On Tue, Apr 7, 2009 at 10:44 AM, Scott Taylor <sc...@railsnewbie.com> wrote: > > On Apr 7, 2009, at 1:32 PM, James B. Byrne wrote: > >> >> I have run into a minor glitch and would like to know what others >> think. I am working on a test/expectation and as part of the >> process of debugging I am rolling back db:migrations one step at a >> time to discover where the problem was introduced. >> >> My workflow looks like this: >> >> $ cucumber ... feature:xx >> fail >> $ rake db:rollback >> $ rake db:test:prepare >> fail ... you have pending migrations. >> >> Now, it seems to me that while testing one is interested only in >> those migrations that have been applied, not which ones are >> available. It is nice to be notified that unapplied migrations >> exist but I am not sure that one should be constrained to apply them >> to conduct tests. >> >> What do others think? > > If you are using git, it might be wiser to create a branch and reset --hard > back to those migrations. > > Or, use git's bisect, and migrate:reset from a clean db each time
This also might be useful: http://outofti.me/post/90028018/rake-db-rollback-to-common _______________________________________________ rspec-users mailing list rspec-users@rubyforge.org http://rubyforge.org/mailman/listinfo/rspec-users
