Pierre Wolff raps out why the DLA model is ideal for evolving open standards

My buddy - Pierre Wolff left this comment about DLAs -  and the tradeoffs between hard wiring to specific services and supporting and developing open standards.  Evolving open standards is a tricky game - especially when it comes to handling this kind of data movement, verbs, actions and authentication.

Marc, you might have pointed out that none of what your discussing here is new, it’s just that it would be best to systematize it and have it become a ubiquitous set of functionality rather than one-offs.  For example, today users on Yahoo!’ IM can communicate seemlessly w/users on MSN’s Messenger.  Today one can use LinkedIn or any of several social networks and import one’s e-mail address list either from Outlook or from Yahoo! Mail, Google’s Gmail or MSN’s Hotmail, in order to facilitate the invite functionality.

But as we see over and over again, this is dependent on the developers of each service to build this functionality from scratch, each works differently, and there’s tons of development and data redundancy all over the place.

The DLA metaphor appeals to me because it elaborates an architecture to normalize of all of these common activities that no longer gain any advantage from being silo’d.  There’s no longer a strategic advantage in being able to easily grab my Outlook contact database into any social network since they all do it now, so why not open that up and make a generic way for any of service to get at it.  Same with moving or inviting my friends from one social network to another, or even if it only moves a representation of them while keeping their profiles on the social network they prefer.  Let’s blow them walls up!!! :)

Thanks for sharing.

I bolded that sentence myself - as I think Pierre really hits the nail on the head.

The world will never have ONLY open standards - but all sorts of hodge podge combinations of proprietary, open and weird hybrids - like Flash and Java.  End-users will demand integration and that’s what DLAs are based upon.

Pierre totally groks the effect of normalizing across such a complex set of inter-connection requirements.  Simple communtiies like MyBlogLog can easily evolve, but its hard as hell to evolve a set of 10-15 standards - which is what’s gonna happen in theDLA space.

Its inevitable technology gravity. Its bound to happen - its just a matter fo when.

Comments are closed.