Re: Recommendation for a CL data structures library

[Raffael Cavallaro]

On 2010-03-21 22:14:30 -0400, Pascal J. Bourguignon said:

> Sure.
>
> And the question remains why you should imposes your choices on me?

Not only am I not imposing anything on you, I've already offered to pay 
you for a commercial license. So you can have your cake (GPL licensing) 
and eat it too (paid commercial licensing).

My principal objection to the GPL is that its license requirements 
regarding opening source code make it very unpopular with many 
commercial developers, and therefore whenever possible, they choose 
non-GPL alternatives.

In short, I don't think GPL licensing gets you anything additional in 
terms of getting code open sourced. Users who need to keep their source 
closed either won't use it, or will use in in a way that allows them 
not to open the source (e.g., Paul Graham's viaweb and their use of the 
GPL CLISP).

Meanwhile, users of LLGPL or BSD, etc. licensed code frequently open 
source whatever they are able as contributions back to the relevant 
project. Giving users the choice of what they will and won't open 
source results in more users, and just as many open source 
contributions.

More use means the library is more of a community standard, more 
contributors, more bug fixes, more extensions, more additions, etc. And 
non-GPL license authors can still sell support (e.g., priorty for 
bugfixes or priority for adding new features, etc. - The Clozure folks 
did this for their IDE, and Clozure is LLGPL licensed).

I think people should avoid GPL licensing their work as a pragmatic 
means of ensuring maximal adoption. Ironically, the FSF understood this 
dynamic which is why they created the Library GPL, now known as the 
Lesser GPL.

warmest regards,

Ralph


--
Raffael Cavallaro