Re: Choice of programming language
- From: "Anthony Garcia" <agarcia658@xxxxxxxxxxxxx>
- Date: Thu, 16 Mar 2006 05:26:35 GMT
"NoIdea" <foru@xxxxxxxxx> wrote in message
news:3vBRf.7160$dw4.1238@xxxxxxxxxxxxxxxxxxxxxxx
"Damien" <damien.hocking@xxxxxxxxx> wrote in message
news:1142350353.235407.53340@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
It's still mostly Fortran, with some C/C++. New codes are likely to be
Fortran 90/95 rather than Fortran 77. What's likely is that the
numerical part is Fortran and the simulation or modeling framework that
poses the problem is C++. Nobody does serious numerical computation in
Java. It sucks for that, it's too slow.
yep agreed on that - i'm writing some linear algebra applications in java
at
the moment and they dont exactly rocket along. im only using java because
it has a arbitrary precision arithmetic feature which few other languages
have built in.
with regards to the OPs question, undoubtedly fortran
Damien
There are actually plenty of people who perform numerics in Java. As an
example consider the rotordynamic code FEMRDYN which performs FEA and
various integration schemes for time domain analysis of rotor systems. I
think portability, availability of tools, programmer efficiency and other
considerations are often the determining factor in numerical tools. At my
company we often use visual basic for programming many analysis functions
because it is portable within the wintel framework and we can integrate it
with other widely available software from CAD to word processing documents.
.
- References:
- Choice of programming language
- From: Mark B
- Re: Choice of programming language
- From: Damien
- Re: Choice of programming language
- From: NoIdea
- Choice of programming language
- Prev by Date: Re: Choice of programming language
- Next by Date: Re: Choice of programming language
- Previous by thread: Re: Choice of programming language
- Next by thread: Re: Choice of programming language
- Index(es):
Relevant Pages
|