Traq Project Forums

I've been thinking about moving the Subversion repository off my server over to Google Code for a few reasons.

1. The Google Code servers are more stable and updated more than my server.
2. Extra features.
3. More accessible.

I'm 100% sure if I will move the repository, but if I decide to do so it wont take long, I wrote a shell script that will sync the repository into a local folder, change the commit author name to the one I use on Google Code and then sync it to Google Code.

I did a small test and it worked pretty well, What I have not tested is if it can be used to keep two copies of the repository in sync. For example, take the changes from the current repository and send them to the Google Code one after the initial sync has been completed.

I will post an update when I have decided if I will be moving it to Google Code or not.

I've decided to completely move the source repository to Google Code.

Anyone using the vizziweb repository should switch to traq.googlecode.com/svn

Google Code is a great idea - for visibility and features, apart from stability
IMO, strictly, you could be "committer" and so others send patches and you check them in if you think they're OK. Alternatively, you could keep a "patchy" branch where you just check in if it *looks* OK, it is totally unpredictable, good only for patch viewing, while your "core" branch has all the solid stuff that works ( alpha / beta / rc / stable ).
The other model would be github / dvcs, but then, only if we use SVN can we claim to be "eating our own dog food" .
Just my thoughts, feel free to ignore

I love the idea of being able to send patches for review by Jack, that would make development faster and Traq better and Jack would still have control of his script.