Irreal

At first glance, this post might appear to be of interest only to Mac users but most of the ideas also apply to other systems. As I’ve written many times, Emacs is at the center of my computing workflow. Almost everything I do, I do in Emacs. The only major exception is browsing for which I use Safari. Of course, other things are always going on—my continual Backblaze backup, for example—but I don’t normally interact with those.

Because I run Emacs and Safari in full screen on their own pages, it’s important that I have a quick and easy way of switching between them. I also want to bring up an Org capture buffer from anywhere in the system. I do all these things by assigning systemwide function keys to the desired action. Thus F6 switches to Emacs and F7 to Safari.

I used to do all that through Quicksilver but it’s really more heavy duty than I need since all I’m interested in is mapping shortcuts to actions. I’ve long considered getting one of the key mappers available for the Mac but as I mentioned in my Emacsclient and macOS post, I recently learned about iCanHazShortcut, which does just what I need. I specify a key combination and the desired operation and iCanHazShortcut takes care of the rest.

One of the tricky things on a Mac is telling the system to switch to an app—Emacs, say—rather than starting a new instance. It turns out that the proper command line invocation is open -a emacs. Now I have a simple and lightweight way of switching between Emacs and Safari.

That leaves only invoking a capture buffer from outside Emacs. This is the thing I’ve had the most trouble with historically. I used to use a bit of applescript to do it but it didn’t always work and other methods I tried worked even less well. Then I found Alphapappa’s yequake. It does just what I need and I can easily invoke it with iCanHazShortcut: I have it mapped to F9.

Finally, I had to deal with the touch strip on my MacBook. It likes to present different “buttons” depending on what app has focus but you can set it to always display the function keys and bring up the other buttons with the fn key.

Naturally, most of this will work on any system. You’ll have to find a key mapper appropriate to your OS but other than that, the ideas in this post should travel well.

Alex Schroeder is an Emacs user who, like me, does as much as he can in Emacs. Also like me, he spends most of his tube time in Emacs or a browser. Unlike me, he uses EWW for a good part of his browsing. I’ve tried to do that too but could never make it stick. Something always seems to go wrong.

Schroeder has an interesting post in which he documents some of his configuration. Especially interesting is his Dired configurations. He use Dired as his file manager so it gets a lot of use and he has some tweaks that you might not know about. If you’re an EMMS user, you might be interested in how he uses Dired as his play list.

If you’re looking for ways to subsume more of your tasks under Emacs, Schroeder’s post may have some ideas you can use. Take a look; if nothing else you’ll enjoy his rant about open plan offices.

I just saw this tweet from Vivek Haldar

Little-known fact: Common #Emacs Shortcuts (C-{a, e, k, y, n, p, b, f}) work *universally* in Mac OS X. In any text box in any app.

— Vivek Haldar (@vivekhaldar) July 18, 2020

and thought that it’s probably time for Irreal’s semiperiodic reminder that the reality is actually much better than a few default Emacs keybindings.

As I wrote back in 2011, you can customize the keybindings for many of macOS’s editing commands. It’s really easy to setup. You just download a file, rename it, stash it in your ~/Library/KeyBindings folder, and enjoy Emacs editing keybindings across your entire system. Even if we can’t stay in Emacs all the time, at least we can take our muscle memory with us when we’re forced to leave. All the details are in the 2011 post, so take a look at that if you’re interested.

By the way, I checked the links and they’re all still good even after the 14 years since Rus originally wrote about this.

As I’ve said many times, Emacs is the first thing I start after a boot and it stays running until the next boot or I update packages that need a restart. That means that Emacs startup time is a non-issue for me. If you’re one of those people who like to start Emacs when you need it and shut it down afterwards, Emacs can be painfully slow to get running. Fortunately, Emacs has you covered. You can start Emacs in server mode and use emacsclient to bring up a frame whenever you need one.

Sadly, it’s hard to start a GUI Emacs frame without going through the terminal first. Dante Catalfamo has a handy post that discusses his method of starting GUI frames with emacsclient. He uses iCanHazShortcut to map hotkeys to actions that would otherwise require invocation on the command line. I’ve always used Alfred or Quicksilver for things like that but iCanHazShortcut seems like a nice lightweight solution if you don’t want or need the power of Alfred or Quicksilver. Really, macOS should have the hotkey capabilities built in but the above utilities are free and more than capable.

In any event, if you’re an Emacs user on macOS, you should take a look at Catalfamo’s post. It’s a short and easy read.

Websockets are a way of providing two way communication between a Web server and a browser. Their benefit is that the server can send data asynchronously to the browser. That differs from the normal protocol of the browser sending a request and the server responding. With websockets, the server can send data whenever it needs to. A good example of their use is in the Web-based chat applications that many companies use for customer support.

Naturally, Emacs has a way of interacting with Websockets. You can write code to act either as a client or server. Adrien Brochard has a very nice post that illustrates the use of emacs-websocket package for using websockets.

Brochard’s post is sort of a minitutorial. He shows how to use the basic interface to send a message to the server and to write a server side function that refreshes a page on the browser whenever it’s code on the server is updated. The interface is simple and easy to use. Brochard’s examples include some code written in Javascript that talks to Elisp code using websockets.

If you have any interest at all in websockets or working with them, be sure to take a look at Brochard’s post. It’s an excellent introduction and well worth the time to read.

UPDATE [2020-07-15 Wed 13:53]: Fixed link to Brochard’s post.