<div class="gmail_quote">On Mon, Sep 5, 2011 at 15:26, Barry Smith <span dir="ltr"><<a href="mailto:bsmith@mcs.anl.gov">bsmith@mcs.anl.gov</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
Come on, it is total bullshit (as Jed notes) that these systems change the mtime of files they don't need to touch. That is moronic, I don't care how much "easier" it makes their life in writing hg and git.</blockquote>
</div><br><div>Both implementations "expect" the worst case (a conflict), so they update the working tree so that conflicts can be resolved in-tree.</div><div><br></div><div>I agree with resolving conflicts in-tree, but I think they should attempt the rebase without touching the working tree and only modify the working tree in case of a conflict. I don't know how difficult this would be to implement, it depends how rebase is currently implemented and what data structures are available.</div>