| Re: Compiling IDL ... ever likey ? [message #5608 is a reply to message #5518] |
Mon, 22 January 1996 00:00   |
Ken Knighton
Messages: 44
Registered: May 1995
|
Member |
|
|
On the subject of compiling IDL, I thought I would just throw a little
gasoline on the fire and see what happens:
Languages that compile to the host machine level are rapidly becoming
obsolete and investing programming time into turning a very efficient
pseudo-compiled, array oriented language like IDL into a true compiled
language is a waste of resources that could be better used to improve
other features of IDL. For example: I haven't heard anyone suggesting
that JAVA be turned into a compiled language.
Computers are simply getting so fast and memory is getting so cheap and
abundant, that a few cycles here and there or a little misuse of memory
doesn't matter for the average application. On the other hand,
programmers are not getting faster or cheaper; therefore, the payoff is
bigger if time is invested in making IDL more programmer friendly.
IMHO, RSI should concentrate on making IDL a terrific application
development platform so that serious applications can be developed for
mass distribution. What IDL really needs is a very cheap run-time
system license for MACs and IBM-clones, an option for compile time type
checking, rapid application development tools, cleaner APIs for plotting
and widgets, ...
If you need to do a calculation that requires looping and will take a
while, and you need to do it over and over, day after day, week after
week, then I agree, you need to use fortran or C. If this is the case,
then it is worth the time that it will take to write a FORTRAN or C
routine to do it and to use CALL_EXTERNAL to interface with IDL.
Sincerely,
Ken Knighton knighton@gav.gat.com knighton@cts.com
Fusion Division
General Atomics
San Diego, CA
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
