Lex Spoon <lex at cc.gatech.edu> wrote:
> One way we can reach some sort of consensus, is to edit this page until
> it reflects something we are all satisfied with.
That sounds like a good idea. 

> I've added a section on "who owns the code", in response to recent
> discussion.  It is clear that Apple is the main owner to worry about at
> this stage; everyone else is likely to be extremely agreeable about any
> license proposal that the rest of us agree on.
Sorry, but unless SqC's contracts at Disney said specifically that
copyright is retained by the individual, it is owned by the company.
There is nothing vague about this part of the law. I'm fixing this.

[Lex Spoon sort of likes the proposal]
Maybe you could relicense some of the below as SqueakL/MIT?

(SMSqueakMap default cards values select: [:e | '*Lex*' match: e
author]) collect: [:e | {e. e categories detect: [:c | '*MIT*' match: c
name] ifNone: ['not mit']}] 

(SMCard[Large Lists] 'not mit') 
(SMCard[ProgrammingMorphs] 'not mit') 
(SMCard[BehavioralInspector] 'not mit') 
(SMCard[Filtering Celeste] 'not mit') 
(SMCard[Scheme] 'not mit') 
(SMCard[Nebraska] 'not mit') 
(SMCard[Screen Shot Morph] 'not mit') 
(SMCard[Scamper] 'not mit'))


> I am not completely clear on how the license bookkeeping will work.  For
> most things we can track the license that SqueakMap claims to apply, but
> are there things other than packages we need to worry about?
Yup, but packages are where we can make the most difference, the most
quickly. Stuff in the image has to be traced to see who first licensed
it as SqueakL, and then talk to him and all the other contributors. VM
support/plugin code is up to the VM contributors to relicense, most of
us are not involved.

> Also, I'm still not clear that MIT is the best ultimate goal.  Is it
> truly better than APSL?  
MIT doesn't limit sharing. With APSL, once you're thinking about using
it in conjunction with another non-trivial license, you need to start
thinking about interactions. So MIT is a better ultimate goal than APSL,
IMO. Of course, "no-interactions" and "strongly encourages some code to
be returned" are probably mutually exclusive features. I happen to think
that "no-interactions" is a more important feature for code to be shared
for long periods of time, and this is the main reason I think we won't
do better than MIT. I added links to the licenses text.

> Is it enough better that it worth rewriting
> portions of Squeak that otherwise would get left alone?
Definitely not - I think nobody should rewrite anything just to change
the license. If you happen to write something that supercedes an
existing subsystem, however, you might as well choose its license
carefully, not inherit what is given. 

Daniel