Not to open a can of worms here ... but it would be nice to have a 
better distinction between that which should be shared and that which 
can be owned.

My own notions are probably hopelessly idealistic, since they were 
formed during the sixties and the wonderful public domain support of 
ARPA and then Xerox PARC. This was basically a patronage model in 
which the Medicis supported the artists to create what the artists 
needed to create and the resultant art belonged to the public. 
For-profit companies formed all the time from this process but didn't 
damage it. This was overwhelmingly successful but has not been taken 
up in current times by governments or companies, primarily, I 
believe, for "being out of control" reasons.

Today, the patrons don't exist on the necessary scale to make this 
work -- so everyone is struggling. If we look to the history of 
Science as a model, we find that open sharing and preservation of 
citation have been key to its success. Distinctions have been made 
between basic knowledge, which is to be shared, and specific 
invention, which can be owned for a time. This has always been a bit 
messy, and large corporations have usually had the advantage 
regardless of the merits of the case.

It's not clear quite how to do this with computer stuff, especially 
in an image model employing objects that are *all* instances of 
subclasses of base classes. Those of us who invented and made Squeak 
decided to share all, and not try to make any direct money from any 
part of Squeak. For those who want to own parts of their creations 
and to sell them without being ripped off, I say "Yes, you should be 
able to do this". But I also say, if you come up with some new 
principle or mechanism that advances the state of software 
engineering and/or computer science, you really owe it to the larger 
world community to find a way to share the new idea. And the world 
community needs to find ways to protect the citations of the 
contributors (as Ted Nelson pointed out long ago) through use and 
reuse of these ideas.

I think part of the solution path here is to deal with Ted's point 
that in the hyperlinked world of the future it is far more important 
to protect citation than ownership. He proposed a variety of ways to 
do this, and some of today's capability systems (like e) could be a 
nice start on that route.

Cheers,

Alan

-------

At 8:45 AM -0500 2/20/03, Jon Hylands wrote:
>  >At 8:47 AM +0100 2/20/03, goran.hultgren at bluefish.se wrote:
>
>>>- Making sure how we want it: Should improvements be published
>>>regardless of distribution? AFAICT Andrew got to this conclusion given
>>>the current language. (Which btw probably means some of us are
>breaking
>>>the license currently - do you have improvements that you haven't
>>>pubished?)
>
>This was one of the things we talked about at Interval Research.
>Glen Eden's interpretation of that clause was that we had to publish
>improvements to the base classes, but the license doesn't say anything
>about *when*...
>
>Later,
>Jon
>
>--------------------------------------------------------------
>    Jon Hylands      Jon at huv.com      http://www.huv.com/jon
>
>   Project: Micro Seeker (Micro Autonomous Underwater Vehicle)
>            http://www.huv.com


--