On Tue, Nov 27, 2018 at 8:13 AM Augie Fackler <r...@durin42.com> wrote:

>
>
> > On Nov 27, 2018, at 11:09, Pulkit Goyal <7895pul...@gmail.com> wrote:
> >
> > I am +1 on using black.
> >
> > Adding to what Augie said, we plan to preserve the annotate history by
> commiting the changes with something like "# skip-blame" in commit message.
> We can then skip those revisions while annotating using the `--skip` flag
> of `hg annotate` which accepts a revset, something like `hg annotate --skip
> "desc('# skip-blame')"`.
> >
> > Maybe we should use "# skip-blame black-changes" to mark those changes
> as black specific and to differentiate between py3 and black changes. (We
> already use "# skip-blame" for py3 related trivial changes.)
>
> Indeed. Note that if we do the black transition, we should probably also
> do the mass-addition of b prefixes on our string literals at the same time
> (in a separate commit, obviously) so that people only have painful rebases
> once and not twice.
>

IMO the annotation should be "# auto-reformat" or something along those
lines: you should be annotating the action that was taken, not the
side-effects you purport you want to take from it. When skipping revisions
for annotate, we can use a revset with all the relevant annotations. e.g.
`desc("# auto-reformat") | desc("# py3-compat")`. But we've already started
using "# skip-blame" so perhaps this ship has sailed.

Speaking of ships that have sailed, I would prefer we use "annotate"
instead of "blame," as the former doesn't have the negative connotations of
the latter.

I agree with Augie that a mass-addition of b prefixes should be done at the
same time in a separate commit from reformatting.

I have no preferences whether we should have a single massive commit or
split things up by file.

I also have some issues with black's formatting. I dislike black's use of
double quotes for strings (maybe a holdover from my knowledge of shell,
Perl, and PHP, where different quotes matter). And I wish it didn't collect
imports on the same line. But my objections are insignificant compared to
the benefits of having consistently formatted source code. The research in
this area shows that the actual style matters less than consistency. If you
really care about a style, you can configure your editor to apply your
personal source format on file load and the official style format on save.
You don't get to do that on remote code review, hgweb, or when looking at
diffs from `hg` output. But it's a mitigation if you care that much.
_______________________________________________
Mercurial-devel mailing list
Mercurial-devel@mercurial-scm.org
https://www.mercurial-scm.org/mailman/listinfo/mercurial-devel

Reply via email to