To be fair, almost all the problems after 6 are just variations on BFS and DFS, just different ways to ask the question.

For example, (11. Check if an undirected graph contains cycle or not) can easily be solved by just performing a DFS with some exit conditions.

> Love graph algorithms have coded many myself in JavaScript to prep for interviews.

...but have you coded any in JavaScript after the interview as part of the job? I suspect for most people the answer is no. It has always seemed strange to me to ask a topic that you're never going to encounter during your day to day work (yes, I know for some roles this isn't true).

I think there's a general consensus that mastering the specific material you would encounter in an undergraduate CS program proves you have the more general cognitive skills necessary to succeed in any entry-level programming job.

We just don't trust the university to say that you've done the projects and sat the exams and demonstrated competence (what it's doing when it issues a degree), so each company unwraps that abstraction and runs its own projects and exams on the same material. But it is fundamentally based on college curricula.

Trusting universities is likely out of the question. Eventually, I expect the industry will create something like the Bar Exam so each candidate can pass a trusted test once upon graduation (or even without going to college), rather than creating the burden for each company to do all that grading.

Lawyers and doctors don't​ do LSAT and GMAT questions on the job. It's not a problem specific to tech. An interview is a performance test. It's popular to say an interview should be like a day at work, but when we performance test our code, we don't throw an average load at it and say it's ready.