Wirt revises my back of the envelope estimate:
> Actually, 398,448.86 years (or more precisely: 145,533,447 days, 14 hours,
> 33 minutes and 48 seconds, at the rate of one password per six seconds,
> presuming no bathroom breaks).

Actually, it will take six seconds longer than that.

> [...]the exhaustive value being calculated as 26*36^7 + 26*36^6 + ... +
> 26*36^0)

There are 2,095,681,645,539 possible passwords of the MPE format.  Wirt's
calculation misses the most obvious one.

> On the other hand, Gavin didn't quite get an "A". His answer may be a little
> high if he forgot that the first character of an HP3000 password has to be
> alphabetic (thus the 26 above), while all of the remaining characters may be
> alphanumeric (36 possible characters).

I took 36^8 as a first order approximation after deciding that it's silly
to require that the first character of a password be alpha only.  I suspect
that at some point (maybe back when MPE had to run in 128KB of memory) it
was simply convenient to use the same validation routine that was used for
user/account/group/file names.

Greg mentions:
> The PKZip manual has a handy little table to explain password "hacking" time
> for a brute force attack on their encryption key method of protecting a ZIP
> file, at the rate of 10,000 keys attempted per second.

Impressive numbers unless you know that PKZIP uses a crappy encryption
algorithm that can be attacked directly.  It's my understanding that any
ZIP encrypted file can be broken very quickly without any guessing
whatsoever.

G.