happy Friday everyone! We’re very excited to set up Pact tests in our company. We have already everything in a place but we’re still unsure with this functional vs contract tests thing. I’m looking for an advise if we should have just one or multiple interactions for an endpoint. The endpoint GET

/profile

expects a query parameter

code

and returns a user profile object. User can be in one of three statuses:

new

,

lead

, and

user

. When it’s

new

, the profile must have

emailAddress

property. If it’s

lead

or

user

there must be also

phoneNumber

and

username

properties present in the response. There are also optional fields like

firstName

that might be null or a string. Now, the questions are: 1. Should we have only one interaction for all the user statuses or an interaction for each status separately (wouldn’t it be testing a business logic?) 2. Should we have a separate interaction for the optional properties to be given (not null)? I hope that’s a good place for such questions, Best regards

Good philosophical question, and, in the end, not actually a Pact question :) You can go either way on all the aspects you've mentioned, but you'll be trading-off specificity.