My whiteboard coding interviews now mostly consist of asking someone to write a for loop with one state variable. That's my bar for low skill. If you can't write 6 lines of code in 45 minutes with one loop and one variable, then you and I are going to have problems working on logic problems on a whiteboard on the actual job.
I've interviewed enough people who don't even know the syntax of a for loop for the language that they chose to use.
As a general matter for how the industry as a whole interviews, I agree though; I am no longer a fan of tricky algorithmic or esoteric data structure questions in whiteboard interviews. The problem I find with other people's interviewing questions is that most people are still asking things that are way too difficult, and they don't end up judging anything other than "Did this person just cram hundreds of hours of algorithms questions?" or "Did this person get lucky and happen to know this type of problem ahead of time?".
For better or for worse, my company schedules me for like >90% system design questions now, which I prefer anyways. Start with a very simple problem and system, and then just throw wrenches.
I'm only 90% sure I know the syntax of a for loop in my chosen language, C#. The reason for the missing 10% is that I never write this syntax.
Most loops are covered by a foreach or by LINQ (think: map and reduce, buy nicer).
The few times I actually need a for loop I type: "for" <press enter> <type the limit for i> <press enter> and then I focus on the loop body. If my needs are more complicated, I _modify_ the for loop my IDE gives me.
I've interviewed enough people who don't even know the syntax of a for loop for the language that they chose to use.
As a general matter for how the industry as a whole interviews, I agree though; I am no longer a fan of tricky algorithmic or esoteric data structure questions in whiteboard interviews. The problem I find with other people's interviewing questions is that most people are still asking things that are way too difficult, and they don't end up judging anything other than "Did this person just cram hundreds of hours of algorithms questions?" or "Did this person get lucky and happen to know this type of problem ahead of time?".
For better or for worse, my company schedules me for like >90% system design questions now, which I prefer anyways. Start with a very simple problem and system, and then just throw wrenches.