Broken HTML Support

Fri Feb 10 21:48:05 UTC 2006

On 2/10/06, Joseph Holsten <pantosys at gmail.com> wrote:
> In particular, what's the position about handling
> broken html, because Josh thinks it should be handled, I think it
> doesn't meet the spec, and should be ignored.
[...]
> I'm not saying we should demand the page have a <!DOCTYPE /> or pass
> the validator without any  errors, but a "SHOULD be well-formed"
> seems acceptable, if not implied.

Joseph,

I think it's perfectly reasonable to reject broken HTML. I also think
that it should not be forbidden for a consumer to accept broken HTML,
provided that the consumer is careful to only accept the <meta> tag if
it's definitely in the <head> of the document. I see this as a best
practices question and not a specification question; broken markup is
implicitly out-of-spec because it's not "really HTML." The question is
basically "how out-of-spec is it OK to accept?"

The only things that I feel strongly are:
1. The spec should not take a position on what a relying party does
with broken markup -- except perhaps mentioning that it has to be
dealt with somehow, because:
2. A relying party must be careful to avoid parsing <meta> tags that
are not in the HTML head, regardless of whether the document is valid.

I'm glad that someone besides me is thinking about this :)

Josh