develooper Front page | perl.perl5.porters | Postings from July 2022

PSC #071 - 2022-07-01

Thread Next
Paul "LeoNerd" Evans
July 8, 2022 17:03
PSC #071 - 2022-07-01
Message ID:
PSC #071 - 2022-07-01

Present: Paul, Philippe, Rik

## Some questions from Philippe

We started with some question notes Philippe had pre-written:

  * Should `perlgov.pod` be updated with PSC?

    It already is, effectively. The version that's in blead just isn't
    released yet until 5.37.2. It's slightly unfortunate the way timing
    works, means that perlgov.pod in a real release is always going to
    be out of date. Perhaps the document should contain a link to
    "latest" in revision control so people can look for more up-to-date

  * Would "inclusion of Corinna in core perl" be a good condition to
    consider a version bump to Perl 7?

    It could, but there's no timing guarantees yet. Rather than set a
    target for "Perl 7 should be ...", best to just continue working on
    features and at some point say "This'll do for 7". Though in
    practice we need a bit more of a leadup towards it, because bumping
    that version number is nontrivial and lots of things would break.

  * When would it be considered good to nominate new core members?

    Any core member can nominate new members. Just do it.

## Review of past actions

  * Rik - on `source::encoding` - might be best to arrange a chat with
   khw at this point and hammer out some details before reporting back

  * Paul - on Yves' UNIVERSAL::import thread - waiting to merge the
    `lexically_export` builtin first, which would then put us in a
    better place to consider the remaining ideas.

  * Rik - empty path and taint. ((I forget what we discussed about this
    one, but I think Rik said he'd follow up.))

  * Paul - tables in Pod. Yes I should write another reply to that.

  * Rik - some handover tasks left from Neil's departure. We still need
    the key to, and Neil needs to send the post he's

## Started work on `use feature 'class'`

Mostly an announcement from Paul that work has begun on the `use
feature 'class'` code, which will hopefully eventually bring a real
object system to core perl.

This did bring up the question on how to handle the "deep internals"
code design questions that will inevitably (or rather, have already)
come up. The RFC process has been good for solving the
Perl-programmer-visible interface sorts of design issues, but we don't
have a good way to solve design questions of internals and inner
implementation of these new large pieces.

A private email discussion between a few interested parties might seem
good, or might draw accusations of "operating in secret".
Alternatively, an open discussion on perl5-porters@ tends to end in
Warnock silence. A good solution to this problem has not yet become

## Backporting `builtin`

Back when `builtin` was being added, we discussed the idea that it
would be dual-life released on CPAN as a way to offer polyfills for
older perls. Now that 5.36 is out of the door it seems a good time to
think about this.

It's likely we won't get the semantics quite right at first, in terms
of `$VERSION` numbering, and how CPAN modules can declare dependency on
particular functions, but as the entire thing is still all marked
experimental, we can iterate on that to find where the problems are and
how best to solve them.

Paul "LeoNerd" Evans      |  |

Thread Next Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at | Group listing | About