Friday, November 28, 2008

Notes on Kranakis: Constructing a Bridge

- There were no more than 30 civil engineers in the US in 1815.
- Finley used various information sources for his early suspension bridge. Historical precedent -- Chinese bridges -- provided the inspiration and the Encyclopedia Britannica (1797) provided the strength of materials. He also built a model, perhaps getting inspiration form Abraham Rees's Cyclopeadia and it's entry on bridges. He also may have consulted a work like Gadson's Geometrical rules made easy for the use of mechanicks concerned in building (1739).
- He was primiarly working before 1812 and the formalization of engineering practice.
- Navier used -- but was critical of -- Belidor's Architecture hydraulique.
- "Relatively little has been written about the history of the maintenance and repair of engineering structures, but this history is at least as long and complex as the history of their initial construction, and equally important." p. 189
- "Finley used models, and experimental apparatus comprising cables, pulleys, and weights. He also sought undertanding through encyclopedia articales on bridges and on the strenth of iron, and through the construciton (in 1801) of a 7o-foot trial span over a stream close to Uniontown. Navier's tools were mainly mathematical: calculus, Fourier series, elasticity theory, mathematical statics, and a variety of specific equations such as the equation for belt friction. Navier also sought understanding through inspection of existing British bridges, through discussions with the builders of some of those bridges, and through published literature, not only on suspension bridges, but also on mathematics, theoretical mechanics, and elasticity theory." p. 199
- Manuals by Dennis Hart Mahan and Charles Storrow came from European sources.
- Check: Charlton, History of theory of structures in the nineteenth century.
- West Point, established 1802. Rensselaer, established 1825. A number of engineering schools opened between 1830 and 1860: Brown, Harvard, Yale, Wesleyan, University of Michigan, New York University, Dartmouth, Rutgers, Indiana University, Cincinnati College, University of Pennsylvania, University of Virginia, University of Georgia, University of Maryland.
- Morrill Act opened up another slew of schools after 1862.
- Note: American inventors followed the patent literature which quickly adopted standard presentation. Not so much for the French.
- CHK: Dorn (Dissertation), The art of building and the science of mechanics: A study of the union of theory and practice in the early history of structural analysis.
- CHK: Danko (Dissertation), The evolution of the simple truss bridge, 1790-1850.
- CHK: Hindle and Lubar, Engines of Change.
- CHK: Weiss, THe making of technological man: The social origins of French engineering education.

Thursday, November 27, 2008

SaaS, market making, and the Cloud

I've recently begun to increase my coverage of virtualization, particularly application virtualization. The intersection of virtualization and my typical enterprise application coverage has afforded some interesting insights. Many of the major vendors are now positioning a Carr-esque vision of utility computing. While the applications themselves may not necessarily live somewhere out on the cloud, extra capacity and provisioning might.

So it seems like there is now a viable vision for utility computing. There is actually some promise in the Cloud -- providing, of course, that enterprises make the requisite investments in virtualization to encapsulate and consolidate their application stacks.

I'm now wondering if there will be a fluid market in computing capacity. Will it trade the same way that we currently trade electricity or other commodities? What has to happen to fulfill this vision?

My thoughts:

1. Standardization. There has to be some sort of standardization on the types of services provided by different producers. There could be no universal market for electricity until there was some sort of standardization on service e.g., AC vs. DC, current, etc. Different cloud providers are currently offering very different types of services so it's certainly not a commodity.
2. Ratings. Assuming that we get standardization -- driven either by industry standards or by legislated standards -- then it becomes possible to compare different offerings. But there must be a mechanism that accounts for risk in these comparables. We would need some sort of D&B, S&P, Moody's.
3. Sufficient liquidity. There has to be enough market activity to support some sort of open exchange.
4. Market makers. Somebody is going to have to be the market maker for the emerging market in order to set prices, increase information, and improve process efficiency.

But could this process be bi-directional? Could anyone with a data center sell additional capacity into the web? This notion seems to resonate with some of the work currently being done in the electrical markets.