Linked by Thom Holwerda on Sun 31st Jan 2010 14:20 UTC, submitted by lemur2
Internet & Networking Despite the recent interest in adopting HTML5's video tag, there is still one major problem: there is no mandated standard video codec for the video tag. The two main contestants are the proprietary and patended h264, and the open and free Theora. In a comment on an LWN.net article about this problematic situation, LWN reader Trelane posted an email exchange he had with MPEG-LA, which should further cement Theora as the obvious choice.
Thread beginning with comment 407167
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE[3]: Correction
by lemur2 on Mon 1st Feb 2010 03:00 UTC in reply to "RE[2]: Correction"
lemur2
Member since:
2007-02-17

The license doesn't have to be transferable, the product itself carries the license.

Mozilla could provide h.264 with every firefox download, but it would involve having to provide a binary blob and paying the royalty cap ($5M or something like that). Mozilla could certainly afford it if they wanted to go that route, with the pot of gold that Google has been providing these many years.

It wouldn't impact re-distribution, either. The distros could still include the blob.

Not that this is an ideal solution, and certainly flies in the face of Mozilla's OSS roots, but it's not un-doable. Mozilla could provide an optional binary blob if they choose to.


This doesn't make any sense. What exactly is supposed to be achieved by having a "binary blob"? MPEG-LA themselves, as I understand it, publish source code for a reference codec. It isn't a question of the open source vs closed binary blob nature of distibuting a h264 codec that is the problem viz a viz licensing, it is the third-party redistribution that is the problem.

In short, if Mozilla pay for a license, and distribute their software, and their license allows recipients to re-distribute the software, then those who re-distribute it haven't paid any license fee to MPEG-LA. Mozilla could "get away" with paying only a small license fee by distributing their browser to only a very few recipients (such as SourceForge and ibiblio perhaps). Once SourceForge and ibiblio re-distribute it, Mozilla's browser still ends up on as many end users systems.

This problem of unpaid-for re-distribution still occurs regardless if Mozilla distribute the codec as an included binary blob or as open source.

It is the patents on h264, and the expectation of MPEG LA that they should get paid for every copy (and for every encoding, and for every transimission of data), that is the problem, and not the fact that Mozilla Firefox is open source.

Edited 2010-02-01 03:04 UTC

Reply Parent Score: 3

RE[4]: Correction
by elsewhere on Mon 1st Feb 2010 07:01 in reply to "RE[3]: Correction"
elsewhere Member since:
2005-07-13

This problem of unpaid-for re-distribution still occurs regardless if Mozilla distribute the codec as an included binary blob or as open source.

It is the patents on h264, and the expectation of MPEG LA that they should get paid for every copy (and for every encoding, and for every transimission of data), that is the problem, and not the fact that Mozilla Firefox is open source.


MPEGLA caps the royalty fee, they don't demand an infinte per product charge (as opposed to some other patent pools that demand an absolute charge per every implementation.) You pay per implementation, up to a fixed ceiling, beyond that, it is gratis.

Mozilla (or Opera or anyone else) could simply pay that cap fee, and not worry about it. Granted, it amounts to several million dollars, but hey, it's only money.

How do you think Adobe et al. are able to distribute h.264 compatible players, even allowing re-distribution? Do you think Adobe is paying a fee for every download of flash everywhere in the world? openSUSE, Ubuntu et al. aren't paying a fee for providing the player as a download separate from Adobe's site. HP, Dell et al. aren't paying a fee for proving Flash pre-installed on their systems. Do you think Google is paying a fee for each download of Chrome? No, they simply pay the cap and don't worry about having to account for individual downloads.

The MPEGLA license applies to an implementation within a specific product. If you distribute a binary blob, that's an implementation within a single product, so there's no problem.

The model doesn't work for source code, because source code can be modified to become an infinite number of products.

The model sucks, and I'm not justifying it, but it's just the way it is.

Reply Parent Score: 3

RE[5]: Correction
by lemur2 on Mon 1st Feb 2010 07:41 in reply to "RE[4]: Correction"
lemur2 Member since:
2007-02-17

"This problem of unpaid-for re-distribution still occurs regardless if Mozilla distribute the codec as an included binary blob or as open source.

It is the patents on h264, and the expectation of MPEG LA that they should get paid for every copy (and for every encoding, and for every transimission of data), that is the problem, and not the fact that Mozilla Firefox is open source.


MPEGLA caps the royalty fee, they don't demand an infinte per product charge (as opposed to some other patent pools that demand an absolute charge per every implementation.) You pay per implementation, up to a fixed ceiling, beyond that, it is gratis.

Mozilla (or Opera or anyone else) could simply pay that cap fee, and not worry about it. Granted, it amounts to several million dollars, but hey, it's only money.

How do you think Adobe et al. are able to distribute h.264 compatible players, even allowing re-distribution? Do you think Adobe is paying a fee for every download of flash everywhere in the world? openSUSE, Ubuntu et al. aren't paying a fee for providing the player as a download separate from Adobe's site. HP, Dell et al. aren't paying a fee for proving Flash pre-installed on their systems. Do you think Google is paying a fee for each download of Chrome? No, they simply pay the cap and don't worry about having to account for individual downloads.
"

I know that the license fee is capped ... it doesn't matter, it still applies.

The MPEGLA license applies to an implementation within a specific product. If you distribute a binary blob, that's an implementation within a single product, so there's no problem.

The model doesn't work for source code, because source code can be modified to become an infinite number of products.

The model sucks, and I'm not justifying it, but it's just the way it is.


Nope. Nice try, but no.

Your explanation doesn't account for the observation that MPEG LA themselves distribute an open source reference implementation.

It is not being open source that attracts a fee ... it is the mere act of distribution of a h264 codec that attracts the fee.

Reply Parent Score: 2