kbjr / events.js Goto Github PK

On Chrome and Safari (and possibly other browsers, but I haven't tried), with the following snippet of code, I run into a type error.

Code:

Events.ready( function() {
  Events.unbind( window, "resize" );
  Events.bind( window, "resize", function() { console.log( "I'm handling it." ); } );
} );

Error (Chrome):

Uncaught TypeError: Cannot read property 'removeNamespace' of undefined
self2.removeEvent                                            events.js:929
self.unbind                                                  events.js:1203

If I call Events.unbind after Events.bind, this issue no longer occurs.

Also, if I change line 929:

        // Remove the namespace
        events[event].removeNamespace(segments);

to

        // Remove the namespace
        if (events[event] !== undefined) { events[event].removeNamespace(segments); }

The error goes away.

Great work on the library!

You code seem to replace mouseleave with mouseout, witch is bad. If you have child element in the element you binded the event to, mouseout will trigger for every child as well as the targeted element. Mouseleave fix that and trigger only one when you exit the element you binded the event on.

removing this code fix the issue for me :

    mouseenter: {
        attachesTo: 'mouseover',
        eventTest: function(evt) {
            return (! withinElement(evt, evt.originalTarget(), 'fromElement'));
        }
    },
    mouseleave: {
        attachesTo: 'mouseout',
        eventTest: function(evt) {
            return (! withinElement(evt, evt.originalTarget(), 'toElement'));
        }
    },

why is this there?

i just test it on IE 6, and its not work,
could you making it to support IE 6, and other browser?
i like this framework,
hope you can make it compatible with other browser

thx

Title kinda says it all, I would like to have it included in the NPM :)

Thanks

to be super-awesome, Events.js have to be fully unit test covered with test results in all browsers/operating systems

This is causing event binding to fail in IE. A fixed has been developed and tested and will be in the next push out.

Looking through HashChangeFix, it seems it defaults to poll every 25 ms. It only stops if it receives a native event for a hash change, which never comes if the hash never changes.

It seems to me that the default should be to check if the browser supports the native hashchange event. Something like: ("onhashchange" in window.document.body)

For example, change:

supportsOnhashchange = (function() {
        var nativeSupport = false;
            return function(support) {
                if (typeof support === 'boolean') {
                    nativeSupport = support;
                }
                return nativeSupport;
            };
         }()),

Into:

supportsOnhashchange = (function() {
        var nativeSupport = "onhashchange" in window.document.body;
            return function(support) {
                if (typeof support === 'boolean') {
                    nativeSupport = support;
                }
                return nativeSupport;
            };
         }()),

I tried the following code:

Events.bind(window, 'resize', function(e){
    console.log('Hello World');
});
Events.invoke(window, 'resize');

But Chrome threw the following error:

Uncaught [object Object] on events.js:1015

Inspecting line 1015 of events.js showed me that you called a function called despatchEvent instead of dispatchEvent. I think this is a typo? Anyway, fixing this (as well as the same typo on line 60), did prevent the script from throwing an error, but the event still wasn't invoked / executed.

On the demo page, you can click on the "Try Me!" button a lot and it makes the animation calls stack. It should be pretty easy to fix this, I could even do it if the source was available.

Any chance the site source is available somewhere?