The Fall and Rise of Google Reader
I've been seeing a number of hits on my search feed for "NewsGator API" recently. There seems to be a couple of significant events: first, that Microsoft's WebFeeds API is getting some attention, and second, that Google Reader seems to be gaining some traction, and even a loyal userbase. This is pretty interesting because Google Reader was universally skewered when it was released, and from what I can see, while the performance is vastly improved, Google Reader itself hasn't changed much.
Google's big issue seems to be the lack of a public API. For example, this Google Reader user bemoans the fact that NewsGator charges for the service and also that he simply likes Google Reader better than NewsGator Online. Fair enough criticisms. Andrew Grant has another gripe: FeedDemon doesn't synchronize clippings and doesn't handle hierarchies of folders. His choice of Google Reader would be a bit odd, then. GR just has labels for feeds; if you import a hierarchical OPML, GR will apply one label for each level of hierarchy. You can select feeds by label, which amounts to selecting a folder in NewsGator Online or similar readers. Google's method is somewhat more flexible than a hierarchy, in that you could create graphs of labels. I haven't lived with GR enough to know whether I'd take advantage of that capability. In any case, what I've come to realize is that all of these organizational features - folders, labels, stars, tags, flags; all of these are congruent features. What I mean by that is that you can implement all of them in terms of the others. There's some more basic concept of organization - we assign other labels onto things - and all these features are expressions of that.