Re: Vast SiteREF: Performance Difference between Firefox, Chrome and Safari

Message has been deleted

Rav

unread,

Oct 28, 2012, 3:11:39 PM10/28/12

to

On 10/28/2012 2:42 PM, Sailfish wrote:
> REF: http://mrdoob.com/lab/javascript/threejs/css3d/periodictable/
>
> Environment: i7 2600K, 8GB RAM, ATI HD 6800 series graphics card
> Browsers: Fx 17.0b3, Chrome 22.0.1229.94, Safari 5.1.7
>
> This referenced site is dog slow (even with Hardware Acceleration
> enabled) using Firefox (and even crashes Firefox when I select the GRID
> option.) However, the site works swimmingly when using either Chrome or
> Safari.
>
> I suspect it's a Javascript issue but, honestly, the performance
> difference seems to be an order of magnitude in difference, at least.
>
> Ideas?
>

Slow for me too, and crashed for me too when I selected GRID (crash
submitted to Mozilla). Fx 16.0.2, Windows XP SP3.

goodwin

unread,

Oct 28, 2012, 3:30:26 PM10/28/12

to

On 10/28/2012 11:42 AM, Sailfish wrote:
> REF: http://mrdoob.com/lab/javascript/threejs/css3d/periodictable/
>
> Environment: i7 2600K, 8GB RAM, ATI HD 6800 series graphics card
> Browsers: Fx 17.0b3, Chrome 22.0.1229.94, Safari 5.1.7
>
> This referenced site is dog slow (even with Hardware Acceleration
> enabled) using Firefox (and even crashes Firefox when I select the GRID
> option.) However, the site works swimmingly when using either Chrome or
> Safari.
>
> I suspect it's a Javascript issue but, honestly, the performance
> difference seems to be an order of magnitude in difference, at least.
>
> Ideas?
>

Sorry for the long URL

> http://validator.w3.org/check?uri=http%3A%2F%2Fmrdoob.com%2Flab%2Fjavascript%2Fthreejs%2Fcss3d%2Fperiodictable%2F&charset=%28detect+automatically%29&doctype=HTML+4.01+Strict&group=0&user-agent=W3C_Validator%2F1.3

Checking for HTML5 yields fewer errors.

or just bad chemistry? :)

WaltS

unread,

Oct 28, 2012, 3:36:52 PM10/28/12

to

On 10/28/2012 02:42 PM, Sailfish wrote:
> REF: http://mrdoob.com/lab/javascript/threejs/css3d/periodictable/
>
> Environment: i7 2600K, 8GB RAM, ATI HD 6800 series graphics card
> Browsers: Fx 17.0b3, Chrome 22.0.1229.94, Safari 5.1.7
>
> This referenced site is dog slow (even with Hardware Acceleration
> enabled) using Firefox (and even crashes Firefox when I select the GRID
> option.) However, the site works swimmingly when using either Chrome or
> Safari.
>
> I suspect it's a Javascript issue but, honestly, the performance
> difference seems to be an order of magnitude in difference, at least.
>
> Ideas?
>

Not really.

One "Unclosed div element" at Line 124, Column 17, and "End tag for body
seen, but there were unclosed elements" at Line 489, Column 8.

Darn Firefox for being so standards compliant.

Didn't work at all in Opera.

No crashes here. Also slow in Nightly.

--
Fedora 17 (64-bit)
Thunderbird Beta (17.0) Install and test it.
One state should not determine an election.
http://www.nationalpopularvote.com/

WaltS

unread,

Oct 28, 2012, 3:47:09 PM10/28/12

to

On 10/28/2012 03:36 PM, WaltS wrote:
> On 10/28/2012 02:42 PM, Sailfish wrote:
>> REF: http://mrdoob.com/lab/javascript/threejs/css3d/periodictable/
>>
>> Environment: i7 2600K, 8GB RAM, ATI HD 6800 series graphics card
>> Browsers: Fx 17.0b3, Chrome 22.0.1229.94, Safari 5.1.7
>>
>> This referenced site is dog slow (even with Hardware Acceleration
>> enabled) using Firefox (and even crashes Firefox when I select the GRID
>> option.) However, the site works swimmingly when using either Chrome or
>> Safari.
>>
>> I suspect it's a Javascript issue but, honestly, the performance
>> difference seems to be an order of magnitude in difference, at least.
>>
>> Ideas?
>>
>
> Not really.
>
> One "Unclosed div element" at Line 124, Column 17, and "End tag for body
> seen, but there were unclosed elements" at Line 489, Column 8.
>
> Darn Firefox for being so standards compliant.
>
> Didn't work at all in Opera.
>
> No crashes here. Also slow in Nightly.
>

Since it was written with the HTML5 doctype <!DOCTYPE html>. I tested it
as HTML5 at http://validator.w3.org/

Message has been deleted

WaltS

unread,

Oct 28, 2012, 4:18:22 PM10/28/12

to

On 10/28/2012 04:06 PM, Sailfish wrote:
> My bloviated meandering follows what WaltS graced us with on 10/28/2012
> 12:36 PM:

>> On 10/28/2012 02:42 PM, Sailfish wrote:
>>> REF: http://mrdoob.com/lab/javascript/threejs/css3d/periodictable/
>>>
>>> Environment: i7 2600K, 8GB RAM, ATI HD 6800 series graphics card
>>> Browsers: Fx 17.0b3, Chrome 22.0.1229.94, Safari 5.1.7
>>>
>>> This referenced site is dog slow (even with Hardware Acceleration
>>> enabled) using Firefox (and even crashes Firefox when I select the GRID
>>> option.) However, the site works swimmingly when using either Chrome or
>>> Safari.
>>>
>>> I suspect it's a Javascript issue but, honestly, the performance
>>> difference seems to be an order of magnitude in difference, at least.
>>>
>>> Ideas?
>>>
>>
>> Not really.
>>
>> One "Unclosed div element" at Line 124, Column 17, and "End tag for body
>> seen, but there were unclosed elements" at Line 489, Column 8.
>>
>> Darn Firefox for being so standards compliant.
>>
>> Didn't work at all in Opera.
>>
>> No crashes here. Also slow in Nightly.
>>

> I saved it to the desktop and corrected those minor errors as follows:
>
> o Inserted "</div>" after line 128
> o Removed </div> in last line "</div></body></html>"
>
> Validated fine. No change in performance.
>

Well then that wasn't the reason.

Message has been deleted

Ron Hunter

unread,

Oct 29, 2012, 2:39:37 AM10/29/12

to

On 10/28/2012 1:42 PM, Sailfish wrote:
> REF: http://mrdoob.com/lab/javascript/threejs/css3d/periodictable/
>
> Environment: i7 2600K, 8GB RAM, ATI HD 6800 series graphics card
> Browsers: Fx 17.0b3, Chrome 22.0.1229.94, Safari 5.1.7
>
> This referenced site is dog slow (even with Hardware Acceleration
> enabled) using Firefox (and even crashes Firefox when I select the GRID
> option.) However, the site works swimmingly when using either Chrome or
> Safari.
>
> I suspect it's a Javascript issue but, honestly, the performance
> difference seems to be an order of magnitude in difference, at least.
>
> Ideas?
>

I see the same, except it doesn't crash on 'grid' here. Definitely
something going on with the Javascript.

Ron Hunter

unread,

Oct 29, 2012, 2:42:28 AM10/29/12

to

It is an amusing display, but hardly useful. I can't see any benefit in
having a spherical or helical display of the elements. The table
reveals some interesting information regarding common properties.

Message has been deleted

Ron Hunter

unread,

Oct 29, 2012, 11:16:50 AM10/29/12

to

On 10/29/2012 3:12 AM, Sailfish wrote:
> My bloviated meandering follows what Ron Hunter graced us with on
> 10/28/2012 11:39 PM:

> Yeah, the difference is so dramatic, it makes me question all of the
> recent proclamations about the strides Mozilla has made in Javascript
> performance improvements.
>
> With the number of confirmation here, I've filed a bug:
> https://bugzilla.mozilla.org/show_bug.cgi?id=806295
>
Is it Javascript, or HTML5? At least Firefox manages to make a try at
it, unlike Opera, and IE9, which do NOTHING.

WaltS

unread,

Oct 29, 2012, 11:33:03 AM10/29/12

to

The Doctype used is the HTML5 Doctype, with the content in the Body
element mostly JavaScript.

Ron Hunter

unread,

Oct 29, 2012, 1:26:56 PM10/29/12

to

That may lead to a solution. Maybe the two functions aren't
communicating well, or playing nice with the processor time.

Message has been deleted