[whatwg] The political and legal status of WHATWG

[Ian Hickson]

On Sat, 15 Dec 2007, Shannon wrote:
>
> It's clear the opinions of all parties cannot be reconciled. The current 
> text has not reconciled the views, nor did the previous, nor can a 
> future one. It doesn't take a genius to figure out that this will not 
> end well. I am quite certain the issue at stake here cannot be solved at 
> the technical or legal level at all. This is truly a political/financial 
> matter.

Indeed, it is an unfortunate situation. There are people at the political 
and financial levels working to address the situation to everyone's 
satisfaction, but obviously such work is by its very nature not 
transparent, and I am not able to uphold the usually high level of 
transparency afforded to the rest of HTML5 with this issue.


> Which brings us to the hard questions at the crux of the matter:
> 
> 1.) When a browser vendor acts clearly outside of the public interest in 
> the making of a public standard should that vendors desires still be 
> considered relevant to the specification?

Their technical arguments, where valid, should be considered, as should 
all technical arguments, yes. Of course, if someone or a some group is 
acting in bad faith, then their credibility is reduced and their 
non-technical input is likely to be ignored more. In the WHATWG I do not 
believe we have seen any browser vendors act in bad faith; indeed on this 
issue in particular all the browser vendors have been quite forthcoming 
(though, unfortunately, not very compromising).


> 2.) When an issue is divided between a vendor (or group of) and 'the 
> public' (or part of), what relative weight can be assigned to each?

The needs of the end users are paramount. Unfortunately, since the browser 
vendors have the final say (in a very pragmatic sense, since they 
ultimately decide what their browsers implement), the needs of the end 
users can often only be met by satisfying the needs of the browser vendors 
as well.


> 3.) When a vendor makes false or misleading statements to guide an 
> outcome should there be a form of 'censure' that does not involve a 
> public flame war?

The reaction to someone or some group acting in bad faith have to be 
catered on a case-by-case basis. In one case I have banned someone from 
the list for a couple of weeks; in other cases issues have been resolved 
simply by open debate. I work closely with the browser vendors in the 
WHATWG to ensure that they do not consider dishonesty a useful policy.


> 4.) If the purpose of the group is to build interoperability should a 
> vendor be 'censured' for holding interoperability to ransom without 
> sufficient technical or legal grounds?
> 5.) What methods exists to define a disruptive member and remove them 
> from further consideration?

These questions have basically the same answers as questions 1 and 3 
above.


> 6.) Should a standards body make a ruling even though some members claim 
> they won't obey it?

I would have to say no -- doing so would make the relevant standard 
irrelevant, which seems like a poor policy. Speaking for myself, I want my 
work to be relevant and helpful to the human race. I'm not doing this to 
satisfy an innate desire to write a specification. If people are going to 
ignore me, I have failed.


> 7.) Should a standards body bow to entrenched interests to keep the 
> peace?

Standards bodies aren't intended to keep peace, generally speaking. In 
fact, one of the big underlying goals of the WHATWG is to encourage active 
competition and revitalise a market that, at the time of the WHATWG's 
creation, was stagnating.


> 8.) Does the WHATWG consider itself to be a formal standards body?

No, the WHATWG is just a community of people interested in evolving the 
Web. However, I don't think it matters if the body creating a 
specification is a formal standards body, an open community, or something 
else, all should hold themselves to the highest standards of quality and 
relevance that I am referring to herein.


> 9.) Should HTML5 be put back under direct control of the w3c now that 
> they have expressed interest in developing it?

It is "under direct control of the w3c". It just happens that I'm editor 
of the spec in the W3C as well as the WHATWG and I'm editing the two specs 
in the exact same way at the same time, and am taking input from all 
sources while editing both documents.


> 10.) Is is appropriate for members to have discussions outside of this 
> list, via IM, IRC or physical means not available or practical to the 
> public?

I don't think it's possible to stop it. I also don't think it's 
desireable. Everyone should be able to contribute, but it would be very 
difficult to enforce a policy saying that nobody was allowed to discuss 
the specification unless it was on the mailing list. Many good ideas come 
from informal and out-of-band discussions.


> 11.) Does the group consider HTML5 to be a 'public standard' or a 
> 'gentlemen's agreement' between vendors?

It's a specification. I don't really know what the two terms you use mean 
or how they differ.


> 12.) Is it even legal for the majority of commercial browsers to form 
> any agreement that could (directly or indirectly) result in higher costs 
> for end-users?

The idea that a working group could be considered a cartel is a novel idea 
to me, it hadn't occured to me before. Given that since the WHATWG has 
started, competition between members of the working group (and other 
vendors) has actually dramatically increased, I don't think there's any 
fear that it will happen. Specifically when it comes to video codecs, 
consider that at least two of the three main vendors actively involved in 
the WHATWG are extremely against anything that isn't royalty free, and the 
third agrees that a royalty free codec is the only solution. I don't think 
any of the vendors could legitimately be argued to be acting against the 
user or against competition in this regard.


> How do you prevent a 'working group' from becoming a cartel?

I spoke to a lawyer about this and, informally, he said that so long as 
the working group is not making people agree not to compete, it's pretty 
safe. Agreement to follow standards is not considered anti-competitive (if 
anything it helps encourage and foster competition -- the whole point of 
HTML5 is to allow new browser vendors to enter the field without having to 
perform expensive reverse-engineering efforts).


> I've read the charter but it doesn't define many rules.

Basically the operating model of the WHATWG boils down to "the editor has 
the ultimate say, and if the members disagree with the editor, they can 
remove him from office". The editor is me, the members are the people 
listed at the end of the charter.


> To cite a theoretical example: if some altruistic billionairre was to 
> write the 'missing codec' that exceeded h.264 in compression, used 64k 
> of ram, ran on a 386 using 50 lines of code and he or she paid off all 
> the trolls and indemnified vendors - what actions, if any, would WHATWG 
> members take to ensure it was accepted by members with vested interests?

I imagine if such a codec existed, there wouldn't be anyone opposed to 
having it, and it would solve our problem neatly.


> If answers to the above questions exist then please don't just answer 
> them here. Anything short of a formal document on the WHATWG can't 
> possibly represent the group as a whole and is just going to be raised 
> again anyway. In other words the mailling list is not the best place to 
> archive these answers (if any are forthcoming).

I'm not sure the questions, as phrased, really work for the FAQ or for a 
blog entry... if you could phrase a question that covers the issues, I 
would be happy to add that question to the FAQ and answer it there. I'm 
not sure what else I can do.

You should also feel free to write a blog post covering this issue, if you 
so desire. For example, you could summarise my responses above. I would be 
happy to work with your to get such a blog post written, if you like.

HTH,
-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'