« December 30, 2003 | Main | January 05, 2004 »
December 31, 2003
Announcing Cowster
No, its not a dating service for cowboys. Cowster is the nickname I've given to the notion of using Web services and XML to build a cattle tracking system. As Dave Fletcher points out The USDA is moving full speed ahead on a plan to implement a tracking system.
Under the plan, now in draft form, individual animals would be tagged with a unique number allowing officials to more quickly determine where it came from, where it has been and herds that it may have come in contact with. It took U.S. investigators at least a few days to trace the infected Holstein from a Yakima County farm to its birth herd in Alberta, prompting calls for a national tracking system that would have enabled officials to response faster. Investigators reviewed cattle sale records to locate the origin of the infected Washington cow, but were slowed by lack of complete records. Information about the animals and its owner would be entered into a national central database to be accessed by state and federal officials in case of an outbreak such as mad cow.From USDA to speed up implementation of tracking system
Referenced Wed Dec 31 2003 11:44:39 GMT-0700
Its that last piece that has people worried. Quoting from an Associated Press story:
Resistance to the plan has come from meat producers who don't trust the idea of establishing a central database that would allow the government or rivals to know detailed information about their operations.From deseretnews.com | Scare could hone cattle tracking
Referenced Wed Dec 31 2003 11:46:38 GMT-0700
This is a classic identity and privacy problem, although, of course, its not the cows themselves that are worried about their privacy.
I used to work in the nuclear industry and we knew the history of every single part that went into a submarine power plant. You pick out the smallest bolt and follow its history back to the day the ore was pulled out of the mine. Think about what that entailed in the 50's, 60's and 70's. There were (probably still are) large buildings full of people in Mechanicsburg, PA whose job it was to manage the supply system for the Nuclear Navy. Doing this for cattle is huge in scale, but modern technology has made it possible as long as we don't try to recreate those buildings full of people.
The answer is for the Government to avoid building a big centralized database---that's so 20th century. Rather, the USDA ought to establish XML schemas and Web services interfaces and let private industry establish the data repositories. If I'm a rancher, I have established relationship with entities whom I trust. Let those entities keep the data for their customers. Don't aggregate the data about a specific cow in one spot, but let the data and the systems stay distributed until its queried. If we can use FOAF to keep track of relationships between people, why not a similar scheme to keep track of a cow's history from birth to bistro?
12:06 PM | Recommend This | Print This
Steve Gillmor on 2003
Tis the season of best and worst lists and Steve Gillmor has one that I pretty much agree with on both sides. One thing he bring up that I don't think has gotten enough attention is Sun's news licensing model:
Sun licensing model÷You may argue whether Sun has a chance against Office with its Java Desktop System (JDS), but there's no doubt McNealy and software chief Jonathan Schwartz have altered the enterprise licensing landscape with their $100-a-seat licensing model. The real game-changer: that's per employee, not user, internal and external. A hundred or a million÷same price. It's a marriage made in heaven for Web services, an eBay, an Amazon, a salesforce.com or even a virtual McDonalds. "Over a billion served" takes on a new meaning for the bean counters.From Best and Worst of Messaging & Collaboration in '03
Referenced Wed Dec 31 2003 10:01:06 GMT-0700
As Steve says, this is especially important for new service-based business models. One of the things that really frustrated when I started building big systems was that everyone wanted to be my partner. Not in the sense that they wanted to help me, but they wanted a cut based on how many customers I had. I thought that was the deal shareholders had, but enterprise software vendors want that deal too. Moreover, they want their cut first. Flat licensing changes that and makes some new businesses economically feasible.