One of the (many) challenges in #testing the #Fediverse is that some of the standards say one thing, but what is widely implemented is another.

Example: RFC 7033, WebFinger, section 5, requires presence of a CORS header.

So far we've not seen any Fediverse application that actually does this. Arguably it's not needed in the Fediverse, but the standard says MUST.

So we can either have all applications fail that test or not test against the standard. Your thoughts?

rfc-editor.org/rfc/rfc7033#sec

@feditest maybe it's good to focus on the real world implications of the incompatibilities one by one?

For example, will this lack of a header actually impact anybody in the real world? I don't know the answer.

@volkris @feditest Agreed, especially for something like CORS where it is a lot easier to find information about how it works than about what problem it solves (especially if you want details on the latter).

Follow

@soaproot yeah, so to answer @feditest 's prompt, that's my thought.

Focus on testing against the things that have real world impact either today or in the near future, and that way the "shaming" of noncompliant implementations might carry more weight as more than pedantry.

Or have a separate level for basic vs perfect implementations.

Sign in to participate in the conversation
Qoto Mastodon

QOTO: Question Others to Teach Ourselves
An inclusive, Academic Freedom, instance
All cultures welcome.
Hate speech and harassment strictly forbidden.