Slashdot Mirror

The best list I've seen in the past year was the one published by Marc Andreessen. I've worked my way through almost all of these now and, aside from one or two clunkers, its a stellar list of books and authors I had not heard of. http://blog.pmarca.com/2007/06/top_10_science_.html

http://blog.pmarca.com/2008/03/an-hour-and-a-h.html

An interesting read, rather than just the snippet in the article.

If not more so for not actually linking to HIS blog post, rather quoting it and then linking to someone else's....

Ugh

tm

Agreed, but you'll have to place them in different positions to be effective:

I agree that to the extent your team specializes, specialist knowledge is useful. But my view is that treating creative work like assembly-line work is a regrettably common error, and that software development should always be creative work. (When it's not, people almost always just buy a copy rather than coding a duplicate from scratch.)

In a creative work, seemingly minor changes can have systemic effects. In fact, they should have systemic effects; if a developer spends a day coding and it has no effect on end users and the business, it's probably a wasted day. Given that, the more a developer understands about the context, the better they are able to achieve product/market fit. So along with a solid CS grounding, a liberal education is essential.

Even if you follow a strongly specialized model, however, people will still need the basics to integrate well. Unless I understand a bit about psychology, a designer's requests will seem retarded to me. Without some understanding of business, I won't be able to give product management good cost/benefit tradeoffs. Without an inkling of the law, I may not know enough to ask my project manager about libraries I'm including. And so on.

If you have a person that spent time taking macrame and ice fishing in college, do not expect that person to be able to solve driver problems, most likely does not have the knowledge or the inclination to fix it in the first place.

Could be. Depends on the person, really. One of the most important characteristics in a good programmer is intellectual curiosity. This often leads people on strange paths. I'm much more afraid of the applicant who got a narrow, vocational CS degree as the quick path to a good job than I am one who has a CS degree plus a lot of apparently unrelated coursework. In my experience, the former often turn out to be clock-punchers. The latter group, if you can get them hooked on the job, will be reading things in the off hours just for the fun of it.

Andreesen did alright with Loudcloud, sold it to HP for $1.6 billion:

http://blog.pmarca.com/2007/07/hp-buys-my-comp.html

Goahead and read Marc Anderson's(Netscape, Ning) rebuttal to Economist's article. You will know how worst is Economist analysis.
http://blog.pmarca.com/2007/12/when-non-techno.html

This is kind-of a follow-up to his in-depth thoughts on the Facebook platform that I found really useful, too.

This is kind-of a follow-up to his in-depth thoughts on the Facebook platform that I found really useful, too.