Engineering Political Compass

Rate each engineering principle. We'll plot you in political vector space.

Each rule projects onto one or more political dimensions. Your answers are aggregated per-axis using magnitude-based normalization — strong and consistent opinions move you further along each axis.

Skipping a rule is fine. Rules you skip don't contribute to any axis.

Avoid Single Points of Failure (SPOF)
Distribute critical functions so no single component can bring down the whole system.
Organizations design systems that mirror their communication structure
System architecture inevitably reflects the social and organizational structure that produced it.
Program defensively: assume everything will go wrong
Validate inputs, handle failures gracefully. Trust nothing you didn't verify yourself.
Don't Repeat Yourself: eliminate duplication through abstraction
Centralize logic in one place to ensure consistency and reduce maintenance overhead.
Fail fast, fail often
Surface problems early and loudly. Hidden errors compound; visible errors get fixed.
If it ain't broke, don't fix it
Proven, surviving systems carry hidden value. Unnecessary change introduces new failure modes.
Keep It Simple, Stupid (KISS)
Complex systems have unpredictable failure modes. Favor minimal, understandable architecture.
Principle of Least Privilege
Every actor gets the minimum access required. Default deny; trust is granted narrowly.
Given enough eyeballs, all bugs are shallow
Transparency and collective review surface problems faster than any closed expert group.
Measure twice, cut once
Deliberate planning up front prevents costly irreversible mistakes later.
Move fast and break things
Rapid iteration and creative destruction beat preserving existing structures.
Premature optimization is the root of all evil
Solve today's real problems; don't build rigid structures for hypothetical future needs.
Redundancy equals reliability
Overlapping backup systems prevent catastrophic failure when individual components fail.
Technical debt is better than no product
Accept imperfect solutions to deliver value; refactor later if it earns the right.
Worse is better: simple working solutions beat perfect ones
A flawed solution that ships and evolves beats an elegant one that never sees daylight.