Sorry to re-open this thread, but it was mentioned to me by EricM last night, so I searched for it.
Below is part of my application for I-L, in which I mentioned another member here who is a member there, and also, and I quote:
K Bohn (Varjeal) of Azor Lockworks in Alberta can also vouch for me.
so either it wasn't read properly, or, more likely, whoever was kicking up the stink has calmed down a bit.
My $25 is going towards trying to calm, trying to reduce the militant attitudes of a few of the members. Of course, many if not most of them are perfectly ok. I just bring a different perspective to the debates about whether it is right or wrong for things like the Matt Blaze paper on safecracking to be out in the open, and so on.
A large part of the fear is that someone could take the entire site and rip it, then a lot of sensitive stuff would be out in the open.
Like IT security, obscurity is generally not a great thing to rely on. However, the analogy falls down when we look at things like your PGP key - without the obscurity of the passphrase and private key, there is no security. In fact, it breaks down further when we consider that updating a set of 100 servers hosting 100,000 websites can be done in a day for almost no cost. A large masterkeyed lock suite is far, far more difficult to update, it's disruptive, and the costs will be in the 6 figures due to re-issusing keys alone!
This makes the publication of a flaw, even a major flaw, in a system far more difficult to justify releasing. How long would it take you to go out and replace every lock you've ever played with due to a security flaw? Years? Probably a few days, even for the members who merely play with locks a few days a year and have never even posted! And there are many locks that are sucure because of the obscurity, too. There might be some ultra-secret flaw in a few locks, that no-one has found yet. It's of no use to anyone, though, and, if it does get published somewhere, then that lock design will, eventually be changed. But re-tooling costs a lot.
Due to the mul-t-lock pin-in-pin weakness, there is now a simple change which has been reported to Eric from the factory which stops the attack from working. It took a while, and may not be true of all the locks world-wide, but going round to change the thousands of mul-t-lock cylinders that have this weakness is going to never happen. It simply isn't cost- or time-effective.
So, hopefully by now I have shown why security through obscurity *is* still security, at least in the real world, where automated attacks cannot be carried out at the rate of a billion a second. That's why safes are still secure against auto-diallers, at least for a time. And the safe combo, like the password, is simply obscurity.
Loading pithy, witty comment in 3... 2... 1...
