develooper Front page | perl.perl5.porters | Postings from January 2020

Re: mononomial versus binomial branch names.

Thread Previous | Thread Next
From:
demerphq
Date:
January 31, 2020 15:44
Subject:
Re: mononomial versus binomial branch names.
Message ID:
CANgJU+Vq6wAWLW-YdKSvfqnFT-VUe1xArFJp+2TdfdO5DHxMBw@mail.gmail.com
On Fri, 31 Jan 2020 at 16:39, James E Keenan <jkeenan@pobox.com> wrote:

> On 1/31/20 10:00 AM, demerphq wrote:
> > I noticed that we have a few mononomial branch names. For instance Todd
> > pushed 'Tie-File-strict'.
> >
> > In theory there is nothing wrong with this, but in practice it is asking
> > for trouble with collissions etc, not necessarily knowing who to ask
> > about a branch, etc. At $work we have a rule that all such branches
> > should be binomial, usually with the first component being the "owner"
> > of the branches name/username. Eg a small selection is as follows:
> >
> > ap/baseincguard-old/3
> > ap/perlfunc-do
> > arc/smaller-toke
> > atoomic/installdocs-man
> > autarch/modernize-perlopentut
> > avar/no-UNIVERSAL-import-routine-TODO
> > avar/perldisthist
> >
> > Most of our branches follow this convention, and we might even have had
> > a push check to prevent pushes that weren't binomial as we copied a lot
> > of our work rules to the perl5 git repo when we hosted it. If so then I
> > assume we lost this when we migrated to github.
> >
> > Personally I think we should follow this convention, and ideally github
> > should enforce the rule.
> >
> > Thoughts?
> >
>
> I don't see a strong need for this at this time.  We have remarkably few
> such mononomial branch names.
>
> #####
> $ git branch -a | ack '(remotes/origin)' | ack -v '(blead|HEAD)' | sed
> -e 's/^  remotes\/origin\///' | grep -v -E '^(maint|release)-' | ack -v '/'
> 133800-carp
> Conf-qah2016
> Tie-File-strict
> add-smoke-d64ld
> aix-doubledouble
> corion-5.15.8-zlib
> devel-ppport-reorder
> goto-doc-fix
> grok
> guts
> improved_charclass_generator
> locale
> maint
> mauve
> newmetaconfig11
> no_hv_fill
> perldelta
> post-5.28
> smartmatchtoo
> snprintf_l
> strtod_l
> wcrtomb
> #####
>
> I think the one you cited was just an oversight on toddr's part.


Sure, I wasn't trying to beat Todd up, I just happened to notice it when I
saw his branch.


> Yes, I
> do think any committer should include their handle in the branch name.
> But, no, I think git hooks and such are overkill for this.
>

Ok. Fair enough, but personally i think if we want to have a policy we
should have it enforced mechanistically...


>
> Several months back I proposed (off-list) to the pumpking that we
> conduct a branch cleanup project, but I haven't received the green light
> for that.  IMO that would be helpful.
>

I have started trying to clean up some of mine, so I am in.

Yves

Thread Previous | Thread Next


nntp.perl.org: Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at ask@perl.org | Group listing | About