.ietf.org/mail-archive/web/codec/current/msg02022.html http://www.ietf.org/mail-archive/web/codec/current/msg02022....

vu3rdd 16y ago

I "heard" that the reason why this cannot be a standard is that the hardware vendors are already making silicon out of the currently published bitstream spec of VP8 and any discussion and a later change in the bitstream can upset such plans. Hardware support is crucial for the success of any multimedia format.

robin_reala 16y ago
3 of the last 4 posts on the WebM blog ( http://blog.webmproject.org/ http://blog.webmproject.org/ ) have been about upcoming hardware support. Sooner rather than later I’m sure. Course, these seem to be mostly discrete parts. For real success WebM needs to make it into SOCs.
ZeroGravitas 16y ago
They can easily get a tame standards body to bless an existing implementation without any changes if they want. This is basically ECMA's business model. The question is, if there are already multiple interoperable hardware and software implementations, including a couple of independently produced open source ones then what benefit does a further English language description bring? If I ruled the world then two independent open source implementations plus a spec would be required for "standards". If I had to give up one of those, it would be the the spec that went first.

eklitzke 16y ago

So, is "standard" here being used to differentiate something approved by an official standards body like ANSI/ISO vs. the IETF/RFC system? I can kind of see how that would make sense, since RFCs are just memoranda published by the IETF that happen to carry a lot of technical weight (v.s. the long, formal, arduous, and bureaucratic standardization process bodies like the ISO are known for). It's still weird to think though that by this definition TCP, IP, etc. aren't really standards. The other explanation that I can think of is that this is "not a standard" since this is just a working draft, and not an approved/numbered RFC, but clearly the draft here is the first step along that process.

robin_reala 16y ago
Add to that all ‘Web Standards’ – the W3 aren’t an official standards body and as such only make recommendations.