Eric Kow <ko...@darcs.net> writes: >> - makeBundle2 the_s (mapRL info ps ++ [info t]) to_be_sent to_be_sent >> + makeBundle2 the_s (ps +<+ (t :<: NilRL)) to_be_sent to_be_sent > > QUESTION: do functions like newset2RL do what you want here, or is that > something totally different? Not in this case, since here we cut off everything after the first clean tag. With newset2RL, we would include all of the repo history in the RL.
>> applyPatches :: >> - forall p C(r u t). (RepoPatch p) => [DarcsFlag] -> Repository p C(r u >> r) -> >> + forall p C(r u t a). (RepoPatch p) => [DarcsFlag] -> Repository p C(r u >> r) -> > > QUESTION: Does the extra a serve some kind of purpose here? I'm guessing it > just > snuck in by accident? Uh oh. An accident. Let me see if I can amend those away. >> makeBundle :: >> - forall p C(r u t) . (RepoPatch p) => [DarcsFlag] -> >> + forall p C(r u t x) . (RepoPatch p) => [DarcsFlag] -> > > QUESTION: Same question about the 'x' Same as above. (I guess this was from the time before I sealed off the end-state of common.) > QUESTION: So then there's the question of why we have to seal off the > ending state of the common patches in fetchPatches. Shouldn't we be > able to keep track of the fact that the common end in the same place as > the us/them patches begin? I don't think the type of :\/: allows us to keep track of that. _______________________________________________ darcs-users mailing list darcs-users@darcs.net http://lists.osuosl.org/mailman/listinfo/darcs-users
