funding is bad, has been for a while, unlikely the internet backbone would get funded today
nice thing about computing in the 60s and 70s is you couldn't make a living at it, so it attracted people who were actually interested in it
the problem with computing is that it's not yet a field
there's nothing like detail to attract people who design a curriculum
consider music; teach the C major scale, it's easy to teach, easy to test, but has nothing to do with music (?!?)
a project that has almost no hope for success, and will take us five years... to fail...
the important thing is outlook, perspective, epistemological foundation
discussion about why universities don't build their own computers and operating systems
Touretsky: "yeah, we don't grow our own food, either"
kay's claim that computers today are 1000 times less efficient than some indeterminate time in the past?
Kay criticizes the slowness of memory architecture, saying it's abysmally slow because it doesn't satisfy moore's law --- and yet his condemnation of today's computers as being "stuck in the past" is based on computations involving moore's law. Wtf.
Admission policy: If anyone's remotely interesting, let 'em in, give 'em two years.
crazy erlang australian hacker?
He thinks FPGAs are a way to change fundamental thinking about programming? I feel like this betrays a mistaken attachment to hardware as determining conceptual structure.
Talks about "doing experiments" but I doubt these are "experiments" in the same sense I expect, and more like merely "some things we tried".
Tradeoff between "coping" and "programming" ~ the reality tends to be that we have to cope with others' bullshit, and that we consider good programmers those that are good copers, but Kay argues good programmers that are those that become physically ill at badly designed things.
evolutionary misunderstanding: "we must be at the optimal spot, because look at how many millions of people are participating" but evolution satisfices, doesn't optimize.
A good point: since UIs are not set up to teach people how to use them (video games being an exception) interface designers are forced to rely on existing systems.
Butler Lampson's Draper Award talk
NSF proposal
"If there's one thing I would try to get every child to [learn and] embed deeply, it would be anthropology... it's the best way to not take yourself and your culture too seriously"
main idea of ARPA/PARC research community: "no center". But we have root name servers!
I like talking about the internet since I didn't have anything to do with it... it's not like saying smalltalk is better than java.
Another good point: "If an object (i.e. a node on the internet) crashes because of a message sent to it, it is the object's own fault. It can interpret the bits any way it wants, and it's not like we're sending a thousand volts over the wire."
"The most important thing you can do on a computer today is confinement."
"The reason we don't have a field is you don't know this [about the B5000]"
his admonishment of computer scientists for not paying enough attention history is completely appropriate in principle ~ but physics knows to discard phlogiston, while keeping Newton.
"the purpose of any educational system is to help the students understand the threshold of quality"
ridiculous argument by etymology that architecture is the making of arches, and infuriating argument that any large piece of software "lacks architecture" merely by virtue of its size.
some bullshit aphorisms about artists being those who don't "merely" see the present in terms of the past, as if it's not inevitable that some predictions will succeed by virtue of luck alone