Re: Digital and Ada ?

[kenner@lab.ultra.nyu.edu (Richard Kenner)]

In article <3d48s8$4af@cliffy.lfwc.lockheed.com> l107353@cliffy.lfwc.lockheed.com (Garlington KE) writes:
>If I understand Robert Dewar's correspondence, it could actually run
>on both VAX/VMS and Alpha/VMS, just that the Alpha would have to run
>slower that it could since it would be emulating the VAX instruction set.
>However, I also got the impression that porting to VMS would not be
>quite as easy as Unix, although it could be done.

No, you misunderstand.  I don't think Robert was proposing running VAX
code on Alpha (though I suppose that *is* an option).

There are two issues:

(1) Given a GCC port to a VMS target, there are some system-dependent
issues, such as filename syntax, that have to be addressed to port
GNAT to that system.

(2) There is a GCC port to VAX/VMS, but none currently to Alpha/VMS.
The latter port is neither trivial nor exceedingly complex.  So far no
one has comitted to doing it but someone (you know who you are!) told
me at Tri-Ada that he wanted to do it.


So, GNAT for VAX/VMS requires only (1) above to be done, but GNAT
for Alpha/VMS requires both (1) and (2).  Note that when (1) is done,
that part will also be usable for Alpha/VMS and so once that GCC
port is done, GNAT would then run without any further work.