search

matc17 / FS2020NMEA

MIT License
0 stars 0 forks source link

Date stamps appear to have wrong format #2

Open jeffchrisope opened 3 years ago

jeffchrisope commented 3 years ago

On the Facebook for MSFS Gliders (https://www.facebook.com/groups/1803376689823050), the following issue was reported: 'Hmmm...exported some IGC files from XCSoar but can't get anything to analyse them - all the tools I've tried say something like "Missing date record" ???'

I'm not sure if the data needs to contain an actual ZDA sentence, or whether the datestamps on the RMC sentences are sufficient for IGC reconstruction later, but the latter appear to have 8-digit datestamps, and the reference I see requires 6-digit (a 2-digit year, not 4-). Maybe that's the issue?

matc17 commented 3 years ago

The converter currently only provides the minimum GPS sentences, this is correct. However in properly coded tools this should be sufficient. For example Santiago Hollmann's IGC Viewer works as expected.

Also literally all nmea documentation shows 6 digit timestamp for GPRMC, for example here: http://www.nmea.de/nmea0183datensaetze.html#rmc.

The only issue is off course that IGC file is pretty useless meanwhile since it was demanded for 0.2 to also have permanent GPS fix (also prior to starting a flight) so I needed to set some coordinates before knowing where the flight would happen. As a consequence you will have jumps of thousands of kilometers in your log file which imho makes it pretty useless.....

jeffchrisope commented 3 years ago

I was referring to the actual ‘Date’ element of the RMC sentence - i.e. item #9 in the reference you linked to. That ref is confusing, the xxxx makes it seem like it’s just ddmm, but the note says ddmmyy. However, unless I missed something when monitoring FS2020NMEA, you appear to be outputting ddmmyyyy, which possibly violates the GMC spec? Or maybe some tools just don’t tolerate all forms of the spec, wouldn’t surprise me.

matc17 commented 3 years ago

OK, now I got your point, sorry.

I pushed 0.3a test release with 2 digit as year. Maybe you could try with that? As I cannot reproduce the issue I can only guess :(