Jonathan Blow

(For the record, I don’t place myself in the top tier re understanding of speed or anything else. I am somewhere in the middle of that gradient between the 99.9% and the People Who Really Know.)

We see all this bad rhetoric claiming “system X is only 2x slower than native code therefore it’s fast”... but one must ignore rationalizations and look at the actual output, which is several orders of magnitude inefficient. Few people are willing to put 2 and 2 together here.

The most common objection to these points is "we write slow software because it lets us make things faster and more easily". I agree this is the common belief, but it's wrong. If development is so much easier, why is productivity approaching 0 over time?

I’m curious if you have any sources or stats about this.

Just look at how many engineers a web company has (like Twitter), and look at how much the functionality of the site changes over time. Divide (dfunctionality/dt) by number of engineers to get average productivity.

I think you underestimate the work done at companies like this. My guess is that 75% of projects never get in front of users (because they’re deemed not good enough in some way), and 90% is invisible to users anyway (internal tools, ad delivery/markets, custom work for clients).

You could call these % lacking productivity but I don’t think so. In game dev most stuff isn’t released, it’s tests to see what works before the final version. Would The Witness really take 5 yrs? Not if you already knew what the final product looked like, but to get there? Yes!

So what did the Witness ship after 7 years of development? -A custom, high performance game engine -A rich 3D environment and soundscape, built with guidance from real-world architects -Approx. 500 puzzles that require keen awareness of the environment to solve