Irreal

Some good news from Eli. The third pretest for Emacs 28.1, Emacs 28.0.92, is now available for testing. As usual, if you can help with the testing by installing and using the pretest, please do so. The heroes who selflessly labor on our behalf will thank you and you’ll be helping speed the release of Emacs 28.

Arjen Wiersma has been an Emacs user for a long time: twenty two years. He has a lengthy post about that time that describes those years and also his current set up.

It all started for Wiersma when, after drifting among several editors for a while, his mentor told him to pick one and devote sustained effort into mastering it. He chose Emacs and all these years later he’s glad he did. The story of those years is an entertaining read and worth your time.

The last half of his post is about Wiersma’s favorite packages. No real surprises there; his list is pretty much like mine. It includes the big two—Org mode and Magit—of course but he also uses Org-roam, MU4E, Which-key, Eglot, and Projectile. Those are all packages that can have a huge influence of programmer productivity. Wiersma makes a case for each so if you were wondering if you should try one or more of them, take a look at his post for some reasons you should.

Wiersma’s closing thoughts include one that I’m always pushing: Emacs, far from being defunct, is getting better and better. As he says, “Here’s to another 22 years [of] Emacs.”

Steve Yegge, who is always interesting, has a nice video that he describes as a guided tour of Emacs. Yegge characterizes it as covering “roughly 12 features of Emacs, out of approximately 18 billion hillion jillion total”.

If there’s an overreaching theme to the video it’s that everything is a buffer and that therefore all the usual editing functions are available to work with it. The other important point Yegge makes is how easy it is to customize Emacs. Not only for trivial things like binding key sequences but for writing functions to automate your work flow or even writing an entire mode if needed. As he says, this would be an essentially impossible task with something like IntelliJ but is trivial with Elisp.

The video covers things like calc, shells. keyboard macros, and Dired. He also shows opening a gzip file, selecting and opening a file in the archive, editing that file, and saving it back to the archive. That’s always seemed like magic to me but turns out to be useful more often than you’d think.

The thing that stands out about the video is how enthusiastic Yegge is about Emacs. As he says, he’s still using Elisp he wrote 20 or 25 years ago. That shows, he says, that Emacs is a worthwhile investment that will pay dividends over a lifetime.

The video is just over 45 minutes so plan accordingly. It’s definitely worth watching.

Marcin Borkowski (mbork) has a short post that brings news of an upcoming feature in Emacs: transient input method. It’s much like toggle-input-method used to switch to another input method but it’s only in effect for the next character. That can be handy if you need to enter, say, a LaTeX character but don’t want to stay in TeX input method all the time. Of course, you can always just toggle the TeX input method off again but that’s a bit of a pain.

I use the TeX input all the time. It’s usually because I want to use --- to enter an em-dash. That works without TeX input method but the Unicode character it uses messes up the RSS feed. The em-dash character used by TeX input method works correctly so I have to switch input methods to use it. The trouble with the TeX input method for me is that my key chords (as defined with the key-chord package) don’t work so I have to either toggle the input method again or use some other way invoking the function bound to the key chord.

Transient input method isn’t available in Emacs 27.2 so I haven’t had a chance to use it yet but it seems like it could really be handy much of the time. As Borkowski points out, transient input method is an example of polishing Emacs to make editing a bit more pleasant and once again puts the lie to the claim that Emacs is old, defunct technology.

Today, I saw this tweet from Tory Anderson:

I just experienced #emacs rectangle-mark-mode for the first time, though I’ve been using rectangles for a long time. This allows me to make a rectangle easily even when neither first nor the last line are the longest. Nice! pic.twitter.com/yGVjxWNiWz

— (webdev “Tory”) (@Endless_WebDev) March 7, 2022

It reminded me that I used to use it all the time and thought I even had a shortcut for it but when I called describe-function it said there was no binding associated with it. I searched my init.el and found that it was actually defined as part of a hydra dealing with rectangles.

I was happy to rediscover this and am writing about it in case some of you may have forgotten about it too. In a sense, it doesn’t really bring anything new but it does make it much easier to visualize the rectangle and how the text will be affected when you issue a command on it.

I stole the hydra from this post by abo-abo. The thing with rectangles is that most of us don’t use them enough to burn the keystrokes into our muscle memory, so the hydra is really useful. Of course, you still have to remember that you have the hydra and what its keybinding is. Something not all of us, apparently, can do.

Tony Aldon wondered how hi-line-mode worked so he looked into it. He was so enchanted with what he found that he decided to share it on reddit. That might not sound too interesting but hi-line-mode depends on two Emacs features, post-command-hook and overlays, that are interesting in their own right. Once you understand them, understanding hi-line-mode is trivial.

The post-command-hook variable, as its name suggests, contains a list of functions to run after every command. Since virtually everything Emacs does—including inserting characters into a buffer—is through a command, the hook gets run every time Emacs does something. In particular, hi-line-mode can be called after every command to update the current line’s highlighting.

The other important aspect of hi-line-mode is the overlay. This is pretty simple and allows you to, among other things, change the text face of a region in the buffer. In the case of hi-line-mode, it changes the background color of the current line to whatever color you’ve chosen.

Once you grok hooks and overlays, it’s easy to understand how hi-line-mode works. Aldon’s post is useful not because it explains how hi-line-mode works but because it explains post-command-hook and overlays. Those are things you can use in your own Elisp code.

Irreal regulars know that I’m a huge fan of Christopher Wellons’ Elfeed. It’s an Emacs-based RSS reader that allows you to manage and read your RSS feeds from within Emacs. If you’re already an Emacs user, there’s no reason not to be using it.

Ramces Red has a post that explains how and why to use elfeed. Ramces covers the installation and configuration of Elfeed. Other than specifying which feeds you want to follow, you can get away with no configuration at all but Elfeed allows extensive customization. There’s also a thorough and flexible tagging mechanism that you can use to sort posts according to interest.

To me, the most useful thing Elfeed brings to RSS is the ability to store the feeds forever. They all go into a database that is searchable by date and content. Even years later, you can search and find an entry and as long as the original post is still online, you can revisit it.

The one thing that his post doesn’t mention is that you can maintain your list of feeds in Org-mode. To do that you need to load elfeed-org. After that, it’s easy to add or delete entries from your feed even if you aren’t an Elisp expert.

I can’t recommend Elfeed enough. Take a look at Ramces’ post to see why.