jtsage
commented
12 years ago Hmm - interesting. fwiw, the purpose of that line is to assure that a pure-time driving datebox does not run into daylight savings time, thus losing and hour (or repeating one I suppose). I am more than happy to make that switch if it's screwing something else up, but both of those dates should be "safe". (I think) (although, I bet when I initialize it earlier I did it right, which could mean this introduces some oddness....)
Any operation internally calling _makeDate in a time formatted box will return Feb 1st 1901 instead of Jan 1st 1900, I think line 491 (v1.1.0) is where the problem is, it reads: date = new w._date(w._n(d.year,1),w._n(d.mont,1),w._n(d.date,1),w._n(d.hour,0),w._n(d.mins,0),w._n(d.secs,0),0); when it should read: date = new w._date(w._n(d.year,0),w._n(d.mont,0),w._n(d.date,1),w._n(d.hour,0),w._n(d.mins,0),w._n(d.secs,0),0); since both month and dates start at base 0 (actually year could be set to anything, so 1901 vs 1900 is a matter of taste here).