Re: MIPs at control level

>
> I have been reconsidering MIPs at control level recently.
>
> First of all, the well-known trick ".[relevance_predicate]" should better
> be explicitly written with a @relevant attribute.
>
Agreed.

> When a trigger has to be grayed on a specific condition, instead of having
> to create some element in some "work" instance, it would be nicer to just
> add a @readonly attribute.
>
Agreed. (I think we had a discussion a long time ago about whether
`readonly` might imply a "disabled" mode for controls. HTML has both
notions, by the way.)

> Even @type might be useful for overwriting the effective data type
> because, for example, for a date of birth, using a not-constrainable date
> widget is not as convenient as a simple text input.
>
I am less sure about this one.

-Erik