In article <BOBO.94Sep27093605@avogadro.arc.nasa.gov> bobo@avogadro.arc.nasa.gov (Mark Friedman) writes:
A few clicks later I also saw the following:
Note that functions created with lambda forms cannot contain
statements.
Supposedly these features were created by popular demand. So which one
of you demanded these eviscerated lambdas?
While browsing through the reference manual I had the same thought... But
presumably this doesn't lead to any atrocious restrictions as you can
just have your lambda call a function. Any comments from the python fans ?
It's not really the lack of statements that makes these functions
eviscerated; it's the lack of ability to reference variables in
non-global outer scopes (i.e. closures).
I was going to underline that section of the quote. I figured that it
would stand out to the functional programming people. I forgot that I
was also cross-posting to some groups where people where less familiar
with passing around closures.
Oops, I knew I was reading too fast ... Half my code uses first-class
closures, and yet I didn't notice :-( I really have no excuse, as I
implemented a scheme-like extension language for a MUD, with closures, GC,
etc, though with a more conventional syntax and imperative style (I did
suggest using Scheme directly, but there were "religious" objections -
incomprehensible, nobody will use it, ie all the usual (unfounded?)
complaints).
David Gay
dgay@di.epfl.ch