I am running a VM on a Linux host using qemu+kvm and libvirt, the guest system is CentOS 7. I wish to share a folder between the two such that host:/some/dir becomes accessible as guest:/some/other/dir. I wish to minimise effort and cannot re-compile the kernel on either side. I experimented with plan9 but cannot get it to run (the p9 dkms module won't compile on the CentOS 7 guest). There must be a better way? Performance is secondary as it does not work at all right now.

I am running the editor sam as a standalone application on Linux. I am using this version: git clone https://github.com/9fans/plan9port plan9 So far I have not been able to figure out a way to paste text into the *bitmap* display from an external Linux application. Is there a way to achieve this?

I downloaded a Plan 9 (4th Edition) ISO from Bell Labs and gave the system twice the recommended memory and hard disk space. I followed the installation prompts, choosing the default options where possible. However, the installer warned that it ran out of space, despite completing the installation. When the machine tried to boot, it got multiple kernel panics and aborted. How do I get Plan 9 installed and running on VirtualBox?

I'm trying to run a virtual machine off of a writable 9p root filesystem. When the default caching mode is used, apt and some other tools fail; likely due to limitations with 9pfs itself. The error message produced by apt is:

Apt: Unable to determine file size for fd ... : No such file or directory

A quick search online suggested that I should set caching to loose when mounting the filesystem. Mounting the filesystem with cache=loose on my system did indeed allow me to use apt normally. However, now I am worried about filesystem consistency. [Kernel documentation](https://www.kernel.org/doc/Documentation/filesystems/9p.txt) states that this cache=loose is only meant for read-only systems and sadly doesn't elaborate further. So I present my question as follows: If I only access the filesystem through a single mount point where caching is set to loose, and I do not access the filesystem or its underlying directory otherwise, would I face any issues with consistency/coherency? Or stated in another way: Does loose caching effect the consistency of the filesystem when viewed exclusively from a single mountpoint? I'm using the following mount options: rw,trans=virtio,version=9p2000.L,cache=loose in my kernel command line and fstab.

I try to install Inferno on my Debian 10. $ git clone https://github.com/doublec/inferno $ cd inferno $ sh Mkdirs $ nano mkconfig ROOT=/root/of/the/inferno/git/clone SYSHOST=Linux OBJTYPE=386 $ export PATH=$PATH:pwd/Linux/386/bin $ ./makemk.sh > removing old libraries and binaries gcc -m32 -c -I/home/user1/inferno/Plan9/386/include -I/home/user1/inferno/include -I/home/user1/inferno/utils/include regaux.c regcomp.c regerror.c regexec.c regsub.c rregexec.c rregsub.c After last operation I got it output Everything lib.c with fatal error: >...**fatal error: /386/include/u.h**...

I was wondering, since it seems to be one of the only websites where I can *seemingly* legally download historic Unix documents.

When using mk , GNU sed, I find that variables do not expand at all. Problematic code: VAR=qux ... build: sed -i "s|FOO = .*|FOO = $VAR|" bar.file This seems to run the following: sed -i "s|FOO = .*|FOO = $VAR|" bar.file when I want it to run sed -i "s|FOO = .*|FOO = qux|" bar.file I have tried using single quotes and double quotes, as other stack exchange posts have mentioned. The only instance I can get $VAR to expand is when there are no quotes around it. Which of course means that sed no longer sees the expression as an expression.

When you watch presentations about Plan 9 and its acme editor you might notice that the name for _copy_ is _snarf_ (I wasn't able to find any meaningful explanation unfortunately). Why is it so? Is there a reason for that? Was it done to create a unique name for this functionality as _copy_ might not be the best name? --- ###References: - Wikipedia (link) : > Snarf, a term used for the "copy" operation in the Blit and Plan 9 windowing systems. - cat-v.org (link) : > There's another menu available with a mid-click; Cut and Paste are pretty self-explanatory, and Snarf is just the Plan 9 word for Copy. - 9fans mailing list (link) : > > Using either/both acme under Plan 9 and/or Mac OS X via plan9port, is > > there a "file" that can be opened in acme that displays the current > > contents of the acme snarf buffer? > > /dev/snarf So it looks like that there is something like a _snarf buffer_ and a file (obviously there must be a file for that buffer) called /dev/snarf. - 9term manual page (link) : > Characters typed on the keyboard replace the selected text; if this text is not empty, it is placed in a snarf buffer common to all windows but distinct from that of sam(1). > > (...) > > Editing operations are selected from a menu on button 2. The cut operation deletes the selected text from the screen and puts it in the snarf buffer; snarf copies the selected text to the buffer without deleting it; paste replaces the selected text with the contents of the buffer; and send copies the snarf buffer to just after the output point, adding a final newline if missing. Paste will sometimes and send will always place text after the output point; the text so placed will behave exactly as described above. Therefore when pasting text containing newlines after the output point, it may be prudent to turn on hold mode first. - _Exploring Acme_ tutorial (link) : > 'Snarf': called copy in most other editors - dictionary.com definitions of _snarf_ (link) : - > (informal) to eat or drink greedily - > (contemporary) to grab something in greed, esp. without permission - > (Unix) To fetch a file or set of files across a network. See also blast. - Inferno Part 2: Let's Make a Cluster! (link) : > The clipboard is called by the unfortunate name “snarf buffer” in Inferno and Plan 9.

I have QEMU virtual machine. Linux host and linux guest. I want stable and fast solution with minimal performance cost. I'm considering two technologies for shared folders (between guest and host): NFS and VirtFS (Plan 9 ). There is small amount information on VirtFS from one side and there are some evidences that NFS isn't good option for virtualization from another side. So my questions are following. What is more reasonable for my purpose? What is more effective? What solution is consume less (speaking about CPU)? What solution is more fast? Are there some benchmarks on subject?

I have folder on host machine which have symlinks to another device. Like this $ ls -all ~/folder_to_share ... file1 file2 linked_dir -> /mnt/bdrive/real_dir/ ... I setup QEMU virtual machine via libvirt and vagrant-libvirt . And also setup sharing folder which is mapped to host dir ~/folder_to_share using VirtFS (9p protocol ). I want my virtual machine to walk inside shared dir and follow symlinks as it is real directory. Is it possible or I should switch to another protocol? P.S. Part of vagrant-libvirt config : config.vm.synced_folder "~/folder_to_share/", "/home/vagrant/shared_folder", type: '9p', owner: 1000

Plan 9 was developed by Bell Labs as a successor to Unix. Although for various reasons it never quite materialized as such, a fair amount of development still went into Plan 9. My question is, what - if anything - from Plan 9 has made its way into modern Unix?

I would really like to obtain a book on Plan 9's design and ethos. Do such things exist? Anybody have a recommendation? (Asking here on the basis that Plan 9 was once meant to be Unix's successor, AIUI)

I'm trying to use rc shell (Linux port from Plan9 OS) and get 'syntax error' messages when using programs with arguments like **arg=val**. ; dd if=/dev/zero of=/dev/null syntax error ; ls --color=auto syntax error My guess is that rc shell sees **=** sign and treats the statement as variable assignment, not a program run. May be there's no dd program or other programs using this type of arguments in Plan9. But how can I use it in Linux where they are widely used? I tried escaping **=** like dd if\=... of\=, it didn't help. I tried quoting arguments like dd 'if=/dev/zero' 'of=/dev/null' 'bs=1M' 'count=1' and it worked. But it's not very comfortable. Is there any other more handy way? I'm using rc shell from Ubuntu 11.10 repository: package version 1.7.1-3ubuntu1.