January 04, 2006

Digital fortress

Well, up we come on another academic term, but first I need to warn you about an absolutely execrable book. As my faithful readers will know, I like getting audiobooks for roadtripping; I got started "reading" them on the long Providence-Chicago runs, and kept the habit. Of course, now I'm making shorter trips, generally, but that just means that one book gets stretched over a few round trips.

I've read many of the CD audiobooks in the public library that aren't juvenile or non-fiction, so every time I'm in there I feel like I have to look a little harder. This time, a few weeks ago, I picked an offering from renowned author Dan Brown: Digital Fortress.

What. A. Mistake.

I mean, it seemed like a reasonable idea. The Da Vinci Code wasn't fantastic, and it had some plot holes, and it made stuff up. But still, it wasn't that bad. I'm ok with a certain amount of mediocrity in my hack fiction. This novel is so much worse than that, though.

The characters act without motivation, even routinely acting counter to motivation. The descriptions are spare, and then suddenly florid, neither tactic particularly working. The plot is entirely predictable (and not very interesting) from just a short way in. And, most crucially, the characters are really, really stupid, and computers just don't work even remotely like he claims they do. When two—no, three—world class cryptographers can see two names and not spot an obvious anagram that I could catch without even seeing it in print, what the hell am I supposed to think? The book is littered with examples like this: Dan Brown is not a smart individual, but he wants to seem like one, so he builds characters with great résumés, bills them as really smart, and then has them act like morons.

He also tries to give a patter of technobabble that is detailed and convincing, but is in fact just painfully and obviously wrong. And in ways that even a single reading by a mediocre computer scientist would have caught—and been able to propose an alternative that would be correct and still let Brown's unimaginative plot proceed apace. He would have been better off staying really vague on the computer details, or taking the Star Trek route and inventing new words (though I imagine he'd do a job on those, as well). But after catching a lot of the run-of-the-mill stupidity, the computer scientist would be forced to explain to him a few very important things:

  • There is no way to send an email to an address and cause some "probe" to automatically emit a "tracer signal" that says where that email was forwarded to. The best you could hope would be to send runnable code and hope the recipient was stupid enough to run it, but this would still require them to open the email first, and you'd have to know their OS in advance, and it still wouldn't be able to report the destination email address, just the computer address (at best).
  • When you decrypt an encoded message, you are not running it. It could say "Attack at dawn" or it could say "rm -rf /" but either way it's just a bunch of text; merely viewing it does not cause it to execute.
  • Cryptographers and system security specialists would know this.
  • Computers may do many things when they unexpectedly divide by zero, but returning 999999999.99 is not one of them.
  • When someone puts up an encrypted message for universal download, password to follow on later payment, it is not useful to substitute your own version of the file later, because the relevant recipients will have already gotten it, even if they haven't decrypted it yet. Changing your copy will not change their copy. Furthermore, in any modern system, even if you had the password to decrypt it, you would be wholly unable to encrypt your own message to use the same decryption password.
  • Cryptographers would also know this.
  • If the entire encryption/decryption system is itself encrypted and posted, having a password will not be sufficient to decrypt it: you also need the decryption program, because the encrypted executable is, y'know, still encrypted.
  • A world-class cryptographer will not have a five-character alphanumeric password, and if she did, she sure as hell would not think of it as "highly secure", whether it be made up randomly or not.
  • Even assuming that a message were decrypted, and turned out to be runnable, and someone ran it, for a "virus" to do anything on a custom-built, custom-programmed machine, it would have to have been designed specifically to run on that machine, by someone who knew that machine inside and out.
  • And cryptographers and systems security specialists would definitely know this.

The whole book is so maddening. Infuriatingly and needlessly wrong things are on practically every track of every disc. (In an early dabble in "making shit up about the Catholic Church", he informs us that in Spain, they take Communion at the beginning of the Mass. WTF? And at a communion rail, to boot.) I've still got two CDs left (out of ten), and I am just completely uninterested in actually finishing them. I don't normally stop books in the middle like that. This is even what is (presumably) supposed to be the exciting part. But, gahh. Avoid this book at all costs.

"For those who read the Lewis books as a Christian parable, Aslan fills the role of Christ because he is resurrected from the dead. I don't know if that makes the White Witch into Satan, but Tilda Swinton plays the role as if she has not ruled out the possibility." --Roger Ebert

Posted by blahedo at 4:44am on 4 Jan 2006
I gave you an out. Posted by lee at 7:30am on 4 Jan 2006
So next you're going to tell me that Alias - my favorite TV show - isn't entirely realistic? Incredible. Oh, and I never claimed to be secretive ;) Posted by Theresa at 8:15am on 4 Jan 2006
Yikes. Thanks for the warning. Posted by Chris T. at 9:21am on 4 Jan 2006

Lee; the problem was that I still needed to download the other book, which took for-freaking-ever and hadn't finished when I left for Urbana. :P

Theresa: Hee. But there's "unrealistic", and then there's this. I'm actually not normally a nitpicker except for fun; I'm more capable than most at a healthy suspension of disbelief; and yet Dan Brown's research on this was so bad that parts of it just did not make sense. This is among the worst books I have ever read.

Posted by blahedo at 11:19am on 4 Jan 2006
You see, I felt all that and more about the Da Vinci Code too. But then, I've done a lot of dabbling in renaissance art. I can't even imagine how infuriating Digital Fortress would be... but from your overview, it sounds awful. Posted by yukino at 8:34pm on 4 Jan 2006
Is it bad that I actually want to read this book now, simply to see how bad it is? Posted by Ian M. at 2:09am on 6 Feb 2012
Yes. Posted by blahedo at 9:08pm on 6 Feb 2012
Post a comment

Add one to this number: 842

Remember personal info?

Valid XHTML 1.0!