Those are the things that have literally no value me. You'd think the internet would be large enough to address pretty much everybody's needs, including mine, but it's getting pretty clear that it's not.
Those are the things that have literally no value me. You'd think the internet would be large enough to address pretty much everybody's needs, including mine, but it's getting pretty clear that it's not.
It'd be great if everyone started making their own Neocities site. Or even just drop contemporary social media and join the new Friendster (https://news.ycombinator.com/item?id=38021802) or Spacehey. But the cultural impetus just isn't there.
If I have a bias about audience size, it tends to be toward "smaller is better".
Now, given the near-zero up-front costs to making software, there should still be enough room for all the niche needs, but the annoying thing is, the computing ecosystem itself - the hardware, the software platforms (OS, browsers) and even the tools used to make them - it's all being optimized for the mass market / lowest common denominator. As silly as it is, professionals can't get good tools, because the tooling is caught in the gravity well of more generic, mass market products.
The part that really gets me is how this starts making effective use of computers ("bicycle for the mind" stuff) impossible. My go-to example: it doesn't matter if you figure out how to make fully open source & open hardware smartphones for nerds. Even if you make them competitive on price and power with mass-market products. It doesn't matter that you somehow hired John Ousterhout and Edward Tufte to make the maximally ergonomic and functional apps for the platform. I'm still going to buy a regular locked down smartphone, because I need one to be able to use my bank account, and my bank - like all other banks - demand you use a locked-down device from a major vendor, with full device attestation ("because security").
The freedom of computing stops at the network.
This also happened in hardware.
Back in the day we could design wide range devices using quite widely available "mil-spec" semiconductors and components - a common difference was in logic circuits where you had 74-series in consumer and commercial (0 °C to 70 °C and −40 ° to 85 °C temperature ranges) and milspec 54-series that would suffer −55 °C to 125 °C.
Following something in the 90s called the Perry initiative IIRC, the rules changed to test-based performance that enhanced "market supply" rather than prescribed manufacturing methods, so after STD-883 almost all wide range components disappeared.
Sure if you're NASA or the US DoD you can get stuff made, but increasingly everyone has to source from the same few commercial suppliers. The upshot is that if you're organising an Antarctic survey, or going into the desert it's almost impossible to kit our with modern gear that won't fail. You're stuck trawling eBay for some 1980s Soviet stuff.