Front page | perl.perl5.porters |
Postings from February 2022
PSC #055 2022-02-18
Thread Next
From:
Neil Bowers
Date:
February 20, 2022 23:24
Subject:
PSC #055 2022-02-18
Message ID:
8ce9dcc5-c8a4-4b52-a2b7-aa0b14baca0d@Spark
Most of the hour was spent talking about releases.
Release Schedule
We decided that we'll aim for a 5.36 release on 2022-05-20, which leaves us the following schedule:
• 5.35.9 20th Feb: contentious changes freeze (Renée, done)
• 5.35.10 20th Mar: no changing user-visible behaviour of correctly functioning programs (Sawyer)
• 5.35.11 20th Apr: total code freeze, apart from crashers, regressions, and security
• 5.36.0 20th May: it's soup
We are looking for a volunteer to do 5.35.11.
Things we think need resolving before 5.36
Paul's de-experimentalising signatures pull request (https://github.com/Perl/perl5/pull/19424) has now been merged.
Nick's PV vs IV pull request (https://github.com/Perl/perl5/pull/18958) needed resolving at the time of our meeting. I don't really understand it all, but my summary of the discussion would be: "it doesn't fix everything", "it wasn't intended to, it moves us forward". We nudged the discussion, and Nick has now merged it into blead.
Karl's implementation of Rik's source::encoding (https://github.com/Perl/perl5/pull/19406) needs wrapping up.
Rik wants to continue the evolution of the idea that started with the `kv` RFC and has written a new RFC to do so.
Changing the behaviour when 5.36 sees code that has `use v5.12` (or any version from then up to 5.36) followed by `no strict`.
De-experimentalising variable-length look-behind (`experimental::vlb`). See https://github.com/Perl/perl5/issues/18756.
De-experimentalize :const subroutine attribute, which was added in 5.22 (see "Constant Functions" in perlsub).
We discussed builtin::trimmed, and agreed that it should go in.
Less pressing: documentation updates and release notes for 5.36.
pragma for using experimental features non-experimentally
Let's say you want to use the bitwise feature in your CPAN module. It was experimental from 5.22 to 5.26, and became non-experimental in 5.28. But retrospectively, from 5.22 onwards it was fine. So it would be nifty if you could include it with something like
use trustworthy::feature 'bitwise';
This would let you use it in Perls from 5.22 onwards. But if it had changed in 5.26, and that version ended up being the one that was trustworthy, then the pragma would only let you use it from 5.26 onwards.
This could be done by dual-lifing the feature pragma, but that would make it a lot hairier. We decided that this would be better started off as a CPAN pragma, to see if there is much interest in something like this.
Neil
Thread Next
-
PSC #055 2022-02-18
by Neil Bowers