Is Clojure the next C ?

[Armando]

Interesting article on software vs hardware and... language "evolution".
The Barbarians are at the Gates

[Eóin]

Let's see

• He applies a human linguistic issue to computer languages without ever justifying that it is applicable.
• He keeps going on about us being "trapped" without every really justifying that we are or what the trap is.
• He then proposes a "solution" without any really explanation of why it will solve things other saying "it's different".

I suppose it's an interesting diary entry, but it's not much of a supported argument.

[f0dder]

+1, Eóin

I also find he trivializes that changes that has been in software development over the years. Yes, there's probably some magnitudes more complexity in the hardware engineering field, but a lot has been happening in software engineering as well.

And then there's the premise that we have a problem. Really? Most users aren't going to need a fraction of the computing power a modern CPU offers. Handling massive loads of users on a web backend is a completely different programming task than running desktop software - a 2048-core laptop sounds plain old lame to me.

[Armando]

Thanks for your insight guys ! Much appreciated.

[wraith808]

a 2048-core laptop sounds plain old lame to me.

And 640K of memory is all anyone will ever need. 

[f0dder]

a 2048-core laptop sounds plain old lame to me.

And 640K of memory is all anyone will ever need. 

Stop perpetuating that false quote

Also, that situation today is vastly different from how things were back then. Yes, we have scalability issues on the back-end side of things that need to be fixed, but we're pretty well off on the end-user side of things, apart from a (relatively) few specialized niches.

[wraith808]

Stop perpetuating that false quote

Yes, I know it's false, but it's never going away, so why not use it... it's hilarious!

Also, that situation today is vastly different from how things were back then. Yes, we have scalability issues on the back-end side of things that need to be fixed, but we're pretty well off on the end-user side of things, apart from a (relatively) few specialized niches.

Well, when you consider that one of the specialized niches is pr0n (VR)... and everyone knows that the internet is for porn...

[Eóin]

It's worth noting that the new highly parallel future doesn't automatically need a new language. Indeed the current approach to utilizing the extra power offered by GPUs and the like is to stick with the old bare metal languages, e.g. CUDA and OpenCL are basically C programming.

[Renegade]

Meh... Not sure.

Software languages (or many at least) have innovated with structures and logic expressions, but I think he's trying to point to something else. What that is? Well, not entirely sure. I think it's that lack of innovation/paradigm shift that he's talking about, and without it, it's kind of hard to point to.

At the heart of it, it seems to me that he wants a hardware agnostic language that uses all available resources.

For his comment:

The Sapir-Whorf hypothesis tells us that our view of the world is strongly affected by the languages we use. When we think in a given language, that language acts as a filter. Concepts it can’t express are removed from our awareness. Our mode of expression constrains us to only those thoughts and concepts that can easily be expressed within it.

Seems pretty reasonable to me. Seems to work in both human and machine languages from what I've seen.

Dunno. I don't need to really worry about any of that stuff. I'm just a small fry. It is interesting though.

[Tuxman]

"The next C"? What's wrong with the existing C?

[Renegade]

"The next C"? What's wrong with the existing C?

Hehehe~!

It seems like language enthusiasts are always looking for a new language for one reason or another.

[Tuxman]

Oh, yeah, like, "ok, about 90% of all common applications are written in C or a derivative, that's boooooring!"?

[Armando]

"The next C"? What's wrong with the existing C?

Nothing, really. 1- It's just a title. 2- "Next" doesn't imply that "previous" is wrong.

[Armando]

I think it's that lack of innovation/paradigm shift that he's talking about, and without it, it's kind of hard to point to.

At the heart of it, it seems to me that he wants a hardware agnostic language that uses all available resources.

That's exactly what I found "interesting" in this "article". Basically : to what extent languages are / need to be projections of the hardware's own structure ? (By nature, certainly more than pure mathematical language(s).) And how much these ties between languages and hardware actually constrain the languages' own potential...? Are the author's thoughts mere unfounded speculations ? Maybe.

There's a definite exchange/mutual "influence" between languages (in the general sense of systematic means of communicating) and media, which in turn changes the way we think, perceive and understand the world around us. It certainly happens in the computer world.

In any case, I don't know clojure/lisp enough to be able to know how valid or legitimate the author's hypothesis/views are. Is Lisp/Clojure data types, control structures, mathematical like flavor etc. really an advantage ?

That's why I posted here -- I knew I'd get informed opinions.

Simple curiosity on my part. Nothing too serious.