Urgent: W3C Patent Policy Issue
Adam Warner
I have prepared a document dicussing the W3C's Patent Policy Framework
that is available on my web site at:
http://www.openphd.net/W3C_Patent_Policy/draft.xhtml
(Plain/print version)
and
http://www.openphd.net/W3C_Patent_Policy
(Web site version)
I have also included the document below in the event that my web site
becomes inaccessible.
Please give the issues I have raised careful consideration and provide
the W3C with a comment by Sunday if you have concerns about this issue
(by emailing www-patentpo...@w3.org).
*** Everyone, this is not an operating system advocacy post. There is
no reason Linux and Windows users cannot unite to protect royalty-free
web standards ***
Regards,
Adam Warner
W3C and the Promotion of Fee-based Standards for the Web
Adam Warner
On 16 August 2001 the W3C made public a proposal to substantially change
their patent policy framework. Amongst the changes is support for a new
licensing model (called RAND) that legitimises the W3C's role in
developing and promoting standards that could require the payment of
royalties.
This is a substantial shift in the philosophical direction of the W3C
and should be of extreme concern to anyone who values being able to
implement W3C standards in a royalty-free manner. In particular this has
profound implications for the support and implementation of future W3C
standards by the free software community. It is likely to extinguish
free software development and deployment in the areas where the payment
of royalties is required.
The last call review period closes on 30 September 2001 (two days from
the time I am writing this abstract). The W3C is aware of the importance
of this issue and states "As the policy has ramifications on the Web
community at large, and as the Web Community have consistently helped
W3C in its efforts, views from this diverse community are essential."[1]
However, as evidence of how well this issue has been publicised, only
two relevant public comments have been made to the W3C archive to date.
It is a matter of urgency that you make your views known. A final policy
is expected from the W3C by February 2002.
Please email all comments or suggested corrections to this document to
com...@openphd.net.
This draft is copyright Adam Warner, 28 September 2001. It may be
distributed freely.
Table of Contents
An Overview of the W3C
W3C Recommendation Process
RAND Licensing
Legitimising RAND
Back-door RAND
RAND in Action
What You Can Do
Essential Reference
An Overview of the W3C
The World Wide Web Consortium (W3C, http://www.w3.org) has been highly
successful to date in its pursuit of "leading the Web to its full
potential". It actively promotes vendor neutral open and universal
standards. Its membership is to be commended for its ability to achieve
consensus and coordination with other standards bodies and consortia.
The W3C has over 500 member organisations and approximately 66 full-time
employees. Even large and influential companies have only one vote at
the Advisory Committee level.
Tim Berners-Lee, the Director of the W3C
(http://www.w3.org/People/Berners-Lee) is also the inventor of the World
Wide Web. The W3C is a distinguished organisation producing quality
specifications, guidelines, software and validation tools.
The W3C is involved in these important areas:
* The architecture domain (e.g. DOM, the Document Object Model).
* Document formats (e.g. HTML, mathematics and graphics).
* Interaction (e.g. multimedia).
* Technological and societal issues (e.g. privacy, encryption and
the legal issues).
* Web accessibility initiatives (e.g. for user agents and authoring
tools).
Crucially the work of the W3C is available to all.
The W3C has an ongoing role in the development of the World Wide Web
from purely static document hosting to dynamic documents, application
services and automated applications.
W3C Recommendation Process
The W3C recommendation process typically follows a five step procedure:
* Interested parties submit notes to the W3C.
* A working draft is produced (these typically come with big
disclaimers, and their citing as anything other than work in progress is
inappropriate).
* Candidate recommendations are made.
* A recommendation is proposed (this means the working group has
reached consensus and the work has been proposed by the Director to the
Advisory Committee for review).
* Recommendation. These have been ratified and can be relied upon to
not change.
The W3C's Patent Policy Framework is at the Working Draft stage. The
Working Draft plainly states: "This Last Call period will be the only
opportunity for public comment."[2]
Remember that "The Last Call period closes 30 September 2001."
Furthermore, "As we have begun to use portions of the policy in the
day-to-day operations of W3C, we plan to skip the Candidate
Recommendation and move directly to an Advisory Committee Review of a
Proposed Recommendation draft." Later in this article I will show some
of the consequences of this in the release of the Scalable Vector
Graphics (SVG 1.0) Recommendation.
RAND Licensing
This is the new licensing model the W3C is proposing that will allow for
non-royalty-free standards to become W3C sanctioned recommendations.
RAND stands for "Reasonable And Non-Discriminatory" terms. "RAND means
that someone may or may not need to pay a fee, and that it is at the
discretion of the license holder."[3]
In essence it requires that any company that imposes licensing
restriction must impose those restrictions uniformly (the
non-discriminatory part of the definition). It appears to follow that
non-commercial organisations cannot be given any preferential treatment
over commercial organisations since that would be discriminatory
licensing.
The Working Draft (http://www.w3.org/TR/patent-policy/) (reproduced in
the Patent Policy Frequently Asked Questions,
http://www.w3.org/2001/08/16-PP-FAQ) also states that RAND allows for
licensing audits (RAND "may include reasonable, customary terms relating
to operation or maintenance of the license relationship such as the
following: audit (when relevant to fees), choice of law, and dispute
resolution.")
Legitimising RAND
The W3C states that "Recommendations addressing higher-level services
may be appropriate for licensing on reasonable and non-discriminatory
(RAND) terms." It is clear that "patent processes will increasingly
affect the Web. These factors make it clear that the W3C must have an
effective policy to address the inevitable increase in patent issues
that will come before the W3C Membership and the development community
as a whole." What isn't clear is that the appropriate response is for
the W3C to condone RAND licensing terms and to actively promote non-free
licenses.
As part of the theoretical underpinning of this new policy we are told:
"On the other hand, there are other technologies, typically higher
level, where it might be appropriate to accept fee-bearing requirements
in a Recommendation. It is worth restating that, as of today, W3C is not
aware of any fee-based license required for any of its Recommendations.
Thus, there is an established history of RF [Royalty Free]."
This distinction between lower and higher level technologies appears to
be somewhat arbitrary and misleading. Any technology that becomes
sufficiently used on the World Wide Web will become a part of everyday
infrastructure. For example it might be considered that a moving picture
format is sufficiently high level for RAND licensing to be appropriate.
But if that moving picture format becomes an integrated baseline
technology in future products then the chance of a future fee being
associated with that technology could be devastating.[4]
The W3C has recognised the pressures from (some of) its members to be
able to exploit the potentially lucrative Internet-related patents they
have been accumulating. There appears to be a resignation that it may be
better for the W3C to promote standards that have non-free conditions
attached rather than to receive no consensus on potential
recommendations.
However by doing this the W3C is diminishing the significant tool they
have to encourage royalty-free licensing: their official stamp of
approval on Internet technologies and credit to the companies that
provide those technologies. The support of the W3C is an important
factor for a web-based standard to achieve dominance. A company might be
willing to provide their intellectual property on royalty-free terms to
receive W3C approval and thus an increasing chance for their sponsored
standard to become widely adopted. Now those same companies may think
they can get the best of both worlds: A W3C recommendation and the
reserved right to charge licensing fees in the future.
The prospect of future fees could also have a chilling effect upon
free/open source software development. Standards that require licensing
fees to implement are, for obvious reasons, totally incompatible with
the use of free software. If the free/open source software communities
will not be able to rely upon the W3C to pursue royalty-free standards
the question has to be raised whether the support of a new institution
is appropriate. Given my admiration for all the W3C has contributed to
the development of the World Wide Web this would be a tragic
development.
Back-door RAND
If an Advisory Committee Representative to the W3C (each member
organisation of the W3C has an ACR) fails to respond to requests for
patent disclosures by default "they will commit their Member company to
license all Essential Claims needed to implement W3C recommendations on
at least RAND terms. This is true whether any personnel from the Member
company participates in a WG or not."
This means oversight, negligence or perhaps deception is rewarded by
requiring the commitment to a RAND license rather than a royalty-free
one. If a relevant patent was disclosed at the appropriate time it might
have been worked around, or the working group may have even disbanded.
For members to face a financial incentive to disclose there should be a
deterrent in the form of royalty-free licensing. Few things would be
more lucrative than being entitled to charge RAND fees on an established
W3C web standard though a simple oversight.
RAND in Action
Even though RAND is only a Working Draft and public comment has for the
first time been solicited (and very shortly closes) the W3C has already
begun using RAND in its day-to-day operations. This can be seen in the
recently released Scalable Vector Graphics Standard (SVG 1.0):
http://www.w3.org/2001/07/SVG10-IPR-statements.html
Apple, IBM, Eastman Kodak and Quark have all only been willing to supply
their intellectual property or potential future intellectual property
under RAND licensing terms. This means that in the event that one of
their patents overlap the SVG specification they have reserved the right
to start charging royalties or set other licensing restrictions upon a
non-discriminatory basis.
Presently the SVG specification is free to use. The uncovering of a
favourable patent or a legal reinterpretation could change that. For
example it is stated: "Kodak does not believe it currently has any
essential claims that fall within the specification of the
Recommendation as currently understood and interpreted by Kodak for
implementors of SVG. However, Kodak hereby identifies U.S. Patent
5,459,819 and affirms that in the event that any claim of this patent is
interpreted as an essential claim within the specification of the
Recommendation in its current or later amended form, Kodak agrees to
provide a RAND License as set forth in the previous paragraph."
The significant change here is that Kodak (as a particular example) are
not giving standards users an assurance that they will be able to
continue to use SVG on a royalty-free basis in the future. A windfall
judgement and we could have a problem of GIF-style proportions--even
though SVG is a W3C sanctioned standard and the company potentially
doing the enforcing helped create the standard for people to freely use
in the first place. Users could feel far more secure that SVG will
remain a free standard if for example Kodak said that in the event that
any claim of their patent is interpreted as an essential claim within
the specification of the Recommendation in its current or later amended
form, Kodak agrees to provide a royalty-free license.
What You Can Do
1. As a matter of urgency, send a comment to
www-patentpo...@w3.org before 30 September 2001. You can check
out the current archive of responses here:
http://lists.w3.org/Archives/Public/www-patentpolicy-comment/.
2. Spread the word about this issue as soon as possible.
3. Ask companies that are members of the W3C to give an undertaking
to only support the development of royalty-free standards. This will
require significant change to the Working Draft of the W3C's Patent
Policy Framework.
4. Do you have a professional relationship with any of the authors or
companies of the Working Draft? If so it may be appropriate to send a
message to the relevant organisation. These are the listed authors of
the Working Draft:
* Michele Herman, Microsoft, mich...@microsoft.com
* Scott Peterson, Hewlett-Packard, scott_k_...@hp.com
* Tony Piotrowski, Philips, tony.pi...@philips.com
* Barry Rein, Pennie & Edmonds (for W3C), ba...@pennie.com
* Daniel Weitzner, W3C/MIT, djwei...@w3.org
* Helene Plotka Workman, Apple Computer, plo...@apple.com
It is also stated here: http://www.w3.org/2001/08/patentnews that
"W3C Members Apple, AT&T, Hewlett-Packard, IBM, ILOG, Microsoft, Nortel
Networks, The Open Group, Philips Electronics, Reuters, and Sun worked
on this draft together with W3C Team members."
Essential Reference
W3C Patent Policy Framework, W3C Working Draft 16 August 2001:
http://www.w3.org/TR/patent-policy/
Backgrounder for W3C Patent Policy Framework:
http://www.w3.org/2001/08/patentnews
Patent Policy Frequently Asked Questions (FAQs):
http://www.w3.org/2001/08/16-PP-FAQ
SVG 1.0 Patent Statements:
http://www.w3.org/2001/07/SVG10-IPR-statements.html
Notes
[1] http://www.w3.org/2001/08/patentnews
[2] http://www.w3.org/TR/patent-policy/
[3] http://www.w3.org/2001/08/patentnews
[4] For example, a scenario where the majority of future web appliances
included this decoding ability in their ROM.
Rex Ballard
While it is true that Tim Berners Lee did successfully implement the
earliest
HTML web Browser using the URL notation, this was not entirely his
idea, nor
was this single innovation significantly responsible for the success
of what
we now know as the Internet.
The technology was established, designed, and even implemented via
discussions
on usenet newsgroups. Furthermore, the standards chosen were
diseminated and
distributed BECAUSE it was General Public License and Patent Free. In
fact,
SSL was initially rejected until Netscape agreed to put the source
code for
SSL into GPL.
Thousands, perhaps even millions of people worked for the success of
open source
software. Keep in mind that most of the earliest servers were based
on GPL software
(Linux, CERN, HTTPD, Apache) and GPL browsers Mosaic, Arena), and were
conducted
under the public trust.
Furthermore, most of the implementations were committed under the
General Public
License. In addition, when Unisys started demanding royalties for
GIF, the web
developers pushed the GPL implementations of PNG and JPEG.
If IBM, Kodak, and others want to try and promote proprietary
standards, they are
free to do so. Some companies such as Adobe, have tried, with limited
success,
to promote PDF document standards through the distribution of free
browser "readers",
Macromedia FLASH has also been promoted with limited success.
Microsoft tried to push ActiveX, Microsoft Media, Word, Excel, and
Powerpoint formats,
other exclusive technologies, in hopes of getting exclusive control of
the Web. For
the most part, this effort has been unsuccessful.
Unfortunately, the W3C has now begun attempting to promote proprietary
licenses, without
disclosing their contents
(http://www.w3.org/2001/07/SVG10-IPR-statements.html follow the links
if you can).
Those who are attempting to intermingle patented and royalty based
proprietary protocol into
the W3C standards, are attempting to get a "Free Ride" on the backs of
the hundreds of thousands
who have given freely.
At one point, was working with a number of Newspaper publishers who
wanted to put their content
on the Internet. I showed them how to create the earliest commercial
Web servers on Linux servers.
I showed them how to price advertizing, how to provide accurate
demographics, and how to create
a multi-billion dollar industry. I asked that rather than paying me
1/10th of 1 percent, that they
contrute that amount to the Free Software Foundation and other
organizations who provided Open Source software, such as W3C. If they
had done so, I suppose that makes me the biggest philantripist
in history. In fact, many of these publishers eventually returned
technology, promotional efforts,
and other non-monitary support worth billions into this technology
base.
There are legitimate uses of proprietary technology. Encrytion
techniques used for electronic
commerce requires idemnification, security, and compliance with
national and international
Linux users. IBM has successfully used MQSeries as a means to conduct
electronic commerce
ranging from communication between banks to "Electronic Brokerage
Exchanges" for the trading
of various goods and services.
Keep in mind that IBM had pioneered GML, Generalized Markup Language,
several years before
the creation of the Web. But it was only when they submitted SGML to
public standarization
and implementation under GPL licensed software under the Project
Athena Andrew Project
that SGML became widely accepted as a viable exchange. HTML was a
very limited, but effective
implementation of a very specific SGML DTD. XML resolved the issues
of dealing with the interrelationship between the DTD, the Content,
and Style Sheet used for converting
the body into viewable form.
Had IBM attempted to retain GML licensing and patent rights, it's very
probable that
standards based on EMACS such as Info-TeX would have been adopted
instead. Dow Jones
was working on a standard which was an "Object Oriented RPC".
Eventually, many of these
technologies were incorporated into CORBA, RMI, and XML/SOAP.
Perhapse Dow Jones should
be paid for their intellectual property?
But what if Dow Jones hadn't shared their technology with the other
publishers? We would
have probably switched to CORBA based on MICO, a GPL implementation.
Had Sun demanded
royalties for all Java and RMI use, the industry would probably have
adopted Linux,
PERL, and Python, which actually has dominated the world of Web
Servers.
The standards implement in the World Wide Web required the agreement
of numerous forces,
including several thousand corporations. Publishers, Web site
managers, telecommunications
companies, Points of Presence (PoPs) including thousands of FIDO and
Wildcat BBS managers
had to agree to a set of standards which would give no one excessive
control. Many companies
tried to promote proprietary alternatives. IBM, DEC, and HP tried to
promote ISO/OSI, AT&T
tried to promote ATM, MCI and GTE tried to promote Frame Relay all the
way to the NIC card.
TCP/IP won out, not because it was the best protocol, but because it
was available in numerous
GPL and Open Source forms. Net3 from Becker was GPL, the BSD version
was Open Source but
could be integrated to proprietary kernels, KA9Q was available for
PCs, and Trumpet Winsock
was available for rediculously cheap.
Sun tried to promote Adobe's NeWS presentation protocol, which was
based on PostScript. But
the market opted instead for standards that had been adopted by the
X11R4 and R5 distributions,
including GIF, JPEG, PNG, and SGML/HTML.
Kodak actually does have a possible legitimate claim. Kodak pioneered
image management technology
in their KIMS (Kodak Image Management System). They created their own
proprietary technology,
but the overall industry opted for JPEG instead because it wasn't
encumbered with copyright
and patent royalty expectations.
The GIF standard was almost rejected outright in favor of the PNG
standard. It was only
when Red Hat purchased and bundled the rights to GIF technology with
each commercial copy
of Red Hat Linux that the GIF standard prevailed. Eventually, the
Debian group, who refused
to pay the royalties, made an issue of GIF royalties and pushed for
the adoption of PNG.
Soon there were so many sites using PNG that all browsers were
supporting it.
At the same time that this complex network of agreement was being
woven into the World Wide Web,
Microsoft was trying equally hard to push it's own standards. They
tried to push NetBUI over
proprietary dial-up lines. The servers would all be Microsoft
Exchange servers, and all
documents would be stored on Microsoft's servers (where all content
could be monitored
by Microsoft). And Finally, every user of the "Microsoft Network"
would only be able to
view documents in Microsoft format. Compression was handled by
putting the compression
technology into the hard drive (Stacker and later DriveSpace), and
using MNP for slower
modems.
How quickly would the big server makers like Sun, IBM, HP, and DEC
align and agree to give
Microsoft a monopoly on the only access to on-line users (network or
dial-up).
Now, the W3C has attempted to "Sell out to the highest bidder",
offering to include
proprietary technologies as mandatory components of the web, to the
exclusion of
other non-proprietary technologies.
Unfortunately, there has been a sudden rash of attempts to patent
well-known and
well documented internet technology simply because no one has ever
tried to
patent it before. They are hoping to grab a few $billion before the
patents
are nullified by archives of Usenet news.
For those who don't remember, Microsoft tried to patent the use of a
spell-checker in a word processor. Appearantly, they didn't think it
would matter that Richard Stallman had introduced this feature into
EMACS
nearly 2 years later.
Prior to the W3C, the IETF set all of the internet standards. The
IETF insisted
that the specifications be implementable, which meant sufficiently
detailed that
a third year CS student could implement a GPL version on a UNIX system
using only
the available specification. If the student needed to add details,
those details
were added to the specification.
When the IETF started comprimising it's principles to accomodate the
wishes
of IBM, Novell, and Microsoft, the W3C began to dominate as the
standards
body incorruptable.
Today, Linux compatibility seems to be the most widely accepted
standard.
When RealNetworks wanted to establish their technologies, they
provided
Linux ports. Before long, the Linux community had accepted
RealNetworks
as a viable standards. Several others have offered both Open Source
and
closed source software for Linux. This has reduced resistance to these
products.
Of course, the flip side of this is that Linux can often trivially
implement
those technologies which are offerend in proprietary form. For
example,
Linux had the ability to server and recieve streaming MPEG video
without
help from Real Networks. Linux was also one of the first servers to
offer
MP3 audio.
I understand the desire of corporations to try and promote their
proprietary
technology, and to prevent the use of GPL software to implement
competing
technology. The problem here is that most of these proprietary
technologies
are simply more "eligent" (proprietary) implementations of technology
which
has often been available, but not mass-marketed, in Open Source, and
has been
discussed, designed, and claimed in usenet discussions.
The only question is whether a company is willing to risk class-action
lawsuits
when they try to patent ideas that have been developed, marketed,and
popularized
under the terms of Open Source licenses.