I remember first seeing Gemini a while ago, and thought, "cool project, but I probably won't ever use it". And I haven't. If I'd want to build a browser, I want there to be a lot of content for me out there already for me to browse.

It's funny, because I think designing a new protocol and markup language, and then building both server and client software for it, and then actually convincing millions of people to use it... well that sounds orders of magnitude harder than writing a web browser, even with all the crazy and terrible protocols and APIs I'd be expected to build.

And sure, I could set my sights lower and try to build something with much more modest ambitions, only hoping for a few thousands of people to use it -- at most! -- but that doesn't sound interesting to me. I'd still have to use the web for all the other things I need and want to use the web for. Maybe this sort of thing does sound interesting to some people. That's great, go for it!