A puzzle for pmj (and anyone else)

First a bit of background:
I have lots of images, of local architecture, maps, people, census returns,
etc. Some are digital photographs, some scans. Over the years they have been
copied into various places for various reasons, such as making themed
groups.

I decided to try to keep track of what was where, and in particular which
were duplicates, so I wrote a little Perl program to read through directory
listings and produce a giant listing (actually several, with various
conditions attached), sorted first by filename, then filesize, then file
date and time, then the path.

Now the puzzle - a lot (around a quarter of all the files) of the copies
(same name, same file size) had date & time that differed by exactly one
hour, plus one pair that differs by exactly two hours, and another by one
hour and one minute. There are also 5 pairs with totally different dates and
times, but the same size and name - presumably these were re-saved from an
application rather than actually copied.

The dates of these files seem to be distributed throughout the year, and
some have rolled over from 23:xx to 00:xx the following day (or possibly the
other way). The copies may have been made by drag and drop in Windows
Explorer, or by the f8 copy facility in IrfanView - or both.

I'm just curious: how did making a copy change the time by exactly one hour?

Tickettyboo...

BST ? or the other way round ?

Ali...

That probably has something to do with it, but why?

Parrotfish...

No.......... that would be a Bank ;o)

Tickettyboo...

lol, evenin' P-f ..thought you weren't allowed anything sharp?

Any Ideas?

Jeff Gaines...

I had the hour problem once backing up to my NAS - it hadn't kept its time
up to date when the clocks changed. It decided everything was out of date
and backed up about 300,000 files instead of the few dozen it normally
does, can't remember the details, brain is too sticky. It didn't change
the file times though - copying a file shouldn't change its time as
displayed by Explorer.

Ali...

That's what I thought.

Were they all saved to the same PC? If some were first saved to a PC with
the time correct and others to a PC/NAS with the time out by an hour that
*could* do it.

Ali...

I don't see how. Yes, the timestamp would be an hour wrong, but subsequent
copies should remain consistantly wrong.

Jeff Gaines...

I thought from you original post they were???

Ali...

The actual timestamps were in some cases years wrong, but the the intriguing
thing was that some copies were showing a difference of one hour from their
originals.

See my reply to pmj for what had happened.