anutron/behavior

Auto-instantiates widgets/classes based on parsed, declarative HTML.

behavior's Issues

mootool keep event alive for behavior

Hi,

I have a span element on my page ... test,

i need to addEvent to pay attention this element, so in my mootool js is kind of like this
var mytest = $('test')
test.addEvent('click', function (e, el) {
.... do something
};

whenever i click the , it will do something ...
on my html ... sometimes i need to toggle to remove that and show that

As what i know ... once i remove that , the event tracking is gone ... i think that's the purpose to avoid memory leak, correct?

But in my case, i kind of need to keep that event tracking even though the is gone ... any suggestion? any options in the mootool addEvent to keep it alive?

hongclub

MooTools 1.3 syntax

Hi Aaron,
will you accept a patch with MooTools 1.3-only compatible syntax?

Broken links

Hello,

The three links in the readme.md are broken, leads to a Github 404 page.

Tibor

Delegator.register('mousewheel', [...] will not be triggered in FF

See the following example

http://jsfiddle.net/Q6VGy/1/

works fine with Chrome but not in FF.

Tested in FF 13.0.1, FF. 12.x, FF 9.x

Error catcher supresses errors

Hey
I wondered if it would be possible to cancel the error catcher behavior applies. I'm having 2 major issues with it:

1. It cancel the stack trace, thus I can't see where the error originated from
2. It actually suppresses errors, making it impossible to know something is actually broken and where.

JSON.isSecure is too secure

http://jsfiddle.net/ZEcg2/

actually no JSON did get past it... even the example from the docu1 won't work :/

"{opt1: 'foo', opt2: 'bar', selector: '.selector'}",
"{opt1: true, opt2: false}",
'{opt1: "foo", opt2: "bar"}'

Allow filters dependencies

Hey
So, I've been using Behavior quite heavily for the last few months (and having great fun at it :) ), and I've come across one issue a few times -
There are cases where one filter needs to use some functionality that is created by another. So far I've used several ways to work around it. Usually all my Classes have a load event of some sort, so I can check for it and use it, but what I could really use is a way to specify filter order (right now filters are using for-in loops - which don't have a specific order on all browsers, and I've noticed some inconsistent behavior on FF as well).
I haven't created a pull request because I think this is something that would require some change to the inner architecture, but I have several ideas for this.
Is it something you might implement or should I create my own implementation for this?

Conflict with addBehavior method name

I tried using Behaviors in our system, but ran into a conflict with out history manager:
http://plugins.jquery.com/project/history-js
history.js has in its amplify.store.js file a section that begins

if ( div.addBehavior ) {
div.addBehavior( "#default#userdata" );

It appears to be related to compatibility for non-standard IE--it seems that addBehavior was the name of a method in some older version of IE's js, or something, and that the compatibility code in history-js is causing a conflict. If that addBehavior method was ever widely used, you could run into other conflicts as well.

Behavior API _getOptions bug with IE

The BehaviorAPI._getOptions method uses:

if (options && options[0] != '{') options = '{' + options + '}';

In IE 8/9, 'myString'[0] returns undefined, instead of the expected 'm'.

I suggest not using IE of course, but for those who must, changing the subject line to:

if (options && options.substring(0,1) != '{') options = '{' + options + '}';