[darcs-users] Re: definition of "depends on"?

[Adam Megacz]

"Yitzchak Gale" <gale at sefer.org> writes:
> Let x, y, and z be changes such that x and y do
> not commute, but x and z do commute. Let -y
> be the inverse of y. Let
>
> A = {x}
> B = {y}
> C = {-y, z}
>
> Is this the case you are looking for?

Yes, it is.  Thank you!

So, I guess the conclusion here is that in its current form, darcs is
going to suffer from a lot of false dependencies -- situations where
it is possible (under the theory of patches) to [un]pull A without
[un]pulling B, but where [the current implementation of] darcs does
not notice it.

Anecdotally, I've run into this quite a few times now; at least now I
understand why darcs' notion of "depends on" is so much more
conservative than one would hope.

Are there any plans for darcs' future evolution which might change
this?

Perhaps darcs could offer a command which would let the user
explicitly ask it to work very hard (ie O(n^2) or worse) to find a
commutation that removes B from A's context (the user specifies A and
B).  But I can see why optimal context minimization is too expensive
to do for every patch.

  - a

-- 
PGP/GPG: 5C9F F366 C9CF 2145 E770  B1B8 EFB1 462D A146 C380