Re: Passing shell arguments to Ada programs the Linux way

[anon@att.net]

Ada is defined as an "Operating System independent" language. Which means 
if it works in Linux it MUST work for all Operating Systems not just Linux.

And most want their programs to work in Linux, Windows, etc. So, for Ada 
users we process the "command line" by the definition set in the RM by 
using the package "Ada.Command_Line".

In other way, is a waste of time for Ada users.

Note: As for "GNAT.Command_Line", well it can only be found in Adacore's 
GNAT Ada. Which some will say it not a standard RM Ada package, so why 
use it. because it might disappear next year. Which means the programmer 
may have to support that package that they do not fully understand. The 
problem with Adacore's non RM packages (GNAT.*, and a lot of System.*) 
is that they are to "Operating System" dependant. 

Also, "Ada.Command_Line" will work for Linux 3.x. Will these others 
packages, that's the question. This is the problem when designing packages 
for Linux at the end of a major design steam such as Linux 0.x, 1.x and 2,x 
and now 3.x. 


In <m2zkkwy9l0.fsf@pushface.org>, Simon Wright <simon@pushface.org> writes:
>A famous IT technical writer <Philippe.vouters@laposte.net> writes:
>
>> You may find a solution on how to pass shell arguments to an Ada
>> program the Linux way. The code has been tested on Linux. As it makes
>> use of the GCC suite, the code is usable under any operating system
>> which implements the GCC suite.
>
>GNAT users could also look at GNAT.Command_Line.