Re: Q: access to subprogram

[rogoff@sccm.Stanford.EDU (Brian Rogoff)]

bobduff@world.std.com (Robert A Duff) writes:
   Brian Rogoff <rogoff@sccm.stanford.edu> wrote:
   >Anecdotal evidence, but useful.

   Anecdotal is all you ever get in this business.  ;-)

Correct. But by collecting this anecdotal evidence and analyzing it we 
can hopefully get some idea of what we're doing :-).

   >   Ada's generics can do *almost* everything that Pascal can do.  They
   >   can't recurse, in Ada, like they can in Pascal.  (I needed that once.)
   >   And they're verbose, as I said.
   >
   >Do you feel that they are the right solution though?

   No.  They work in many cases, but they're not the best solution.

That's my feeling too, although when I thought about it more I couldn't 
really rationalize why I disliked generics so much for this problem and 
yet like the use of null bodied generic formal package parameters for 
signatures. Maybe in a few days the generic approach won't seem so bad.

   >Was this the "why are they any better than wrapping up the environment and 
   >func in a tagged type" question?

   Yes.  Not just "why are they better", but "why are the a LOT better"
   (which is the claim I often hear).

I think that you posted that you like Smalltalk blocks a lot, which are 
essentially just anonymous closures, right? I think the beauty of this 
construct is that you can implement lots of features cleanly in terms of 
it, like control structures. Also the implementation of closures on objects 
seems heavy/verbose to me. I think in a language like Smalltalk, which 
already has GC, that upward funargs and objects are (a LOT) better than 
objects alone. On the other hand, if you are asking if they are so much 
better than explicitly managing the heap that Ada should adopt GC :-), I'd 
say no.

-- Brian