Neither mini-apps nor a monolith
The hard part isn't making lots of apps: it's deciding where one ends and the next begins.
The monolith
The mega-suite where you buy 200 features to use 3, and every upgrade is all-or-nothing.
The fragment
Micro-apps that do nothing useful alone and dump the integration burden on you.
Our craft is the boundary
Every AppLibra app is scoped to one complete job: AL MES isn't "a timer" - it really runs your shop floor and gets the cost right - but it stops there, it doesn't try to be a warehouse too. Neither too little nor too much.
What makes the balance possible
A free base
AL Foundation App and AL Message Passing (the bus) gather the common infrastructure: apps neither reinvent it nor absorb it. And you don't pay for it.
They compose over the bus
AL MES and AL Processing Charge stay separate but collaborate: MES passes real time and cost, Processing Charge turns them into a charge.
They extend the standard
They post to Business Central's standard journals and ledgers: leverage on the platform, not a parallel engine to audit.
The proof
Real composition at work: MES feeding Processing Charge (time and cost), the scale sending weight to Catch Weight, GS1 decoding the barcode for the warehouse. Each is a complete app; together they're a system.