[xquery-talk] [xml-dev] Mistakes made in the design of XQuery 3.1

daniela florescu dflorescu at me.com
Sat May 9 14:25:04 PDT 2015

> You can be assured that we had lively debates on this topic,

After “living” in the XQuery W3C working group for 15 years….. I don’t doubt you had. I am just happy I was not there.

But, sorry, W3C took the wrong decision: in one instant with this decision, you lost the JSON community.

> and there were good arguments to be made both ways: none of them at all silly.
> We were aware, also, that XQuery has two equality operators, and Javascript also has two, and none of them quite do the same thing. We didn’t want four equality operators. 

Do you really think that XQuery has only TWO equality operators ? 

No. Only two are in the external syntax. Every user I know of XQuery adds more as functions, depending to his/her own application needs.

So, two, four, six, ten, what’s the difference ? (it’s like that saying “the wife had only two lovers..it’s not that bad..four would have been really bad...” )

When you see that it is more then two, three, then you make it parametrable.

Making null a separate value would have enabled any programming language which is a consumer of JSON (and they ALL are..) to add their OWN semantics, 
according to their wishes.

It’s called flexibility, and making things parametrable, (especially for something as important as equality for semi-structured data, whose semantics is in the eye of the beholder),
instead of rigid, monolithic, authocratic design, which invariably makes 90% users unhappy.

But that’s only one of the things that I think XQuery 3.1 got wrong. There are others.

The major OTHER problem is that it hadn’t been designed (as JSONiq did..) to be a syntactic superset of XQuery 3.0 , but which can be subsetted NICELY into a 
JSON-only language that would have been syntactically and semantically pleasant to a JSON-only community.

For JSONiq this was one of the major, major design goals.

Too bad, because such W3C decisions have implications on splitting the XML and JSON communities, and THAT, has implications on decades of technology……and billions
of dollars lost in industry by unnecessary frictions and fragmentation of the market.


More information about the talk mailing list