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

[davidf(at)dfanning.com (David Fanning)]

Greg (ushomirs@my-deja.com) feels like this discussion
is getting a little hot. I'm sorry he feels that way
and I'm sorry if he feels like I'm contributing some of
the heat. 

Let me just say, again, that I am completely in favor of
discussions of IDL's weaknesses. Many people at RSI
read this newsgroup (I know this from my personal experience
of feeling a little heat myself from time to time) and
this is a useful forum for making our thoughts, feelings,
and even our wish-lists known to them. They DO pay attention.
And they DO care about what you have to say. Their goal--as
even the most cynical among us must acknowledge--is to sell
software. They can't write software that is indifferent to
the needs and wishes of their users if they want to be
successful.

But what is not helpful--indeed it might even be
counterproductive--is to couch our disappointments
and displeasure and whatever it is that is not working
for us in invective. Better, and MUCH more productive,
to voice our concerns and offer a few constructive
suggestions, as Mr. Vukovic did a day or so ago.
You might even try wit and humor if you are capable
of it. Although humor sometimes goes awry in
print form, it at least makes for more interesting
reading than a rant of the things you don't like
and it can be every bit as sharp.

Will RSI respond to everything we want? Not likely.
While the software costs more than Microsoft Office,
RSI probably sells the same number of licenses in 
a year that Microsoft sells in an hour. Pockets
and staff (whatever you may have heard) are not
nearly so deep. So wish-lists have to be balanced
against time, priorities, and staff considerations.
What people have asked for and what they want DOES
figure into that equation. More now, I think, than
it did several years ago when I was more familiar 
with the process.

The bottom line is this: keep those cards and letters
pouring in. RSI *needs* our feedback. But nobody--from
dogs to children to software companies--responds well
to shouted criticism. Show them the stick (and your
wallet), but give them the carrot, too.

> So i don't see how unsound programming practices
> that would have been acceptable in a prototype could pass in a real,
> shipping product. 

Several years ago now RSI realized that what seemed like
a good program structure 16 years ago wasn't working so
well anymore. It was becoming more and more difficult
to graph new features onto the underlying program
scaffolding. So they took nearly one and a half
years to completely rewrite the IDL internals, using
the best programming practices at the time. The idea 
was to create an internal structure that would allow 
things like objects and pointers and the new development 
environment, which would probably not have
been possible with the old structure.

What I remember most about that time was how often
I had to face irate people who were screaming
at me that for a year and a half they had gotten
NOTHING for their maintenance dollars! Why was RSI
ripping them off! Was David Stern talking the whole
company on vacations to Mexico with their money, etc., 
etc. I felt like *I* was under a lot of pressure,
and I wasn't even writing the code. I can imagine
what the engineers were feeling, working 10-12 hour
days. I think the fact that a fair number of them 
left the company at the end of that effort is testament 
to the strain they were under. 

But you see, it is the nature of people that they want to
have their cake and eat it too. I wouldn't be a
programmer for a software company for anything. You are
always in the hot seat, no matter what you do. And
who among us can write bug-free code under that kind 
of pressure?

> And boy, does Microsoft get major flak for bugs in their stuff!

Yeah, well, they deserve it. Gates is rich as hell. :-)

> SO while my original complaint about bad documentation for LSODE started
> this big flame war, nobody even bothered to take on my second question,
> how the heck do i call an IDL function from an external (linkimage or
> dlm) module??

Donno. Not my job as Defender of the Realm. :-)

Cheers,

David

-- 
David Fanning, Ph.D.
Fanning Software Consulting
Phone: 970-221-0438 E-Mail: davidf@dfanning.com
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Toll-Free IDL Book Orders: 1-888-461-0155