|By Roger Strukhoff||
|August 7, 2014 11:45 PM EDT||
A couple of days ago, I noted the fundamental enterprise IT change driven over the past generation by user impatience - nobody likes waiting in line.
We can point to Moore's Law, Metcalf's Law, and a general business trend toward disintermediation and decentralization as well, taking note of the millions of hard-working people all over the world who've been making things faster, better, cheaper, and more effective for several decades.
All of this effort has led to the big change in IT during my career-the invention of plumbing.
Going to the Well
The bad, old way of MIS meant you went to the well to get your water. You visited the wellmaster, er, MIS manager, and hoped your ration, er, report, was ready. Sometimes MIS would bring the water to you, but only if you didn't bug them for it very much.
The minicomputers came, and with them, plumbing. Information flowed in two directions, and more or less when users wanted it. It had moved from the well into several silos, and you needed a different terminal to get at the stuff in each separate silo.
So soon enough, user impatience re-appeared. They wanted several types of information, and they wanted it to be available on their desks, in one place, all the time. In came PCs, approved or not, and forced IS to get better plumbing, usually a miraculous local network called Ethernet.
What Did They Ever Do For Us?
But users are wont to grumble and complain, and as with Monty Python's Biblical revolutionaries asking what the Romans had ever done for them, modern-day computer users didn't appreciate all the plumbing that IS had so graciously provided.
Once the world at large discovered the Internet, then the Worldwide Web, the old ways were doomed. There's been a constant back-and-forth battle for control between IT and users since that time, but modern-day information plumbing-much of it wireless-continues to be this nihilistic force that trumps the most rigorous IT efforts to control the flow.
With the great plumbing afforded by the Internet and Web came the battle to destroy those silos that had replaced the original well. This battle will no doubt continue for many more decades, as each new type of information flow creates new silos.
Data types are data types, archives are archives, and an enterprise's information-one of its most valuable assets-just can't be tossed around willy nilly, like the records, books, newspapers, magazines, and laundry in my old college dorm room. It must be organized somehow.
Yet user demands for access to any information, anytime, anywhere, on any device are an inexorable force. The nearly 20-year odyssey of today's @CloudExpo @ThingsExpo and everything that goes with the event started with application service providers and web services.
The idea of deconstructing (or decoupling) what applications do from their silos was a profound, radical thought. It was also easier said than done.
Once developers and users started to think of software as a service rather than a product, though, they also started to think about how to measure it. An application is bought in a package or a download; a service is something that comes streaming in as you need it.
Following that thought, why not, then, combine specific services with one another to create a new service? Set theory and a facility for high-level mathematical proofs started to influence the thinking of how best to develop and deliver services.
SOA, Then Cloud
Web services extended into Service-Oriented Architecture (SOA), and a further decoupling at the hardware level (promulgated by software) brought us into the modern age of cloud computing.
The Internet of Things and Big Data are now possible with cloud computing, now that compute resources can appear infinite and highly flexible.
I wrote a piece once about how this sort of deconstruction was similar to revolutions in 20th century music, art, literature, and philosophy. A friend of mine said at least six people in the world would be interested in my piece-I can now say that, in fact, almost 5,000 people have read it over a time period of only five years. Here it is if you need some sleep.
Modern-day distributed computing can also be said to be simply a result of a human desire to improve things. It's commonly accepted wisdom that we shouldn't try to re-invent the wheel, but that's exactly what people do.
The ironic consequence of all this is now people have questions about the plumbing. Can the Internet withstand the onslaught of the coming zettabytes? What protocols, specifications, and techniques will be required to keep local information flows as local as possible, so as not to burden the bigger pipes?
We all know that the late Sen. Ted Stevens was right. The Internet is a series of tubes (or pipes), if structured with a little more subtlety and complexity than he articulated.
Now that the plumbing has addressed user impatience and improved user experiences for decades, will it become the bottleneck itself? And will a great concentrations of information among a small number of cloud services providers some day bring us back to having to depend on the well and its keepers for our cyber-survival?
Apr. 30, 2016 04:00 AM EDT Reads: 1,700
Apr. 29, 2016 09:15 PM EDT Reads: 320
Apr. 29, 2016 09:00 PM EDT Reads: 1,545
Apr. 29, 2016 08:30 PM EDT Reads: 1,750
Apr. 29, 2016 07:30 PM EDT Reads: 2,403
Apr. 29, 2016 05:15 PM EDT Reads: 1,698
Apr. 29, 2016 04:30 PM EDT Reads: 1,779
Apr. 29, 2016 04:30 PM EDT Reads: 641
Apr. 29, 2016 03:45 PM EDT Reads: 1,630
Apr. 29, 2016 03:30 PM EDT Reads: 1,671
Apr. 29, 2016 03:15 PM EDT Reads: 1,552
Apr. 29, 2016 03:00 PM EDT Reads: 866
Apr. 29, 2016 02:45 PM EDT Reads: 876
Apr. 29, 2016 02:30 PM EDT Reads: 1,410