Re: Plea for IDL 2000 (was: a plea for more reliable mathematical routines)

[Craig Markwardt <craigmnet(at)cow.physics.wisc.edu>]

davidf@dfanning.com (David Fanning) writes:
> What I don't understand is the heat behind these
> feelings that IDL is a big hack. Heck, go use something
> else if you feel that way. It's a competitive marketplace
> that IDL lives in and you are free to buy (or build)
> anything that does the job for you. IDL only exists because
> *somebody* keeps buying it.

Hi David--

I love IDL.  It's got a wonderfully expressive language, powerful
vectorizable operators, and mountains of library software (my own, and
from others).  As an interactive analysis language it has profoundly
changed how I work (for the better).  I have made software the beats
the socks off its C/FORTRAN equivalents.

I could hate IDL. It's got a quirky language with objects tacked on.
I have made a large investment in mountains of software that won't run
on any other system.  IDL is not very friendly to the
programmer/maintainer and has introduced and obsoleted several
language features over the course of a year or less.  Software bugs
persist through several versions.  When a bug appears we have no
recourse in fixing it, since we don't have the core IDL source code.
For example, witness the arguments about mathematical functions.  I
have spent a profound amount of my time working around IDL bugs.
Making a simple hardcopy in direct graphics is a serious
inconvenience.

Love/Hate, that's what it is!

Craig

-- 
--------------------------------------------------------------------------
Craig B. Markwardt, Ph.D.         EMAIL:    craigmnet@cow.physics.wisc.edu
Astrophysics, IDL, Finance, Derivatives | Remove "net" for better response
--------------------------------------------------------------------------