Site Meter

Monday, November 19, 2007

client/server vs. p2p

I realized that in some ways, what I'm been struggling with lately is a matter of architecture.
In a client/server approach, the resources are collected together in centralized places and server programs are your interface to those resources. In a peer to peer approach, there are many sources of information. In fact each node is on the same level. I feel like in each of the projects I was working on, I was being pushed to either be a client or a server, and I didn't particularly want to be either.

I know that hierarchy has its place, but on the whole, I'm much more comfortable with a p2p approach to many aspects of work and life.

Monday, November 12, 2007

back to physics

Ok, I need to do some integrals. I'm trying to find the number of electrons scattering per unit time into a given longitudinal velocity.
I need to do the integral:

where is typically a Gaussian.
I can do the integral first which seems like the right thing to do. But then to check it, I can integrate over from the momentum acceptance to infinity and get the Touschek lifetime. But I can't see how it gives the right answer. Anyway, its fun to get back to this stuff.

Saturday, November 10, 2007


I've been trying to understand a certain piece of technical history that is important for accelerator physics. It involves people at each stage and their own skill sets. I'll leave it somewhat vague for now, because I don't understand it well enough. On one end we have various math ideas which from a certain perspective join numerical and analytical approaches, but from another, its just some math. These go under the names differential algebra, non-standard analysis and truncated power series algebra. At the other end of the bridge we have collaborative work on designing a particle accelerator.

The thing is that in some ways this is really just a personal bridge. I am comfortable reading math. And I am comfortable working in a team with an open environment. But in between has been a huge mess.

I was just trying to find references to Foucault's The Order of Things. I found this essay.
What the author says is that Foucault's grand schemes aren't particularly new, and his referencing is pretty poor, but when discussing particulars, he adds new depth.

Continuing my free association: I was just listening to this radio show starring Richard Stallman!