Why startWorker() and not worker()?

[Tri0L]

Hello!

Why are you using startWorker() and not worker()?

workerpool/workerpool.go

Lines 192 to 200 in 85cc841

	case p.workerQueue <- task:
	default:
	// Create a new worker, if not at max.
	if workerCount < p.maxWorkers {
	wg.Add(1)
	go startWorker(task, p.workerQueue, &wg)
	workerCount++
	} else {
	// Enqueue task to be executed by next available worker.

Maybe you have some examples, how I can pass db connection to worker? Not to specific task.
Main idea is: Up some workers with 1 connection per worker and do sql requests(tasks) using worker sql connection.

[F21]

I am interested in doing something similar to this as well. Maybe there needs to be a worker interface that we can use to create our own custom workers that will then be started and stopped by the pool.

[gammazero]

Why are you using startWorker() and not worker()?

When starting a new worker, that new worker to run the new task directly, instead of pulling it off the work queue. Since the behavior was slightly different when starting, I split it into two functions. Other than that, there is no reason it cannot be the same function. That change to use the same function and goroutine was made in #59.