[darcs-users] *practical* differences between darcs' patch model and git/mercurial's?

Florian Weimer fw at deneb.enyo.de
Sun Oct 21 20:48:20 UTC 2007


* Adam Megacz:

> Sartak <sartak at gmail.com> writes:
>> The patch algebra does a lot to make that calculation less
>> conservative. If two patches can commute, they're totally independent.
>
> Are you sure?  I'm almost certain that I've run into situations where
> two patches commute, yet darcs believes that one of them depends on
> the other.  I can probably construct a case if this is considered a
> bug...

This is a bug.  Of course, there can be dependencies at the darcs level
that are not there at the semantic level.  For instance, both patches
might touch the same dependencies line in a Makefile, adding different
source files.


More information about the darcs-users mailing list