Schneier on Security

Clive Robinson • May 5, 2017 8:30 AM

@ Ron, TS,

Maybe because it’s easy to make paper into nitrocellulose? But the regular chemical sniffers would pick that up…

Innstead of thinking nitrates for “explosives” how about other chemicals that when burnt would release blood/nerve agents (cyanides / organo phosphates).

Due to the closed air system on most aircraft by the time anyone realised what was happening quite a few people would be dead… Pulling down those little yellow oxygen masks would not stop you breathing in such chemical agents nor would it stop skin absorbtion.

The thing is this chemical weapons risk has been known since the last century as it was sugested as a possible way for hijackers etc to make “dead men” defenses and got serious attention after the Japanese deth cult released sarin onto the subway system.

Clive Robinson • May 5, 2017 1:37 PM

@ Chris Abbott, eureka,

I think what you’re saying about a chemical is more likely.

Seems like an actual concern

As I said it once was, and may well still be.

But experience tells us that in general the TSA is reactive not proactive.

Which I suspect some passengers might be gratefull for… After all it was our host @Bruce that joked about a terrorist stuffing a bomb up “where the sun don’t shine” then somebody actually did it. Much to the supprise not just of Bruce but a Prince, over whom the blood and guts of the attempt descended, a sight I’m sure he will not soon forget.

But in general the terrorists are no where as inventive as smugglers or oil rig workers with a thirst on them for some booze.

Back when I worked in the oil industry I had to go onto platforms more often than I liked. Part of the problem was back then I had a fondness for a certain well known caffinated soda that came in a can preferably just begining to freeze. I’ve only been on one platform where you could get it from the bond or commissary. So one hot summer I decided I’d take a large bottle with me onto a gas platform that was getting an upgrade. At the heliport I was stopped and told “sorry you can’t take it”…

The guy that stopped me was quite friendly and we got chatting whilst I filled out the paper work so I could get it back on my way back. It turns out that rig workers with a thirst on them were quite inventive about how they would try to smuggle booze onto a “dry” platform. One such was jokingly refered to as a “booze bomb”. Basically what they did was to drill a small hole in the bottom of the can and drain out the soda then fill it with vodka or some such and seal the hole again. But these failed either a weight or “squease test”. Those with a thirst thus got a bit smarter and weighed the can prior to sealing but the can still failed the squease test because there was no preasure in the can.

Well one thirsty lad worked out that if you drilled the right size hole in the older steel based cans you could solder in a hypodermic needle without melting the syringe end fitting. Thus the correct weight of vodka was put in then using a modified “home soda maker” he preasurised the can and sealed the needle in a way it could be cut flush and polished flat to the base. Thus the can passed the squeeze test. He was getting away with it till someone squeezed to hard and the can shot out the remains of the needle and covered every one in the area with a fine mist of vodka which caused no end of problems, and apparently nearly caused a fire due to someone smoking, hence the blanket ban on soda drinks etc.

Any way being of a curious and creative mind I and a friend who worked on the more curious patents you don’t get to see decided to see just how much of a problem such a can would be. Well lets just say it was a practical entry into home made Fuel Air Explosives. Though technically a Boiling Liquid Expanding Vapour Explosion (BLEVE)[1] it was defiantly quite impresive. Both of us being practical jokers as well decided that there was potential as an ink bomb. Thus the vodka got replaced with a dark mauve food colourant[2]. The can shaken vigourously then the ring pull pulled with a long piece of string. The cloud and stream of colourant was quite spectacular. However in another test we decided to see what would happen if the seal on the hypodermic needle failed… The resulting stream of purple had a very extensive range.

Any way some years later I was doing a paid security gig for an official agency part of which was a red team “MacGyver Test”. Where we were chalenged to come up with a new threat weapon made of innocent household items that somebody might have around them, that not just looked innocent but would be stable unlike home made explosives belts etc. I suggested using the “booze bomb” but using one of a number of “kitchen sink” liquid chemical agents… It caused a bit of consternation for the agency concerned, I gather it was not something they had thought of though unpreasurised glass jars/bottles they had.

I’ve yet to see it come up in a book or movie, so keep your eyes open.

[1] https://en.m.wikipedia.org/wiki/Boiling_liquid_expanding_vapor_explosion

[2] Cochineal, back then still madr with crushed “scale insects” https://en.m.wikipedia.org/wiki/Cochineal

Clive Robinson • May 8, 2017 5:57 AM

@ itsa,

Book ciphers are a bad idea as it is not truly random data. A good OTP data stream must come from a good PRNG or random source.

Book ciphers in their original form as I indicated are an idea that went back over a hundred years. Further I also pointed out that at various points in time people have looked at how to improve them.

There is three things you need to consider,

1, Reduce to an alphabet the same size as that in the plaintext and importantly fixed in width.

2, Reduce any statistics to a suitably low value.

3, Apply a mixing function to shuffle the result to reduce statistics further.

All three of these can be done by pencil and paper methods that have been around for some time.

One way is to apply a mixing function with two books. An overly simple idea would be to arrange a start point in both books. Then using the ten most common letters (remembered with “eat on irish lid) only from the first book count along in the second book and write down the letters to form a base string.

To flatten the statistics of the base string use the conversion process from base28 to base10 used in the Nihilist cipher system to make a base keystream

Likewise flatten the statistics of the plaintext. Add the flattened plaintext to the base key stream along with an additive mod 10.

Then using the Nihilist process in reverse convert the base10 ciphertext to base28 ciphertext.

It’s a simple process that makes a near flat statistics ciphertext that by the last process converts it back to what statisticaly looks like –but is most definitely not– a simple substitution cipher.

The only question is how you would make the additive value, of which there are quite a number, that again can be easily done by hand. In a way you could view it as a stream cipher with super encryption. To see how to knock out a few annoyances such as base28 to base26 conversion have a look at how the UK and later NATO Rocket super encipherment system worked.