I think it's significant, though admittedly minor, because readableJS is one of the goals of the project.
Speaking of readable JS, by far the biggest problem I encounter
is when PS bundles a complex expression into comma-delimited
Javascript like "(foo(),bar(),baz())" which is basically a progn
that evaluates to baz(). These things often create monstrous
one-liners. I don't have a good idea of what to do about it --
I recall that Vladimir tried to pretty-print them at one point
and this turned out harder than it seemed -- but it's definitely
the #1 thing that would prevent me from trying to persuade
a JS programmer that my compiled PS is readable.
On Thu, Jan 17, 2013 at 9:44 AM, Boris Smilga
<boris.smilga@gmail.com> wrote:
Also, there is this baddie:
On 16 Jan 2013, at 00:18, David Sargeant wrote:
(function () {
var collect66 = [];
for (var num = 10; num <= 1; num += 1) {
collect66['push'](num);
};
return collect66;
})();
Not at all a major issue, but collect66.push would fit rather better in this context. There are a couple more instances of this kind in the code of ParenScript, they all boil down to keywords being used with @ instead of literal symbols:
src/lib/ps-dom.lisp:6: `(@ ,el :inner-h-t-m-l))
src/lib/ps-dom.lisp:12: `((@ ,el :get-attribute) ,attr))
src/lib/ps-loop.lisp:102: (then (if (numberp by) `((@ ,var :slice) ,by) `(,by ,var))))
src/lib/ps-loop.lisp:167: (:collect `((@ ,var :push) ,item))
src/macros.lisp:177: (if (eql ,funobj (@ __PS_MV_REG :tag))
src/macros.lisp:178: (@ __PS_MV_REG :values)
src/non-cl.lisp:189: `((@ (list ,@things) :join) "")))
Do you think we should fix them, or is this totally insignificant?
— B. Smilga.