[whatwg] Patent Policy and Process

[Ian Hickson]

On Mon, 14 Jun 2004, Jim Ley wrote:
> > To answer your specific questions, the structure is two tier. The active
> > tier is this mailing list, the contributors being anyone who sends
> > comments on the mailing list, those comments being taken into account by
> > the spec's editor (myself, at the moment) with all feedback being
> > responded to.
>
> So the editor of a spec. has a requirement to respond to all feedback?  Can
> this be specifically mentioned in the charter.

No, the members as a whole do. It is mentioned specifically in the
charter.


> Is there any time limit for such response?

No, but it is pretty obvious when someone is being ignored.


> Is there any process for raising issues that are not responded to, to
> some sort of arbritration?

No.


> > These members have the final say on what goes in the specification,
> > and on when the specification hits its various milestones. I
> > personally only have one vote in this group, so I'm obviously not
> > all-powerful. :-)
>
> Er, except you (as editor) have the casting vote under the "can't come
> to agreement" clause, that's more than a single vote.

The charter doesn't say the editor has the casting vote. Quite the
opposite. It just says the editor has to be the one to get the others to
agree on something.


> Can there be a requirement that such disagreements between the group be
> documented publically in response to any decisions.

The whole process happens on a public mailing list.


> > According to the charter, all issues raised in this mailing list
> > should be addressed. I have been doing this so far.
>
> Is there any process for achieving this? - an issue tracking log etc.

Nope. Doesn't seem necessary at this time. If it becomes a problem we can
set up a Bugzilla or somehting.


> > The copyright status of the specification is that it is copyright Opera
> > Software. This is because an Opera employee has been the sole editor of
> > the spec and WHATWG cannot legally own any copyrights as it is not a legal
> > entity.
>
> You stated above that you were operating as an individual, now you say
> you're operating as part of Opera, which is it?

Legally anything I do for Opera while under employment at Opera is
Opera's, not mine. Hence they own the copyright on the work I've done
here. However, WHATWG doesn't recognise my affiliation.

To be honest the copyright is just because I originally wrote Web Forms 2
before WHATWG and the issue hasn't been reexamined because frankly, who
cares. It doesn't make any difference. The whole thing should probably be
put into the public domain anyway.


> Please put WHATWG on the level of a legal entity

Highly unlikely to happen. The whole point of this work is to prepare a
proposal for a real standards organisation, not to become one.


> I currently do not believe that you're operating as individuals, as you
> noted this came directly out of Opera and Mozilla's proposal to the W3
> Web-apps workshop.

The announcement I referred to was unrelated to the workshop.


> I don't particularly see how you can address this lack of good faith in
> the charter, perhaps it would be better if the companies involved either
> specifically got involved or made statements specifically distancing
> themselves from the actions of their employees.

I don't really understand what lack of good faith you are talking about.


>> The specification is open, thus not proprietary.
>
> This is an interesting definition of open - it's proprietary to being
> more generous to the members of the WHATWG, but it looks currently to be
> proprietary purely to Opera.

I have no clue what you mean by this.


> > It's just a draft of a proposal that is being developed openly in this
> > mailing list based directly on input from anyone who wishes to e-mail
> > this list.
>
> So why not give such people a vote in what enters the spec, that would
> be truly open, at the moment you could (I know _you_ wouldn't) just go
> WIBBLE in response to every issue, and the members could still vote in
> the complete opposite direction.

Because specs that are designed by committee largely suck. In fact of all
the specs I've been involved with, I can pretty much state categorically
that all those that were written by a single editor taking input from
others but being basically autonomous are better than all those where the
design was driven by committees and votes.


> > > for answering implementation issues
> > Any issues should be raised on this mailing list, whether design, author,
> > or implementation issues, and all will be addressed in the same manner.
>
> So a statement on the mailing list asking to resolve how to implement a
> particular feature should be considered final?  Does this need to be from
> the editor, or any member of the Working Group?

Nothing is final unless it's in an official standard (which WHATWG has no
plans of ever doing).


> > > Are features sections or subsection - ie do you require 2 full
> > > interopable implementations of every section
> > This is fully defined in the charter, which I won't quote here due to its
> > length. However, as far as I can tell it does address your wish above.
>
> I'm sorry I don't see at all how it's defined in the charter, let me re-word
> it as an example Does Section 5 require 2 interopable implementations or
> only subsection 5.6 or only subsection 5.6.1 ?

That rather depends on the precise sections in question. In practice, the
kinds of tests I write involve the interactions of anything that is in any
way interrelated, so sections 2, 3, 4, 5, and 7 would all have to
interoperate together for section 5.6.1 to be considered interoperable.


> > > What intellectual property do the organisations have in the various
> > > areas, what licensing terms would that IP be made available to
> > > others on?
> >
> > No announcement regarding patents has yet been made. This is being studied
> > by the members at the moment. It's likely that the royalty free version of
> > the W3C patent policy will be used wholesale.
>
> You've said above that you're acting as individuals, whilst it's entirely
> possible that you hold patents, it's much more likely that your companies
> do, the W3C RF patent policy seems wholly inadequate when it's just 7 guys
> stating they don't hold patents in the specification.   What's relevant is
> if Apple or Opera who are much more likely to actually hold a patent do or
> not.  Again as above I don't see how the distinction between individuals and
> their companies really works in the organisation.

I don't have anything to say on the patent situation, except that
obviously it's in all our interests if this work is not patent encumbered.
Unfortunately all patent issues have to deal with lawyers and lawyers
aren't known for their swift actions.


> > > What does "shipping" and "unofficial" mean w.r.t. to implementations -
> > > is a beta release of Opera shipping or development,
> >
> > Development (that's what "beta" means).
>
> So a FireFox implementation would not (currently) count, this seems rather
> harsh, and seems to suggest it will be a very long time before you actually
> get 2 releasable implementations.

It is much more likely to take a long time to get interoperability simply
due to bugs than it is due to technicalities in version numbering.


> > > is an extension to IE official or unofficial?
> >
> > Official (it's the extension that is the user agent, and if the
> > extension's vendor says a certain build is official, then it is official).
>
> Right, I'm not sure I can particularly agree with this, could you make this
> clear in the charter, and state exactly how such extensions would need to be
> configured - for example a script included on every page would only "work"
> if certain other scripts weren't on the page, which would mean it's always
> possible to show a valid Web-Forms 2 document that that implementation
> failed to render. So such an extension should never be usable for an
> implementation in the spec.

It is not expected that such a user agent would form any significant part
in showing interoperability for these purposes, no.


> > > how long does the WG undertake to provide the archive?
> >
> > No guarentee is given. Contributors are encouraged to make their own
> > private copies if they require long term availability of the archives.
>
> Individual users do not [need] permission from the copyright holders of
> the messages to make copies hence:
>
> > > Will you solicit approval from contributors for other groups to archive
> > > the mailing list and specifications to ensure continuation?
> >
> > No policy on this matter has been decided at this time.
>
> Can you please decide on this soon.

If pressed, I imagine the answer will be "no" (we won't solicit such
approval). In fact I wouldn't expect the members to get in any way
involved with any third party archiving.

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