Clojurians Log v2

That's a great explanation, thx!

for us it was that we have a system of entities that have relationships to each other. we were creating API routes that exposed them, but then e.g. entityA related to entityB, I would need to fetch it from the A API and then the B API, which was slow. The other thing we would do is create ad-hoc API routes for each use case, that would return parts of entityA and entityB as was required by the UI, but that led to a proliferation of endpoints

pathom will allow us to declaratively select exactly what data we need from our entire system and load it on the front end, from one endpoint.

I've been waiting for some motivation to try pathom/eql for a while, so two (noob) questions: 1. Does introducing pathom/eql require any server-side changes? 2. Do they play well with re-frame?

you can implement pathom on the client but we decided to implement it in our API layer for simplicity

since we did that, we use re-frame to send POST requests with an EQL query and then update the app-db with the results. it works well

we went an extra step further and are putting the results into a datascript db in the app-db to do normalization for us. this is OK for short term, but long term I'm interested in replacing it with something like https://github.com/lilactown/autonormal

Thanks! one more question, what is the major motivation of introducing pathom/eql into your system?