The Mozilla project was wrongheaded from the start. Anyone who genuinely believed in open source could have known that KHTML was better quality code and so it proved, despite vastly greater resources being poured into the terrible Netscape codebase. (I can't help thinking this was largely jingoistic Americans preferring an American project).

Some good things have come out of Mozilla-the-organisation - I very much hope that Rust/Servo is a success. But when it actually comes to developing an open-source browser, the incentives of a donation-funded foundation like Mozilla are all wrong.

I don't know what the right way to fund open-source development is. Dual licensing has its share of failures. So does trying to make it a direct business. So do research grants. Partly it's just the tragedy of the commons. In my darkest days I wonder if open source is fundamentally doomed because it simply can't make the monetary incentives line up with good engineering practice.