Hey devs, as we work toward 0.8 Restful API compliance, it is critical that we have a dataset that fully exercises the code for compliance. What we have now in our little xml file (javascript/samplecontainer/state-basicfriendlist.xml) is woefully inadequate. I'd like to have a set of tests that fully tests the underlying source data and use that test suite as our public 'proof' of spec compliance. The first reasonable goal is to make sure we can GET all the data and then we'll test that we can write and do round-trip operations.
currently the 0.8 Rest spec is using the 0.7 data spec: http://code.google.com/apis/opensocial/docs/0.7/reference/opensocial.Person.Field.html http://code.google.com/apis/opensocial/docs/0.7/reference/opensocial.Activity.Field.html Other data types that I could pull out are the group object and the simple schema of the app data. These need to be exercised as well. These are the most important, but are there others? In addition, spec compliance read tests should exercise the auth and search systems that are applicable to all these resources. Are there any people who have done their own implementations who would like to volunteer some of their testing datasets? Not only is the test data missing for many fields in the spec, but some could have a number of potential formats, depending on how the container chooses to represent them. If you look at the new XML representations in the spec docs for Activity, for example, you'll find that there are some subtleties express regarding the formats of the various id's some are strings but the data is a URL, some, like author are URI but the data is a string and so on. thanks for any help! davep
