On Fri, Oct 30, 2009 at 10:15:53PM +0100, Jean-Philippe Bernardy wrote: > As I try to point out, this is not as bad as it sounds, if you make GUID > local. > To be more specific, the hunk gets a GUID. Each element in the Hunk > gets an implicit GUID: (GUID, index in the thunk). Hunks are attached > at a given > (GUID, index in the thunk).
Described this way, it seems reasonable to use the hash of a hunk as its GUID? (Correct me if I'm wrong, but it seems like the GUID based representations of hunks minimize the number of times the hunk needs to be rewritten in order to be contextually correct -- e.g. line numbers -- right?) --nwf;
pgpnq9WErA15T.pgp
Description: PGP signature
_______________________________________________ darcs-users mailing list [email protected] http://lists.osuosl.org/mailman/listinfo/darcs-users
