Jonathan Blow

I'd take a thousand silly startup ideas for the chance we land on one Wikipedia. Everyone just does what they can. Mechanical sympathy isn't the only way.

Okay, but ... it doesn't actually work. The theory is that the way these things are programmed is a trade-off, you give up efficiency, you get productivity and reliability etc. But productivity and reliability are both extremely low. And the amount of efficiency traded off was

tremendously higher than claimed. At some point, isn't someone supposed to notice this and question wtf is going on?

The same spread of productivity and reliability exists in web programming as in all others. We gained a boatload of accessibility on the way though. Fairly sure that most of programming is frequently reflective and this isn't news, there's no prize for being annoyed about it.

I find your first assertion here very difficult to agree with if I just look at the numbers.

You keep on mentioning "the numbers" and "these things" like stuff on the web is all the same and it really isn't. I get the ire and frustration, but different categories of solutions fit different problem spaces. There's plenty of utilitarian web things that just work.

Like what? I can't think of a single web site that works for me consistently. By "the numbers" I mean, for example, the number of employee-years per unit of functionality implemented, across various fields.

The implication there is that there's an alternative, non-web, way to implement that same functionality that would somehow magically be better. As if targeting something that isn't the browser is somehow cheaper (which, well, market rates on the app market indicate otherwise).