[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

*Subject*: Re: a plea for more reliable mathematical routines*From*: Mirko Vukovic <mvukovic(at)taz.telusa.com>*Date*: Tue, 14 Sep 1999 15:14:24 GMT*Newsgroups*: comp.lang.idl-pvwave*Organization*: Deja.com - Share what you know. Learn what you don't.*References*: <7r90g4$rqb$1@nnrp1.deja.com> <37D82EA9.BA62A369@wellesley.edu> <7r9jbl$aem$1@nnrp1.deja.com> <37DCCE9A.F1AC4BF1@zedat.fu-berlin.de> <MPG.1246a891f3c895e19898eb@news.frii.com> <37DE1600.5E99BDE4@zedat.fu-berlin.de>*Xref*: news.doit.wisc.edu comp.lang.idl-pvwave:16534

In article <37DE1600.5E99BDE4@zedat.fu-berlin.de>, fit@functional-imaging.com wrote: > I definitely disagree. It is inferior to Java, Python, C/C++ (if You're able to > program a little bit of OpenGL and Motif yourself) to name only some, far too > expensive, introducing new bugs with every release (maybe a merger with Micro$ > would be adequate), lacking hooks for any reasonable development environment (or > have You ever managed to get it to work with Rose or SNiFF+ to name only a few). I would agree if you compare them as general purpose languages. But for data analysis and writing imaging routines, I presume that IDL beats these, since it was designed (with flaws) for that purpose. You can accomplish the same with the languages you mentioned, but with how much effort. I restrict my comment for small and medium sized applications. For a huge application with millions of lines of code, it may be more worthwile to go to Java/C++/..., simply because of the ruggedgness and the development tools. Regarding the above issues I would prefer a comparison of IDL with PV-Wave, matlab, mathcad -- none of which I use. > Secondly, I definitely did not characterize objects as childish but the way > they're used and implemented in IDL (look folks, now we're object oriented !). > What has been done there to the object paradigm is pretty much the same as they > did to numerical mathematics (look folks, we've the numerical recipes > implemented, ok the results are shaky at best, but look we have them > implemented). To incorporate an object oriented paradigm (encompassing, yes > David, a development process as well) is a little different to providing a syntax > of o->x() form. > I agree that 5.2 is not up to C++ regarding oop, but with some programming conventions, can you achieve much of the same results? Like, you cannot define a private/public interface, but can you as a programmer label an interface as such and use it in a consistant way. I agree it is inferior to an explicit declaration, but better than nothing. (here I am threading a ``tiny bit'' beyond my expertise) Mirko Sent via Deja.com http://www.deja.com/ Share what you know. Learn what you don't.

**References**:**ODEPACK***From:*ushomirs

**a plea for more reliable mathematical routines***From:*Richard G. French

**Re: a plea for more reliable mathematical routines***From:*ushomirs

**Re: a plea for more reliable mathematical routines***From:*FIT

**Re: a plea for more reliable mathematical routines***From:*David Fanning

**Re: a plea for more reliable mathematical routines***From:*FIT

- Prev by Date:
**Re: a plea for more reliable mathematical routines** - Next by Date:
**Re: a plea for more reliable mathematical routines** - Prev by thread:
**Re: Plea for IDL 2000 (was: a plea for more reliable mathematical routines)** - Next by thread:
**Re: a plea for more reliable mathematical routines** - Index(es):