The documentation on this page http://scripty2.com/doc/scripty2%20fx/s2/fx/parallel.html suggests that effects within a parallel can have individual delay and duration properties.

In the example shown you would expect the first effect to run for 0.9 seconds after a delay of 0.1 second and the second effect to run for 0.75 seconds after a delay of 0.25 seconds.

Running the code results in both effects running for a duration of 1 second with no delay on either.

Also worth noting that the individual 'after' functions from each effect also fail to run.

Is this a bug in the implementation of S2.FX.Parallel, or a bug in the documentation. If it is a bug in the documentation, then could this parallel be extended to use these options?

Thanks.

It looks like firefox is dropping the style.MozBorderRadius in favor of style.borderRadius which breaks VENDOR_PREFIX detection AND VENDOR_MAP.

We should try to find another way of populating the VENDOR_MAP. Maybe by iterating through a list of style properties scripty2 depends on by testing the presence of the property and fallback appending possible prefixes (webkit Moz) and cache it.

I'll have a look at it tomorrow.

I've created windowing system based on S2.UI.Dialog. You can see the demo here:
http://wrzasq.pl/chilldev.js/application.html

I'm trying to implement as many features in original S2.UI.DIalog class as I can and commit it to upstream to contribute with Scripty2, but I don't want to provide code with such a nasty "hacks". The problem is with additional icons on title bar (those to iconize and collapse dialogs). We follow jQuery UI styles, but unfortunately they do not support multiple icons there (just a closing element). Does anyone have an idea how to put additional items there in celever way? At the moment i'm just setting "right" style for each element. If we work that out, I could commit at least toggling/collapsing mechanism.

Hi everybody
I'm not shure if it correspond to the project philosophy
but i think it could be very cool to have a method "setContent" to refresh the content of dialog box.
is it a good idea ? i've made some code to illustrate :

setContent: function(new_content) {
    if(new_content){
      this.content = Object.isElement(new_content) ? new_content :
      new Element('div').update(new_content);
    }else{
      new_content = new Element('div');
    }
    this.element.select('.ui-widget-content').first().update(new_content);
},

there is no "memo" var when you capture the "ui:dialog:before:close" event.

it can be fixed with a copy/paste the line memo, before the element.fire "ui:dialog:before:close"

var memo = { dialog: this, success: success };

and replace

var result = this.element.fire("ui:dialog:before:close", {dialog: this);

with
var result = this.element.fire("ui:dialog:before:close", memo);

thx a lot for the fix

Hello,

On the manual it says that calling cancel "immediately halts execution of the effect"

http://scripty2.com/doc/scripty2%20fx/s2/fx/base.html

Tried that http://jsfiddle.net/cGDYE/ I was expecting the effect to not even start but it continues animating.

thank you

http://scripty2.com/hardware-acceleration/ displays the following message, "NOTE: Your browser supports CSS transitions, but not accelerated compositing. The demonstrations on this page will work properly, but cannot take advantage of hardware-accelerated animation."

Hardware acceleration was added Sept. 30 2010 in the dev build. http://googlechromereleases.blogspot.com/2010/09/dev-channel-update_30.html

Autocompleter.js
The zIndex are not set at the creation time so the list could appears behind other elements :

// Position the menu to appear directly below the input.
(function() {
var iLayout = this.input.getLayout();
this.menu.element.setStyle({
left: iLayout.get('left') + 'px',
top: (iLayout.get('top') + iLayout.get('margin-box-height')) + 'px' ,
----------->>> zIndex : 9999 <<<------------ just add this line here to fix it
});

http://github.com/madrobby/scripty2/blob/91c9a51e163dbec9da7ff226898f2f242d26db3a/src/extensions/element.js#L114

This should be :
return element.effect('morph', Object.extend({

Do you prefer we post issues here or on Lighthouse?
I might commit a patch when i leave work.

... by making the first curve an inverted parabola instead of a parabola, and merging with the second curve. The movement will then model a ball being thrown onto a high ledge and bouncing, giving the effect a more natural feel.

Hi,

Wishing to add autocomplete support on the new HTML5 input types, like email, tel, ..., I realized that it's blocked to "text" types so far : https://github.com/madrobby/scripty2/blob/master/src/ui/controls/autocompleter.js#L23

I guess the main goal being to avoid "submit" types right? Would you consider the following selector input:not([type="submit"]) proper as a replacement ?

Hoping that helps,
Best regards,

• Sébastien Grosjean - ZenCocoon

The prototype.s2.js and prototype.s2.min.js files contained in the zip on the homepage are both the a6 versions.

hello,
I've checked out S2 today and try to build it.
I've found some errors because build system search file in SCRIPTY2_DEBUG_DIR but theses files are now in SCRIPTY2_DIST_DIR since commit 81e0e05

I've generated a patch but don't know how to to send it here ?
in case here the gist of the patch http://gist.github.com/234479

On form submit I disable all button on the page with:

$$('.ui-button').each(function(button){ button.retrieve('ui.button').setEnabled(false); });

For buttons that are <a> elements, they have the correct visual style but can still be clicked as if they were enabled.

When I try the demo in Minefield I get the following error:
$("panel").morph is not a function from line 87 on the html page:

I’m not sure if this is intentional, but S2.UI.disableTextSelection() prevents elements from getting focus because event handling is stopped in _textSelectionHandler for mousedown events.

Please see http://stuff.vandervossen.net/temporary/focustest/ for a test page and http://stuff.vandervossen.net/temporary/focustest/focus.mov for a demonstration of this issue.

When trying to use the auto-complete I encountered a number of IE errors. I have forked the project on github and committed fixes to each of the issues. They are:

• Error sending custom event to anonymous menu element - eric1234/scripty2@8e1d4e411a84039b40101db8365289fe02689cb0
• Placement of auto-complete menu needs deferment to get accurate layout - eric1234/scripty2@12dacf62fa335363dfe3f44cd27c601bfafb7947
• Check for selected item when setting input -
  eric1234/scripty2@96b63c63b4d76d385c4f4639a24001dccc659854

I am putting this all under one ticket as it is really just about getting Auto-complete to work in IE. But each change is as a separate commit so you can cherry-pick the commits you want and/or provide a different implementation if you prefer.

I will also send a pull request to this project.

I noticed interesting issue with combining slideUp and morph, which breaks following animations when they're css-transition -based.

I stripped it down to this simple test case — http://jsfiddle.net/xBgtm/6/

Notice 2 implementations and how 1st one (slideUp + morph) breaks when CSS-transitions -based effects are enabled, whereas 2nd one works as expected.

I didn't look deep down into it, to see what exactly is going on, but I did notice transition-related styles (-transition-property, -transition-duration, etc.) NOT being removed from an element after animation is finished. This happens only when combining morph and slideUp/Down. When using standalone morph, transitions styles are removed properly. That's probably what makes the following animation fail.

Finally, I'm not fully sure, but it also looks like only morph uses CSS transitions for animation, and slideUp/Down doesn't. Is that so? And if yes, what's the reason?

Thanks!

I'm working on an app that uses multitouch transforms heavily. I was developing with a3 and not having any problems. When I moved over to the master a5 (with no changes to multitouch, just lots of extra UI features), I see the manipulate:update events coming much more slowly. As a result, the scaling becomes wildly inappropriate. This is on a fairly new Mac 10.5 with both the latest WebKit and FF3.6. I would appreciate if someone else could reproduce the bug to make me feel I'm not crazy.

How to reproduce the issue:

Download the source of the flickr demo ( http://scripty2.com/demos/touch/touchshow/ ) and run it with a local copy of scripty2_a3. When you resize an image, note that it resizes proportional to the distance dragged, and can be resized smaller by dragging back toward where you started. Now run the page against a local copy of scripty_a5. When resizing an image, notice that it gets bigger much faster, and when you drag back toward the starting-point, it still gets bigger.

Is this project discontinued? If so, what alternative would you suggest?

button.js line 189 should be:
primaryIcon ? 'ui-button-text-icon-primary' : 'ui-button-text-icon-secondary';

Simple solution:
if a button "options" property exists then pass it to the UI.Button constructor. Resulting example for API would be something like:

new S2.UI.Dialog({
   ...,
   buttons:[ {
      primary:true,
      label:'OK',
      action:someAction,
      options:{icons:{primary:'ui-icon-check'}}
   } ]
}

This behavior could be a bug if the content of the Dialog is a form. You only could focus the form elements using tabs, click on they using the mouse is impossible.

Dialog by default closes on return keypress with success status. However, there doesn't seem to be a good way to prevent this if you want to use an ajax request to determine the success.

Example: I have a sign-in dialog that I want to appear throughout the site and have different success callbacks in some cases so I observe the container's ui:dialog:close:after event and success means the sign-in was successful and otherwise not. But if the user presses enter the dialog is closed with success and there is no way for my callback to distinguish a success from my ajax callback versus the return keypress.

Suggestion: at minimum, add a closeOnReturn option so this behavior can be easily disabled. Otherwise, it seems like the most logical action for the return keypress would be to trigger the click action of the primary button so that the actions attached to the button are executed instead of just always closing the window.

Using git clone of repo and ruby 1.8.6 (2008-08-11 patchlevel 287) [universal-darwin9.0].

Running the default rake task which executes:

[:clean, :dist, :unified, :doc, :package, :clean_package_source]

fails running the :package task with this error:

** Invoke package (first_time)
** Invoke /Users/stephen/dev/javascript/scripty2-git/pkg/scripty2-2.0.0_a4.tar.gz (first_time)
** Invoke /Users/stephen/dev/javascript/scripty2-git/pkg/scripty2-2.0.0_a4 (first_time)
rake aborted!
Don't know how to build task 'README.rdoc'

Running:

rake package

by itself works fine.

With S2.UI.Autocompleter and having highlightSubstring to false brings an issue, no text displayed in the proposition list

Replacing "text" by "result" fix it all.
http://github.com/madrobby/scripty2/blob/master/src/ui/controls/autocompleter.js#L178

P.S. : Brilliant start with this lib, I truly hope to see more work on it as it looks really promising

Scripty2 's morph() function is not working in firefox while in webkit (tried Comodo Dragon and Chrome Canary) it works fine. In Opera also it works perfect.

In firefox, console reports " $('main').morph is not a function " error all the time.

Can anyone tell me what's going on??