Brief article on benchmarks of Python repository with leading DVCSen

[Stephen J. Turnbull]

John Arbash Meinel writes:
 > Stephen J. Turnbull wrote:

 > > Yup.  If you do this script (the echos should be exact, content
 > > matters) in a fresh directory,
[elided]

 > I'd just like to point out that if you do the natural thing of:

 > I believe git's auto-detection becomes even less reliable.

Slightly.  Not much in real, moderate-sized examples as Teemu points
out.

 > I realize the "workaround" is to commit inbetween.

Actually, that wasn't a workaround, that was intended to make the
contrast evident.  If you tell git that it's a move with "git mv" it
will grok that until the commit, but it seems to forget afterward.

 > For git to infer a rename after-the-fact you generally have to commit
 > unmodified texts, so that the sha1 exactly matches.

No, git does fine inferring the renames unless the changes are huge
(> 50% according to Teemu).  

 > the user's intent may differ from what was inferred.

This is the problem.  It is not necessarily true that the majority of
the code of file "A" will end up in in what the user considers the
successor to A (which might not even be named "A"!) after a
refactoring, and only the user knows for sure.  Even tracking
containers and their names is only an approximation.