At the time when we had Wagon as the storage mechanism for the backend
as well, I would have said switching them was enough.
I'm not sure if the current code is 100% encapsulated in the
repository layer, but some refactoring around that, particularly
making sure artifact storage is transactional, and then providing an
implementation that wraps and "commit" with a checkin would do the
trick.
Alternatively, a cron that crawled a checkout and did any necessary
add/deletes on a regular interval, but that would not capture every
change. NEither of these take into account updates from an external
source.
I don't really know that subversion is the right tool for the job here
- we're largely storing unchanging binaries - timestamped backups
would make more sense here I think. Is that the use case they have?
Cheers,
Brett
On 12/03/2008, at 8:42 AM, Wendy Smoak wrote:
What would it take to use Subversion as the back-end for Archiva?
I noticed we have JSPWiki incubating now, and they can do it...
http://www.jspwiki.org/wiki/SubversionProvider
--
Wendy
--
Brett Porter
[EMAIL PROTECTED]
http://blogs.exist.com/bporter/
