On Thu, 11 Feb 2016 08:04:16 +0100
Gergely Polonkai <gerg...@polonkai.eu> wrote:

> even though the SHA1 commit ID can be looked at, a changing commit ID
> doesn't mean changed code. What if you hat two commits since the last
> build, one that adds a change and another that reverts it? You are
> seeing the same code, yet you rebuild it.
> If you don't have a decent build system that can do incremental
> builds, you may want to stick with git diff. You already have the
> previous build's commit ID, so you can do git diff old-id HEAD. If
> anything changed, do a build. If you pass --stat to git diff, you
> will get the list of changed files, so you can make a decision if you
> really need to build (e.g. no build if only the README file
> changed.). This technique may be actually applied to other VCSes.

I'd vote for this answer, too.

To the OP: it worth keeping in mind that in Subversion, revision
numbers are repository-wide, so to say that if a revision has changed,
the content of a branch has changed, too, is not really correct.

You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to