Felix Paulusma
Felix Paulusma
We actually created an "add-on" that does exactly this, but has a horrible API and isn't easy to use (or safe) at all, exactly because `aeson` inherently doesn't keep track...
\* `instance HasAuth (AuthHeader "X-Token") User where` I'd also like to help make the authenticating with `servant` better (see #1494 ), so I'll have a good think about this as...
Some brainstorming, just wrote it down, no idea if this would even compile. ```hs -- | When the 'Lenient' mod is set, the handler will be passed the entire 'AuthResult...
I might be a bit too new to `servant` to immediately get this, but what is this "generalized auth" you all are talking about? It sounds like it's a different...
Ah, it is labeled as `EXPERIMENTAL` everywhere, which is why I didn't pay much attention to it. :thinking:
I've been playing around and have something that technically only needs a `Request -> Handler (NewAuthResult a)` in the context. The `auths` aren't used at all (yet), but I think...
I'd like to see what you come up with. 😃 I've taken this approach mainly because that's how `servant` seems to do things (i.e. `ReqBody' mods ctypes a`). If you...
Hmm, maybe an idea to do both the term level and type level solutions? [Type families might speed up significantly in the (near) future](https://gitlab.haskell.org/ghc/ghc/-/issues/8095#note_404797)
> A lot of the compilation time cost with servant APIs isn't due to type families themselves, has more to do with how servant stuffs are represented in Core and...
@tchoutri Added it in an extra comment in the original issue. (#986) :+1: