After consulting DF8ZR and DF6NM, who both are also active in DFCW, I would
like to suggest a kind of temporary DFCW standard in order to enhance DFCW
QSO's and avoid 'collisions' with other modes :

1. To use a DFCW 'subband' from 137650Hz to 137695Hz, this will avoid
collosions with QRSS actovity that is mainly between 137700Hz and 137800Hz.

2. To use a frequency shift of 4Hz. At this moment this seems to be narrow
enough to minize the bandwidth of DFCW and wide enough to be properly
separated by receiveing software as GRAM. Als a 4Hz shift is easy to
implement with PLL-VFO systems that work in 1Hz steps (as you just have to
toggle the 3rd byte).

3. To use a 'channel spacing' of 10Hz. This would mean that we have 5
'channels' going from 137650/137654Hz to 137690/137694Hz. Working with
'channels' will minimize interference between DFCW signals while the 10Hz
'step' is enough to separate 2 DFCW signals even if there is a slight drift
on one of the transmitters.

4. To use a 3 second 'element length' separated by a 1 second 'gap'. From
QRSS we know that 3 seconds dot length is optimal for most receiving
software. The 1 second 'gap' will make visual detection of the signal
easier and also reduce the dutycycle a bit, avoiding overheated PA's.

5. To introduce following practice for weak signal work :
If you receive a CQ from a station and you are not sure to be 'seen' at the
other end, just send a 'KK'. If the other station can read the 'K' it
should reply the with 'RR' as confirmation. After that the station that
sent the 'K' can reply to the CQ. This procedure is suggested by Markus,
DF6NM, and can pervent idle transmissions and useless QRM.
eg.
CQ ON7YD K
KK
RR
ON7YD DF6NM OOO K
etc...

73, Rik  ON7YD