On Wed, Jun 12, 2013 at 07:53:30AM +0200, H.Merijn Brand wrote: > On Tue, 11 Jun 2013 22:44:06 +0100, Tim Bunce <tim.bu...@pobox.com> > wrote: > > > On Tue, Jun 11, 2013 at 04:09:13AM -0700, H.Merijn Brand - Tux wrote: > > > Branch: refs/heads/f_dir-ref > ^^^^^^^^^^^^^^^^^^^^ > > > I'd *really* like features to be developed on feature branches and only > > to land on trunk, in a single merge, once they have reasonable test > > coverage. > > Isn't that exactly what I am doing now?
Ooops! I'd *really* like to pay more attention to commit messages :) > FWIW, I finished some tests in DBD::CSV yesterday, and the basic > thought seems to work as intended. What is left to do is rework the > tests into tests that can be used in DBI itself, improve docs and > examples and merge back to master (there is no trunk anymore :) :) > In hindsight, the name I chose for the branch isn't that great. Do we > need a kind of standard, like feature_something? That by itself might not help much. The goal should be to communicate the purpose and scope to someone who isn't familar with the branch. Something like "dbd_file_multiple_dirs" expresses the scope and purpose reasonably well. Tim.