On Sat, Mar 16, 2013 at 01:33:42PM -0700, Gavin Sharp wrote:
> On Sat, Mar 16, 2013 at 12:07 PM, Andrew Sutherland
> <[email protected]> wrote:
> > My personal experience in mozilla-central and comm-central was that holding
> > blame/annotate sacred provided a lot of stop energy to improving the state
> > of code, leaving the code generally worse off.
> 
> Indeed. Determining the original source of a particular line already
> often involves looking multiple revisions back. Adding one more
> because of a whitespace fix doesn't make a significant difference.
> There should be no hesitation to fix whitespace due to fear of
> "breaking blame". (Though of course there may be other reasons why
> "fixing whitespace" isn't a good thing, and it should generally always
> be done as a separate changeset).

Also: `git rebase -X ignore-space-change` can be helpful if you have to
rebase across a change that alters a lot of whitespace, but beware that
the resulting state may not be indented correctly and need a fixup; for
me this was a definite improvement over having to hand-resolve spurious
merge conflicts (and possibly get them wrong).

(The thing I needed it for was a large change in Rust that added a level
of syntactic nesting around a lot of code, rather than a whitespace-only
change, but it should be generally applicable, I think.)

--Jed

_______________________________________________
dev-b2g mailing list
[email protected]
https://lists.mozilla.org/listinfo/dev-b2g

Reply via email to