Computer systems research is even more dead and buried than it was when this was written in 2000. Most of the once innovative work that makes up the undead, crumbling foundation that all modern software is based upon was made possible by state money and the massive hoards of gold big tech companies had in the late 20th century. This capital that allowed the development of things such as Unix, the Internet, HTTP, e-mail, even things we take for granted like libc can now be found flowing largely into consumer product development and speculative technologies (like NFTs! :so-true:). Companies are now content building whatever slop they want to sell to people upon the rotting foundations of all this technology. Not only are they content but it directly benefits them; you're going to need their engineers/support staff to figure out and maintain whatever bespoke thing they've cobbled together. And good luck ever migrating away.
We've been stuck with Unix/Linux, Windows NT, PCs, and the Web for decades now to the detriment of humanity's technological development. Systems incrementally developing into such massive and overcomplicated morasses as we are stuck with today has serious reliability, security, and cost implications. A modern Microsoft "Patch Tuesday" has users downloading hundreds of patches at a time for serious security and reliability issues. Alternatively, sit at the console of any modern Linux system, type 'ps ax' and gaze into the intricate internal workings of what once was a simple and elegant design. I'm not blaming the Linux people here, its impressive what they've managed to achieve with the resources they've had.
During the construction of socialism worldwide I think we will again see massive shifts in the way we think about and do computing. Part of the reason I do work on experimental projects like Plan 9/9front is the hope that it will eventually be useful to the creation of a better society. Thoughts?
This was written while on a lot of substances so sorry if this is nonsense x3
:soviet-heart:
Maybe not that surprising, but Pike went on to join Google and develop Go.
I don't deny that Plan9, uxn, or any of the other projects mentioned here in this thread are interesting or useful, but is the issue with the modern operating system complexity, per se? It seems to me that that complexity is sort of necessary to meet the (current, materially contingent) needs of the majority of users. In the era of "elegent" OS design, the average individual user had a degree in computer science or mathematics, and the average computing enterprise could afford to develop bespoke software in-house.
final line of the paper:
The community must separate research from market capitalization.
wow I wonder why that didn't happen
There may be ways to improve the situation, but they will require a community-wide effort.
ooh boy, got some bad news for you Rob
When I type in 'ps ax' in my console the output is fine. Of course, that could be because it's usually a server without the bloated desktop GUI running on top of it that that the nerds on this site seem to think is what makes or breaks a good Linux system. Operating systems have become insanely more complex trying to make the operation of them by us dumb users that much more intuitive.
