Live Mesh: Microsoft hews to open standards rule
As reported pretty much everywhere already (I rarely "scoop" things, a trait that would make me a bad journalist were I ever to become one), Microsoft has provided more details about the new platform that it is calling "Live Mesh." They will likely have the same problem explaining it that they did with .NET in its earliest days, which to a certain extent is understandable. How do you explain essentially developer-oriented technology to non-developers, and further, how do you convince them that they need this technology? Hopefully, marketing has learned its lessons from the .NET days, and won't start slapping the "Live Mesh" label on everything from server platforms to lunchboxes, a situation that unnecessarily confused the .NET story.
Live Mesh is supposed to be a common framework to enable cross-device interoperability. It also includes a bunch of shared services that can be used from any Mesh-compatible device, such as network storage space and photo-sharing services, among others (others likely include many of the "Live" properties) . This makes sense given the direction that the world is moving in, with an ever-growing proliferation of computing devices both on one's person and within the home that, currently, are too much like islands of processing power. A true mesh platform that standardized cross-device communication and synchronization in the same way HTML / CSS / Javascript has standardized user interfaces on the web would surely be a step forward from an IT evolutionary standpoint.
Perhaps it was a Freudian slip, but I think the use of the term "standard" was the essential part of the previous sentence. Microsoft won't get anywhere if they tried to peddle a closed-protocol environment to developers in 2008. That might have worked in 1995, but that certainly won't work today. Devices are simply too varied, and computing is too essential for large numbers of consumers to buy-in to a lock-up strategy.
Live Mesh, at least from what I understand of it, seems to get that. They aren't proposing protocols for data exchange that would be alien to most web developers. HTTP, RSS, REST, ATOM and JSON are are standard protocols, and though FeedSync may be new-ish, it is an XML protocol based on ATOM and RSS that is scoped, according to the spec, "to define the minimum extensions necessary to enable loosely-cooperating applications to use Atom and RSS feeds as the basis for item sharing – that is, the bi-directional, asynchronous synchronization of new and changed items amongst two or more cross-subscribed feeds."
From a data standpoint, Microsoft is hewing to the open standards rule. That openness seems to extend higher up the stack as well. Not surprisingly, the want to make it as easy as possible for Microsoft-developed technologies to access and consume those feeds, which explains the object model libraries for .NET (consumable from Silverlight) and WIN32 in Mary Jo Foley's developer stack diagram. They also, however, include libraries for Javascript, Java, Cocoa, Flash, Ruby, Perl and Python.
This is a real shift, in other words, for Microsoft. They are starting to realize that it is better to be the generalized platform used even on non-Microsoft platforms in a world where computing device proliferation creates ever more niches within which competitors might thrive. This is more in line with Microsoft's core skillset and competitive advantage, as it leverages Microsoft's surplus of platform designers to create something that is very different from what Apple (as the best example) tries to do with its products.
Microsoft will still actively encourage Windows development, to be sure, but the door is wide open for others to plug into the same infrastructure Microsoft uses for its own products. Even if you don't like the libraries Microsoft creates for your platform, the fact that the data layer is completely open would mean that you can roll a custom library more to your liking (a fact, I think, which would encourage Microsoft to make those libraries quite attractive on other platforms - the better to keep developers using the Microsoft version).
Live Mesh is interesting stuff, and a sign to me that Ray Ozzie really could be the force to turn Microsoft into a completely different sort of company. If he succeeds, it will be a company more organized around Microsoft's competitive strengths, while at the same time shifted in a direction more in tune with the "ubiquitous computing" reality of 2008.