Ian Jeffries

Agreed, no need to rush it. I still think it's useful to track though to minimize confusion.

No idea if this is a good idea, but I thought it was at least worth discussing.

Sounds great. I may take a stab at this after I finish SVG support. A few thoughts: + I'm curious if we can make an API with read-only, write-only, and...

And another thought: Do you remember why the "Delta" variable contains a channel? I wonder if we could get away with just a TVar.

Interesting. Multiple listeners can also react to a single TVar using `retry` to listen for changes, so it's possible we might be able to get away with just that. I'll...

> I think there's no getting around using channels unfortunately. If so that's fine, it doesn't seem like a big cost. However.. > Yeah, but with `takeTVar`/`retry` the first woken...

> Also, I think `onChange` shouldn't take a `current` value as an argument, but rather get it from the `TVar` first before going into the loop. Nice, good idea!

Sweet, that worked! > But I think this is still a conceptual problem with `stack` (and `cabal`). You have similar issues if you have e.g. one `hspec` and one `doctest`...

> I don’t plan to publish/maintain this myself, but @seagreen if you want to take responsibility of this code, then please go ahead (no attribution required). I'll pass on that,...

@rightfold: Field names will definitely appear in serialized data. My plan is to make Plate useful for describing meaningful-on-their-own formats like https://en.wikipedia.org/wiki/ICalendar, not to compete with Protocol Buffers-style formats that...