Why the fuck would CS graduates be afraid of DP problems, NP-completeness or approximation algos??? That’s literally the table stakes of our profession.
The problem is rewarding rote memorization in a whiteboard interview, at the expense of actual understanding, and ability to research the problem.
> Why the fuck would CS graduates be afraid of DP problems, NP-completeness or approximation algos??? That’s literally the table stakes of our profession.
If the profession being discussed is "academic work in Computer Science", sure.
If it is "software development", those things absolutely are not really "table stakes".
If it is "software engineering", then I don't think there is broad consensus on what that profession even is, much less what table stakes in it are.
> Are you seriously claiming Guido van Rossum was hired for software development/engineering..?
No, nor do I think he was he hired for the other thing I discussed, academic computer science. The post I was responding to made a general comment about "CS graduates" and "our profession"; I was responding to that. Whether that post itself was material to, or merely tangential to, the discussion of GvR's hiring at Microsoft is an argument that, while perhaps interesting to some, was not the focus or concern of my response.
Because these problems can have some subtleties that are hard to get right in a high-pressure environment. Some interviewers will completely write you off for small mistakes.
That sounds interesting! Do you have any examples?
I would have thought that if you actually needed people to perform under pressure, you would design your test explicitly around that, instead of using “comfort with the whiteboard” as a proxy...
That might be a fair statement for some roles, but what about the second part where you code it up and write the code on a whiteboard w/o running/debugging and have to get it right in 45min?
The problem is rewarding rote memorization in a whiteboard interview, at the expense of actual understanding, and ability to research the problem.