Front page | perl.perl5.porters |
Postings from January 2012
Re: What to do with POSIX::SigRt?
Thread Previous
|
Thread Next
From:
Nicholas Clark
Date:
January 9, 2012 06:54
Subject:
Re: What to do with POSIX::SigRt?
Message ID:
20120109145410.GK9069@plum.flirble.org
On Sun, Jan 08, 2012 at 10:59:13AM -0500, Ricardo Signes wrote:
> * Nicholas Clark <nick@ccl4.org> [2012-01-03T05:49:15]
> > I don't see merit in keeping it.
>
> What is our path to deprecate SigRT if we choose to?
Depends if we want to kill it dead, or simply push it out onto CPAN much
like Switch or Devel::DProf.
Pushing it onto CPAN seems "nicer", but a bit harder.
Currently POSIX.pm has
tie %POSIX::SIGRT, 'POSIX::SigRt';
I think we'd need to break out the code implementing this, ie that in
package POSIX::SigRT, out into its own .pm file.
For a transition period, to keep things working, we'd then need to have
POSIX.pm tie %POSIX::SIGRT to a bunch of proxy code that loads POSIX::SigRT
but warns you that you should have done this yourself.
Then, after a release or two we drop the proxying code.
Likely, in parallel, one would have POSIX::SigRT be in dist/ dual-life
on CPAN, and going through the "evicting module from core" process.
Ultimately this culls 61 lines of clean but dense code, almost comment
free, and 35 lines of documentation, which end
B<NOTE:> whether POSIX realtime signals really work in your system, or
whether Perl has been compiled so that it works with them, is outside
of this discussion.
which likely *also* means that we can't test it.
> from users, and it would be more polite to deprecate in 5.17.0.
I don't know whether more users would notice because we document it and they
read it, or because code $n levels down starts warning because it didn't
before. In this case, judging by the hit level in the core, on CPAN and
elsewhere (RIP Google Codesearch) there aren't any users.
For this and a few other things, I think it would be best to have a plan for
what's next, mention it in the release documentation for 5.16.0, as a
"future direction", and see if we get any feedback.
(positive, negative or lynch mob)
Nicholas Clark
Thread Previous
|
Thread Next