jQuery Forum

I think you can file tickets on <a href="http://dev.jquery.com">dev.jquery.com</a> and then upload patches, John Resig is now maintaining this plugin.<br clear="all">
Best regards,
Cloudream

Hi polyrhythmic,
It would be great, if you would share your improvements and bugfixes.
can you report the ticket number of your bug reports here, too?
There are some dependencies between the animateClass- and the color-
plugin. If you fix the borderColor-propterty all browsers except
firefox are trying to animate borderColor + borderLeftColor etc. This
can make the animation very sloppy.
You can find the branch @ http://jquery-ui.googlecode.com/svn/branches/dev/effects
regards
alex

Hello,
I've been wanting to properly separate MVC, and keep my CSS properties
in my classes and out of my JS, so I have been hacking away at the
class animation methods. I have fixes for the WebKit getComputedStyle
object bug and the IE6+ 'undefined' color animation bug.
Ticket 4324 (http://dev.jqueryui.com/ticket/4324) is where I attached
the file, and added a link in Ticket 4328 (http://dev.jqueryui.com/
ticket/4328), since it's the listed color animation 'critical bug'.
The color animation fix is based off of 1.6rc6 as I haven't found any
advantage in the 1.7 code. I'm not sure of the root cause of the
bugs, but I have fixed the symptoms.
I saw a thread on the jQuery dev list where Paul Bakaus prefers to
animate all possible CSS properties:
http://groups.google.com/group/jquery-dev/browse_thread/thread/fd4ce58916323e93
That is why I didn't use a whitelist (a la Scriptaculous), although
performance may suffer.
Code is tested in IE6 and 7, Firefox 3+, Safari 3+, Opera 9.5+, and
Chrome on XP and Vista. Mac and IE8 testing is needed. Performance
with multiple concurrent class animations in Opera 9.5+ is very bad.
Let me know if you have any questions! I'm a bit green at submitting
fixes so I'm open to suggestions.
Thanks,
Charles

Hi polyrhytmic,
thx for your patches. all these bugs were already fixed in the branch
version. I mentioned earlier (http://jquery-ui.googlecode.com/svn/
branches/dev/effects).
there are still things to do, before it can be go final. (get involved
@ http://groups.google.com/group/jquery-ui-dev/browse_thread/thread/6e28e0a8f9899380?hl=en)
About some of your bugfixes:
1. The scrollbarBase-fix (is NaN)
I don´t know, wether you are right here, but NaN is a fals-y value and
should already been filtered with the following line: && newStyle[n] /
* No functions and null properties */
2. margin & Dean Edwards styleLeft fix in curCSS
This is a bug in jQuery and should be handeld there. But your are not
realy accurate here. The same bug appears, with other properties like
border-width, padding, background-position etc., too.
The best way to workaround this issue is to make the number-parsing
more accurate: ((!isNaN(parseInt(opts[i].newStyle[n], 10)) && opts
[i].newStyle[n].match(/\s|,/) === null) // is a number and has no
whitespaces
3. Your color-animation fix
This is also an jQuery bug. (Sometimes in IE and especially) if
$.fx.off is set to true, the fx.state property is not initially 0. A
workaround that is more oriented to this problem can look like the
follwoing lines:
if ( !fx.colorsInitialized ) {
    fx.start = getColor( fx.elem, attr );
    fx.end = getRGB( fx.end );
    fx.colorsInitialized = true;
}
It would be great, if you could test the branch script with your demo
(http://jquery-ui.googlecode.com/svn/branches/dev/effects/ui/
effects.core.js). I´m also interested in performance issues (you
mentioned something with opera).

Hello Alexander,
Glad to see these bugs are being solved. I have setup a test page
using my demo and the current branch version @
http://www.doublerebel.com/index.demo-branchver.php
Great work, I get the expected behavior in Firefox and Safari.
Internet Explorer has some type of margin, padding, or height bug, as
can be seen by comparing the demo page in IE to the demo page in
Firefox/Safari. Also, the bug/reflow in IE is even greater using your
branch version than in my version (index.demo.php).
> there are still things to do, before it can be go final. (get involved
> @http://groups.google.com/group/jquery-ui-dev/browse_thread/thread/6e2...)
Great new features. FWIW I'm checking out your code to see if I can
be of any help.
> 1. The scrollbarBase-fix (is NaN)
> I don´t know, wether you are right here, but NaN is a fals-y value and
> should already been filtered...
I know it doesn't make sense, but that's the bug I got. Thankfully, I
don't get this bug in your branch version.
> 2. margin & Dean Edwards styleLeft fix in curCSS
> This is a bug in jQuery and should be handeld there. But your are not
> realy accurate here. The same bug appears, with other properties like
> border-width, padding, background-position etc., too.
> The best way to workaround this issue is to make the number-parsing
> more accurate: ((!isNaN(parseInt(opts[i].newStyle[n], 10)) && opts
> [i].newStyle[n].match(/\s|,/) === null) // is a number and has no
> whitespaces
Unfortunately, attempting to animate _any_ of the margin properties
using your branch version gives me the same curCSS error in IE. I
have setup a test case @
http://www.doublerebel.com/index.demo-branchver-ietest.php
The test case works flawlessly in Webkit & Gecko browsers, but fails
in IE. The test case will crash in the curCSS function in IE, I use a
modified jQuery core script in the test case, it will log the property
and value to the console. (You should see in your console "property:
margin ret: 0px 0px -1em")
> 3. Your color-animation fix
Your branch version is a more appropriate fix, I agree. On a great
sidenote, that same fix also applies to the clip animation plugin,
which was based off the color animation code. Until I get the fix
properly posted, all are welcome to pick mine up @
http://code.doublerebel.com/jQuery/jquery.animate.clip.js
> I´m also interested in performance issues (you
> mentioned something with opera).
Yes, running multiple class animations simultaneously in Opera is
noticably slower than on any other supported browser. I haven't done
any precise testing yet, but on my Vista and XP machines, the
animation becomes jerky whenever I trigger another mouseover or
animation.
Thanks again,
Charles
>
> On Mar 29, 4:32 am, polyrhythmic <cwp4...@gmail.com> wrote:
>
> > Hello,
>
> > I've been wanting to properly separate MVC, and keep my CSS properties
> > in my classes and out of my JS, so I have been hacking away at the
> > class animation methods. I have fixes for the WebKit getComputedStyle
> > object bug and the IE6+ 'undefined' color animation bug.
>
> > Ticket 4324 (http://dev.jqueryui.com/ticket/4324) is where I attached
> > the file, and added a link in Ticket 4328 (http://dev.jqueryui.com/
> > ticket/4328), since it's the listed color animation 'critical bug'.
> > The color animation fix is based off of 1.6rc6 as I haven't found any
> > advantage in the 1.7 code. I'm not sure of the root cause of the
> > bugs, but I have fixed the symptoms.
>
> > I saw a thread on the jQuery dev list where Paul Bakaus prefers to
> > animate all possible CSS properties:http://groups.google.com/group/jquery-dev/browse_thread/thread/fd4ce5...
> > That is why I didn't use a whitelist (a la Scriptaculous), although
> > performance may suffer.
>
> > Code is tested in IE6 and 7, Firefox 3+, Safari 3+, Opera 9.5+, and
> > Chrome on XP and Vista. Mac and IE8 testing is needed. Performance
> > with multiple concurrent class animations in Opera 9.5+ is very bad.
>
> > Let me know if you have any questions! I'm a bit green at submitting
> > fixes so I'm open to suggestions.
>
> > Thanks,
>
> > Charles
>
> > On Feb 26, 11:27 am, alexander farkas <a.farkas...@googlemail.com>
> > wrote:
>
> > > Hi polyrhythmic,
>
> > > It would be great, if you would share your improvements and bugfixes.
> > > can you report the ticket number of your bug reports here, too?
>
> > > There are some dependencies between the animateClass- and thecolor-
> > > plugin. If you fix the borderColor-propterty all browsers except
> > > firefox are trying to animate borderColor + borderLeftColor etc. This
> > > can make the animation very sloppy.
>
> > > You can find the branch @http://jquery-ui.googlecode.com/svn/branches/dev/effects
>
> > > regards
> > > alex
>
> > > On Feb 26, 10:32 am, Paul Bakaus <paul.bak...@googlemail.com> wrote:
>
> > > > On Wed, Feb 25, 2009 at 8:54 PM, polyrhythmic <cwp4...@gmail.com>

The IE reflow bug appears to be related to hasLayout. I created a
workaround by defining height and margin@
http://www.doublerebel.com/index.demo-branchver-hasLayout.php
Nevertheless, it is strange that your branch version produces more
extra whitespace due to the hasLayout bugs than the 1.6rc6 code.

Just chiming in halfway through the convo here, but wont
element.style.zoom = 1; force hasLayout?

Hello polyrhythmic,
this really great help thanks .-).
I fixed the margin, padding etc. bug in the branch. I had to test also
the oldStyles against whitespaces.
the scrollbar-color bug is related to the fact, that the old filter-
function only tests wether there is a style-property with the string
color in it.
i only test, wether we have a special plugin for this. this means the
clip-effect should also work with the animateClass-feature.
(i´m using a mix of filter, blacklist and whitelist)
I can´t really test performance @ home. but i will look in to this
problem.
regards
alex