License Committee Report for September 2009

[Giancarlo Niccolai]

Russ Nelson wrote:
> Giancarlo Niccolai writes:
>  > sent with the submission I explained why I thought it is preferable to
>  > stop the "custom license / GPL, but with my own uncontrolled/uncertified
>  > exception" rage, and to offer a standardized license covering this
>  > general need.
>
> I believe it is wiser to wait until the need is made clear by having
> many adopters of this language.  
I don't know if the relevance of this license has anything to do with 
the relevance of THIS language. The need is felt because every language 
needed to stick with specialized licenses and/or with totally unrelated 
exceptions, or, as a last resort, to dual licensing. This includes PERL, 
SWI Prolog, Python, PHP, and even a good deal of GNU foundation 
software. Also, the license is not project specific (as i.e. Python's 
and Ruby's).

> A GPL ... with modifications ... can
> be consider as the GPL unless you particularly care about the extra
> freedom.  Yet Another License needs to be understood in its totality.
>   
On the point that evaluating the exception is easier than evaluating a 
license, read the evil #1 to #4 in the rationale. -- The fact that the 
text of an exception is shorter than the text of a license doesn't help 
in evaluating the breadth of its legal consequences. I could show many 
examples in the Italian regulatory system where the change of a single 
adjective (with a quasi-synonym) with respect to a base law WILLFULLY 
caused criminals (friends of the law proposer) to go free, and public 
money initially thought to go in green energy to be diverted into 
nuclear energy and incinerators. The law changing the adjectives in the 
first case and adding a word in the second was exactly one sentence 
long; structurally identical to an "exception". And no one was able to 
understand how maliciously dangerous they were in advance.

The danger with exception proliferation is exactly that of letting 
similar legal bombs to sneak into the open source world, or into your 
economic entity, as they are simply too many to be considered carefully 
one by one; and so, they are often accepted with nothing more than "oh, 
ok. Seems a reasonable exception".

So I strongly disagree on the fact that "GPL ... with modifications ... 
can be considered as the GPL". This was a shortcut taken by the 
community to spare the activity needed to regulate the rights on some 
shady regions (code helping other's code -- when not a library that 
could be linked-by-name), as that of embeddable scripting engines. IMHO, 
any exception, no matter how sensed it looks, totally invalidates any 
previous analysis on the whole of the license, no matter how trivial it 
seems. Especially when the exceptions become so many that they may start 
to clash on each other.

On this account, the scheme of the Creative Commons, with its modularity 
of pre-defined rights you can chose to preserve, it's a better solution 
than a monolithic core that the community accepts to be amended on one's 
needs -- and with one's personal and more or less careful or willfully 
malicious wordings -- without too much fuzz.

The point I want to make clear is that this license is NOT aiming to 
protect some specific specific work, but is aiming to provide an 
open-source regulation on a previously shady area; so shady that 
practically every other similar software in the open source community 
needed to provide an ad-hoc solution. Ad-hoc solutions, even when they 
are "just exceptions", doesn't work in the direction that drives the OSI 
into granting open-sourceness certifications. Rather, they are a 
constant nuisance for other developers willing to go open, and for 
commercial entities willing to adopt those products but not being sure 
about what they can or cannot do with the product (a simple search on 
the net shows that this concerns are present on both sides of developers 
and users, and shows also how much they are tiresome).

Hope what I say makes sense.
Giancarlo