We held a pair of shootouts this weekend at the luxurious Temple of the Dog Studios in beautiful Decatur, Alabama: AmphetaDesk vs FeedReader, and w.bloggar vs Movable Type’s web interface.
The final score: one vote for a native Windows interface, one vote for a web interface.
AmphetaDesk vs FeedReader
AmphetaDesk and FeedReader are both servers that run on your Windows box and periodically go out to check RSS feeds. Both of them can be configured to listen for subscribe requests on Amphetadesk’s and Radio Userland’s ports (8888 and 5335, respectively).
AmphetaDesk uses your default browser for just about all its interaction. FeedReader is a classic 3-panel Windows design: the leftmost panel lists the feed sources; the top right panel lists the article titles, authors, and dates; and the largest panel on the bottom right contains an Internet Explorer control for displaying the article summaries and, if you click “Read On” in that windows, the articles themselves.
Because FeedReader doesn’t rely on your browser to display the feeds and articles, it’s relatively easy for it to, for example, show unread articles in boldface type and read articles in normal type (changing from one to the other as you click) and to show a running count of unread articles. There are several ways for AmphetaDesk to be as responsive to user inputs, but none of them looks particularly attractive from here.
More importantly, FeedReader can display just the summary from the article you just clicked on. AmphetaDesk displays a single page with all the summaries. If you have a lot of feeds, or if you’re taking feeds from folks who publish big summaries, that page can grow to be pretty fat.
These programs share some common faults. There’s no “undo”. Once you delete a feed, it’s gone. It was far from obvious how either program managed articles you’d already read – though FeedReader at least showed them in a different font weight. And neither of them can read opml files or any other multi-feed list format I’ve tried. Either there’s a secret I haven’t figured out yet, or both programs make you enter feeds one at a time.
In its favor, AmphetaDesk has a gigantic alphabetical list of raw feeds to choose from. FeedReader’s offerings are skimpy, requiring you to go out and hunt feeds down yourself. And when a site offers multiple feeds through its “link rel” headers and you use the “subscribe” bookmarklet, FeedReader concatenates the URLs for all the feeds into one string. Oops.
FeedReader, which, to be fair, is advertised as alpha, has a few functionality problems. For instance, quite often in FeedReader, when you click on a feed name in the left-hand column, you’ll have to click on some other feed name and then go back to the feed you want to get the article names to display in the top right column.
Two elementary lacks in FeedReader drive me crazy: the absence of a “back” button for the Internet Explorer control and the lack of integration with the default browser. If you hit shift-click, you’ll open a separate full-featured Internet Explorer window, which is a step in the right direction, but a tremendous portion of the time I find myself right-mousing on the link in FeedReader’s panel, selecting “copy shortcut”, and pasting into Opera or Moz.
Despite these advantages for AmphetaDesk and the fact that I never had a problem with AmphetaDesk, either with Opera 6 on this Win98 laptop or with Mozilla on a Win2K desktop, FeedReader’s ended up as the RSS aggregator I’m using all the time. That’s because of:
- FeedReader’s ability to handle a large number of feeds without producing an absurdly large web page
- its somewhat quicker responsiveness
- the ability to tell easily which articles I’ve read, and
- the ability to control the order in which feeds are displayed. Whoever decided to constantly change the order of feeds on the summary display page in AmphetaDesk made a huge mistake, imho.
w.bloggar vs Movable Type
w.bloggar (formerly known as just bloggar) is a Windows application that replaces the web-based interface for Blogger, b2, MovableType, Nucleus, BigBlogTool, BlogWorks XML Blogalia and Drupal. It works (at least on MovableType) through the XMLRPC interface built into MT.
My test with this blog was short and not particularly sweet. I created a dummy blog entry and posted it, but it didn’t show up on the blog. Then I selected “post and publish” which brought the article up on the blog (though I was unable to tell if it pinged the hosts I notify). Then, I told it to delete the article, but the article itself was left behind in my archive where it will doubtless stay forever.
The editor window offers several convenience features, like highlighting a section and selecting “italics”, which you’d find on a basic web editor. But it seems to be all old-style HTML, rather than the XHTML “style”.
When I went to enter another article and found that there was no way to add another article category, I pulled the plug on the test. Granted, I have to enter XHTML by hand typing it this way, and I’ve got to pay a call on a validator from time to time, but Movable Type’s user interface is far from awful, and there weren’t enough convenience features to motivate me to use w.bloggar regularly. I can easily see how, if I had several blogs and was using two or three different blogging systems, or if I had a slower connection or a less responsive domain host, w.bloggar would rise dramatically in value.