[APICHANGE] Remove some Cajita/Valija cruft.

* Remove ___.extend in ES5/3.
* Remove .../demos/playground/setup-valija.js.

@r5357

[felix8a, 2013-04-16 00:46:10]

markCtor looks like it's still being used by others. let's fix those before
deleting it. (otherwise, cutting a new release becomes complicated)

[kpreid2, 2013-04-16 16:56:36]

* Remove ___.extend in ES5/3.
* Remove ___.markCtor in ES5/3.
* Remove .../demos/playground/setup-valija.js.

[kpreid2, 2013-04-16 16:57:05]

On 2013/04/16 00:46:10, felix8a wrote:
> markCtor looks like it's still being used by others. let's fix those before
> deleting it. (otherwise, cutting a new release becomes complicated)

New snapshot which deprecates instead of removes markCtor.

[kpreid2, 2013-04-16 16:58:21]

* Remove ___.extend in ES5/3.
* Deprecate ___.markCtor in ES5/3.
* Remove .../demos/playground/setup-valija.js.

[kpreid2, 2013-04-16 17:07:45]

* Remove ___.extend in ES5/3.
* Remove .../demos/playground/setup-valija.js.

[kpreid2, 2013-04-16 17:08:59]

On 2013/04/16 16:57:05, kpreid2 wrote:
> New snapshot which deprecates instead of removes markCtor.

On second thought, given that there are outside parties using this, there's no
strong reason to explicitly deprecate it and not also deprecate the rest of the
Cajita-compatibility API, and I'd rather just let all of ES5/3 die out (or at
least not touch this can of worms now).

New snapshot doesn't touch markCtor at all.

[felix8a, 2013-04-16 17:38:53]

In general, I'm very wary of deleting anything that's a published public
interface, but I'm unable to find external users of ___.extend, and it seems
unlikely that some random person we don't know about will be using ___.extend,
so ok. lgtm.

[kpreid2, 2013-04-16 17:40:05]

On 2013/04/16 17:38:53, felix8a wrote:
> In general, I'm very wary of deleting anything that's a published public
> interface

Ah, I should also tag this [APICHANGE], because it is.