4 ms·

There need to be real consequences for this kind of nonsense for people. Liberal estimate of missed earnings multiplied by 10 paid to the victim. Some fininte

• renewiltord 6y ago
  Perhaps if we shot everyone who wrote a bug there would no longer be any bugs.
  • harry8 6y ago
    Perhaps if we program our computers to commit crimes we should be totally beyond the reach of the law. See, snarky comments like that are easy, not particulalry clever, amusing nor helpful and bring down the tone of discussion. Maybe make your point a little better than that here?
    • renewiltord 6y ago
      "Off with their heads" comments are precisely at the level of discourse where my response is adequate. And you've outlined many good reasons in that second paragraph for that.
      • harry8 6y ago
        Financial implications as a consequences of committing theft with jail for repeat offence is actually the norm in civilized society. It's pretty radical to suggest that we have a situation where you should avoid consequence for it. Where on earth are you seeing "off with their heads?" Not from me. Are you really supporting financial incentive to make and not fix errors in your own favour? Should we apply that principle to bank tellers and shop assistants too or only if a computer is involved? Does it need to be a statistical algorithm? Or is an if statement good enough to shield you? I think maybe you're having a completely different argument with someone else?
        • renewiltord 6y ago
          Oh yeah, I'm not buying what you're selling. There isn't any theft here. And "in your favour"? Give me a break. Failing to complete payments is not in PayPal's favour, and I don't even mean some "in the long term no one will trust" crap. They literally make money off transactions. Transactions that cannot be completed do not make them money.
          • harry8 6y ago
            It's been my experience that paypal take the money and then sit on it as a matter of policy. That's theft. Taking money under false pretense agreeing to a contract to perform a service then wreaking havoc on a business not performing it with no ability to even discuss the non-performance. Do you prefer the word fraud for that? In principle, imagine your government wrecking your business doing the exact same thing. Hey why not the IRS! Equality before the law. No consequences. No restitution. The algorithm has spoken. Because your principle has been established and you have to stick to it.
  • ithkuil 6y ago
    Indeed, bugs will become features! (Goodhart's law)
  • knolax 6y ago
    Manslaughter's a capital offense in some places, so yeah if your bug kills people a la 737 Max or Therac-25 maybe you do deserve to be shot.
    • renewiltord 6y ago
      Yeah, but it's no surprise that those places of yours didn't birth Boeing or Atomic Energy of Canada.
      • NoOneNew 6y ago
        I sense you have a deeper issue and/or gripe. Care to elaborate?
        • renewiltord 6y ago
          In the sense that a bunch of people who haven't read anything around organizational dynamics making outraged comments annoys me, yeah, I've got a gripe. All these comments are like 1960s level of conversation around safety engineering. Everyone thought back then that if you just ramped up punishments, things would get better. Turns out they didn't. Anyway, to save you from getting it from an inferior source (me), I'll give you some good ones: - The Field Guide to Understanding Human Error by Sidney Dekker - The Challenger launch decision : risky technology, culture, and deviance at NASA by Diane Vaughn You don't have to but you can also read how the FAA only pursues civil actions in severe cases. The equivalent penalty for a thing like this would be the lowest level action, either administrative action (pretty much a notice and a record on a file) or require proof of correction. Maybe the thing we should actually issue fines for is commenting on subjects without first reading about how they're handled well. Maybe 10x the hourly wage of each person who read it. Yeah. That'll be good. Comments will probably be great then.
          • harry8 6y ago
            Again you are arguing against a case that has not been made. A straw-man. There were consequences for the Challenger disaster - there was a full, televised enquriy. Feynman was there. That's a pretty deep and expensive consequence. Obviously NASA differs from Paypal. Beyond that I'm pretty sure those involved would feel consequence about their colleagues having died. Total sociapathy at that level is relatively rare in my experience. This is a quite different thing to make peace with to screwing some very small percentage of customers. To apply exactly what I said to _this_ case the consequences are: Estimate the loss to this person, 10k seems like a reasonable liberal guesstimate but yes, actual data would be better to get a harder number. Paypal pay her $100k as compensation. She's happy and paypal have a cost function to optimise sesibly. That's a start on proper incentive. Then if they do it more than N times then they're not taking the consequences seriously and they are only optimising that cost function considering breach as merely a business expense with no regard for ruining people. Someone needs to have a real, non-financial consequence for making a decision like that wittingly or unwittingly. The CEO is good because the buck stops there. It's his or her job to make sure that kind of "we'll screw over this many people because it's in our financial interests not to fix that" does not happen. Now we can discuss what the value of N should be. Should it be 10, 100, 1000? What should constitute a breach? Nowhere is a guillotine part of this. I can't actually see /anything/ you've argued a part of this. I see the outraged comments as being yours. You seem utterly outraged at my holding an opinion I do not hold and did not advance.
  • Nextgrid 6y ago
    Shooting is an obvious extreme but wouldn't you agree that there has to be some consequences to try and prevent bugs that have major financial consequences? I think it's totally fair that if your company's development process allows bugs like this to happen then it must have some consequences to discourage it from doing that in the future.
  • Barrin92 6y ago
    there's very harsh punishments for physicians or say biomedical or military researchers or engineers who fail to do their due diligence yet we still have those. I think it says enough about the industry that there's so many people who are dead afraid of rigorous quality standards.
    • renewiltord 6y ago
      The magic here is that software engineers are some of the best engineers. We kill barely anyone in comparison to other engineers. We don't need standards because even without them, we don't kill anyone at anywhere near a few orders of magnitude near those other fields.
      • Barrin92 6y ago
        I don't know if this is sarcasm but that might be because most engineers don't build anything more exciting than SaaS dashboards and it's pretty difficult to kill someone with an angular component. Given the amount of software outages on this planet, if software engineers were building planes they'd collectively fall out of the sky once per week
        • renewiltord 6y ago
          Ah but that's how you know we're good. If you can't do it right, don't do it. And we know when we can't do it right. Sadly aeronautics and structural engineering haven't caught up to the level of ethics that software has. They build bridges that fall down and planes that crash. But I know I would crash a plane or break a bridge, so I don't make them. Much more ethical to not go around killing people like those guys. Killing people just because you want to build something "more exciting". Enough to make one's gorge rise.