Comment by lycopodiopsida | Hacker News Reader

Comment by lycopodiopsida | Hacker News Reader

Parent Full thread

For a text editor mediocricity is fine. It has to solve many tasks for many people. It has to be shown that structural editing makes an editor simpler, faster and in suited for a broader range of tasks. Until then, text is the king.

> Yeah, but is it even good?

Yes, it is. Is the only solution which was able to replace several outlines + highly specialized solutions like OmniFocus for me. And that, being a simple text outliner. It has no point to compare it to some hypothetical solution which does not exist and may never exist at all. We need it here and now.

> and that's written in lisp.

It may be more of an artefact of the time when emacs was created than some intrinsic advantage of lisp in general. The way I see it, the advantage of emacs is that it is mostly written in the own script language. Would it be written in lua instead (a more common choice these days) not much would change as long as you can eval code on the fly.

> Or are you saying that Common Lisp itself is a problem?

I would not say that CL itself is a problem, but on the other side the lisp community likes to write essays about the perfect software in the sky but delivers mostly nothing. I cannot do otherwise than compare it rust, which is very young language compared to CL, but the community cranks out amazing and useful software on a weekly basis.

So, yeah, while in theory nothing speaks against amazing software being written in CL (as in any other language apart from brainfuck, really) in praxis it raises a lot of red flags for me.

> For a text editor mediocricity is fine.

No, not for the text editor, but may be for some users. If a tree falls, does it make a sound?

After all, maybe you are just not a power user and don't have the same frame of reference that power users have.

> It has to be shown that structural editing makes an editor simpler, faster and in suited for a broader range of tasks. Until then, text is the king.

Text might be the king until then, yes, but so is C++ and the children running with AKs in Africa. Doesn't mean it's right or acceptable.

To me, I have argued extensively, that yes, structural can be absolutely simpler and faster and more versatile. Without a question.

> I would not say that CL itself is a problem, but on the other side the lisp community likes to write essays about the perfect software in the sky but delivers mostly nothing.

I am sorry, but I am not a part of that generation, if there's one. Yes, I can see how strange it is to not see what's claimed to be a very powerful language overtake the world. I agree on that, it's a bit mysterious.

But some things just take time, I think. For the people to get accustomed to it; to go through some pain first, perhaps.

Look, the only reason I am using CL is because it's an image-based language. That's simply crucial for interactivity. What are my other options? Smalltalk isn't bad at all, but I like macros too much, and the stability too.

> in praxis it raises a lot of red flags for me.

You have to account that the community hasn't been as nearly as large as for any other mainstream programming language. So, aren't your expectation, perhaps, a bit too high for the output of that community? It only has something like 3k packages online. Python has 420000+. Where is all the amazing python userland software that has lived for >5 years?

And, you know, as small as it is, people have published some pretty cool libraries. Stable libraries. Sublanguages. Macro libraries. FWIW it's pretty cool as it is. A game was published two days ago: https://store.steampowered.com/app/1261430/Kandria/

Is it amazing? Haven't played it, don't know. Maybe? But what exactly is all that software you are expecting to begin with?

I don't know the name for it, but there ought to be some cognitive bias that says: "things that don't happen, can't, and won't".

Of course, I get it. I would be sceptical too. Maybe even very sceptical. But there's no other way to deal with sceptisism other than to analyze what's actually proposed, and whether it's doable.

https://project-mage.org/why-common-lisp

> After all, maybe you are just not a power user and don't have the same frame of reference that power users have.

Maybe. Or maybe your ideal of an editor has nothing to do with what everyone else wants from it, including power users.

> Text might be the king until then, yes, but so is C++ and the children running with AKs in Africa. Doesn't mean it's right or acceptable.

I will ignore the "but the children" polemics, because it is true for C++ and that is what Rust shows. The important part is "shows" - rants do not. In the end, talking is cheap - show me a better editor than emacs first, because a promise cannot edit text.

> Yes, I can see how strange it is to not see what's claimed to be a very powerful language overtake the world. I agree on that, it's a bit mysterious.

Yeah, indeed, if you have a wrong premise about a language being "very powerful" per se and refuse to re-evaluate it, then it is "strange". Or you accept that the powerful parts of lisp do not matter in the end, the disadvantages are real and that it comes from an era with little collaboration and small-scale software in general. Then you shrug and move to, you know, actually delivering something.

> Is it amazing? Haven't played it, don't know. Maybe? But what exactly is all that software you are expecting to begin with?

Ah, the inevitable CL debate :) You can find a bunch of libraries, half-written and abandoned decades ago, but the fact is also - I have software written in rust on my mac right now, and be it the inevitable RipGrep, but I cannot remember ever using or having something installed in CL. And it will be true for most people. It is a pretty much dead language which is, on the other side, made undead by claims about some perceived PL-superiority which does not result in any significant number of useful software being written. When you stop drinking the lisp-superiority-coolaid, not much remains to see, such is my perception after some diving into CL and Clojure.

Alright, I get it, you don't like CL. Coolio.

> show me a better editor than emacs first

I will. Just one correction: it's not going to be an editor. It's going to be many editors forming a cohesive whole.

Rune: Rune Is Not an Editor.

Other than that, what can I say? Yes, I have to prove my words, even though, in my opinion, the writings have enough to convince you that it can be done. And I will do my best to deliver.

I wish you to succeed, of course :)

Thank you : )

> I cannot remember ever using or having something installed in CL

I get you, of course. There are not a lot of user-facing, open-source software. When there is, they tend to be big and not CLI utilities like ripgrep, and not well advertised. A few ideas: pgloader (from python to CL), pgchart, OpenMusic, Opus Modus, the Scorecloud app, CEPL to play with graphics, regex-coach for windows, the Lem editor, Maxima, Axiom, Sucle (Minecraft clone), PTC's Creo Elements multi-million lines 3D CAD software with a free version, StumpWM…

but also you are ignoring all the big historical projects (some still in activity like Maxima).

Try https://ballish.margaine.com/ for a fast code search.

Reply on news.ycombinator.com