Re: New Module Proposal. libseed

[Xan Lopez <xan gnome org>]

On Thu, May 14, 2009 at 4:05 PM, Havoc Pennington
<havoc pennington gmail com> wrote:
>> So perhaps it would be a good idea to just stick to a JS defined in
>> some standard widely used for all GNOME code, in order to avoid future
>> headaches, and consider other languages with real self-extension
>> capabilities if we are really serious about "using whatever dialects
>> make our life easier" (<mandatory plug for the Lisp family of
>> languages>).
>
> The problem is that JS-with-a-few-basic-enhancements is just _so_ much
> better than least-common-denominator-web-JS. There's no need to go
> down a slippery slope of a million enhancements, just to fix some
> basic stuff... like variable scoping. Web JS is why people think "ugh,
> JavaScript"

While this might be true (although personally I remember 'let' more
like "that's a nice improvement" more than "OMG that makes the
language usable") I think that as long as these are vendor-specific
extensions it's risky to go that path unless the benefits *really*
make a difference. Today it's JSC vs SpiderMonkey, but in 3 years
there might be two more good engines, and perhaps we'd like to move to
one of them. I guess in the end it's just a matter of deciding if
those extensions are worth locking yourself in a particular
implementation of the language or not.

Owen has said that he'd only really miss destructuring assignment I
think, your opinion is that 'let' is a deal breaker?

Xan