« RSA-704: C++ Port | Main | Airfoil For Windows »

May 12, 2006

Bloglines Plays Chicken

I'm one of Tim Bray's thousands of subscribers (via Bloglines) and I prefer his Atom feed. Bloglines recently redirected all of Tim's Atom feed subscribers to his RSS feed. (Tim's take.) Bloglines would rather add marginal features like embedded YouTube applets instead of hunkering down and supporting RFC4287.

They would probably defend themselves by saying that they're just a consumer choosing equivalent feed formats. In this case, they're not. Tim made the decision to make his Atom feed full text a while back; his RSS feed is just summaries.

Pathetic. I suppose I could call myself that, too, for continuing to use Bloglines. I'm not prepared to jump to Newsgator Online just yet. I'm actually tempted to roll my own Planet and call it a day.

Posted by Alex Harden on May 12, 2006 9:11 PM |

Comments

Not to defend Bloglines (I don't use it anymore-- since Newsgator synchronization has allowed me to get to know great software like Feeddemon and NetNewsWire again), but...

I don't think the decision of a centralized aggregator to eliminate the duplication of pulling two feeds with the same content is so unreasonable.

They probably made the *wrong* choice in this case, since the atom feed appears to be much more useful.

One idea I've fooled around with on paper is scoring feeds on completeness and amount of metadata, so that a machine could make a reasonable choice when presented with multiple, similar feeds.

Posted by: Ross Karchner | May 13, 2006 2:23 PM

If you get Planet working, let me know. I'm trying to run it under OSX and have all the Python requirements squared away, but it won't actually parse feeds - I get "Error 500"...

Posted by: Don | May 14, 2006 10:33 PM

I'm getting the same errors under Windows - trying to trace the code to find out why. If I invoke feedparser.parse manually against a particular feed, it works fine.

Posted by: aharden | May 16, 2006 3:42 PM

Hmm... I compared my config.ini with Sam Ruby's and he didn't have the feed_timeout variable defined. Commented that out. I also set activity_threshold from 0 to 90 per Sam's .ini. After that the feed parsing start to work. The only thing I'm seeing as weird right now is that I have to manually delete the *.tmplc files that get created with each run before the script will successfully run again. I'm going to customize my planet's look a bit before I do the initial copy to my web host and share.

Posted by: aharden | May 16, 2006 4:04 PM

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)