How about,
OK we can have up to 13 Charaters right?
How about, 65?
The Transmit transmits in FOX mode 5 streams.
the recv decodes all 5 and re assembles them on order of frequency. low
to high.
Joe WB9SBD
On 5/23/2024 1:02 PM, aa8sh--- via wsjt-devel wrote:
Hello Fellows from Clark AA8SH,
I'm looking for feedback on enhanced chat functionality for WSJTX.
Rather than using 13 characters of free text to form a single
expression, please consider a system that uses 13 characters to
represent 12 CHOICES OF expression. This would allow a single pass of
free text to "send" as many as 12 expressions without actually sending
the expressions, greatly reducing throughput.
The system would allow end users to sequentially navigate through a
decision tree of available expressions. As the end user does so, the
software compiles an alphanumeric word which is 36 characters deep
(A-Z 0-9) and 12 placeholders wide, accommodating as many as
4.738381338×10¹⁸ choices of expression.
The rough proposal:
1) An online expression lookup table is populated by end users with
expressions--words, phrases, or sentences. Updates are made available
to end users weekly so that users can QSO while offline.
2) An online decision tree refers to the expressions in the lookup
table. The decision tree is continuously optimized by AI for maximum
intuitive navigation by end users. Updates are made available to users
weekly so that users can QSO while offline.
3) Local installed software writes into and reads from the WSJTX Free
Text pane. Use of 12 sequential conditional dropdowns of 36 choices
each results in an alphanumeric word that is 12 placeholders wide and
36 characters deep.
4)The first placeholder of the created alphanumeric word is to be
reserved for the wildcards +/- which are used to let the receiving
software know which updated release was used for encoding. + might
represent this week's release, - this month's release, / this
quarter's release. This would allow a station that has been offline
for up to three months to use the system.
5) If an end user is online, the local software forwards each
transmitted alphanumeric word to AI for decision tree optimization.
If an end user is offline, all created alphanumeric words are logged,
then transmitted to AI when next online. In my opinion, this is the
most powerful aspect of the entire proposal. Probably a huge lift to
implement, but extremely powerful, once implemented. If end users
populate the database with expressions and AI arranges those into an
intuitive decision tree, the resulting system would be robust.
6) I leave matters of software functionality to the experts. I dimly
imagine resting a cursor on a received alphanumeric word, then seeing
a dialog box of text pop up. That would be nice!
So, my questions to the group. Is it worth doing? Would anyone use
it? Is it a waste of time? Does it run counter to the purpose of
WSJTX? I'm looking for brutally honest feedback. If the consensus is
this is a non-starter, then it's a non-starter. But if there is a way
forward, please direct me in that way.
Very Respectfully Submitted
Clark AA8SH
_______________________________________________
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel
_______________________________________________
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel
