On Tue, Jan 27, 2009 at 12:59:06PM -0700, Mark J. Nelson wrote: > This defers discussion of -r. Like Bill, I think that webrev should > behave as hg diff. Because, after all, what is webrev, other than a way > to intuitively display diffs?
To me, it's a way to intuitively display changesets, which is slightly different. It's the duality inherent in any graph -- nodes and their connections can be interchanged depending on how you look at the problem. I tend to consider the changeset the primary unit in the history of a repository, and (probably erroneously) think of the diff as being a part of the changeset, rather than the path between the two. Obviously this breaks down a bit when a changeset has two parents, but that happens so little in my workflow that it hasn't caused me to rethink my world. I don't have an issue with reworking the way the tools work with the way most people think (I'll miss the quick'n'easy formulation for MQ, but I hope to be moving to pbranch soon, anyway). I just want to make sure that most people actually do think one way or the other before I make it final. I can change it in the future if I have to, but I'd rather not. Danek _______________________________________________ tools-discuss mailing list tools-discuss@opensolaris.org
