The Good News

1. It's easy to get started. After two years of waiting, unreplied emails and heartbreak - developers should now be able to get an API key within minutes and start building on the LinkedIn platform. That's great news and not something that could have been taken for granted.
2. The API allows search. That's great because with a little disambiguation done on the client side you can find the LinkedIn accounts of people you're connected to on other networks. Unfortunately, no one is doing exactly that yet - but isn't that the biggest value proposition here? I see a person on Twitter, on Facebook, on some other social network and I want to see what they do for a living. Let the app collect and expose that data from LinkedIn!

   Disambiguation of people with the same name and privacy limitations regarding who gets to see who's information are both complicating factors. The coolest use of the search API we've seen so far is Salim Ismail and Rohit Khare's Knx.to. That service is limited to your own connections so far, but it's definitely a keeper.



3. The API uses OAuth. That means that 3rd parties can offer fast, secure, standardized authentication into your LinkedIn user account. That's great.
4. Activity updates are now parsable by type. The API allows developers to pull in just one type of the many updates a person gets on LinkedIn. Will someone please build an app that just shows me when my contacts change jobs and leaves out all the status messages, friend connections and other cruft? That kind of granular control has a lot of potential and is reminiscent of the vision behind the proposed user activity data protocol Activity Streams.

And Now For the Bad News...

1. The first use-cases make it look like LinkedIn is trying to be Twitter. Tweetdeck and Posterous are the most high-profile early adopters of the new API; Tweetdeck will give you a LinkedIn column (too bad LinkedIn contacts can't be integrated into other columns) and Posterous will let you publish links to updates on that platform over to your LinkedIn contacts' streams. Jobdash looks like Tweetdeck just for LinkedIn and job-hunting, but it doesn't yet offer features like limited display of notifications by type - it's just a big stream of updates.

   LinkedIn is not Twitter! LinkedIn's Adam Nash told us this morning that he loves the Twitter and Twitter-like integrations but "integrating messaging isn't the goal, there's a wide range of business applications that will benefit from it. Twitter is hot so people are jumping to that but there are far more compelling business cases."

   Two years after the business-oriented platform was announced tiny Tweetdeck was just so hot it out-maneuvered all the business applications that could have been built to showcase? I don't buy it. Just like the formal partnership between Twitter and LinkedIn earlier this month, I worry that this API is built with marketing, promotion and broadcast functions best served.
   

   

2. Terms and Conditions are unclear, restrictive and changing. The API terms say that you can't build applications that compete with LinkedIn. API management service Mashery CEO Oren Michels (disclosure: RWW sponsor) had this in response to say: "It appears that you can't create a new experience around LinkedIn, an iPhone app for example. You might create some interesting bolt-ons to other services that might drive users to linkedin.com - but that's a very 5 years-ago approach to an API."

   "The signal from this is that they aren't encouraging developers to take the social graph and deep knowledge of peoples' professional lives and create new UIs for interacting with LinkedIn because they are explicitly concerned about competition," Michels said. "LinkedIn has amazing assets and a great business model - get out of the UI business!"

   Likewise several developers have expressed concern around the commercial limitations on the API. LinkedIn's Nash clarified with us that those terms simply prohibit charging people extra money for access to the free LinkedIn service and building an advertising network on top of LinkedIn profile data because of privacy concerns.

   Finally, the terms of the API aren't always clear. Michels points out that rate limits on accessing the API aren't made explicit - only that there will be rate limits and that a developer can email LinkedIn to request a personal expansion of their limit.

3. Not playing nice with others: LinkedIn is exposing what it calls an Activity Stream, but it's not at all related to the standardized format that Facebook, MySpace, Netflix and others are now publishing. LinkedIn publishes some Microformats but has been entirely absent from the wide-ranging community discussion of Activity Streams formats, we're told.

Michels may have said it best: "There are some really smart people over there at LinkedIn. If this is what we waited 2 and a half years for, it's a bit disappointing."

It is a bit, but not entirely disappointing. We look forward to seeing how the platform evolves and what kinds of applications are built on top of it. The web has been waiting a long time for a LinkedIn platform - now let's see what happens.