On 9/2/21 1:35 AM, Jim Brown via wsjt-devel wrote: > On 9/1/2021 11:32 PM, Phil Karn via wsjt-devel wrote: >> I don't see that that necessarily follows. > > At this point, I think it is appropriate to note that the developers > of this software are VERY bright, have carefully observed how it works > in the real world, test extensively themselves, get lots of input from > other smart hams, and allow the software to evolve as they learn more. > And the guy at the front of the Vee is this team is K1JT, a Nobel > laureate. I can fault them only on their choice of operating > frequencies for FT4 on 40M, which is a train wreck. And I suspect a > simple reason why -- that none of them work CW on that band, or know > anyone who does. > Reputation carries (or should carry) no weight in science. Only the evidence matters. Amateur radio is (or should be) an experimental service, always willing to try new things. I said (or should have said) right up front that I don't *know* if this will work, only that it seems an interesting idea worth trying, and trying is the only real way to know if it'll work. The main ideas I've suggested, namely slotting and randomizing the transmit frequency with each transmission, are backward compatible with the existing software so they could be made options. Then we can see if or how well they work. My other, more radical ideas like hopping on each symbol, would not be backward compatible. They would have to be a new mode. It's not like this group is unaccustomed to new modes.
But if you want to talk credentials, Joe has repeatedly acknowledged the original work that Tom Clark (W3IWI/K3IO) and I did on a paper we published at the 1996 Central States VHF Society on DSP waveforms for EME. I based them on what I'd learned working for Qualcomm on their CDMA project. I proposed the use of m-ARY orthogonal modulation (e.g., m-ARY FSK) plus FEC, which all the JT modes now use. Of course, we just came up with the idea; Joe and many others made them actually work, which in any project is the bulk of the effort. Phil _______________________________________________ wsjt-devel mailing list wsjt-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/wsjt-devel
