Porting Tailscale to Plan 9(tailscale.com) |
Porting Tailscale to Plan 9(tailscale.com) |
A bunch of us are currently in https://meet.google.com/qre-gydb-mkv chatting about this. (Edit: the hour is over; we all left)
The earlier Apr 1st blog post was https://tailscale.com/blog/tailscale-enterprise-plan-9-suppo...
Either way it makes VPN easy between 9 and non 9 machines. Otherwise Plan 9 can do it's own VPN-like over tls or ssh tunnels and bind remote network stacks to a local namespace. But that makes seamless Unix and Windows comms difficult.
I'd also like to point out that most users of Plan 9 dislike web technology because it's a giant nightmare of code. No one human can even begin to comprehend the code base of Chrome, let alone Firefox - programs that are as big, if not bigger than the kernels they run on. That is an absurd state to be in - your runtime requires a billion dollar company to maintain. Even open source Firefox needs millions in funding.
Whereas a single human can grasp plan 9 code from the kernel to user space. That's the runtime I want, something I can understand. The process is the container on plan 9 so you have everything you need to build distributed apps without a web browser. It's human scale distributed computing. I'd like a future without the "modern" corporate scale web.
Given the huge maintenance cost of immature computer architectures such as mips, 386, arm, arm64 and amd64, we decided to put our focus on the more mature and stable achitectures:
power64 and itanuim.
Therefore, all architectures other than power64 and itanium are thereby frozen, conserved and promoted to end of life.
https://copy.sh/v86/?profile=custom&m=768&vram=16&hda.url=ht...
You can start tailscaled and tailscale inside the VM. It may take a while to come online sometimes due to limited proxy availability.
Edit: alt gives you the third button. To start a terminal, hold alt and right click, select new, release alt, and right click drag to size the terminal window.
That OS fascinates me.
I remember early in my career when an expert I worked with could sit with me and patiently show me how to do something and let me ask questions for however long it took me to understand well enough what to do and how to swim if I fell in the deep end of whatever they wanted me to do. It was some of the fastest upskilling that I have ever done in my career, like getting a bachelors degree worth of very specific knowledge in three hours.
I don’t know C and I don’t know enough about Plan 9 to use it productively for anything, but it has some extremely cool and useful features that I want to know more about and learn how to use, even if it is only so that I can lament the non-existence of those features in the big three operating systems today.
If I had the money I would probably pay to get face time with all three of those folks for expanding my Go knowledge and rsc and rob pike for the plan 9 understanding that I have always wanted, but have never been able to give myself.
EDIt: I reserve the name “chaos10” for this project, since - like SerenityOS - there will be no plan.
I kinda expected this paragraph to continue with
> This has made a lot of people very angry and been widely regarded as a bad move.
For now only 9legacy (with all the latest changes) works.
Please do! Just be careful with your sysupdate.
> April 1, 1999
>
> FOR IMMEDIATE RELEASE
Golang on RISC OS would be a truly ludicrous porting effort.
I wonder what your employer/policy doesn't like.
Fixing Go to not special case Plan 9 benefits all platforms--- all operating systems use the same code paths now, making the code simpler.
Most scripts are write-once:read-never, especially if you actually implement -h/--help
> The most important principle in rc’s design is that it’s not a macro processor. Input is never scanned more than once by the lexical and syntactic analysis code
I worked at a unix shop that deleted most of a working drive because a shell script was modified while it was running. Luckily they kept daily backups on tape. This was about 17 years ago.
[1] https://www.scs.stanford.edu/nyu/04fa/sched/readings/rc.pdf
For example, consider the following change:
-echo $x; rm -rf /n/foobar/
+rm -rf /n/foobar/
^^^^^^^^^^^^^^^^
I am unfamiliar with the read-buffering done by either of the 2 main implementations of rc, and so am unable to comment on whether it does things to avoid this problem. But if it does do things to avoid it, those things are orthogonal to the "not a macro processor / input is never scanned more than once" thing.
The biggest thing is the heavy reliance on union file systems (and file systems in general) and an extremely simple syscall API. It's a heterogeneous-networked-node OS so it handles realistic workloads natively with primitives designed for it instead of piling complexity on top of Unix-like APIs (ie. Linux). I dunno, I just think a lot of the modern "cloud native" stack is unnecessary if you had an OS actually built for the workloads we have.
I don't really get the 'joke'? Porting a full web browser to Plan 9 would seem like a cool project - where's the humor?
Their April Fools' jokes are real and work as you can see in the submitted link.
So basically, a Plan 9 web browser, would be a great April Fools' prank! (because, again, their "pranks" are real and work)
Note that one of Tailscale's main party tricks is NAT traversal, when both machines are behind different NATs and can't otherwise get a connection open to each other. (And then Tailscale ultimately falls back to a relay server on the internet if it can't get a direct connection for IP packets)
Though, 9front lets you run your own NAT giving you an Internet facing 9 machine you can serve a TLS tunnel from directly. So the server side is solved making the client side NAT a non issue.
I'm talking about two machines deep in somebody else's network or where you don't control the router/NAT.
I guess the answer is always “it depends”, but that generally has never been my experience with most things. Are you over-engineering the shit out of every script to the degree the script itself is a Turing complete machine and with enough —-help flags anything is possible? Most 40+ year old Unix tools with a thousand flags have their limits and you have to script around them to achieve things you want.
In my experience, eventually a business need will arise that require you to change a script. Are your coworkers comfortable changing these scripts or are you in the mind set of “that’s a simple enough change, I’ll do it”
No. They're 30 line scripts with 0-5 or so flags. They mostly exist to remove choices from other utilities. Put another way: create named (and namespaced) abstractions by making choices and slapping a name on it. They're functions.
> In my experience, eventually a business need will arise that require you to change a script. Are your coworkers comfortable changing these scripts or are you in the mind set of “that’s a simple enough change, I’ll do it”
They're small enough that they can be ignored if they don't do exactly what you want. I'm fine changing them, but it's most likely they'd just get rewritten or bitrot after I'm gone.
Even if you aren't super familiar with rc, it's not that weird to look at. I find it way more readable than (ba)sh syntax.
I would avoid laptops altogether, honestly. Not a great fit.
Is that true? I cloned the 2015 release of plan 9 a week or so ago and it had around a million lines of C. Can a single person hold all of that? I sure as hell can’t.
And which plan 9 release and when? Ghostscript and Python were originally distributed with 9front which are both HUGE compared to the rest of the system. Remove those and its much, much smaller. Unsure if ghostscript was included in vanilla 9 from the labs. Python was included in 9front because it was necessary for mercurial. Once git9 arrived python was nuked from base and removed many lines of code. Ghostscript is next to go from base once pdffs is running (patches welcome.)
Not only is the VNC redirection unnecessary, so it is the entire filesystem. You could just render the vm directly to the window and boot a read only image. Plus then you don't have to deal with VNC.
In regards to using it for a "cloud native" stack, the issue is that people want to run code that isn't designed for Plan 9. You could build whatever backplane type thing you want out of plan 9 but the end goal is still likely to be to run some web app or REST api server. Unless someone does a great deal of effort to port all of those environments that people want (nodejs, modern python, etc) you're going to be stuck using a VM and losing a lot of the benefit.
This feels similar to what Joyent did with lxzones in SmartOS, where the backplane was solaris based but the apps they were running for clients were using Linux. It's hard to make the plan 9 backplane better enough to warrant dealing with integrating the guest and host environment.
It should not be a huge deal of effort since as you mention the plan9 syscall API is simpler than on Linux. The added plan9 support could then also serve as a kind of "toy" backend that could make the rest of the code more understandable in other ways.
I'd even argue that OP's early experiment with such a port of tailscale shows precisely such an outcome.
Yes, this is what I'm referring to. It's really many filesystems unioned into one namespace that is controllable per-process.
> In my opinion part of the reason why the userspace tools can be as nice as they are, are due to the use of file system interfaces and the simplistic syscall API. Could you elaborate more on the issues you see with the use of these?
I didn't say I had any issues, I said I preferred them! Aside from a lack of familiarity and needing to install plan9ports on other systems, I haven't had issues.
> In regards to using it for a "cloud native" stack, the issue is that people want to run code that isn't designed for Plan 9. You could build whatever backplane type thing you want out of plan 9 but the end goal is still likely to be to run some web app or REST api server.
Right, language support is the biggest issue with running on Plan 9 from that perspective, at least for "server" workloads. Excluding graphical APIs, the basic stuff (file IO, networking, etc.) isn't all that hard to add to a language (it of course depends). The real trouble is things that have no equivalent in Plan 9, such as mmap and shm.
> This feels similar to what Joyent did with lxzones in SmartOS, where the backplane was solaris based but the apps they were running for clients were using Linux.
This is also what Oxide is doing. Their rack's OS is IllumOS but their customers are expected to only interface with the OS via their tooling and instead provision VMs.
> It's hard to make the plan 9 backplane better enough to warrant dealing with integrating the guest and host environment
If I were doing it, I would do it the other way! Run Plan 9 in a backplane/hypervisor and target it from the language level. The nice part is the systems programming model!
You could implement a modern graphical API on top of virtio-gpu, which would give you low-level access to accelerated graphics.
> The real trouble is things that have no equivalent in Plan 9, such as mmap and shm.
Some uses of mmap and shm actually seem to have a near-equivalent already in plan9's segattach. Other uses would require some implementation of distributed shared memory (i.e. implementing the usual CPU concurrency model over the network) to be made feasible while keeping to the usual networked-OS focus of plan9.
The general problem stands though, almost no languages support Plan 9