Re: public opinion on W3C

Chris Lilley wrote:

> > Before, I hacked my own incompatible crap out of
> > spite, with all the expected usefulness (none).
> 
> ;-) I'm sure you do yourself a disservice there. I
> expect you did what most folks do, in fact, hope
> that there is no problem and as long as no lawyer
> sends you a lteer, you are fine.

Of course I'm doing myself a disservice. I'd rather implement something
standard that people could use, but I like hacking and I will do so in
different ways if a patent prohibits me from a given way.

> > Of course, having up-front notification about
> > patents is nice.
> 
> Ok, good. And the best way to get this upfront
> notification is to require the members of the
> working group while the spec if in develomnent) and
> the W3C in general (once the spec is stable) to
> declare their patents and the license terms.

Agreed.

> > In particular, I find an "open standard" that has
> > a supposedly "non-discriminatory" licensing policy
> > pretty closed and discriminating when it just
> > prevents me from implementing it.
> 
> As opposed to just not knowing at all, which was the
> situation before and apparently did not stop you or
> anyone else from coding. Wheras now you know which
> of the companies that developed the spec already gave
> you a license to develop whatever code you want even
> if it does infringe on their patents, and which ones
> did not.

I agree with this. BTW, I coded *around* patents, but that made me
incompatible with the applications that licensed the patents.

What I'm saying is that those standards that don't say anything about
patents, like the ISO ones, are not *pretending* anything about being
non-discriminatory or whatever, they are just ignoring the issue.

I don't agree with that, and RAND is clearly better than that, but this
is not my point.

My point is that RAND says it is non-discriminatory while in fact it
could be discriminatory.

I will *know* that a specific standard is protected by patents, and I
thus will not waste time implementing it rather than doing it and
getting a love letter from a lawyer preventing me from doing or
distributing it.

But the end result is the same: no open source implementation.

I welcome the time saving and avoided wastage, but if my goal was to
save time, I wouldn't hack at all, saving all the time for those other
things.

If the W3C stance would be something like "all patents are to be made
available 100% royalty-free to those implementing the standards", then
we'd be in business.

It doesn't eliminate software patents. It doesn't even eliminate
standards that depends on software patents. It just allows me to
implement the standard, something that I think everyone should be able
to do regardless of the amount of money they possess (the
"non-discriminatory" part).

> For case 1), any company that gives an RF license
> gives you one less company to worry about. You will
> not be paying them anything unless you
> want to sue *them* for patent infringement.

We all love them. :-)

> b) A company gave RAND but they have no current patents
> that are relevant. Any future patents would be after the
> spec in question was already there, so prior art should
> stop that future patent being granted at all, supposedly

That's the best case involving the RAND (the actual best one is RF).

> c) A company gave RAND and cited a patent by number.
> 
> In case c) you can work round it, as you mentioned
> in your earlier post. Or you can decide that there
> is enough prior art that the patent is worthless
> and you and everyone else can go on ignoring it.

That's my bitch. I'm no one-man Fraunhofer Research Institute.

If a company makes a patented technology and uses it in its proprietary
products, that's okay (I might not like it, but still, they're not
knocking to my house and raping me).

But when you have something like JPEG2000, where the key technology is
patented (I think, not sure), and they have the guts of calling
something like this an "open standard", I'm pissed. And everyone is
supposed to use that? I guess the term "everyone" doesn't include me.

Or you have the even better and more subtle ones. For example, MPEG4 is
based on the QuickTime format, which is widely available and known. But
the most popular codec for QuickTime is the Sorenson codec, throughly
patented. If this extends to MPEG4, I'll be there, with a bunch of "open
standards compliant" files that I won't be able to play.

The current MPEG is simple. You have an MPEG file, you implement an MPEG
decoder and you have an MPEG player. No patents, no nothing.

Then again, I'm no software patent expert. I only want to continue being
a Linux user in a Windows world without being too much of an excluded.

> > But it open the door, and once it is open, I bet
> > you'll see companies gushing in.
> 
> No, it does not open the door. It recognises that
> the door is open, that there are two doors, and
> requires people to say which door they are using.

And the Right Thing for W3C to do in my opinion would be to say that if
a company wants to be part of a working group, they have to take the
Right Door or go away.

I understand that a number of companies and their technologies will not
go to the W3C, wanting to keep a stranglehold on the market. But I'll be
able to point to the W3C proudly, as it used to be and should be.

> So if there is a W3C working group set up on
> streaming media, you and others will question loudly
> if it is chartered to be RAND as opposed to
> Royalty Free. Good.

Yes, but the reason I am posting these comments is that I'd rather spend
my time hacking and implementing those W3C recommendations rather than
questioning loudly.

I scream and complain too much, help me cut the habit! ;-)

> > At least I would get sued because I unknowingly
> > used patented technology, that's the good side?
> > All right...
> 
> No, its a bad side - but its a bad side of software
> patents, not a bad side of the W3C having a policy to
> deal with software patents.

There was a typo in my original message. I meant "at least I would NOT
get sued". This would be the result of using a RAND charter. The "all
right" was ironic, of course.

> > RAND doesn't look like this to me.
> >
> > Regarding your example about GIF, times have
> > changed. Now, patents aren't used as much in
> > submarine kinds of way.
> 
> Really?

Well, ok, they still are, but companies are more forthright with this
nowadays. They issue a patent on something idiotic that a monkey could
think up in a day then fire away patent lawyers at everyone like so many
photon torpedoes, imposing their proprietary products by way of market
control (for example, Windows Media Player).

In the old time, they had to be subtle, so that everyone would start
using their stuff and *then* hit us. If Unisys had been clear with its
LZW patent, GIF would be a dead format today, development of PNG or its
moral equivalent would have been started much earlier and it would be
king today (because in the old time, hackers had some influence on the
technology mainstream, since there was almost no mainstream).

> Right. I agree that the 'circulating patent pool'
> where a few large companies pay each other large
> sums of money to keep their smaller competitors
> out of the game is a very bad aspect of software
> patents. If W3C were to produce sucha specification
> there would be lots of early warnings because the
> working group would be RAND, the spec would say it
> was RAND and everyone would know that the intention
> was to make a fee-bearing spec so they would not
> waste time implementing it. And all W3C specs have
> to go through a Candidate Recommendation phase
> where they get implementation feedback. No
> implementatiosn means no progress. W3C would not
> be well served by making specs that the open source
> community cannot implement; it would just delay
> progress of the spec and increase the chance that
> it never made it to W3C Recommendation at all.

Ok. What I fail to see then is why a RAND policy would be applicable in
an helpful way over an RF one. Maybe that's my whole problem and if I
just understood that, I'd agree with RAND?

What if a company makes the initial implementation? No open source, the
Candidate Recommendation phase goes through and I'm fucked (which I
don't like by people I'm not intimate with).

> As I mentioned, W3C did not invent software patents.
> We are just trying to come up with a way to deal with
> the fact that they exist; a way more intelligent than
> sticking out head in the sand like an ostrich and
> hoping no-one bites our ass. ;-)

Simple: require RF on the charters. There. Next problem please?

> >  W3C didn't call me a sucker in the past, will it
> > do so in the future?
> 
> Well, I certainly have no plan to ....

Excellent! :-)

> > There are "open standards" like JPEG2000 and MPEG4
> > in development, where it is widely known that they
> > are patent-ridden. How "open" is that?
> 
> I agree that they are patent ridden and I agree that
> people respond to that by ignoring specs or
> implementing parts of them (see JPEG for example - it
> is used a lot but the parts that are patented are
> just not used at all).

I think that in the case of JPEG2000, the main encoding algorithm is
patented, so no license, no JPEG2000 software.

> On the plus side, ISO does require that patents be
> disclosed by the folks makinmg the spec. W3C is going
> to do that, too, because it helps avoid some nasty
> surprises.

And ISO is developing JPEG2000, right? What's your point?

Photoshop and Internet Explorer are going to support JPEG2000, and if
the space savings are as they claim, it could very well become very
popular, but Gimp and Mozilla won't be able to. Nice. Next thing I know,
I'll be browsing the web without images, not because I choosed to, but
because I simply won't be able to.

Photography is one of my hobbies, and I visit a number of photo-related
sites where JPEG2000 would make a big difference in their ability to
support a large number of visitors (large images that are supposed to
compress much better with JPEG2000 than with JPEG). I would feel quite
discriminated.

> > Non-discriminatory?
> 
> No, it is not non discriminatory. Perhaps someone
> from ISO can correct me if I am wrong, but I believe
> that ISO just requires necessary patents to be
> declared. It makes no restrictions on the licenses
> on those patents. W3C will require as a bare minimum,
> RAND. And will encourage RF.

What does RAND means more than the ISO policy then? From experience,
"reasonable" is in the eye of the beholder (for a large company,
"reasonable" as in "we have reasonable chances of raking in huge profits
on this"?). What about the non-discriminatory? Is there something to
make it more "solid" and less prone to interpretation?

More specifically, will it discrimine me from doing an open source full
implementation of a Recommendation?

> W3C did not invent software patents. We don't have a
> magic wand to make them go away. What we can do, for
> specifications that we produce, is to make it clear
> for developers what the status is in terms of patents
> and licenses. Then developers can make an informed
> decision about what to implement, how to implement
> it, and what to avoid.

I understand. I totally support your encouragement of RF, I only wish it
was sole policy for W3C.

I'll end this with a well-known quote...

"Anyone who slaps a 'this page is best viewed with Browser X' label on a
Web page appears to be yearning for the bad old days, before the Web,
when you had very little chance of reading a document written on another
computer, another word processor, or another network." -- Tim
Berners-Lee in Technology Review, July 1996 

Is the W3C itself now yearning for the bad old days? I guess not, but
not yearning is a thing, and taking steps (like requiring RF) to avoid
getting back there is another.

-- 
Pierre Phaneuf

Received on Thursday, 4 October 2001 12:25:18 UTC