Post

what if part of sync state was a small per-repo LRU cache of block CIDs? since most updates (at least with tid rkeys) stay on a few hot edges of the mst, maybe you could avoid sending most CIDs with most commits, just referencing a cache index instead

…oops i’m gonna make an alternate firehose format, aren’t i

I think the issue with something like this is that the sender needs to know the receiver's cache size, and I don't think there's a good one-size-fits-all answer to cache size