> Have to ever looked at Jira's database structure? The XML is > nothing more than a reflection of their structure. Tables, > keys, normalized relations, etc.
Yes, it is very normalized. I've done a few hacks myself to make our transition from bugzilla a bit better, and find it was quite easy to do with SQL. I would think doing so in XML is going to be fine as long as you know which are primary and foreign keys and I have found that to be fairly obvious. If you do an import from XML, JIRA isn't going to care that you've twiddled all the IDs, and as far as users are concerned they only remember the keys which remain the same. > > We possibly need a way to do it consistently as other projects join > > ASF and bring their previous issue tracking with them too. > > Jira needs a good import/export tool, but that doesn't help > with the near term. I think their focus is on end-user tools right now, and rightly so. > Jira seems like it could use additional administrative > features. Pruning tools, for example. We will have over 25K > users after importing bugzilla. Ideally, we would like to be > able to purge users who aren't associated with any issues, > who don't have any special permissions, and haven't logged in > for some period (say a year). This raises another issue - how are you going to extricate all of the non-ASF projects from the codehaus export? I'm not trying to throw a spanner in the works here, I'd love to see the Apache bugzilla move to JIRA, but in a practical way that inconveniences the least number of people and projects. Cheers, Brett
