Gemini is probably not exactly what GP wished for, but it has something resembling a critical mass of users and while I do not think the text format (gemtext, i.e. gmi) is perfect I find it good enough for what it does.

I even use gemtext now and then offline just as an even simpler markdown. Since it has so few features it is trivial to convert gmi to markdown or to any other format without losing anything. It works as a lowest common markup language for when something that minimal is enough.

Not tens of millions of lines of code?

So many choices!

You mean tens of millions of consumers?

That would actually make sense. But it would also point at the larger problem (the one we're not actually looking into because we're too busy with solving the unsolvable a.k.a. with C++).

It goes like this: are "we" building a browser for it to be reckoned with, or are "we" building a browser in order to let people browse webpages?

Because only one of these two requires collecting tens of million of... pretty much anything where ten million is a large number I guess? Yet people conflate the two, thinking the same approach holds for both goals, so let's put it sideways:

Which exact problem does a new browser (engine) solve, besides people saying there are too few browsers? What's the purpose of having this problem, its underlying nature? Can we solve it a way that doesn't require reimplementing the last 30 years of computing history? Can we even go look for such a way or will someone show up to stop us?

If the goal is to become a browser vendor, obviously there's no workaround to building a browser (or rebadging one lol); if the goal is not that or not only that, anyone building a browser is gonna have to expound a little more on what exactly they're trying to achieve. That's complicated by how the vast difference between a new browser engine and, say, a new model of TV set, can't really be expressed in beancounts.