Fork me on GitHub
#ring-swagger
<
2017-03-02
>
mpenet07:03:37

@ikitommi personally I wouldnt' wait on core. I doubt things will change (other than having s/form conforming). Some/most things could be built on what's available and as long as it's modular this should be ok for the long run (ex about form "parsing")

mpenet07:03:55

Maybe feature parity with what's available now with Schema shouldn't be a goal tho. I guess this should be incremental and shouldnt be afraid to break the design

ikitommi08:03:16

true that. Listed the differences into a gist (for Macchiato) some time ago (https://gist.github.com/ikitommi/fb3e0200504dd8b635ed7edd0cdbc768): stuff could be finalized into a intial support for next week - both coercion & docs.

ikitommi08:03:07

now on vacation (hiihtoloma we call it in Finland :)), will write all I know that is missing into issues into github repos early next week, we do open source on next week fri, try to glue things together then. Will tag all issues somehow, contributions welcome.

ikitommi08:03:41

and the coercion… did a small spike on spec-tools how to do coercion without the dynamic binding, looks good: instead of wrapping things itno generic Spec records, a more typed records are returned (like Spectrum does, KeysSpec, etc.) - these still proxy calls to underlying clojure.spec instances but also implement a new Coercion protocol taking the 3rd argument. Easy to implement for all the specs. Usage looks the same as with the dynamic conformers, but without any magic or surprises.

mpenet09:03:04

that does sounds good, nice

mpenet09:03:31

I ll try to help but I am a bit busy with personal stuff atm (wife expecting to deliver in a few days)