I've been concerned for a while that, although I do generate both P5 and PDF versions of all articles and reviews and store them in svn as a matter of course, I'm not able to do that for XHTML versions; and in addition, the structure of the ancient site, based as it is on query strings containing ids, is not friendly to archiving tools. I've now started creating a static XHTML build process which will hopefully result in archivable versions of the web pages and a better URL structure (which should presumably be integrated into the old webapp in the medium term, so that archiver URLs are rewritten). Ultimately, we want to have a rich static archive we can pass on to any future editor at another institution.