[teampractices] Code review social norms

[Gergo Tisza]

On Wed, Mar 16, 2016 at 12:48 AM, Kevin Smith <ksmith at wikimedia.org> wrote:

> I would mention that in some cases, I would prefer to accept the commit as
> is, and then perform minor refactoring, such as changing a name, fixing a
> typo, or rearranging the code. Not only does that clearly separate
> authorship, but it would also encourage those changes to be reviewed by
> someone other than that author.
>

Git log is the public record of what changes happened to the code, and is
regularly read for various reasons (creating changelogs, identifying
commits that caused regressions, looking for relevant changes that happened
to master since a feature branch was split, getting a sense of recent
changes). Lots of trivial formatting commits add an overhead to that. Same
with git blame, which is often needed in refactorings to understand why a
certain line of code is there.

I note that if someone wants to get input about the effect certain workflow
changes would have on developers' productivity, wikitech-l is probably a
better place to ask than teampractices.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.wikimedia.org/pipermail/teampractices/attachments/20160318/fd23ca1d/attachment.html>