5 ms·
That is an assumption that is not stated or implied. But, if you want to go down that tangent the tech industry would benefit from a set of standard minimum qu
by austincheney 5y ago
That is an assumption that is not stated or implied.
But, if you want to go down that tangent the tech industry would benefit from a set of standard minimum qualifications. Such would reduce employer liability and result in higher quality products pushed to market in shorter time at lower expense. That last sentence is technobabble really meaning employers could more easily discern candidate maturity and thus require a lower headcount for an equivalent task. The great discriminator there isn’t financial but the time and cognition required.
- satellites 5y agoI've had this argument on HN before and I'm generally against that - I don't think the ability to pass a standardized licensing test is any more of a marker of a good engineer than an interview is. If anything, it's less effective than a good interview. I've never taken a test in all my education that I felt was truly and fairly assessing my skills as a software engineer. And I went to a well-ranked CS program. Long-term projects, and the ability to talk in depth about the technical aspects of those projects, is the best marker of ability to me. But, if a testing system were proposed that could demonstrably add value to the industry and wouldn't select for privilege, then I'd be for it. I'm just very skeptical, and I think lots of licensing programs are onerous and only exist to enrich the licensing organization without adding real value to society.
- austincheney 5y agoHave you gone through a professional licensing program? Nearly 100% of the people on HN that make this argument have not. The big difference is that a licensing process is objective in a way an interview can never be. That alone adds value in a way you cannot possibly imagine until you have worked in a different industry. There is no greater institutional barrier or discriminator than implicit bias.