Add missing variations of sha2

[warpfork]

Per discussion in #205 .

I've inserted these entries at the first available numbers following the other sha2 variants. It's unfortunate we can't put them all in a row, but no space is available since some sha3 variants immediately follow the already-entabled sha2 variants.

[vmx]

I would move sha2-512-224 and sha2-512-256 to higher ranges. But I don't have a strong opinion on that.

[rvagg]

IMO these should be up in the 0x1012 range. The single byte range is precious. sha2-224 is unlikely to get wide usage too so I'd prefer that to be higher as well. I can see the case for 384 though.

fwiw I don't think FIPS 180-4 is a strong claim to their level of importance at this stage, in crypto terms it's ancient, if you're making new software then it's not highly likely that you're reaching for it as your standard. And FIPS-202 is all about SHA3 I believe.

[warpfork]

Oh I'm definitely not claiming a FIPS number expresses anything about their importance :) That's just there to remove any possible ambiguity about the naming vs what the thing is.

[mvdan]

I also don't have a strong opinion, but I have to admit I'd avoid taking spots in the single-byte range unless the added multicodec will be very common :)

[vmx]

I also lean towards putting it into higher ranges (as @rvagg suggested) and it seems most people involved lean towards higher ranges, lets put it into higher ranges and be done. I guess that most important thing is that it is somewhere.

[warpfork]

I'm pretty strongly held that I want sha2-384 to be in the single-byte range because it is the "best" out of these hashes (see discussion in issue) that is also available in the standard libraries we tend to notice first (js, in node and browsers, and golang). So I think it will be quite common.

I guess I'll bump up the sha2-512-* numbers to two-byte if we like.

I could also see bumping up sha2-224 to a higher range because I don't think it'll be commonly used. Fragments things that are related, but I guess avoiding that isn't really a goal we need to prioritize, so okay.

[warpfork]

... I seem to have started this PR using the github edit feature, and then deleted the fork it creates in the meanwhile, so I no longer know how to amend this PR. Going to close it and open a new one shortly.