« Guidelines for Austin Bloggers Metablog Postings | Main | Meetup Venues for 2007 »

Bloglines Freedbacking: Redacted RSS Entries

Bloglines says they will find posts that use the words "bloglines" and "freedbacking" in them. OK, here goes. If you aren't into blog tech weenie stuff, feel free to move along.

Sometimes spammers find the Austin Bloggers site and sometimes their posts make it to publication. Bloglines picks up these posts from the RSS feed. That's bad. I want a way not only to get them off the site, but also out of cached RSS stores (such as Bloglines).

The problem is that you can't simply delete the entry. Most RSS tools (including Bloglines) cache the entries, so the deleted entry never goes away.

I've come up with a clever (and, I believe, standards-complaint) technique to handle this. Each entry in the RSS feed is given a unique identifier in the guid field. When an entry is deleted, it is replaced with an "entry removed" message with the same guid value. An RSS reader should replace the earlier version of the entry with this one.

Bloglines, unfortunately, does not. I fear they may be indexing on the link field, instead of preferring guid when present.

To demonstrate the problem, I published a test entry that appeared in the Austin Bloggers RSS feed this way:

    <title>Test Post</title>
    <description>Please ignore this entry. I am using it to demonstrate an RSS issue.... (via Austin Bloggers News and Announcements)</description>
    <guid isPermaLink="false">entry:5205@austinbloggers.org</guid>

Within moments, the Bloglines display for Austin Bloggers showed this entry:

Bloglines screenshot

Next, I removed the entry from Austin Bloggers. The RSS was updated so that the entry now appeared as:

    <title>(this entry has been removed)</title>
    <guid isPermaLink="false">entry:5205@austinbloggers.org</guid>

Notice that the guid is unchanged. The desired result is that an RSS reader replaces the earlier version of the article with this one. Bloglines, unfortunately, does not. The earlier version remains, and this version never appears.

So, Bloglines, please either support this technique or provide an alternate way to redact unwanted articles from an RSS feed.