Remove/deprecate/document sneak-dependency methods

[dmethvin]

Some methods are ambiguous, creating "sneak" dependencies that are going to confuse Closure Compiler and also thwart manual attempts to reduce size.

• jQuery.fn.load() - Either needs event.js or ajax.js; only the arguments tell for sure. Can we deprecate/remove one or both to make it unambiguous? At minimum need to document effect of its use.

• jQuery.fn.show(), jQuery.fn.hide() - Needs effects.js only if a speed is provided. Can we deprecate the animation-related ones? Again, we at least need to document it.

As a starting point I suggest we deprecate the event behavior of .load() and the animate behavior of .show()/.hide(), removing them in 2.0. Perhaps we could put in a jQuery.error() if .load() detects it's being called all event-like.

Are there other examples?

[Rick Waldron]

With jQuery.fn.show/jQuery.fn.hide they are "members" of effects.js... Maybe I've misunderstood?

[dmethvin]

Sorry, what I meant was that for the no-args case they are just css display changers, but for all other cases they need .animate(). It would be good to be able to keep the simple show/hide but not require all of effects. Or maybe we should tell people who want to do that to use classes and avoid show/hide unless they want to include effects.js?

[Timmy Willison]

+0, I don't think this can actually happen. Not having to included effects for the simple usage of show/hide would be nice and .load() is unfortunately ambiguous, but backwards compatibility will always be important. Also, I thought John said there was never going to be a 2.0.

[jzaefferer]

+1, animated show/hide always looked bad - overloaded load is just bad.

[jaubourg]

+0, Let's investigate properly first... also, I'd like to see if we could use Closure to help here.

[mikesherov]

-1

[Rick Waldron]

-1

[dmethvin]

Voted down, but I may do some of this as we investigate the Closure integration.