Hello Gary, On Friday 27 January 2006 03:30, Gary V. Vaughan wrote: > Morgen Andreas! > > Andreas Gruenbacher wrote: > > On Thursday 26 January 2006 16:22, Gary V. Vaughan wrote: > >>I'll generate a patch after the release. > > > > Why not do it now? There's still enough time left. > > Okay. Attached. The tests are adjusted to use the newly consistent > -P syntax too. I'm off to Holland until Sunday evening now, so > apologies if I drop off the net until then...
we obviously were in a little misunderstanding here: I assumed you were talking only about unifying the existing options, while the patch converts all patch arguments to options as well (e.g., "quilt push some-patch" would become "quilt push -P some-patch"). Interestingly, you did not change "quilt push 5" to "quilt push -N 5" or similar. I see an advantage to always being able to specify a patch with -P when in doubt. I also see an advantage in not having to specify -P for those commands that don't require it right now though: with those commands, a patch name is the parameter that makes the most sense, at least IMHO. People also sure have grown very used to how many of those commands work. What do you think about the following alternative: convert all existing patch options to -P, and keep all current arguments as they are. We could also allow to specify those patch arguments with a -P option as an alternative for those cases where a user is unsure, but I'm not sure that this wouldn't create more confusion than it's worth. The affected commands are: applied, delete, files, graph, header, next, pop, previous, push, refresh, unapplied. In fact I don't really care so much about changing applied, next, previous, and unapplied because I've never really grown to use their optional patch argument, but I do care about the rest. Thanks, Andreas _______________________________________________ Quilt-dev mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/quilt-dev
