There's a JavaScript arms race going on (as I'm sure many of you are aware). The competing JS engines of the major browsers are leap-frogging each other in performance every few months it seems.

http://www.conceivablytech.com/4472/products/chrome-10-posts-huge-performance-jump/

It would be so cool (and a huge showcase for the D Programming Language) if DMDScript was in that performance race and beating the big guns.

Possible?

At the moment I do not think it is possible for DMDScript to compete. It needs a major rewrite.

node.js is going to win, it uses Google's V8, but its really not important which JS compiler is uses, it could use any. node.js is going to become a replacement for PHP and will also be Javascript's version of what Perl's CPAN, Ruby gems and Python eggs do. Microsoft is contracting out to Joyent to port node.js to windows, this should bring about a big jump in Javascripts use as a multipurpose scripting language.

Brad,

I've been a long long time fan of JavaScript, and have done alot with programming the language and discussing it's features and programming.

I like its C notation, library simplicity and its prototypical inheritance model.

JavaScript is terrible with Strings, almost as bad as VB.  I think this is a very serious limitation as processors keep getting better at numbers and never improve on strings, making them a major bottleneck.  I really like D's slice model.

The language's dual meaning for the '+' operator is also a major problem preventing some rather serious optimizations; since we can't predict what needs to be done with 'a+b' in code, we can't compile it without including branching for types AND type coercion.

Also, without the '+' concat operator, you could literally transcribe all the math
operators down to 64-bit double math operations in machine code, as long as you
could guarantee the variable names and/or properties didn't move without you noticing.

Also, the Date library is horrid.

Further, JavaScript needs native parallelization, not some tack on like Node.js; and a standard file interface that hides the FS details and also allows a dynamic equivalent to import.

So what really needs to be out there, is a hybrid of Lua, V8, and D.

On Friday, 23 December 2011 at 18:19:49 UTC, Daniel wrote:
> So what really needs to be out there, is a hybrid of Lua, V8, and D.

Maybe more like a jit for D? I use JavaScript a lot, and love it for what it is (and forgive it for what it's not). That said, D has the potential to be a major web language, but needs to have a Node.js like web server framework that can compile and cache code on the fly.

Even just patching in a way to write D modules for Node.js could be brilliant. Writing C/C++ extensions for PHP was always a nightmare and even though over time there ended up being a really large number of extensions available, most people didn't utilize that functionality for their own projects. D could (should) be the go-to serious system-level problem solving tool for web applications, but it needs some integration work so web developers don't have to start from scratch to build an app.

- Trevor

Vibe,d is the D version of node.js, check it out here ...
http://vibed.org/