callAjax: function() {
var me = this;
new Ajax.Request(‘ajax.php’, {
onComplete: function(transport) {
me.result = transport.responseText.evalJSON();
}
}
}

or

callAjax: function() {
new Ajax.Request(‘ajax.php’, {
onComplete: (function(transport) {
this.result = transport.responseText.evalJSON();
}).bind(this);
}
}

In the former case “me” is set to the callAjax’s containing object and since “me” is not “this” it’s unaffected by onComplete’s bindings. The latter uses Prototype’s bind() method to state that onComplete’s “this” refers to the same object as callAjax’s “this” (effectively by encapsulating the former idiom).

In either case, the advantage is that “myobj” isn’t explicitly mentioned, so it doesn’t need to be globally scoped, and the same method can be aggregated by other objects without them mistakenly messing with myobj’s result.