One of the many potential feature requests in the queue for Friki is the ability to export the content as a single archive. I'm sure most of us have encoutered problems when a hosting operation "goes dark" or simply gets too flakey or expensive to continue using. These thoughts were in the back of my mind when I read Lasse's plea (also here if/when that blog goes away) for information about getting his posts out of JRoller. I also noticed that Simon Brown wrote that he had recently added such a feature to Pebble.

This is also one of my main worries about the continuing effort to create a new Java forum solution for Java Ranch. JavaRanch has tens of thousands of registered users and hundreds of thousands of messages tangled in its old UBB directories. Migrating to a new solution is not just a matter of writing a little web application. The commmunity is the users and messages, and without an effective migration strategy it could all be a big dissapointment.

All of this is interesting, but without some sort of interoperability standards, actually moving any significant quantity of such content between systems would be a nightmare. Entry naming, grouping and sequencing of entities, format and markup, in-content local references between entries, entry metadata, user metadata, change histories ... There are a huge amount of things that would need a non-trivial "porting" effort.

Several blog, wiki and forum systems currently offer a "better than nothing" export feature. Currently Pebble and Friki both store their data in regular, plain-text files (or at least XML), and "zipping" then into an archive is relatively easy. But what about the ones that use databases? or that use a binary markup, or that format entries as they are created and store complete HTML pages ...

I write something on the internet pretty much every day. Blog entries, posts at JavaRanch and other forums, edits to Wiki pages, changes to my own web sites, etc. etc. I worry more and more about losing all the effort I've put into on-line content creation over the years, and would love to be able to "back it up" somehow.

So. If anyone knows of any standards (or work being done toward such standards) for any sort of application-neutral external representation of such page/message storage systems, I would love to hear about it/them. If not, I may just have to start such a process myself...