Re: A little different approach to discovery

["Bill Kearney" <wkearney@syndic8.com>]

--- In syndication@yahoogroups.com, Roy Osherove <fireride@n...> 
wrote:
> Hi folks.
> I don't pretend to know very much about this problem, but here's a 
little
> something I've written that talks about solving syndication from a 
different
> direction

While what you propose is certainly do-abe, but not all sites lend 
themselves to doing things in that manner.  I'd go so far as to say 
perhaps not even a statistically significant minority.  

There have been quite a few past threads on how a machine processable 
HTML page alone would eliminate the need for anything like an RSS 
file.  Yes, technically that's correct.  But the hurdle presented by 
that is so high as to be nearly impossible for sites to overcome.  

That and there are actually good reasons to use a feed instead of 
such a page.  A feed will undoubtedly be smaller.  It could contain 
different information.  Think of how a magazine table of contents 
often has descriptive paragraphs that aren't actually in the 
referenced articles (teasers, call outs, etc).

There's also the issue of economy.  Generating a static RSS file is 
quite scalable.  Generating reliable regexes and having them work 
reliably comes nowhere near the likelihood of success a feed would 
offer.

So yeah, for sites that are prepared to be 'perfect' it might work.  
But for the rest of us things like feeds really fir the purpose quite 
nicely.

What your article does suggest is that if people undertake the time 
to make their site /very/ machine processable there may well be some 
excellent possibilities.  If we could get that message across then 
we'd really have something!

-Bill Kearney
Syndic8.com