Lisp-style Macros for Ruby | Hacker News Reader

Lisp-style Macros for Ruby

github.com

29 pointsraganwald12 comments

I wish matz didn't remove macros in the first place.

I really think that when a language designer is contemplating removing features, they should be mindful of the inevitability of people greenspunning those features right back in. There's a lot of value in making a language simpler and more elegant, and if the language gives you great tools for building stuff in people might surprise you by designing better implementations than what you intended.

But still, it's inevitable that anything remotely lisp-like will get sucked into Lisp's black hole.

Can you explain your last sentence? I don't understand it.

The Black Hole Theory of Design:

http://blogs.sun.com/jag/entry/the_black_hole_theory_of

Another way to put it is that Lisp might be a local maximum: It may not be the globally optimum language for a particular problem, but once you approach Lisp you find that improving your language requires either becoming more Lisp-like or making a huge leap away from Lisp (C, ML, Haskell, Smalltalk, Erlang, &c.)

I'm not sure he removed it, I think he just didn't want macros from the start as I recall.

He has said he took lisp and removed macros. But it could have just been conceptually, I don't know. Does anyone?

Matz was talking about concept, not implementation; MRI is not literally a Lisp interpreter with macros removed.

I've read that, but I still don't grasp it. Even though he calls Ruby "MatzLisp," it feels a lot more like "MatzSmalltalk" to me. Even without macros, the homoiconicity of Lisp makes eval-ing lists a completely different beast than eval-ing strings or blocks. And everything is an object, there are no functions. And there are magic syntactic constructs like while...

But I digress. I guess he did start with Lisp and remove macros :-)

the 'MatzLisp' thing was a joke. He never understood Lisp, Lisp was too complicated for him and wanted to do his own thing anyway.

I know the whole MatzLisp was a joke, but where are you getting the idea that he never grokked Lisp?

I'm pretty sure writing an interpreter is pretty complicated as well.

watch his keynote from some last Ruby conference. He described that he tried to learned Lisp and that he thinks it is too complicated. He is relatively clueless about it.

Interpreters exist in dozens. In the Lisp community it is a sport to write those. Scheme has more than a hundred implementations. Many of them are much more ambitious than Ruby.

The Ruby implementation technique is not very advanced - it is a very simple interpreter, hacked together. The Ruby language may be convenient to use for scripting, but from a language implementation standpoint it is far away from the state of the art. It shows. The Ruby implementation is about the slowest language implementation out there.

There are dozens of Lisp implementations that are much more advanced, include compilers, etc..

Sure, Ruby is simpler and its implementation is simpler - that's part of the reason of its success.

He was asked about macros at some recent RubyConf (maybe the one in Charlotte). He said (as I recall) that he was against having macros in Ruby because it would allow for too much variance in program syntax.

Seemed an odd argument from someone who defends other "dangerous" language features (e.g open classes) by simply saying that Ruby is a sharp knife and you just need to be careful with your tools.

Reply on news.ycombinator.com