Darcula and other novels aside, there are applications where text volume
definitely matters.
One I've come across in my work is transaction-log filtering. Logs, like
http logs, can generate rather interesting streams of text data, where
the volume easily becomes so large that merely attempting to convert
between character encoding forms can become too cost prohibitive in a
given implementation.
E-mail and novels may be produced and consumed at human-limited rates,
but the same is not true for all data streams that are text or text-like
data.
Just something to keep in mind,
A./
