xml-dev - Re: Frontline report from the Desperate Fgrep Hacker

[John Cowan <cowan@locke.ccil.org>]

Sean McGrath scripsit:

> [John Cowan]
> >I happened to know that the element always appeared on a single
> >line of the file: the start tag, the character data, the end tag.
> 
> How could you possibly know that?

Because the documents were 100% generated by a single
program whose behavior was entirely predictable.
I should have mentioned this in my original post.

> Later on in your post you say that this processing
> mode is "(sufficiently) reliable". Sorry, no use
> to me. I do medical systems. I do e-commerce systems.
> "Nearly right" equates to plain wrong in these
> environments.

"Sufficient" means "sufficient to the circumstances."
In this case, 100% is both necessary and possible.

> John, you are one of this lists gurus on XML. I am
> missing something here right?

*blush*

It's sort of the same point I've been making with
RDF: the fact that RDF has lots of flexibilities doesn't
mean that *your* metadata specification has to exploit
them.  It can be far less general, but if you
design it to be RDF-compliant, then general purpose
RDF tools will also be able to handle it.

Similarly, just because XML 1.0 is a highly flexible
format doesn't mean that all that flexibility *must*
be allowed for.  When you know what you are getting,
you can use faster, less flexible processing models.


-- 
John Cowan                                   cowan@ccil.org
       I am a member of a civilization. --David Brin