Subject: Re: Proprietary but flexible licenses.
From: Danese Cooper <Danese.Cooper@Sun.COM>
Date: Mon, 7 May 2001 17:23:12 -0700

Well, as the former SCSL Goddess, I feel I must comment on this thread ;-)

The SCSL has gone through various incarnations.  What angered the Free Software 
and Open Source communities was first the assertion that SCSL was 
somehow an improvement on already accepted licenses and then the implementation 
of Java source licensing, which uses SCSL.  Even though there is that word 
"community" in the SCSL, there never was a significant initiative to actually 
create and sustain a Java community (eg. accept bug fixes, release code more often 
than once every half-year, answer technical questions about the source sent in 
by mere mortals...).  All of this is why I wasn't the SCSL Goddess for long.  
I wanted to work on community, which Java's SCSL implementation was not.

The license itself isn't inherently evil, however. The SCSL was in fact created to 
distribute Jini source code, and there is today a community built around Jini that 
has repeatedly voted against the statement that SCSL is not meeting their needs.
In the case of Jini, Sun worked really hard to create community.  They held wonderful

community meetings in America and Europe.  Ken Arnold and Jim Waldo actually engaged

with community members on community mail lists.  They answered questions and solicited

feedback.  They welcomed and acknowledged contributors.  They allowed the community

to govern itself and focused on the technology instead.  They allowed zero-cost access

to the sources, and eventually even made the Jini brand (and hence the compatibility

requirement) optional (and zero-cost as well).  None of this was hampered by the SCSL.

SCSL has also undergone some rework over the last two years, and the most 
recent version is significantly shorter, easier to understand, and more friendly 
to researchers.  Anyone who is willing to agree to the license terms can download 
source code and play with it for research purposes.  They can even share the code 
with their modifications to other researchers who have also agreed to the SCSL (eg 
those who are members of the community).  Only if they plan to make "productive use"

of the code or of a larger work that includes some or all of the original code must

they first contact the original donor (Sun) for a "commercial use" agreement (which

*may* include royalties, but in the case of Jini and Java2, Standard Edition doesn't).

 
Where SCSL doesn't work for the Free Software and Open Source communities:

1. Doesn't allow free redistribution (you have to make sure that everyone has agreed
to 
obey the community rules before you distribute).

2. Doesn't allow forking, requires compatibility as a condition of productive code use
(mainly a Java problem now, as Jini has switched to optional branding).

3. SCSL is implemented as a click-through license so that privity of contract between
Sun 
and the licensees can be maintained.  Free and Open licenses don't require click-through.
 
Some potential licensees don't wish to be added to a database and resent the click-through.

Now, there are a couple of Java implementation issues that are sometimes confused with

licensing issues.  Jini's community doesn't have these issues, however:

4. For Java, the Compatibilty Kit is only available from Sun, and there is a cost 
associated with it.  So not only MUST you use it (see 2 above), but you have to pay

money to use it.  Sun has made it pretty difficult for the common man to find out 
exactly what the cost is, also.  You have to find a Sun salesman to tell you.

5. The Compatibility Rules for Java2, Standard Edition identify some code that you 
cannot modify (because Sun feels that any other implementation would break the "Write
Once, 
Run Anywhere" value proposition of Java).  For this reason, a "cleanroom" clone of 
Java2, Standard Edition couldn't possibly pass compatibility.  You MUST use some Sun
code.

Anyway, the point of all this is to say that SCSL can be made to work, if you pay 
attention to setting up a program with community goals at the core.  I do agree that

at this point you also have to overcome the "stigma of SCSL", but writing your own license
also comes at a price IMHO.  SCSL is at least well written from a legal point of view
and 
enforceable.

Danese Cooper
Sun Open Source

>
>Crispin Cowan writes:
>> What you're describing is pretty similar to Sun's Community Source License
>
>ah, thank you. i'm going to look at the license now. see how close it is to what
>i'm wanting.
>
>> Note that SCSL is largely reviled in the free/open source communities.
>
>hm, understood. thanks again. i take it that the SCSL is definately *not*
>considered Open Source, then, even if Sun *may* have tried to push it as, or
>just as good as, it? not claiming they did, just trying to clarify, since as i
>understand it, there are a couple of different definitions of what open source
>is floating around (opensource.org and the FSF, etc.).
>
>> So while SCSL has the semantics that you want, it also has some unfortunate
>connotations.
>
>hm, again. yeah, that could be bad. while i want to avoid those bad
>connotations, i would like to go with the model i have in mind. i guess i may
>have to make my own license, if that is feasible. and make it so it has the good
>aspects of the SCSL, but does not try to force itself the way the SCSL does.
>
>thanks mr cowan, i now have somewhere to start, at least.
>
>>>    ||  "Did you ever get the feeling the world was a tuxedo,
>>>    ||     and you were a pair of brown shoes?"
>>
>>How aprapos :-)
>
>yes :) i heard this a few months ago, i am pretty sure on TV or a movie
>somewhere, but for the life of me cannot remember where. it just stuck in my
>head and i decided it was so good i had to put it in my sig.  :)
>
>now i'm going to reply to another response in this email, conserve paper ;-).
>
>Dave Turner writes:
>
>>Why don't you ask instead on unfree-software-business?  You're not
>>really talking about any sort of freedom at all.
>
>my apologies. i wrote to this list because it had a very high level of
>professionalism and no slander or degradation (high-quality members, in other
>words), and also had a wide range of opinions and standpoints. i do understand
>that the list is named free-software-business. as far as not having any sort of
>freedom at all, if you mean compared to BSD-style, no, you're right. but i am
>not looking for traditional proprietary licenses that heavily constrict the end
>user and third party developers. i am, however, looking for something that is
>much more free than those, allowing end users a lot of power to view the source
>and know exactly what they are getting, modify it, possibly even trying it out
>before paying for it, and allowing third parties to build software that enhances
>or works off of this program, if not allowing them power to re-distribute it.
>
>if i am in error to post this thread to here, then i am very sorry, and will not
>do so any longer, keeping my posts well within the bounds of opensource.org's
>and the FSF's definitions.
>-- 
>   /\    --- Adam Theo ---
>  //\\   Theoretic Solutions (www.Theoretic.com)
> /____\     Software, Politics, and Advocacy
>/--||--\ email: theo@theoretic.com   AIM: Adam Theo 2000
>   ||    jabber: adamtheo@jabber.org   ICQ: 3617306
>   ||  "Did you ever get the feeling the world was a tuxedo,
>   ||     and you were a pair of brown shoes?"