Irreal

It’s really handy to have a thesaurus handy when you’re writing, and of course if you’re writing in Emacs, you want to have that thesaurus available and easily accessible from Emacs. I’ve been using power thesaurus and its Emacs interface, powerthesaurus.el. It’s nice and works well but merely provides a list of synonyms from which you can choose a replacement for the word you’re looking up.

Recently, I saw a pointer to Ag Ibragimov’s mw-thesaurus. It’s an interface to the Merriam-Webster Thesaurus and is also quite nice. It’s strong point is that rather than a just list of synonyms, it provides a copy of the thesaurus page in Org format. That provides a lot of context and helps you zero in on the best substitute. On the downside, there’s no way to automatically substitute your choice in your text. That might seem like an odd omission but given the way it displays its results, it’s not clear how such a substitution would work.

In any event, I have them both installed at the moment. It’s not really much of an inconvenience to replace the word with its synonym manually. It’s certainly no harder than navigating to the proper synonym on the thesaurus page. On the other hand, it is convenient to be able to simply choose the word you want from a list with something like Ivy and have it automatically replace the word you’re looking up.

I’ll use them both until I decide which I like better. Or perhaps I’ll keep both installed and use the one that makes the most sense in a given situation.

Alain M. Lafon has an interesting post on spellchecking in Emacs. Lafon is Swiss and therefore has to deal with four official languages as well as German and English in his day-to-day life. The Swiss with their usual efficiency handle that just fine but their polyglot ways means that a single document may contain text in two or more languages. The question is: what’s a good way of handling spellchecking in Emacs?

Well, it is Emacs so of course you can do that but it’s a bit trickier than you might expect. Lafon’s post gives step-by-step instructions for setting up spelling checking using multiple dictionaries.

The only potential wrinkle is that you have to use Hunspell if you want to have dictionaries for more than one language. If you’re on macOS (or Linux), you can install hunspell with Homebrew so it’s not much of a problem. Lafon gives the magic spell for apt users. If you use Homebrew, you’ll have to download the dictionaries separately but they’re available from the OpenOffice Extensions page.

Most of you—at least if we’re not Swiss—probably won’t have to worry spellchecking in multiple languages in the same buffer, but if you do, Lafon’s post tells you how to handle it.

UPDATE [2020-08-24 Mon 17:10]: Added link to Lafon’s post.

Just a quickie today. Those of you who read yesterday’s post but are running on Linux might find this Reddit post by catsol_v2 on compiling gccemacs on Linux useful. It’s not quite as detailed as Ramírez’s post but has enough information to get gccemacs compiled and running.

Again, this is for people who don’t mind living on the edge or who want to build a separate instance of Emacs to play around with native compilation. The rest of us are going to have to wait for Emacs 28 (at least).

If you’re like me, you’re probably excited about native compilation coming to Emacs. In case you don’t know, the project, gccemacs (started and headed by Andrea Corallo), brings native compilation to Elisp. That means that not just will Elisp code—and hence Emacs—will run faster but also that some of the C code in Emacs can be replaced with Lisp.

Corallo gave a presentation to the European Lisp Symposium, which I’ve written about before, that discusses how things work and gives the results of some benchmarks. The work is ongoing but the system is already usable. As far as I can tell, the plan is to include it in Emacs 28 but if you can’t wait to try it out, the code is in the Emacs git repository.

If you’re on macOS, Álvaro Ramírez has step-by-step instructions for compiling Emacs with the native compilation enabled. You’ll need to load gcc with jit support if you don’t already have it. After that, it’s pretty simple. Ramírez provides a configure script to set everything the way you’ll need it so all you have to do is run the script and call make to build Emacs. See Ramírez’s post for the details.

This project is a great thing for Emacs and even for making Elisp a more general programming language. I’m already inclined to use Elisp when I can for small utilities and one-offs so improving Elisp is a win for me.

The other day, I wrote about Bozhidar Batsov’s TL;DR take on the Emacs 27.1 NEWS file. Batsov’s post hit the highlights but if you want to know everything that’s new, you have to go over the NEWS file itself. Happily, Mickey Petersen has provided his traditional exegesis of the NEWS file. The NEWS file is already quite long so with the added commentary, Petersen’s post is even longer but I find it easier to read than the plain NEWS file.

It’s incredible how much the developers put into this release. A lot of it are things that are mostly of interest to developers rather than end users but even for the average emacser there are a lot of improvements. There are, for instance, some changes that provides better control of the startup process.

Petersen does a good job of explaining what the portable dumper is and why it’s necessary. This has been a long-running problem since the glibc maintainers said they didn’t want to support the capability any longer because Emacs was the only software that used it.

There’s much, much more than I’ve mentioned so you really should read it yourself. A nice thing about Petersen’s post is that it explains why some seemingly uninteresting changes are really a bigger deal for the average user than they appear. It’s a long read but worth going over if you’re a serious Emacs user. It’s divided into sections so you can read it piece-by-piece if you like.

Alain Lafon, who, as I’ve said before, is an Emacs master, has a very interesting post on Emacs and messaging apps such as Skype and Slack and many others. One of the major problems with them from his point of view is that their text entering facilities are terrible. If you’re going to be entering a lot of text—or even a little, actually—the last thing you want is a brain-dead interface with unfamiliar keybindings for editing. What we Emacsers really want is to be able to enter text in Emacs with its powerful and familiar editing commands.

His idea is to link these apps to IRC—and hence to ERC—and deal with them through Emacs. The way to do that is to use BitlBee and libpurple to act as a bridge to IRC. BitlBee is the bridge between IRC and messaging apps and libpurple provides the core of IM apps, dealing with the IM protocols and other low level details. Follow the links if you want more information.

Although you can install BitlBee and libpurple directly, Lafon prefers to run them from a Docker image. His post gives detailed instructions for setting up the Docker image. Once you’ve got things set up with either method, you can conduct your conversations with the IRC client of your choice. If you pick one of the Emacs IRC clients, you’ve moved another function—or perhaps several other functions—inside Emacs.