Irreal

Like most longer term Emacs users, one of the first things I did was to get rid of the annoying long-form Yes/No questions that required you to answer “yes” or “no” explicitly. That was easy to do. I simply aliased y-or-n-p to yes-or-no-p. The same code is in virtually everybody’s init.el if they started before Emacs 28.

Starting in Emacs 28, the variable use-short-answers was introduced. When non nil it causes yes-or-no-p to call y-or-n-p to ask the question instead of following its normal path.

Lennart Karssen over at Polyomica recently discovered this and wrote a short post about it. He, like almost everyone else hates the long answers. They’re there for a reason, of course, and the documentation for the variable recommends against setting it non nil but I doubt very many folks follow that advice.

I’ve had my original aliasing code for at least 17 years and see no reason to change it. When use-short-answers was first introduced, I thought I should adopt it but I never did and as far as I can see, it doesn’t make any difference.

In the unlikely event that you weren’t aware of how to turn off the requirement for entering long answers, take a look Karssen’s post or just check the documentation for use-short-answers.

I just discovered a useful trick over at Ruslan’s Tech Blog. It’s about finding the file at point. Ruslan does this a lot. I, on the other hand, hardly every do it but I have gotten used to typing Ctrl+c Ctrl+o to follow a link in an Org buffer. Despite using the goto-addr package, this doesn’t work reliably everywhere.

Ruslan was also bothered by the difficulty of opening a file at point—although his normal usage didn’t involve Org. Like many of us, he thought he’d just write a quick command to do what he wanted but before embarking on that, he read the documentation for find-file and discovered that the functionality he wanted was already there.

If the point is on a file path, you can invoke Meta+x find-file-at-point but that’s a lot of typing. It’s a bit less typing if you use its ffap alias but perhaps still too much. Instead, you can just type Ctrl+x Ctrl+f as usual and then Meta+n to follow the link at point. It’s a nice solution because you do what you always do when you want to open a file and then type Meta+n instead of a file path to open the file at point. Best of all, you don’t have to do anything to enable this. It’s the default Emacs behavior.

As longtime Irreal readers know, I am an Avy fanatic and use it as my principal means of buffer navigation. For a long time, I mostly used avy-goto-word-1 but it can only be used to search for the beginning of words. Four years ago, Karthink Chikmagalur published a post that recommended using avy-goto-char-timer exclusively. For some reason, I ignored his post at the time but a couple of years ago I was reacquainted with his recommendation and resolved to use avy-goto-char-timer more.

One day I suddenly realized that I was, in fact, using it exclusively and couldn’t even remember my binding for avy-goto-word-1. It solves the problem of avy-goto-word-1 being able to find only the beginning of words.

As the name suggests, avi-goto-char-timer will accept input chars until there’s no input for the timeout value. Then it will present a tree of potential targets having those characters. The default timeout value is 0.5 seconds so if you stop typing for half a second, avy-goto-char-timer will present you with a tree of targets.

Lately—perhaps because I mostly work with my laptop on my lap or perhaps because of declining reaction times—I’ve found that half a second isn’t quite long enough. It often produces suboptimal outcomes, resulting in spurious characters being added to the text in the worst case.

As often happens, I put up with this infelicity for some time before I finally got fed up enough to do something about it. The “something” in this case was trivial. I simply increased the timeout value to 1.5 seconds. That may be a little long but it seems fine so far. I can, of course, always decrease it a bit if it turns out to hold things up.

If you’ve been using Org-mode for a while, you’re probably familiar with its table editing functionality and may even know that you can use that functionality anywhere in Emacs with orgtbl-mode. For most non-Org applications, of course, those tables will be in some sort of comment.

Andrew over at the Listful Andrew site has a useful post on how to do that. It’s a little more finicky than you might think because the editing won’t work if the table lines are commented out. It’s easy to work around that, of course, but Andrew offers some other strategies.

When you’re in Elisp, the easiest way to add a table to a comment is to add it do a Docstring. That’s easy and the table can be edited at will. Andrew offer another method using his custom description macro. Most of us won’t have that macro, of course, but it’s another strategy.

Finally, Andrew considers adding tables to Bash comments. The problems are similar to those encountered with Elisp: you can’t edit tables whose rows are commented out. As with Elisp, that’s easily worked around but again there are better ways.

One way is to use the null command : to introduce the comment string. See Andrew’s post for the details. Another way is to use a here document or a here string. Again, see Andrew’s post for the details.

Tables in comments isn’t something that you need all the time but it’s nice to have to ability to add them when you do. Andrew’s post is a nice demonstration of how to do that.