[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [syndication] Scraper code?



S. Mike Dierken wrote:

----- Original Message -----
From: "Seth Russell" <seth@robustai.net>

Well I'm a bit beyond my deapth here, but here goes anyway.   How about
a combination of a  <a href="foo?action=subscribe&feed=bar">common
button</a> and a user preference selection at a web site.   Then foo is
the URI of a script that interacts with a user preference for a specific
local or golbal resource.   The script page would display a simple
selection of known aggregators and the user simply selects the radio
button and the resultant post is to the local or global resource.  The
foo could be common to all blogs,  so that the user's preference for foo
could be kept in their cookie and a user would need to select it only
once.   Whould that work?

Would that be similar to each publisher adding a form on the page with a
'subscribe using [list of aggregators]'
The publisher's server could:
a) send the request on behalf of the user directly to the aggregator (don't
know how the user would be identified),
b) send a redirect response to the browser that then navigates to the
Web-based aggregator, the user then fills in something
c) send a response to the user with a Content-Type and body appropriate for
the client based aggregator service.

I don't know how your comment of 'the resultant post is to the local
resource' is done though.

Hey, I'm pretty sure this could be made to work, and work very well. I don't have my cgi script tools together or I'd do it myself. If there is anyone out there with a server who can write a quick cgi script that does redirects and can access cookies? A live implementation will answer all of these questions.

Anybody want to collaborate and get this job done?

Seth Russell
mailto:seth@robustai.net