Irreal

When you open a file, do you like Emacs to remember where you were when you last visited the file and return you to that place? I’m probably a minority but I find that behavior annoying. Of course, I can see why people would like that capability. The nice thing about Emacs is that you get to choose. As with most things in Emacs, you can have it your way.

James Dyer is one of those people who like having Emacs take him right back to where he left off when he opens a file. It’s a small part of his larger workflow for dealing with recent files. What he wanted was a list of recently used files, ordered by most-recently-visited, that would allow him to easily revisit a file and be positioned at the cursor’s last position. He discovered that recentf-mode and save-place-mode did almost what he wanted. The only problem was that recentf-mode opened its list of files in a separate buffer. Dyer wanted a more completion-like interface. It turns out that recentf-open-files does that but, annoyingly, the files get reordered.

Dyer solved this with a tiny bit of Elisp. You can read the details in his post. If you, too, like being able to revisit recently edited files and be taken back to the exact position where you left off, take a look at his post.

To me, the most interesting part of Dyer’s post comes at the end. His changes, he admits, are a small optimization but one that makes his workflow a tiny bit smoother. As he says,

That’s the kind of workflow improvement that makes every day at the keyboard just a little bit more pleasant, and those small pleasures add up to something significant over months and years of coding.

Who can argue?

Eli Zaretskii writes that Emacs 30.2 has been released and is ready for download. It’s a bug fix release but still worth installing. As usual, I built mine from source. I had a bit of difficulty downloading the tar and signature files because the mirrors hadn’t yet updated. By the time you read this, the mirrors should be up to date.

I’m writing this in Emacs-30.2 and haven’t experienced any problems. Once I got everything downloaded, the compile and install experienced no problems.

As usual, I want to express my thanks to Eli and all the other devs who selflessly contribute their time and efforts to keep Emacs the best in class editor. These guys are heroes and if you ever find yourself in the same bar as one of them, drinks should be on you.

I’ve been reading Peter Provo’s’ book Emacs Writing Studio and one of the points he makes early on is that “word processing” is different from what it used to be. The traditional paradigm considers word processing from the point of view of the printed page but that’s no longer a useful way of looking at it. Today the written word may be published as HTML on the Web, as an e-book, as a text file, or even as a traditional printed book. Prevos’ point is that GUI word processors are no loner fit for purpose.

Oddly, Emacs never embraced the traditional paradigm, probably because it started life as a code editor and only later came to be used for writing books and other “published” entities. That raises the question of what an Emacs “page” is and how you work with it.

Many Emacs users are probably unaware that there is such a construct. The closest thing they’re familiar with is the “current screen”, which can be paged forward and backward with Ctrl+v and Meta+v. But how much data is displayed on the that screen depends on how many lines there are in the current window.

A “page” is different. It doesn’t depends of the window size and doesn’t even necessarily have a uniform size even in the same file. A “page” in Emacs is simply the text between two special markers—line feeds, ^L, be default. Hardly anyone uses them these days but they can be useful for organizing a text file into sections for easy navigation.

Charles Choi notes that although there are forward page (Ctrl+x ]) and backward-page (Ctrl+x [) commands, they don’t act in the way you might expect, which is the put the top of the page at the top of the screen.

Naturally, Choi has a solution. It’s a bit of simple Elisp that calls forward/back-page and then recenters it so that the top of page is at the top of screen. If you’re interested in working with pages, take a look at Choi’s post.

I like the posts of Randy Ridenour because he’s (definitely) a non-technical user who nevertheless has managed to learn and use Emacs in an advanced way.

One indication of this is his post on inserting bible passages. As Ridenour says, it’s not a capability that many people need but what impresses me is that he managed—despite not being a programmer—to write some Elisp to help with this. Sure, maybe you could do better—even Ridenour admits this—but, again, he’s not a programmer. I find this inspiring and a testament to the power of Emacs and the ability of any reasonably intelligent person, even if they lack technical expertise, to make effective use of even the advanced features of Emacs.

Last week, I wrote about Ridenour’s contribution to the Emacs Carnival Writing Experience. This week, Ridenour accepts the Emacs Carnival Elevator Pitch challenge. I think his pitch is the response that best adheres to the restrictions of an elevator pitch. It’s short, non technical, and will make perfect sense to anyone, technical or not. Doubtless there’s someone who can adduce a counter example but all-in-all I find his argument persuasive.

As I’ve said before, I think the elevator pitch challenge is particularly difficult but Ridenour’s solution seems like the perfect response: short and persuasive.

As I mentioned last week, this month’s Emacs Carnival challenge is to make an elevator pitch for Emacs. It’s a pretty tough assignment. Realistically, you have less than a minute to convince your captive audience that they should give Emacs a try. I’m not sure I’d be up to it.

We do, however, have two bloggers willing to try. Christian Tietze and Jakub Nowak have each offered their take on the matter. Although they use slightly different approaches, their pitches are remarkably similar. They both emphasize that with Emacs you get a uniform interface and a minimum of context switching. Rather than dealing with separate applications and their different interfaces for your tasks, you can bring the majority of them under the Emacs umbrella and enjoy a single interface for all of them.

There are menus if you want them but, for me, a chief strength is that you can do everything without using a mouse. After all, Emacs was developed for use on a terminal long before the ascendancy of mice.

It’s interesting to me that both Tietze and Nowak chose to stress interface consistency. It is, for sure, an important aspect of Emacs use but, surely, there are more important aspects to discuss. The problem is that it’s hard to discuss those other aspects in the time required for an elevator ride. Perhaps interface consistency and lack of context switching is the best approach.

I’d guess that most people don’t think much about different interfaces and context switching until someone points the problem out to them. Perhaps that makes them a good elevator pitch. What do you think?