Comment by AdmiralAsshat

> Had you let Carmack talk to the lead engineer for about fifteen minutes, I'm sure he would've come around entirely for objective reasons

I wish this were true. Yet I've been so many projects around me go months and even years down some very-clearly wrong direction, because the engineering lead refuses to course correct. Typical patterns are:

- The system design attempts to address "future" issues, making it two orders of magnitude more complex than necessary.

- The eng lead wants to "get it right from the beginning." So months later, you have a beautiful CI pipeline, coding standards, base classes, etc, and still no usable prototype.

- Trying to solve all possible use cases, with one system.

- The above, but then focusing on just one use case so narrowly, the rest have to be crammed in.

I could keep going. These are cases when you need a decisive (and technically sharp) manager above to call bullshit and rip the bandaid if necessary.

And this story leaves unanswered the question of why they compromised on a suboptimal architecture in the first place. In my experience, 95% of the time it's due to some other constraint imposed by an executive. E.g., a made-up ship date.

So it's perfectly possible that had the head of graphics gone for what he knew was the ideal solution in the first place, he would have been yelled at by Jobs for that. It wouldn't be the first time that a HiPPO caused a boss to do a 180 and implicitly blame the underlings rather than own up.

That's actually what happened according to the story. The lead admitted that Carmack was right, and then Jobs got angry at the lead for not making a good technical decision from the outset.