From f6f246031d4c8bec4eb3694629d2ac6bc128e598 Mon Sep 17 00:00:00 2001 From: Rod Vagg Date: Thu, 26 May 2022 15:19:25 +1000 Subject: [PATCH] remove bmt entry This is an ongoing pain-point because it serves as an example of the very weird edges of the multicodec table. I'm not aware of any real use; or even how it could be practically used. It's not a "hash function", but describes a hash process that needs additional "function" information. I suspect it has its genesis in looking at Bitcoin et. al., but for those binary merkle trees we use `bitcoin-tx` coupled with `dbl-sha2-256` which gives us all the information we need. `ssz-sha2-256-bmt` is a newer example of an entry that's more descriptive, it stretches definitions a bit here but describes a hashing process within a format. It's probably not an awesome example either, but it's closer to a standard (current) understanding of multiformats than `bmt` is (not just because it can actually be implemented and used). Can anyone think of a good reason to keep this entry, or can we remove it as cruft? --- table.csv | 1 - 1 file changed, 1 deletion(-) diff --git a/table.csv b/table.csv index 015a147..e884530 100644 --- a/table.csv +++ b/table.csv @@ -79,7 +79,6 @@ stellar-block, ipld, 0xd0, draft, Stell stellar-tx, ipld, 0xd1, draft, Stellar Tx md4, multihash, 0xd4, draft, md5, multihash, 0xd5, draft, -bmt, multihash, 0xd6, draft, Binary Merkle Tree Hash decred-block, ipld, 0xe0, draft, Decred Block decred-tx, ipld, 0xe1, draft, Decred Tx ipld-ns, namespace, 0xe2, draft, IPLD path