[xquery-talk] Izzit Bcos I is functional?
ihe.onwuka at gmail.com
Wed Jun 17 04:22:54 PDT 2015
On Wed, Jun 17, 2015 at 8:49 AM, Michael Kay <mike at saxonica.com> wrote:
> > But you can pretty much model the semantic equivalent of JSON data
> without having to use a Schema, namespace, PI or any traverse any other of
> these treacherous terrains. Call it XML the Good Parts (Hmmmm where did I
> get that idea from). JSON is a very expensive over-reaction. There was no
> need to invent another format and then make in non-interoperable with XML.
> Actually, I don’t agree. The cost of doing simple things with XML, like
> configuration files, is far too high (because of all the other things it is
> capable of that you don’t need for such cases) and there was a very real
> need for something simpler for that kind of application.
But that is not the same as saying that the replacement of XML with JSON as
a format for generic data interchange was warranted, and making it not
interoperable with XML was spectacularly short sighted. Now true that
probably wasn't the intended use case. Sadly the lack of interoperability
and the consequent inability to deploy a common set of tools is probably
seen as a virtue in that community.
> Restricting yourself to a subset of XML doesn’t greatly reduce the cost of
> writing code to process it, you still have to use the same APIs.
When they are done implementing everything with JSON they will find they
will have something that will look just like XML, have most if not all of
it's atttendant overheads but be even more expensive as it will have been
re-implemented from scratch.
I'm just waiting for the next straight faced instalment.
What NOSQL needs now is a transformation language.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the talk