[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: About Turkish (what is the importance to the VMS)

    > [Dennis:] I used this a long time ago; as I recall, it's a
    > mathematical optimization technique.

It is a technique for discrete optimization problems, e.g. the
"traveling salesman problem" (find the shortest tour that visits N
given cities).

In the "downhill method", you try all sorts of small changes to the
current best solution (e.g. remove three edges from the tour, and
reconnect the pieces in a different order). Whenever the change
leads to a better solution, it replaces the current one. 

The downhill method will often get stuck in a false minimum, where all
possible "small changes" only make the solution worse. The "simulated
annealing" method gets around that problem by allowing occasional
changes for worse; so, even if it falls into a false minimum, it still
has some chance of climbing over the surrounding walls.

The probability of such "uphill moves" is controlled by the user, and
should slowly decrease to zero as the computation proceeds. That
probablilty knob is traditionally labeled "temperature", because of a
vague analogy with the physical process of metal annealing.

    > [Julie Porter:] Well, back in '92 the PS printer was powered
    > with a RISC chip AMD 22XXXX something or other. It was quite a
    > bit faster than the 68020 then in the desktops. We used to joke
    > the fastest computer Apple sold was the printer.
This may still be true... Not long ago, one of my colleagues
here was using a laser printer's processor to do image processing.
(It is easy to program a Postscript printers to write data back to the
computer; usually the hardest part is getting the computer to listen.)

    > [Julie Porter:] If I had more time I would love to do some
    > analysis with the positions of the pen strokes in the Vms.
    > [Dennis:] One fellow has a web page on this subject, but I don't
    > remember who.

You are probably thinking of Mik Clarke:


It is possible that the Voynichese letters were designed to be drawn
with a brush, as Mik suggests. However, looking closely at the best
VMS images we have, it is clear that this particular copy was written
with a pen: one can clealy see the trace of a split, squarish nib.

In fact many (if not all) of the the colors seem seem to have been
applied with a broad pen, too.

All the best,