On Tue, May 01, 2018 at 01:00:54PM +0200, Ævar Arnfjörð Bjarmason wrote:
> So didn't we use diff heuristics to begin with, and then regressed? I've
> only given this a skimming, but it's useful to have that sort of
> historical context mentioned explicitly with commit ids.
Sorry for not making this too explicit: I traced wt-status.c to its
beginning in c91f0d92ef ("git-commit.sh: convert run_status to a C
builtin", 2006-09-08). Here I lost track of other changes - but the
commit you gave is earlier also has rename detection hard coded in the
status command. Should I add this as a starting point instead of the
commit mentioned so far? And this also seems like the very beginning of
git status.
The point I wanted to make, is that git show showed you renaming of
files out of the box whereas other commands did not. At least I remember
this form 5+ years ago.
This changed with 5404c116aa, so that the out of the box behaviour is
the same, but this is more a coincidence that the hard coded flag is the
same as the default configuration value.
My comment was targeted at the hard coded rename detection flag - the
other two just have seem to pile up on that and I wanted to clean them
up, too. Maybe a phrasing like "While at it, also remove the two other
hard coded values concerning rename detection in git status." is better?
Is my intent clearer now?
Greetings,
Eckhard
