JOVE – Jonathan’s Own Version of Emacs | Hacker News Reader

JOVE – Jonathan’s Own Version of Emacs

github.com

72 pointsnanna46 comments

Wikipedia: https://en.wikipedia.org/wiki/JOVE

I used to have a government job for a while, a desk job, nothing related to IT. But of course we had computers. It was all Windows machines with really old versions of Windows and very restricted in what software they allowed. Also, when I asked the IT folks if I could have Emacs on my machine, they just asked: "what is Emacs?", and I knew I was probably in the wrong place.

As government jobs go, there was a period of multiple weeks with veeery little to do for me. I would have really liked to kill the time with some programming, but of course, the computers had no compilers or anything like that installed, like I said, it was not an IT position. I'm not much of a Windows person myself, but a google search informed me that MS had its own version of JavaScript called jscript.

So after some playing around with it, I convinced myself that it should be possible to build an Emacs clone in jscript, and set out to do so. I wanted it to be fully compatible to Emacs Lisp with the goal that I could eventually install genuine Emacs modes.

Suddenly, my job was fun! As a testament to government jobs, I was working on this almost fulltime for a couple of weeks, and no-one noticed. And I got pretty far, but eventually, I did actually get more tasks to work on from my boss and development stalled. But I think that were probably the best few weeks during my whole time in that job (I quit not too soon after).

Sounds like you would have had the passion for coding and engineering in a private sector job!

Yeah, I'm looking right now. But unfortunately, I live in a rather rural area now and there isn't much in terms of SWE jobs around here.

Emacs is a great outlet for a bored developer. Any time at work when things aren't moving along fast I try to build some useful tool or try to port some internal tool to Emacs Lisp. I always learn a lot and have a blast doing it. Plus I feel it's a much healthier outlet then trying to bring in each and every tech fad to the companies stack.

Hah, on a much smaller scale I had a phase early in my career where I had to wrangle an autoconf build system to work on many unix variants used in HPC. I got pretty fluent in the minimal, portable subset of Bourne shell that actually worked everywhere.

Also, some of the HPC platforms had terrible performance with autoconf and similar because they were terrible at fork/exec latency. Rather than doing copy-on-write virtual memory and fine grained time-slicing, like a typical workstation, they seemed to be doing stop-and-swap of whole processes or something. Something that took minutes on a typical workstation could take hours on the HPC service node.

Anyway, home sick with a fever, I got the weird impulse to figure ways to avoid forking subprocesses from the shell. A common one was invoking `expr` for arithmetic, so I started writing my own arithmetic library in pure Bourne shell.

It was decades ago, so I can't recite the actual code. But I remember that I tried a couple versions. My fever broke before I got around to multiplication and division, but I assumed I would just use naive iteration on top of addition/subtraction.

The first version was direct ASCII-encoded decimal operators that exploded the input strings into sequences of digits and iterated them using a case statement encoding the 10x10 sum and carry results. It used function call arg list to iterate the digits and reverse them, so it could then process sum digits from least to most significant.

Since I was processing operand strings from least to most significant digit, it was actually an arbitrary precision operator. I mapped the empty string operands as implicitly zero padded.

After that, I think I experimented with larger bases instead of decimal. E.g. split the string into groups of 2 or 3 decimal digits and encode 100x100 or 1000x1000 lookup tables. If I remember correctly, t wasn't very practical, as the larger case statements didn't really perform well.

> Unlike GNU Emacs, JOVE does not support UTF-8.

If this is still true in the latest versions, I find it pretty amazing that something like this has been maintained all the way until 2023.

Still maintained. There was an update in May of this year.

Well, that's basically a deal breaker in 2025.

But the real question is: Can it run evil mode?!

No. It lacks elisp. It offers some familiar keyboard shortcuts to appease your muscle memory, multiple buffers, screen splits, but apparently not much more.

ASCII is still adequate for great many programming tasks, especially in highly confined environment where JOVE can make sense.

Found this intro to Jove helpful:

https://opensource.com/article/17/1/jove-lightweight-alterna...

Think I might actually use it when I need to make a quick edit to something in the terminal, instead of `nano` or `emacs -nw`

> How small is it? The Jove executable is roughly 150K

Which counts as small now that nano has ballooned to ~400KB?

For comparison, Turbo Pascal packed an entire x86 IDE into ~40KB.[1]

[1] https://prog21.dadgum.com/116.html

You should look into mg. It's included in OpenBSD's base system, and plenty of other systems have it present in their package manager.

https://en.m.wikipedia.org/wiki/Mg_(text_editor)

Hadn't hear of mg. Thanks!

Pre-installed on macOS since 2019 (when it replaced Emacs), too.

I used to use JED when I was stuck on DOS. I was surprised to find out that it's still being maintained on github and the author recently made a commit to improve vms (!) support. https://www.jedsoft.org/jed/

I used to use jed all the time, many years ago. But a package for it isn't available for the os (opensuse) I'm using these days and I couldn't get past the configure script when I tried building it myself. Maybe I should try again.

OpenSuse is such a great base OS, everything is cohesive, but I really wish it had more packages. Flatpak covers most of its shortcomings but sometimes you need something a bit unusual or that does not pair well with Flapak.

I am currently using Fedora mostly because of this.

My first Linux computer was a laptop-ish sort of thing with 4 megs of RAM. Emacs was a bit too heavy duty, so I used Jed quite a bit.

I still use both emacs and jed.

jed starts very fast and has much lower memory usage, so it is well suited for quick edits of configuration files and scripts and other workflows, where you start your editor in your shell, instead of the other way around.

emacs (through packages) can be turned into a custom IDE for a lot of languages, but takes more disk space and uses more RAM.

It comes standard on a default install of Slackware. Even in current, as of now, jove is installed in version 4.17.5.5. Of course, standard emacs is also provided.

Now, if I have to use an emacs-like editor I'd go with Jed. Somehow it seems much less daunting and much more friendly than the real thing.

How does it fare against OpenBSD's mg?

Jove was my first editor on Unix. Emacs took up too much resource and was too slow back then.

Back when?

most of the 80's and 90's?

around 1998, I was working at a regional ISP, my main workstation was a sparc 5, but I had picked up a conversion box from ps-2 to sparc so I could use a Microsoft natural keyboard. emacs was still considered "eight megs and constantly swapping", but I had 96mb of memory in my sparc, and was able to run it.

Microsoft paid a visit to our isp, trying to make a deal, saw no windows anywhere but were excited to see my Microsoft keyboard. they asked about my computer, I told them I was running solaris on a sparc, and they were excited to ask me if I had run internet explorer, that they had just released for solaris. I looked at them horrified, and said, "I only have 96mb of ram in this, I can't run internet explorer!" - but I was able to run multiple windows of emacs, many terminals, a window manager, and netscape (just not the web server, because we ran apache)

I also started using Jove back when 30 of us shared one PDP 11/44 running BSD Unix, and it was antisocial to use something as heavyweight as Emacs. 40 years later, I'm still using UNIX and Emacs.

It was the same for new CS undergrads at UC Berkeley back in the early 90s. There were still labs full of VT220 or similar serial terminals all hooked up to a shared computer.

On reflection, it probably explains why I've used Emacs for my whole career but never really got into any of the elisp customization or other advanced features. I still base my work in the shell (and filesystem) and launch ephemeral Emacs processes rather than living in it as some folks do. I never got interested in IDE functions like controlling compilers nor debuggers from within Emacs.

I never even wanted Emacs to split a terminal window into smaller "screens". I learned the key combo to abort that, much like I learned only enough vi to kill off an unintended launch. But, I do get a lot of mileage out of the XEmacs "frames", i.e. independent X windows all fronting the same set of editing buffers. But I also have terminal windows alongside that to do all the other things from the shell that some people prefer to do from inside the editor...

Eight Megs And Constantly Swapping! :D

Jove and Jed were (are still in some cases!) my go-to options when Emacs was too heavyweight

There are many versions and derivatives of Emacs (Torvalds famously uses another one), so my question is: what makes an editor an Emacs-clone?

Is it the usage of a Lisp? The idiosyncratic keybinds? The "everything-is-a-buffer" paradigm?

In order to be an Emacs clone, an editor has to be extensible and customizable in an actual programming language, not necessarily Lisp, but not just keyboard macros or json configuration files.

JOVE is by definition not an Emacs clone, it's just an Erzatz Emacs, a "nonextensible imitation".

Please see my other post with the "authoritative definition" of Emacs and the RMS quotes explaining the difference:

https://news.ycombinator.com/item?id=44632654

https://www.emacswiki.org/emacs/ErsatzEmacs

>A ‘nonextensible imitation’ of a supposed implementation of an Emacs; by authoritative definition in RichardStallman’s 1981 publication, ‘EMACS: the extensible, customizable self-documenting display editor’, such a product is a contradiction in terms, a literal absurdity.

https://www.gnu.org/software/emacs/emacs-paper.html

What is Joe's Editor then?

And what about Jed's editor?

Jed's editor is customizable and programmable, through SLANG scripts.

my first emacs experience was emacs over 2400 baud - it was amazing for what it was, but painful at that rate, plus the resources it was taking on the host, even though I was running in a terminal, were crazy, but the experience was magnificent!

later, I ran memacs on my amiga locally, which was a better experience, had most of what I used, and seemed to work well - that was my introduction to writing code that would run on unix, locally (dcc).

enter a world of unix and x11, real life emacs, and xemacs became my thing (xemacs mostly, later), but jove was still useful: lighter weight on my sparc, seemed to just work, but I'll be darned if I didn't return back to emacs.

now, don't use emacs, when connected to a *nix box I drop back to vi (happy if vim is present), but since the advent of modern ide's that don't suck, I haven't opened emacs. I still miss zippy, Eliza, and the Hanoi towers though.

Not to be confused with JOVIAL, Jules' Own Version of the International Algorithmic Language, which ran the US air traffic control systems for the longest time.

I can only assume that it's a reference/homage. My first reaction on seeing the headline was "Deep cut, dude!"

The first emacs I used at SUNY Stony Brook circa 1989 and pretty much the last. Vi for me

Technically and officially, JOVE is only an "Ersatz Emacs" because it lacks an extension language, and is merely a "nonextensible imitation" of Emacs.

https://www.emacswiki.org/emacs/ErsatzEmacs

ErsatzEmacs

A ‘nonextensible imitation’ of a supposed implementation of an Emacs; by authoritative definition in RichardStallman’s 1981 publication, ‘EMACS: the extensible, customizable self-documenting display editor’, such a product is a contradiction in terms, a literal absurdity.

“Many other editors imitate the EMACS command set and display updating philosophy without providing extensibility. Despite that deficiency, and despite the greatly reduced set of features that results from it, these can be useful editors, though not as useful as an extensible one. For a computer with a small address space or lacking virtual memory, this is probably the best that can be done.

“The proliferation of such superficial facsimiles of EMACS has an unfortunate confusing effect: their users, knowing that they are using an imitation of EMACS, and never having seen EMACS itself, are led to believe that they are enjoying all the advantages of EMACS. Since any real-time display editor is a tremendous improvement over what they probably had before, they believe this readily. To prevent such confusion, we urge everyone to refer to a nonextensible imitation of EMACS as an ‘Ersatz EMACS’.“

– Richard M. Stallman. EmacsTheExtensibleCustomizableSelfDocumentingDisplayEditor. MIT AI Memo 519a, 26 March 1981.

https://www.gnu.org/software/emacs/emacs-paper.html

(I don't have a dog in this fight, but I do have a cat named Emacs, and have known RMS and used and programmed and developed display drivers and user interfaces for Emacs for a long long time.)

Nelson Spins Pip While Emacs Watches

https://www.youtube.com/watch?v=aRaD5zH3Qdg

HCIL Demo - HyperTIES Authoring with UniPress Emacs on NeWS

https://www.youtube.com/watch?v=hhmU2B79EDU

https://news.ycombinator.com/item?id=26113192

https://news.ycombinator.com/item?id=28419139

>I worked at UniPress on the Emacs display driver for the NeWS window system (the PostScript based window system that James Gosling also wrote), with Mike "Emacs Hacker Boss" Gallaher, who was charge of Emacs development at UniPress. One day during the 80's Mike and I were wandering around an East coast science fiction convention, and ran into RMS, who's a regular fixture at such events.

>Mike said: "Hello, Richard. I heard a rumor that your house burned down. That's terrible! Is it true?"

>RMS replied right back: "Yes, it did. But where you work, you probably heard about it in advance."

>Everybody laughed. It was a joke! Nobody's feelings were hurt. He's a funny guy, quick on his feet!

Reply on news.ycombinator.com