Earlier this week, I wrote about Atom being a key component of Basic Content Management services. Google is poised to make this happen with their unveiling og GData. Web 2.0 Explorer wrote:
Google data APIs ("GData" for short) is a new Google protocol for "reading and writing data on the web". It’s described here: "GData combines common XML-based syndication formats (Atom and RSS) with a feed-publishing system based on the Atom publishing protocol, plus some extensions for handling queries." So GData is a new protocol, but "based on Atom…"
Very interesting. My first perusal of the GData docs gives me hope that Google is doing good things here. Some thoughts:
- Queries – this is a big missing piece from Atom, which seems stuck in a File/Folder organizational mentality. The query expressions in GData aren’t very deep, but there’s more than enough to be very functional.
- Categories – The query syntax for categories (http://deeje.com/musings/-/xacti) is different than for other fields, and is just shy of being like technorati tag queries (http://technorati.com/tag/xacti). Ug, why not just go with tags and be done with it?
- Optimistic concurrency – APP has purposefuly sidestepped the question of author locking, and GData’s solution is elegant: build a version number into the Edit (action) URL, and if a newer version shows up, then you can’t perform your action anymore.
- Authentication – GData provides authentication methods for desktop apps, and authentication methods for webapps… but what about app models that start to blur that distinction?
As a client developer, my first question is… when/does this work with Blogger? And my next question is, who else is on board? SixApart? Roller? WordPress? My third question is, what does the APP IETF working group think of this?