Bootstrap efficiency

[aquaticrna]

boot strap is currently very naive and is simply sending the whole db across, in order to avoid duplication distinct is being use to filter incoming id's that should be cleaned up a fair bit, might be worth transacting schema data separately from non schema data

[metasoarous]

The ids are being fetched right now directly from the index in such a way that extracting the eids gives us duplicates. Hence the need to process with distinct.

The reason I did this stupid thing was to get around the fact that Datomic doesn't let you run a datalog query for all eids in the database. Oy. The thing to do I think is to pull from [?e :db/ident] for schema, and maybe query for everything that doesn't have an ident as "app data". But both should be specifiable, the latter in particular. Once we have posh exporting it's queries to the server for scoping though, we'll use that to compute the bootstrap.

These default functions should be part of datsync.