Fixes to mutations, also some others
@pospi I went through all the graphql stuff, made some adjustments to the layer docs. And a lot of fixes to the mutations. I am missing some technical knowledge, so please review carefully! Although at least I think it is better than it was.... :)
Questions (that might imply some re-work):
- I noticed sometimes the commented section in each mutation was
interface
and sometimestype
. Didn't know what to do with that, above my pay grade. - There were missing things from the
interface
section that sometimes made sense to me and sometimes didn't. Didn't know if it was OK to have missing things, seemed like from the comments that the create and update params needed to implement something? If not, maybe we should take outfinished
for example on creates. Or on the other hand, why not leave it, and allow the option of creating data after the fact? The resolvers could default to false as needed. -
created
I decided shouldn't be just the machine time, but should be able to be input since it has some additional semantic meaning where it is included. You might disagree, let me know. (I don't think of it exactly as the timestamp we talked about adding to all records.)