Re: [xml-dev] processing instruction with 'xml' target

[<juanrgonzaleza@canonicalscience.com>]

Tatu Saloranta said:
>
> So what relevance does it have if parser documentation
> and/or error messages use wrong terminology? Or even
> if browser's understanding of what a PI is was indeed
> wrong (which it may or may not be -- web browsers
> don't care, xml is as irrelevant to html browser
> rendering as html is for xml specification).
> Either people who wrote those were sloppy, or
> misguided; big deal.

From a theoretical point of view, there is a subtle difference between
'anything' and 'representation of that anything'.

From a practical point of view, there is big relevance. As explained in
some of quotes i introduced. If the parser i am interested treat <?xml ?>
as a PI, sorry to say this, but it is said in the formal spec will be of
none importance FAPP.

If you think that difference wetbeen a formal spec and a real world
implementation of that spec is of no relevance, then take a cofee with
some CSS hacker and talk about the 'old times' when MSIE interpreted the
box model in a non-normative way.

> That doesn't change the
> specification at all: with xml, specification is not
> just de jure standard: there is no de facto deviation
> of it. There is no big schism over "what exactly is a
> processing instruction".
> The specification very clearly explains what is a PI,
> and what is not.

Since i already remarked the difference between using a formal spec as
arbiter and a real implementation and since i already said was a PI
according to the spec, not need to reply this again. Try to read i already
wrote... no wait better try to understand i said.

>
> -+ Tatu +-
>
>

Juan R.

Center for CANONICAL |SCIENCE)