News feeds vs Data feeds (and Yahoo)

Some feeds have no use, at least for the majority of us, outside of a regular feed reader like Bloglines or FeedDemon. I call these "news feeds" whether they are feeds of world news from CNN.com or Shelley Powers' personal blog postings. Perhaps "news feeds" is not the best name and I should change this habit. Now that I'm an a-lister I should be more precise, maybe. ;)

I don't like to just use the word "feeds" to describe them all, because I think feeds that have more granular data in them need to be distinguished. I call these "data feeds". Most feedreaders won't be able to parse the additional data in the feeds, so they should still have title and descriptions, etc. Some feed readers will (For example, Nick Brabury's FeedDemon supports Yahoo's media RSS extension).

Previously, I pointed out that many of the Yahoo RSS feeds could be much more useful to developers if they were "data feeds". These feeds are documented on the Yahoo Developer Network pages, but lets face it - the majority of them are NOT for developers at all. They're for feed readers. Jeremy Zawodny asked for more suggestions, so here's a quick rundown of the "big list" of Yahoo RSS feeds.

Of course some of these feeds, IMO, are supposed to be "news feeds". They're supposed to be read inside a feed reader and have no real use for developers. For example: Ask Yahoo!, Yahoo! Answers, Yahoo! Autos Custom - not to be confused with the Yahoo! Autos feed, Yahoo! Education, Yahoo! Health, Yahoo!Message Boards, Yahoo! News, Yahoo! Next, Yahoo! Picks, etc). Jeesh...I'm tired of typing "Yahoo!" so I'm just going to stop now.

Some feeds could have use for developers, but their usefulness is limited by the content of the feed itself. For example:

  • Autos - This only provides a title, description, and link to an auto ad. Think of all the possible applications that could be developed if with this feed if it contained xml elements for manufacturer, model, year, msrp, price, fuel economy, blue book value, location, enclosure (photos), etc.
  • Classifieds - Same as above. The extension elements would need to be more generic, such as msrp, price, condition, location, etc.
  • Buzz - This provides a title, description, and link. Why is this search term so hot? How hot is it, how many people have searched for it? How long has it been hot? What is the category of this search term (entertainment, technology, finance)?
  • Finance - I already commented on this here. To me, this missing data is the biggest example.
  • HotJobs - Gives me a title, description, and link. Location, salary, education level, required skills? None of that appears, but is all information that should be in the feed. Any consumer is going to want to see this information before clicking through to the full listing.
  • Local - The Events feeds of Yahoo! Local are just pitiful. Fortunately, the Upcoming.org feeds are great, though. Worst of all, the Events feed is not even a valid RSS 2.0 feed. It uses a "cat" element, which doesn't exist in RSS 2.0 (it should be "category", of course). It provides an event date in the "date" element, which also doesn't exist in RSS 2.0. They do provide a "pubDate" element but its value is in the wrong format. The Yahoo! Local Restaurants feed is missing info like the location, price range, and cuisine category (Italian, Greek, Indian, Eastern NC BBQ :P).
  • Movies - There are a variety of movie feeds available. Sadly, only a couple of them provide useful information using Yahoo's own media namespace (media category, credits, thumbnail, etc). Most, such as the new releases and coming soon feeds, don't provide anything other than title, description, and link. None of the feeds provide rating, genre, box office earnings, release date, dvd release date, or enclosures (trailers).
  • Musicmatch - Just the name Musicmatch already has me turned off - but maybe I'm the only one who thought the MusicMatch player install single handedly reduced the value of my PC by half. A variety of feeds are available - all useless to developers. Strangely, a couple of these feeds use an mmrss namespace, but only for a product image. There is no useful information about the media itself. The Yahoo! Music feeds and the Yahoo Music Engine are much more useful.
  • Shopping - So much more can be done here, but the full Shopping API takes care of that for developers.

Other Yahoo! feeds are useful for developers, but the data that they provide fits nicely inside existing syndication formats without any extensions (blo.gs, del.icio.us, 360 blogs and comment feeds, Directory, and Groups).

Also, some of the Yahoo! feeds are already very useful for developers because of their use of RSS module extensions to provide more information about each item (Flickr, Upcoming.org, Yahoo! Traffic Alerts, Yahoo! Music makes good use of the media and ymusic namespaces, Search and My Search, and Weather).

Feed developers:

  • Run your feeds through a validator before publishing them!
  • Please give your items a guid.
  • Make your feed more useful by providing as much detail as you can. If bandwidth preservation is an issue, let your feed accept inputs (via querystring, for example) that let the consumer ask for a more verbose feed.
  • If you're not sure how to create your feeds, or you want to find a way to do it quickly, take a look at RSSBus.



Technorati : , , , ,

Print | posted on Monday, August 21, 2006 10:29 AM

Feedback

# re: News feeds vs Data feeds (and Yahoo)

Left by Brent at 1/3/2007 9:07 PM
Gravatar Generous investment of time on your part - nice breakdown...

Have you had any success correcting error laden parameters of Yahoo Local 'Feeds' for the sake of publishing?

Be Well,
Brent

# re: News feeds vs Data feeds (and Yahoo)

Left by Lance at 1/22/2007 4:37 PM
Gravatar Regarding the broken Yahoo Local feed, it is not valid, but most readers should be able to handle it. They dont be able to parse out the info that isn't namespaced, but. One way around things like this is with the use of RSSBus, you can rip out whatever you want and reformat it as a new feed (since they wont do it themselves).

My suspicion is that they dont plan to change it since they have Upcoming.org now.

Your comment:





 

Copyright © Lance Robinson

Design by Bartosz Brzezinski

Design by Phil Haack Based On A Design By Bartosz Brzezinski