I can't tell if this is in jest or not but blockchains are prolific in software.

Nearly every company uses them to prove the authenticity of deployments in production at one or more layers.

I feel like signatures and maybe even Merkle trees are used, but not blockchain.

Nearly every company, what is your source for that? I've never even heard of this before.

No, no they do not.

IIUC a blockchain is a structure where each node has a content addressable hash that includes the hash of the previous node in the chain. If you change any historical node, every subsequent hash is updated.

This structure is used inside your .git directory, your docker manifest, etc.

Blockchains are incredibly useful structures.

In context blockchain is referring to a distributed set of hash structures + a consensus algorithm

That's a Merkle tree. Blockchains are a Merkle tree plus some form of consensus algorithm, often a trustless one. It's the usefulness of the consensus algorithm which is usually called into question by critics (generally because it's often expensive and cannot reference anything outside the system).

IIUC a merkle tree is a data structure where the leaf nodes are the hashes of data. The inner nodes of the tree are hashes of the child hashes up to a root hash.

Merkle trees are helpful for quickly validating the integrity of a chunked file. Both IPFS and BitTorrent use merkle trees to validate files.

I do not understand how git could represent its history using a merkle tree.

git commits themselves are merkle trees (or more like, DAGs) that contain filesystem trees, a set of parent commits, and metadata

each parent commit in turn contains their own parents etc, until you reach an initial commit

the funny thing here is that parent commits don't have references to children, it's children that have references to parents (like the union find data structure) so the relationship here is inverted. that's because git objects are immutable

It was in jest because blockchain is a solution looking for a problem, but as others have pointed out what you say isn't really the case. I'm glad it generated some discussion. Hashing previous steps for verification isn't blockchain.