principle-concurrency

Prevent race conditions and deadlocks by applying structured concurrency rules, ensuring child lifetimes nest under parents with clear error propagation, and selecting appropriate synchronization primitives like mutexes or channels to manage shared state safely.

Choose between mutex, channel, actor, semaphore, and atomics based on ownership semantics, performance requirements, and correctness needs, ensuring the selected primitive aligns with your concurrent system's specific data access patterns.

Structured concurrency handles cooperative cancellation by enforcing lifetime rules where child tasks nest under parent lifetimes, ensuring no orphan work remains and errors propagate cleanly through the cancellation tree without lost wakeups.

You need backpressure in producer consumer pipelines when production rates exceed consumption capacity, requiring synchronization strategies to pause or slow producers, preventing unbounded queue growth and subsequent system instability.

Concurrent code suffers from starvation or livelocks when synchronization primitives are poorly selected, causing threads to indefinitely wait or repeatedly change states without progressing, requiring a correctness-first framework to diagnose.

Yes, this concurrency framework applies across both threads and goroutines, providing a correctness-first approach to diagnosing issues like ABA problems, selecting appropriate primitives, and structuring cancellation trees regardless of the runtime.