Thoughts: Leopard and OpenSocial

Final thoughts on Leopard:

I’ve not seen universal happiness with the Leopard UI. Barely visible icons and menu bars, and excessive CPU required for unnecessary reflection/3D geegaws figures at the top of the list. Compatibility with applications, including ones like Apple’s own Aperture is spotty. The Ars Technica review was, hands down, the best. Though the language is *little used, support for Java is missing (Java 1.6) or broken (Java 1.5). Problems with installation, but liking some of the new features, such as web widget thing. The new Safari has irritated me, and I haven’t even moved to Leopard.

Since two of my three computers are outside of the minimum required for Leopard, and the other is currently configured just right, I don’t see an upgrade in my future for months, perhaps longer. I may way until I can afford an Intel-based Mac, and get the OS already installed. After all my applications have been ported to Leopard.

On the new OpenSocial initiative:

I find it humorous that a lot of people are jumping up and down on this, without really knowing what OpenSocial is or does. Marc Andreessen, whose company, Ning, is participating in this initiative, probably has the best hyperbole cutting take.

OpenSocial is a defined set of APIs, agreed upon by a consortium of companies with Google being the major instigator.

Google is not serving as a gatekeeper for this capability, other than its role in defining the API, providing preliminary libraries, and a sandbox.

The OpenSocial does not enable ‘the social graph’. This API is meant to be hosted in different social network applications, using a combination of HTML and JavaScript. This is not a web services API, which is what you would need with the so-called ‘open social graph’.

This is a way for 3rd party application developers to create an application and only have to worry about integrating it with a couple of different platform APIs, such as Facebook’s and OpenSocial. One social platform can develop widgets to another social platform, but that’s nothing more than a direct link between two applications–it’s not ‘open’, there is no universal pool of data goodness from which to suck, like bees and nectar.

This. Is. Not. An. Operating. System.

This is a way of combining several smaller (or less US-centric) social platforms in such a way that developers will find it worthwhile to port their Facebook apps over to the new platform, and Google can then sell ads.

It’s a dumb-as-rocks API: not a whiff of the semantic.

The upside to OpenSocial is that 3rd party application developers don’t have to develop for a bunch of individual platforms. Of course, most are developing now for Facebook, and the other platforms want you to play in their playgrounds.

The downside? Who controls the API? It’s not ‘open’. Currently, it’s vendor controlled and closed. Will this change? Hard to say. The other downside is that this will start another round of dueling specifications. Flickr is not part of the original list of companies supporting OpenSocial, and being part of Yahoo, this isn’t surprising. Will Yahoo then group with Microsoft and Facebook to create a competing “open API”? Or perhaps, create a third? Are we looking at the beginning of Social Font?

Hyperbole cutting facts: OpenSocial is not being released yet, does not enable the ‘social graph’, isn’t necessarily ‘open’, and only adds to that part of the web that is utilized for social networking. It doesn’t provide a thing for the overall web. Anyone using ‘Balkanization’, ‘Facebook’, and ‘walled garden’ in the same sentence should be beaten with Peeps. Especially when one considers that the instigator behind Open Social is Google.

Most significantly: Thursday is just an announcement of the OpenSocial effort. The API isn’t scheduled to go live until November. Only Orkut supports the functionality at this time. Other partners are ‘in development’. Stop peeing your pants.

*Joking! Don’t hit me.