Inferno Operating System • Hacker News Board

Re: Inferno Operating System

#41

Post by _wmd » Sun, Jul 21, 2019, 1:07 AM UTC

nickpsecurity wrote: post #40

_wmd wrote: post #36

Earlier quoted context omitted.

Transparently gluing boxes together over a low bandwidth fabric died as an active research area right around the time Plan9 was seeing its first development. By the late 80s shared bus SMP had demonstrated its practicality and quickly became the predominant architecture. Today we don't spawn processes on remote CPUs because the whole act of scheduling on multiple CPUs is entirely transparent to us, that's a competing…

Although no downvote, there continued to be plenty of research and dollars in grid computing that did stuff like that on top of "distributed, shared memory" that did stuff like that. Then, all the research in HPC clusters that tried to create a "Single, System Image" running stuff across machines like it was one machine. The MOSIX quote doesn't change the fact that various researchers kept attempting this and making…

I think the keyword here is transparency -- later designs (even stuff like MPI) explicitly expose the topology of the available hardware. SMP is the closest thing we've ever got to true transparency, and then only for 80% of cases, and for those only because the compute nodes have very similar locality and e.g. memory bandwidth

Even SMP requires careful control if you want to get anything close to the actual performance of the underlying hardware, and the topology is once again very explicit

Re: Inferno Operating System

#42

Post by adam_gyroscope » Sun, Jul 21, 2019, 1:11 AM UTC

Back when I was in university we had a small course on Limbo/Inferno because of a connection my university (RIT) had to bell labs. I took the course in... 1998? and wrote a version of Tetris for limbo. The code was awful, but did find a bug in the Tk implementation. This lead me to my proudest/most embarrassing programming moment - apparently, my code ended up in front of Dennis Ritchie, who thought the code was terrible. Whatever, Dennis Ritchie looked at my code!

Anyways, here's the awful code: https://github.com/adamf/limbotris

Re: Inferno Operating System

#43

Post by chungy » Sun, Jul 21, 2019, 1:14 AM UTC

henesy wrote: post #32

panpanna wrote: post #4

In reality, this was an attempt to turn Plan 9 into a commercial product. Plan 9 despite it's crazy UI (the shell and editor have been ported to Linux so some people must have liked it) brought a lot of new ideas which eventually found their way into other projects such as Linux containers and Go (for obvious reasons). One really good idea that died with Plan 9 was transparent distribution of CPUs across the network.…

Linux containers feel like a very weak imitation of what they could be under an environment like Plan9 imo. Linux lacks a lot of core abstraction properties that would make containers elegant to implement under something like the Plan9 model, at least. Cool project inspired partially by Linux containers: https://doc.9gridchan.org/guides/spawngrid

Linux containers feel like weak imitations of jails and zones found in FreeBSD and Solaris/illumos, for that matter.

Re: Inferno Operating System

#44

Post by henesy » Sun, Jul 21, 2019, 1:18 AM UTC

chungy wrote: post #43

henesy wrote: post #32

Earlier quoted context omitted.

Linux containers feel like a very weak imitation of what they could be under an environment like Plan9 imo. Linux lacks a lot of core abstraction properties that would make containers elegant to implement under something like the Plan9 model, at least. Cool project inspired partially by Linux containers: https://doc.9gridchan.org/guides/spawngrid

Linux containers feel like weak imitations of jails and zones found in FreeBSD and Solaris/illumos, for that matter.

Yeah zones especially, I agree.

Re: Inferno Operating System

#45

Post by spc476 » Sun, Jul 21, 2019, 1:43 AM UTC

panpanna wrote: post #4

In reality, this was an attempt to turn Plan 9 into a commercial product. Plan 9 despite it's crazy UI (the shell and editor have been ported to Linux so some people must have liked it) brought a lot of new ideas which eventually found their way into other projects such as Linux containers and Go (for obvious reasons). One really good idea that died with Plan 9 was transparent distribution of CPUs across the network.…

QNX can do that as well. You can, from the command line, run a program on computer A referencing a file from computer B and pipe the output to a program on computer C and send that output to a file on computer D.

Re: Inferno Operating System

#46

Post by MisterTea » Sun, Jul 21, 2019, 1:56 AM UTC

dang wrote: post #6

2015: https://news.ycombinator.com/item?id=9876217 2014: https://news.ycombinator.com/item?id=8801372 2011: https://news.ycombinator.com/item?id=3007069

And we won't stop posting until Glenda and her labs bread hoard have crushed all others!

Re: Inferno Operating System

#47

Post by dang » Sun, Jul 21, 2019, 2:34 AM UTC

MisterTea wrote: post #46

dang wrote: post #6

2015: https://news.ycombinator.com/item?id=9876217 2014: https://news.ycombinator.com/item?id=8801372 2011: https://news.ycombinator.com/item?id=3007069

And we won't stop posting until Glenda and her labs bread hoard have crushed all others!

There's nothing wrong with posting an interesting story after a year or two. Especially when the topic is as interesting as this one! The current thread is the best so far.

Re: Inferno Operating System

#48

Post by nine_k » Sun, Jul 21, 2019, 2:37 AM UTC

m_sahaf wrote: post #11

There are times when I wish Plan9 and/or Inferno took off and won some respectable market share. Having a laptop, smartphone, tablet, Synology storage at home, servers on DO, and remnant external HDD from prior era becomes frustrating because my files end up being everywhere and nowhere. There were times when I want to be able to start working on some code on my laptop, pause, then continue work on my tablet later on…

You can have centralized / common auth using LDAP, even across platform.

You can have file sync with syncthing, or similar, again across platforms.

Re: Inferno Operating System

#49

Post by soraminazuki » Sun, Jul 21, 2019, 3:28 AM UTC

sprash wrote: post #38

This OS doesn't require an MMU or protected memory. I wonder if a processor architecture that lacks those features could yield much higher performance due to less complexity, reduced memory latency or less cache miss penalty. Will it be enough to mitigate the perormance penalty of runing code in the Dis VM?

You'd have to compensate for the lack of memory protection by running all unprivileged code in a VM, which would very likely introduce its own class of complexities and performance penalties.

Re: Inferno Operating System

#50

Post by ori_b » Sun, Jul 21, 2019, 3:52 AM UTC

_wmd wrote: post #36

panpanna wrote: post #4

In reality, this was an attempt to turn Plan 9 into a commercial product. Plan 9 despite it's crazy UI (the shell and editor have been ported to Linux so some people must have liked it) brought a lot of new ideas which eventually found their way into other projects such as Linux containers and Go (for obvious reasons). One really good idea that died with Plan 9 was transparent distribution of CPUs across the network.…

Transparently gluing boxes together over a low bandwidth fabric died as an active research area right around the time Plan9 was seeing its first development. By the late 80s shared bus SMP had demonstrated its practicality and quickly became the predominant architecture. Today we don't spawn processes on remote CPUs because the whole act of scheduling on multiple CPUs is entirely transparent to us, that's a competing…

Plan 9 is not a single system image OS. And we spawn processes on remote CPUs more than any point in history, often using tools like Kubernetes.