Hello everyone, Sometimes I'd like to investigate the commit history of a (large) git repository, but I don't actually (yet) care about the actual files in the repository. I know about shallow clones, but it covers kind of the opposite problem: a person may just want the files, but not the history. 1. Is there a way to download only the git commit objects, and not the corresponding tree (or file) blobs? 2. Is this a use case that anyone else cares about? I suppose a ton of code assumes it can follow any references, so implementing this as a feature means a ton of code would need to be adapted to handle the reference-not-locally-present case. Maybe it might be interesting, but would make the git code-base way too complicated? I don't really know what keywords to look for, and man git-fetch did not superficially seem to answer this, so I apologize if I wasn't able to find this feature in the documentation. Best, Antonio
