What is a random sequence?
In Cryptography papers there are lots of statements like
Alice choses a random number k
Bob choses a random element of F_p
Can one recognize a number or a sequence of numbers as random?
Which of the following sequences is random:
Answer: all of them are equally likely outcomes of 23 coin-flips.
Sérgio B. Volchan tells the history of the concept
of randomness in mathematics
in an article
for the American Mathematical Monthly.
It is quite fascinating IMHO how seemingly resonable definitions
of randomness were put forward and shot down later to be replaced
with the next definition. The most recent definitions preclude
meaningful checks for randomness by examining finite parts of
a sequence, so the conundrum remains: Is 7 a random number?
That's how to write manuals
The Jupiter ACE
was a home computer produced in the UK in the 80ies.
It had a FORTH interpreter instead the usual BASIC of the C64, BBC micro, etc.
Their Manual explains
the inner workings of the machine in an accessable way. Compare that
to the thousands of VBA books that keep the reader totally in the
dark what goes on behind the funny icons.
Surprising results with IPv6
Spamfilters add complexity, which in turn makes v6 transition harder.
Host A (running OpenBSD) has dual stack v4/v6 with routable v4 address
Host B (running Plan9) has dual stack v4/v6 with a subnet-local v4 address
Both machines have a routeable v6 address and run an MTA.
So I assumed that it should be possible to send mail from A to B.
Turns out to be not that simple. The Plan9's MTA uses various heuristics
to find out if incoming mail is spam (as do other MTAs). One of the checks is to connect
to the MTA listed in the MX record for the sender's address' domain.
Host A's MX record is v4-only, so B cannot connect to the
MTA, so it rejects the mail. Not only the sender and the receiver have
to be v6-enabled, but also the sender's MX (and probably the blacklist