Why we need an information utility
- Summary:
- Operating standards among high tech vendors are rare. The DOD kerfuffle puts a bright light on the need for standards so that the winner-takes-all issue is resolved.
I’ve previously offered my belief that the evolution of an information utility is just around the corner but it’s not really much more than that. Wishful thinking, I suppose. Other than me, I’ve heard Oracle co-CEO Mark Hurd utter those words at least twice, once at OpenWorld and again last week on the record at the Oracle media day.
An information utility would bring together disparate offerings from numerous vendors the same way an electric utility does or any utility for that matter because that's how utilities operate. They set standards for how big entities interact to produce a common good for society—something with great, well, utility.
We have the appearance of continent-spanning electric grids today but the reality is much different. In the U.S. it’s a lot of independent contributors generating electricity, transporting it, and delivering it to consumers. There are also numerous maintenance companies involved in all phases of the business.
The thing they have in common is a set of standards that to which they all work. Delivering 120 volts of alternating current oscillating at 60Hz to the American consumer has significant implications for how you build transformers, generators, and transmission lines for instance. Those standards would be much different if we had gone with the European standard of 240 volts at 50 Hz or even more different if we’d gone in Thomas Edison’s direction (rather than George Westinghouse’s or Nicolai Tesla’s) and decided on direct current. But we didn’t do any of that. We set one set of standards and a whole industry, or multiple industries grew.
Now the same thing is happening in information. So far we’ve been cocooned inside the idea that Google is Google and Facebook is it’s own thing and commercial enterprise cloud vendors have their own fiefdoms that now and then share data but all that is changing quickly.
Pentagon procurement
It’s possible that the current heated discussion about the winner take all competition for cloud technology in the US Department of Defense (DOD) could be the last such contended cloud technology procurement in history. Why? Because there will be too many losers. Companies with deep pockets and resources—and big egos running them—that simply won’t stand for losing such a big prize.
Those companies include Amazon, IBM, Microsoft and Oracle. Oracle has been the most vocal in all this deciding that Amazon is a mortal enemy. Amazon is the 800-pound gorilla in cloud infrastructure right now and Oracle, despite its size, is punching upward.
Amazon looks to have the inside track so Oracle is balking at the idea of winner take all. Microsoft and IBM are also in the mix though they are working more behind the scenes. DOD doesn’t want the hassle of spreading information “across a multitude of clouds” because it could hinder their ability to analyze huge amounts of data and present computing and storage issues. At stake is a $1.6 billion pool of money between now and 2023.
Right now the missing link between the vendors appears to be a set of standards and technologies that could break down walls or build bridges and enable better interoperation. Oracle especially has been working to produce advanced componentry that would drive the utility such as its autonomous database and security systems and solid-state storage systems. These would be important elements in a standards-based information utility. But other vendors haven’t been standing still either.
However, none of the major vendors has done much to initiate the standards-setting process. Last week I got a boatload of insight from Don Deutsch Vice President, Chief Standards Officer Oracle. You don’t know this guy but he’s affected your life. He was a participant at the seminal meetings that set the relational database standards in the 1980’s. That process, which gave us SQL among other things, got going before there was a viable relational database product.
IBM was there for the meetings but so were representatives from a lot of computer companies you’d need to search Wikipedia to remember like GE (Deutsch’s company at the time) and Control Data and several others. Among the participants who weren’t hardware companies was a nascent database company that would eventually become Oracle.
Today, Deutsch rides herd on literally hundreds of standards whose boards Oracle participates in. To be precise, Oracle currently has 257 employees participating in 314 technical working groups, 67 administrative or policy committees, and 216 leadership roles across 103 standards-setting organizations.
At its simplest a standard sets protocols for how disparate systems communicate and trade data
says Deutsch. The better the standard the better the communication. That’s precisely what we don’t have right now and it is holding back deployment of the 21stcentury information utility.
We’re at the very beginning, back at the point where we could have another meeting of smart people setting the standards for the next big thing, the information utility.
My take
A big government procurement like the one at DOD is often awarded to a constellation of vendors all working together to produce the product because there are standards. Without standards there’s a great deal of risk for everyone.
It’s worth noting that Astronaut Alan Shepard, the first American to orbit, once said that the hardest part of waiting for liftoff was knowing he was sitting on top of a million parts all of which had been built to a standard and delivered by the lowest bidder. But it all worked because there were standards.
Today’s DOD spending is like that but the Pentagon doesn’t have much choice about how it makes the purchase given the downside of insufficient interoperability. Of course the downside Oracle mentions (while couched in terms of 'unfairness') is equally risky.
No doubt the participating vendors see this too and that’s why I think this procurement is the last of its kind. Regardless of who wins, in the near future every vendor that participated as well as some who didn’t will need to begin wrestling with the consequences and the need to deploy an information utility. This is an inflection point of a major order.