Reply from SM2CEW to the article by SM5BSZ in QTC 8 2008.
(Translated by Leif, SM5BSZ, Sept 16 2008)
Leif makes it easy for himself in his critizism of my article. He mixes boldly his personal opinions, fluffy comparisons, facts and fiction. But when Leif considers my opinions harmful to our hobbu and counter-productive for the survival of CW I realize that he has not familiarized himself with - or does not want to understand - my reasoning. This can be seen in the article, Leif seems to think that if the Deep Search decoder writes the same thing two times, then everything would be ok. I also find several references to Leifs own experiences on 144 MHz CW EME 10 years ago, something that is totally irrelevant for the issue that I address in my article.

The Deep Search concept disturbs me since it opens the way for more protocols where the information takes other routes than through the radio channel that was referred to when our QSO definitions were created.

The following facts remain: - A QSO via Deep Search can not happen unless ALL information is already present in both of the computers used. The computer never receives a complete message, it uses fragments to make a guess (match) against known information from a data base. In my article I show that one can write anything in the data base, the program will match received fragments to the texts in the data base and print the contents on the screen as if it were received by radio (see the example SP6GWBluff).

- A third party station who listens to a QSO when the signals ar at the Deep Search level must inevitably have all the information in his computer to be able to see a QSO. This third party station also has to fool the computer by typing the call of one of the two QSO partners as if it were his own to get messages of the QSO on his screen.

- A very large number of EME-operators tune their guessing further by setting the sync level to -2 and to switch on "aggressive decode". One of the German operators (antenna 2 x 12 el yagis) who recently obtained his DXCC on 144 MHz considers this a fantastic way to become sucessful; "If one uses these program settings, the desired message comes within a couple of minutes. Then I can revert to normal settings".

- The amount of information that then is transferred through the radio channel is approximatly equivalent to two characters. The computer could equally wel have declared that it had received the whole Holy Bible in case one had chosen to put that text into the data base rather than two call signa and a locator. The computer actually did not receive the message, it just made a match of received fragments to a data base.

As we can see this has no connection at all to whether Leif has been able to work stations in CW or not. That information is totally irrelevant in relation to the issue of how a computer program produces text on the screen.

The idea with Deep Search is to create sucess without having to rely on the radio channel except for transfer of fragments. This is also shown in the report exchanges in the JT65 program. It relies on "now the report is transmitted" not any message about what the report really is. And the fact that the call sign of expeditions is never transferred, the receiving station decides what prefix he wants to see by typing it into the computer before the qso.

These are the circumstances that I am warning about in my article. It is important to me that the radio channel carries the information that shall be received and that the decoding is done in a way that does not renounce this.

For this reason I can not understand why we should use computers to mislead the operator and make him believe that all the information has been received rather than give the computer the time to dig into the noise until it can decode an unknown text and present it like it was transmitted.

The reason I take this standpoint is in that we are radio amateurs and that when the definitions for a QSO was made, these obviously built on the fact that all information was exchanged via radio waves. My opinion is that we should stay with that idea. It goes without saying that it also includes QSO requirements, top lists and contests.

To SM 5 BSZ Main Page