[eeps] EEP 049: Value-Based Error Handling Mechanisms

[Jáchym Holeček]

On Fri, Dec 7, 2018, at 14:08, Anthony Ramine wrote:
> Are you encouraging the use of non-local returns to work around a 
> limitation of the language?

They are merely stating what the standard altenative to value-based
error propagation with explicit case analysis along all of call stack
has been for a very long time. The `throw` feature has, by convention,
always been used as a private non-local return mechanism in some very
contained body of code.

> Literally all new programming languages made these days have such a 
> short-circuiting construct some way or another that doesn't translate to 
> non-local returns. I don't see what makes Erlang so special that you 
> would argue it doesn't need one.

This statement is so broad, vague and subliminally aggressive that
maybe it deserves expanding upon with specific technical content
once you've calmed down a little bit.

BR,
    -- Jachym