Still no new license -- but draft text available

[Grant Griffin]

Piet van Oostrum wrote:
> 
> Many people do m\not like GPL but if python had originally been released
> under GPL, CNRI wouldn't have been able to change that, and all this fuss
> wouldn't have been necessary.

But without its generous CWI license, many of the commercial uses of
Python (which have undoubtedly contributed to its success) would not
have been possible.  So it seems to be a tradeoff.

I personally dislike the GPL because it has repeatedly left me unable to
use technically-meritorious and free (!) software.  So I then have to
buy something, or, more often, have to write my own.  That just seems
wasteful.

It amazes me that people who want to give away their work would want to
place restrictions on its use.  If one has a "gift" mentality, the gift
means more if given without strings.  Alternatively, if one has a
"paternal" mentality, one would want to see one's software child be as
widely adopted and loved as possible.  So go figure.

But that being said, the *empirical* evidence seems to indicate that the
open/free aspect of software is much more important than whether (or
not) it is copylefted.  If you consider the success of Python (with its
generous CWI license), Perl (with its generous Artistic License--and the
Grinchly GPL available as an alternative), and the Linux Sysem (not
"GNU/Linux System"--na-nya-nana-nay Mr. Stallman! ;-) available only in
GPL, it all seems to add up to this: the presence or absence of copyleft
probably just doesn't matter.

As evidence of the merit of the GPL, we hear the example of Next having
to contribute Objective C so they could use the rest of the gcc
compiler.  This example is notable for its singularity: I haven't heard
others.  Also, notice that Objective C never really caught on.  And Next
itself is dead (I think).  So at best, there seems to be very little
empirical evidence that copyleft thing actually accomplishes its goal of
making otherwise closed/commercial software become open/free.

If Open Source has intrinsic economic value, the copyleft concept isn't
needed; if it doesn't, copyleft won't be enough to make Open Source a
force.  In other words, you don't need to require people to make their
improvements to software open/free if the openness and freeness of those
improvements is truly economically beneficial to them--they'll do that
without being forced; likewise, if you _try_ to force them but they find
no advanatage in it, it probably won't work; they'll just go elsewhere,
as I have done many times.

Likewise, one can find no obvious correlation between the copyleftness
of a license and the amount of contributions a given package receive. 
Python is a good example of Open Source software that receives
considerable contributions without a copylefted license.  Ditto Perl.

Personally, I think of the whole copyleft thing as being primarily a
clever marketing gimmick (whether or not Richard Stallman realizes it.) 
It definitely has a strong appeal to a certain segment of the
population--especially young people.  But then again, it turns off
another segment.  So, from strictly a marketing point of view, it's a
tradeoff.

Notice, though, that copyleft seems to thrive primarily in cases where
somebody is *re-implementing* an existing system--most noticably the
GNU/Linux System (OK, it's probably about time I threw His Royal Root of
All Square Meanness a bone ;-).  Specifically, the appeal of a
re-implemented free system is primarily that it is _free_.  (Or, as
Linus Torvalds has said, the problem with UNIX was that it was "so
expensive".)

But if you're trying to popularize a _new_ system (as Guido and Larry
Wall once were--and still are to an extent), it makes a lot more sense
not to limit the ecologies that your system can live in (by putting in
license terms that discourage commercial uses); broad usage terms give
it a better chance of thriving and prospering.  Put another way,
packages which have a copyleft license restriction are at a competitive
disadvantage compared to those who don't.  (Even within the world of
free/open software, competition is the rule.)

But I think that the success of a re-implemented system like GNU/Linux
can be entirely explained by the fact that it is free, open, and very
high quality--in other words, by the merits of the product itself.  The
fact that it is copylefted is incidental--except that perhaps Richard
Stallman would never have applied his considerable drive and technical
talents to helping create it if he wasn't able to create a legal
mechanism to enforce his strange obsession with making it difficult for
people to make money on the work he "gives" away.

(which-reminds-me:-i-bet-that-red-hat-ipo-thing-last-year-must-have
   -really-burned-his-butt-<wink>)-ly y'rs,

=g2
-- 
_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation	      http://www.iowegian.com