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

I wonder if Arc is the tortoise to Clojure's hare.

Would you care to elaborate? Your assertion implies that arc is going to surpass clojure in popularity. Why do you think so? Personally I think that's unlikely to happen given the traction that clojure has been getting already. Clojure has been designed for practical use and is already being used in real world projects whereas I get the impression that arc is sort of a research project where Paul is just experimenting with some ideas and is not constrained by the desire to do practical work.

I didn't make any assertions. Clojure is wildly successful and follows a pattern of development that seems fruitful to emulate. But in the story, the tortoise does win.

I am always on the lookout for interesting stories about people using functional programming on the job. Are you currently using Arc to this end?

I don't use Arc at all.

ROFLMAO! I think we are taking your comments too seriously.

It's not a story. It's a paradox. The tortoise loses, but the paradox pitches logic against intuition and is thus educational, like a koan (from a different cultural tradition). I am not sure if it is applicable to Clojure (as well as Arc and lisp in general) where intuition follows logic rather than contradicts it.

Worse is better and in this case Clojure is even more so 'The zombie-reanimated corpse of Lisp'.


Besides being a Lisp, Arc does not exist anywhere near the the same territory as Clojure.

Lisp introduced us to: code == data

Clojure introduces us to: code == concurrent high-performance immutable data

Yes! The functional aspects of lisp have always made it feel more parallelizable, but outside of particular lisp dialects (*Lisp), effective parallelization has been difficult. Pervasive immutable data structures and front page concurrency mechanisms make clojure decidedly parallel.

(I hope that parallel being the only way to gain additional performance should be obvious by now.)

In a hundred years, is that something that will matter?

In a hundred years, will any of the stuff in Arc matter? What exactly? The innovation in Arc is minimal, Clojure does much more in new and better ways.

concurrent high-performance immutable data sounds pretty much like the fabric of reality. So I would hazard a YES :) In fact it'll probably matter in a million years from now.

"Multi-core is here to stay", says Rich Hickey

Hopefully not. Multi-core should be seen as a stepping stone on the way to better technology.

Why not have multi-core version of other tech, unless you are talking Quantum Computing, though isn't that in some ways the ultimate form of multi-core?

PG lists 9 points that made Lisp awesome. After using Clojure for a while, I would argue that a good approach to State, Identity & Value is point #10.

Arc's problem is that it is not the only tortoise in the race.

I wonder if Arc to Clojure isn't more like GNU Hurd to Linux.

Is there anything promising about Arc other than being PG's pet language?

I'm not knocking PG here - I just don't see Arc having any real differences from all the other lisp variants. There's nothing revolutionary about it. Which is fine - it is a 100-year language, not a revolutionary language.

I just wonder what the odds are that anyone remembers it in a decade. People use langauges that are unique, and as far as I can see, Arc is not all that unique (unless you're the sort of person who becomes highly excited about the minutia of Lisp semantics)

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact