On Fri, Jul 01, 2005 at 10:27:57PM +0000, Stephane Bortzmeyer wrote: > On Fri, Jun 10, 2005 at 08:07:25PM -0400, > Mark Stosberg via RT <[EMAIL PROTECTED]> wrote > a message of 36 lines which said: > > > To make it a more convincing wishlist item it would be valuable to > > gather a consensus on darcs-users that others are interested, > > OK, good idea, so, please, now that I added details to the idea, could > people on the list tell if they find the idea interesting / mandatory > / stupid / harmless / useless / trivial / funny ? > > http://bugs.darcs.net/Ticket/Display.html?id=368
The trouble is that not all repositories are created by init or get. They can reasonably be, e.g. tar'd, ftp'd, or rsync'd, which would result in two repos with the same id, after which point collisions would be guaranteed. So, while I like the idea of being able to refer to a local patch by its index within the repo, I don't think this is a good strategy globally. However, provided no functionality depends on absolute uniqeness, there is no reason not to use a truncated patch hash as a unique id, since the collision probability is low enough to be quite useful. I explained more here: http://www.abridgegame.org/pipermail/darcs-users/2005-April/006695.html -- Jamie Webb _______________________________________________ darcs-users mailing list [email protected] http://www.abridgegame.org/mailman/listinfo/darcs-users
