The Canary Trap

Those of you familiar with the early Tom Clancy novels may have guessed at my attempt to get round this. The canary trap was a computer generated piece of a document, unique to a numbered copy of that document. Should a copy of the document ever appear in print, it could be matched up to the numbered version and the canary caught. I believe that Terry Pratchett uses a similar method to keep track of proof copies of books in progress.

My idea was based on the above. During the game Jingo, where I played Russia, I signed all my letters with a different name for each recipient. As far as I could tell, the uses were limited. I reckon that there could be only two instances where it might be useful:

1. If an email had been claimed to be from a third party, quoting my email, I could immediately see whether it were indeed from the third party in question, or whether the sender had made it all up. I fail to see the advantage here. Surely you'd be able to tell from the quoted text itself? (Unless you're in the habit of sending identical mails to many recipients).
2. If one of my emails was actually forwarded to a third party, they might spot the different signature and believe it to be fake. (they might even have checked with me!) To my mind this is the main advantage.

What do you think? Was it a good idea, or was I trying to be too clever?

Mike

p.s. It was bloody difficult keeping track of who was to get which signature. (Murky could probably do it automagically!) Well, semi automagically, using Pegasus email I'd define a different sigfile for each player and label the sigfiles 'Austria', 'England' etc. I'd create a 'neutral' sigfile which would be used as default.

p.p.s. There are only so many ways of Russianising Mike - Mikeovsky, Mikel, Mikelovitch etc etc This was possibly the most taxing part of the whole thing!

Kevin Litwack Responds:

A moderately clever idea, but I spot an immediate problem. When I forward or reply to messages, I usually only include the quoted text, and possibly the headers for the message. I almost never bother to quote the signature too, as I usually say who the message was from at the top. A fair point

Therefore, whether I were faking a forwarded message or actually forwarding a real one, there wouldn't be any signature anyway. Also, I wouldn't notice much less care if there were a signature on a "forwarded" message sent to me, so I certainly wouldn't suspect a falsehood based on it. I might be in the minority on this point, but nevertheless it might be an issue. It is an issue, the method is not infallible, but is an extra weapon to sow the seeds of doubt in the mind of the recipient of the forwarded message.

Another problem that occurs to me as I'm writing is this: Wouldn't the signature be different in a forwarded message anyway? What I mean is that if you sign your messages to me "Mikel", then whether I'm making up a message or actually forwarding one the signature would still be Mikel, thus making it impossible for the party receiving the message to differentiate. Kevin has missed the point, the signature will be different to that which the recipient of the forwarded message is used to receiving, and therefore they could view anything forwarded which is in a different style to be suspect.

Just some thoughtm. Although personally, I prefer not to have any sneaky tricks like this. IMHO, it's much more fun to lie and cheat fair and square, rather than resorting to this sort of trickery. ;)

To which Howard Dawson Responds:

IMHO anything upto and including a large telescope set up in the house opposite to view your plans is acceptable............

Return to the files index.