In one particular case when a process will live for a second or two, it's indeed a visible delay. But it's just one of many scenarios and it's definitely not th

In one particular case when a process will live for a second or two, it's indeed a visible delay. But it's just one of many scenarios and it's definitely not the top priority one for applying all those engineering resources. When you build a server application, you just don't care about startup time much - in situations where it matters, green-blue deployment will do a better job to reduce downtime (and, anyway, it's not always the platform startup which contributes to the delay before full availability).