Hacker News new | past | comments | ask | show | jobs | submit login

I think you might be right, but the greater context is highly ironic.

The company resurrecting the Emacs concepts is Microsoft. They're the ones that created OmniSharp, bringing the idea of language servers to the programming mainstream. They then created the Language Server Protocol for what now is Emacs's biggest long term rival in the flexible editor arena: Visual Studio Code.

Visual Studio Code is for the moment on a somewhat bloated and shaky foundation because of Electron. But otherwise its design is quite solid. And even that foundation will probably become a lot stronger in the next few years as WebAssembly gains wide adoption. Visual Studio Code itself is written in Typescript and it's not hard to imagine Microsoft adding a WASM backend to it.




I think Emacs's greatest strength comes from the ability to use a single scripting language throughout the whole editor. I can quickly write up some ELISP statement to do anything, if I like it I can put it in a file and load it. If I want to share it, I put it in an extension.

I've just looked up the official tutorial on making an extension in VS Code. It seemed cumbersome. You need a code generator to lay the foundations, you need to consult the API docs etc.

With Emacs you can learn as you go and the editor itself guides you if you ever get stuck. You can read the source of any part of Emacs you like, change it on the fly, evaluate and try it (except for some core modules written in C). And I think that's the beauty of Emacs. Emacs is easy to tinker with.


Yeah, but I have a million monkeys^Wdevelopers at my back for Javascript/Typescript. How many divisions does Lisp have?


Out of curiosity I looked it up. It's a factor of 200: 3788 packages on MELPA, vs 650000 packages on NPM.


Here is how Lucid implemented their C++ knowledge repository for Energize C++.

"Foundation for a C++ Programming Environment", chapter 7, Protocols

https://pdfs.semanticscholar.org/84a4/8824fc7dd872414efa0ad6...

And how it looked like with their Emacs customization in 1993 on UNIX.

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


Lucid went bankrupt as far as I remember :)


And that proves what?

Sadly the best ideas aren't always the ones that get widespread adoption.

Here we are now, about 30 years later still trying to get major C++ environments to offer some of those ideas.


It proves nothing, it's just about my original comment about Microsoft:

> bringing the idea of language servers to the programming mainstream

Lucid failed to do this, Microsoft succeeded.


The simplicity of you statement ignores what it meant to get a computer running Energize C++ properly in 1993 and an Electron (VSCode) app in 2017.

Specially the money part.


The Microsoft Tablet PCs and the Apple Newton failed because the hardware wasn't there yet. C'est la vie...


Just financially, but never intellectually at least. ;)




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: