Mark Hammond

I thought this was just related to functions, but it's also a problem for records with default values. Eg, consider [this app-services record](https://github.com/mozilla/application-services/blob/da9301f829de67149d50dce5b709aed42f1bfe1b/components/places/src/places.udl#L419-L427) - Python generates a constructor for this:...

I think this would be solvable if we insisted on `kwargs` - eg, if the sig was `__init__(self, **kwargs):` and lots of boilerplate that unpacked the `kwargs` dict etc. This...

Thanks! > The downside of that is that we lose the self-documentation: argument names act as documentation if properly chosen. Right - but I'm saying that's a downside of the...

(and to be clear, I'm talking here about *just* constructors for records and *not* for arbitrary functions. I think there's a case to be made to do this for functions...

> (Side-note: Got a link to that app-services record? I'm curious) The 3 structs [here](https://github.com/mozilla/application-services/blob/main/components/places/src/places.udl#L409) all do something similar. I only noticed this when I tried to create Python bindings...

I don't think anyone plans on working on this, but we'd certainly look at a PR!

"background thread" implies "low priority thread" to me, which I don't think is the actual intent. The actual use-cases here tend to be more about thread blocking than thread priority....

using the same terminology as tokio is certainly appealing!

This issue title is broad enough I think it's fine to cover both cases. If it turns out a patch is submitted which only solves some inefficiencies but leaves others...

At a very high abstract level, this makes perfect sense to me and is roughly the approach @bendk took for the Desktop Firefox JS bindings (although there we generate .webidl...