My impression is that the better class of desktop RSS readers can handle
a feed that comes in with http authentication over an https connection,
but that the web-based readers (such as Bloglines) can't at the moment.

The problem would be that you'd have to trust a remote web service with
your id and password (or else reenter them every time you accessed
Bloglines). We've built an RSS feed that presents everything on your
Sirsi MyAccount page (charged items, fines, holds) in the form of an RSS
feed, using Cocoon's screen-scraping functions. This makes it easy to
embed your barcode and pin in the url - not exactly secure, but not much
less secure than sending the barcode and pin in clear text over http,
which is what happens when you access the native Sirsi MyAccount page.

But what if you put this in your Bloglines account? It becomes part of
their social-bookmarking world, so the books I've signed out become
searchable ... So we haven't rolled this out, we're still thinking about
it. We also have an http-authenticated https proxy page in front of
Cocoon that we can use, but we haven't decided what to do with all this.

Peter


> -----Original Message-----
> From: Code for Libraries [mailto:[log in to unmask]] On
> Behalf Of Walter Lewis
> Sent: Friday, January 28, 2005 08:15 PM
> To: [log in to unmask]
> Subject: Re: [CODE4LIB] Individually Customized RSS feeds
>
> Edward Iglesias wrote:
> > Well there have been some postings recently on SIRSI adopting RSS
> > technology.  See
> >
> http://www.theshiftedlibrarian.com/archives/2005/01/19/sirsi_breaks_op
> > en_the_rss_flood_gates.html
> >
> > Eric Lease Morgan wrote:>
> >> If I understand the concept, I have tried to do this as a
> part of the
> >> (incomplete) Ockham Alerting Service. The Service allows
> you to query
> >> an index, and the search results are then available as
> HTML, email,
> >> or an RSS feed. When the content in the index gets
> refreshed, the RSS
> >> feed will return different results every time. The operative word
> >> here is "when". See:>>
> >>   http://alert.ockham.org/
>
> I think I *get* the notion of search results as RSS feed.
> And I respect Eric's "when" qualifier.  I even like some of
> the things being suggested in the SIRSI announcement and
> discussion posted on Shifted Librarian.
>
> What I was particularly thinking of was stuff that we would,
> in the pre-Spam/phish era of email, have only considered
> emailing (and, in many cases, still do of course):
>         hold notifications
>         overdue notifications
>         profile based SDI reports (e.g. new books or articles in *)
>         targeted announcements (by location, status, etc.)
> When dealing with individual patrons, I'm assuming that we
> need to authenticate in order to respect (or at least
> acknowledge) their right to privacy. (how much privacy with
> email is a different thread)
>
> Could any/many of the different classes of RSS readers deal
> with a standard web authentication scheme? standalone?
> embedded in email clients? embedded in browsers?
>
> Walter Lewis
> Halton Hills
>