Intent of the Squeak License

Alan Kay Alan.Kay at squeakland.org
Tue Jan 13 08:05:25 UTC 2004


Hi Folks --

The intent of the original Squeak group was to make Squeak available 
in a manner that was as analogous to scientific knowledge as 
possible. That is, to make it a moral obligation to share basic 
advances in software engineering, while allowing ownership of 
specific content, including the right to sell one's own content. This 
means that a difficult line has to be drawn between "fundamental" 
knowledge, which is best shared, and "organized forms" that could be 
ownable products.

It is possible for humans to mess up any set of good intentions (and 
quite a bit of this has happened to intellectual property in the 
sciences over the last 50 years). But, as has so often been pointed 
out over the years, that simply trying to write down any kind of 
principles is not enough. One of the most often used examples is the 
American Constitution. It's important that it is written down, but 
the citizenry has to both understand the intent and what "good faith" 
means in order for the system to work. Both the intent and the good 
faith have been under attack many times, including now.

I think the most important part of our intent that didn't get handled 
legally (I'm not sure what we should have done) is the distinction 
between citation and copyright of code. We did have every intention 
that contributors to the Squeak code have attribution and citation 
(because that is the nature of reward in science beyond the basic 
pleasure of finding things out). We had *no* intention that 
organizations such as Apple or Disney or HP etc., or individuals, 
should wind up with copyrights or any other form of ownership of any 
of the shared code. (Ted Nelson wrote quite a bit about these 
distinctions, and especially why it is important to preserve the 
provenance of contribution while allowing sharing and use.)

The nature of programming (especially with late-bound systems like 
Squeak) makes it difficult to draw the distinctions that I think are 
important here. But we did so at Disney. We told Disney to pretend 
that Squeak came from a vendor (like Director coming from 
Macromedia). Here it is quite clear what is "basic" and what is 
"content". We left quite a bit of this Squeak content behind when we 
left Disney, because we believed that it was the kind of stuff that 
was reasonably ownable. The tricky part (but not to us) was to decide 
what new things done to Squeak counted as fundamental and sharable 
and what counts as ownable content. For example, a video game using 
B2D or B3D was ownable content, while B2D and B3D counted as 
fundamental and sharable.

So our intent was that, once we had decided that code was fundamental 
and to be shared, it would cease to be owned or copyrighted by Apple 
or Disney and would instead belong to the world. To me that is what 
the spirit of Squeak is about, and should be the spirit of free and 
open source software.

Now for the complicated opinion. Regardless of the current practice 
(which to me is all over the map) and current "laws" (which to me are 
poorly drawn and dubious at best), I consider that the fundamental 
parts of Squeak "belong" only to the world and to no smaller entity, 
just as do Maxwell's Equations. The finest thing we can do is to come 
up with new principles that add new powers in this high range and 
give them away to the world. I think that guidelines for "reasonable 
ownable content" can help make the distinctions needed here, but only 
with enough "good faith" to avoid the kinds of legal arguments that 
can confound any attempts at good faith and progress.

Finally, I understand a little bit the concerns about meshing with 
other open source movements (like Debian). But, it seems to me that 
just as "science means you don't have to trust the experts", "open 
source means that you don't have to put other software people in your 
critical path". That is, in the end, it is about controlling one's 
own SW destiny. Since Smalltalk predates all of this stuff, and most 
of the important fundamental principles in Squeak were funded by ARPA 
and PARC, I just can't get too worked up about having Apple and 
Disney in the chain of funders. To us a really important part of the 
Squeak implementation process was to start with a Smalltalk in which 
the IP connection and relationship to Xerox PARC was completely clear 
-- and this was possible because Apple actually was granted real 
ownership of its Smalltalk as a first round outside implementer in 
the early 80s.

The bottom line for me is as I've stated it before: the relevant law 
is yet to be written and the current laws are yet to be tested. This 
means that a litigous society such as the US can cause incredible 
pernicious difficulties and delays over almost any issue, including 
those that are legally well posed. I have seen this many times in IP 
lawsuits over the years in which I was an expert witness. It's a 
mess. So I like the "let sleeping dogs lie" principle for most 
things. As far as change, I think it would be useful to try to make 
the above principles into a new kind of characterization of ideas and 
their usage. This could lead to laws and licenses that might actually 
be of some use.

Cheers,

Alan


-- 



More information about the Squeak-dev mailing list