The case for small diffs in Pull Requests

[H. S. Teoh via Digitalmars-d]

On Wed, Jul 20, 2016 at 04:39:33PM +0000, Vladimir Panteleev via Digitalmars-d wrote:
> On Tuesday, 19 July 2016 at 08:08:21 UTC, qznc wrote:
> > I would like to see squashed commits in D. History looks polluted by
> > merge commits to me. This is useless for single-commit pull requests
> > at least.
> 
> Without merge commits you can't easily trace a commit back to its pull
> request, which contains the feedback, review, and often a
> significantly better description of the change.

Yes, many a time while git bisecting, I've found that merge commits are
invaluable in providing a "paper trail" of exactly what happened with
the code that caused the problem, even if the PR in question is only a
single commit. Without the github PR number, it would be very hard to
reconstruct the history of exactly what went on behind the change -- any
discussions, rationales, etc..


T

-- 
Без труда не выловишь и рыбку из пруда.