# Nathan J. Williams 2006-07-04:
> "Garrett D'Amore" <garrett_damore@tadpole.com> writes:
> 
> > Again, I'm not beholden to XML, but I _abhor_ the idea of creating
> > another one-off, non-extensible special purpose file format.
> 
> XML makes me grumpy because it "solves" the problem of low-level
> syntax but doesn't help at all with the higher-level semantics. As far
> as I'm concerned, saying that XML makes for a uniform file format is
> about the same as saying that ASCII makes for a uniform file format
> (but with more complexity).

Exactly. That's what I meant by "disguising *new* format as existing
one".

> If you're limiting the XML to some kind of key-value dictionary, you
> still don't know the legal sets of keys and values, and you get into
> trouble when you want more grouped data; if you aren't so limited, you
> have an entire hierarchy to explain, with an awful lot of visual
> clutter (I don't like that XML forces everything into a tree
> representation. Not everything wants to be a tree).

Agreed.

> XML documents need a DTD to be interesting; given the possible
> complexity of a DTD, this is kind of like saying that a config file
> needs a parser to be interesting. You haven't helped anything by
> moving the problem down into a thicket of angle brackets.

Agreed. What we may be missing are a few well-written parser libraries
people could readily use instead of inventing their own. If the choice is
well thought out, I think we can come up with like two or three formats
that would suit most needs...

	-- Jachym

BTW: we're getting off-topic here -- what about moving to tech-userlevel?