Telokis
Telokis
## Update: Events definitions and emit/broadcast are 100% strongly typed based on a mapping Below is the same example as the one used in my last comment but updated to...
Yes, I understand, thanks for pinging! Regarding the existing TS projects my goal was to not break anything (hence the multiple conditional types to make sure the defaults are the...
@Wallacy > I'm going to do a test tomorrow on the project that I use the moleculer to look at this issue of backwards compatibility. Thank you very much, I...
@icebob > The legacy event handler signature is (payload, sender, eventName) but after 0.14 there is a new with (ctx). But for backward compatibility the parseServiceSchema tries to find out...
It won't be a breaking change if they don't intentionally "enable" the strong type-checking by providing an action mapping and/or event mapping. (If I properly kept the backwards-compatibility, of course)...
> I have to admit, much of this is over my head at this point. Conceptually, I love the idea of automatic inference of types on remote services. As long...
> By the way, as I see, the value is stored, just not return it, right? Yes, from what I noticed, the value is properly stored and retrieved but the...
Thanks for your answer. To make sure I understand: - We add a configuration option `enableNullValues` to [cachers/base.js#L29](https://github.com/moleculerjs/moleculer/blob/d6cbe63abf436d0792ca614edb3129008137c001/src/cachers/base.js#L29) - If `enableNullValues` is set to `true` - [Cacher] Return a specific...
@intech Hi! No, I didn't work on/with Moleculer since around then. I still need to properly finish my PR #829 and won't consider tackling this issue until that PR is...
> > @intech Hi! No, I didn't work on/with Moleculer since around then. I still need to properly finish my PR #829 and won't consider tackling this issue until that...