Re:Ask HN: What do companies WITHOUT a whiteboard interview ...
by muzani on Saturday October 10, 2020 @11:45PM
(#24743572)
I never hire or have been hired from whiteboard interviews. They seem irrelevant in our situation - we're hiring people to write bug free code that works but is being optimized by someone more senior.
One of my favourite is architectural discussions. Tell them how the app is and ask them how they would do it. Show them a screen we plan to make in the app: How would you implement pull to refresh? What about this complex layout here? What about pop up dialogs? How do you prevent the pop up from crashing the app if a phone call comes in here? How would you suggest caching on this kind of screen? Now assume with have an architecture of X, how would you implement a cache within 4 hours without changing the architecture? Okay, on this page we'd like to take photos and make sure they're not tampered with; you can ask help from the back end to verify but keep the workload to a minimum.
The good answers will propose a solution. The really good answers will offer some solutions and say which don't work and why, and what factors might sway the decision. Acceptable answer is googling and showing the thought process of selecting/rejecting that solution.
I think the only way to really prepare is to get good at making apps, maybe browse subreddits on the topic.
One of my favourite is architectural discussions. Tell them how the app is and ask them how they would do it. Show them a screen we plan to make in the app: How would you implement pull to refresh? What about this complex layout here? What about pop up dialogs? How do you prevent the pop up from crashing the app if a phone call comes in here? How would you suggest caching on this kind of screen? Now assume with have an architecture of X, how would you implement a cache within 4 hours without changing the architecture? Okay, on this page we'd like to take photos and make sure they're not tampered with; you can ask help from the back end to verify but keep the workload to a minimum.
The good answers will propose a solution. The really good answers will offer some solutions and say which don't work and why, and what factors might sway the decision. Acceptable answer is googling and showing the thought process of selecting/rejecting that solution.
I think the only way to really prepare is to get good at making apps, maybe browse subreddits on the topic.