| Re: Announcing GDL 0.7, now with PLOT command [message #38563 is a reply to message #38285] |
Sun, 14 March 2004 10:49  |
George N. White III
Messages: 56
Registered: September 2000
|
Member |
|
|
On Tue, 2 Mar 2004, H C Pumphrey wrote:
> I shall begin testing GDL on some real-world code .... roll on contour!
> And remember that a few years ago, the main implementation of S was S-Plus,
> while R was a toy. R has now taken much of S-Plus's mindshare. OTOH, octave
> has stagnated as a clone of a very old version of Matlab. Which will be the
> main implementation of IDL five years' time: RSI's or GDL?
There have been open-source APL and J interpreters, but my impression is
that commercial interpreters are more widely used.
Octave wasn't just a clone, it introduced some ideas (structures) that
became part of real Matlab. Now Matlab has incorported ideas from octave
and many other extensions to the orginal language. R addressed
fundamental problems with S-Plus in student lab environments on multi-user
machines (it allowed you to limit the memory a user could allocate so one
user couldn't monopolize all the resources of a large server). I suspect
this meant that a large number of S-Plus users ported software to R for
use in courses and then found it suitable for "every day" use.
I started using R for statistical summaries and reports because it has
excellent support for dealing with missing values as well as the
statistical tools. Recently I've been doing more things in R because they
can be used by more people than if they are done in IDL, but GDL could
reverse that trend.
The grad students and recent PhD's I encounter generally have experience
with Matlab. A few have used R and a very few have used IDL. GDL would
need some compelling advantage to get it into eduational institutions
before it can have the level of success of R.
--
George N. White III <aa056@chebucto.ns.ca>
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
