Re: RSS vs. NNTP
- From: "Steven E. Harris" <seh@xxxxxxxxx>
- Date: Thu, 14 Apr 2005 19:06:32 -0700
"Brandon J. Van Every" <mylastnameruntogether@xxxxxxxxxxxxxxxxx> writes:
> Clearly, blogs could 'push' to subscribers.
Yes, but consider the cheerleading for RSS as an improvement over
pundit-pushed mailing lists. These were "pull-once" by way of
subscription, then push thereafter. People were always complaining
that the messages were spam, couldn't filter them into buckets, and
couldn't figure out how to unsubscribe. In other words, too many
people either use inadequate tools to manage their incoming e-mail or
don't care enough to figure it out.
> They could even be less specfically subscribed, if someone offered a
> front-end aggregation service for many blogs.
You just invented Bloglines: http://www.bloglines.com/myblogs
That's the only way I use RSS. I hit that site a couple of times a
day, seeing what amounts to my own little site-of-sites. Bloglines
does the polling, pulling, caching, and bookkeeping.
Like the transformations mentioned elsewhere in this thread, it takes
Web sites referred to by RSS and turns the collection back into a
single Web site. If there were multiple Bloglines-like sites I polled,
I'd probably have to tie those together again too.
> The Usenet Big 8 most certainly is. The mechanism may be
> distributed, but the view is highly centralized.
True. But even beyond Gmane, it's possible for specific news servers
to host small sets of groups, even if other servers don't propagate
those groups. I concede that the overhead in getting one's writing
onto a news server is much higher than sticking Web pages and
attendant RSS files on a Web server, or pouring one's heart into some
Roach Motel system like Blogger or LiveJournal.
> By all means, show us these web browsers that package up exactly the
> same functionality of scouring the web for content of interest.
I was referring to features in older versions of Internet Explorer
(and perhaps Netscape) around 1997 that allowed one to designate
certain bookmarks as "live" or something. The browser would keep track
of the last update time for the resources pointed to by these
bookmarks. Based on some periodicity or upon explicit command, the
browser would poll for each bookmark to see if the respective resource
had been updated. If so, the browser would highlight the bookmark and
perhaps download the resource, caching it for later viewing.
Again, it's not quite the same thing as RSS, but it's partway
there. An author publishing a periodic journal would have to call each
updated entry by the same name, such as "current.html", perhaps
redirecting to a more stable name for the actual entry.
> I think the problem here is we're talking at cross-purposes, with you
> taking the role of Techie and me taking the role of Suit.
Guilty as charged. That's my assumed mindset in a group like this.
> You are saying, RSS protocol is not a particularly interesting
> technology.
Picking nits, and only because I want to say something complementary
right afterward, RSS is not a protocol, but rather a data format. It's
usually fetched over HTTP and has certain semantic assumptions about
what the consumer intends to do with the data conveyed, but it's more
akin to, say, RFC 822 than NNTP.
On the upside, RSS is one of the first embraced examples of separating
data from the consuming application. By that, I mean that there's a
great variety in style and intent of the applications that consume
RSS. Sure, mail and news clients can vary greatly in looks and
features, but they all exist with more or less the same intentions.
Fetching binaries with a special purpose NNTP-aware program may be a
counterexample.
> I am saying, RSS is a particularly interesting social nexus for
> marketing and community formation.
I can't speak to that because I'm not quite sure what you mean. Can
you elaborate? It does sound interesting.
> You're probably vastly more interested in tech than world conquest.
Yes, conquest is my /next/ project, unless something more pressing
comes up first.
--
Steven E. Harris
.
- Follow-Ups:
- Re: RSS vs. NNTP
- From: Brandon J. Van Every
- Re: RSS vs. NNTP
- From: Ulrich Hobelmann
- Re: RSS vs. NNTP
- References:
- Why Lisp instead of Scheme?
- From: David Steuber
- Re: Why Lisp instead of Scheme?
- From: Tayssir John Gabbour
- Re: Why Lisp instead of Scheme?
- From: Christopher C. Stacy
- Re: Why Lisp instead of Scheme?
- From: Ulrich Hobelmann
- RSS vs. NNTP (was: Re: Why Lisp instead of Scheme?)
- From: Brandon J. Van Every
- Re: RSS vs. NNTP
- From: Steven E. Harris
- Re: RSS vs. NNTP
- From: Brandon J. Van Every
- Why Lisp instead of Scheme?
- Prev by Date: Re: OT: Newsreader Killfiles
- Next by Date: Re: Web Application Frameworks in Smalltalk and Common Lisp
- Previous by thread: Re: RSS vs. NNTP
- Next by thread: Re: RSS vs. NNTP
- Index(es):
Relevant Pages
|
Loading