More than one Quartz server can run on a machine: one from QuartzControl, one from a terminal, one from a session that was killed. The card Servers running on this machine looks for them and lists them.
The detection, here with nothing found.
What is in the list
Per entry the port, the project — as far as it can be recognised — and the origin:
- From this app — started by QuartzControl.
- Started outside — by something else.
Plus whether the server answers a request. A process that is still there but no longer answers is a different case from none at all.
“Unknown” is not “none”
On some systems the process list cannot be read. QuartzControl then says exactly that:
The process list cannot be read on this system. Whether servers are running is unknown.
The same attitude as everywhere in QuartzControl: what cannot be determined does not become an all-clear.
Stopping is a decision every time
QuartzControl never stops a foreign server on its own. The sentence on the card says why:
Servers started outside QuartzControl belong to someone else (a terminal, a second window, a session that was force-quit).
It asks accordingly differently. For a server of its own: the preview in the browser will no longer be reachable. For a foreign one: whoever started it may be working with it right now.
At startup
If QuartzControl finds servers left over from an earlier session at startup, it asks once what should happen to them. Whether it does depends on the setting *On quit*: choose “Keep running” and you get this question on the next start.
Why stopping sometimes does not work
If it then says “could not be stopped”, there are two possibilities, and QuartzControl names both: the server was already gone — or the process number now belongs to somebody else. Operating systems reuse numbers, and between showing a list and clicking on it lies any amount of time. That is why QuartzControl checks again immediately before the signal that the number really means the same program.