I apologize in advance if this opens up wounds for some people.
I've searched the forum and wiki without success to determine how WC handles the transition from Daylight Saving Time to Standard Time. Since the .cat file is storing local times I can see the jump from 01:59 to 03:00. But what about in the fall when the time rolls from 01:59 to 01:00? Are there two sets of entries for 01:00 to 01:59, or just one set of entries? If just one, then which set, the first hour or the second hour?
Unfortunately, this is my first year using WC so I have no historical data to look at.
This is important to know because I'm importing WC files into MySQL. I chose not to use the Custom CGI to call a PHP page because some data is missing from the parameter list, and it floods the web server logs. Thus, I'm processing .cat files.
Already, I've encountered problems with the yyyydbase.csv files from Virtual Weather Station. It too uses local time in the files and did not account for DST being moved from April to March in 2007. Perhaps I didn't update the Vantage Pro firmware in time. In the fall for any year, I cannot find two sets of values from 01:00-01:59 so I presume it overwrote or ignored one set of data.
Surprisingly, the files exported from the Weather Underground web site have the same problem.
This is very disappointing because it means one hour of data has been lost for each year. This is why I always,
always store timestamps in files or databases in UTC format and then convert only for display. Totally avoids the problems of storing a local time and it doesn't matter where the data was gathered and if the locale used DST, double DST, or whatever. It also makes comparing two sets of data much easier.
So, what can I expect to see in November in the 11_WeatherCatData.cat file?
Thanks.
