Proper tracking of file-level operations: rename, directories, merges

[John Szakmeister]

On Wed, Oct 26, 2011 at 10:23 AM, Barry Warsaw <barry at canonical.com> wrote:
> On Oct 26, 2011, at 09:13 AM, Mark A. Flacy wrote:
>
>>I'm curious about the use case where this information is considered useful.  I ask because I honestly can't think of one.
>
> Refactoring is the most important one that comes to mind.  Say I have a file
> that I want to split into two files.  Ideally, I'd `bzr cp foo bar` and then
> do the respective edits.  The value in having a native copy function is that
> I'd now have history on both files, going back to the original creation point,
> and being shared up to the point of the copy.

This is *exactly* the scenario we've hit a few times (maybe 12-15
times over the last year).  In our case, we were using bzr against
svn, so we'd do the copy in svn, and keep on trucking, but it was
painful.  We wanted to keep that history, so it was really our only
choice.  No, it didn't come up often, but when it did we jumped
through hoops to do it.

-John