There’s very little WebKit / KHTML code left in Chrome.

There's very little NT 3.1 code left in Windows, but it's still clear how old that project is.

All unixes are descendants of original Unix from PDP-7 days. Why does it matter?

...ok the /bin /usr/bin nonsense still stays after decades, maybe you have a point

Windows, macOS and Linux all have legacies of the late 80s and early 90s.

> KHTML+KJS released in 1998, via WebKit from 2001, in 2008 it gained the Chrome name, but the code has more than 20 years of legacy.

Why do you want the oldest code to be less than 20 years old? Why is that "nice"?

Because 20 years is like half the history of modern computing. A lot has changed in a small amount of time

So what? Code doesn't rust.

It doesn't but it accumulates cruft and since then new libraries emerge which you might be able to reuse instead of writing your own thing. Just as an example: Boost first appeared in 1999 so very likely at least early on no one used it.

Of course you could do that, but the existence of a bit of code that's survived that long within a project that's been around for 20 years doesn't mean nothing new has happened. Mozilla invented a whole language to make it easier to write browser in; I don't think they won't have considered using Boost or whatever much less radical approach we might come up with here won't have been considered and invested in.

Funny you mention "rust" because many modern things are written in rust to prevent bugs and generally to use modern programming paradigms.

The last time a major browser originated, RAM was measured in MB, CPU freq in MHz, and the iPod was the thing that the one trend hunter your friend knew was about to buy.

The major browser platform today, smartphones, did not exist. PDAs did not even have wireless internet yet.

The basis for the functionality of the browser is due for a reimagining.