data in monthly log files

[seattle-dave]

I'm a little confused how these files are created. My file for Sept. has 7 data points for the 23rd at the top; then in the correct order, on the 23rd, there are different sets of data. I've included a text file showing this. The first 7 rows were at the very top of the log file; the rest were where they should be. (first column is day; 2nd column is time)

[steve]

Could you attach the complete file, please? And also zip up your diags folder and attach it - although it's probably too long ago for there to be any evidence of what went wrong.

[seattle-dave]

OK, here they are (note the 9/23 diag file is large)

[steve]

Did you edit the today.ini file? When you closed Cumulus at 15:03 on the 23rd, the last update time was (correctly) 3pm on the 23rd:

9/23/2010 15:03:17.799 : Writing todayfile, LastUpdateTime = 9/23/2010 3:00:00 PM raindaystart = 4.57086610794067

But when you started Cumulus again a couple of minutes later, the last update time (the 'timestamp' in today.ini) was way back in July:

9/23/2010 15:05:08.219 : Read today file: Date = 9/23/2010, LastUpdateTime = 7/27/2010 2:00:00 PM

So Cumulus re-downloaded all of the data back to that date and time, resulting in duplicate/out of sequence data in your log file.

[seattle-dave]

I may have edited the today.ini file (for reasons that escape me now). That answers the question.

Another question: my station is a ws-2080, so I started out using easyWeather. When I compare the easyWeather.dat file with the Cumulus monthly log files, the times are different (the data agree). I assume both programs use the data stored in the station, so why aren't they the same?

As you may have surmised, I'm new to this, so have a lot to learn.
Thanks,

[steve]

I don't completely understand what you're comparing with what, but while it's running, Cumulus logs its data at the intervals you specify, independent of anything else. When it logs data that it downloads from the logger at start up, it has to try to work out what the timestamps are based on the current time and the station's logger interval (the station unhelpfully doesn't timestamp the logger entries). So there could easily be a minute or two difference in the timestamps compared to Easyweather.

[seattle-dave]

I'm comparing the easyWeather.dat file with the Cumulus monthly log files. I assume the easyWeather.dat file contains the actual data stored in the station. It does contain time data. The times in the Cumulus monthly logs are shifted about 10 min from the ones in easyWeather.dat. Cumulus takes its data from the station memory, so I guess I'm missing something.
Thanks

[steve]

I know very little about how Easyweather works, I'm afraid. I do know that the station doesn't timestamp its logger entries and they have to be calculated by whatever software downloads them. The timestamps in Cumulus depend on whether the data was logged 'live' or from the logger. If you're comparing timestamps in Cumulus for data from the logger with timestamps in Easyweather for data from the logger (which in the case of EW I think it always is), then in theory the timestamps should be within a minute or two. But I don't know how EW calculates the timestamps.