Monday, 21 January 2008

[LazyWeb] using Subversion as a front end to Confluence?

I hack a lot of Confluence spaces as part of my job. I've been working on the UFace project at Google Code so I've been getting used to the Google way.

One of the great things about google code is that the wiki is in subversion as *.wiki files; so you can edit them in any text editor (and TextMate comes with great MoinMoin syntax support, you just have to install the bundle...). So you can grab your wiki content, jump on a plane, hack it with your favourite editor then check in your changes when you're next online.

I wonder how hard it'd be to sync Confluence content with a subversion repo; so folks could still use the online Confluence site to edit content, which would get mirrored to an svn repo as *.wiki files - then folks could commit to the svn repo to update the confluence database? Anyone ever tried?


Don Brown said...

There is the WebDAV plugin, which would expose the content via the same protocol SVN (mostly) uses. One way syncing pages would be easy; two way with collision detection and handling would be tricky.

James Strachan said...

Ah cool - thanks for the heads up Don!

mcannon said...

Mr Poejoe! :) Doesn't SVN have some sort of pluggable storage system? If so, I bet you could backend it onto Confluence via some sort of web service (instead of storing the file on disk).

Know anyone who is good with web services in Java? :)