Add test for relative priority of nested and referential styling

[palemieux]

Nested styling has higher priority than referential styling

[skynavga]

I am concerned that this is not a bug fix to an existing test, but is an attempt to expand the set of tests beyond what was used at the time of specification development. What is driving this expansion of tests? Is it intended to address a new feature? If not, then it is not appropriate to add, even though it may be a legitimate test for interoperability testing. Please keep reminding yourself that these tests are not intended to be used for interoperability testing.

[palemieux]

This particular combination of nested and referential styling is not currently exercised.

The IMSC test suite is not intended to be used solely as an exit criteria for the IMSC specification, but more generally as a means for implementers to exercise their implementations.

[skynavga]

I don't recall this being agree to by the TTWG? This repo is in the domain of the WG and not implementers. I might agree to adding tests submitted for use by implementers if they are segregated from the primary tests used for specification verification as part of the W3C process. However, there are other options as well, such as creating additional repositories for that purpose.

[palemieux]

The IMSC test suite was always primarily intended to be useful to the industry, and has always been described as such -- below is the result of a quick search.

https://www.w3.org/2018/08/09-tt-minutes.html
https://www.w3.org/2018/09/13-tt-minutes.html

The test suite is under the control of TTWG and always has been. Nothing new here.