Hi Thomas,

Thanks for your positive feedback and the great discussion! Yep North of Europe can bring along really new thinking in the field IT. Linus and Guido for instance did that already in their time :-)

I have made some sort answer to part of your comments in my post to both you and John Ryan. But, since you really about db technology, you could possibly spend sometime evaluating Hannes and colleagues research work (a couple of 2017 to 2020 university-level research communication pdf-s).

I am a dummy in this field. And my catch is more a feeling backed by a couple of successful tests on limited benchmarking. Just basic aggregates and arbitrary raw selections on tables from 100K to 20 millions records, no heavy sql join currently, no recursive operation, no udf-s, no heavy multi-thread sql-calls hammering for a start. But IMVVVVHO, their basic target is just plain right at first. And, even better, they depart from the traditional discussions in this field. This means possibly innovation. I do use this word so often. I remember using the word for smalltalk (yep that's mid-eighies) for python (that's nineties) and rared since that time.

This project - or projects with similar targets - "may" really bring some "very real improvement" in sql engine speed and the way we do maths on computers. I say "may" not "will". But I stick to "very really improvements" ... not continuation of current trends. Currently the so-called bigdata community is at times quite regressive in the way they handle their data. Bringing sql into the(their) loop will help them manage their.... csv-s.

It could in return be helpful to us, IT specialists of management-related data, as well in return as it could provide the warp analytical speed that a couple of us still miss. In my case, replace VFP not with sqlite or a dammed costly sql engine that will NOT deliver or an arbitrary nosql engine - no, no, no!! - but with some sort of duck. Yes it flies better than some think it does:-)

Cheers, Daniel