I’ve been working on a project for some time (which we will open source as it nears completion) to port the fdb-record-layer to go, based on experience we have with a similar closed source library that I wrote with similar goals to the record layer. If anyone else has interest in this effort please DM me!
I have a question on the preload cache. I’m wondering why the preload cache isn’t actually a load cache, which is then populated on all record loads (ie: loadRecord, queries, and so on).
This benefit is that workflows that do:
- load a bunch of records
- update records
- save updates
In the save portion, the existing loaded record (the raw data) can be fetched from the load cache, and we don’t need to hit fdb again.