If Starfleet has gone a long time with minimal problems caused by networking of computers, then there are three valid answers to this question.Tribble wrote:There is a difference between "should there be a fail safe that can decompress the bridge in the event of a fire?" and "let's run all 7 "separate" safety systems in the same network together, as well as everything else on the ship ". You'd think at least one of them might be seperate? There is also a difference between "should there be a warp core ejection system?" And "let's network that system to literally everything else on the ship". Why should the warp core ejection system be networked with things like the holodeck, turbolifts, food replicators etc.?
1) Why not network all the computers? It might well be that under some circumstances, having the ship's main computer able to decide to eject the warp core is safer, because it removes human reaction time from the decision loop in an extremely urgent situation.
2) As noted, wiring fail-safes into the main networks allows people to remotely activate those fail-safes in case the crew nearby are dead or incapacitated.
3) When it comes to normal emergency conditions, a ship with no computer support is as doomed as a ship with no power source. Without a computer, the ship's engines and reactors are likely to be uncontrollable, and you can't even call for help without the navigation computer to point your antenna and communications software to turn it into a usable signal.
Under abnormal conditions that were likely not foreseen by the ship's designers, yes, the single massive integrated computer network can become a liability. But it is far from obvious that this makes the massive integrated network an especially stupid or bad idea.
If the emergency fail-safes in question are systems that do not need to be activated remotely, that's fine- but many such systems would need remote activation to do their jobs under a variety of foreseeable emergencies.Yes, when literally everything on the ship is networked together it's reasonably foreseeable that when something goes wrong with the one part of the network it could potentially bring the whole network down with it. I'm not calling for additional fail safes as much as ensuring those fail safes are seperate from the main network. If they are capable of running independently anyways, that shouldn't be a problem.
Note that I am referring to emergency fail safes, not regular ship functions. There' are certainly good reasons for the tactical systems to be networked together, for example.
If the fail-safes do not need to coordinate with normal ship systems in order to work, that's fine... but many fail-safes would. For example, if your ship has shields up and ejects the warp core, what happens if the warp core collides with the interior of the shield bubble? Is that okay, or will the core go 'splat' and explode in close proximity to the ship? Maybe you need a way to tell the shield generators to open a hole in the shield for the core to fly out through- in which case there has to be a connection between the emergency core ejector and the shield generators.
Likewise, maybe you need to remotely shut down antimatter feed lines or other systems- the fail-safe mechanism says "welp, ejecting the core, better make sure there isn't plasma or antimatter spraying around from the lines I just severed." Again, you need a way for the fail-safe to send messages to the routinely operational systems of the ship.