Heavy petting for the contextual component RFC

[mixonic]

@mmun for you to consider.

[workmanw]

👍 I could definitely see several places in our we could use this.

---

One question, what about proving class references, rather than strings, to the component helpers. For private component classes? For example:

// app/components/special-button.jsvarMyPrivate=Ember.Component.extend({/* ... */});exportdefaultEmber.Component.extend({myPrivateComponent: MyPrivate/* ... */});

{{!-- app/templates/components/special-button.hbs --}}{{!-- passed by class, not string --}}{{componentmyPrivateComponent}}

We've got a lot of components in our app that aren't used in block form and have "private" sub-components. It gets kind of messy to have every component in it's own .js file sometimes.

---

I was secretly hoping to see some mention of a splat-like-operator, but I realize it's probably outside the scope of this RFC.

[mixonic]

@workmanw splat is being discussed at handlebars-lang/handlebars.js#1050 We're waiting for wycats to clear his stack. Hopefully we green light soon :-)

@mmun has concerns about how we handle the container (and therefor injections etc) on bare classes. I think it may be surmountable. If he is hesitant I would at least like to get the un-controversial version of this rolling so we can land it soon, maybe 2.2 if core approves on Friday.