Could that be done with this build ?
I think we should do a review there first, there are some features, which are done, others that are 99% done and could be finished, and others, which would require so much work, that they would only delay the release.
I can't decide this alone, this is a team process.
I am pretty sure, we could get to a stable build in a month, if we drop some half-finished features instead of trying to finish them. And it would mean a month without working on new features.
My personal current feeling about this is, that we need about 16 man hours for cleaning things up. With two coders, this would mean two lazy weekends.
---------- Post added at 09:22 PM ---------- Previous post was at 09:15 PM ----------
Apologizes if this is coming through as "outsider talk", but I would advice against using feature branch work-flows with SVN. Even with newer version's "merge-tracking" feature, it just isn't worth the effort.
You are right about the management effort vs coding ratio there, but I think it could solve a lot of trouble regarding the scenarios, since having functional scenarios would be first of all the job of the feature branch maintainer.
Also, the effort with SVN isn't that high. With TortoiseSVN, I need a minute to create a new branch, and merging the branch back is generally not taking longer than 30 minutes, if you merged the trunk into your branch on a regular base.
Also tags are not read-only implicitly in SVN. If you are not taking counter-measures on the server via hooks, users can commit in the tag directory.
Never had found a way to commit to a tag, unless you mess with the server directly. At least in the past two years.