Comments on: RSS Bandit mis-using RSS guid?

By: l.m.orchard

l.m.orchard — Mon, 31 Jan 2005 01:47:29 +0000

Sam: The phrasing “URL of the item” goes back to mixed usage of RSS– that is, some RSS feeds are lists of commentary on links pointing at other sites. And, more recently (relatively speaking) RSS feeds contain and refer to full-blown, independant pieces of content.

The “URL of the item” in the first case refers to a news item (say from CNN or the BBC) about which the RSS feed’s author was commenting.

The “URL of the item” in the second case refers to the full-blown piece of original content.

If I recall, it’s this mishmash of uses that called for the GUID tag’s isPermaLink distinction.

By: Sam Newman

Sam Newman — Thu, 23 Dec 2004 09:47:42 +0000

Well, Is described as the “URL of the item”. If isn’t a permalink, then at some point it _stops_ being a link to the item, therefore is no longer correct according to the specification.

I can understand the being used as some internal marker – and certainly some aggregators use as a way of determining if a post is new or not which seems perfectly sane. However I’m trying to think of a situation where I would use a as a permalink, but have it point to something other than what the is pointed to.

By: Brent Simmons

Brent Simmons — Thu, 23 Dec 2004 00:35:39 +0000

To answer the question “if we already have , why do we need ?”

It’s because there’s nothing that says that the is a permalink. It may be, but it may not be. It may not even be unique in the feed. It might even link to some other site entirely.

The tag addresses the problem of item identity in a way that doesn’t. When a guid is also a permalink, that’s even better — because otherwise there’s *no way* for an aggregator to know for sure what the permalink for an item is.

By: Phil Wilson

Phil Wilson — Thu, 16 Dec 2004 10:19:59 +0000

Luke Hutteman (author of “SharpReader”:http://www.sharpreader.net ) made a post last year about this issue with a real-life use case. “http://www.hutteman.com/weblog/2003/03/14-51.html”:http://www.hutteman.com/weblog/2003/03/14-51.html

Hopefully this is all fixed in Atom ;)

By: Sam Newman

Sam Newman — Thu, 16 Dec 2004 09:48:39 +0000

Confirmed that yes, bloglines also incorrectly uses my feed’s guid. I’ll file a bug with them also. As for @guid@’s working after a domain move, I’m struggling to think how this would work in the real world, but can at least understand it. I still think the RSS spec is horribly bloated though :-)

By: Sam Newman

Sam Newman — Thu, 16 Dec 2004 09:36:56 +0000

Just confirmed, this happens in rss bandit 1.2.0.117. I’ll report a bug.

By: Sam Newman

Sam Newman — Thu, 16 Dec 2004 09:17:49 +0000

Yeah – I wondered about it being an old version as an item in the support forums mentioned in passing that RSS Bandit should only use the guid as a link if @isPermaLink@ is set to false, so someone over there obviously knew. Anyway I’ll grab the latest and retest.

By: Phil Wilson

Phil Wilson — Thu, 16 Dec 2004 00:15:12 +0000

You could move domain and republish the same items. This would break your [link] elements but you could keep your GUIDs the same so that aggregators wouldn’t duplicate all your entries.

Among other things.

I’m pretty surprised that this is broken in RSS Bandit as Dare should know the spec inside-out. Maybe it’s an old version? It’s a sourceforge project though, so you can “submit a bug”:http://sourceforge.net/tracker/?group_id=96589&atid=615248 .

By: jsled

jsled — Wed, 15 Dec 2004 17:36:05 +0000

bloglines.com does this as well, unfortunately. They provide the content at the bottom of the post, too, but use for the anchor used for the title of the post, which breaks the UI. :/