On 07/04/2017 09:07 AM, Steve Hay via perl5-porters wrote: > On 4 July 2017 at 11:38, Dominic Hargreaves <dom@earth.li> wrote: >> Great stuff, thanks! >> >> I note that the base.pm commit from Aristotle does not appear in >> blead yet (or maint-5.26). It's lesss urgent there given that the >> defaults changed but it still has applicability there given that >> people could unset -Ddefault_inc_excludes_dot. >> >> Of course none of this should block the release of the 5.24.2 or >> 5.22.4, but would be good if it didn't get forgotten - and indeed >> perlpolicy at least implies that all maint fixes should come via blead. >> > I think this was deliberate, e.g. on 17 March Sawyer said (in reply to > Dave M) that nothing else on perl #127834 needs doing for 5.26.0 > (https://rt.perl.org/Ticket/Display.html?id=127834#txn-1453921), and > on 24 March Dave M reverted the various dot-in-INC changes that were > then current in base.pm (commit 6ee05a9b91). > > The latter was done under perl #128769. Originally, the intent there > was to revert the patches thus far in favour of Aristotle's > forthcoming patch, but then it ended up going like this: > > Graham Knop: >> Can't we just revert the base.pm changes without putting anything >> extra in place? We're already removing . from @INC globally. There's >> little need for base.pm to have its own protections, especially given >> how broken the current implementation is. > Sawyer: >> This makes sense. >> I spoke to Graham more about this on #p5p. There's no need to have an altered base.pm with 5.26.0 (because safe @INC is turned on by default anyway). When base.pm is released separately, it will simply be a different (and up-to-date) version. >> So we should undo the changes to base.pm in blead. Post 5.26 we could sync it with CPAN. > However, I agree that it would make sense for the final base.pm > changes to be ported forward to 5.26 (and blead, pending a new CPAN > release of base?) because of the possibility of people unsetting the > default exclusion of dot in @INC. > > If there is agreement on that (or at least, no disagreement) then I > will go ahead with that shortly, when I get things moving on 5.26.0. I agree. Thank you, Dominic, for raising this, and Steve, for handling it.Thread Previous | Thread Next