A boring enterprise programmer mindset
I don't want another event loop.
I don't want another scheduler.
I don't want another async framework.
I want to solve business problems.
I have customers.
I have orders.
I have invoices.
I have payments.
I have users.
I have permissions.
Sockets are not my business.
Networking is infrastructure.
Databases are infrastructure.
Timers are infrastructure.
Files are infrastructure.
I don't want them leaking into every function.
When I open the source tree, I want to see
Not
When a message arrives,
I don't care if it came from
- TCP
- UDP
- QUIC
- IPC
- Shared memory
- A file
- A timer
I care that
arrived.
I don't think in sockets.
I think in business events.
I don't want fifty items touching the same state.
I want one owner.
One place.
One decision.
I don't optimize first.
I optimize after measuring.
Most engineering time is spent understanding code.
Not making it 3% faster.
When somebody joins the team,
I want them to understand the architecture in a week.
Not the scheduler in a month.
Five years later,
I want to add a feature.
Not rewrite the execution model.
Performance matters.
Architecture matters longer.
For most enterprise software,
architecture wins more often than microbenchmarks.
That is the mindset behind a "boring system."
Not slow.
Not old.
Just predictable.
Easy to understand.
Easy to change.
Easy to keep running.