Why Google is extending RSS
My ZDNet post today explores the reasons behind Google's new syndication format, GData. Like most people, Jeff Jarvis isn't sure what this means - and neither is Dave Winer. I'm in the same boat, but what I do know is that Google has taken a sudden interest in extending RSS and Atom. Check out this Google Base documentation, for RSS 2.0 and for all syndication formats. This is all about enabling bulk upload of items into Google Base, which you'll recall is Google's potential giant database of structured data on the Web. Google is obviously eyeing RSS (or syndication in general) as a means of getting people to upload data to Google Base. But why did Google feel the need to create a new protocol, called GData?
My initial reaction was that GData is a way to mix RSS/Atom with their APIs, in order to better integrate their increasing number of web applications. I'm not sure if this points to less of a walled garden, or paradoxically more of one because Google is defining the protocol now.
3 TrackBacks
TrackBack URL for this entry: http://www.readwriteweb.com/cgi-bin/mt/mt-tb.cgi/2654
Comments
Subscribe to comments for this post OR Subscribe to comments for all ReadWriteWeb posts
Minor point, GData isn't really a new protocol. It's an implementation/specialization of the Atom Publishing Protocol.
Interesting, because Google is calling it a protocol: "The Google data APIs ("GData" for short) provide a simple standard protocol for reading and writing data on the web."
It seems to me to be an extension of the Atom Publishing Protocol, rather than just an implementation. But I'm by no means a tech expert on this.
I guess that depends how you define "protocol"... personally I wouldn't get hung up on the label.
Yep good point Michael. My main interest is in *what* GData will be used for.
The protocol is HTTP - GET / PUT / POST / DELETE (in other words, CRUD).
The payload is Atom, chosen because it happens to do lists of things really well.
The list items carried in GData Atom are the interesting things - they'll vary according to the application. In the case of Google Calendar, each Atom entry contains a calendar event.
So, in a nutshell, GData is HTTP-based CRUD using Atom entries to represent discrete records.
There is really nothing sudden about this. They have been busy making big web enabled database clusters and helping define the Atom protocol.
This meshes perfectly with all that.