| Re: What does an optimal scientific programming language/environment need? [message #36526 is a reply to message #36427] |
Thu, 02 October 2003 17:08   |
donotreply
Messages: 1
Registered: October 2003
|
Junior Member |
|
|
In article <k6Meb.443$ye2.217564282@newssvr11.news.prodigy.com>,
unixmonster@hotmail.com says...
>
>
> bv wrote:
>
>> grunes wrote:
>>>
>>> I'm working on creating an optimal scientific programming language and
>>> environment. My hope is that people who use current environments have
>>> specific things they love about it, that need to be included. For now
>>> I'm trying to combine the best concepts from FORTRAN, BASIC, C, APL,
>>> IDL, PV-WAVE, and possibly MATLAB.
>>
>> Before you embark on what is bound to be a long and winding road you
>> might want to consider a recent quote by "DB" from sci.math.num-analysis
>> ng which would invariably apply to whatever you might come up with.
>>
>> "To get any chance of succeeding new programming languages should
>> from the beginning provide a huge advantage compensating the loss of
>> decades of expertises contained in the already available libraries, in
>> the trained people, as well as in the compiler technology. Now to make
>> the situation worse, the many functional languages compete with each
>> others."
>>
>> --
>> Dr.B.Voh
>> ------------------------------------------------------
>> Applied Algorithms http://sdynamix.com
>
> I would prefer to see APL extended with operator overloading and with
> defined primitive numeric types - so that one could model things like
> Grassmann algebras, moving frames etc. and maintain the concise syntax.
>
> I see little point in inventing another syntax.
>
> The most useful math machine I have is my TI-92+, because I can take it
> anywhere and it has a "good enough" symbol manipulation capability. I use
> it mainly for doing calculations in 6-d space. The syntax is based on
> "Derive" and I find it quite acceptable.
DERIVE has been my favorite computer algebra and ad-hoc calculation
language for a long time. The fact that it is now sold by Texas Instruments
through their education department belies its power. It's LISP-based
(although the LISP is almost entirely hidden) and, in its current
incarnation, quite programmable. However, there's little that's procedural
about its programming (not unexpected, given its LISP roots); instead, one
writes a number of functions that reference each other.
DERIVE and its ancestor, MuMath, has actually been around for a LONG time
-- IIRC, since the late 1970s. By the standards of most anything found in
the computer world, it's remarkably bug-free. It also allows symbolic
results to be output in Fortran syntax.
Highly recommended.
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
