4 ms·

"I don’t have figures at hand but it’s my impression that Emacs development and the introduction of major new features has accelerated lately. Regardless, it’s

by buster 3y ago

"I don’t have figures at hand but it’s my impression that Emacs development and the introduction of major new features has accelerated lately. Regardless, it’s clearly not the case that development has slowed or stalled." For such a post, it'd be much nicer to have some statistics. It does come around as a post by some heavy user preaching that everything is fine and well but that's just a subjective view from a bubble. How many commits did emacs have over the last ~10 years? How many updates and packages in ELPA l/MELPA over the last years? I find it hard to believe that development has not stagnated a bit in times of all the new editors and lots of young new developers not having touched a terminal and growing up on full fledged IDEs. I'm saying that as someone liking emacs (I do occasionally use it), but using sublime text and intellij most of my time. I find emacs an incredible piece of software and which for emacs to stay around!

iLemming 3y ago
One also has to consider other factors. Despite the evolution of the core of Emacs not being impressive and possibly even stagnating, it is remarkable how many external packages are available. GitHub alone hosts numerous Emacs Lisp packages, making it the largest body of Lisp code among various Lisp dialects, surpassing Clojure, Common Lisp, Racket, and others. Isn't that impressive? Can someone name a single scripting language used exclusively for configuring a specific editor that is as widely used as Emacs Lisp? Elisp is not a general use programming language; it's not very popular, even among Lisp-hackers; it's not even a great Lisp - it lacks tons of nice features other Lisp dialects had for decades. Yet it is undeniably, the most used Lisp of them all. How can anyone observe all of this and claim that Emacs is dying?
- charcircuit 3y ago
  >How can anyone observe all of this and claim that Emacs is dying? Because LISP is also dying. Compare JavaScript vs Emacs Lisp on Google Trends. It doesn't make sense to have an editor exclusively have it's own language instead of reusing a commonly existing one.
  - iLemming 3y ago
    Lisp is not "dying". Lisp is constantly evolving and keeps influencing other languages, like Elixir, Julia and others. Even modern Javascript borrowed tons of features from Clojure - a language that many big and small companies succesfully use in production. There are constantly new experimental things happening in Lisp. People building their Hammerspoon, AwesomeWM and Neovim configs with Fennel. Researches keep publishing papers, experimenting in Racket. Even writing books about type theory, which may sound strange since most Lisps are dynamically typed. You seem to be completely oblivious to the world of modern Lisp, we stopped capitalizing each letter around 1980s. No actual programmer who wrote at least a few Lisp functions would write it today as "LISP". > It doesn't make sense to have an editor exclusively have it's own language It may not make sense to you because you don't seem to see the benefits of Lisp. Some extensions for VSCode are written in Clojurescript. Even with obvious overhead, some people see the benefits of choosing that path. While it is possible to go the other route - writing Javascript or Python that emits Lisp, I don't know why anyone would do that. Nobody stopping anyone from writing a Javascript transpiler/compiler that emits Elisp code to configure Emacs. But why would you do that, if Elisp (admittedly not even being the greatest Lisp) is still much more powerful? > Compare JavaScript vs Emacs Lisp on Google Trends. What's your point? Javascript is a general-purpose language, it's used in browsers, cloud computing, embedded devices, Game Dev, IoT, Desktop Apps, etc. Elisp is not. Its sole purpose is to be used as a configuration language for one, specific piece of software, that (at least officially) only runs on desktop, and that's it. Yet despite all that, there's quite a lot of it. And people keep writing more. There are few new packages get announced every month or so on r/emacs. That's what I'm saying.
    - charcircuit 3y ago
      >Lisp is constantly evolving and keeps influencing other languages, like Elixir, Julia and others. A language doesn't have to be alive to evolve or influence other languages. >Clojure - a language that many big and small companies succesfully use in production. This sounds less dead, but Clojure can get away with being semi dead due to its interopt with normal java libraries. No one at my university or the people I have worked with have used Clojure, so it is at most niche. >People building their Hammerspoon, AwesomeWM and Neovim configs with Fennel. These are configured with Lua. They do not support Fennel by default from what I can tell. >Researches keep publishing papers, experimenting in Racket Researches alone can't make a language reach mass adoption. >Even writing books about type theory, which may sound strange since most Lisps are dynamically typed. And such languages produced in those are incredibly niche and likely dead. >Nobody stopping anyone from writing a Javascript transpiler/compiler that emits Elisp code to configure Emacs. This misses the point. Javascript is the most well known language among developers. Emacs should be making their extension platform attractive to developers, allowing people to use a language they already know is better than forcing them to learn a new, unfamiliar language. >>Compare JavaScript vs Emacs Lisp on Google Trends. >What's your point? The probability that a developer knows JSON and Javascript is much higher than that developers knowing Elisp, or even a Lisp at all.
- buster 3y ago
  "there is a lot of X" does not imply "X is increasing". That's what i am pointing out. There is also a lot of COBOL code. It doesn't mean that COBOL code is increasing. The Pyramides are huge and impressive. It doesn't mean that more pyramids are built.

by buster 3y ago

iLemming 3y ago

One also has to consider other factors. Despite the evolution of the core of Emacs not being impressive and possibly even stagnating, it is remarkable how many external packages are available. GitHub alone hosts numerous Emacs Lisp packages, making it the largest body of Lisp code among various Lisp dialects, surpassing Clojure, Common Lisp, Racket, and others. Isn't that impressive? Can someone name a single scripting language used exclusively for configuring a specific editor that is as widely used as Emacs Lisp? Elisp is not a general use programming language; it's not very popular, even among Lisp-hackers; it's not even a great Lisp - it lacks tons of nice features other Lisp dialects had for decades. Yet it is undeniably, the most used Lisp of them all. How can anyone observe all of this and claim that Emacs is dying?

charcircuit 3y ago
>How can anyone observe all of this and claim that Emacs is dying? Because LISP is also dying. Compare JavaScript vs Emacs Lisp on Google Trends. It doesn't make sense to have an editor exclusively have it's own language instead of reusing a commonly existing one.
- iLemming 3y ago
  Lisp is not "dying". Lisp is constantly evolving and keeps influencing other languages, like Elixir, Julia and others. Even modern Javascript borrowed tons of features from Clojure - a language that many big and small companies succesfully use in production. There are constantly new experimental things happening in Lisp. People building their Hammerspoon, AwesomeWM and Neovim configs with Fennel. Researches keep publishing papers, experimenting in Racket. Even writing books about type theory, which may sound strange since most Lisps are dynamically typed. You seem to be completely oblivious to the world of modern Lisp, we stopped capitalizing each letter around 1980s. No actual programmer who wrote at least a few Lisp functions would write it today as "LISP". > It doesn't make sense to have an editor exclusively have it's own language It may not make sense to you because you don't seem to see the benefits of Lisp. Some extensions for VSCode are written in Clojurescript. Even with obvious overhead, some people see the benefits of choosing that path. While it is possible to go the other route - writing Javascript or Python that emits Lisp, I don't know why anyone would do that. Nobody stopping anyone from writing a Javascript transpiler/compiler that emits Elisp code to configure Emacs. But why would you do that, if Elisp (admittedly not even being the greatest Lisp) is still much more powerful? > Compare JavaScript vs Emacs Lisp on Google Trends. What's your point? Javascript is a general-purpose language, it's used in browsers, cloud computing, embedded devices, Game Dev, IoT, Desktop Apps, etc. Elisp is not. Its sole purpose is to be used as a configuration language for one, specific piece of software, that (at least officially) only runs on desktop, and that's it. Yet despite all that, there's quite a lot of it. And people keep writing more. There are few new packages get announced every month or so on r/emacs. That's what I'm saying.
  - charcircuit 3y ago
    >Lisp is constantly evolving and keeps influencing other languages, like Elixir, Julia and others. A language doesn't have to be alive to evolve or influence other languages. >Clojure - a language that many big and small companies succesfully use in production. This sounds less dead, but Clojure can get away with being semi dead due to its interopt with normal java libraries. No one at my university or the people I have worked with have used Clojure, so it is at most niche. >People building their Hammerspoon, AwesomeWM and Neovim configs with Fennel. These are configured with Lua. They do not support Fennel by default from what I can tell. >Researches keep publishing papers, experimenting in Racket Researches alone can't make a language reach mass adoption. >Even writing books about type theory, which may sound strange since most Lisps are dynamically typed. And such languages produced in those are incredibly niche and likely dead. >Nobody stopping anyone from writing a Javascript transpiler/compiler that emits Elisp code to configure Emacs. This misses the point. Javascript is the most well known language among developers. Emacs should be making their extension platform attractive to developers, allowing people to use a language they already know is better than forcing them to learn a new, unfamiliar language. >>Compare JavaScript vs Emacs Lisp on Google Trends. >What's your point? The probability that a developer knows JSON and Javascript is much higher than that developers knowing Elisp, or even a Lisp at all.
buster 3y ago
"there is a lot of X" does not imply "X is increasing". That's what i am pointing out. There is also a lot of COBOL code. It doesn't mean that COBOL code is increasing. The Pyramides are huge and impressive. It doesn't mean that more pyramids are built.