At last, a voice of reason amid the vulgar crowd.

I've held several management positions where I've carried a pager. At one employer I helped keep trading databases operational in 7 time-zones on 3 continents. At another I helped fix a backup issue on Christmas eve. Helping those customers was a core part of my responsibility. I fully understood that and took great pride in it.

But as a developer I too will never accept another on-call rotation.

Companies which assign on-call duties to developers make the mistake that development, management and operations are different kinds of work which require different environments and skill-sets. Other engineering tasks include testing, documentation, training, and maintenance. At small startups the founders and early employees may do some or all of these but that becomes impractical at larger established businesses.

Engineers should learn and do all these things in the course of their career but not all at the same time unless quality isn't a concern.

My experience at a unicorn a few years ago convinced me companies which assign developers on-call rotation either don't understand or don't care about the quality or sustainability of their business. In that company senior management was replaced by folks from Google and Facebook shortly after I joined. I was moved into a team where I had no role in the design, develop or deployment of its services. I had no say in the hiring or firing of the so-called engineers who rushed failing services into place past a wholly ineffective QA department.

I should have seen the writing on the wall when I began to be pressured by managers and recruiters to rubber-stamp candidates who couldn't pass our coding tests but had spent lots of time on-call. The company's priorities slowly became clearer to me as they grew evermore desperate to live up to their promises. Ultimately I suffered an ischemic attack from the stress of this environment and left the company to focus on my health.

Oh and the company? It let go of most of its engineers a year later and was eventually acquired by competitor for a few hundred million after having raised over a billion dollars.