Options for adding docs and test fixtures for multicodec entries

In the space of a few weeks we have two separate discussion about needing a place to define entries in the multicodec table, including test fixtures:

• Murmur3 specifics needing to be pinned down properly: https://github.com/multiformats/multicodec/pull/236#discussion_r730573053
• A new proposed RSA pubkey compact binary representation for the did:key spec: https://github.com/multiformats/multicodec/issues/235#issuecomment-957071667

We don’t really have such a space currently. Some options:

• We have https://multiformats.io which is currently just a very simple static 3 page site - hasn’t been touched in a year and uses Hugo. It'll probably need some work to make it easier to hold specs and fixture data, including moving off Hugo.
• We have https://ipld.io which is much more actively maintained and we’re shoving test fixture data into it these days too (mainly using testmark). The problem here is that even the above two entries push uncomfortably at the definitional scope of "IPLD" - do either of those really belong on an IPLD site?
• We have this repo, which we could just add a new directory to "specs" which contains a directory for each new spec we want to add, and not worry about fancy web frontends.
• We have https://github.com/multiformats/specs which contains some initial drafts of general multiformats IETF specs (to define multihash, multiaddr, multibase etc.) that we wanted to propose (we still need to do this but the work is a bit stale!). We could expand the scope of that repo to allow it contain arbitrary specifications for any entry in this table that doesn't have a logical home elsewhere for a specification.

At this stage I'm inclined to consider the last option - just do it in the repo, no web frontend, encourage the use of testmark for fixturing but not be too restrictive in formats or content of a "specification" initially but let that evolve over time.

Start by reading the repository's existing specs area and the linked Murmur3 and RSA discussions, then compare the four possible homes described in the issue. Done would require a decided location and agreed format or process for multicodec specifications and test fixtures; the issue currently leaves that design unresolved.