Yves Orton

@jkeenan why did you add a "do not merge" to this /draft/ PR? Doesn't that kinda defeat the point of it being draft?

@kid51 the point is that is a note to me, before I change it from draft ill probably split some of those patches apart, and I will certainly remove a...

> I've had a look over it, but I'm not especially familiar with the regexp engine. Your feedback is already super useful, thanks. I need to think on what to...

@iabyn - fair enough. I thought I had explained well enough in the PR documentation. I didnt bother in the commits because I am fairly certain I am not doing...

I have updated the PR by breaking up the main changes in the PR. The two patches I would really like some feed back from @iabyn are https://github.com/Perl/perl5/pull/20677/commits/ff43150a169745a295340ff18960efbbfa839be4 and https://github.com/Perl/perl5/pull/20677/commits/99429446d08cb71cbcb3eb05d36d88e6a95a0c86

@PhilipHazel could you please check this over? This is an attempt to make Perls capture buffers behave more consistently. It hopefully brings PCRE and Perl into agreement in more cases.

Patches from this PR have been pushed independently as: https://github.com/Perl/perl5/pull/20703 https://github.com/Perl/perl5/pull/20704 https://github.com/Perl/perl5/pull/20705 https://github.com/Perl/perl5/pull/20706 https://github.com/Perl/perl5/pull/20707 As they get merged I will rebase them away from this PR. For now they are...

> This looks like a really nice enhancement. > > The only thing I saw is that some of the values are cast, either redundantly, or inconsistently. So ` arg=...

Id like to +1 this feature request. I always forget the recipe to say "use parallelism, use ccache, etc etc", which makes builds and tests vastly faster. EG, during install...

FWIW, I am with @happy-barney on the question of whether undef is a valid bool. We only introduce "true" booleans to perl recently, and for most of the existence of...