Re: Plea for IDL 2000 (was: a plea for more reliable mathematical routines)

[Mirko Vukovic <mvukovic(at)taz.telusa.com>]

In article <7rsidq$j5s$1@nnrp1.deja.com>,
  ushomirs@my-deja.com wrote:
> In article <MPG.124ad31447ec3ccd9898f7@news.frii.com>,
>   davidf@dfanning.com (David Fanning) wrote:
>
> > New in IDL 5.3 (according to the on-line documentation in the beta
> > version is a "compile option" routine that can change the default
> > integer size from 16-bit to 32-bit:
> >
> >    IDL> Compile_Opt DefInt32
> >    IDL> a = 0
> >    IDL> Help, a
> >       A    LONG   =  0
> >
> > Cheers,
> >
> > David
> >
> > P.S. Note there is NO comma after the COMPILE_OPT
> > command! Took me about 10 minutes to realize that. :-(
>
> see!  that's yet another example of how poorly thought out IDL is!!
> other directives (such as .RUN, .COMPILE) don't need a comma after
> their names.   Why not make it .COMPILE_OPT, so that the lack of comma
> would at least make sense?  I guess that would be too reasonable and
> well thought-out for RSI.. sigh..
>
> greg
>
> Sent via Deja.com http://www.deja.com/
> Share what you know. Learn what you don't.
>
Agreed

It seems to me that they should have a comp.lang.
specialist that would help them with routine names,
and parameter and keyword names and usages.
Sometimes a parameter is used to signal two different things.


Mirko


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.