[xquery-talk] some thought about NoSQL query processing in the form of a talk

[Ihe Onwuka]

Some observations, mostly inspired by your other posts which are still
giving me food for thought.

History always has valuable lessons to teach us. If you look at the NOSQL
landscape in terms of players it is not dissimilar to the landscape in the
early relational era (which includes the non-relational vendor).

Many of those companies had success, rapid growth and then either got too
uppity or made a strategic error and were acquired, often  by Computer
Associates (Cullinet, ADR, ingres) but IBM swallowed up Informix and Sybase
became SQL/Server. The few  vendors survive standalone today that I can
think of, Software AG, Cincom and Information Builders were mostly set up
in late 60's early 70's. There is no reason not to expect  a similar
rationalisation (of both product and vendor) in the NoSQL arena. Commercial
considerations aside too much of it is built on the two biggest hype
machines of the moment - Big Data and JSON.

JSON advocates have been allowed to present the format as an alternative to
XML and more appropriate in certain use cases (I forget which because I
don't really care). This is not true (actually it's a big fat lie) because
in many domains (e.g where precision is required, wherever there is
regulatory oversight or where your use of data could lead to an expensive
lawsuit) JSON is simply not an option. I don't think this is widely
appreciated yet but if and when the tech bubble bursts the money will be in
servicing these areas that are no go for JSON .

RDBMS took off when the biggest kid in the playground IBM jumped in with
DB2. IBM made the mistake of leaving the relational market on Unix and
Windows markets open to other vendors (like Oracle), but it's questionable
if there would ever have been an Oracle if IBM had not paved the way for
the acceptance of relational products with DB2. One thing IBM never did is
bet against it's own technology which is what MarkLogic seem to be doing by
with it's support for Javascript.

True times are different and IBM did not have to cope with the offerings of
an open source movement  but a moments thought will reveal some possible
potholes here. We all heard the mutterings against MarkLogic because of
difficulties with the (Oba.....Affordable Care Act) website and inviting
every Javascript coder off the street to build on your platform (because
that's effectively what they have done) could lead to other bad PR
debacles. As the XML community have found with JSON PR machine, it doesn't
have to be true, or even justified, it just has to stick.






2015-05-30 14:13 GMT-04:00 daniela florescu <dflorescu at me.com>:

>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://x-query.com/pipermail/talk/attachments/20150531/cdeca046/attachment.html>