[Martin Taylor 940730 11:11]

Martin Taylor 940727 15:00

The following is a confession of two-fold stupidity, with no real excuse.

I guess my mind must have been on many other things at the time, but there's

no excuse for not reading a message before posting it, as I did.

These stupidities were found when the message arrived back here and I had

a moment to look at it again. Appalling!

Not rereading was the second stupidity. The first is in what I said,

the error of which should have been self-evident on a rereading:

In information theory, if a construct z is a function of two variables x and y,

then H(z) <= H(x) + H(y). In fact, it is given by H(z) = H(x) + H(y) - H(x|y).

The first inequation is correct, the second equation is incomprehensibly

wrong. I can't imagine why I wrote it. Please accept abject apologies,

and an intention to reread my writings before posting in future. That,

I suppose, means that I should not write at all when pressed on all sides.

The temptation is hard to resist, but I shall try.

Sorry again.

Hope the Durango meeting went well.

Martin