<@U619WRNJE> expanding on what <@U0RM4EPC7> said t...

# arrow

[raulraja]

@ersin_ertan expanding on what @simon.vergauwen said that is a polymorphic function that can compute over any higher kind for which a MonadError<F, NukeException> exist. In the case of Option it can't provide such an instance because there is no way to store the NukeException, since None which is the unbiased case can't store that error. Same for Try since it's Failure it's fixed to Throwable and NukeException is not in the Throwable hierarchy. The only choice here is Either which is parametric in it's Left case.
attack may be invoked in two different ways:https://gist.github.com/raulraja/40ac0d0f03f4a1a85f6d249464f30a95

[ersin_ertan]

@raulraja Still getting a compile error. Is monadError a variable or is it suppose to be ME.buindingE? https://gist.github.com/ersin-ertan/ec0674dc1160158d3eaccc6b2eaf7374

[raulraja]

@ersin_ertan this should compile https://gist.github.com/raulraja/40ac0d0f03f4a1a85f6d249464f30a95

[simon.vergauwen]

Haha omg it took me a good while to spot the difference.

[simon.vergauwen]

@ersin_ertan perhaps it can save you some time. bindingE is only available on MonadError<F, Throwable> as it catches any exceptions thrown within the bindingE block and then calls raiseError(e: Throwable) to model the thrown exception is correctly.

[simon.vergauwen]

@raulraja correct my if I am wrong.

[raulraja]

that is correct but here in this context if you wanted to catch runtime exceptions you could still use ME.catch(...)

[raulraja]

which automatically calls raiseError underneath if the computation fails.

[raulraja]

But yeah bindingE is only for general exceptions.

[ersin_ertan]

Great, is something that should be reflected in the docs? If so, I can do that.

[raulraja]

@ersin_ertan any improvements to the current state of the docs is welcomed!

[raulraja]

https://github.com/kategory/kategory/issues/311