@Garrett Johnson:
On your "es-harmony-part-one" post I think you should stress that
while JavaScript 1.8 (since 2007), currently 1.8.2 (since 2009), may
have influenced Harmony it is a dialect of ECMAScript developed by
Mozilla.
https://developer.mozilla.org/en/New_in_JavaScript_1.8
Also on your
@Garrett Johnson:
I noticed this in your summary of what Harmony is:
"It will ultimately form what will become sixth edition of ES."
Check out this post on how /be describes Harmony:
https://lists.webkit.org/pipermail/webkit-dev/2011-December/018914.html
- JDD
--
To view archived discussions f
> > I tried in Chrome and deleting just doesnt work...
> > Deleting hasOwnProperty just... deletes it
>
> Ack, yes, my bad. That only works for the properties on, e.g.,
> Object.prototype, not the ones on Object itself.
It doesn't work in Chrome on properties of Object.prototype either:
http://js
> In some browsers, Safari and Chrome at least, deleting the bad value will
> reveal the original value again.
Just to clarify this special behavior does *not* happen in Chrome (old
- canary) or Safari < 5.
- JDD
--
To view archived discussions from the original JSMentors Mailman list:
http:/
> Really? I'm asking because I didn't actually test that and it
> surprisese me. If the TABLE contains a TR then a TBODY must be created
> implicitly. According to DOM spec.
Yap, that's what the tests suggest.
What's interesting is that though some browsers allow setting the
innerHTML of a table
@Lasse Reichstein
> The test checks whether the value set as innerHTML is the same as
> the one read back afterwards. There is nothing that suggests that
> this must be the case. The formatting of the values read from innerHTML
> is not specified. E.g., if there is an element with multiple attribut
> You've given a test, in singular, which is not the same as test*s*,
I gave 5 tests (colgroup, optgroup, select, scripts, tables).
> table created with a tbody, that can hardly be considered a bug, when tables
> have been being created automatically with tbodys for ages
The spec says they c
@Jorge We have given you tests, plz troll elsewhere.
--
To view archived discussions from the original JSMentors Mailman list:
http://www.mail-archive.com/jsmentors@jsmentors.com/
To search via a non-Google archive, visit here:
http://www.mail-archive.com/jsmentors@googlegroups.com/
To unsubs
@Jorge:
Shows Chrome 11 and Safari 5 failing at least 1 of the tests:
http://jsbin.com/ipowa4
> LOL, March, 7, 2007 ? That must have been Safari 2.x :-))) Are you
The poster said that the bug was old but still exists (maybe in
mobile).
- JDD
--
To view archived discussions from the origin
> That url gives me "Sorry, no posts matched your criteria."
>
Remove the trailing `-` from the url.
http://bit.ly/eDSgvi
-JDD
--
To view archived discussions from the original JSMentors Mailman list:
http://www.mail-archive.com/jsmentors@jsmentors.com/
To search via a non-Google archive, vis
> But I wanted to point out that unlike in IEs there's not any problem(s) with
> innerHTML in Chrome nor in Safari nor in Firefox nor in Opera. Only in
> Microsoft's Internet Explorer(s).
Not so, while working on FuseJS (I also see checks in jQuery) I
noticed several browsers, not just IE, have
@Adrian Olaru: To reiterate what Peter and Miller said, these kind are
differences are insignificant.
I remember running across this jsperf test a few days ago.
The results are indistinguishable from each other and are of such a
high ops/sec rate that any minor difference is overlooked.
http://jsp
> I wasn't reading closely enough. I see you add the div if the
> documentElement.clientWidth is not 0.Though it still has the effect of
> making the scrollHeight 1 in that document. I'm not sure why that
> is.
My detection method was actually part of another feature test looking
for the root
I took a combo approach on a recent project:
https://github.com/mathiasbynens/benchmark.js/blob/master/benchmark.js#L81
// I allow `fn` because test `fn` is required
new Benchmark(fn);
// or a name first because other devs require a name too (an options
object for just a name is bulky) and...
new
I remember @kangax made a bookmarklet to do this that uses an iframe's
untouched global for comparisons.
https://github.com/kangax/detect-global
--
To view archived discussions from the original JSMentors Mailman list:
http://www.mail-archive.com/jsmentors@jsmentors.com/
To search via a non-Goo
> Yes, I modified the test to switch between standards and quirks and it
> leaks on both.
I meant, no.
No difference between quirks and standards mode.
They both leak.
--
To view archived discussions from the original JSMentors Mailman list:
http://www.mail-archive.com/jsmentors@jsmentors.com/
> Yep. But is there a different result in standards mode?
>
> That test case is in quirks mode.
Yes, I modified the test to switch between standards and quirks and it
leaks on both.
- JDD
--
To view archived discussions from the original JSMentors Mailman list:
http://www.mail-archive.com/jsm
> I'm busy. Misinformation's no good right? What should I do?
No worries. I am a night owl and created a test and a screencast.
http://screenr.com/UBz
http://dl.dropbox.com/u/513327/ieleak/test.html
So the leak happens with a meta refresh, manual refresh, and
navigating between multiple pages.
@Garrett - I would rather you present a test case where it doesn't
work.
Asking someone to do a web search is kind of a copout.
So far you have just presented doubt based on nothing (no test case),
while others have presented working test cases with reproducible
results.
-JDD
--
To view archived
The reduced test case (using a single page refresh) is fine.
It shows evidence of the memory leak and it's removal.
It is understood that users don't refresh a page over and over (though
it might happen with sites like ebay),
and will more likely be visiting a site that uses the script on every
pag
Hiya,
Instead of having `noodles` return a new manually augmented object you
can have it utilize noodles.prototype
https://github.com/fitzgen/noodles/blob/master/noodles.js#L47-56
// private
function Klass() { }
noodles = exports.noodles = function (items) {
var instance = new Klass;
instanc
@Daniel Donaldson
Assuming the content of '.take li' is only text and you are working
with an HTML (not XHTML) doc you could get away with:
$('.take li').click(function() {
// I avoided using innerText or textContent for simplicity
$('.place').append(this.innerHTML);
});
If you're us
> It has a long history and wide support, but I'm not sure I think it's
> necessarily that great.
It provides a functionality that is not matched by any existing
proposal or implementation.
I gave one example of a creative use for it that helps solve a
problem, augmenting native prototypes,
that
> But I've heard Brenden Eich doesn't think __proto__ is a good idea anymore.
__proto__ is great, has a long history of support (at least 13yrs),
and is now supported by every major browser (Safari, Chrome, Firefox,
Opera), except IE.
It's also available in server-side JS enviros like Narwhal, Ri
> But I've heard Brenden Eich doesn't think __proto__ is a good idea anymore.
__proto__ is great, has a long history of support (at least 13yrs),
and is now supported by every major browser (Safari, Chrome, Firefox,
Opera), except IE.
It's also available in server-side JS enviros like Narwhal, Ri
25 matches
Mail list logo