A proper law

[Owen Lewis]

> -----Original Message-----
> From: ukcrypto-admin@chiark.greenend.org.uk
> [mailto:ukcrypto-admin@chiark.greenend.org.uk]On Behalf Of Ian G Batten
> Sent: 06 March 2003 13:48
> To: ukcrypto@chiark.greenend.org.uk
> Subject: Re: A proper law
>
>
> On Thu, 06 Mar 2003, Owen Lewis wrote:
> > I agree with you. Much can be achieved both by refusing to be
> over-awed by a
> > problem and setting aside the received wisdom as to how such a
> problem can
> > be approached or that the problem cannot be solved.
>
> That's clearly true.  However, the point about Enigma was that it would
> fail the most basic tests for a crypto system today --- the input is
> correlated with the output and is not flat random --- and it was used
> incredibly badly.  And the bad use wasn't just a matter of individual
> stupidity, some of it was systematic to the communications network
> (stereotyped openings and closing in a system already prone to
> conjectured-plaintext attacks) and even to the key setting process (the
> reuse of stecker sets over time was bad, but in some senses far worse
> was the policy of not repeating a the choice of wheels within a month,
> which meant that by the end of the month the number of options to be
> tested was significantly reduced).
>
> It's clearly the case that Enigma was broken by very clever people, but
> we must not forget that it was used monumentally badly.

Were the true history of cryptanalysis ever to be written, I'd expect to
find that the large majority of successes have come from exploitation of
weaknesses in the systems implementation rather than through a frontal
attack on the crypto algorithm. That or some entirely different process,
such as the subornation of personnel.

With regard to flawed implementation of operational requirements, there is
an amusing and true story from WWII that indicates why the German
implementation of Enigma coding may have had the types of flaw that it did.


Western Desert 1942. Extended FEBA from the sea to hundreds of kilometres
inland, all across desert. Rommel needs to push on to Cairo, the British
need (at this point) to hold the line. Vital to the British ability to hold
the line was a water pipeline laid over the surface of the desert, sand
dunes & all, behind and for the length of the front. Without that water
supply, the British would have had to withdraw in less time that that
required to repair the pipe.

Rommel knew of this vulnerability. Orders were sent out along the length of
the German front that raiding parties were to be sent out and at the ordered
day and hour, the British pipeline was to be breached with explosives at
every kilometre along
its length. This was done with great German efficiency. Jubilation and
preparation for the big push forward as the British retreated. But the
retreat never happened. Why?

To the Wehrmacht, ordered were to be obeyed and neither questioned nor
interpreted according to circumstances. It had obeyed explicitly and the
pipeline was breached at every kilometre along its length. However since the
surface of the desert rose and fell by many metres at irregular distances,
an estimated 50% of the water in the pipe remained, puddled in the pipe
between the regular breaks. With strict water rationing, this proved
sufficient for the British to hold their positions until the pipeline was
repaired.

Discipline has its place and also its limitations. In a strictly disciplined
and hierarchical organisation if a superior makes a flawed decision, the
juniors tend to comply blindly, even where their eyes tell them that there
is a better and more certain way of achieving the aim.

Should this little tale seem too pointedly anti-Teuton, it should not be
taken so. Similar stupidities occur everywhere. But, if the Teutons do have
an identifiable failing it is, occasionally, that they so develop a virtue
that it turns to vice.

(....Supersonic travel)

> I didn't know that that theory was advanced.  It'd be interesting to
> know when: the muzzle velocity of a rifle in .303 is well over 2000fps
> (speed of sound is of the order to 1000fps, depending upon altitude) and
> those were available in the late 19th century.  Even .38-40 is
> supersonic under certain circumstances.

Downhill and with the wind behind it :-)

More seriously, weren't all munitions sub-sonic until the introduction of
nitro-glycerine/nitrocellulose based powders? If I'm right in this, then
that leaves all but about 130 years of human history before the possibility
could be simply demonstrated.

Even then, in the 1940's there was considerable talk about a 'sound barrier'
through which aircraft would not be able to travel. AIR some of the WWII
prop-driven aircraft could approach this 'sound barrier'  and vibration
induced by turbulence in the airflow (and reduced response in the controls?)
threatened such aircraft with destruction. Then came 1947, Chuck Yeager and
the Bell X-1. However, it was some years more before the requirements for
supersonic flight were properly mastered. Remember the disintegration of the
DH-110 over the Farnborough arena in 1952? John Derry had just gone
supersonic, flat on the deck,  as a crowd-pleaser.

As for the theory of suffocation if one travelled faster than horse could
gallop, AIR that proposal was seriously debated at the time when
Stephenson's Rocket and others pioneer railway engines were planning their
speed trials. But I was very young at the time and my memory may now fail
me.

Owen