On Fri, Feb 26, 2010 at 08:13:40 +0100, Petr Rockai wrote: > it seems we need a decision whether we should use some sort of > abstraction over "_darcs/foo" (like a global constant, darcsdir and > concatenation, or something a little smarter or whatever).
> I agree that littering unnamed constants around source is more often a > bad idea than not, but I am not convinced about this case. On the other > hand, if there is to be a lengthy flamewar, that outweighs any benefits > not going through this exercise might have. So to avoid bikeshedding, > there's a couple ideas, you may add your own, but preferably cast a vote > in favour of one of these: You put some effort into anticipate a noisy thread, and pro-actively engineered the discussion to maximise our signal to noise ratio. Bravo! > 1) "_darcs/pristine.hashed" > 2) darcsdir "pristine.hashed" > 3) darcsdir hashedPristineDir > 4) darcsdir HashedPristineDir > 5) darcsdir HashedPristine > 6) hashedPristineDir -- includes "_darcs" > 7) darcsdir ++ "/" ++ "pristine.hashed" > 8) darcsdir ++ "/" ++ hashedPristineDir > 9) darcsdir ++ "/" ++ hashedDir HashedPristine > 10) any of 5-7 with </> instead of ++ "/" ++ > 11) Some variation of the naming. (dir_hashedPristine, ...) I don't have much of an opinion, but I'll prefer #10 to #5-#7 as there is probably no good reason to use foo ++ "/" ++ bar when we have foo </> bar instead (is there?). Also, in his darcs branch, David went in and removed the abstraction, presumably because it annoyed him. That said, in the long run, there are still people who want to replace "_darcs" with ".darcsrepo", so we should probably anticipate that. -- Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow> PGP Key ID: 08AC04F9
pgpGOa9KtFZYc.pgp
Description: PGP signature
_______________________________________________ darcs-users mailing list [email protected] http://lists.osuosl.org/mailman/listinfo/darcs-users
