OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [xml-dev] [Fwd: W3C ridiculous new policy on patents]

> Is there any public discussion of SVG and IP issues? 

There was substantial discussion on the SVG developers list: 
These and other threads from that time frame.

An optimistic(?) summary from Chris Lilley (post 5471) had some
explanation from his perspective:

r_diblasi@hotmail.com wrote:
> W3C Team and community,
> In Status of this document there is a paragraph
> discussing patents:
> http://www.w3.org/TR/SVG/#status
> <snip>
> 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</snip>
> http://www.w3.org/2001/07/SVG10-IPR-statements.html
> Can anyone comment of how or if the patent disclosures and
> license commitments may affect developer of products based
> on this specification and any of the patent mentioned in this
> SVG specification?
> This patent disclosure and license commitments paragraph
> has never been in any of the SVG specifications ....until now

That is correct. In fact, they have never been in W3C specs, until now.
Well actually SMIL 2.0 had them when it went to Proposed rec, a few
weeks before SVG did.

There is a problem, that W3C pretty much ignored until recently, calle
dsubmarine patents. Someone has a patent, and they sit on it ad watch
people build dependencies on it, maybe even join a working group and
encourage people to depend on this patent. Then, when the standard is
final (or perhaps years later) , they produce this with a flourish and
demand lots of money (perhaps retrospectively).

So, W3C has been working on ways to deal with this problem.

There is legal precedent, for example the celebrated Rambus case, and
another one involving Dell, where a patent was withheld and then
produced with the expectation of generating a lot of money, and the
courts ruled that the non-disclosure of the patent was illegal and that
no fees were to be paid *because there was an existing agreement for
early disclosure of patents*.

W3C is thus putting into place such a system of early disclosure of
patents, to guard against precisely this sort of problem.

All members of the SVG WG were asked to disclose patents which wrre
'essential technology' before SVG went to Proposd Rec. You (well,
everyone) noticed there was a bit of a delay before SVG went to Proposed
rec; now you know what we were doing.

The W3C membership at large is similarly being asked about essential
patents, before SVG becomes a REC.

This is good news for developers. It means you are not going to get a
submarine patent surface under you years down the line, as happened to
some developers of GIF reading and writing software when Unisys suddenly
started demanding royalties, retrospectively, on what people had assumed
was unencumbered technology.

Notice that in many cases, members of the SVG working group offered
royalty fee licenses.  This term has various meanings depending on who
uses it. in W3C, it means a RAND (rasonable and non discriminatory)
license which is non-fee-bearing. In other words, even if these
companies have patents which are necessarily infringed by all SVG
implementations, they are no going to charge any fees to anyone who does
not come after *them*. Thus, any patents that those companies hold are
irrelevant *if* you are implementing SVG, For doing anything else, well,
talk to a lawyer we can't help you there but at least for SVG we can
make the situation clear.

This is intended to foster a good, open developer environment and to
allow free software implementation of W3C specifications, which is a
good thing, to continue.

In some cases, members of the W3C SVG WG declared patents that *might*
apply. For example Apple, very early on, declared a patent (basically to
do with a system were there ids a separate alpha channel for each color
chanel, so for RGB that wuld be R, G, B, Ar, Ag, Ab. The SVG WG, without
either confirming or denying the validity of this patent claim, simply
stepped around it and ensured that SVG did not infringe it.

Kodak has a patent, similarly - you know those photo booths that take a
picture of you and when it comes out, ther you are standing besides
mickey mouse? well its a system of combining digital photographs at
fixed positions in a background image with multiple copies and alpha
channels.. and again, it doesn't look like an SVG implementation
necessarily infringes this patent at all. 

So, thats what this is all about. Early disclosure of patents that
*might* affect all SVG implementation, and upfront declaration of what
the licensing terms would be (ideally, royalty free) and declaration of
the actual patent numbers so that developers, and in particular open
source developers who are more exposed because anyone can inspect the
code, can check exactly what is claimed and either avoid it or decide it
doesn't apply to SVG.

This, I hope you will agree, is a lot better than some company five
years down the road demanding back royalties on every copy of your free
SVG software....