Almost everything on computers is perceptually slower than it was in 1983 (2017)

Can't deny this. But there's a few other factors that I feel when using a computer.

- linearity: logical steps are simple (boot, os, login, byte fetching)

- predictability: it can be non instant but I know what's coming

I had an old pentest distro with linux 2.4 on a key and a few recent ones using systemd and tailored DE. Whenever I used the old 2.4 era distro, maybe I had less capabilities overall, maybe a few glitches here and there, but somehow I felt in contact. Everything reacted as-is, throughput felt large (console output, file IO). Even the old sequential boot had a 'charm' to it. Even though systemd is so vastly superior in all aspects .. there was something that appealed to me. Weird.

>I hypothesize that this is because programmers and other various Computer People consider "being on the computer" to be a satisfying goal in itself that sometimes has positive side effects in the real world, while everyone else simply has stockholm syndrome.

Oh believe me, as much as I love being on the computer for the sake of it, I don't enjoy having that screen time utterly wasted by shit software. There's only so many hours in a day after all.

I work w/ ERP software, so I see people still using text-mode UIs on a daily basis (hell I track my time w/ one), and I also support a "modern" ERP that is GUI based and can "run inside a browser." (Which nobody actually does, because it sucks, doesn't support all features, and swallows up tons of keyboard shortcuts.) One of these packages can be run comfortably from a $5 Linux VPS. The other package asks for two _very fast_ SAS storage arrays, 32GB of RAM, minimum of 6 CPU cores, etc. Of course you've gotta license Windows for all of that, which thankfully is not my job. (That or you spin it up in "the cloud" I guess, and pay thousands of dollars annually to rent somebody else's computer, since this software is not "cloud native" at all, no matter what their sales people say.)

I try to leave the software better/faster/more usable than I found it, but it's hard when the upstream vendor is just piling shit on the fire so they can pitch their half-implemented features on the sales brochure: with absolute no regard for the added operational overhead of the garbage code.

Can the mods un-dupe this? That was 2 months ago, and it's generated fruitful discussion on a hacker-relevant matter. Certainly more relevant than the Atlas Obscura "look at this quirky history" articles that get upvoted so much.

That's probably because he had his network connection priorities ordered so that the TB networking had priority and there wasn't a valid route on the connected network to the TB display. That's very likely a fixable problem, as I have encountered similar and resolved it.

I think if something is possible to do given better hardware, no matter how bloated or useless, developers at large will end up doing it. I don't think RAM usage or performance are concerns for the people that go straight to Electron because in the present it's easy develop for and nobody has the right to force them to chose otherwise. Electron isn't going away; the only ways it improves are Chrome getting more performant or computers getting even faster. Either one is nontrivial to work on as an outsider - there are only two significant browser engines in the present age because they've become so complex and accumulated so many features that they require the support of well-funded corporations to maintain them, and computer processors can only become so small.

Browsers on the whole will not lose features. The web is ubiqutous and cutting down the number of features to maintain would break existing webpages. The resulting web standard establishes a hard baseline on the amount of complexity needed to engineer a standards-compliant browser. That in turn means a hard baseline on the amount of processing power required to run it acceptably, save for more performance gains that will surely be buried immediately under even more features.

I once tried using Firefox 1.0 and found it was impossible since it has no understanding of modern security protocols. The web as viewed by Firefox 1.0 in 2003 no longer exists. You can't practically divorce the features necessary to use the web such as new protocol compatibility with the extra features like Pocket in the code changes. And trying to do so anyway leads to Pale Moon where a scant few contributors are responsible for merging in dozens of security patches - which themselves come about due to the growing complexity of the web platform. Numerous people have called using it irresponsible from a security standpoint. In the end the endless growth of web standards and the code needed to support them has made that fact a necessary one.

For all practical purposes, you either use one of the two major browsers or get left behind by coders and organizations that have orders of magnitude more resources than you.

Society and the economy don't incentivize creators with being happy with a modest amount of features if there's a glaring increment in performance that ought to be seized on. It's about growth and innovation. Finding new ways to reinvent the wheel and get further away from the hardware, because it's now easy to.

Sometimes I wish we could undo growth. It would be nice if Electron and the like fall out of favor and we have a renaissance of software like it was designed in 2004. But enough developers have decided that the performance tradeoff is worth it, and average users have become desensitized to SPAs and their expectations are low enough that they would rather put up with it all in order to do what they need to instead of demanding better.

I feel like it can only keep growing. That's one thing the collaborative nature of software pushes forward. More features. It's just that some are better at limiting the scope of projects than others.