I’m sitting here again listening to the Today programme - my god that’s starting to sound like a litany - and heard the usual mile high piece on viruses, spurred on by the latest and particularly virulent Mydoom strain.
They strayed off viruses themselves and talked a bit about hackers/crackers and the mindset of people who write them. The usual questions came up about whether we could ever have systems that were un-crackable, and it was interesting to hear from the interviewer that they completely understood that it was an ongoing and escalating battle between crackers and security folks.
Many views of security are that one should put in lots of effort to make whatever it is as secure as possible. My view is that all you need to do is make things slightly more secure but visibly more secure.
If I want to lock up my bicycle I can’t ever make it secure enough that it wont get stolen. A van full of power tools can cruise up and even with the best locks that money can buy they’ll have my bike in 5 minutes, or even worse just take off with bits. The only way to make it fully secure is to cement the whole thing 5 feet under the ground. Computer security is similar. To make it completely secure is to actually make it unusable, if people can get at it then by definition it’s unsecure.
But no bike is an island - unless it’s been thrown in a canal. If you make your bike more secure than its neighbours, and visibly more secure than its neighbours then prospective tea leafs will choose easier pickings. Computer security should work on the same principles of competition to ensure that they personally don’t get broken into, and play the principle by making sure that you only need to be more secure than easy pickings, and somehow communicate this to prospective crackers. Maybe they should send back a list of easier IPs to hack with every exploratory ping.
Now of course this doesn’t apply to viruses which are indiscriminate. And it wont help if your bike is much much nicer than the ones around it. Tall poppies need very strong stems.
