AEP-154: proposal: only support one etag type.

[toumorokoshi]

AEP-154 has a discussion around strong and weak etags.

This Is referencing the existence of the concept in RFC 7232.

I worry that allowing support of both strong and weak etags increase the complexity in a client's ability to handle these features.

In addition, the guidance on weak etags (being that it's "differences that the server feels is unimportant") leaves a lot of room for interpretation that may include fields that should never be excluded (e.g. user-settable fields).