Linked Data is finally picking up serious speed. The stepping stone into a World Wide Database. As you can see on most articles out there, there is strong emphasis on the "open" aspect of this technology. You can constantly hear TimBL talking about "Open Linked Data" as the next big thing, just like he once empasized the "open" aspect of the Linked Documents.
From a macro/social perspective the "Open" aspect of the web is in fact very relevant. We now know that the words "open" and "link" have ultimately brought something huge upon us. Just by putting our documents out one by one magical, valuable and unpredictable things continue to happen. A revolution that has touched the very essence of our life. Such is the power of the Open Web in general.
However, this "Open" aspect is only one part of the story ( and the one that will take over the buzz in the short term for sure ). There is also a more serious and practical side to Linked Data that you should start paying attention to: the corporate story, where this new technology is now being used as the latest contendor in the pursuit of enterprise agility.
This is in fact quite predictable. We already saw a similar story with the emergence of "portal" and intranet solutions right under the growing buzz of the Web of Linked Documents ( WWW ) over 10 years ago. The Open scenario is more visible, while the other, behind closed doors, is silent but massive in terms of the ongoing investments it drives. Both happen together and both feed on each other.
However, despite the similarity with the "portals" phenomenon, please don't assume the "corporate" branch of Linked Data will be as big as "portals". That would be a terrible mistake. This will be a THOUSAND times bigger. Yes. I am making that number up, but I believe I might even be running short.
The reason for this huge difference simple: For every document there is roughly a thousand times more structured data. If you don't like mine, just go ahead and make up your own numbers...
You see, structured data is at the core of our lives, feeding and being produced by all operational applications and systems ( including your phone right there, the systems that run your bank account and the subway and that old LDAP server in your company ). Structured Data is a growing treasure we have accumulated over time. It is everywhere, and it is the life blood of organizations. The lifeblood of our semi-automated semi-online society in fact.
We care about Data and we cherish it. Most of the times we are unaware of its omni-presence, but when it goes wrong. We feel it.
Data is like Air.
From a corporate perspective, we already knew that our Data ( in general ) was important when we first launched our intranets years ago.
But why have we still failed to achieve success? Why is this data still sitting there, rusting, and costing us millions to move around just to answer simple questions?
Clearly we haven't built a suitable answer yet. You see, the Document Web has been a terrible underachiever in the enterprise. This has accounted for a lot of frustration and thus you might be thinking: There is no silver bullet solution to this problem..
After all, when we first opened our doors to "Internet technologies", we thought we were getting an information super highway, but it turned out to be a piddly trail to a land of excessive costs, vendor lock-ins and little results.
But please don't loose faith. The reason behind this massive "perceived" failure is that we picked the wrong solution to the answer and everything else has been a consequence of that mistake. We have been trying to fit a square into a circle for 10 years
What did we get wrong?
Simple: ultimately, what we need most is to virtualize Concepts, not Documents. We need to integrate the rows inside our Databases and Spreadsheets, and the web is, at its core, an address space for Documents. Yeah, there is no space for those columns, so we have to create all sorts of contrived APIs and exchange formats. This fundamental mismatch in the addressing space granularity has become our worst enemy, and quite a headache to say the least.
And this is also why most organizations today, after a thousand different APIs, millions and millions spent in one middleware generation after another, are still surrounded by replicated and fragmented systems.
And we know it, and it hurts. Yet we have even embraced the constant struggle this brings to our lives as a necessary ( and massive! ) activity in the IT realm. And we have coined a word for it: "Integration".
Hmm.... word is not right. It should be called "Moving data around in buckets". We owe the misconceptio to the marketing tricks of big IT companies. Of course they come up with prettier buckets each year to keep your boss happy.
I will soon tell you how to jump on board of the revolution. It's just around the corner.
But, for the curious:
The Web now accepts rows and columns in its core infrastructure ( URI x URI = Cell ) so you can directly publish your structured data sources down to the cell level, and start kissing your prehistoric "buckets" middleware good bye ;)
