RE: Sticks, carrots, real XML, ultra-liberal parsing

["Danny Ayers" <danny666@xxxxxxxxxxx>]

> "The point is that if the very folks advocating ATOM
> are also advocating parsing tag soup to discover
> feeds and encouraging people to support tag soup ATOM
> feeds then why bother with the pretense that ATOM is
> an XML technology?"
>
> 2 points -
> 1.) Once I started using a tag soup parser in RSS
> Bandit for RSS autodiscovery it was very tempting to
> use the same parser to parse RSS feeds especially
> since evry other week I hear some complaint about how
> RSS Bandit should support malformed RSS feeds like
> other aggregators do.

I note that you're talking about RSS, which (viewed as a variety of specs,
bunch of apps, whatever) does very little to encourage good markup. The same
needn't hold for Atom.

> 2.) Mark Pilgrim is an advocate of parsing tag soup
> ATOM feeds based on his most recent blog posting which
> is counter to the spirit of XML and a detriment to XML
> on the Web. He is also a major advocate of the ATOM
> format.

I wouldn't have thought promoting Postel's law was quite as heretical as
that suggests. There is some conflict between it and XML (Aaron did a nice
clear blog about it) at this point in time, but we're in a very dynamic
environment. The explosion in test-based methodologies suggests to me that
the way we looked at the validity of data 5 years ago won't be the same 5
years hence.

Back to the present - I do think we need more work on the production side.
Your own stance appears to be to accept bad data and not make any effort to
encourage quality data - at least Mark's backing the production of good
stuff as well (see under "tests").

> Hope that is more coherrent.

Parsable ;-)

Cheers,
Danny.