On Thu, Jan 27, 2011 at 08:47, Echlin, Jamie <jamie.ech...@credit-suisse.com> wrote: > Afternoon, > > We've enforced a rule whereby tags cannot be copied. I know this sounds > backwards but I need some help in justifying it. > > The reason why we're preventing it is because it's not clear where to merge > back to, and in fact encourages merging back to the tag, which is also > blocked. > > What we're telling people to do, in the event of a bug where they need to > make a patch to a released version, is find the rev the tag was created > from, then branch whatever branch the tag was created from at that rev. Then > when the work is finished they can merge back to the parent, if necessary. > > Of course this only works if the tag was created from a rev and not a > working copy, but we're enforcing that too. > > Sound reasonable, or are we being too cautious?
Sounds overly cautious to me. And what happens if the branch you tagged from is ended? What I have done in the past is to copy the "flawed" tag to a branch, make my changes in the branch, then merge back to trunk. My process starts w/ code changes in trunk, then creates tags each time I promote to an environment; trunk to developer testing, developer testing to integration testing, integration testing to production.
