• sj_zero@lotide.fbxl.net
    link
    fedilink
    arrow-up
    42
    ·
    1 year ago

    The worst is when the problem is something that will only manifest at scale, so your bosses are going “You should have tested it before putting it in production!” And I’m like “DO YOU THINK I DIDN’T TEST IT BEFORE PUTTING IT IN BLOODY PRODUCTION!?”

    • Windex007@lemmy.world
      link
      fedilink
      arrow-up
      25
      ·
      1 year ago

      The worst is when you’re actively raising alarms that your testing infrastructure has glaring, easily pluggable gaps that would catch certain types of common issues, and you explain that if you could have a few extra days you could get that done… And management says no, it’s not worth it, and THEN what you said

      • Raltoid@lemmy.world
        link
        fedilink
        arrow-up
        19
        ·
        edit-2
        1 year ago

        Always remember the magic words when it comes to corporate bullshit: “Can I get that in writing?”

        • tool@r.rosettast0ned.com
          link
          fedilink
          arrow-up
          4
          ·
          1 year ago

          I don’t even ask for that anymore because it rarely leads to good ends. What I do now is send an email summarizing the dumb bullshit that they want me to do, describe the detrimental effects that it will have in excruciating detail, ask if there are any corrections and if my understanding is correct, and say that if I don’t get a reply from them by X time, I’ll do $DumbBullshitThing at Y time/date. It gets CC’ed at least one level higher than them in the food chain and also to my personal email address for CYA.

          It puts the onus on them, creates a paper trail, and also places the blame on them when shit blows up because they asked me to do $DumbBullshitThing when the consequences were clearly laid out.

  • TheGreenGolem@lemm.ee
    link
    fedilink
    arrow-up
    12
    ·
    1 year ago

    Once I heard a (maybe) quote: “If it works on your machine, why don’t we just put it in the fucking datacenter?”

    • Johnny@feddit.de
      link
      fedilink
      arrow-up
      14
      ·
      1 year ago

      As long as you don’t change host platforms…

      There are lots of things that can break in Docker between Windows and Linux. Not to mention ARM and x86

      • tgv@programming.dev
        link
        fedilink
        arrow-up
        1
        ·
        1 year ago

        Even between Linux on x86 and Linux on x86. Thing was running fine for months, I make a small change, test, deploy to staging (same machine as prod, so that helped), and suddenly some library deep in the bowels of the app says that it can’t start a thread. Runs on my machine though. Turns out to be some incompatibility that had to be fixed by excluding seccomp. Version updates can be a bitch.

  • lugal@sopuli.xyz
    link
    fedilink
    arrow-up
    3
    ·
    1 year ago

    And then it turns out the code on the machine isn’t the same as in git and one line has less intend which kicks the function out of the loop and that’s why “only the last input is imported”

  • Da_Boom@iusearchlinux.fyi
    link
    fedilink
    arrow-up
    1
    ·
    1 year ago

    At work we’re converting our Dev environment to docker… and for about two weeks “it works in production” was a fact of life.