Do people hava any recommondations here about using schema or spec for your coercion and generating swagger schema's?

It seems the docs say spec might be too alpha but I am wondering if there are people who have some more experience with it

@mitchelkuijpers we use both in projects and they are mostly on par feature-wise. Internally: Schema was built to support coercion & walking, so the implementation is proven and final. For Spec, coercion is an add-on built with the clojure core team members have stated many times that it’s a bad idea.

in the current project, we moved back to Schema from spec as (as a consultancy) we can’t leave the client with “unofficial, might break in the future” solution.

Ah ok, that is good to know

but, both work. I prefer Schema.

I think I'll just go for schema right now, it also seems the errors are a bit more user-friendly out of the box

I'll rather be on the safe side for now

we are btw desinging a new tool for this space, as we are mostly building multi-tenant data-driven systems: just data, can be read & written over the wire & db, first class transformations & human errors. A fusion on all our internal tools & best parts of schema & spec. can be transformed to specs too.

Very cool, we are building a CRM and are looking to create a first version of our REST api. We will not focus for a while on the errors.. let's first ship something!