Hm. If it is content, then it should work in such a way that it can be accessed no matter what tries to get it: webcrawler, javascript incapable browser or user

Hm. If it is content, then it should work in such a way that it can be accessed no matter what tries to get it: webcrawler, javascript incapable browser or user with a "normal" browser but JS switched off. If user agent is ajax-capable, then use ajax. Google for "hijax". If it is a webapp, then making it crawlable does not make sense anyway.