foxitrot wrote:The base idea is good. What about the data gaps/overlaps because of Daylight Saving Time jumps twice a year, interpolate? Or let's ignore DST changes and use the base time zone shift (usually during the winter time)? Neither option is ideal.
I agree that neither option is ideal, but even if you stay with a single timezone, the graphs will be offset by an hour when DST kicks in for the summer - so there'll be either a one hour gap in the data or an hour's worth of data with hour on the clock to put it.
I suggest dropping the hour's worth of extra data when DST ends at the end of the year, and inserting a gap in the graph (or even just a straight line) when DST starts and the clock moves forward.
In any case, the DST problem is a separate problem to supporting multiple timezones - unless you tie all the graphs to UTC (and avoid DST changes completely).
++David \ davidg666