Unfortunately the planned meeting is in my holiday. I have not been very present on the list recently, my excuses - many other obligations - but I will try to keep track of the future query project happenings.
I thought maybe an old document I wrote (of not very high status, on a early version of MMBase when there was no taglibs and all usage was Scan) could be useful input for part of the problems to be addressed, which would definitely not be found in the archives unless I shamelessly plug it again: http://www.cwi.nl/~arjen/mmbase/mmci-18102000.pdf (see http://www.mmbase.org/archive/developers/msg00632.html) What I'd like you all to consider is whether a language like XQuery/XPath (extended with ranked text query functionality) could be used to specify all your queries against the cloud directly, without iterating over intermediate result sets. If not, I'd be very interested in examples that are difficult; if so, using XQuery/XPath might be a good basis for addressing the query language problem. Now I'm off for holiday, Arjen
