Re: [xml-dev] The limits of XML mean the limits of my data world

[Norman Gray <norman.gray@glasgow.ac.uk>]

Michael and Marcus, hello.

On 30 May 2022, at 12:02, Michael Kay wrote:

> Lots of people have invented non-XML syntaxes for XSLT; they have all died a quiet death without achieving any signfiicant adoption.

And I was one such [1], and yes, 'a quiet death' is accurate.

[1] https://nxg.me.uk/dist/lx/

>  I think there are a number of reasons:
>
>  (a) the use of XML syntax is off-putting to beginners but no problem at all to experienced users

I wouldn't say I'm a heavy user of XSLT, but I've been I suppose a moderately sophisticated user of it for a number of years, since XML was new.

I still hate the syntax, and it sets my teeth on edge every time I come near it.  The syntax makes it relatively easy to include literal XML elements in the output document (and that's a relatively rare requirement, for the uses I put it to), but that's the only good thing I can think of to say about the choice of XML for the syntax of XSLT.

I appreciate not everyone has this same reaction to it.

>  (b) using a preprocessor makes development a lot more complicated - diagnostics are less helpful, and helpful diagnostics are much more important to developers than pretty syntax.

For that very reason, the Lx thing I put together includes a (Java) XMLReader implementation which generates SAX events directly from the input syntax, so there was no preprocessing, and the line numbers were correct.

Ah, the loneliness of an itch no-one else has....

Best wishes,

Norman


-- 
Norman Gray  :  https://nxg.me.uk
SUPA School of Physics and Astronomy, University of Glasgow, UK