Srsly guys, stop with the quote warfare.
.
.
.
Ok uh... would a setup such as the following work?
1. An observer (could be sensor drones, or ships, or whatever) detect an enemy ship.
2. Regardless of whether the observer has been detected by the enemy, the observer then transmits to a friendly control center (ship or station where data is collected), which preferably is and will remain undetected by the enemy. If the observer matters (that is, you can't afford to lose it), then this transmission can be done via laser. If you don't give a damn what happens to the observer after it transmits, then you might as well just do an omnidirectional squawk. Or maybe it uses some sort of "hand gesture" system, readable from any direction but only if you've already got your telescope trained on the observer. Either way, the control center is still safe.
3. Now the control center knows where the enemy is. It can track the enemy as well as the observer, if not better (because it's designed to). It now begins tracking the enemy contact. The observer may now go on with it's business.
Note that the observer is not necessarily a 'focused' detection device like a telescope. It may be capable of focusing in on something like a telescope does, and it may do that whenever it spots a new contact, though.
The control center, on the other hand, is like a whole bunch of independent telescopes (or some similar sort of device(s)), which upon hearing about something, will point themselves at that something and not let it out of sight. Only one telescope (out of all of those on a control center) needs to be assigned to any one contact, except possibly as the contact moves out of view of one and into view of another. If there are multiple such control centers (which is probably a good idea), they can share some of their targeting data with each other (provided they can use some of the previously discussed "low profile" communication methods), so that if one goes down, the rest of them are still able to track (most of) its targets.
As for actually doing the attacking, the weapons platforms (lasers, missile pods, etc.) could and probably should be separate from the control centers. That way, you can attack without risking the loss of any more than that single platform.
Distributed and redundant systems FTW... then again, you probably won't trust an army of tiny robots to declare war or negotiate peace, so you still need to have some way of telling them what to do. Maybe just a failsafe, so if they're still floating around 10 years after the last friendly contact, they self destruct.