Re: [Oorexx-devel] RFE #446 Enhancement to rexx.cpp

[Rony G. F. <Ron...@wu...>]

Given the name "SysCArgs" already implies that this is not a [oo]Rexx language related piece of
information, but one of the runtime environment that started the Rexx script. As the discussions on
<news:comp.lang.rexx> showcase, there are people (especially those with a Unix background, but now
more and more realize the same possibilities to take advantage of under Windows as well) who would
be helped, if all of the C argv was made available to them. The prefix "Sys" already indicates that
this is something outside the Rexx language, but valuable for programmers.

However, if there is a name used like SysCArgs then it implies that a Rexx programmer gets access to
the C argv array. It would not make sense for those users who know about the C argv argument to
leave out the first two arguments, as then this does not represent the C argv. It would add a
surprise, an oddity to that particular function.

Independent of how SysCArgs gets implemented, the Rexx language itself remains totally unfaffected
by it, i.e. Rexx coders, who are not in need of the C args, can very successfully use PARSE SOURCE,
PARSE ARG, USE ARG to get at the information they usually are interested in. ( SysCArgs won't change
that at all. )

---rony




On 03.06.2013 17:03, Rick McGuire wrote:
> I have several objections.  Since this is an array, this will actually shift the arguments to
> non-intuitive positions (there's no way to have an element 0 in an array).  I don't actually see
> that this provides much useful function, and really entangles the language too much with command
> shell issues.  
>
> Rick
>
>
> On Mon, Jun 3, 2013 at 10:55 AM, Mark Miesfeld <mie...@gm... <mailto:mie...@gm...>> wrote:
>
>     All,
>
>     The current implementation of the SysCArgs array purposively leaves out the C arg 0 and arg 1.
>
>     Currently some users seem to be interested in having access to those args.  Well, arg 1 for sure.
>
>     Since the enhancement has not been released yet, it seems to me it would be okay to change the
>     implementation to include all the C args.  I'm in favor of that.
>
>     Is there any objection to changing the enhancement to include arg 0 and arg 1 at this point?
>
>     --
>     Mark Miesfeld
>