254 karma · joined August 11, 2017
I would much rather some native syntax like K<T> where K is some generic though.
I think this boils down to a roll of the dice. Its possible that some code is simple enough, or some engineer is smart enough, that tests are not really needed, but given the choice to not unit test would be made early, and its very hard to know the complexity of the solution at the beginning, you're really just rolling the dice with regards to whether it would work out. Sometimes you get a huge payout by delivering good code fast, but sometimes it can really blow up in your face with lots of prod errors.
Personally, I'd rather smooth out that large payout vs large explosion curve by just writing tests. If we could have pushed out the code faster, that is a loss, but if the code would have had many errors, thats a gain. So probability distribution with regards to payoffs is smoothed out.
For example, even though passing a function to an SQS listener is many abstractions away from handling internet traffic from SQS, I'd much rather the parsing of bytes be hidden in a place where I never have to see it and I can focus on business logic that actually provides value.
It goes without saying that 60 minutes is too short a time to ask someone to solve a real life business problem, but real life business problems tend to be broken up into many small problems, which can in turn be simplified and presented to the interviewee.
The discussion I've been having with colleagues is how to do precisely this - eschewing the typical leet code/hacker rank problem, what problem can we come up with that has high fidelity to real life work, without being too complex for the 60 minute interview? I definitely think this is possible, and normally you get to that kind of question by working backwards - what is a real life feature I'd implement, and how can I ask a version of that during an interview. One example would be a graph of similar categories. Categorization is a common issue, and graphs are a common data structure.
Side rant - I've seen a lot of hate for things like data structure questions and algorithms associated with data structures. It's definitely possible to have an entire career of programming where all you do is build CRUD endpoints and design relationships between objects, but sitting down and learning about these amazing data structures that have been refined over decades of work can seriously improve a lot of code. A common problem I've seen is people building rigid object hierarchies (by composition, not inheritance) where a tree would have been way more flexible and easier to understand.
The primary point of the article is that there are very common inefficiencies in our industry that, if we tackled responsibly, would greatly reduce the turn around time of producing changes to the code base.
The point I took away the most from is how much having a single point of change for a single concept and periodically cleaning the code base to keep it this way can dramatically increase productivity.
This is not to defend Japan's actions, they did totally unjustified things in China, and as it turned out after Japan modernized, the west never posed a threat (as WWII would wipe out the Europeans ability to colonize and the US had no interest in it.
The point being, now our perspective on history is very different. Colonization is a thing of the past. Respecting borders is by and large an unimpeachable norm across most of the world. I don't think Japan will feel the same pressures.
Also bit of unsolicited advice, never think you are out of the learning phase of your career. There are few things more frustrating than an engineer who has been working for 5 years, think they no longer have to learn, and refuse the entire team's attempts to move them from a bad idea.