RSS Online subscription Service: Discovering social networks from subscriptions

rss| Network | online

Today see: Webleon's Blog:longhorn embrace RSS, before the endless stream of client RSS Subscriptions tool to be careful, "default" power is still very large. In contrast, I feel that online RSS subscriptions will undoubtedly become a more promising model in the future. Bloglines is currently one of the more popular RSS online subscription services.

I use Bloglines less, bloglines the default according to the author classification of the RSS reading mode is not high efficiency, or more accustomed to according to the time inverted flow of news, so has been trying to use Lilina daily production of an RSS newspaper. But beyond that: you can still feel some of the advantages of Bloglines's central service: It's easy to find the popularity of RSS in a subscription relationship and help subscribers generate more interaction between them. You know, Many Bloglines users themselves are also bloggers: so let you see how many people subscribe to their RSS is no doubt satisfying their egosurfing needs (FeedBurner also has a similar RSS subscription counter function): If the subscriber's RSS bookmark is publicly published: then you can still see To the specific subscriber who is to find more "like-minded": it is hard to imagine that there are people who subscribe to my del.icio.us.

Similar to del.icio.us: Discover the functionality of other users who have the same URL as you.

Using Bloglines, you can also find more useful features: such as posting blogs, subscribing to weather forecasts, and so on. Although the central service is somewhat dangerous: Bloglines provides an import/export function that conforms to the OPML specification, I tried it to swap with Lilina bookmarks.

But on the other side of the good side: like FeedBurner, bloglines RSS feeds can help you see websites that have some routing damping from home: like Googleblog.

2005-06-26

About going to the center: How to build a service similar to Bloglines?

My own attempt to use Lilina is actually trying to build a technical preparation for a bloglines similar service, and now I have the following points to realize:

1 caching mechanism: Caching mechanism can greatly reduce the load on RSS publishing source;

2 RSS Error correction/fault tolerant mechanism: Many RSS is not even strict XML, and timestamp problem may be a very common problem.

3 Dynamic update mechanism: Because most RSS updates are not very frequent, so how to smart some dynamic adjustment of the RSS synchronization cycle also need to consider: for example, according to the average interval of the latest 5 content for RSS updates frequency.

4 of the Dead chain of RSS check: Many domestic BSP RSS address is not very stable, when the service upgrade, how can the dynamic upgrade to the new RSS publishing address?

In the development of RSS search engine also encountered the discovery of RSS data source: Rely on machine capture need to increase the identification more trouble, and it is difficult to see the importance of each RSS, if there is a bloglines the same community to help dynamic discovery/sequencing, will undoubtedly be a very good supplement.