I compeletely agree with that. It would be a complete waste to have chad and others' work from the past 5 months thrown away at the last minute for a solution brought up this late. We can initially use git for the migration however if we decide to later, Phabricator is still available. :-)
Indeed. We should continue to evaluate phabricator, and see if it fits our model better than Gerrit in the long run. Also, we should start keeping a list of things phabricator needs to be usable (like we are doing with Gerrit).
- Ryan