On 2005.10.14 15:22, Adrian Irving-Beer wrote:
On Fri, Oct 14, 2005 at 07:50:55AM +1300, Martin Langhoff wrote:
> No -- GIT doesn't only look for a filename match, that'd be naive.
> Once it finds teh file, it tries to figure out if it's the same one
> (SHA1 matches or a high similarity index). Same for identifying the
> rename: it looks for new files in that commit with a high
similarity
> index.
Ah, joy, so it makes an educated guess instead.
I fail to understand why the developers would go to such great
lengths to develop fuzzy matching, which is *going* to fail by
definition in certain circumstances, just to avoid adding a single
command where the user explicitly records a rename and gets 100%
accuracy.
Because fuzzy matching is an "interesting" problem. Any fool can
implement a rename command, and they just aren't any fools.
If the only solutions out there are going to make a policy of
assuming I'm that lazy and trying to second-guess my every move,
then thanks, but no thanks. I'll pass on the silver hammers until
someone comes up with a (better) real toolset.
_______________________________________________
Gnu-arch-users mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/gnu-arch-users
GNU arch home page:
http://savannah.gnu.org/projects/gnu-arch/
_______________________________________________
Gnu-arch-users mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/gnu-arch-users
GNU arch home page:
http://savannah.gnu.org/projects/gnu-arch/
