Chinese timezones

[Jesper Norgaard Welen]

The chinese timezones don't have much description in the tz database. I am
currently trying to make a map of the worlds timezones base on the Olson
database, and so I miss some more information about what these timezones
cover.

* Asia/Harbin seems to cover the Heilongjiang region, no less no more.

* Asia/Chongqing is a city lying in the Sichuan province, which is huge, so
does the timezone represent all of the province?

* Asia/Urumqi covers the Xinjiang province, but represents a definition
problem since Kashgar also lies in the far western part of this province, so
what does timezone Asia/Kashgar represent? Asia/Urumqi timezone according to
the comments in the 'asia' file also covers Tibet, which lies in the Xizang
province, that is huge stretching from Kashmir to Myanmar, so I don't know
if all of this is covered by the Asia/Urumqi timezone alone.

* Asia/Shanghai covers all the rest of China as I understand it. This is the
timezone there was recently a discussion about changing to Asia/Beijing, I
believe.

Can anyone shed some light on this?

- Jesper N?rgaard Welen

[Zhe Su]

Hi,
Actually those are legacy timezones which are not used anymore. Now
only one time zone is used for whole China, it's now officially called
"China Standard Time". And is known widely as "Beijing Time".
According to the rule of timezone database, Asia/Shanghai is used to
represent "China Standard Time", because Shanghai is(was) the most
populous city in China.
Correct me if I'm wrong.

Regards
James Su

Post by Jesper Norgaard Welen
The chinese timezones don't have much description in the tz database. I am
currently trying to make a map of the worlds timezones base on the Olson
database, and so I miss some more information about what these timezones
cover.
* Asia/Harbin seems to cover the Heilongjiang region, no less no more.
* Asia/Chongqing is a city lying in the Sichuan province, which is huge, so
does the timezone represent all of the province?
* Asia/Urumqi covers the Xinjiang province, but represents a definition
problem since Kashgar also lies in the far western part of this province, so
what does timezone Asia/Kashgar represent? Asia/Urumqi timezone according to
the comments in the 'asia' file also covers Tibet, which lies in the Xizang
province, that is huge stretching from Kashmir to Myanmar, so I don't know
if all of this is covered by the Asia/Urumqi timezone alone.
* Asia/Shanghai covers all the rest of China as I understand it. This is the
timezone there was recently a discussion about changing to Asia/Beijing, I
believe.
Can anyone shed some light on this?
- Jesper N?rgaard Welen

[Paul Eggert]

Post by Jesper Norgaard Welen
* Asia/Harbin seems to cover the Heilongjiang region, no less no more.

It also includes Jilin.

Post by Jesper Norgaard Welen
* Asia/Chongqing is a city lying in the Sichuan province, which is huge, so
does the timezone represent all of the province?

Yes. It also includes east Gansu, Guangxi Zhuang, Guizhu, Ningxia
Hui, Sichuan, Yunnan, and perhaps some other areas (I haven't checked
Qinghai yet, for example).

Shanks & Pottenger write that the time zone boundaries did not follow
provincial boundaries: Gansu was split, for example. This may help to
explain Asia/Urumqi versus Asia/Kashgar versus whatever: I suspect
Xinjiang and Tibet were split though I haven't had the time to find
the boundaries. Also Inner Mongolia.

Post by Jesper Norgaard Welen
* Asia/Shanghai covers all the rest of China as I understand it.

Yes.

I'll try to pry loose some time and find out more precisely what the
boundaries were, according to Shanks & Pottenger's data. It's not a
trivial task, I'm afraid.

Which provincial boundaries have moved since 1980? This knowledge
would help me in my task. For example,
<Loading Image...> says
that the Qinghai-Gansu border has moved but gives no details.

Post by Jesper Norgaard Welen
Actually those are legacy timezones which are not used anymore.

Yes, if you don't care about pre-1980 time stamps, then you can use
TZ='Asia/Shanghai', or even TZ='CST-8' on POSIX hosts.

However, if you do care about old time stamps, then Asia/Harbin
etc. are not "legacy", because they have a practical difference for
programs running on today's computers.

[Andy Lipscomb]

"

Post by Zhe Su
Actually those are legacy timezones which are not used anymore.

Yes, if you don't care about pre-1980 time stamps, then you can use TZ='Asia/Shanghai', or even TZ='CST-8' on POSIX hosts.

However, if you do care about old time stamps, then Asia/Harbin etc. are not "legacy", because they have a practical difference for programs running on today's computers.
"

Perhaps a better term would be "historical" time zones. This would in fact be a useful distinction, since only a relatively small subset of the timezones are necessary for applications that look only forward--these would be the first set to present to a user for selecting a time zone, with a "show all" option to get the complete list. (For example, the 50 United States can be covered on a forward-looking basis by eight zones--New York, Chicago, Denver, Phoenix, Los Angeles, Anchorage, Adak, and Honolulu. And if one takes as irrevocable the power of the EU over DST dates, then--for example--London, Paris, and Helsinki would cover that entire area, except for the different abbreviations used in the three +0 countries.)

[Robert Elz]

Date: Mon, 26 Jun 2006 14:21:24 -0400
From: "Andy Lipscomb" <AndyLipscomb at decosimo.com>
Message-ID: <E3827447A4261D4C80192E07BB7E649102438A39 at chaexc01.decosimo.local>

| Perhaps a better term would be "historical" time zones.

No, time zones including historical information. They're not obsolete.

| This would in fact be a useful distinction, since only a relatively
| small subset of the timezones are necessary for applications that
| look only forward

That may be true, but ...

| --these would be the first set to present to a user

That would be foolish. Users typically select their timezone just once
(and certainly they should need to, assuming they don't relocate, and zones
don't split - both of which are fairly high visibility activities).
If the user selects a zone that seems to work just fine for most applications,
but mysteriously gives incorrect answers for that relatively small subset
(like anything that prints the modification time of a file that might
have been last modified in the past) then we have achieved a poor result.

It may be a bit harder to explain, and we may not yet have found the best
way to either do that, or allow users to select the correct zone, but doing
that additional work at the time when the user is consciously considering
the issue is almost certainly going to pay off in the long term.

kre

[Robert Elz]

Date: Mon, 26 Jun 2006 14:21:24 -0400
From: "Andy Lipscomb" <AndyLipscomb at decosimo.com>
Message-ID: <E3827447A4261D4C80192E07BB7E649102438A39 at chaexc01.decosimo.local>

| Perhaps a better term would be "historical" time zones.

No, time zones including historical information. They're not obsolete.

| This would in fact be a useful distinction, since only a relatively
| small subset of the timezones are necessary for applications that
| look only forward

That may be true, but ...

| --these would be the first set to present to a user

That would be foolish. Users typically select their timezone just once
(and certainly they should need to, assuming they don't relocate, and zones
don't split - both of which are fairly high visibility activities).
If the user selects a zone that seems to work just fine for most applications,
but mysteriously gives incorrect answers for that relatively small subset
(like anything that prints the modification time of a file that might
have been last modified in the past) then we have achieved a poor result.

It may be a bit harder to explain, and we may not yet have found the best
way to either do that, or allow users to select the correct zone, but doing
that additional work at the time when the user is consciously considering
the issue is almost certainly going to pay off in the long term.

kre

[Robert Elz]

Date: Mon, 26 Jun 2006 20:36:43 +0800
From: "Zhe Su" <james.su at gmail.com>
Message-ID: <278a3d0606260536y10e368c0w2ee8afb2bb7e3370 at mail.gmail.com>

| Actually those are legacy timezones which are not used anymore.

That's probably correct if you're only considering what happens today,
and probably (or perhaps) tomorrow. But the timezone database wants
to also consider the past, so timestamps made years ago can be correctly
interpreted.

To answer Jesper's question, you should (if you are able) try to give
information on where those old timezones applied, when they were in use
(just what were the boundaries of applicability).

The timezone database itself doesn't care (we're not cartographers).
but others do.

kre

ps: could we please delete from the list anyone who has a mailer that
(insanely) sends disposition notifications (or return receipts, or whatever
you want to call them) to the list, instead of to the address requested.
Then once we've found and eliminated all of those, we can delete from the
list anyone stupid enough to send messages to a mailing list requesting
disposition notification, and so feeding the bugs in all those broken mailers.

[Jesper Norgaard Welen]

I have two resources, one is a map with the following 30 regions of China:

Hainan
Hong Kong
Anhui
Zhejiang
Jiangxi
Jiangsu
Jilin
Qinghai
Fujian
Heilongjiang
Henan
Hebei
Hunan
Hubei
Xinjiang
Xizang
Gansu
Guangxi
Guizhou
Liaoning
Nei Mongol
Ningxia
Beijing
Shanghai
Shanxi
Shandong
Shaanxi
Sichuan
Tianjin
Yunnan
Guangdong

The other is a map with around 2800+ counties of China, but I have
absolutely no idea how to combine it to get the corresponding timezones of
tz database (or Shanks and Pottenger, if you like). I can combine the maps
and see which counties lie within a region. I can also copy-paste map
fractions to combine the final timezone map that I want. But I don't have
any information how to combine it apart from the tz database itself, which
has very little detail (Paul, your email was more helpful than the tz
database itself!). So I have a good opportunity to a fine-granularity
approach.

Clearly, it must also be possible to improve the text in the tz database
that defines the timezones, for instance to include Jilin explicitly etc.

I can't help with the changes of chinese administrative boundaries since
1980, or earlier. If the timezones did not follow administrative boundaries,
that will make my task more difficult of course.

I send you a GIF file Gansu.gif within Gansu.zip that contains a screenshot
of a portion of the Gansu region with its counties (many mail programs only
accept zip files and no other types of attachments, for fear of imbeddeded
viruses, trojans etc. Nowadays, you can't even trust your browser!).

Post by Zhe Su
Actually those are legacy timezones which are not used anymore.
Paul Eggert writes
Yes, if you don't care about pre-1980 time stamps, then you can use
TZ='Asia/Shanghai', or even TZ='CST-8' on POSIX hosts.
However, if you do care about old time stamps, then Asia/Harbin etc.
are not "legacy", because they have a practical difference for
programs running on today's computers.

I guess I forgot to mention that it is the historic timezones that interest
me, while the current division of China in only one timezone is rather
trivial. The idea is to be able to calculate what was the local time in
Harbin or Lishui when it was 10:15:00 GMT 14.th. of August 1941 (as an
example). And no, the answer is surely not 18:15:00!

Regards,
- Jesper
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Gansu.zip
Type: application/x-zip-compressed
Size: 25752 bytes
Desc: not available
Url : http://mm.icann.org/pipermail/tz/attachments/20060627/28846e42/Gansu.zip

[Paul Eggert]

it is the historic timezones that interest me, while the current
division of China in only one timezone is rather trivial. The idea
is to be able to calculate what was the local time in Harbin or
Lishui when it was 10:15:00 GMT 14.th. of August 1941 (as an
example). And no, the answer is surely not 18:15:00!

For that sort of thing the Shanks & Pottenger database is the best
that I know of, though of course it's not infallible. They do attempt
to cover the wartime changes. Shanks was born during Eastern War Time
and notes this fact in the author's bio on his book.

You should be able to determine the zone boundaries yourself, if you
have good maps, by consulting the Atlas Query
<http://www.astro.com/cgi/aq.cgi>, which is an online version of the
Shanks & Pottenger database. For example, I just verified that
according to that database, Harbin was at UTC+0830 on 1970-01-01
00:00. As far as I know, the only post 1970 change in China was in
May 1980, so if you use any time stamp from 1970 to 1980 you should be
able to wander around the Chinese countryside using that database and
determine where the time zone boundaries were. (Good luck with the
spellings, though!)

(Paul, your email was more helpful than the tz
database itself!).

Yes, I had started to update 'asia' when I answered your email.
Here's its current (incomplete) state.

--- asia 2006/05/01 13:44:41 2006.6
+++ asia 2006/06/24 21:29:46
@@ -220,6 +220,7 @@ Rule PRC 1987 1991 - Apr Sun>=10 0:00 1:
# historic timezones from some Taiwan websites. And yes, there are official
# Chinese names for these locales (before 1949):
# Changbai Time ("Long-white Time", Long-white = Heilongjiang area)
+# Heilongjiang, Jilin
Zone Asia/Harbin 8:26:44 - LMT 1928 # or Haerbin
8:30 - CHAT 1932 Mar # Changbai Time
8:00 - CST 1940
@@ -227,10 +228,12 @@ Zone Asia/Harbin 8:26:44 - LMT 1928 # or
8:30 - CHAT 1980 May
8:00 PRC C%sT
# Zhongyuan Time ("Central plain Time")
+# most of China
Zone Asia/Shanghai 8:05:52 - LMT 1928
8:00 Shang C%sT 1949
8:00 PRC C%sT
# Long-shu Time (probably due to Long and Shu being two names of that area)
+# east Gansu, Guangxi Zhuang, Guizhu, Ningxia Hui, Sichuan, Yunnan
Zone Asia/Chongqing 7:06:20 - LMT 1928 # or Chungking
7:00 - LONT 1980 May # Long-shu Time
8:00 PRC C%sT

[Paul Eggert]

it is the historic timezones that interest me, while the current
division of China in only one timezone is rather trivial. The idea
is to be able to calculate what was the local time in Harbin or
Lishui when it was 10:15:00 GMT 14.th. of August 1941 (as an
example). And no, the answer is surely not 18:15:00!

For that sort of thing the Shanks & Pottenger database is the best
that I know of, though of course it's not infallible. They do attempt
to cover the wartime changes. Shanks was born during Eastern War Time
and notes this fact in the author's bio on his book.

You should be able to determine the zone boundaries yourself, if you
have good maps, by consulting the Atlas Query
<http://www.astro.com/cgi/aq.cgi>, which is an online version of the
Shanks & Pottenger database. For example, I just verified that
according to that database, Harbin was at UTC+0830 on 1970-01-01
00:00. As far as I know, the only post 1970 change in China was in
May 1980, so if you use any time stamp from 1970 to 1980 you should be
able to wander around the Chinese countryside using that database and
determine where the time zone boundaries were. (Good luck with the
spellings, though!)

(Paul, your email was more helpful than the tz
database itself!).

Yes, I had started to update 'asia' when I answered your email.
Here's its current (incomplete) state.

--- asia 2006/05/01 13:44:41 2006.6
+++ asia 2006/06/24 21:29:46
@@ -220,6 +220,7 @@ Rule PRC 1987 1991 - Apr Sun>=10 0:00 1:
# historic timezones from some Taiwan websites. And yes, there are official
# Chinese names for these locales (before 1949):
# Changbai Time ("Long-white Time", Long-white = Heilongjiang area)
+# Heilongjiang, Jilin
Zone Asia/Harbin 8:26:44 - LMT 1928 # or Haerbin
8:30 - CHAT 1932 Mar # Changbai Time
8:00 - CST 1940
@@ -227,10 +228,12 @@ Zone Asia/Harbin 8:26:44 - LMT 1928 # or
8:30 - CHAT 1980 May
8:00 PRC C%sT
# Zhongyuan Time ("Central plain Time")
+# most of China
Zone Asia/Shanghai 8:05:52 - LMT 1928
8:00 Shang C%sT 1949
8:00 PRC C%sT
# Long-shu Time (probably due to Long and Shu being two names of that area)
+# east Gansu, Guangxi Zhuang, Guizhu, Ningxia Hui, Sichuan, Yunnan
Zone Asia/Chongqing 7:06:20 - LMT 1928 # or Chungking
7:00 - LONT 1980 May # Long-shu Time
8:00 PRC C%sT

[Gwillim Law]

You can find basic information about provincial boundary changes on my
"Statoids" site at http://www.statoids.com/ucn.html. The only major changes
since 1980: the formation of Hainan province and Chongqing municipality,
and the annexation of Hong Kong and Macau. I'll see if I can find anything
about the Qinghai-Gansu border.

-- Gwillim Law

----- Original Message -----
From: "Paul Eggert" <eggert at CS.UCLA.EDU>

Post by Paul Eggert
Which provincial boundaries have moved since 1980? This knowledge
would help me in my task. For example,
<http://en.wikipedia.org/wiki/Image:China_administrative.png> says
that the Qinghai-Gansu border has moved but gives no details.

[Jesper Norgaard Welen]

Thanks for this magnificent link. I had already been using
http://www.astro.com as one source for determining latitude/longitude for
cities. But I had no idea that you could actually use it as a timezone
converter, even for historic times, and indeed implements the Shanks &
Pottenger database.

I have been following your advice and wandered around to find possible
boundaries of timezones. In the "Beijing" area I have not found any
deviations, but I have also not pushed it far. So Asia/Shanghai seems to
cover the regions Hebei, Shandong, Fujian, Jingxi, Guangdong, Hunan, Hubei,
Anhui, Henan, Shanxi, Zhejiang, Jiangsu and Liaoning. All are on GMT+8 in
1970.

In Qinghai there was a zone on GMT+7 with Haiyan, Gonghe, Tongren, Jigzhi,
Xining, and another zone on GMT+6 with Yushi, Zhidoi, Qumalai, Nangqen,
Delingha, Golmud, Lenghu and Mangya.

In Tibet (Xizang) there was a zone on GMT+6 with Lhasa, Chamdo, Shigaise,
Jimsar, Shawan and Hutubi. Another zone on GMT+5 with Pulan, Aheqi, Shufu,
Shule .

On GMT+7 there were also the regions Shaanxi, Nei Mongol and Hainan.

In Gansu it seems the vast majority of the area was on GMT+7 right up until
its westernmost part, where I found a single city Dunhuang on GMT+6. I'm
sure there most be more cities in this area on GMT+6 but I haven't got
around to check it. The cities on GMT+7 were Lanzhou, Linxia, Tianshui,
Pingliang, Wuwei, Zhangye, and Yumen.

My major effort has been on the Xinjiang area, which has a lot of counties,
so it is not easy to determine a map. The cities on GMT+5 were Aksu, Atushi,
Yining, Hetian, Cele, Luopu, Nileke, Zhaosu, Tekesi, Gongliu, Chabuchaer,
Huocheng, Bole, Pishan, Suiding, Yarkand. The cities on GMT+6 were Urumqi,
Turpan, Karamay, Korla, Minfeng, Jinghe, Wusu, Qiemo, Xinyan, Wulanwusu,
Jinghe, Yumin, Tacheng, Tuoli, Emin, Shihezi, Changji, Yanqi, Heshuo,
Tuokexun, Tulufan, Shanshan, Hami, Fukang, Kuitun, Kumukuli, Miquan, Qitai
and Turfan.

The cities in Xinjiang I could find lies in the upper and lower part of the
region, while there is a central zone where I could not find any cities that
coincided with astro.com, but this might also be because it is a big desert
area. Many counties have a name that corresponds to the city name also, so I
could often find it that way, but not for the following counties:

Awati 40.11 80.57
Shaya 40.53 81.43
Kuche 41.77 83.51
Xinhe 41.34 82.13
Baicheng 42.07 81.97
Wensu 41.59 80.38
Wushi 41.20 79.13
Luntai 41.70 84.65
Yuli 40.83 87.00
Hejing 42.88 85.32

The coordinates above correspond to the approximate center of the county.
I'm looking for cities near these county centers that would lie inside the
county and exist in www.astro.com to produce a GMT offset for 1970. Can
anyone help?

Unfortunately there is apparently no "show nearby cities" or map in
astro.com, so it is a bit of guesswork. The amount of different spelling for
each name is also amazing, but perhaps not if you consider "anglified name",
yughur, pinyin, traditional chinese, kazakh and uzbekh as valid variations,
probably flavored with a nice collection of typos too.

Regards,
- Jesper

Paul Eggert writes:

-----Original Message-----
From: Paul Eggert [mailto:eggert at cs.ucla.edu]
Sent: Martes, 27 de Junio de 2006 12:05
To: Jesper Norgaard Welen
Cc: tz at lecserver.nci.nih.gov
Subject: Re: Chinese timezones


For that sort of thing the Shanks & Pottenger database is the best that I
know of, though of course it's not infallible. They do attempt to cover the
wartime changes. Shanks was born during Eastern War Time and notes this
fact in the author's bio on his book.

You should be able to determine the zone boundaries yourself, if you have
good maps, by consulting the Atlas Query <http://www.astro.com/cgi/aq.cgi>,
which is an online version of the Shanks & Pottenger database. For example,
I just verified that according to that database, Harbin was at UTC+0830 on
1970-01-01 00:00. As far as I know, the only post 1970 change in China was
in May 1980, so if you use any time stamp from 1970 to 1980 you should be
able to wander around the Chinese countryside using that database and
determine where the time zone boundaries were. (Good luck with the
spellings, though!)

[Oscar van Vlijmen]

From: Jesper Norgaard Welen
The cities in Xinjiang I could find lies in the upper and lower part of the
region, while there is a central zone where I could not find any cities that
coincided with astro.com, but this might also be because it is a big desert
area. Many counties have a name that corresponds to the city name also, so I
Awati 40.11 80.57
Shaya 40.53 81.43
Kuche 41.77 83.51
Xinhe 41.34 82.13
Baicheng 42.07 81.97
Wensu 41.59 80.38
Wushi 41.20 79.13
Luntai 41.70 84.65
Yuli 40.83 87.00
Hejing 42.88 85.32
The coordinates above correspond to the approximate center of the county.
I'm looking for cities near these county centers that would lie inside the
county and exist in www.astro.com to produce a GMT offset for 1970. Can
anyone help?

As is probably known the GeoNames database lists nearly every rock and creek
with a name on it.
http://earth-info.nga.mil/gns/html/cntry_files.html
The China file is more than 20 MB large and has more than 160,000 entries.
If you're interested, I managed to extract all 961 names lying in the
rectangle with longitude 79-87 and latitude 39-44 deg.
The list contains several alternative spellings and codes with type of
location like populated place, main administrative division etcetera. The
codes are explained on pages, found at earth-info.nga.mil.
I can't say that a list with 960 names is a real help, but that's what I can
offer.

[Clive D.W. Feather]

Post by Oscar van Vlijmen
As is probably known the GeoNames database lists nearly every rock and creek
with a name on it.
http://earth-info.nga.mil/gns/html/cntry_files.html

However, it doesn't list the village (in UK terminology - the population is
over 5,000) where I live.

I can't seem to find a link on that site for sending them corrections.

[Clive D.W. Feather]

Post by Oscar van Vlijmen
As is probably known the GeoNames database lists nearly every rock and creek
with a name on it.
http://earth-info.nga.mil/gns/html/cntry_files.html

However, it doesn't list the village (in UK terminology - the population is
over 5,000) where I live.

I can't seem to find a link on that site for sending them corrections.

[Jesper Norgaard Welen]

The chinese timezones don't have much description in the tz database. I am
currently trying to make a map of the worlds timezones base on the Olson
database, and so I miss some more information about what these timezones
cover.

* Asia/Harbin seems to cover the Heilongjiang region, no less no more.

* Asia/Chongqing is a city lying in the Sichuan province, which is huge, so
does the timezone represent all of the province?

* Asia/Urumqi covers the Xinjiang province, but represents a definition
problem since Kashgar also lies in the far western part of this province, so
what does timezone Asia/Kashgar represent? Asia/Urumqi timezone according to
the comments in the 'asia' file also covers Tibet, which lies in the Xizang
province, that is huge stretching from Kashmir to Myanmar, so I don't know
if all of this is covered by the Asia/Urumqi timezone alone.

* Asia/Shanghai covers all the rest of China as I understand it. This is the
timezone there was recently a discussion about changing to Asia/Beijing, I
believe.

Can anyone shed some light on this?

- Jesper N?rgaard Welen

[Zhe Su]

Hi,
Actually those are legacy timezones which are not used anymore. Now
only one time zone is used for whole China, it's now officially called
"China Standard Time". And is known widely as "Beijing Time".
According to the rule of timezone database, Asia/Shanghai is used to
represent "China Standard Time", because Shanghai is(was) the most
populous city in China.
Correct me if I'm wrong.

Regards
James Su

Post by Jesper Norgaard Welen
The chinese timezones don't have much description in the tz database. I am
currently trying to make a map of the worlds timezones base on the Olson
database, and so I miss some more information about what these timezones
cover.
* Asia/Harbin seems to cover the Heilongjiang region, no less no more.
* Asia/Chongqing is a city lying in the Sichuan province, which is huge, so
does the timezone represent all of the province?
* Asia/Urumqi covers the Xinjiang province, but represents a definition
problem since Kashgar also lies in the far western part of this province, so
what does timezone Asia/Kashgar represent? Asia/Urumqi timezone according to
the comments in the 'asia' file also covers Tibet, which lies in the Xizang
province, that is huge stretching from Kashmir to Myanmar, so I don't know
if all of this is covered by the Asia/Urumqi timezone alone.
* Asia/Shanghai covers all the rest of China as I understand it. This is the
timezone there was recently a discussion about changing to Asia/Beijing, I
believe.
Can anyone shed some light on this?
- Jesper N?rgaard Welen

[Paul Eggert]

Post by Jesper Norgaard Welen
* Asia/Harbin seems to cover the Heilongjiang region, no less no more.

It also includes Jilin.

Post by Jesper Norgaard Welen
* Asia/Chongqing is a city lying in the Sichuan province, which is huge, so
does the timezone represent all of the province?

Yes. It also includes east Gansu, Guangxi Zhuang, Guizhu, Ningxia
Hui, Sichuan, Yunnan, and perhaps some other areas (I haven't checked
Qinghai yet, for example).

Shanks & Pottenger write that the time zone boundaries did not follow
provincial boundaries: Gansu was split, for example. This may help to
explain Asia/Urumqi versus Asia/Kashgar versus whatever: I suspect
Xinjiang and Tibet were split though I haven't had the time to find
the boundaries. Also Inner Mongolia.

Post by Jesper Norgaard Welen
* Asia/Shanghai covers all the rest of China as I understand it.

Yes.

I'll try to pry loose some time and find out more precisely what the
boundaries were, according to Shanks & Pottenger's data. It's not a
trivial task, I'm afraid.

Which provincial boundaries have moved since 1980? This knowledge
would help me in my task. For example,
<http://en.wikipedia.org/wiki/Image:China_administrative.png> says
that the Qinghai-Gansu border has moved but gives no details.

Post by Jesper Norgaard Welen
Actually those are legacy timezones which are not used anymore.

Yes, if you don't care about pre-1980 time stamps, then you can use
TZ='Asia/Shanghai', or even TZ='CST-8' on POSIX hosts.

However, if you do care about old time stamps, then Asia/Harbin
etc. are not "legacy", because they have a practical difference for
programs running on today's computers.

[Andy Lipscomb]

"

Post by Zhe Su
Actually those are legacy timezones which are not used anymore.

Yes, if you don't care about pre-1980 time stamps, then you can use TZ='Asia/Shanghai', or even TZ='CST-8' on POSIX hosts.

However, if you do care about old time stamps, then Asia/Harbin etc. are not "legacy", because they have a practical difference for programs running on today's computers.
"

Perhaps a better term would be "historical" time zones. This would in fact be a useful distinction, since only a relatively small subset of the timezones are necessary for applications that look only forward--these would be the first set to present to a user for selecting a time zone, with a "show all" option to get the complete list. (For example, the 50 United States can be covered on a forward-looking basis by eight zones--New York, Chicago, Denver, Phoenix, Los Angeles, Anchorage, Adak, and Honolulu. And if one takes as irrevocable the power of the EU over DST dates, then--for example--London, Paris, and Helsinki would cover that entire area, except for the different abbreviations used in the three +0 countries.)

[Robert Elz]

Date: Mon, 26 Jun 2006 20:36:43 +0800
From: "Zhe Su" <james.su at gmail.com>
Message-ID: <278a3d0606260536y10e368c0w2ee8afb2bb7e3370 at mail.gmail.com>

| Actually those are legacy timezones which are not used anymore.

That's probably correct if you're only considering what happens today,
and probably (or perhaps) tomorrow. But the timezone database wants
to also consider the past, so timestamps made years ago can be correctly
interpreted.

To answer Jesper's question, you should (if you are able) try to give
information on where those old timezones applied, when they were in use
(just what were the boundaries of applicability).

The timezone database itself doesn't care (we're not cartographers).
but others do.

kre

ps: could we please delete from the list anyone who has a mailer that
(insanely) sends disposition notifications (or return receipts, or whatever
you want to call them) to the list, instead of to the address requested.
Then once we've found and eliminated all of those, we can delete from the
list anyone stupid enough to send messages to a mailing list requesting
disposition notification, and so feeding the bugs in all those broken mailers.

[Jesper Norgaard Welen]

I have two resources, one is a map with the following 30 regions of China:

Hainan
Hong Kong
Anhui
Zhejiang
Jiangxi
Jiangsu
Jilin
Qinghai
Fujian
Heilongjiang
Henan
Hebei
Hunan
Hubei
Xinjiang
Xizang
Gansu
Guangxi
Guizhou
Liaoning
Nei Mongol
Ningxia
Beijing
Shanghai
Shanxi
Shandong
Shaanxi
Sichuan
Tianjin
Yunnan
Guangdong

The other is a map with around 2800+ counties of China, but I have
absolutely no idea how to combine it to get the corresponding timezones of
tz database (or Shanks and Pottenger, if you like). I can combine the maps
and see which counties lie within a region. I can also copy-paste map
fractions to combine the final timezone map that I want. But I don't have
any information how to combine it apart from the tz database itself, which
has very little detail (Paul, your email was more helpful than the tz
database itself!). So I have a good opportunity to a fine-granularity
approach.

Clearly, it must also be possible to improve the text in the tz database
that defines the timezones, for instance to include Jilin explicitly etc.

I can't help with the changes of chinese administrative boundaries since
1980, or earlier. If the timezones did not follow administrative boundaries,
that will make my task more difficult of course.

I send you a GIF file Gansu.gif within Gansu.zip that contains a screenshot
of a portion of the Gansu region with its counties (many mail programs only
accept zip files and no other types of attachments, for fear of imbeddeded
viruses, trojans etc. Nowadays, you can't even trust your browser!).

Post by Zhe Su
Actually those are legacy timezones which are not used anymore.
Paul Eggert writes
Yes, if you don't care about pre-1980 time stamps, then you can use
TZ='Asia/Shanghai', or even TZ='CST-8' on POSIX hosts.
However, if you do care about old time stamps, then Asia/Harbin etc.
are not "legacy", because they have a practical difference for
programs running on today's computers.

I guess I forgot to mention that it is the historic timezones that interest
me, while the current division of China in only one timezone is rather
trivial. The idea is to be able to calculate what was the local time in
Harbin or Lishui when it was 10:15:00 GMT 14.th. of August 1941 (as an
example). And no, the answer is surely not 18:15:00!

Regards,
- Jesper
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Gansu.zip
Type: application/x-zip-compressed
Size: 25752 bytes
Desc: not available
Url : http://mm.icann.org/pipermail/tz/attachments/20060627/28846e42/Gansu-0001.zip

[Gwillim Law]

You can find basic information about provincial boundary changes on my
"Statoids" site at http://www.statoids.com/ucn.html. The only major changes
since 1980: the formation of Hainan province and Chongqing municipality,
and the annexation of Hong Kong and Macau. I'll see if I can find anything
about the Qinghai-Gansu border.

-- Gwillim Law

----- Original Message -----
From: "Paul Eggert" <eggert at CS.UCLA.EDU>

Post by Paul Eggert
Which provincial boundaries have moved since 1980? This knowledge
would help me in my task. For example,
<http://en.wikipedia.org/wiki/Image:China_administrative.png> says
that the Qinghai-Gansu border has moved but gives no details.

[Jesper Norgaard Welen]

Thanks for this magnificent link. I had already been using
http://www.astro.com as one source for determining latitude/longitude for
cities. But I had no idea that you could actually use it as a timezone
converter, even for historic times, and indeed implements the Shanks &
Pottenger database.

I have been following your advice and wandered around to find possible
boundaries of timezones. In the "Beijing" area I have not found any
deviations, but I have also not pushed it far. So Asia/Shanghai seems to
cover the regions Hebei, Shandong, Fujian, Jingxi, Guangdong, Hunan, Hubei,
Anhui, Henan, Shanxi, Zhejiang, Jiangsu and Liaoning. All are on GMT+8 in
1970.

In Qinghai there was a zone on GMT+7 with Haiyan, Gonghe, Tongren, Jigzhi,
Xining, and another zone on GMT+6 with Yushi, Zhidoi, Qumalai, Nangqen,
Delingha, Golmud, Lenghu and Mangya.

In Tibet (Xizang) there was a zone on GMT+6 with Lhasa, Chamdo, Shigaise,
Jimsar, Shawan and Hutubi. Another zone on GMT+5 with Pulan, Aheqi, Shufu,
Shule .

On GMT+7 there were also the regions Shaanxi, Nei Mongol and Hainan.

In Gansu it seems the vast majority of the area was on GMT+7 right up until
its westernmost part, where I found a single city Dunhuang on GMT+6. I'm
sure there most be more cities in this area on GMT+6 but I haven't got
around to check it. The cities on GMT+7 were Lanzhou, Linxia, Tianshui,
Pingliang, Wuwei, Zhangye, and Yumen.

My major effort has been on the Xinjiang area, which has a lot of counties,
so it is not easy to determine a map. The cities on GMT+5 were Aksu, Atushi,
Yining, Hetian, Cele, Luopu, Nileke, Zhaosu, Tekesi, Gongliu, Chabuchaer,
Huocheng, Bole, Pishan, Suiding, Yarkand. The cities on GMT+6 were Urumqi,
Turpan, Karamay, Korla, Minfeng, Jinghe, Wusu, Qiemo, Xinyan, Wulanwusu,
Jinghe, Yumin, Tacheng, Tuoli, Emin, Shihezi, Changji, Yanqi, Heshuo,
Tuokexun, Tulufan, Shanshan, Hami, Fukang, Kuitun, Kumukuli, Miquan, Qitai
and Turfan.

The cities in Xinjiang I could find lies in the upper and lower part of the
region, while there is a central zone where I could not find any cities that
coincided with astro.com, but this might also be because it is a big desert
area. Many counties have a name that corresponds to the city name also, so I
could often find it that way, but not for the following counties:

Awati 40.11 80.57
Shaya 40.53 81.43
Kuche 41.77 83.51
Xinhe 41.34 82.13
Baicheng 42.07 81.97
Wensu 41.59 80.38
Wushi 41.20 79.13
Luntai 41.70 84.65
Yuli 40.83 87.00
Hejing 42.88 85.32

The coordinates above correspond to the approximate center of the county.
I'm looking for cities near these county centers that would lie inside the
county and exist in www.astro.com to produce a GMT offset for 1970. Can
anyone help?

Unfortunately there is apparently no "show nearby cities" or map in
astro.com, so it is a bit of guesswork. The amount of different spelling for
each name is also amazing, but perhaps not if you consider "anglified name",
yughur, pinyin, traditional chinese, kazakh and uzbekh as valid variations,
probably flavored with a nice collection of typos too.

Regards,
- Jesper

Paul Eggert writes:

-----Original Message-----
From: Paul Eggert [mailto:eggert at cs.ucla.edu]
Sent: Martes, 27 de Junio de 2006 12:05
To: Jesper Norgaard Welen
Cc: tz at lecserver.nci.nih.gov
Subject: Re: Chinese timezones


For that sort of thing the Shanks & Pottenger database is the best that I
know of, though of course it's not infallible. They do attempt to cover the
wartime changes. Shanks was born during Eastern War Time and notes this
fact in the author's bio on his book.

You should be able to determine the zone boundaries yourself, if you have
good maps, by consulting the Atlas Query <http://www.astro.com/cgi/aq.cgi>,
which is an online version of the Shanks & Pottenger database. For example,
I just verified that according to that database, Harbin was at UTC+0830 on
1970-01-01 00:00. As far as I know, the only post 1970 change in China was
in May 1980, so if you use any time stamp from 1970 to 1980 you should be
able to wander around the Chinese countryside using that database and
determine where the time zone boundaries were. (Good luck with the
spellings, though!)

[Oscar van Vlijmen]

From: Jesper Norgaard Welen
The cities in Xinjiang I could find lies in the upper and lower part of the
region, while there is a central zone where I could not find any cities that
coincided with astro.com, but this might also be because it is a big desert
area. Many counties have a name that corresponds to the city name also, so I
Awati 40.11 80.57
Shaya 40.53 81.43
Kuche 41.77 83.51
Xinhe 41.34 82.13
Baicheng 42.07 81.97
Wensu 41.59 80.38
Wushi 41.20 79.13
Luntai 41.70 84.65
Yuli 40.83 87.00
Hejing 42.88 85.32
The coordinates above correspond to the approximate center of the county.
I'm looking for cities near these county centers that would lie inside the
county and exist in www.astro.com to produce a GMT offset for 1970. Can
anyone help?

As is probably known the GeoNames database lists nearly every rock and creek
with a name on it.
http://earth-info.nga.mil/gns/html/cntry_files.html
The China file is more than 20 MB large and has more than 160,000 entries.
If you're interested, I managed to extract all 961 names lying in the
rectangle with longitude 79-87 and latitude 39-44 deg.
The list contains several alternative spellings and codes with type of
location like populated place, main administrative division etcetera. The
codes are explained on pages, found at earth-info.nga.mil.
I can't say that a list with 960 names is a real help, but that's what I can
offer.

[Oscar van Vlijmen]

From: "Clive D.W. Feather"

Post by Oscar van Vlijmen
As is probably known the GeoNames database lists nearly every rock and creek
with a name on it.
http://earth-info.nga.mil/gns/html/cntry_files.html

However, it doesn't list the village (in UK terminology - the population is
over 5,000) where I live.
I can't seem to find a link on that site for sending them corrections.

At the bottom of the page I see:
Point of Contact: geonames at nga.mil

[Clive D.W. Feather]

Post by Oscar van Vlijmen
Point of Contact: geonames at nga.mil

Ah. Thanks.

[Clive D.W. Feather]

Post by Oscar van Vlijmen
Point of Contact: geonames at nga.mil

Ah. Thanks.

[Oscar van Vlijmen]

From: "Clive D.W. Feather"

Post by Oscar van Vlijmen
As is probably known the GeoNames database lists nearly every rock and creek
with a name on it.
http://earth-info.nga.mil/gns/html/cntry_files.html

However, it doesn't list the village (in UK terminology - the population is
over 5,000) where I live.
I can't seem to find a link on that site for sending them corrections.

At the bottom of the page I see:
Point of Contact: geonames at nga.mil