> From: Brent Dax [mailto:[EMAIL PROTECTED]]
> Henrik Tougaard:
> # > ------------------------------
> # > CORE PLATFORMS
> # > ------------------------------
> # > OpenVMS (Alpha)
> # > ??
> #
...[snip]...
> As a stopgap measure,
> 
>       ($^O eq 'VMS' ? " /Include=[.include]" : " -I./include")
> 
> ought to work (or something like it--I'm not a VMS guy).
It does work. We get a tiny bit further....
The same call to $c{cc} has a 'hand-written' -o flag. That's
wrong again - it should be "/output=" instead.
By fiddling I can get past the compile, but then it fails
miserably at the point where the test-progam is to be
executed (in backticks). This will fail on VMS for two
reasons:
  1) a separate link has to be called
  2) running programs is done by the 'RUN' command, not by
     just writing the program name.

> You may not be aware, but on [EMAIL PROTECTED] they're discussing this
> sort of thing--apparently mm[ks]'s (the make equivalents) Makefile
> syntax is sufficiently different that it has to be done differently.

Not only Makefile syntax - everything is different. Which makes it a good
check for portability.

The entire configure script for perl itself was rewritten for VMS, hopefully
that wil not be neccessary for parrot. Unfortunately I don't have the
neeeded tuits for the task.
--
Henrik Tougaard, Copenhagen, Denmark
"The first rule of Perl Club is, we don't talk about Perl Club"
    - Dave Cross at YAPC::Europe 2.0.01 

Reply via email to