Jim Roskind had a really nice talk [https://www.youtube.com/watch?v=_uaaCiyJCFA] at AWS Reinvent 2022 which highlighted the fence post problem. Basically, AWS was targeting P50 and P90 latencies in many services, and sure enough there were big clumps of latencies right under those target fence posts as engineers were basically gaming the metrics and nerfing the >P90 latencies to make it happen.
I don't think that counts as gaming the metric. It's an example of a metric that even when targeted in good faith results in consequences management didn't intend.
Good faith starts with understanding that "90" in P90 is an arbitrary number, and that the exact sampling threshold is not important to why anyone cares about that metric.
The more interesting part of the talk is how using what he calls "fenceposts" naturally results in bunching up against the fenceposts because even small regressions that cross the fenceposts are blocked but regressions of any size that don't cross a fencepost go unnoticed.
It's just a plain bad metric. Useful due to brevity though.
A decent metric for testing against might consist of a target mean, the sigma specifying a gaussian about that mean, a minimum sample size, and a hard upper cutoff.
I'm of two minds about this. Facebook and Instagram keep me connected with friends, new and old. True, they're filled with algorithmic crap now, but when I tried ditching them for a while, it didn't improve my connections with people. If anything, I lost touch with people because that's where they were. I'm much better off knowing what my friends -- many who have scattered -- are up to, and giving casual updates myself.
On the other hand, yes, absolutely put the phone down at the dinner table and engage with those who are present. No questions there.
Finally, my kids are getting to the age where their friends have phones and use them to communicate. My kids don't yet have that and are reliant on me to text parents, which is a lot of friction for everyone involved. Their summer is very boring because they hardly get to see their friends. So, sure, fight boredom by entertaining yourself, but that's a lonely existence.
Note that by leaving them by the front door I don't mean abstain. I relent that it is necessary for communication now we don't use letters or have landlines, but I need to actively walk to the front door if I wish to do that. Notifications are something I check when I leave the house in the morning, or briefly after the gym in the car perhaps, rather than being all I do for the majority of the time I sit in the house. The act of it being at the entrance and allowed no further is enough.
Sadly, this bill has been around since 2007, introduced by the very same Hank Johnson! I would love if Congress would put an end to the forced-arbitration bullshit that so obviously does an end-run around your 7th amendment rights and makes a farce of informed consent, but alas, they couldn't even get it through when Democrats held the presidency, the house, and 60 seats in the senate.
The big two, x86_64 and arm64, have 64-byte cache lines, so that's a reasonable assumption in practice. But I was surprised to discover that Apple's M-series laptops have 128-byte cache lines, and that's something a lot of people have and run, albeit not as a server.
Yup, this is how stuff ends up getting lost and delayed, even if the crews on either side are competent and efficient. Most people can't fill up an 18-wheeler, so some bin packing has to take place for the long haul.
I've done a couple of cross-country moves with a full house and had fairly good luck -- North American Van Lines if anyone is interested -- and by far the best experience was when my stuff was last into the truck making the long distance trip. It was delivered a few days later by the same truck and driver. Nothing was lost or broken. There was no chance to screw things up with intermediate transfers to depots.
"Work 5 shifts per week to monitor the NERSC HPC Facility, which includes 2 - 3 OWL (midnight - 8am) shifts. Some days may be onsite, some may be offsite. The schedule will be determined by staffing needs."
40 hours per week, full salary, full disclosure about the night shifts, but none of this 24x7 wake up in the middle of the night on top of your regular job bullshit that the tech industry insists on.
there's nothing wrong with shift work, and there's nothing wrong with the graveyard shift (for people who want that job), but there is a LOT wrong with alternating day and night shifts in the same week every week, and nobody should agree to do that. (perhaps if you are in your early twenties you feel like you can handle it, but I'm not sure that's a good idea either)
I tried graveyards and swing shifts in my twenties. I only did it for a year, but it wrecked my sleep schedule for years afterward; I struggled with insomnia because of it. Not sure there is any age where this works.
Some people can handle it better than others. I'll never do it again though.
some of the people I know who've made it work, and I'm not recommending this either, are "hardworking ambitious immigrant" types who do it to hold down two full time jobs. tends to be in industries where the overnight shift is a quiet shift rather than something like running the same factory at full tilt overnight.
My natural rhythm makes me a nocturnal being, while the society works 9 to 5. Once I realized this, it became obvious why I feel like fuck all the time. Can't wait to retire.
Shift work sucks more than on call for me. I like having a flexible schedule that I can work whenever I want, and I'll happily take the remote risk that I'll get paged during my one week per quarter on call rotation in order to guarantee that no one expects me to be on during set hours the rest of the year.
I think the problem isn't on call itself, it's that a lot of companies suck at on call. If your rotation is every 3 weeks and you get woken up in the night at least once per rotation, then yeah, that's awful. But the problem you have in that case is that stuff is always on fire and you don't have enough people on the rotation, not with on-call as a concept.
I'd take a night-shift job over being on call ontop of a 9-5.
I think the real answer here is to have team members on multiple continents. Have some team members NA/EU/Asia then it's always reasonable hours for someone to deal with the production problem. High priority issues can be worked on around the clock without anyone working overtime.
That's still shift work. It's still the company assuming that I'll be available during N specific hours of the workday in order to fix issues.
Look at it this way:
If I work a job where I'm expected to be on 9-5, 46 weeks per year, 40 hours each week, that amounts to 1840 hours of scheduling my life around my employer.
If I work a job where I can schedule my work however I like and also have on call 1 week out of every quarter, the worst case scenario there is 672 hours of scheduling my life around my employer (and in practice the demands of on call in my current rotation are far less than that). The rest of my life I can schedule as I please, so long as I do my job.
I would rather take the option that minimizes the number of hours where my employer gets to tell me where to be.
I know plenty of friends (mostly medical) who've had "on call" shifts on top of their 9 to 5, but in those cases it was pretty exceptional for them to actually receive a call, and the "interruptions" would be a small number of hours.
"This was the first leadership interview loop in the past 12 months (20+ companies) in which anyone had asked me to write code."
Interesting data point for the question: at what point in your career will you stop being asked to write code on a whiteboard to prove you aren't lying on your resume.
I’ve never been through an interview cycle where candidates didn’t lie about this, so I assume pretty high. It always boggled my mind that people would claim programming skills but not be able to write hello world.
If its not the pressure cooker, this is trick question and i am going to blast you for missing the return and void in parenthesis types question then we can partially blame the candidate.
Many of us have faltered like this once or twice unless its an outright lie kind of situation.
I’m not a fan of trick questions. It’s usually something as simple as “you have 10 minutes write out a program in any language you like that takes in a string and prints out the reverse.”
The idea is just a litmus test and I don’t think it’s useful for filtering anything other than liars. I never had someone completely freeze up.
You got 21 recruiter messages _with salary_?? Amazing.
I have had almost no luck getting that information from recruiters without hopping on a call. None have ever provided that information upfront.
I only managed to get that info from an Amazon recruiter after pushing the issue pretty hard. At first he bragged that Amazon had upped their bands, without mentioning any numbers. When I asked how much, he pointed to a reddit thread and levels.fyi, still not mentioning any numbers. I finally straight-up asked him to stop being coy and just name the range, after all it's law in California and I'm going to keep asking until I get an answer. It took him a couple of days to respond, then I never heard from him again.
And this is the only recruiter that has ever actually given me an answer. Ever.
