Irreal

Andrew Hyatt has a new video that describes the changes to EKG version 0.2. I wrote about EKG back in February. The TL;DR is that it’s a note taking app that links notes similar to, say, org-roam but with some interesting differences in operating principals. EKG is available on Melpa if you want to try it.

Version 0.2 adds a couple more functions, info documentation, and some general bolt tightening. The two most significant additions are

Templating

Templates are a way of adding text to any note sharing the tag of the templates. The implementation is different from other templating systems you may be used to. Take a look at the video for the details.

Embddings

This is the most interesting change. Embeddings are a way of capturing similarities among notes. It’s more general than notes sharing terms. Hyatt gives the example of noodles and pasta. The embedding functionality will consider these similar even though noodles and pasta are not the same thing. Generating embeddings is a complex operation beyond the capabilities to Emacs so the chore is outsourced to OpenAI. That fact has some possibly negative consequences that Hyatt discusses in the video.

The video is 17 minutes 27 seconds long so it should be easy to fit in but it won’t make any sense if you haven’t watched the first video that describes EKG and explains what it does.

The minions, ever vigilant for opportunities to engage in another skirmish in the light vs. dark theme war, brought this to my attention and insisted that I run with it.

Maybe it’s time to make another #emacs color scheme. I need to do my part to fight the 20 year olds creating all this low contrast garbage. Code comments should not be low contrast. Nothing should be low contrast for that matter.

— Ryan (@Pixel_Outlaw) February 25, 2023

It’s not really a light/dark theme issue but it is related. It’s possible, of course, to have low contrast text with a light theme but you almost always see it with dark themes. A horrible example is dark gray on black or—even worse—dark blue on black. It’s essentially unreadable but some folks think it looks cool. Those of us with aging eyes are not as enthusiastic.

The minions are jumping up and down with glee about that tweet but maybe we should stop feeding the trolls: quit rendering low contrast text already.

Keith Waclena has a great Introduction to GNU Emacs. It’s an ongoing project that started back with GNU Emacs 19.29 in 1997 and has been in continual development since. This latest version is for Emacs 28.2. Waclena works for the University of Chicago as an applications systems analyst but his tutorial reads like (in a good sense) it was written by a professor trying to help his students learn Emacs.

His view of Emacs is much like mine: it’s not really an editor but rather “the last remaining Lisp Machine”. This means, among other things, it can be usefully though of as an operating system in that there are applications to perform almost any task that you need to perform.

It’s pretty clear that Waclena intended this to be a book or at least book-like. In several places he refers to it as a book and it is pretty comprehensive. He even covers such esoteric matters as dealing with the Meta key, how many spaces there should be after a period, Emacs versus the Unix philosophy, and usual myths about Emacs being a resource hog and slow.

This is a large document that covers at lot of material. It’s the best introduction to Emacs for beginners that I’ve seen. The document is also available in PDF or EPUB formats. If you’re looking for a comprehensive introduction to Emacs, this is a good place to start. Some of the topics are still TBD—as befits a work in progress—so it will doubtless continue to improve. It’s worth a look even if you’re an experienced Emacser.

When people talk about the killer apps of Emacs, they always mention two: Org Mode and Magit. They’re both uniquely Emacs apps—even though Org is seeping into other environments—and an essential part of the Emacs experience. For the working developer, Magit is a must-have. These days, Git is the default VCS and the absolute best interface—porcelain in Git-speak—is Magit.

Magit is maintained, and constantly improved, by Jonas Bernoulli (tarsius) who, unlike many open source developers, has no other job other than supporting his open source projects and who is completely dependent on community support for his livelihood. Due to a change in policy from GitHub, he is losing many of his supporters and is desperately in need of our assistance.

There are several ways of supporting him. He prefers, of course, a continuing pledge that you can arrange here but if you feel more comfortable with a one-time contribution—especially if you want to use PayPal—take a look at this page for some of the options.

It’s a scandal that large organizations that depend on Magit don’t step up to provide Bernoulli with a living wage but the majority of them don’t so it’s up to us. Most of us can’t afford significant recurring donations but we can make a one-time contribution even if it’s minimal. If you’re able, try to do that.

FreeAd7233 has an interesting reddit post on Emacs startup time. He’s recently moved from Doom Emacs to his own configuration and is amazed at how quickly Emacs loads. It’s less than a third of a second on his local (Apple M1) machine and less than three quarters of a second on a VPS. That’s fast enough for all but the most anal but as I’ve said before, for me, at least, it doesn’t matter.

I was a long time Vim user and followed the usual procedure of invoking the editor whenever I wanted to work on a file. When I moved to Emacs many years ago, one of the hardest things for me was adapting to a new way of working: rather than start Emacs when I needed it, I simply left it running and switched focus to it when I needed it. Later, as Emacs became more central to my workflow, this wasn’t even an option. Of course I left it running because I was almost always in it.

So the point of this post—if there is one—is why should I care how long Emacs takes to start? I do it only once when I start my system or possibly when I upgrade my packages. You can even make this part of the boot process if you like but I’ve never felt the need.

Still, not everyone has my workflow. FreeAd7233, for instance, treats the idea of leaving Emacs running as a sort of strange idea practiced only by aliens. Even those folks who want to start Emacs on demand have a way of accomplishing that. Of course they do. It’s Emacs. You can simply start Emacs in demon mode and invoke emacsclient to popup a “new” instance when needed. Of course, it’s not really “new” because it still has your previous state remembered so it’s even better than a fresh start.

The takeaway is that there’s really no reason to obsess about Emacs startup time but, of course, folks still will. It’s just that it’s not a reason to avoid Emacs.

Vernon Grant over at Discovering Emacs has an interesting video on using the mark ring. Most Irreal readers already know about the mark ring but its exact operation is a little mysterious. That’s especially true about the global mark ring. When, exactly, are locations put on the mark rings?

Grant discusses these and other issues involving the mark ring. He gives his understanding of when entries are added to the global mark ring—a definitely mysterious set of rules—and some of the most usual events that cause them to be added to the mark ring.

Most of us know that you can visit the locations on the mark ring with Ctrl+u Ctrl+Space and the global mark ring with Ctrl+x Ctrl+Space. What I didn’t know is that by setting set-mark-command-repeat-pop to non nil you can type the first Ctrl+u Ctrl+Space or Ctrl+x Ctrl+Space for the first entry and then simply hold down the Ctrl key and press Space to cycles through the other entries.

The video is only 7 minutes, 14 seconds long so it’s easy to find time for it. Unless you’re already a mark ring expert, it’s definitely worth your time.