Necessity of $ / manual context control about spect

[dy]

Possible problem of manual context control is:

• no destructor - not clear how to detach effect
• async fx figures out current aspect via callstack

yes, that's the problem. Since calling spect is not limited to current scope, we have to push and pop context, that is manual control over context, that can possibly come along with scope-bound control.

• very similar to incremental-dom. That's actually a high-level extension of it.

So what would that be?

// attach aspect per fn-scope
spect(target, el => {
})

// manual control over scope
groupStart(target)
groupEnd()

What about setting state on manual context? How should that trigger redraw?
Maybe we should provide scope(el), scopeEnd() methods, but only as low-level machinery. state should be able to call defined scope at any moment. For that purpose tiny scoping functionscope(el, () => {}) is the most natural choice.

Whatever good the idea is, that's just a round-walk path: scopeStart is nice, but we have to rerender scope. Defining scope from inside is impossible, must be scope(target, () => {}), which is essentially the same as spect(target, () => {}).

To make spect compatible with jquery, we can make it a default export, so that user would just import $, {} from 'spect', so that if want to rename $ to something different - just import spect from 'spect'

[dy]

Ok, so the initial issues are addressed:

• to resolve jquery conflict - import $, { fx } from 'spect' to resolve conflict
• to resolve naming inconsistency - see prev.
• selector usefulness - it is shortcut.