> > This is a very interesting finding. > Especially since the time zone seems not have changed but > rather the file > modification date returned by the OS as of today is different today. > > So > Sun Oct 23 18:05:07 UTC+0200 2011 > seems have changed to > Sun Oct 23 17:05:07 UTC+0200 2011 > > Especially it puzzles me why it's not returning > Sun Oct 23 17:05:07 UTC+0100 2011 > > now which would be the same time normalized to time zones. > > Not sure if there is any API to deal with this issue. I will have to > investigate. It might be able to detect whether a data is in > DST but this > introduces other problems since DST range might be different > in various locations. > Rainer, File time is usually stored as UTC, which does not take DST into account. Only FAT would give problems, but I am using NTFS throughout my systems. That is a really strange thing. Seems that time comparison was never used much in the past, even so it was available for packages. --- Stefan |