There will always be uncertainty about what the other guy is thinking and doing, whether our messages have been received and understood etc. Exchanging RR73, 73 and free text messages at the end of a QSO extends the QSO sequence so reducing any remaining uncertainty with each additional message. Even after all that, it is still possible the other guy won't have logged the QSO 'correctly', updated his online log pages, uploaded it to LoTW/QRZ/eQSL/Twitter, sent a QSL card, emailed us, popped round to say hello, blogged about it ...
So, too bad. That's just how it is. No amount of laying down the rules about what constitutes a completed QSO is going to eliminate all doubt. The rules are barely relevant, and are not universally agreed or imposed anyway. That's the point. Luckily, the requirement is not to remove ALL uncertainty but to increase assurance that both parties have exchanged information sufficiently accurately and completely to consider it a QSO. For everyday run-o-the-mill QSOs, very little assurance is needed since the consequences of being wrong are negligible. For special, rare or remarkable QSOs (whatever that means to either party), greater assurance is welcome. Unfortunately, without additional info beyond the typical minimalist exchange, it’s not easy to tell whether the other guy is sufficiently assured ... but continuing to send any message, or re-starting the QSO, or sending an email, QSL card etc. are all clues that more assurance is needed or would be welcome. Likewise, ending the QSO by not sending further messages or starting the next QSO are clues that the earlier QSO is over (whether complete or not). There are dynamics here. Good luck to anyone coding the autosequencer state table for all those possibilities, and more besides (messages-sequence-out-of, QSB, QRN and QRM corrupting messages from either end, duplicate/early/late/incomplete/corrupted messages, wrong format, hash collisions and misses, etc.)! For anyone obsessed about working actual living breathing hams rather than their computers and autosequence robots, there is plenty of latitude for deliberately meddling with the message contents, sequencing and timing to see whether and how the other guy responds on the basis that people and computers react differently. I'm blabbering about a digimode version of the Turing test which is, again, about reducing not eliminating uncertainty. Just as people can be trained to ace lie detector tests, computers can be programmed to appear human. What makes you think I am a human? I'm not even sure myself. 73 Gary ZL2iFB aka HAL -----Original Message----- From: Gary McDuffie <[email protected]> Sent: 21 May 2020 09:01 To: WSJT software development <[email protected]> Subject: Re: [wsjt-devel] Clicking on RR73 produced wrong TX response msg > On May 20, 2020, at 10:28, Neil Zampella <[email protected]> wrote: > > The 73 from you is a courtesy ... the RR73 is saying "Roger Roger - > BYE'" basically ... we're good. This is why the program then > switches to the Tx6 to send the next CQ call. If you call letting him know that you got his acknowledgement of your report a courtesy, I guess so. I call it a necessity in order to allow him to move on. If I don’t send the 73, it means I didn’t get his RR73. If he doesn’t finish, he isn’t in the log. Gary - AG0N _______________________________________________ wsjt-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/wsjt-devel _______________________________________________ wsjt-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/wsjt-devel
