Oleg Grenrus
Oleg Grenrus
This is a side-effect of candidates not removed when version is published, so maybe these candidates can be just removed too. E.g. `base` ![Screenshot from 2021-06-22 19-03-12](https://user-images.githubusercontent.com/51087/122960863-824e8c00-d38c-11eb-900e-269ae7d5d5f8.png) ![Screenshot from 2021-06-22...
e.g. https://hackage.haskell.org/package/basement-0.0.11/revisions/ ``` base:{} >=4.9.0.0 ``` probably the change is `base >= 4.9.0.0`....
But its bounds advertise they would. Please either test with GHC-7.8.4 or explicitly drop the support, i.e. adjust lower-bound.
The title tells it all.
A problem I find in improving `aeson` is dealing with generic and TH code. It's monolithic, trying to cater for all needs at once. Issues like https://github.com/haskell/aeson/issues/824 and https://github.com/haskell/aeson/issues/827 is...
:+1: / :-1:, @bergmark ?
```haskell decodeThrow :: (MonadThrow m, FromJSON a) => ByteString -> m a ``` this generalises `decode` as there is `MonadThrow Maybe` instance, but it's also usable in `IO`. Now as...
The idea is dumb and simple: (I speak only about `Encoding`, `toJSON` / `Value` is analogous). - Add a `toFieldEncoding :: a -> Maybe Encoding` to `ToJSON` class (and `liftToFieldEncoding`...
While working on the sax-like stream based JSON parser, I noticed that parsing `Value` with it is faster than with `attoparsec`. It started as a separate library, but as it...