I think that's an important point. People see plan9 as another "fun" "side os" for OS lovers.
Plan9 is actually a good OS. It certainly offer features that are much ahead of OSX, GNU/Linux, Windows, you name it.
The only reason why we aren't using Plan9 _right now_ is because nobody found a way to make money with it. Sad, but true.
Microsoft had similar ideas and pushed them with the Singularity project. It stopped for the same reasons. People are ok buying and developing for "more archaic" systems, that will probably (very) slowly evolve into something similar to what Plan9/Singularity have been offering for quite some time now.
BTW, someone was running OS/360 under Hercules on a Pi. I once tried to get it running on Symbian, without success . It'd be cool to have a maintain in my pocket
As I understand it, Unix was a research OS that was accidentally sold. Then they made Plan 9 and that stayed a research OS, because Unix was good enough in the market.
It wasn't accidentally sold. It was developed as a side project. I believe Bell Labs was required by a court ruling to license it for free, which aided in the initial growth in popularity.
If I had to guess why Plan 9 stayed a research OS- well, Plan 9 was first released to universities in 1992. Linux was first released in 1991, 386BSD was launched in 1992, FreeBSD and NetBSD forked in 1993.
If you're not on the 9fans mailing list there's a pretty good chance you would have missed this. Hats off to Chris for catching this and taking the time to try it out. I had a lot of fun toying around with acme earlier today as a result.
The link to the image was posted here nine days ago[1], it wasn't missed. These sorts of 'Execute a few basic commands to do something on the RPi' posts are becoming a bit ridiculous and fluffy.
I probably should avoid responding to this, but here goes: If you don't like it, submit something you do like and let people vote it up or down. Don't like the community? Go elsewhere. Friends of mine are always talking about how the quality of comments has gown downhill on Hacker News, and your comment is proof. Please stop turning everything fun into a referendum on karma.
The original posting was a bare announcement in comp.os.plan9 for people who were already familiar with Plan 9. The newer article adds references to some introductory Plan 9 documentation which would be very helpful to someone new to the OS.
An advantage of going native rather than VM, especially with an OS as small and simple as Plan 9, is that you can have complete control and understanding of what's running on your machine - no mystery processes you didn't start, no little webservers opening ports behind your back...
It's usually good enough, however I've found a number of lower quality USB enclosures that handle things better when setting it as high as 32MB so that as much data is in buffer as possible for them at all times. any higher and I've not seen any speed difference but below 16MB and they'll run at nearly half speed for some reason.
Don't think so, but it seems Wirth is revising the Project Oberon book. This time he'll include the design of his own RISC CPU, targeting a $100 FPGA board.
https://www.youtube.com/watch?v=BJIqHIYSDrk
I get a good laugh out of the "Oooh, my RPi is a web server! " type posts. Of course it can be a web server - it's more powerful with more storage and RAM than some of the servers we ran a 30000-user ISP with back in the mid-90s.
The "amazing" bit is that they're $35 + junk-box parts to have a working desktop or server config.
Plan 9 is significantly more recent than Unix: Back when Unix was first invented, graphics existed but there were more kinds of graphics devices, and it wasn't entirely clear that vector graphics, for example, wouldn't become the way forward and deserve the majority of developer attention.
38 comments
[ 3.1 ms ] story [ 85.8 ms ] threadEspecially its consistency is something worth having.
Plan9 is actually a good OS. It certainly offer features that are much ahead of OSX, GNU/Linux, Windows, you name it.
The only reason why we aren't using Plan9 _right now_ is because nobody found a way to make money with it. Sad, but true.
Microsoft had similar ideas and pushed them with the Singularity project. It stopped for the same reasons. People are ok buying and developing for "more archaic" systems, that will probably (very) slowly evolve into something similar to what Plan9/Singularity have been offering for quite some time now.
edit: http://web.mit.edu/multics-history/source/
BTW, someone was running OS/360 under Hercules on a Pi. I once tried to get it running on Symbian, without success . It'd be cool to have a maintain in my pocket
If I had to guess why Plan 9 stayed a research OS- well, Plan 9 was first released to universities in 1992. Linux was first released in 1991, 386BSD was launched in 1992, FreeBSD and NetBSD forked in 1993.
And it wasn't free.
Why? This isn't even "how I got X to run on RasPi" - it's "I ran the image that someone else created". Period. That's it. Um, wow.
[1]: http://news.ycombinator.com/item?id=4812749
1. bs was already included in the post (FWIW the shorthand is '1M' in Linux, and that's 1048576 :-))
2. I was referring to the use of disk vs rdisk (i.e buffered vs unbuffered/raw, see details[0]), which IIRC is a BSD-ism.
Using rdisk gives a tenfold increase in speed here. Without it you're nowhere near USB2 speeds.
[0]: http://lists.apple.com/archives/filesystem-dev/2012/Feb/msg0...
Genode would be useful, since it can run on the Pandaboard.
http://news.ycombinator.com/item?id=593323
Step 2) Run the Inferno-OS VM on your primary machine
Step 3) Let your imagination run wild.
The "amazing" bit is that they're $35 + junk-box parts to have a working desktop or server config.