Re: Formal definition of HTML5

Hash: SHA1 (Ian Hickson) wrote:

>It is a huge problem, IMHO, because most developers aren't 
>competent by that definition. Also, I _want_ my tools to catch 
>as many errors as possible. Having the spec artificially limit 
>what errors can be caught seems like an unnecessary limitation. 
>(And if we do have a spec schema, and it doesn't catch 
>everything, you know people will claim that conformance 
>checkers that catch mistakes the spec schema wouldn't flag are 
>buggy and are reporting bogus errors.

The classic case being for SGML Validators reporting things not 
defined in the DTD because the relevant criteria was not 
expressible using that “schema”.

However, nothing prevents — in fact it's quite common in IETF 
specs — the normative definition in a schema language (EBNF 
for the IETF case) being augmented by prose descriptions when 
the schema language makes the constraint awkward or impossible 
to express. A simple note informing the reader that this is the 
case is usually sufficient for these exceptions.

There is no good reason why a normative definition in a schema 
syntax should not be accompanied by good prose and the two are 
certainly not mutually exclusive options (quite the contrary).

- -- 
I have lobbied for the update and improvement of SGML. I've done 
it for years.
I consider it the jewel for which XML is a setting.  It does 
deserve a bit of
polishing now and then.                                        
- -- Len Bullard

Version: PGP SDK 3.8.1


Received on Monday, 16 April 2007 22:38:41 UTC