3 ms·

It's an unsolvable problem, because any sufficiently complex job will therefore be complex to measure. I mean, if you could measure it, then you could just do i

It's an unsolvable problem, because any sufficiently complex job will therefore be complex to measure. I mean, if you could measure it, then you could just do it, no? Why spend all this time measuring how good an engineer is? If you really want to know, you have to look at the code, sit in on the meetings, write the code and see the challenges with integrating with the existing system, submit a PR to see the pushback and... oh wait now you're just an engineer. If we hire two engineers to act as one then maybe, maybe, we could measure performance. At the cost of the largest inefficiency imaginable. It's not unlike using generative AI for coding. Writing code is the easy part! Making sure it works is another beast. Honestly, I'd much rather write code than read it. If it takes X time to write code and > X time to read it, which in my book it does, why bother with the reading?

danaris 2y ago
"Unsolvable" in the sense that there's no one magic algorithm you can use, sure. But every job has some ways to determine if the person in it is actually, y'know, doing it. And off the top of my head (and it's early, so YMMV), I'd say that remote jobs tend to have more easily-auditable products than many in-person-only jobs (particularly the type where the main purpose is direct interaction with the public). > If we hire two engineers to act as one then maybe, maybe, we could measure performance. At the cost of the largest inefficiency imaginable. And why should "efficiency" be our most important criterion? "Efficiency" is why we have lean staffing that forces the "bus factor"[0] of nearly every position to 1. "Efficiency" is why we have stretched, fragile supply chains that break down the moment any one supplier or middleman experiences a crisis. Efficiency is the enemy of resilience. The only people it really helps are shareholders. [0] https://en.wikipedia.org/wiki/Bus_factor https://en.wikipedia.org/wiki/Bus_factor
- consteval 2y ago
  Efficiency is perhaps not the right word, but I was under the impression that attempting to measure performance is a cost saving tactic. I think you lose the cost saving if you just double your engineers while keeping the same amount of work.
  - danaris 2y ago
    I mean, from a certain perspective, nearly everything management tries to do is a cost-saving tactic, but I wouldn't say that measuring performance is primarily about "cost savings". It's more about making sure that people are doing the job they were hired to do—which, on the face of it, is definitely a valuable thing to be doing. The problem is when it's being used as a whip to get people to work ever harder with no more pay and no more support than before. Doubling engineers while keeping the same amount of work is something that can increase resiliency—both in the engineers themselves and in the business. Making sure you have at least, say, 125%-150% coverage of all duties makes a huge difference in everyone's stress levels, and makes sure that if someone gets sick, gets hit by a bus, or just goes on vacation, their duties neither stop getting done, nor get piled onto other people who are already at 100% (or more) capacity.

danaris 2y ago

"Unsolvable" in the sense that there's no one magic algorithm you can use, sure. But every job has some ways to determine if the person in it is actually, y'know, doing it. And off the top of my head (and it's early, so YMMV), I'd say that remote jobs tend to have more easily-auditable products than many in-person-only jobs (particularly the type where the main purpose is direct interaction with the public). > If we hire two engineers to act as one then maybe, maybe, we could measure performance. At the cost of the largest inefficiency imaginable. And why should "efficiency" be our most important criterion? "Efficiency" is why we have lean staffing that forces the "bus factor"[0] of nearly every position to 1. "Efficiency" is why we have stretched, fragile supply chains that break down the moment any one supplier or middleman experiences a crisis. Efficiency is the enemy of resilience. The only people it really helps are shareholders. [0] https://en.wikipedia.org/wiki/Bus_factor https://en.wikipedia.org/wiki/Bus_factor

consteval 2y ago
Efficiency is perhaps not the right word, but I was under the impression that attempting to measure performance is a cost saving tactic. I think you lose the cost saving if you just double your engineers while keeping the same amount of work.
- danaris 2y ago
  I mean, from a certain perspective, nearly everything management tries to do is a cost-saving tactic, but I wouldn't say that measuring performance is primarily about "cost savings". It's more about making sure that people are doing the job they were hired to do—which, on the face of it, is definitely a valuable thing to be doing. The problem is when it's being used as a whip to get people to work ever harder with no more pay and no more support than before. Doubling engineers while keeping the same amount of work is something that can increase resiliency—both in the engineers themselves and in the business. Making sure you have at least, say, 125%-150% coverage of all duties makes a huge difference in everyone's stress levels, and makes sure that if someone gets sick, gets hit by a bus, or just goes on vacation, their duties neither stop getting done, nor get piled onto other people who are already at 100% (or more) capacity.