And likewise, it's easy to keep a service "Up" if you don't deploy any code changes. Unfortunately, feature launches, bug fixes, one-off backfills, and all the other nonsense that comes with actually running a business tends to require deploying code.
You can fire everyone and fully re-staff the firm, and eventually get to the point where you can do all of those things safely. But it's going to take a hell of a long time and downtime to shake out all the things that can go wrong in the myriad of unplanned interactions between different services.
I don't expect Twitter to crash and burn in the sense of the website being down and not coming back up. But I do expect feature velocity to grind to a halt, until all of these self-inflicted wounds are staunched.
You can fire everyone and fully re-staff the firm, and eventually get to the point where you can do all of those things safely. But it's going to take a hell of a long time and downtime to shake out all the things that can go wrong in the myriad of unplanned interactions between different services.
I don't expect Twitter to crash and burn in the sense of the website being down and not coming back up. But I do expect feature velocity to grind to a halt, until all of these self-inflicted wounds are staunched.