[darcs-users] Formal documentation

Michael Olney mpolney at gmail.com
Fri Feb 4 05:08:52 UTC 2011


<ganesh at earth.li> wrote:
> I think the important pieces are an appropriate set of definitions, a set of
> properties that all patch system should obey, a proof that the existing
> implementation of "primitive" patches in darcs obeys those properties, and a
> proof that anything that obeys those properties does indeed behave
> consistently when you commute and merge patches in it.

It may be my own ignorance at work, but it's not clear to me how much
mileage you would get out of this scheme. The basic properties of
patch theory seem to be extremely general. Consistency with these
properties alone doesn't appear to tell me much about how the system
is going to behave when I use it.


More information about the darcs-users mailing list