LWN: Comments on "Making Emacs popular again"

Making Emacs popular again

innocentoldguy — Fri, 03 Dec 2021 19:46:51 +0000

Vim isn't the only modal editor. Kakoune is another one. I use Kakoune quite a bit because I like its simplicity and adherence to the Unix philosophy. For example, Kakoune doesn't do screen splits but instead recommends you use tmux for that.

I've recently been learning Emacs because I like Org Mode. My only issue with Emacs is that the plugins seem to range from unpolished to non-functional (adoc-mode). Plugins for both Vim and Kakoune feel more polished and well-thought-out to me.

Vim's popularity

CRConrad — Sun, 07 Nov 2021 11:07:49 +0000

> And yet, vim is still the most popular editor for programmers, isn’t it?

I think that strongly depends upon whether you define "most popular" as "most beloved" or "most used".

My guess Vim is "the most popular editor for programmers" in the latter sense, because vim is installed as default on pretty much every single Linux system where developers have to occasionally edit git messages, config files, or shell scripts; but that doesn't necessarily imply the former.

Making Emacs popular again

nix — Sat, 23 Jan 2021 18:10:19 +0000

It's very similar to other Lisps (if archaic in many ways), so learning it helps learning them, and if you know any of them it's pretty easy to pick up elisp.

Also... if you live in Emacs, its' likely that learning it is useful for its own sake, even if it *is* useful nowhere else. It's like being able to reprogram your brain.

Making Emacs popular again

xiaoxing — Sat, 12 Dec 2020 22:09:56 +0000

The biggest problem for me is that elisp is useless elsewhere. It’s hard for me to invest in learning and mastering the language if it’s not useful outside of one tool. I love emacs and has been using it for years since I switched from vim. But I’m still in the realm of copy & paste other people’s snippet and stitch it onto my own config.

Making Emacs popular again

ceplm — Tue, 14 Jul 2020 22:25:52 +0000

> … less likely to be interested in spending the time to learn their tools, which vim and emacs both require.

And yet, vim is still the most popular editor for programmers, isn’t it?

Making Emacs popular again

debacle — Wed, 08 Jul 2020 08:03:15 +0000

If the device supports FTP, you can just using Emacs with TRAMP, I assume?

What about adopting Hemlock?

pabs — Sun, 31 May 2020 02:12:19 +0000

Surely "it could be faster" is a good problem to solve?

What about adopting Hemlock?

jem — Sat, 30 May 2020 18:34:11 +0000

Everybody is free to use a text editor of his or her choice. This is the first time I heard of Hemlock, and I have been using different versions of Emacs for decades. I took a look at the website (https://www.cons.org/cmucl/index.html) and was not convinced. All the manual links are broken, and the latest news item is from 2012.

If you are suggesting the GNU Emacs project should abandon their current code base and adopt Hemlock instead, what problem would that solve? Would it solve the usability or the not-so-sleek look of GNU Emacs mentioned in the article? What about the millions of lines of Elisp code? Would Elisp compatibility be possible to implement in Hemlock? Emacs Lisp is not going away ever.

The port of GNU Emacs to use Guile instead of the built-in Lisp implementation has failed so far, but not because Scheme is a weak language or Guile a bad implementation of Scheme. The problem with Guile Emacs has to do with the difficulty of integrating Guile with Emacs and getting the final bits of Elisp compatibility to work, getting the editor to start up fast enough, etc.

The idea with Guile Emacs is not to replace Emacs Lisp with Scheme (not initially, anyway), but to use a better (faster), and more decoupled implementation. Interest in Guile Emacs has slowed down because people are asking the rhetorical question "what problem does it solve?"

What about adopting Hemlock?

jgfenix — Sat, 30 May 2020 17:24:26 +0000

It's a Emacs written in Common Lisp.

Making Emacs popular again

flussence — Wed, 27 May 2020 03:41:42 +0000

> It's quite possible that there would be almost no free software, no linux or lwn.net, no gitlab/github, etc, etc, if it had not been for his unfailing efforts and unwavering belief in free software though the years.

Snuffing out the GPL3 has become one of the foundational pillars of a trillion-dollar industry, so uh, congrats to him for his bit part in drafting a relatively short anti-shibboleth that works on both an economic and social level?

And I'm pretty sure the existence of Git was mostly Tridgell's fault. I've never seen a VCS commit with Stallman's name anywhere near it; in fact I don't think I've seen him write any software this century.

BSD says hi, by the way.

Making Emacs popular again

Cyberax — Tue, 26 May 2020 18:30:03 +0000

> Free software, by its very nature, *can't* become extinct. Even if current trends/fads mean that there is a lull in the number of people using Gnu emacs today, the source code will still be available for future generations to discover and use.
Why would future generations use it?

Sure, even now there are people who now resurrect old IBM mainframes for fun, but this doesn't mean that the old mainframe programming is in any way alive.

Like it or not, tool development is a Red Queen's race - you have to run just to be able to keep up. Emacs code editing ability is already kinda borderline compared to more modern tools, and it's only going to get worse.

Making Emacs popular again

jnorden — Tue, 26 May 2020 18:21:07 +0000

I'm late to this party, but as a longtime user of Gnu emacs, I feel obligated to weigh in. I've been using emacs on an almost daily basis for 30+ years. I use rmail for about 90% of my email. I use a customized version of TeX mode for composing documents, which include exams/quizzes/handout for the classes I teach, research-related work, as well as mundane letters, memos, notes, etc. I use emacs for all of the software that I write and/or dabble with, mostly Perl and C. I use shell-mode about as often as I use a terminal window (currently mate-terminal, a fork of the pre-gnome-3 gnome-terminal).

To start with, the idea that emacs "needs" to have more users to prevent it from becoming "extinct" is basically absurd. Free software, by its very nature, *can't* become extinct. Even if current trends/fads mean that there is a lull in the number of people using Gnu emacs today, the source code will still be available for future generations to discover and use. It's about like saying that we must find a way to make the "Early New English" language of the 17th century more appealing and widely spoken in order to prevent the works of Shakespeare from becoming extinct. Even if, for some reason, people stopped reading and producing Shakespeare's plays for a number of years, they would undoubtedly be re-discovered and become popular again.

This all seems to be part of the current insane attitude that software, and technology in general, is some sort of perishable commodity with the shelf life of milk. Somehow, if it isn't updated every month or so, it just isn't any good any more, even though it still does what it used to and your needs for it haven't changed.

Emacs has never been an editor for "casual" user. It doesn't compete with notepad, any of the various "office" editors (open source or not), or even vi/vim. Gnu emacs is for people that want an extensible editor that gives them complete control over how it operates, and can be easily and freely customized to accomplish any sort of task that they want it to. This sort of freedom comes with a price - you need to invest some time and effort in order to learn how to use it effectively. But for many of us, it is an effort that has been more than worthwhile.

In my opinion, it would be incredibly counterproductive to try to attract people who don't need the functionality that emacs provides, or who aren't willing to put forth the effort required to learn how to effectively use that functionality. I believe this means that any person who's decision on whether or not to use an editor is swayed by the appearance of buttons or rounded corners is someone who should *not* be encouraged to start using emacs. If you are not attracted to emacs by the features it provides and the tasks it can accomplish, then please find an editor that will better suit your needs.

On the other hand, if someone wants to add such features for their own benefit, perhaps because they feel it will enhance their own aesthetic experience while using emacs, then by all means do so. That is the whole point of free software, after all. But adding these in an attempt to attract more users is a bad idea.

My *fear* is that a major effort to increase the "user base" will lead to the transformation of emacs into something that doesn't serve anybody's needs very well. This is happening in many open source projects, where all sorts of functionality has been deprecated and then removed because of the perception that it isn't needed or being used by a large enough fraction of users. The recent loss of malloc_get_state() and malloc_set_state() are examples that are particularly relevant to emacs.

Even in emacs, I personally found it a bit annoying to type "M-x count lines region" only to be told in the mini-buffer that:

‘count-lines-region’ is obsolete; use ‘count-words-region’ instead.

But this was easily fixed by adding a single line to my .emacs file. However, if large blocks of code start disappearing from the source, or changes are made that render existing elisp files unusable, then emacs really will run the risk of becoming extinct.

For example, a package of elisp functions that I wrote 30 years ago for emacs-18, which I use to record and average student grades, still works just fine with emacs-26. TeX is the only other software that I know of with this level of stability. It seems that there are very few people today who, like Knuth and Stallman, take a long-term view of what they are trying to accomplish. I could go on along these lines, but this is probably sufficient.

----

However, I feel that I must respond directly to some of the comments about RMS that have been made, along the lines of "emacs would be better without him" or his "signature tantrums." I'll respond in a way that RMS never would, because he is far too polite:

Do you have any idea who the f*** you are talking about!!?

When Richard founded the FSF, which basically started the free software movement, people tried to write him off as some sort of extremest nutcase. "Nobody will write software and just give it away" was a common criticism. Well, history has shown that Stallman was correct, and his critics were the nutcases. It's quite possible that there would be almost no free software, no linux or lwn.net, no gitlab/github, etc, etc, if it had not been for his unfailing efforts and unwavering belief in free software though the years. My own opinion is that, if anything, Richard's opinions and views are a bit too mild and conservative.

The arrogance of youth is natural. I was certainly guilty of it when I was young. But there is no excuse for disrespecting the people who basically built the universe that you currently enjoy inhabiting.

-Jeff

Making Emacs popular again

moltonel — Mon, 25 May 2020 17:00:00 +0000

That's where a switch to guile seems most appealing : it supports EmacsLisp, Scheme, ECMAScript (aka javascript), and soon Lua. I use Emacs all day every day, but I don't like lisp, and I'm sure many potential Emacs users are turned away by EmacsLisp. Contributing to Emacs would be much more appealing if you could pick and choose between any guile-supported language.

Making Emacs popular again

N0NB — Thu, 14 May 2020 13:06:16 +0000

I Liked CUA mode in Emacs until I set out to use Org-mode earlier this year to track a comprehensive TODO list. It was strongly recommended by Org-mode that CUA mode be disabled so back to default keybindings it was. Especially in the way I am using Org-mode, the Shift+arrow keys are used to set the TODO tags and CUA mode would highlight text instead. If a new user is coming to Emacs for Org-mode then they're likely going to have to learn the default Emacs keybindings.

As already mentioned, anachronisms abound and a new user needs a cross reference to make sense of it. Over time it becomes second nature as one moves between other applications and Emacs.

Some features show that the GUI is a bolted on afterthought rather than well
As already mentioned, anachronisms abound and a new user needs a cross reference to make sense of it. Over time it becomes second nature as one moves between other applications and Emacs.integrated into Emacs. One of the excellent features of Emacs is the kill ring, however, I don't recall it being exposed through the GUI so that text in the kill ring could be selected rather than have the chunk of text be dumped into the buffer at point (another anachronism) each time a new chuck is recalled. This behavior has allowed me to make some rather spectacular editing errors. Fortunately, as I recall, the kill ring is copied to the clipboard and I use Gnome's clipboard extension to recall a text chunk. No, I am not the most efficient at using an editor and I accept that.

I use Emacs because it works very well with Autoconf, Automake, and shell files, Org-mode, GNUS, and because I have a highly customized chunk of my init.el devoted to tweaking the C-mode syntax highlighter exactly to my liking. ATM my init.el runs to 297 lines including comments and I still struggle with elisp. I've tried a multitude of GUI editors and none come close to being able to tweak the syntax highlighter to the detail I'd like. Most use Scintilla which offers a certain amount of control over the highlighter but is still lacking for my preference. Non-free editors such as VS Code need not apply here.

RE: lsp, C++ (and rtags) (Making Emacs popular again)

dfszb — Thu, 14 May 2020 13:01:37 +0000

Just for the record, we had rtags (https://github.com/Andersbakken/rtags) before lsp-mode (and the ccls backend), and rtags did and still does a great job, the symbol usage lookup is still better imho than those offered by lsp-mode backends (ccls or clangd). I used it on several projects, from small to fairly large (500k sloc), with heavy c++ / boost usage. The lsp-mode tools offer much nicer UI and very flexible completion, they can also slow down emacs to be like Visual Studio.

Making Emacs popular again

bmorel — Thu, 14 May 2020 13:01:10 +0000

> What fascinates me is: vim is apparently the last _modal_ editor left.

There is also kakoune (https://kakoune.org/).
I was willing to try to get used to it since ages, but debian's version didn't had syntax coloration embedded by default, which was a blocker. This thread (specifically your comment) reminded me of it, and nice surprise for me is that, by default on current debian, it *have* syntax coloration without having to mess with config. So I'm gonna give it a try. Maybe I'll just revert quickly to vim anyway, since my fingers have some muscle memory with it now.

> I _always_ forget which mode I am in. Anyone knows what's wrong with me?

Dunno. I don't think I have issue with that.
Basically, I see 4 modes: visual, which is pretty... visual, for selections. Command, which put my cursor in the command area with the ':' prefix, 'normal' and 'editing' (this one have 2 variants, yes: replace and insert, and unfortunately, by default, the cursor keeps the same shape).
The only possibility of confusion I can see is between normal and edit modes.

I think why I don't have problems reminding the mode I am in is because I only go into editing for actual writing. When I'm done, I just escape to normal mode to do whatever I want to do.
I would not call me an advanced vi user, and I don't use it since that many years, less than 10.
Vi-like editors... well, that and bash/zsh... are really usable to me *because* I discovered tiling window managers. They are all powerful tools by themselves, but require time to learn. Their main selling point is that they allow to kill the rat, but that's useless if you need that beast to control and place your windows.
And that makes the reason for which I can avoid using an IDE: my desktop environment have mutated to a development environment, still lacks some features compared to actual IDEs, but have mails, calculator, music player, image viewer, crash-resilience and speed. Can also run most of it's parts without X11, useful for sshing, and it's shortcuts, icons, toolbars and popups really rarely change with new versions.
What I know is that, on the 6 dev colleagues I had on my last job, 2 moved to vim+bash+i3, and I didn't said them to do so, they did it because they wanted to. So, learning vi/emacs is *not* related to age and modern techs.

To the topic... I don't see emac's point, exactly because of how I setup my environment: it does too much stuff, so having it in my environment would mean redundant tools, which I try to avoid (well, vim have window manager integrated, too, I never use it and would be happy if it was not here).
If emacs wants to rise anew, then they should not aim at competing with text/code editors, but with IDEs, because the integration of many features seems to be it's selling point, but it is by far harder to use and setup than the average IDE.

Maybe provide some packages that does IDE by default? Some easy way to summon it with different setups depending on the task people want to do? I think Visual Studio did that last time I had to use it: depending on the language, it had a different look, a different way to do things.

Making Emacs popular again

antismap — Thu, 14 May 2020 12:59:02 +0000

I use Emacs every day, but not for serious code editing. For c/c++ I use kdevelop, for python pycharm. Those IDEs are just better for the job.
For all the rest I use emacs.
vscode or other solutions may be nice in code completion and being "smart" but for those tasks, Emacs will always be more practical:
- open a file while browsing directly in the editor
- copy / paste files around, create folders, directly from the editor
- file operations such as grep, filter lines and so on
- git operations such as log, commit, push... (using magit)
- generally doing something in small window divisions, being able to copy/paste part of each division to other ones

IDEs suck at this, it starts with opening a file which is not part of a project. It's just a pain.

Making Emacs popular again

jem — Wed, 13 May 2020 17:43:37 +0000

With this setting I can get rid of the backward jump. But I guess I'd better get used to the standard behavior of (n)vi(m), even if it feels like I am dragging along a single character selection with the movement commands -- a selection I can extend with the v command, but never wholly get rid of.

There's a lot to like about vi, too, for example how you are able to combine editing commands with movement commands. At least you don't have to reach for the arrow keys all the time.

Making Emacs popular again

karkhaz — Wed, 13 May 2020 12:10:07 +0000

Does writing "inoremap <esc> <esc>l" in your vimrc implement the behaviour you expect?

Making Emacs popular again

jem — Wed, 13 May 2020 11:07:39 +0000

>It's just one setting away on Vim:
>set virtualedit=onemore

Ok, thanks. But this is only a partial solution. As you said, $ does not work, and pressing <esc> in insert mode still moves the cursor one position left. You would think i<esc> (entering insert mode and immediately exiting) would be a no-op, but no. My biggest problem with this setting, though, is that it is not supported in VS Code's vim emulator, which is my main use case for vim: to replace VS Code's Notepad-like default editor with a real editor.

Making Emacs popular again

roc — Wed, 13 May 2020 08:53:11 +0000

I suppose that's true, but for me the GNOME compositor never crashes and I do occasionally have to reboot for kernel hangs, so I want the applications I use to tolerate the latter.

Making Emacs popular again

sjatkins — Tue, 12 May 2020 18:32:28 +0000

For me orgmode is what I most miss lately when working without other editors and IDES. Next up would be very fast rgrep and direct opens from results. I haven't seen it done as efficiently and generally in any IDE/other editor. The C world support is quite excellent. Back in the day I thought well of JVM support also. To be honest I haven't fully deployed all the python (what I most code in in the day job) support that is present. magit would on my list. Also Tramp as finicky as it can sometimes be has been a life saver.

When I am working/debugging against an remote server emacs in a screen session is a godsend. No GUI based tool can do that.

Making Emacs popular again

sjatkins — Tue, 12 May 2020 18:32:04 +0000

I don't think it was altogether snark and I have done fairly extensive modern FE work in both angular and react frameworks and bits of straight jquery based stuff before that. I found the FE javascript world to be a highly fragmented mess for quite some time. I do feel it is getting better slowly. But the npm world of fragmented code bodies originally to limit download sizes and versioning hell is not for the faint-hearted. Even today most projects I have seen choose their base versions once and craft largely by hand what that they need is compatible with it and lock it down. Updating can take a month off the schedule easy to rebalance all the versions of everything used or needed compatibly. That is not a good situation. I have been tempted more than once to create a small AI that given a set of absolute required modules would find the most balanced set of of other modules transitively required automagically. It is not a small task as many modules lock down the version of things they require to strongly, much more strongly than the code requires.

Making Emacs popular again

epa — Tue, 12 May 2020 09:50:52 +0000

M-x quasi-mode

Making Emacs popular again

epa — Tue, 12 May 2020 06:45:33 +0000

I was pleasantly surprised to use a Mac and find that Ctrl-A for beginning of line worked. Are there any others?

Making Emacs popular again

filbranden — Tue, 12 May 2020 04:44:34 +0000

> Qt is only available for GPL 2 and 3, using it would mean that if there is a GPL 4 someday, Emacs could not switch to it.

This doesn't make any sense whatsoever.

If that was the case, then you wouldn't be able to port Emacs to FreeBSD and run on a libc licensed under BSD.

Or maybe run on top of a Linux kernel which is licensed under GPL 2 exclusively. (Ok maybe I'm stretching, but is that really too much?)

Lucid toolkit is LGPL 2.1 or MPL 1.1, not "or later", should those bindings in Emacs be removed then?

I can only take it that this argument against Qt due to licensing was:

* Made by people who don't really know how licenses work, possibly not having a clear understanding of open source, perhaps because they haven't been in close contact with open source and are not keen on it; or was

* Made with malicious intentions, political in nature, perhaps trying to imply that if the Qt project was desperate and naive enough to hand these people a blank cheque in form of allowing them to relicense their code however they wished, they would be given the blessing to allow their toolkit to run the one true and pure editor, which is used by, checks notes, 3% of developers.

Sometimes I really do understand people who dismiss the GPL, especially when seeing this kind of zealotry, particularly by people who *should know better.*

Making Emacs popular again

filbranden — Tue, 12 May 2020 04:21:11 +0000

> A "mega" thread, eh? Make Emacs Great Again?

Haha, my thoughts exactly. And I'm happy to see how you also caught the "subtle" reference in the adjective on the thread.

That, together with one of the main characters in the story last having made the news under the #MeToo hashtag, and the introduction of that character being followed by a discussion on license pedantry made for an article I could almost have expected to find on a tabloid rather than LWN :-)

Surely a fun read though... And the comments too!

Making Emacs popular again

filbranden — Tue, 12 May 2020 04:15:31 +0000

> Maybe Gnemacs?

Neomacs!

Making Emacs popular again

filbranden — Tue, 12 May 2020 04:04:30 +0000

> Surely there are some vi-like editors which fix this off-by-one design error?

It's just one setting away on Vim:

set virtualedit=onemore

You might still find some of the behavior a bit unexpected. For example, $ moves to the last character of the line, rather than one past.

Also, you might just break so many plug-ins this way :-)

But yeah, it's available as a configuration option.

Making Emacs popular again

Cyberax — Tue, 12 May 2020 02:53:10 +0000

VSCode is just faster and more fluid than Atom.

(I prefer IntelliJ editors, they are unmatched in their semantic editing support)

Making Emacs popular again

briangordon — Tue, 12 May 2020 01:15:01 +0000

Atom was there first on the JavaScript point, and seems to have gone even further to make every little thing pointedly modularized and extensible. Yet VSC came out on top. Maybe someone who used both editors can speak to why. My only guess is that the first-class (and conspicuous) git integration in VSC became a killer feature for people who wanted an editor equipped to interact with the burgeoning GitHub/GitLab community but who weren't totally comfortable with the command line interface.

Great support, and lots of cool new tools

tnoo — Mon, 11 May 2020 19:47:49 +0000

> Honestly, Emacs would probably be in a better place if RMS didn't have any say.

Emacs is his baby, and he cared for it his whole life. And he still does: I had a reproducible crash a few years ago. It took one hour for him to send me a first patch that fixed the bug. And after one day of discussion on the mailing list the root cause was found and fixed. No vendor does that.

For me, the amazing thing is that emacs got a lot more momentum thanks to great tools like org-mode, magit, ivy/avy/helm, pdftools, org-ref that were not around ten years ago.

Making Emacs popular again

NAR — Mon, 11 May 2020 08:34:50 +0000

I have to say it took me a couple of days to realize what "rings the bell" (literally) means - I'm so indoctrinated with the tech-speak, I didn't think it could be anything else than the terminal beep :-)

Making Emacs popular again

jake — Sun, 10 May 2020 22:06:44 +0000

Glad you liked the article ...

> Just by curiosity, what are the tasks where you need emacs?

The main way to edit articles here at LWN uses Emacs. I also use Gnus to read mailing lists. Once in a great while I will be doing something in Vim and switch over to Emacs for some useful feature. Though I am not expert with either editor, so Vim probably has the feature and I just don't know how to access it :)

jake

Making Emacs popular again

jerojasro — Sun, 10 May 2020 19:38:39 +0000

AFAIK, it's not needing per se, but just that the functionality available in emacs is better than what is available elsewhere; things that come to mind where that applies:

* editing Lisp code, expanding macros, stuff like that.
* using large plugins, like org-mode.

Making Emacs popular again

jem — Sun, 10 May 2020 19:03:00 +0000

I'm not convinced. Lisp is known for having almost no syntax at all, but this big specification adds quite a lot. On top of that, it is backwards compatible, so you can freely mix the old S-expressions and the new syntax.

The specification also adds "meaningful indentation" à la Python (and Haskell), with all the cumbersome interaction between tabs and spaces, and problems with copy-pasting code.

My fix to the readability problem is to indent the code properly, and split the code into separate functions as appropriate. Use an editor which does the indentation automatically, shows matching parentheses, and warns about the "silly extraneous" parentheses. It doesn't really matter if a Lisp function ends with 13 closing parentheses, as long as they are all there. You don't have to count them.

Now, if I could only think of an editor that is up to the task...

Making Emacs popular again

marcH — Sun, 10 May 2020 15:18:52 +0000

> Once you've got used to them, they're much easier to read and understand than 'readable' expressions.

Thanks for your personal impression but please have a look at the reference.

Making Emacs popular again

jem — Sun, 10 May 2020 14:42:02 +0000

And yet, Emacs could be used on an ADM-3A in the mid 1980s with no problem. Using Emacs back then felt very much like using a modern version of Emacs on a text terminal.

Making Emacs popular again

mathstuf — Sun, 10 May 2020 13:39:52 +0000

That's basically what I do for Windows and macOS development except that `git diff > foo.patch && scp foo.patch` is the normal transport medium. Sure, there's Vim, but that's mostly just good enough to make the change; integrating it into the rest of the workflow ends up back on one of my main machines.

Making Emacs popular again

mathstuf — Sun, 10 May 2020 13:38:26 +0000

> I fell in love with Emacs because it's key combinations were based of the command you want to execute instead of where you're hands are on the keyboard like Vim. I just can't memorize the key shortcuts for Vim because I use Dvorak.

hjkl are the only location-based keybindings in Vim. Everything else is mnemonics (or, rarely, forced elsewhere due to availability like <C-e> and <C-y>). I was able to be decently effective when I was using Colemak because the keys were bound to letters, not location (hjkl were in a nice inverted diamond so at least they were proximal).