Hello to all, I would be interested to study JT65 and, perhaps, add this mode to Multipsk (let's say before the end of 2008 or the beginning of 2009). The JT65 specifications are too much general and don't permit to build a code from them. I saw that the sources (Fortran and C or C++?) of JT65 exist but there are a lot of files and it is impossible for me to extract the specifications from them (I am not programmer). However even for a programmer, I suppose that this must be complex.
I think the only solution (at least for me) would be to work in team. The goal would be: 1) to understand how JT65 is built going into details from the sources (reverse engineering) and to write precise specifications, 2) to share the sources into functional blocks so as to know the role of each one. It would be necessary, afterwards, to organize all these blocks in some diagram so as to permit for anyone to understand the general working (RX, TX, data management), 3) to translate the compressed (many operations in the same line, the general meaning being hidden) source coding in C or C++, into some pseudo-code easy to read, or in Basic as it seems the most readable and it is known by everyone or in non-compressed C source coding. The goal would be that this permit to any Ham having yet written a code for coding/decoding a mode (and so used to a minimum of digital processing) to write his own JT65 code (in Basic, Pascal, C or whatever) and to understand the working. All this might be public and the exchanges done within a Yahoo group. If anybody is interested, tell me. 73 Patrick
