Re: [Chicken-users] faster threading

[F. Wittenberger]

Am Samstag, den 01.11.2008, 09:51 -0400 schrieb Marc Feeley:
> On 31-Oct-08, at 10:11 PM, Jörg F. Wittenberger wrote:
> 
> > Thankful for a suggestion from Marc Feeley I reused rbtree.scm from  
> > the
> > snowfort repository.  The attached version extends the snow version by
> > optionally generating lookup and fold procedures for the generated
> > trees.  (And contains some chicken specific lines on the head.)
> 
> I'm glad you decided to use the red-black tree implementation from  
> Snow.  As you know this code was originally written by the MIT-Scheme  
> team and I made some important changes to improve the efficiency.   
> Please maintain the copyright notice and comments at the top of the  
> code which explain all of this.  Also note that the code has an LGPL  
> license.

Thanks Marc and all for answering.  (Actually I asked Marc before in
private mail mentioning an minor problem I have with the implementation
anyway too.)

Cleaning up this issue was my intent when posting too.  (In fact I found
a "all rights reserved" in the rbtree file together with the MIT
license, which made me wondering what's valid; I failed to spot an LGPL
reference.)

Now the obvious solution - since I consider it a none-solution for any
of us to just know that chicken could be several times as fast as it
currently is - will be that someone has to sit down, understand the
red-black algorithm from the paper and code it under BSD license in
hygienic macros as I suggested (merely for aesthetic reasons to get rid
of the define-macro).

The other solution is the "lawyer way" - "Dünnbrettbohren" in German.
Run the macro and manually tweak the result for inclusion.  As far as I
understand the legal issue, the output of neither a MIT, GPL or LPGL
program is covered by any of those licenses, is it?

So what shall we do?

Best regards

/Jörg