Greetings Pen-'Ellers, KGC writes, But there is an idea floating around geekdom that the Web works (in the sense that it scales 5B+ documents, something which no one really expected) because of various purely technological ideas (most of which get attributed, inaccurately, to Tim Berners-Lee). I want to engage this idea in my book (for my own nefarious, leftie political reasons) and my publisher is cool with me doing a bit of "politics of technology".
Me, Clay Shirky writes about the economics of what makes the web work. Has some theories about various ideas floating around about the IT industry that are a starting place to think about what works and doesn't work about Web Services, etc.,. http://www.shirky.com/ Hal Varian writes a column for the NY Times and teaches the economics of information technology at UC Berkeley. He may have some specifics for you to track down about hardware spending versus, software ideas like Tim Berners-Lee might represent in the public mind. http://www.sims.berkeley.edu/~hal/ Varians University web site. http://www.sims.berkeley.edu/resources/infoecon/ A site of his that gives research sites for information about the information economy. Doug Henwood has a new book out sometime in the next decade (has been promised for more than a year so far) about the new economy in which he gives an economic accounting of the basic area you are interested in. Look up Ian Foster who is chief scientist on the GRID, which is a new internet like technology for super computing on a large scale. The cost of hardware for this project where it is academic related is probably public information. Therefore you can get an idea about the relative cost of building a new internet. And this system probably is a requirement for the success of Web Services in the long run, so it gives some insight now of what Web Services has to bring along to work right. This question is like asking what Open Source software brings in value to business. So you might look at Servers and costs selling them. IBM servers (along with other companies) are driving Sun Microsystems into the ground by utilizing Open Source software. This gives some idea of what theory (or human labor) provides over hardware. Especially look at how the relative updating cost for Sun are higher than the brand new installation of IBM servers. Not easy comparison, but perhaps gives some insight. Look at labor costs overseas like India for IT because that makes theory much cheaper to use. Because that is what you mean by theory I think is labor costs. You might clarify your thinking about that issue of theory versus hardware in technology terms also. For example, historically for a lefty what is the path toward programming? Writing. What about memory in computer? The public libraries. Writing - Let's take color in magazines (being print media closely tied to traditional typescript), which gradually increased from the 1920's onward. Color represents a major increase in costs and production for photographs. Throughout the 20th century color photographs were basically just one big frill on the ass of the printing trade. So when we talk about computing and web services we might ask where the sheer productive volume of writing theory actually is merited by Web Services. Hal Varian gives some bench marks about the sheer volume of information being produced, tv, x-rays, written text etc. So the value of theory can be understood in some ways by the general increase in the volume of produced writing. One can take radio transmission, tv transmission etc. as fancy sorts of writing because they transmit words also. Some people argue that the value of that sort of stuff declines to near nothing in the present computing environment. Copying costs being just about nil as Clay Shirky would argue. However, the value of theory in terms of writing would the vast increase of unit volume of writing. And because of that a transformation of the sheer structure of writing in some analogy like black and white photos going over to color. We don't exactly foresee what makes a big increase in production of information important, because our culture never had this option. Printing in some ways was a big increase, but the volume increase of memory coming, Terrabyte hard discs, allows us to think in terms of tens of thousands of movies stored to use in theory making. So instead of the of few kilobytes on this list, theory would entail a gigabyte structured into meaningful writing or whatever people will end up calling what this points at. So in that sense I am opposed to Doug Henwood's (amongst others) view of the economics of theory and software in Information Technology. I think theory can actually be looked at in terms of volume of product attached to all volume of production in economic terms. Doyle
