comp.lang.idl-pvwave archive
Messages from Usenet group comp.lang.idl-pvwave, compiled by Paulo Penteado

Home » Public Forums » archive » Re: Interesting Rant
Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend 
Return to the default flat view Create a new topic Submit Reply
Re: Interesting Rant [message #51301 is a reply to message #51299] Wed, 15 November 2006 05:25 Go to previous messageGo to previous message
Maarten[1] is currently offline  Maarten[1]
Messages: 176
Registered: November 2005
Senior Member
David Fanning wrote:
> I disagree with much of what he has to say, but I
> can't tell whether this is because I am old enough
> to remember what things were like *before* IDL, or
> whether I've just become inured over the years. :-(

On the contrary, I'd like to imagine what things can become *after*
IDL. While I would list a different set of annoyances, I'd say you have
become inured to the IDL idiosyncrasies over the years. Lovely word,
"inured", had to look it up though (can you tell I'm not a native
speaker?)

* Arrays in structures can not be resized. This also applies to
objects.
* You can have resizable arrays in structures, if you use pointers. The
way to access data in the array changes, you'll have to rewrite your
code.
* Pointers that can get lost in a scripted language. If I want a memory
leak, I'll use C, Fortran, ...
* Default integer size is 16 bits (how long have computers been at
32bits by now; how much code will break if you change that to 32 (or
even 64 bits)? How can code break on such a change? I think the code
was broken in the first place if it relies on this bit-size.
* Individual floating point constants are float, not double.
* Direct graphics seem to be dead, object graphics are not practical
for interactive use. Hello, the I in idl stands for interactive, right?
* Procedure arguments as output parameter and array elements.
* Brain dead for loops. Nice to have a vector engine (which doesn't
work in all cases, idl_validname() only accepts strings, not string
arrays, so for loops cannot be avoided in all cases), but sometimes an
explicit for loop aids readability (histogram jugling, anyone?).
* Stupid end of line behaviour, especially on interactive use.
* Logical test on least significant bit only
* The UI on Mac OS X. Using X11 on Mac OS X does not make a port.
* Special characters in graphs, combined with the butt ugly Hershey
fonts. The graphs look old-fashioned, and like they were made with some
homebrew software, rather than a professional, expensive tool. Yes, you
can do better in IDL, many don't.

So while IDL probably is an improvement over what came before it (I'm
too young to really tell), those programs are truly dead and buried.
IDL seems to survive, despite better solutions, especially for
interactive use. The only things that keep idl alive are the dinosaurs
that use it, and the legacy code that has been written for it. Others
have suggested to use Python as a basis, and I think I agree, although
some array indexing issues will be as annoying as IDL. At least there
is a large community behind it, and the core of python is free.

Maarten
[Message index]
 
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Previous Topic: Color background with PLOT
Next Topic: Re: Image warping in IDL

-=] Back to Top [=-
[ Syndicate this forum (XML) ] [ RSS ] [ PDF ]

Current Time: Wed Oct 08 19:31:14 PDT 2025

Total time taken to generate the page: 0.00500 seconds