| Gordon Sande 2004-06-25, 7:43 pm |
| On Thu, 17 Jun 2004 06:04:47 -0300, Catherine Rees Lay wrote
(in article <eUFvO0Ov6V0AFwId@spamtram.polyhedron.com> ):
> In article <2jcvanFvukarU1@uni-berlin.de>, Madhusudan Singh
> <spammers-go-here@yahoo.com> writes
>
There is also a question of relevance for the typical application of the
language. Fortran when used for simulation of an oil reservoir needs
efficient calculation with little need for precise control of the
nonexistent interactive screen of of the rounding modes of currency
conversion. Cobol will control conversions well even if it is less
efficient in calculation or screen control. Basic may control the
screen well even with its low efficient computation and poor control
of conversions. In short, different tools for different tasks.
The add-on GUI toolkits address secondary uses.
> As I understand it, one of the major reasons that this sort of stuff
> isn't standardised is that for it to be part of the formal standard,
> either
>
> a) it has to work on every system where Fortran programs will be run
> (and clearly stuff like graphics and sound won't)
>
> or
>
> b) the standard has to define "standard behaviour" for every single case
> where the system doesn't support a particular functionality.
>
> This is fine for individual vendors - they can choose to only supply
> their toolkits on systems where the functionality exists, and are free
> to say "routine x doesn't work on OS y". But you can't do it in an
> official standard.
>
> Some vendors do support standards other than their own for graphics
> (OpenGL being the one I'm most familiar with) but I'm not aware of
> anything similar for GUI controls.
>
> Catherine.
>
|