Top Links You Must Click On
Linux Business News
So How Come the New Economy Bombed
So How Come the New Economy Bombed
By: Maureen O'Gara
Oct. 25, 2002 12:00 AM
Well, it seems it wasn't just God's work smiting - like Sodom and Gomorra - the insufferably stupid, vapid, and hollow ideas that flourished under the New Economy's brief sojourn and the insufferably greedy and undeserving people who went along with them that ultimately did the New Economy in.
As comforting as that thought might be, it turns out it might have been the architecture.
Tim Negris, the guy who coined the phrase "thin client" when he was a VP at Oracle, which was before he was a VP at IBM Software, and sat out the Internet bubble thinking, and putting his money into real estate - right away an indication that he's smart - figures the Internet - and with it the New Economy - tripped over its own shoelaces because it was based on an incarnation of client/server technology when it should have been peer-to-peer.
If it had been peer-to-peer, or if it used the newfangled Network Business Model that Tim calls "Peer Services Computing," the New Economists wouldn't have needed all that money to spend buying Sun servers, whose ROI was out to heeere, and pressured them into trying to get lots of traffic in an impossibly short period of time at high enough fees to pay for all the fancy technology.
There was a basic economic disconnect between the client/server model and the Internet.
Client/server widgetry is Old Economy overkill, Tim says, wasting a thousand trillion CPU cycles, one hundred trillion bytes of memory and ten thousand trillion bytes of disk space every second of every day.
Tim compares the New Economy to the Afghan economy and calls it a "chimerical abstraction projected onto a lawless and primitive territory." Client/server didn't so much kill the New Economy as keep it from being born, he says, forcing it into old-style restrictions like consolidating data on centralized database servers - an article of client/server faith that didn't work for the New Economy. Like just who exactly owned the data in those big database servers?
P2P or a "peer grid," on the other hand, seems to be just the ticket. It offers equal or proportionate direct technology cost-sharing among the participants - costs are directly tied to need and the widgetry is relatively cheap - distributed workload sharing across participating systems and owner-controlled information visibility, security, privacy and use.
Of course, all the pieces needed to build Peer Services Computing aren't quite here yet. Web services, for instances, the latest identity crisis-prone techno craze, are currently conceived of as a one-way street rather than as something that will go out and actively round up consumers. Peer Services, Tim says, are active services and consuming and providing are flip sides of the same coin.
Peer Services - for reasons of "invertible security," control, efficiency and flexibility - need some data stored and managed privately on participating peers and other data aggregated on the operator's system, also technically a peer, a situation that will require a new range of information management capabilities that don't exist yet.
They will need, for instance, the kind of data management that can separate out the metadata from the data itself across systems and a singular central set of data definitions and XML structures so Peer Services apps can function properly. And then, since the data needs to be live and capable of advertising itself, describing itself and authorizing itself, it will need new facilities for distributed transactional metadata that allow searching and selecting, say, to be done in a safe shared fashion, which introduces the idea of meta-metadata or data about data about data, but then you just have to read Tim's Manifesto at www. Equinom.com.
Equinom Inc is the new company Tim and folks like CTO Lynne Thieme and VP of engineering David Thieme, her brother, whose combined fingerprints are all over such technologies as IBM's Fast Path transaction processing and massively parallel databases as well as thin clients, have put together in Carmel, California.
Egged on by technical advisor Robin Bloor, head of Bloor Research, business advisor John Burns, a principal in Frontier Risk Capital Management, and marketing advisor Scott Anderson, a principal in The Shadow Marketing Network, they are developing the enabling software for transactional P2P, enhanced Web services and the peer services building blocks that other software developers can use to build Peer Services middleware and applications. And to make sure the notions take - they figure it will only happen through broad participation by industry, government and the academe - they're going to make their software available free of charge to schools, non-profit trade groups and government agencies.
A proof-of-concept should be out in the next few weeks.
Reader Feedback: Page 1 of 1
Enterprise Open Source Magazine Latest Stories . . .
Subscribe to the World's Most Powerful Newsletters
Subscribe to Our Rss Feeds & Get Your SYS-CON News Live!
SYS-CON Featured Whitepapers
Most Read This Week