We are going to completely re-implement the search engine for eZ publish 4.0. Most probably it will be a part of the eZ components. The work will start in a few of months and specifications etc. will be discussed on our mailinglists.
None of the well known search engines (including mngosearch, openfts, ...) is really up to the task we need. But Google sparked quite some activity in the academic world and some concepts/algorithms may be useful to implement in a new search engine for ezp.
But it's a major effort to reach a stage that makes users happy with ezp search. I have a feeling that we even need a paradigm change on search & navigation ...
I'm far to be an expert in the search engines, but if your new paradigms involve good (french) wine and (belgium) beer, I'd be quite happy to investigate a wee bit more ;)
