I think the conversation is more nuanced. In gene...

# compose

[jim]

I think the conversation is more nuanced. In general, across the application architecture (entire stack) you want to ideally have a single source of truth. That is to say, you should never have two places (variables, databases, tables, etc) where the same information is duplicated/synchronized. Attempting to do so inevitably leads to data consistency issues and makes it much harder to perform updates. So as a general principle, I think single source of truth as defined by Compose does in fact apply across the entire stack.

Having said that, there are practicalities where the real world can start to intrude. For example, performing optimistic updates on a local shadow of the single source of truth while you're waiting for a round trip with your database server. Sometimes such things are necessary in the real world, but that doesn't mean it is desirable architecturally. So yes there is some nuance, but I wonder if the slide above may create more confusion among newbies than clarity. Food for thought.