rtksvr send_nmea: send the single solution status

[ourairquality]

When sending a nmea single solution back to a base, include the solutions status rather than always SOLQ_SINGLE and also include the number of satellites used in the solution and the reference station id.

[rtklibexplorer]

I've started using Claude AI to help with the reviews, especially parts of the code I'm less familiar with. It had a couple of minor concerns. Do you think these are valid?

1. nmeareq==1 loses SOLQ_SINGLE — the old code explicitly set sol_nmea.stat=SOLQ_SINGLE inside the lat-lon-hgt branch; the new code leaves stat at 0 (SOLQ_NONE). Some NTRIP casters require quality≥1 in the GGA and may reject the connection in this mode.

2. ns=0 floor removed for nmeareq==2 and ==3 — the old sol_nmea.ns=10 was set before the if/else and acted as a floor for all modes. Now nmeareq==2 and ==3 use svr->rtk.sol.ns directly, which can be 0 early in convergence. The comment on that line ("Some servers don't like when ns = 0") suggests this was intentional protection. Consider sol_nmea.ns = svr->rtk.sol.ns > 0 ? svr->rtk.sol.ns : 10 for those modes.

[ourairquality]

I've started using Claude AI to help with the reviews, especially parts of the code I'm less familiar with. It had a couple of minor concerns. Do you think these are valid?

1. nmeareq==1 loses SOLQ_SINGLE — the old code explicitly set sol_nmea.stat=SOLQ_SINGLE inside the lat-lon-hgt branch; the new code leaves stat at 0 (SOLQ_NONE). Some NTRIP casters require quality≥1 in the GGA and may reject the connection in this mode.

Not seeing where sol_nmea.stat=SOLQ_SINGLE is lost in that path, it's seems to still be there.

2. ns=0 floor removed for nmeareq==2 and ==3 — the old sol_nmea.ns=10 was set before the if/else and acted as a floor for all modes. Now nmeareq==2 and ==3 use svr->rtk.sol.ns directly, which can be 0 early in convergence. The comment on that line ("Some servers don't like when ns = 0") suggests this was intentional protection. Consider sol_nmea.ns = svr->rtk.sol.ns > 0 ? svr->rtk.sol.ns : 10 for those modes.

These are paths with a solution and a solution is only possible with ns>0, and if it did get here with a solution based on no satellites then it would be appropriate for the consumer to discard the position - the change is intentional.