Hacker News Reader

This is a good example of why it is useful to have more than one person or team work on the same problem in parallel. Yes you duplicate effort but you often come up with better solutions especially if the teams regularly meet to compare notes.

As long as their "career performance/review" doesn't depend on making their solution/big project work. Otherwise, you're setting either for failure and creating a culture averse to sharing, collab and positive but passionate debate.

How detached does your management need to be in order to f this up? You’d obviously have to coordinate this in a large corp.

You'd be surprised how common this is. You gotta remember the managers are people playing the career game too. And if it's one of the large corps with a lot of turnover, they have to move fast.

Speak with people who work in large corps and you might here similar stories. Problem is people online naturally can't go too deep to avoid doxxing.

#2 is the biggest problem I've faced. It's nigh impossible to truly revamp what you have while also having to sell/serve/use/change the existing product. You either have to leadership who is willing to take the hit and stop new features/bugfixes from coming in while you're redoing the product, or you have to pad timelines for new features/bugfixes but that's risky too because you can be seen as someone moving too slow

> But there was a huge positive as well. Without worrying about breaking the original use case, we could hack on the new tool all we wanted without regard for stability or reverse compatibility. We got a thorough understanding of how to serve the new use case and why the original tool was designed the way it was

Maybe good for your understanding, but this is pure waste.

Enhancing understanding is not waste.

This is no more wasteful than discarding a prototype, the only possible problem here is putting too much effort into throwaway code.

How is more thoroughly understanding an important service pure waste? I could agree with inefficient, but definitely not pure waste.

What did thoroughly understanding it produce? It consumed a shit ton of resources - what was the product?

What is software development if not thoroughly understanding the problem? Implementation is rarely the hard part