License committee report for January 2008

[Dag-Erling Smørgrav]

I have been ill, so this comes a little late, especially in view of
the fact that I was informed this morning - in quite polite terms -
that the license has actually been accepted.

However, I can not let your statements stand without rebuttal.

Russ Nelson <nelson at crynwr.com> writes:
> Dag-Erling Smorgrav <des at linpro.no> writes:
> > This is a gross misrepresentation of events,
> Oh?  Let's go through what I said, sentence by sentence, compare it
> against the facts, and see who has a better grasp on reality.  Before
> we get too far, may I recommend that you polish your mirror, because
> you'll be looking at yourself before long.
>
> "The board sent this license back to the submittor because"
> The minutes haven't been published yet, but any board member or
> observer can verify what we took that action.

Correct.

However, you omitted certain important aspects, such as the fact that
you alternately ignored and blocked my application for months before
submitting it to the board, then omitted to inform me or anyone else
of the outcome, then refused to answer my explicit request that you do
so, and angrily remonstrated me for suggesting that this was, in fact,
one of the duties inherent to your position.

During that period, several other and far more controversial licenses
were approved, and both the submitter and the general public were
promptly informed, sometimes within hours of the decision, and without
even so much as a heads-up to license-discuss.

I suspect that eight months from initial submission to final approval
is some sort of record...

The rest of my email is mostly concerned with how you blatantly
ignored the discussion on license-discuss and summarized only your own
opinion rather than the consensus arrived at on the mailing list.

> "the license is identical to the FreeBSD license,"
> wget http://www.freebsd.org/copyright/freebsd-license.html
> wget http://www.des.no/simplified-bsd-license.html
> wdiff freebsd-license.html simplified-bsd-license.html

This has been thoroughly refuted in the discussion on license-discuss.
Very briefly, the FreeBSD Project uses a number of different licenses,
one of which happens to be a two-clause variant of the BSD license,
which is used on a relatively small portion of the source code.  It
also uses a very similar but not entirely identical license as its
"compilation license" (which covers the collection of individual parts
comprising FreeBSD, but not the individual parts in themselves).  The
FreeBSD Project does not refer to it as "The FreeBSD License", though
it is published online at a URL which includes the character sequence
"freebsd-license", which - as discussed in further deail below - has
no practical significance.

To pick one random example, I doubt that anyone would argue that the
"Computer Associates Trusted Open Source License 1.1" is properly
named "The CA-TOSL1.1", just because that is the name of the PHP
document in which it is described on the OSI web site.

> "yet the FreeBSD license is more well-known."
> Google reports:
>   284 for "simplified BSD license".
> 9,030 for "freebsd license".

As was pointed out in the discussion, all but a few of these 9,030
hits were false positives, including coincidential use of the words
"FreeBSD" and "license" in sentences such as "Sun revoked the FreeBSD
license for JDK" or "What are the FreeBSD license terms?".  It is also
important to distinguish between "FreeBSD license" (lower-case 'l')
and "FreeBSD License" (capital 'L').  The former can mean anything you
want it to (e.g. "the current FreeBSD license", the license which is
currently used by the FreeBSD Project); only the latter specifically
refers to "the license written and used by the FreeBSD project".

I suspect that subsequent to these discussions, the number of genuine
Google hits for "The FreeBSD License" increased by several orders of
magnitude due to the numerous online archives of the license-discuss
mailing list.

> "The submittor has refused to change the name of his submission:"
> http://crynwr.com/cgi-bin/ezmlm.cgi?3:14760

The proposed alternatives were unacceptable for reasons which have
been extensively discussed.  Most importantly, whichever alternative
was chosen would have alienated a majority of the people already using
the license.

The particular version I submitted did not originate with the FreeBSD
Project.  I first became aware of it while working on PAM support for
OpenSSH in 2001; it was at the time, and to the best of my knowledge
still is, the OpenBSD Project's preferred license for all new code,
and I was therefore asked to relicense by PAM support code under the
two-clause variant.

I have seen it mentioned informally as the "ISC/OpenBSD License" (e.g.
in the Nabble archive of a discussion regarding categorization of a
new project in Savannah), but *never* by the OpenBSD project itself.

It is also used by NetBSD, Dragonfly BSD and other BSD derivatives,
inasmuch as all BSD derivatives regularly borrow code from eachother.

I have also seen it used in projects completely unrelated to *BSD.

It is used by the Varnish HTTP Accelerator, which was the reason for
my original submission; it was selected by the organization that
initiated it, convinced the company I work at to write it, and funded
the first round of development.

I use it myself for all new code I write on my own time.

So what do you have?  You have a license which is used by dozens of
related and unrelated projects, some of which are bitter enemies.  You
have a license which has not been formally named by any of its users.

Having taken the initiative to submit it for OSI approval, I needed to
come up with a name for it.  I chose the most neutral and descriptive
name I could think of.  At no time did anyone suggest an alternative
that would be both descriptive and non-partisan.  You least of all, I
might add.

> > > Recommend: Modifying the New BSD License template so that the third
> > >   clause is optional.  In this manner, both the FreeBSD license and
> > >   the Dag-Erling's Simplified BSD License will fit into the template
> > >   with no license proliferation problems.
> > and your "recommendation" has not been discussed on the lists.

Your choice of words - specifically , "the Dag-Erling's Simplified BSD
License" - is deliberately hostile and does not belong in text which
is meant to assist the board in their decision.

> > Perhaps it is time for the board to consider replacing Russ Nelson in
> > this role, as was done for the list moderator role.
> Mmmmm, I'm still postmaster, and I still moderate the lists.  We just
> decided that Ernie has more patience with morons.

A more detached observer would say that you used grossly inappropriate
language in an ad-hominem attack against a person who was making
reasonable arguments against the position you were taking in the
discussion, and that you subsequently abused your moderator privileges
to prevent him from defending his person or his opinion.

DES
-- 
Dag-Erling Smørgrav
Senior Software Developer
Linpro AS - www.linpro.no