Front page | perl.perl5.porters |
Postings from December 2017
On Wed, Dec 27, 2017 at 8:51 PM, Leon Timmermans <fawaka@gmail.com> wrote: > We can stop people from saying unkind things, but how are we going to > prevent repeats of this utter fuck-up, given that this "process" could have > only resulted in a fuck-up? > > Are we going to politely destroy perl > ? > To expand on these points a bit, I think the process failures we had here include: - introduction of a breaking change in syntax late in the development cycle (the cutoff for contentious changes is in less than a month, which gives little time to resolve breakage downstream - so at this point in the cycle such breakage should be *small* and *contained*) - changes were not apparently pushed to a smoke-me branch first, where we could have assessed the amount of downstream breakage and then discussed what to do next (I would suggest that syntax changes *must always* be smoked first before merging to blead) - changes were not committed in a rebased branch, which (as previously discussed) makes bisections more difficult, and also complicates the now-inevitable reversion process - introduction of user-facing syntax changes to a well-known-to-be-controversial feature without discussion of the proposal in the greater community (outside of just p5p) of the new keywords and their function These are all things we should improve in our process. respectfully, Karen Etheridge ether@cpan.orgThread Previous | Thread Next