Consider Urbit. It’s a brand new … something.
[IMPORTANT: Please, when considering it, try and forget that its primary creator is a racist psychopath; and assume, for the moment, that none of his utterly bullshit ideas about “Dark Enlightenment” are somehow baked in.]
It’s full of some interesting ideas. I like the idea of a small, easily-understood dingus that does what you need it to, and only that, and that is at least in theory designed to last more or less forever.
Then there’s Snap. It’s Google trying to fix all the problems they have in datacenters at their scale. Lots of people have problems with datacenters and scale, though; the catch with Snap is it’s effectively getting rid of TCP/IP, by building a new networking protocol.
It is also interesting. Building things without baggage and with the complete knowledge of the past N years, combined with the resources to act on a “big” design, is a hell of a thing.
Both of these things really start from a clean slate, and present big ideas free of existing limits. They are, in their own way, an attempt to move to the Next Big Thing.
My question then is, since neither of them runs a web browser, who gives a fuck?
OK, OK, I know, a data center networking stack isn’t designed to run a browser. I think my point still stands though: the browser is still down in the muck of more-or-less 90’s technology. Yeah. it connects to big data and that’s where a lot of the “real work” happens, so it can be sort-of 90’s.
And yes, weird moonshots like Urbit are in a way just a side-effect of there being way too much money in Silicon Valley, such that it’s all just one big crapshoot, and we might as well fund more-or-less everything because no one knows fuck-all about what is actually going to work. Remember when Twitter seemed pointless and dumb?
[IMPORTANT: Twitter is pointless and dumb, but somehow we’re all stuck with it.]
Anyway, ordinary people will never run Urbit, and I’d wager that it won’t be able to hold onto the attention of anything but the most niche nerds for more than a bit. It doesn’t run the ubiquitous tool of computing: the web browser. Hell even their docs say they intend to get rid of the browser-based “Bridge” at some point. Naturally, I guess; they’re trying to imagine a baggage-free future, so why keep this one remnant of the past around.
My dad used to say that a computer wasn’t really a computer without a printer. This was back when you used computers to compute, then you printed the results on paper. Obviously before networks were commonplace (and wireless!). And now printers are just for legal documents, flyers, and D&D character sheets.
Same goes for things like USB sticks; I keep a few around for emergency safe/recovery boot. They’re sort of pointless for a lot of workaday problems now.
We let the baggage of the past go all the time. But we also don’t: every computer in my house runs some sort of Unix or Unix-like. We still have files as streams of bytes on a disk (although it’s not a spinner platter of rust any more). We still have graphics drivers and color depths and SMTP.
Does everything depend on a giant pile of yesterday’s baggage, which is why moonshots seem so weird and alien? Can we ever move past this stuff without slow, incremental, buggy, expensive rewrites of … literally everything?