Website founded by
Milan Velimirović
in 2006

12:18 CET



Remember me

Forgot your
Click here!
to create your account if you don't already have one.

Rating lists


MatPlus.Net Forum Promenade The dishonest courier
You can only view this page!
(1) Posted by Siegfried Hornecker [Tuesday, Apr 15, 2008 16:46]; edited by Siegfried Hornecker [08-04-15]

The dishonest courier

See - what is the solution?

PS: Found it on the internet now. Was too hard since I didn't know how padlocks work.
(Read Only)pid=2187
(2) Posted by Vladimir Tyapkin [Tuesday, Apr 15, 2008 20:15]

The idea behind this puzzle is widely used in asymmetric cryptography : (you can read just 'A postal analogy' section where solution is explained).
(Read Only)pid=2188
(3) Posted by Jan Hein Verduin [Friday, Apr 18, 2008 21:11]

I've found two solutions now, but neither one invokes a "wow", so I still may be missing something.

I hate solving these kind of puzzles anyway. It's like with chess problems: I'm just not a solver. I can marvel at the construction or the elegance of the solution, but the mere process of solving doesn't interest me.

(Read Only)pid=2194
(4) Posted by Siegfried Hornecker [Friday, Apr 18, 2008 23:15]; edited by Siegfried Hornecker [08-04-18]

I like Laterals a lot but also not for solving but because they have such interesting solutions.

Reminds me of searching for some more, downloading some websites as a backup and updating my links. When I did it the last time (July 8th 2007), these links worked: (*),Raetsel_Ecke,Laterale.htm

English websites are marked with an asterisk (sadly, there's only one by now, the others are german).
(Read Only)pid=2196
(5) Posted by Milan Velimirović (+) [Friday, Apr 18, 2008 23:49]

The solution is hidden, but
if sending the box is not expensive
(compared to the value it contains)
it is not difficult...
at least for professionals :-)

... and looks very elegant indeed!
1) 'A' puts the value in, locks the box with his padlock and sends it to 'B'
2) 'B' adds his padlock to the box and sends it back to 'A'
3) 'A' unlocks and removes his padlock and re-sends the box to 'B'
4) 'B' unlocks and removes his padlock and opens the box

(one just needs to find it, can you?)

(Read Only)pid=2197
(6) Posted by Siegfried Hornecker [Friday, Apr 18, 2008 23:57]

That's better than the one I found. :-)

Yes, it's simple and elegant.
(Read Only)pid=2198
(7) Posted by Roberto Stelling [Saturday, Apr 19, 2008 19:05]

It is straightforward if you can have more than a lock per latch, but what if you can't ? :)
(Read Only)pid=2201
(8) Posted by Hauke Reddmann [Sunday, Apr 20, 2008 15:55]

In that case, the following *might* work:

1. Send the locked box.
2. Wait for receipt ACK.
3. Send another box, open with the key and a paper
"This key is to a box that already has been sent.
Go and steal it if you are an (_*_) , it's no use for you!"


Nasty word has been replaced with ASCII art to protect the innocent :-)
(Read Only)pid=2202
(9) Posted by Siegfried Hornecker [Sunday, Apr 20, 2008 16:06]; edited by Siegfried Hornecker [08-04-20]

He is an "nasty word" so he steals it. :-)

PS: Same @Milan - he steals the padlock!
(Read Only)pid=2203
(10) Posted by Milan Velimirović (+) [Sunday, Apr 20, 2008 16:31]

The simplest way would be if 'A' sends a box with unlocked padlock an keep the key!
'B' puts the value inside and locks the padlock (simple press, key not needed).
Alas, no guarantee that the padlock will reach 'B' then.

Well, since this is the Chess Problems site, we'd better try to compose a chess problem with similar motivation :)
(Read Only)pid=2204
(11) Posted by Jan Hein Verduin [Monday, Apr 21, 2008 19:59]

Roberto is right, it is straightforward if you can have more than one lock per latch, so I would be less than thrilled if that was the solution. But here's my other way to do it: first send the content in a box locked with a padlock, then after a few days send a second box with a lock through the latch and through the eye of the key of the first lock. That way the key can not be stolen.
(Read Only)pid=2211
(12) Posted by Roberto Stelling [Tuesday, Apr 22, 2008 07:52]

I like this one Jan, provided the key eyes are big enough to fit the lock!
But as the key goes outside the box it is like sending a plain text key to an encripted message you've sent in the past (the first box), it would be ok for the proposed problem (but they'll say that the key eyes are not that big anyway. ;)) but wouldn't really help much on a real world communication scenario.

The two lock per latch solution is exactly how asymmetric cryptography systems work.
It all starts with Alice sending Bob a message encrypted with her public key (a box with her lock) then Bob sends back the original message encrypted with his public key (the same box with her and his lock) then Alice decrypts that message with her private key (she removes her lock using her key) and sends the message back to Bob, who is his turn decrypts it with his private key (opening the lock, and the box, with his key) retrieving the original message (the content of the box).

But what could we do if there were no way of putting two locks on the same latch ? :)
(Read Only)pid=2212

No more posts

MatPlus.Net Forum Promenade The dishonest courier