Several immediate things come to mind: - this is a mechanism for sharing files and directories (e.g., zipped csv files), whereas noms defines its own structure

Several immediate things come to mind: - this is a mechanism for sharing files and directories (e.g., zipped csv files), whereas noms defines its own structured data model that is much more granular - noms has versioning built-in, you can track the history of a particular dataset - this is firmly based on bittorrent. you could maybe run noms on top of bittorrent, but it's more intended to be run like git, where you talk directly to a server that you want to collaborate with