[Info-vax] [OT] Wirth style languages, was: Re: Obscure Ada compiler vendors?

[glen herrmannsfeldt]

Bill Gunshannon <bill at server1.cs.uofs.edu> wrote:
> In article <kjknjk$orj$3 at speranza.aioe.org>,
>        glen herrmannsfeldt <gah at ugcs.caltech.edu> writes:
>> Bill Gunshannon <bill at server1.cs.uofs.edu> wrote:
 
>>> I wish they still taught COBOL at a later point and probably in a special
>>> course, but I see no reason to cover Fortran in anything but a general
>>> Programming Languages Course where you show them differnt languages and
>>> the pardigms attached to them but no one becomes proficient.
 
>> Maybe, but Fortran today isn't anything like the one from 25 years ago.
 
> IMHO hacking OO onto languages desigend for an older paradigm is just
> plain wrong.  Fortran had a specific purpose.  Most o fthe work doing
> that today needs to be very parallel.  I have not wroked much with
> Fortran beyond 77 but I odubt it is optimized for parallel computing.

Well, they did add OO, but so far it hasn't seen much use.

But Fortran 77 didn't even have any kind of dynamic allocation. 
You had to dimension arrays as big as they would ever need to be,
and then usually use a smaller part of an array. So, dynamic allocation
was one of the major additions. Array expressions, structures, 
many more built-in functions, and many other features of more
modern languages.

Fortran 2008 has coarrays, a feature meant for parallel programming.
I don't know about optimized, but coarrays do make programming easier.
I am not yet convinced that it is the best answer to parallel 
programming, though.

-- glen