Fork me on GitHub

I need some predicate functions like ref? to-many?, that check if a keyword is or not a ref? on db. Is a good idea create a mem db at build-time, install schema, run a query, get the list of idents/keywords and make this functions Or is better do it at runtime, with memoize or something like?


This seems a trivial predicate to write with d/attribute. Why are you thinking of caching, memoizing, etc?


I'm using d/entity at the moment. But I wanna go to the client api


oh I didn't realize this was client api

👍 4

peer API is way better then client API but client API is tons of light years better then #sql 🙂


Not sure about "better", but I actually query my "model map" which contains attribute descriptions (and which get turned into schema change txs at system boot).


I actually have a number of phantom attributes, too, that get computed but never committed, and this structure has made that possible.


Hey all, I'm having a bit of trouble trying to get a Lambda function connected to my Datomic Cloud db via the client API. My function keeps timing out with the following error:

Unable to execute HTTP request: Connect to 
connect timed out
I've followed the instructions here: And here: but still not having any luck 😕 My Lambda is configured to access the Datomic VPC via security group $(SystemName)-apps as in the documentation. I have not tried VPC Peering however. Is that a necessary step when trying to access Datomic Cloud from Lambda? Thanks.


I've even tried giving my Lambda function administrator privileges to make sure it's not something silly like that..