Birthdays have changed?
Ian Gregory
birthdays in Address Book in Tiger, if I didn't know the year they were
born I used "1000". When I upgraded to Leopard a few days ago I did an
erase and install and then restored my home directory from a backup. Now
when I look at birthdays in my Address Book I see something odd:
Birthdays where I had entered the correct year, eg 1991-02-23 were fine,
but birthdays where I had entered the dummy year "1000" to signify an
unknown age seem to have shifted back in time by about 6 days. For
example, John's birthday is on May 25th so I had entered it as
1000-05-25 and for some reason when I looked at it in Leopard it had
changed to 1000-05-19. There were 20 people for whom I had entered
"1000" as the year but that was my only record and I can't remember the
actual birthdays of about half of them. However, the other half (about
10 people) have all shifted back in time by about 6 days. It seems like
some might have shifted 5 or 7 days but I can't be 100% sure because I
might have been out by a day on some of them.
Can anyone offer an explanation? All I could think of is that the year
1000 was before the change to the Gregorian calendar, but that was a 10
day adjustment wasn't it?
Ian
--
Ian Gregory
http://www.zenatode.org.uk/ian/
Ian Gregory
> In article <slrnfvine...@ID-256592.user.individual.net>,
> Ian Gregory <f...@prdetfanaaeextna.invalid> wrote:
>
>> For example, John's birthday is on May 25th so I had entered it as
>> 1000-05-25 and for some reason when I looked at it in Leopard it had
>> changed to 1000-05-19.
>
Can anyone point me to a reference? I tried Googling but all I caeme up
with so far is a bug whereby birthdays can change by one day when you
change timezone setting. In my case I have not changed timezone and
birthdays have changed by six days.
Philo D
Gregory <f...@prdetfanaaeextna.invalid> wrote:
The discrepancy between Gregorian and Julian increases by about 3 days
every 400 years. Today that discrepancy is 13 days. (Note the date of
Russian Orthodox Christmas, for example.) Previous differences
occurred February 29 in: 1900, 1800, 1700, 1500, 1400, 1300, 1100, 1000
... So in the year 1000 the discrepancy would have been 5 days (if
before February 29) and 6 days (if after). As you note, this is before
there *was* a Gregorian calendar, but computers may not take that into
account.
Ian Gregory
> The discrepancy between Gregorian and Julian increases by about 3 days
> every 400 years. Today that discrepancy is 13 days. (Note the date of
> Russian Orthodox Christmas, for example.) Previous differences
> occurred February 29 in: 1900, 1800, 1700, 1500, 1400, 1300, 1100, 1000
> ... So in the year 1000 the discrepancy would have been 5 days (if
> before February 29) and 6 days (if after). As you note, this is before
> there *was* a Gregorian calendar, but computers may not take that into
> account.
Thanks, this might be starting to make sense now. I still haven't found
any relevent bug reports but I did find this:
http://developer.apple.com/releasenotes/Cocoa/AppKit.html
Which includes this:
> NSDatePickerCell's date arithmetic implementation has changed
> substantially in Leopard, abandoning use of the obsolete
> NSCalendarDate class (which only supports the Gregorian calendar, and
> yields imprecise results for dates in the distant past -- e.g. for
> years circa 1500) in favor of a fully modern NSCalendar-based
> implementation underpinned by ICU library routines. This fixes
> significant editing issues for Gregorian calendar dates, while
> providing substantial localization improvements for non-Gregorian
> calendars.
So perhaps when Leopard imported the Tiger Address Book it somehow tried
to correct what it assumed was an inaccurately calculated date. Here
are the cases where I am fairly sure of what I originally entered as the
date - the first date is what I entered and the second is what it
changed to:
1000-01-20 -> 1000-01-15 (-5 days)
1000-05-25 -> 1000-05-19 (-6 days)
1000-06-18 -> 1000-06-12 (-6 days)
1000-11-11 -> 1000-11-06 (-6 days)
1000-12-01 -> 1000-11-25 (-6 days)
Martin Frost me at invalid stanford daht edu
> On 2008-04-07, Philo D <do...@earthling.net.invalid> wrote:
>
> > The discrepancy between Gregorian and Julian increases by about 3 days
> > every 400 years. Today that discrepancy is 13 days. (Note the date of
> > Russian Orthodox Christmas, for example.) Previous differences
> > occurred February 29 in: 1900, 1800, 1700, 1500, 1400, 1300, 1100, 1000
> > ... So in the year 1000 the discrepancy would have been 5 days (if
> > before February 29) and 6 days (if after). As you note, this is before
> > there *was* a Gregorian calendar, but computers may not take that into
> > account.
>
> So perhaps when Leopard imported the Tiger Address Book it somehow tried
> to correct what it assumed was an inaccurately calculated date. Here
> are the cases where I am fairly sure of what I originally entered as the
> date - the first date is what I entered and the second is what it
> changed to:
>
> 1000-01-20 -> 1000-01-15 (-5 days)
> 1000-05-25 -> 1000-05-19 (-6 days)
> 1000-06-18 -> 1000-06-12 (-6 days)
> 1000-11-11 -> 1000-11-06 (-6 days)
> 1000-12-01 -> 1000-11-25 (-6 days)
Which matches the above analysis exactly. Leopard probably didn't
*change* the dates. They are probably stored as integers, and Leopard
is simply *interpreting* the stored integers differently, as explained
above.
Martin
Warren Oates
Martin Frost me at invalid stanford daht edu
<nos...@stanford.edu.invalid> wrote:
> Which matches the above analysis exactly. Leopard probably didn't
> *change* the dates. They are probably stored as integers, and Leopard
> is simply *interpreting* the stored integers differently, as explained
> above.
I love it. There were people in England (and the colonies) in
1750-whatever who thought that the new-style calendar was the devil's
work, those days had been stolen, why is the year so short. Mac thinks
that way.
Thomas Pynchon talks about it a bit in _Mason & Dixon_, and of course,
the way he writes, the way his mind works, you start thinking "hmm,
where _did_ thos days go?"
--
W. Oates
Jeffrey Goldberg
> I love it. There were people in England (and the colonies) in
> 1750-whatever who thought that the new-style calendar was the devil's
> work, those days had been stolen, why is the year so short. Mac thinks
> that way.
I think that there was also the issue that people earned daily wages but
paid monthly rents.
This, of course, why the "October Revolution" is celebrated on November 7.
The Russians only went to the Gregorian calendar after the revolution.
There was a joke going around in Hungary after Chernobyl which wasn't
reported in Hungary (bordering the Ukraine) for nearly a week after the
accident. Of course Hungarians listening to Radio Free Europe or BBC
World Service knew all about it, and were taking their iodine. After TASS
(Soviet news agency) finally announced it the joke was this:
Q: Why do we celebrate the October Revolution in November?
A: Because that's when TASS felt fit to report it.
> Thomas Pynchon talks about it a bit in _Mason & Dixon_, and of course,
> the way he writes, the way his mind works, you start thinking "hmm,
> where _did_ thos days go?"
They left in turning taxis from across the sea.
-j
--
Jeffrey Goldberg http://www.goldmark.org/jeff/
I rarely read top-posted, over-quoting or HTML postings.
http://improve-usenet.org/
J.J. O'Shea
(in article <02501d5c$0$24029$c3e...@news.astraweb.com>):
I, Ugo "the Judge" Boncompagni, took them. You can get them back for a small
fee. Don't call the cops or they might have a slight accident. Capish?
--
email to oshea dot j dot j at gmail dot com.
Warren Oates
<alpine.OSX.1.10.0...@hagrid.ewd.goldmark.org>,
Jeffrey Goldberg <nob...@goldmark.org> wrote:
> They left in turning taxis from across the sea.
Oooh, I haven't read that one for a while ...
--
W. Oates
Wayne C. Morris
J.J. O'Shea <try.n...@but.see.sig> wrote:
> On Mon, 7 Apr 2008 08:08:43 -0400, Warren Oates wrote
> (in article <02501d5c$0$24029$c3e...@news.astraweb.com>):
>
> > In article <myod8mb...@Sunburn.Stanford.EDU>,
> > Martin Frost me at invalid stanford daht edu
> > <nos...@stanford.edu.invalid> wrote:
> >
> >> Which matches the above analysis exactly. Leopard probably didn't
> >> *change* the dates. They are probably stored as integers, and Leopard
> >> is simply *interpreting* the stored integers differently, as explained
> >> above.
> >
> > I love it. There were people in England (and the colonies) in
> > 1750-whatever who thought that the new-style calendar was the devil's
> > work, those days had been stolen, why is the year so short. Mac thinks
> > that way.
> >
> > Thomas Pynchon talks about it a bit in _Mason & Dixon_, and of course,
> > the way he writes, the way his mind works, you start thinking "hmm,
> > where _did_ thos days go?"
> >
>
> I, Ugo "the Judge" Boncompagni, took them. You can get them back for a small
> fee.
Per diem?
Kevin Michael Vail
<wayne.morris-50EF...@shawnews.wp.shawcable.net>,
Carpe diem, which in this case means if you make a fuss you could end up
sleeping with the fishes.
--
Bright eyes/burning like fire, | Kevin Michael Vail
Bright eyes/how can you close and fail? | ke...@vaildc.net
How can the light that shone so brightly | . . . . . . . . . .
Suddenly shine so pale?/Bright eyes | . . . . . . . . .
Jeffrey Goldberg
> Jeffrey Goldberg <nob...@goldmark.org> wrote:
>
>> They left in turning taxis from across the sea.
>
> Oooh, I haven't read that one for a while ...
That's the only one I made it through.
Cheers,
Warren Oates
Jeffrey Goldberg <nob...@goldmark.org> wrote:
> That's the only one I made it through.
Well said.
--
W. Oates
Ian Gregory
Not exactly, because the discrepancy should have increased to 13 days
for modern dates and I would then have expected a date like 1990-03-23
to have changed to 1990-03-10, but modern dates are unaffected. So I
still don't completely understand it. However, I think I can confidently
add either 5 or 6 days as appropriate to all the birthdays where I used
year 1000 to recover the original date.
Matthew T. Russotto
Modern dates were already in the Gregorian calendar. What happened is
the system was treating old dates as if they were in the retrospective
Gregorian calendar, and it is now treating them as if they were in the
Julian calendar.
--
There's no such thing as a free lunch, but certain accounting practices can
result in a fully-depreciated one.
Ian Gregory
OK, I think I can accept this as a more or less full explanation now.
The only question is what they use as a cut-off date for defining
"modern", presumably the "official" date on which the Gregorian
calendar was decreed by Pope Gregory XIII on 24 February 1582.
Matthew T. Russotto
It uses October 4, 1582. Which is sort of surprising; I'd have expect
it to be locale-dependant.
Ian Gregory
> In article <slrnfvnq5...@ID-256592.user.individual.net>,
> Ian Gregory <f...@prdetfanaaeextna.invalid> wrote:
>>
>>OK, I think I can accept this as a more or less full explanation now.
>>The only question is what they use as a cut-off date for defining
>>"modern", presumably the "official" date on which the Gregorian
>>calendar was decreed by Pope Gregory XIII on 24 February 1582.
>
> It uses October 4, 1582. Which is sort of surprising; I'd have expect
> it to be locale-dependant.
>
> http://developer.apple.com/documentation/Cocoa/Reference/Foundation/Classes/NSCalendar_Class/Reference/NSCalendar.html
Cool. I have discovered a very useful calendar converter at Fourmilab:
https://www.fourmilab.ch/documents/calendar/
If I pick any date that got changed between Tiger and Leopard, enter it
in the Gregorian Calendar box and click "calculate" then the date shown
in the Julian Calender box is shifted by the same number of days as
between Tiger and Leopard. Clearly the Tiger was displaying
pre-Gregorian dates in the Gregorian clendar whereas Leopard is
displaying them in the Julian calendar.
I changed the dates in my Address Book manually and switched to using
1901 as a dummy year to avoid future confusion (at least during my
lifetime). I don't know anyone older than 107.