Dougal Campbell's geek ramblings » Syndication

WP-Cache fix for Content-Type in feeds

Dougal — Tue, 19 Jun 2007 16:56:42 +0000

If you run a busy WordPress site, or even if your site just has a lot of processor-intensive plugins, then you probably already run the WP-Cache plugin (plugin directory, original announcment, recent security update info). Even though my site isn’t super busy, my server is a little light in the RAM department, and using WP-Cache helps the box keep up with requests better.

One minor annoyance, however, is that with WP-Cache enabled, my syndication feeds aren’t delivered with the correct Content-Type. They are all cached with a “text/html” type. Again, this is only a minor annoyance, since I haven’t run across a feed reader yet that cares very much about the Content-Type header. But the Feed Validator does. If it sees an Atom feed delivered with content type “text/html” instead of “application/atom+xml”, it gives you a warning. It still checks the validity of the actual feed content, but there’s that big ugly warning at the top.

But not any more. I dug a little bit into the WP Cache plugin to try to figure out what was going on. The plugin does attempt to preserve the correct content type of pages when it caches them, but the output buffering that it uses to capture the content gets in the way. You can’t get to the generated headers until after you flush the buffer. But you don’t want to flush the buffer until you’ve got all of the data to write into the cache meta files. It’s a catch-22. (Actually, it might be okay to flush the buffer at the point where you need the headers, but I’m not sure without further analysis, and I just don’t have the time right now. I’ll leave that as an exercise for someone else. )

So, if you’re a stickler about minor things like the Content-Type of your feeds, and you’re using WP-Cache, feel free to check out my patch: WP-Cache Feed Content Type Patch. Note that the changes to the wp_cache_get_response_headers() function don’t actually work for me. But I left them in the patch in case anybody else wants to examine that section more closely. It might just be something that works in some combinations of Apache/PHP versions and not others.

Update 2007-10-12: This problem was officially fixed in wp-cache version 2.1.2.

Feedlounge is live

Dougal — Wed, 25 Jan 2006 22:52:36 +0000

This is old news now, but what the heck — I haven’t posted in a while, and this will be quick. In case you didn’t already know, FeedLounge, that fantabulous online feed reader that all the Cool Kids have been raving about, is out of testing and ready for the world. It’s currently only available with a paid subscription, but to get a taste (the first hit’s free), you can take the Three Hour Tour.

I’m using FeedLounge to track stories from over 100 different sites (over 120 feeds, but a few are different topics from the same site, see?). As an alpha tester, I get a couple of months service for free. But when my free time is up, I’m going to see if I can squeeze a subscription into my budget. This is one of the few web-based applications that I’d ever consider paying for.

The State of FeedLounge

Dougal — Mon, 05 Dec 2005 16:25:42 +0000

Alex has posted a moderately detailed article explaining The State of FeedLounge. For any newcomers who don’t know what FeedLounge is, it’s a web based feed aggregator, not unlike BlogLines. But FeedLounge is a next-generation web service that looks and acts more like a desktop application than a web site.

I’m one of the lucky alpha testers who have had access for the last six months. I committed to using FeedLounge as my only feed reader, dumping SharpReader and the RSS capabilities of Thunderbird. I’ve watched FeedLounge go through some rough spots, and it always came back better than before. Alex and Scott have done an outstanding job with everything from the great AJAXian user interface to the invisible backend. When the time comes, if the price is right, I’m going to be a paying customer.

But you don’t care about all that, do you? What good is it to read about a service that you don’t have access to? All you care about is knowing that FeedLounge will be open for public beta on January 16, 2006.

Google Blog Search

Dougal — Wed, 14 Sep 2005 21:31:24 +0000

I may be a little late coming to this party, but Google has introduced Google Blog Search, which limits searches to just blogs (well, in theory, anyhow). The most interesting thing about this, to me, is that GBS provides the search results for your terms as Atom and RSS feeds.

It seems that Google is leveraging blog feeds to get structured data, and monitoring ping sites for timely updates. This is something that lots of people speculated about way back before Google bought Blogger. RSS/Atom feeds from blogs are such a rich source of structured information, it only makes sense that a “mainstream” search engine should take advantage of it. I can only wonder what took them so long?

The next feature I’m looking for: “Search within X miles of location Y”, using geotagged blogs. This would be great for searching for information written by other bloggers in your vicinity.

New My Yahoo Beta brings RSS to the masses

Dougal — Tue, 28 Sep 2004 20:51:50 +0000

From Jeremy Zawodny’s blog:

Well, we just launched a beta of the next generation of My Yahoo that fits into that reality. Instead of “you can add anything you want, as long as it’s on the list of My Yahoo content” you can now add pretty much any public RSS or Atom feed. In other words, the content model is open.

Let me say that again, just for dramatic effect: the content model is open now.

RSS has slowly been edging into the mainstream for a while now. But with this new exposure, we can expect a veritable explosion of new awareness of RSS from the non-technorati. Creators of feed aggregator software and services should expect a fresh deluge of new users. Fortunately, most of the feed software/services out there are pretty mature now.

One note: for some reason, when I tried to add my feed to My Yahoo, it didn’t like the RSS 2.0 version. It would show me the headlines just fine, but then complained when I tried to perform the final add. But it liked the RSS 1.0 (RDF) version just fine.

The Conversational Web

Dougal — Tue, 28 Jan 2003 17:57:55 +0000

A lot of blog development of late has been directed towards establishing connections between multiple weblogs via mechanisms such as Trackback, Pingback, and friends. Throw in a little RSS and a news aggregator that can tell you where to go to comment on an article. Maybe add a dash of FOAF…. The result is that multiple websites become part of one big conversation.

Well, if it’s a conversation, how about we add a splash of Instant Messaging? Yum. Now we’re cooking!

A couple of days ago, I whipped a quick mod for this site that will notify me via Jabber when a new comment is posted. I’ve got an idea in the back of my brain for setting up a Jabber PubSub component that would let people subscribe themselves to the comments of a particular story. Since not everybody uses Jabber, I figure there will need to be an email notification option, as well. I haven’t had the time to work out the details yet, let alone start any code. Heck, I’ve been typing this entry off-and-on over the last 24+ hours, as it is.

And this is the point where I wish I had already recoded my Trackback implementation to handle autodiscovery and multiple trackbacks.