Correct use of one-second messages

[davidfokkema]

We synchronize with labVIEW by adding a second to the GPS timestamp. Is that correct? And do we use the correct one-second messages? t, t+1, and t+2.

[davidfokkema]

I added a second before looking up the one-second messages, on station 599 (pysparc1). I restarted the software at 2015/10/05 15:50 local time (13:50 UTC).

[davidfokkema]

I was rewriting the timestamp every time a message was processed, so it was continuously advancing into the future. Not so good. I fixed that now, as of 2015/10/06 10:10 local time (8:10 UTC).

[davidfokkema]

This figure shows the 'old' behaviour (calculating the timestamp, and then adding 1 second to synchronize with labview):

This figure shows what happens when you first add one second, and then calculate the timestamp. You then use GPS accuracy information from messages sent one second later. This is different, and seems to not be what we want:

[davidfokkema]

I reverted to the original code on station 599 (pysparc1), with one-second synchronisation with labVIEW enabled, on 2015/10/9 16:24 local time (14:24 UTC).