On 19 Oct 2004, Paul Vixie wrote:
I've been around for decades and know at least a few people who have
written more than 10,000 more or less freely redistributable and popular
lines. Few of them are members of either your or Mr. Stallman's
organizations, although both of you frequently claim to speak for us.
I suspect that if one were to count lines of source or number of packages,
you would find the majority of "open source" code to be under GPL. The
suggestion Mr. Schryver makes above is pretty plainly wrong. If open
source were a parliamentary system, Stallman and the FSF would be the
prime minister and ruling party. He represents a lot of people.
But of course, the definition of "open source" is itself somewhat vague. I
would not, for example, count Paul Vixie or ISC as an "open source"
proponent; because of his support software patents, his advocacy of the
BSD copyright which allows persons to take this code private if it suits
them, and because Vixie/ISC trademarked other people's software (BIND).
This is all pretty antithetical to the open source goals and mindset as I
As a short aside: The BSD copyright wasn't made for "open source", but for
allowing the public to exploit and commercialize the University of
California's government sponsored work. BSD code was originally not open
(though essentially free) due to the inclusion of ATT code. It just
wasn't necessary to pay Berkeley, except a nominal fee for the 9track
tapes. I suspect that if the BSD code wasn't DARPA/HP/DEC (and later OSF)
funded, it would very likely have had a copyright more similar to other
Univerity developed code like AFS. I don't mean to ignore the free
software disposition of people at CSRG, but developers rarely have control
of such decisions. Most universities have technology licensing
departments to obtain money for University work. The BSD Copyright was
appropriate 15 years ago, for the output of the CSRG and other work.
This codebase certainly assisted the open source movement. But the BSD
copyright isn't really "open source" by many definitions. The frequently
cited problem is the profiteering and re-privatization of derived
proprietary propducts. Some groups haven't liked the infectious nature of
the GPL, but nearly all reject the BSD equivalent copyright.
But such terms as open source are often vague and hard to pin down because
they mean different things to different people. I recall that Sun once
tried to trademark "Open Systems"---if successful this would have turned
things upside-down: making the term "Open Systems" proprietary to Sun!
Of course, at about the same time Scott McNeely also said that Motif would
never be shipped by Sun. That turned out to be wrong, too, in small part
to my credit.
in particular, neither i personally nor isc as a company shares the
views expressed here by mr. raymond with respect to ietf's patent
policies. i think of myself, and of isc and the various people who
have written the software isc publishes, as part of what mr. raymond
calls "the 5%", and i agree with mr. schryver that mr. raymond's
claims of representation don't help me and might, by creating
misunderstanding amongst the consumers of "open source" software, hurt
I think that's probably right: You are part of the 5% not least because
you're pro-patent. Mr. Raymond, Mr. Stallman nor other represent you
because you aren't really proponent of open source. You're welcome to
your point of view, but its substantially different from the open source
point of view.
If preventing you from using open source software as the basis for your
proprietary products harms you (and I suppose it would in your case), then
I can forsee your concerns. I suppose if Eric Raymond promotes the idea
that open source software should be free of patents, impossible to take
private, and free of trademark encumbrances, then I suppose that could
hurt //Vixie// and open source profiteers. However, it would prevent open
source profiteers from harming the open source community. None of this is
purely hypothetical: EG what Vixie did with the BIND trademark was very
similar to the Linux trademark schenanigans a few years ago)
I think most open source proponents would agree that the proper way to
make money on open source is to package it or offer consulting services to
improve it keeping the modifications open source or install it or
administer it, etc. I think most would agree that the wrong way is to try
to use open source software as the basis for a proprietary product, by
making propietary improvements that then aren't shared. Some people
disagree. They are welcome to their opinion. I wish those people wouldn't
try to pass themselves off as open source proponents. I can acknowledge
that being pro-patent/closed source is a point of view, and I can even
make accomodation and compromise to work with those people for the greater
good. And if some think I'm a radical proponent of open source, let me
say that I can even acknowledge that there are presently closed source
products that are superior to open source products. It is just my belief
that this will change over the long term.
Eric Raymond's error is not about the fundamental nature of open source
(at least not what I've seen), but his view that the IETF should itself
promote open source. This might be nice to wish for. Eric's error (if it
could be called that) is just that this is not realizable at present as I
On Wed, 20 Oct 2004, Paul Vixie wrote:
somebody asked me...
What is your position on these issues then?
i think that anyone who comments on the mailing list, or in WG meeting
minutes, or as a draft author, should have to disclose any relevant IPR
of which they are then aware or of which they become subsequently aware,
whether or not such awareness is due to prospective benefit by them, or
their employers, or their heirs or assigns.
The above is current policy.
i also think contributors to ietf specifications, whether verbally, or
in e-mail forums, or as authors, should have to quit-claim any
relevant IPR except that which they have disclosed in advance of a
draft being submitted to the RFC editor.
A verbal quit-claim is not worth the paper its not written on. Such
worthless assurances should be viewed very skeptically. An assurance
needs to be more substantial. For example, the G.723 patent holders
formed a licensing consortium to handle the licensing, which was
affordable to small developers and open to everyone.
i think that the ensuing ietf-isoc-malamud hairball should pay for IPR
searches of all final-drafts before they reach the RFC editor, to get some
kind of reasonable belief that all relevant IPR has in fact been disclosed,
even though no warranties as to IPR should be expressed or implied.
This is not current policy. IPR searches are often a bad idea for
developers. (Yes, there is more than a little irony, here) I had an
opportunity to discuss the issue several years ago with Genuity's patent
attorney (I was on the initial list of inventors on a provisioning system
patent). She recommended strongly against doing searches, since that could
be used against you in an infringement case. The fact of having done a
search would be evidence that you were aware of the patent being infringed
and thus the infringement is intentional. The penalties for intentional
infringement are more severe than for unintentional infringement.
if working groups want a standard to use protected IPR, their only
responsibility is to ensure that all IPR claims are properly disclosed.
if implementors want to build products on a standard that uses protected
IPR, they should be able to read the IPR legend in the RFC and make an
informed business decision as to whether they like what they see.
This is also current policy.
Av8 Internet Prepared to pay a premium for better service?
www.av8.net faster, more reliable, better service
617 344 9000
Ietf mailing list