Fork me on GitHub
#datomic
<
2017-03-08
>
dominicm12:03:46

@robert-stuttaford I saw this https://gist.github.com/robert-stuttaford/3bd5240c988f05092504 and your comment about a few issues. Wanted to know if they were with onyx integration or with that strategy as an input?

robert-stuttaford12:03:00

@dominicm can’t remember the issues; but i can tell you that we are happily using this strategy in production to this day 🙂

dominicm12:03:33

@robert-stuttaford so you're doing it with that gist instead of onyx-datomic?

robert-stuttaford12:03:09

no, we’re using onyx-datomic

dominicm12:03:00

@robert-stuttaford ah okay, so you're using tx-range and not the tx-report-queue then

robert-stuttaford18:03:55

TIL in Datomic you can alias :db/idents (the most recently assigned alias becomes the default ident). effect: rename it, but support all past names.

timgilbert20:03:29

Yeah @robert-stuttaford, we've been renaming like :product/name to :deprecated/product.name and the like as we evolve our schema

csm22:03:08

how are t and next-t updated in the client API? I think I’m seeing an issue where one query gets an older t than expected when using (datomic.client/db conn)

csm22:03:44

I’m seeing an issue where one service transacts some facts, and another service performs a query (five minutes later), but the query gets the t from just before the transaction

csm22:03:11

literally the t just before the transaction: the db-before is :t 1343698, :next-t 1343699, the db-after is :t 1343699, :next-t 1343700, and the query side has :t 1343698, :next-t 1343699

csm22:03:49

is (d/as-of db (:next-t db)) ever a good idea? or (d/as-of db (Date.))?