Re: We must fork the SVG standard (was: SVGA 1.0 uses RAND -> DO NOT ! implement it, DO NOT ! use it)

Daniel Phillips wrote:
> 
> On October 5, 2001 03:43 pm, Chris Lilley wrote:
> > Daniel Phillips wrote:
> > > On October 4, 2001 02:00 pm, Chris Lilley wrote:
> > > > > The public review period, if there ever was one, went by quietly:
> > > >
> > > > Please get your facts in order. The public review period of the SVG 1.0
> > > > specification (from first public working draft on 11 February 1999 to
> > > > end of the Proposed Recommendation period on 16 August, 2001) was *over
> > > > 29 months*
> > >
> > > Let me get this straight.  You claim the public review period for SVG did
> > > not go by quietly?
> >
> > I was responding to your casual slur "if there ever was one". I wasn't
> > prepared to let that particular mud be thrown without picking you up on
> > it.
> 
> Please do not blame me if you created a perception that W3C operates in
> isolation from the public.

I did not create it. You claimed it. SVG has benefited from a long and
detailed period of public comment. There has been excellent feedback
from the public over two years.

> If the public really had been properly made aware of the opportunity to
> review and comment on a version of the SVG recommendation that relied on RAND
> licensing 

Ah, so you are now talking about the PPF.

> I'm sure the comment period would have been much less quiet than it
> actually was.

The comment period on the PPF has indded produced a lot of comment.

> Note that previous version of the SVG specification does not include the word
> "patent":
> 
>    http://www.w3.org/TR/2000/CR-SVG-20000802/

Naturally, since at that point we were not testing out the patent
policy. We had no policy or policy-in-the-making with which we could
require our members to tell us about any patents that they held.

> 
> So, from the public's point of view, on July 19 a single line of text was
> added to the document, which included the word "patent" and linked to another
> document.  No wonder the public did not notice.

The single line of text was in the status of the document, a part very
visible on the front page of the spec.

In fact, the svg developer community was pretty much the only public
forum that had a discussion of the patent policy before it got /. ed

> Let's examine the language that appeared in the July 19 document:
> 
>     http://www.w3.org/TR/2001/PR-SVG-20010719/
>     "There are patent disclosures and license commitments associated
>     with the SVG 1.0 specification. These may be found on the SVG 1.0
>     Patent Statements in conformance with _W3C policy_."
> 
> The _W3C policy_ is a link to:
> 
>     http://www.w3.org/Consortium/Process-20010719/
> 
> which does not mention RAND.

It doesn't mention RF either, you notice.

>  So in what way does the SVG recommendation's
> reliance on RAND licensing conform to W3C's policy as laid out in that
> document?

Because the PF proposes to alter the process document and we were trying
it out. Implementing it. So we were able to require the members of the
SVG WG to do patent disclosures or to say that they were not aware of
patents,and to specify licensing terms.


> > If you take the trouble to read the various stages of the SVG
> > specifications (eight working drafts, two last calls, two releases of a
> > candidate recommendation, one proposed recommendation) you can see there
> > was abundant call for public comment and there was, in fact, also
> > abundant comment which was taken into account in the technical
> > development of the specification.
> 
> But it is a fact that no comment was made on the RAND policy of the SVG
> recommendation.

The SVG specification does not have a RAND policy.

The PPF describes a RAND licensing term and an RF licensing term.

>  So I question your use of the word "abundant". 

I question which spec you are criticising. The SVG specification defines
a graphics format. The PPF spec defines a Patent policy. It seems to be
the latter that you have issues with.

>  It was
> apparently not "adequate".  Or perhaps you believe that the public changed
> its mind between August and September?

No, I believe that between August and September the first public draft
of the PPF was released and has produced a lot of public comment.

> > > Consider that the real introduction of the RAND policy was as part of
> > > the SVG specification,
> >
> > There is no "RAND policy". I suspect you are referring to the Patent
> > policy Framework. Either way, that is a different spec to the SVG spec.
> > Please be clear which specification you are actually talking about.
> 
> I am talking about the SVG recommendation.  This most certainly does contain
> references to, and relies on, RAND licensing policy.

It contains references to licensing policies, yes.

>  This recommendation
> even attempts to define RAND licensing.

No, it does not. Point to where it does, please.

> > > and suppose the public really had been aware that W3C
> > > intended to bless SVG as a patent-encumbered standard.
> >
> > it isn't patent encumbered.
> 
> This is an interesting claim.

I find your claim that it is encumbered no less interesting.

> So you are saying that SVG is not encumbered
> by patents? 

That is my claim, yes. Its up to you to examine my claim and the claim
of others and make your technical judgement about what you can actually
implement based on your knowledge of the field. What W3C does is try to
flush out patent claims early, and encouarage developers of the
specification to give RF license, an to insist on a RANDlicense as the
minimum acceptable.

> It would be nice if this was true.  But what are those RAND
> licenses that the SVG recommendation refers to?

There are two things to consider. The first is actual patent claims -
specific, numbered patents - that have been notified to W3C as possibly
applying to SVG. In some cases (that of Apple) a long time ago when it
was not as clear what was going to be in in the final specification.

The second thing is potential patent claims- stuff as yet undiscovered.
In the case of a spec that is not finalised, claims for patents that
cover things that have not even been added to the spec yet, but might
be. In the case of work that has not even been started at W3C yet,
anything at all that we might do, any patents that might claim to cover
that.... you can see how open ended this is getting. For this second
case, W3C will be requiring all W3C members - not just those in the
working group, but all 500 or so - to offer a RAND license on all that
stuff. At minimum. They can also give an RF license of course.

Note that a company can give a RAND license to the open-ended case two
and then, if they come up with a specific patent that they claim might
apply toa specific w3C specification, they can give a RF license on that
particular patent.

> > > Would we not have expected to see the September 30's 726 comments
> > > submitted instead some time before August 16?
> >
> > Again, are you commenting on PPF or SVG?
> 
> I am comparing one to the other.

OK. They cover radically different technical content.

> > There are hundereds of comments
> > on PPF, yes. Many of them contain real content, which is valuable, and
> > some of them are merely bluster and noise. I'm hoping you can constrain
> > yourelf into the former category.
> 
> And how many comments are there on patent questions? 

There are an increasing number of specific patent questions, which is
encouraging, and in some cases these have split off into more relevant
technical ailing lists where the expertise to discus the specific
technical claims can be found and where people have more knowledge of
specific prior art.


> How many comments in
> total were there on the last month?  Are you saying that the public did not
> have anything to say about the SVG patent issues, so did not say it?

No, but you seem to be saying that.

> > > We saw no such thing.  Instead, we saw the close of the comment period go
> > > by quietly, just as I said.
> >
> > Where are the comments for the Proposed Rec for SVG sent? On what basis
> > do you say that it went by quietly? You have not seen all the comments,
> 
> You mean here?
> 
>    http://lists.w3.org/Archives/Public/www-svg/

No. Some of them were sent there. Others were sent to a team only
mailing list.

> Looks pretty quiet to me, maybe two comments per day.  Are there more I don't
> know about?

Yes.

> > you are blustering, and it shows.
> 
> Respectfully, you are covering your ass, and it shows.

Ok, so perhaps that line of argument is not going anywhere - neither
will convince the other - ;-) so lets stick to factual issues.

> > > I submit that, far from my facts being out of
> > > order, you have attempted to spin the facts in a way that suits you.
> >
> > Well I would say you were spinning the facts in a way that makes a
> > simpler story for your audience, by your consistent refusal to discuss
> > or acknowledge the RF parts of the PPF and by consistently using an
> > nvented name for it "RAND policy" and by attempting to drag in other W3C
> > specs as if they were patent policy and not technical specifications.
> 
> Don't get me wrong.  I think the RF parts are wonderful, and I want to see
> more of that.

Ok, good. That would be a helpful thing to mention. People often don't
mention what parts of a spec they find good. This can result in the only
comments on that part being negative and thus, the part that many
silently liked being dropped.
> 
> I think the RAND parts are terrible.

OK.

-- 
Chris

Received on Friday, 5 October 2001 11:04:30 UTC