Closed garbados closed 2 years ago
I wonder too.
I think the hash link
field to reference older messages.
Maybe it should allow multiple "links
" to enable merge messages.
The post/delete
already has two, a link
and a hash
too
But anyway - that's already some sort of topological time sort, right?
=> So, if many messages happened, then depending on which peers you choose to trust or not, you could check messages which according to their "topological sorting" happened at a similar time and then compare the timestamps (e.g. take the mean time stamp?)
Also, at least in post messages (the ones that represent writes and matter), the public_key
could represent a hypercore too. If messages need a nonce
, those could then be hypercore chunk indexes, because combined with the pubkey, they are unique
Typically you could record a time sent and time received, but with the links
we are creating a directed graph, so these records aren't used for ordering messages, the graph is. the timestamps then are simply for user convenience
Fair enough. Closing this issue...
Regarding time range queries, should peers just optimistically trust each other's clocks (bad idea) or is there another approach implied?