Today I want to talk about how amazing emacs is. Not because it is the most feature complete operating system out there or because it fully emulates vi/vim. No, because emacs has a very nice feature called emacs server.
What's an emacs server?
If emacs is not an operating system, at least emacs stands for "eight megabytes and constantly swapping", doesn't it? (This is actually from times where 8 megabytes were quite a lot of memory)
So why do people make fun of emacs and how is it related to the emacs server? Emacs, like any other operating system, loads a lot of things at startup. In my case this is initialising org-mode, starting my mail client, calculating my agenda view to tell me what to do today.
An emacs server creates a special emacs process that listens on a socket for connecting to it. This way the initialisation is already done before you connect to it and all configurations are already loaded. This is the actual "slow" part of emacs. And is a bit similar to starting python, which also needs to load its libraries at start.
With the emacs server running, you can connect to it using the emacsclient program.
As a matter of fact, rxvt-unicode also knows about a server mode (checkout the manpage, look for urxvtd -q -f -o). For rxvt-unicode, you'd use urxvtc to connect to it. So quite simlar.
What is so cool about the emacs server?
Saving a lot of response time and making working with emacs feel much faster is the obvious advantage. However, there is a much bigger one:
With the emacs server, you can connect to it from the terminal and X Window. Because the emacs server also manages the buffers ("open files" for non-emacs users), you can view the same open file from the terminal or an x window.
Turning the notebook into a server
As you might know, we at ungleich are pretty much into IPv6. So all of our devices are generally speaking world-wide reachable. Our work notebooks are no exception from that. In fact, most notebooks even have their own /48 IPv6 network assigned via VPN.
So if I am away from my notebook, but need to check my open (and potentially unsaved) notes or view my emails, I can use any other computer, ssh to my notebook and type emacslient -nw in the terminal.
While my regular emacs is running as an X11 window, I can select, display and work in all buffers that I have previously opened in the emacs server. In the terminal, on a remote computer.
How to configure your system to use the emacs server
In my case I start the emacs server when I start X11 in my .xinitrc:
eval $(ssh-agent) ... urxvtd -q -f -o emacs --daemon ...
Instead of running emacs --daemon, you can also use M-x server-start in a running emacs process.
And because I always want to have my mail client open, just after I start i3, I have i3 launch the following command:
ssh-add </dev/null && emacsclient -c -e '(mu4e)'
The connection to my mailserver is tunneled via SSH to prevent security issues from using SSL/TLS. Thus I need to add all my ssh keys to the ssh-agent, before starting my mail client.
To actually open a new emacs windown (aka "frame" in emacs speech), I use the following configuration in my ~/.i3/config:
bindsym $mod+Tab exec emacsclient -c
How does this look like?
Below you find a screenshot of my writing this article. The upper window is the X11 window, the lower window is a terminal window (they happen to be configured to have the same nice background colour).
emacsclient -a "" -c
is actually smarter than just using
because it will start the daemon if it is not already running. No day that you don't learn something!