Squeak practical use?We run the equity markets through it!

Bijan Parsia bparsia at email.unc.edu
Tue Jan 29 16:56:09 UTC 2002


On Tue, 29 Jan 2002, O'NEEL Bruce wrote:

> Bijan Parsia writes:
> <snip>
>  > 
>  > For "non functional" type looping, I'm growing more and more to admire
>  > Common Lisp's extended loop. A language for looping!
> 
> Um, it looks like a great idea, but it is awful in an environment
> where you have to keep a system written by many different people
> running.

Experienced people differ on this point.

>  The problem is that everyone, and I mean *everyone* uses
> different bits of loop differently so that every single one you come
> across you have to pull out the CLTL2 (and/or the manual for the
> particular lisp since everyone and his brother wanted to extend loop)

But this is well out of date as 1) CLTL2 isn't a standard, 2) The ANSI
spec is, 3) implementations conform to the ANSI spec, and 4) the ANSI spec
is about as extended as it gets :)

The ANSI spec is very very very nice.

Many of the examples I see that are convincing are really a bloody mess of
ad hoc manuvering. I fail to see the advantage in ignoring Greenspun's 10
1/2th law: Any sufficiently complex looping code contains contains an
ad-hoc, informally-specified bug-ridden slow implementation of half of
ANSI Common Lisp's extended loop.

> and puzzle out what in the ^%^&@# they were doing.  I think the basic
> problem is that there are no well known patterns for using loop that
> have been adopted like other lisp idioms.

This seems false. The ANSI spec contains a number of exemplar examples.

> Here's to not changing languages in mid stream :-)

Er...ok ;) Uh...here's to using the right tool for the job!

Cheers,
Bijan Parsia.




More information about the Squeak-dev mailing list