[darcs-users] identical change conflict?
deliverable at gmail.com
Tue Mar 22 05:44:38 UTC 2005
As I described above, I'm arranging for SCM tracking, whereby several
distributed SCMs coexist in the same "local" directory. When you pull
changes from a "remote" directory with similar coexistence, the first
SCM gets to truly update the source files, while the others just
overwrite them, while updating their metadata.
I've noticed a trange difference in darcs' behavior when it doesn't go
first: since the "new" local file is identical, even tla with its
inode tracking (but in pristines only) does not notice any difference,
as there's really none. Yet darcs, when it in its turn gets to pull,
marks that same single test line I change in "local" as a conflct!
Effectively, it adds a row of 'vvvvv' above the line and a row of
'^^^^^' below, no '*****' as they're the same. I have to do idiotic
moving-stuff-around out of the pull way to prevent this.
Is there a semantic reason for the conflict? Yes, the changes are
temporally different and are equivalent "just accidentally", but there
ain't no textual conflict. No matter how you look at it the file is
the same as it was but the time stamp... And if there's a reason, is
there an easier way to hide it from darcs?
More information about the darcs-users