Is it important to distinguish between ::rf/eid
and ::rf/qid
?
Since maps with those in it are used with different fns, would ::rf/id
work for both?
Currently you can't visually tell an event vector from a subscription vector (other than by adopting naming conventions for subs and events, which is what I'm currently doing).
Using eid
and qid
makes it clear what a map is supposed to be used for. However, is that important?
Other variants: :re-frame.subs/id
and re-frame.events/id
?
One more stupid question. Except for making subs not pure. Why is it bad to subscribe in reg-subs
,
What I mean is something like this
(rf/reg-sub
::foo
:<- [:bar]
(fn [bar]
@(rf/subscribe [bar])))
Have to track which components need to rerender when the data changes. And “untrack” when it in lints. There is no time this function can untrack and so it’s just leaked in the tracking. (My understanding. There’s a lib that does this for you and some googling will find it. I’m eating lunch now)
Actually, it should be fine in that regard because it's all wrapped in a reaction.