Well, it's not my thing (I mean whatever Stephan decides is up to him; he's
doing the implementation), so I'm somewhat reluctant to give much comments.
However, I'm afraid that slow block closures will just remain there for
another five years, I guess, until someone Stephan II decides to make them
quicker and has to go through all that learning again...

That said, if Stephane implements 'slow blockclosures' or 'fast ones' I'll
definetely buy him a beer (or whatever) whenever I encounter him ! Go
Stephan !


On 30/04/01 21:33, "danielv at netvision.net.il" <danielv at netvision.net.il>
wrote:

> I've been following this thread with interest, and reading Stephan's
> meta- reasons, I just have to agree. Do the simplest thing that could
> work, and "simple" is relative - if you know how to do it, it's simple.
> 
> If Squeak gets block closures (even slow ones), there'll probably be
> various good things that will happen in consequence (code will break and
> be fixed, ports of software that needed closures will be untwisted, so
> forth).
> 
> If someone then takes on the implementation challange of "The Allen
> proposal", integrating it with J3 for better performance, we'll all be
> happy, but we probably won't be affected much (another X0% performance
> for Squeak? we'll all smile, and continue doing precisely what we we're
> doing before). 
> 
> Stephan, if at some point you see what you were doing does get complex
> and Allen's proposal seems to be getting simpler, you can always change
> directions later.
> 
> Daniel
> 

--
Roel Wuyts                           Software Composition Group
Roel.Wuyts at iam.unibe.ch         University of Bern, Switzerland
Board member of the European Smalltalk User Group: www.esug.org