RE: Auto-discovery revisited

["Danny Ayers" <danny666@xxxxxxxxxxx>]

> On 09/01/2004, at 10:06 AM, Danny Ayers wrote:
> > 99% of blogs may be tag soup. They will stay soup if that's the only
> > expectation. If, on the other hand, tool developers are encouraged to
> > build
> > tools to produce good markup, then that figure could shrink rapidly.
>
> This is precisely the argument that was used for RSS newsreaders to
> reject invalid feeds.

Essentially, yes. But I'm not suggesting that anything reject invalid data,
merely that the production of good data should be encouraged. On a
rudimentary level, if people think they have something to gain from
producing good data, they're a lot more likely to do so. Atom can offer a
lot. The mistake of "reject invalid RSS feeds" is that it's a stick. What we
need are carrots.

And before that, I remember watching long
> arguments over the lengths to which Mozilla should go to interpret bad
> HTML In the end, RSS newsreaders do their damned best to interpret
> everything that they find, web browsers try to be as bug-for-bug
> compatible with their predecessors as they can be, and Mozilla
> implements the damn "marquee" tag.

There isn't a single RSS spec, the specification does little to encourage
validity. A good validator only became available late in the day (thanks
Mark & Sam!). Given these factors it isn't surprising that newsreaders have
to cope with a lot of poor data. These are all factors which Atom can
address.

However I would suggest that there are considerably fewer invalid RSS feeds
than there might otherwise be because newsreaders generally aren't as
flexible in their handling of dodgy data as HTML browsers. It's just not
easy, as Dare described.

I can't find the post, but Bill K. said something recently along the lines
that when Syndic8 notified people that their feeds were bad, the
overwhelming majority of producers did something about it. This suggests
that people can be receptive to such ideas.

> The power lies with the content-producers. Atom is not a big enough
> lever with which to move the ill-formed web.

We don't need to consider the whole of the ill-formed web; only that in
blogs. Given the list of blog tool producers that say they will support Atom
[1], I'd say Atom did have considerable leverage in this domain.

That there's no way we can get 100% coverage doesn't matter. My point is
that by saying "we'll do our best to consume anything" without "we'll do our
best to produce quality data" reduces the possibility of interop.

Cheers,
Danny.

PS. It's been a while since I last saw a <marquee> tag, I'm pleased to say
;-)

[1] http://www.intertwingly.net/wiki/pie/RoadMap