I haven't put any effort into any kind of snapshotting capabilities yet, since I won't want the scope to be too large and there's often ways of designing your system in ways where replaying isn't a big issue (the db scans are fast, designing aggregates to have a smaller scope/less overall events, etc).

But with increased resources, I can see some solutions being considered around snapshotting. For now the goal is heavily unix philosophy inspired: a really fast and purpose built event sourcing database