In article .net, "Dwight
Stewart" writes:

"Len Over 21" wrote:

"Dwight Stewart" writes:
I haven't purchased a program like that recently, but
surely they've gotten better over the years. Is that not
the case?

They have, but the mighty morsemen consider such to
be desecrations of the will of the old radio gods.

Well, I suspected the programs might have gotten at least somewhat better
over the years. The author of the program I had wrote about trying to
compensate for bad code by looking for patterns instead of focusing on each
individual dot and dash as it was being sent.

It is a trivial matter to do decisions on "longer" v. "shorter" once the on-
time has been converted to a numerical value...which is easily done by
a simple comparator gating a much higher repetitive pulse into a small
(8-bit) binary counter. That forms a "width-to-digital" conversion. The
count is then written into working memory and used by the main
program. Once that is in place, the main program can do its "pattern
checking" most anywhich way.

The program also didn't use
hard rules for dot, dash, and space, length, instead interpretating each as
it went along.

That part is the "adaptivity to rate" section. In an ordinary PC it is
fairly
easy to access the internal calendar clock for 100ths seconds time
hacks (actually shorter if you know the calendar-clock software details).

By comparing the overall "on" lengths it is possible to determine the
bit rate per unit time and thus the equivalent WPM rate.

Obviously, I don't know the details, but the program did do a
pretty good job considering it was just a simple programming example
included with an operating system. My only complaint was that it didn't send
code like some of the other programs advertised, but I couldn't have used
that back then anyway.

The sending part is fairly straightforward involving the keyboard addressing
a lookup table in memory and organizing the outgoing "on" and "off" times
at whatever rate is desired.

The interesting part of modifying that is to add some random variability
to the "on" and "off" times, which is selectable to bias those if desired,
and thus create the equivalent length-rate-bias "swing" of a human
telegrapher! :-) Sending is much easier to do and duplicate than
receiving.

As I said much earlier to others in here, I've seen it done and looked at
the source code, heard-seen it in action. The PCTA will not believe
it since they don't WANT to believe it so any argument with them is an
exercise in futility. :-)

As I also said earlier, there just isn't any market for such a program
since there is so little morse code communications being done in radio
now as compared to a half century ago.

LHA