Re: IDL interpreter questions - can someone (D.Fanning) explain -TIA

[Ken Mankoff <mankoff(at)I.HATE.SPAM.cs.colorado.edu>]

On Sat, 19 May 2001, Mark Rivers wrote:

> mankoff@lasp.colorado.edu wrote in message ...
>
> >Now its true that I don't know anything about the actual IDL
> >implementation (though I have written RPC code for IDL). I actually
> >answered based upon the behavior of IDL, not the implementation.  That is,
> >functions won't modify the callers variables, and neither will procedures,
> >unless you add the 'return'.
>
> That's not true.  Here's the proof:
> So the procedure and the function both modified arguments passed to them.
>

apparently you are right, what I said wasn't true. Wow. I stand doubly
corrected *and* get to learn a new feature of IDL I never realized
existed, all in one day (good thing its only 11:54pm in my time-zone :).

I've been coding in IDL for almost three years, and I really thought that
procedures and functions behavior could be modified by use of a return
statement. Furthermore, almost every procedure I've written has a "return"
as its 2nd-to-last line, and an "end" as its last line. I just read the
IDL help and this is completely redundant!

Sorry for any confusion I may have started to spread, especially to the
pour soul who originally started this thread. Thanks Mark & JD for
teaching me some obvious stuff about the language I should have realized
long ago.

The really wierd part, is that i could swear I once tracked down and fixed
a bug with a return... i'll have to see if i can remember what and where
that was, and figure out what the *real* bug was...

-k.

--
Ken Mankoff
LASP://303.492.3264
http://lasp.colorado.edu/~mankoff/