얼마 전에 소개한 글에서 처럼 RSS는 단순히 정보를 알리는 syndication 기능 뿐만 아니라 어떤 대상에 대한 정보를 표현해주는 metadata의 기능도 가지고 있는데.
그동안 내가 RSS가 폭넓게 어떻게 사용될 수 있는가에 대한 부분을 잠시 정리하여 윗글에 comment 로 올려놓은 것을 아래에 잠시 붙여넣기.
Three things regarding the non-blogging use of RSS.
1. Product announcments for the shopping sites. It’s been mentioned already, but too often what I’m more interested in is not the collection of every single product from a single shopping site, but rather a certain category of items or one item from different sites. In order to achieve this, you could have something like, “http://www.amazong.com/feed/q?keyword=rss”, which delivers a feed of all products which are related to rss. Even better, throw in the rip/mix/burn of feeds (yeah!) only on one category from various shopping sites. There you have the best feed which contains most, if not all, up-to-date product announcements related to rss. I’m quite positive, in the next few months, as the number of product-related feeds increases, we’ll be seeing some meta-shopping sites coming into life, which provide the services I just described.
2. Sportcasting feeds and real-time delivery.
I developed for my master’s project a pub-sub model for sportcasting, which basically delivers real-time scores of the games. One problem I had with this model was that the update rate had to be extremely fast in an environment like this, probably the same with stock-tickers. I don’t know how often the Bloglines tracking system updates the feeds, but RSS might not be the best model for “real” real-time updates. For example, every time I “watch” an NBA game on ESPN, the update periods are usually 30 seconds, and I feel even that is a little too long. 30 seconds might not be too bad for a desktop RSS reader, but I assume there’s practically no way that an online aggregator can update those feeds in less than a minute, not knowing which one has to be updated that frequently. I believe Bloglines update period is about 30 min. (if i’m wrong on this, please correct me)
3. RSS within an Intranet
Anything becomes a much more difficult problem when it comes to a business application. The same goes for RSS. The thing about RSS is that it’s ON THE WEB, meaning anyone can access it as long as it has a URL. This cannot be tolerated within a business application. There should be ways to add in access control lists and security mechanisms so that the feed itself, if not the server on which the feed resides, can determine who can read and who cannot. I feel like this will require lots of experimentation and trial-and-errors.