12

What is it with devs who try to bloody "cost optimise" everything to within an inch of its life when there's no reason to do so?! This ain't your personal pocket money project here. This is a real commercial app with real consequences.

Seriously, saving £100 a month might seem like a lot to you, but this is a multi million pound project we're talking about. That's bloody nothing, and no-one will care. If a Fargate spot instance restarts at the wrong time and causes downtime though, or if we need logs going back a week, and don't have them because the log retention period is a few days, then everyone will be royally pissed. All because you thought "it should be ok", or it "seemed like the right thing to do". Sod off.

Comments
  • 3
    We don't care about the money, we want to improve the user experience and start making services that once more don't take 20 minutes to show a page of results or add an item.

    At least that's my motivation when optimizing, thought I'm also against extreme optimizing. Just some basic ones, like batching n stuff
  • 3
    While I think even $100 can be worth saving it all depends on the potential consequences.

    If it can potentially cause a major cost under the wrong circumstances its most likely not the right thing.

    But if it at worst cost the same or similar amount in the wrong circumstances then it might be well worth it.

    Making a hundred such savings a year adds up and if many employees do it it can really add up.

    But saving on critical things are rarely the right thing.

    And if the saving ends up hampering customers or preventing sales its also bad.

    So its not always right or wrong you need to have context

    But it sounds like your dev have the wrong excuses for saving ;)
  • 2
    Priorities.

    Optimization vs Cost (long -/ short-term) vs Human Resources vs Benefits vs Complexity.

    If you save 100 bucks, but codebase looks like dysentery festival, wasted one employees working day x 7…

    Yeah. No brainer. Long term and short-term costs explode and maintenance will be a pain in the butt.
  • 1
    @Voxera Yeah, there's cases where it would be totally valid to save that kind of cash. I'm always open to suggestions if it's a significant saving, very low risk, or the app simply isn't that critical. When an issue is going to cost the company in the order of tens of thousands though, and the change could genuinely introduce stability issues or make debugging / analysis harder - no way.
Add Comment