PSC #057 2022-03-04

Porters,


*PSC #057 - 2022-03-04***
Present: the whole gang: Neil, Paul, Rik

trim

We talked briefly about trim(). The short version is that we reaffirmed our support for "it's trim, it takes one value, it returns the new value, it is optimized for when the call is effectively edit-in-place." If this is too contentious, it can be held until 5.37.0.

experiments and when they end

Neil (at least) will reread the content of the thread on this and see what edits may be useful to make to perlpolicy. We think proposed changes to the text will more productive than fishing for comments on ideas.

One thing we want to do is to clarify that we need some flexibility to move faster when safe. For example, if a tiny change would allow an otherwise-stable experiment to be made a stable feature, we would like to do that without a year-long delay.

things to get into v5.36.0

We talked about a few things that may or may not be ready for v5.36:

 * source::encoding -- I think this will have to wait, as we think the late updates to the proposed design are probably better, but need more work.

 * bidi delimiters (like `q❲Foo bar.❳`) may get in as an experiment, but clearly we need more discussion about the choice of behaviors, and rjbs has been lax in pushing this along

 * the behavior of assigning non-ASCII to $0 — we think the proposed changes are right

 * the builtins for started-life-as-type-X — Paul is thinking about and working on it

 * `indexed` — rjbs hopes we can get this in, but it needs implementation and nobody has yet put themselves on the hook


-- 
rjbs

• PSC #057 2022-03-04 by Ricardo Signes
• Re: PSC #057 2022-03-04 by Oodler 577 via perl5-porters