Gurusamy Sarathy <[EMAIL PROTECTED]> wrote:
> I haven't looked at the rest carefully enough to comment, but this
> caught my attention as it scrolled by:
> On Thu, 18 Oct 2001 09:46:42 EDT, Charles Lane wrote:
> >+++ perlvars.h Wed Oct 17 14:29:00 2001
> >@@ -44,3 +44,6 @@
>> #if defined(VMS) && defined(Drand01_is_rand)
>> PERLVAR(Gmy_inv_rand_max, float) /* nasty compiler bug workaround */
>> #endif
> >+#if defined(VMS)
> >+PERLVAR(Gvmsish_flags, int) /* VMS-specific global flags */
> >+#endif
> You want to put this inside intrpvar.h. Almost no variable other
> than mutexes shared across interpreters, and constants should be
> in perlvars.h. The same probably applies to the PL_my_inv_rand_max
> thing as well.
PL_my_inv_rand_max really is a constant, it's just that it has to
be set at runtime to work around a compiler problem.
But for the vmsish_flags, you're right and I'll see about changing it.
> And you might want to allocate VMS-specific things as
> PL_sys_intern.stuff, rather than PL_stuff and init then via
> init_os_extras().
That sounds like a good way of doing it, thanks for the suggestion
--
Drexel University \V --Chuck Lane
======]---------->--------*------------<-------[===========
(215) 895-1545 _/ \ Particle Physics
FAX: (215) 895-5934 /\ /~~~~~~~~~~~ [EMAIL PROTECTED]
