Collection of changes for the next schema release

[nevrome]

This PR prepares the next major release of the Poseidon Schema, planned as version 3.0.0

[nevrome]

This PR contains a number of changes for the Poseidon schema v3.0. See changelog.md for a summary.
Please have a close look and point out potential issues. Feel free to react with a 👍 to this comment if you're happy with these changes, otherwise please post below.

Pinging some core members and power-users, but anybody is invited to comment. @stschiff, @TCLamnidis, @AyGhal, @wolfgangaroo, @Mattists, @martynamolak

[nevrome]

Looking at this from a bit of distance I now think there are some aspects we could/should still specify in this version, notably some details we previously discussed on an implementation level:

• encode information on group names and genetic sex in VCF poseidon-hs#347 (as a CAN ?)
• ASCII-only for PoseidonID and GroupNames poseidon-hs#331 (as a MUST)
• Amend group-name parser to avoid forbidden forge-string characters poseidon-hs#332 (as a MUST)
• trident validate should check LF-line endings  poseidon-hs#311 (as a SHOULD)
• Validation of accessions from the GSA fails poseidon-hs#350

I'm also aware that there are more ideas about .janno columns we may want to add to the schema eventually, e.g. discussed in #97. But that requires good preparation and discussion. To move forward with this release now we should not open any can of worms here, imho. New columns, that were not mentioned in the review of the Poseidon paper, should be added later.

[Mattists]

The allowed Source_Materials are "petrous;bone;tooth;soft;sediment;other".
We somewhen decided that I create the following types of materials for my network: "Hair, Petrous, Tooth, other Bone".

• I would suggest to also add "hair" as a Source_Material as it appears in the Source_Tissue values.