Perl Steering Council

All three of us attended, but none of us had the time for significant discussion, so we decided to reclaim the time and make some progress on our various to-do list items.

[P5P posting of this summary]

All three of us attended. Other than administrivia we talked about formally documenting our supported platforms, and we intend to start a discussion on the mailing list about this soon. This is also a topic for the upcoming Perl Toolchain Summit.

[P5P posting of this summary]

All three of us were present:

• We agreed on not terrible feature names for PPC0027: keyword_any and keyword_all. These can now be merged.
• This led to a brief excursion on the fact that it would be good to have proper iterators as part of the language. But we quickly agreed that the PSC meeting is too small of a margin for that subject.
• We talked about the failure of “experimental features”, and possible approaches to remediate that problem. It is going to be difficult to address technically. Philippe pushed for a PPC about the subject.
• We took note of the upcoming freeze period and surveyed the changes we think should be merged beforehand. The only thing we weren’t already tracking is the coderef-in-stash optimization, which we decided should be reverted once again and re-attempted earlier in the cycle next time.

[P5P posting of this summary]

Back to the full roster, albeit with Philippe attending from on the road.

We mainly went over the list of PPCs, with an eye toward the fact that this release cycle is coming to a close.

• PPC0033 (ampersand method calls) is approved.
• PPC0030 (equ) does not have an implementation ready for this release cycle and PPC0031 (eq:u) is controversial.
• PPC0027 (any/all) is basically accepted other than needing its feature flag naming figured out.

As an aside regarding PPC0027, we reiterated that we would like use feature ':all' to go away if possible. It was never a good idea anyway, but has become untenable with the introduction of feature flags like indirect and bareword_filehandles, which we expect to have many more of in the future. Since their purpose is to be disabled rather than enabled by default, a simple toggling of all features (on or off) is a nonsensical request.

[P5P posting of this summary]

Only Philippe and Aristotle this time.

We spent most of our time on the PPC process, and started by merging Dave Cross’s PR for a static PPC web site. Many thanks to Dave once again.

We discussed revising the PPC process, and started by picking more specific names for the various statuses of a PPC, which we’ll soon apply to the existing PPCs.

[P5P posting of this summary]