I've seen some projects move to github for code repo and some have started using GH issues; the only thing that really works IMO is to keep the old tracker available in read-only mode and just get going, migrating from one to the other always leads to a loss of information because GH issues is a rather "poor" environment when it comes to features. I think that's also the least work. Optionally an issue could be copied to github by hand when a fix is being made. Some examples: - OpenLayers set up switched their Trac to read only and started using GH issues, - GeoTools persisted in using their Jira which results in a slightly convoluted workflow when doing PR's, also the relation between issue and code is up to the author. - PhantomJS dit a bulk import from google code to github loosing a lot of context. Personally I find GH issues to be generally sufficient, it even allows for things like code sprints, plan board sessions etc. through the API and services like waffle.io, huboard.com The big strongpoint for any tracker for me is two-way integration between issues and code commits, GH makes this really easy Mark -- Disclaimer; This message is just a reflection of what I thought at the time of sending. The message may contain information that is not intended for you or that you don't understand. -- DokuWiki mailing list - more info at http://www.dokuwiki.org/mailinglist