There's some creeping data problem in Confluence that occasionally results in duplicate pages with the same name. While it would be great to fix the root cause of this bug, it's been happening since 1.0.x and we still can't reliably reproduce it.

We really need to work around this. If someone calls getPage() and multiple pages are found, Confluence should pick one of the those pages to survive, and insert the others into that page's history somehow.