Why Ruby is an acceptable Lisp (2005) | Hacker News Reader

Why Ruby is an acceptable Lisp (2005)

randomhacks.net

27 pointsd0mine31 comments

I don't see how Ruby could be called a Lisp. To me, the fundamental concept of Lisp is that the basic data structure is the cons-pair, and that both code and data can be represented and manipulated as graphs of cons-pairs.

The blog is arguing that Ruby is an acceptable alternative to Lisp, which may be true. But that doesn't make Ruby a Lisp.

The crux of the post seems to be that because Ruby has lambdas, it is an acceptable alternative.

In my experience, being that I'm learning Haskell and coming from a Ruby background, merely having lambdas doesn't qualify for functional status.

Well said. The latest version of PHP is going to have lambdas, and I'm quite certain that it's not a Lisp.

Its lambdas don't even close over lexical state, so I'm not sure why they even exist. Yet another concept the PHP designers don't get, I guess, but feel pressured to add. (See also: PHP's object system.)

He's not arguing that ruby is a Lisp, just that it gives most of the power of Lisp in a syntax closer to what most programmers are familiar with.

You're right, but he shouldn't have titled it "Why Ruby is an acceptable Lisp".

Interestingly enough, one of Clojure's selling points is that the basic data structure is not the cons-pair (rather the implementation details of storage are abstracted). It is clearly a lisp, however.

This is so wrong in so many ways, but I think the ur-mistake, the one that leads to all the others, is writing an article like this in the first place. You like Ruby? You like Ruby. Use it! Why write a whole article trying to justify how you feel? It's as if the author is trying to give himself (and others) permission not to use Lisp.

Suppose you're right. He is trying to give himself permission to not use lisp, so he feels better. Then writing this helps him use ruby (as you advise). so it's good that he wrote it. (but no need for you to read it).

Very clever, except that you miss the point, which is that he has to concoct a bunch of bogus arguments in the process, and those are not so good.

Oh, that's bad then.

He is saying something. Something false, I believe, but not meaningless. There is a certain style of programming that was once only possible in Lisp. The author is claiming that you can program that way in Ruby, or close enough.

Oh, I don't think it's meaningless, just that the arguments are bad, and seem to have been invented in order to justify something that doesn't really need justifying.

Now I suppose someone is going to ask me what arguments are bad. Two examples: 1. "The most common use of LISP macros is to avoid typing lambda quite so much" - obviously false; 2. Using Lisp to model a Ruby example of "mini-language" demonstrates Lisp's power, not Ruby's.

It works both ways. Don't you like Ruby? You don't like Ruby. Don't read articles about Ruby!

I for one may use Ruby as a scripting language for my next project, I want to see more articles like this.

Don't read articles about Ruby!

I usually don't! This is an article about Ruby and "LISP". I read articles about Lisp. This one is mistaken, I think, mostly because the author hasn't used Lisp to do much, but writes about it anyway.

But I really don't mean to carp about the article. Ruby makes him happy, that's great.

I see your point. But, please, let's encourage people to write more articles about what they think is right. Even if the article is wrong it's still good that it was written.

I think the definite false part is 'ruby is a[n acceptable] Lisp' - which is false. One could have a language with a ruby like syntax and co-routines and be a Lisp.

One could probably argue that Perl6 is a Lisp with CFGs a first order language primitive (which is in essence what macros allow), c.f. Greenspun's Tenth.

Stating the obvious: there's a difference between "provides much of the value of in practice" and "being". The same logic says that Esperanto is an acceptable English.

For the admins - Dup -

http://news.ycombinator.com/item?id=98438

Thanks. I think 352 days is an excusable time for the repost.

No offense but I think a great deal of hacker news fans have seen this one way or the other. I remember there was response to this called lisp is an acceptable lisp. If you want lisp and are using ruby why not just use lisp? If you want ruby but use lisp just use ruby. It doesn't matter which programming language you express your ideas in. It only matters if you actually get cracking and implement your ideas. Pointless saber rattling about programming languages is quaint and dandy if you aren't actually programming. To quote feynmann "If I were forced to sum up in one sentence what the Copenhagen interpretation says to me, it would be 'Shut up and calculate!"

http://steve-yegge.blogspot.com/2006/04/lisp-is-not-acceptab... ?

Yea I agree with yegge. I think that lisp is what you make of it. Language doesn't matter.

I don't think that's actually Yegge's point, is it?

I misread yegge. lisp isn't a popular language its more of a language where you either use for industry or you use for experimentation or both. I personally think it doesn't matter what you code in blub or not blub it just doesn't matter. What matters is if you ship or not.

Humm... I still think the problems with the platform are about an order of magnitude greater then the problems with the language per se. You first need to have something to work with, and only after that you can think about improving it.

Reading the comments on that blog entry, I can't help but think Lisp is Blub. The Lisp users posting there simply can't comprehend that anything is wrong and have no tolerance for anyone who says anything is.

The lisp vs foo is somewhat special. An accessible, easy-to-learn lisp platform with a full complement of libraries is pretty much a bit of a Holy Grail.

edit: Ups, read the "Lisp is not an acceptable Lisp". Makes the same point...

An accessible, easy-to-learn lisp platform with a full complement of libraries

PLT Scheme? Clojure?

Even CL has some of the best libraries I've used, although coverage can be spotty at times.

This still hasn't killed all the other programming languages, which leads me to believe that libraries aren't actually what matters. I think, instead, that most people pick programming languages based on what's crammed down their throats. For a UNIX app, you "have to" use C. For everything else, you "have to" use Java, since that's what you learned at school (now it's .NET too).

Popularity is overrated anyway. Haskell and SBCL are not exactly popular, but they have great runtimes, important modern features, and plenty of libraries and tools. The only reason not to use them is because you'd rather use Java instead.

Why does a language make such an dichotomy. Why not instead learn tools and apply it in the languages that you know. Use what you know not the dogma of the other languages.

I'm still trying to decide if ruby is an acceptable OOPL. Sure, the language itself is, but the way people use it? not always so.

So far as the idea "Ruby is an acceptable Lisp"? This is as wrong as the idea "Time is Money".

Reply on news.ycombinator.com