5 ms·
> accept risks from intentionally violating the spirit of that limitation. What risk? Since it's the engineers, not the executives that take any oath of profes
by alxlaz 6y ago
> accept risks from intentionally violating the spirit of that limitation.
What risk? Since it's the engineers, not the executives that take any oath of professional conduct, there wouldn't be any risk for any executive. All this would get us is a legal framework for throwing engineers under the bus via an ethical commission if they do something silly like blowing the whistle on an unethical decision higher up in the hierarchy.
I'm very much for greater personal responsibility in the field of software engineering. Until not too long ago, I used to work a in a field (medical equipment) where accepting the burden of potentially catastrophic mistakes came with the job. But I also know -- based on that same experience -- that personal accountability is meaningless without organizational ability.
Unless this hypothetical "Hippocratic Oath for engineers" is backed by a "Hippocratic Oath for executives", a "Hippocratic Oath for product managers", and "Hippocratic Oath for engineering managers", (edit: or by a legal framework that requires companies to enable it) all it'll do is reduce the PR effort involved in cleaning up a mess like Volkswagen's emission test scandal to pretty much zero by providing an exceptional -- and very mythical-sounding! -- framework for scapegoating.
- austincheney 6y ago> there wouldn't be any risk for any executive This isn't a real problem. I take it you have never worked outside of software. You don't need a law license to be the business manager of a major law firm or a medical license to manage a large hospital network. How often does such problems occur in major law firms or large hospital networks? In other words this is a hypothetical complaint of some future condition that doesn't exist in order to avoid accepting responsibility in the present. EDIT: Also see: https://en.wikipedia.org/wiki/Sarbanes%E2%80%93Oxley_Act https://en.wikipedia.org/wiki/Sarbanes%E2%80%93Oxley_Act
- sobellian 6y agoIf a lawyer does something unethical, they could be disbarred. If a doctor does something unethical, their license may be revoked. These end your career. There are already professional organizations for developers and engineers. ACM's ethics code even includes, "avoid harm." You don't need to be in any of them to write software. We would need to pass laws dictating that Facebook et al can only employ licensed software engineers. Do we really want to impose that level of gatekeeping on the profession? The important part is that there's an accountability structure outside the company. Without that an oath just consists of words on paper.
- austincheney 6y ago> These end your career. Agreed: https://news.ycombinator.com/item?id=24106617 https://news.ycombinator.com/item?id=24106617 > You don't need to be in any of them to write software. Then they are irrelevant. > Do we really want to impose that level of gatekeeping on the profession? Yes, absolutely. Everybody benefits except developers who probably shouldn't be writing software professionally in the first place.
- ScalaFan 6y agoThe key to lawyers and doctors losing their licenses to practice is that they have professional bodies that will police/enforce the rules. However, they also have a monopoly to provide those services. I suspect corporations don't want limit software engineers to those with licenses only as there would be less of them available and costs of said licensed software engineers would increase.
- austincheney 6y ago> I suspect corporations don't want limit software engineers to those with licenses It comes down to money, which includes liability. I suspect licensing won't ever happen in software unless law suits increase and licensing becomes a deterrent, at which point licensed developers will cost less after factoring for risk assessments. There are all kinds of unintended benefits for everybody that come from limiting negligence, which pays for itself. As a counter argument who benefits most from allowing the hiring and practicing of negligent developers?
- alxlaz 6y ago> I take it you have never worked outside of software. You take it wrong, and reasoning by analogy with other fields is a very risky line of logic, because they're not at all similar in terms of internal regulation and professional autonomy. First of all, in fields like medicine, that kind of responsibility is necessary because clinical autonomy is far wider than any kind of professional autonomy that software engineers have, at any level. And it works precisely because this autonomy doesn't come only with the responsibility of adequate care, it also comes with the ability to say "no" without fearing for your employment or license. Physicians enjoy a great deal (if faltering, and arguably still insufficient!) of protection for defending the interests of their patients, and firing someone for it is a pretty risky and expensive affair. Software engineers enjoy pretty much zero protection in this regard. If you want engineers to bear that kind of responsibility, that's great, and I'm all for it, too. But in order for it to have any use, you also have to give them the means to defend the standards of professional conduct they must now conform to. Are you also prepared to: * Defer to an external body of experts to decide if a technology is fit for commercial use or not? * Forbid an engineer from arbitrarily overriding another engineer's bugfix? * Have product managers defer to an ethics commission, ran by engineers and engineering ethics specialists, when deciding what features to prioritize or introduce? And if the ethics commission says "no" to a feature, then it's a "no" that not even the board or the executive team can override? * Make it illegal to ask engineers to introduce a workaround with known flaws, rather than the proper fix, without explicitly informing customers about the known flaws? * Make it illegal to fire engineers for not executing a manager's order if that order violates the engineers' standards of conduct? * Defer to an external engineering organization to decide who even gets to call themselves a "software engineer"? Second, one of the reasons why managing large hospital networks by people without a medical license works (more or less...) is that, for example, there's usually no legal way for the CEO to walk into an OR while cardiac surgery is happening and request the surgeon unplug the CPB machine, even if that means killing the patient. Even if there's no internal regulation being broken, criminal law is more than enough to keep that sort of thing in check. There's hardly anything that would currently prevent the CEO of a software company from walking into an office and requesting that an encryption feature be backdoored, for example. Without the kind of autonomy and legal protection that would enable software engineers to defend such an oath, its provisions would be completely meaningless. What kind of benefits do you think the software industry, and society in general, would get from an oath that's more risky to defend than to break?