Name: Anonymous 2005-11-09 2:00
I don't know anything about programming, but here's a very general question:
Is anybody working on (or already knows the answer to) the problem: can one computer, connected to a network of computers, running some sort of operating system, be made absoultely impenetrable to hacker attack, even when the attacker knows the source code of the operating system, as well as every other piece of information about the first computer? A very simple sort of operating system is an acceptable answer. The mere fact of connecting to other computers via a network is the thing which raises the question of remote access in the first place.
Here's my naive take: computers, being after all physical objects, should eventually be made impervious to certain types of attacks. It's a bit like enclosing a jewel inside a solid concrete enclosure: you can pretty easily smash the first enclosure, take the jewel, and reconstitute an indistinguishable enclosure if you have enough information (especially when no one's looking). However, you could build various traps into the first enclosure which would signal a break-in, requiring an ever more clever thief/attacker. Can this game ever end in a definite physical sense? It seems that the presence of 'information' in computer science gives vastly more power to a knowledgeable attacker. The analogy most likely breaks down when you consider that a computer is all about information, and that information can be manipulated nearly any way you want in order to get what was desired from the victim computer-whether merely access, or files, or meta-data, or whatever else.
Is anybody working on (or already knows the answer to) the problem: can one computer, connected to a network of computers, running some sort of operating system, be made absoultely impenetrable to hacker attack, even when the attacker knows the source code of the operating system, as well as every other piece of information about the first computer? A very simple sort of operating system is an acceptable answer. The mere fact of connecting to other computers via a network is the thing which raises the question of remote access in the first place.
Here's my naive take: computers, being after all physical objects, should eventually be made impervious to certain types of attacks. It's a bit like enclosing a jewel inside a solid concrete enclosure: you can pretty easily smash the first enclosure, take the jewel, and reconstitute an indistinguishable enclosure if you have enough information (especially when no one's looking). However, you could build various traps into the first enclosure which would signal a break-in, requiring an ever more clever thief/attacker. Can this game ever end in a definite physical sense? It seems that the presence of 'information' in computer science gives vastly more power to a knowledgeable attacker. The analogy most likely breaks down when you consider that a computer is all about information, and that information can be manipulated nearly any way you want in order to get what was desired from the victim computer-whether merely access, or files, or meta-data, or whatever else.