Fork me on GitHub
#graphql
<
2017-05-17
>
stijn07:05:16

@hlship we are using selection to identify the type of input arguments in mutations.

stijn07:05:43

should I provide this feedback in the issue too? or do you want to discuss it here?

hlship16:05:52

stijn: Discussion about a specific issue is best in the issue; general discussion or planning for future enhancements is good right here.

stijn07:05:30

our 'update' and 'create' resolvers are very generic and use the type information from the graphql schema to convert the input to a datomic transaction

stijn07:05:06

this means that we need to inspect the input args, get their type from the schema and do the conversion

stijn07:05:09

on the query side we don't use selection as with datomic entities you don't have the SQL n+1 problem

stijn07:05:41

@kenny no, we are using Apollo for the moment

hlship16:05:52

stijn: Discussion about a specific issue is best in the issue; general discussion or planning for future enhancements is good right here.

hlship19:05:18

That seems like a bug to me; the spec goes into a lot of detail about handling of input (what we might call conforming of inputs, w.r.t. to clojure.spec). I'm a bit surprised that there's an issue with an variable value of type Int or Long and an argument of type GraphQL Int.

hlship23:05:04

Definitely a bug, and fixed in upcoming 0.17.0.

hlship19:05:27

What release is this?

boubalou20:05:08

0.15.0 since I cannot use 0.16.0 because of the mutations issue.

avindra21:05:34

does anyone know roughly when lacinia 0.17.0 is going to drop?

hlship21:05:04

Probably this week. We follow a process where PRs are reviewed by at least two core committers and half of them are on vacation this week, which slows things down.

hlship21:05:36

I like to keep a regular release about every two weeks or so.

avindra21:05:27

dope, thanks!