Jeroen Demeyer on Tue, 11 Jan 2011 22:13:06 +0100

[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: Path to perl

On 2011-01-11 12:18, Karim Belabas wrote:
> @perl@ is used whenever the script will be run on the end user's machine
> (to provide extended help and prettyprinting, respectively). 
> All other scripts are maintainer-only : they generate files to be
> included in the release, but the scripts themselves no longer need to be
> run by the end user to build the package.
Thanks for the clarification.

> Our releases assume that the person *installing* PARI on a given target
> machine has a sane environment, in particular that perl can be found in
> *his* PATH. Then the extracted scripts can be run by other users,
> themselves possibly plagued with badly broken configurations.
Okay, I understand your point.  However, for Sage I would guess that
being able to move binaries around is probably more important, so it
might make sense to use #!/usr/bin/env perl for Sage.

> This ticket apparently refers to a bug in sage packaging : our gphelp 
> doesn't start with #!/usr/bin/perl but must be extracted from
> at Configure time (@datadir@ must also be positionned correctly;
> @version@, on the other hand, is unimportant).
There actually is a mistake in the bug report: Sage builds PARI using
PARI's Configure (with some modifications), the person making the bug
report just looked at doc/gphelp and didn't realize it was auto-generated.

Considering @datadir@ in gphelp: am I right to say that the value of
@datadir@ doesn't matter if the environment variable GPDOCDIR is set
(which is the case within Sage).