Could somebody explain or share info, how to cover such case: I have some huge dataset like millions / billion records which I somehow displaying with pagination and filters. I need to call spicedb and ask what exact filters are accessible for current user, than pass it to my query and obtain final result But during expanding all this relations i potentially could get around million of ids and pass them to other query. This is kind'a overhead. *Could you advise how to solve this correctly? * I was thinking about like to mix up some groups in crdb and sync all of this, but unfortunately this is not a solution. I need to save flexibility Any help appreciated

If I understand your problem correctly, it reminds me of OPAs partial evaluation solution https://blog.openpolicyagent.org/write-policy-in-opa-enforce-policy-in-sql-d9d24db93bf4

the general rule of thumb is that you filter on one side (or the other), and then apply the other filter. If you expect to only show, for example, a subset of results, you can filter in your search index and then issue checks to filter. Alternatively, you can call LookupResources with cursors and use that for pagination.

This is cool. The 64bit requirement is a bummer for those using UUIDs

see the later messages

Fantastic

not currently