| Re: Giant arrays! [message #9502 is a reply to message #9420] |
Fri, 11 July 1997 00:00  |
David Ritscher
Messages: 30
Registered: August 1995
|
Member |
|
|
Mary Jo Brodzik writes:
>> P.S. Anybody else want to join in a campaign for a system variable that
>> switches the default literal integer to long rather than short? I've
>> been bit too many times...
> Yes, now that would be useful, I've been bitten a number of times,
> myself. And although I try to warn new IDL users, this problem
> continues to bite them, too!
Do you mean that
i = 2
should produce a something other than a 2-byte integer?
I think we should discuss this problem after the year 2000 :-)
IDL 5.0 has finally fixed the 'enhanced feature' of not being able to
differentiate between an array a(3) and a function call
my_function(3). I think it is time for a language reform, where a
system variable tells the interpreter when to interpret code as
vintage code, so, for example,
!LANGUAGE_LEVEL = 1
at the beginning of a routine (i.e., one I wrote years ago) would tell
the interpreter to default to 2-byte integers, arrays with parenthesis
for delimiters, etc. With such a mechanism in place, then RSI and VNI
would be free to update the basic language structure while leaving our
old code in a functional state.
What would other good additions entail? A syntax that follows C++
more closely would be a good start (concerning use of '(', '[", '{',
etc.) Other suggestions??
Tschuess,
David Ritscher
--
David Ritscher
Zentralinstitut fuer Biomedizinische Technik Tel: ++49 (731) 502 5313
Albert-Einstein-Allee 47 Fax: ++49 (731) 502 5315
Universitaet Ulm Internet:
D-89069 ULM
david.ritscher@zibmt.uni-ulm.de
Germany
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
