[tz] [PATCH] tzdata: Asia/Ha_Noi: Add new timezone

[Trần Ngọc Quân]

Signed-off-by: Tr?n Ng?c Qu?n <vnwildman at gmail.com>
---
asia | 20 ++++++++++++++++++++
zone.tab | 1 +
2 files changed, 21 insertions(+)

diff --git a/asia b/asia
index 0be896b..eb7787b 100644
--- a/asia
+++ b/asia
@@ -2769,6 +2769,26 @@ Zone Asia/Tashkent 4:37:11 - LMT 1924 May 2

# Vietnam

+# From Tr?n Ng?c Qu?n (Fri, 03 Oct 2014):
+# Hanoi is the capital of Vietnam.
+# Currently, official timezone of Vietname is Ha_Noi, not Ho_Chi_Minh.
+# References (most of them are in Vietnamese language):
+# * http://en.wikipedia.org/wiki/Time_in_Vietnam
+# * http://vi.wikipedia.org/wiki/M%C3%BAi_gi%E1%BB%9D
+# * http://www.thoigian.com.vn/?mPage=P80D01
+# * [1] http://moj.gov.vn/vbpq/Lists/Vn%20bn%20php%20lut/View_Detail.aspx?ItemID=22004
+# Zone NAME GMTOFF RULES FORMAT [UNTIL]
+Zone Asia/Ha_Noi 7:06:40 - LMT 1906 Jun 9
+ 7:06:20 - SMT 1911 Mar 11 0:01 # Saigon MT?
+ 7:00 - ICT 1942 Dec 31 23:00
+ 8:00 - ICT 1945 Mar 14 23:00
+ 9:00 - JST 1945 Sep 02 # Tokyo Standard Time
+ 7:00 - ICT 1947 Mar 31 23:00
+ 8:00 - ICT 1954 Oct 09 23:00
+ 7:00 - ICT 1967 Dec 31 23:00
+ 7:00 - ICT 2002 Oct 14 # See [1]
+ 7:00 - ICT
+
# From Paul Eggert (2013-02-21):
# Milne gives 7:16:56 for the meridian of Saigon in 1899, as being
# used in Lower Laos, Cambodia, and Annam. But this is quite a ways
diff --git a/zone.tab b/zone.tab
index 084bb2f..cf87c19 100644
--- a/zone.tab
+++ b/zone.tab
@@ -428,6 +428,7 @@ VC [phone] America/St_Vincent
VE [phone] America/Caracas
VG [phone] America/Tortola
VI [phone] America/St_Thomas
+VN +1054+10600 Asia/Ha_Noi
VN +1045+10640 Asia/Ho_Chi_Minh
VU -1740+16825 Pacific/Efate
WF -1318-17610 Pacific/Wallis

[Robert Elz]

It doesn't matter what is the capital, or what the official time zone
name happens to be - tz database zone names are derived from the name
of the city with the biggest population in an area (often country) that
uses the same time.

For Vietnam, every reference I can find puts Ho Chi Minh City as being more
populated than Ha Noi. Hence Ho_Chi_Minh is the name that is used.

It is the same reason that we use Shanghai (not Beijing), and New_York (not
Washington).

Even if Ha Noi's population should overtake Ho Chi Minh City it would
take a long time before we would change the name, now it has been
selected - eg: For India we still use Kolkata, even though Mumbai and Delhi
seem to have gone (way) past it in population count. Kolkata (when we
called it Calcutta) was chosen, and remains...

Unless you can show that Ha Noi and Ho Chi Minh City have not used the
same local time for some period (a few minutes would be enough) since 1970
an additional zone won't be created for it either. (Differences before
1970 could just possibly be used, sometime in the future, so research into
actyal time usage back into history is always useful - if backed by good
references.)

kre

[Alan Barrett]

Post by Robert Elz
Unless you can show that Ha Noi and Ho Chi Minh City have not used the
same local time for some period (a few minutes would be enough) since 1970
an additional zone won't be created for it either. (Differences before
1970 could just possibly be used, sometime in the future, so research into
actyal time usage back into history is always useful - if backed by good
references.)

The Wikipedia page,
<https://en.wikipedia.org/wiki/Time_in_Vietnam>, says that they
were different. For example, from 1968 to 1975, North Vietnam
(Hanoi) was on UTC+07:00, and from 1960 to 1975, South Vietnam
(Saigon, or Ho Chi Minh City) was on UTC+08:00.

--apb (Alan Barrett)

[Tim Parenti]

As it stands right now, the suggested new zone from Tr?n Ng?c Qu?n does not
differ from Asia/Ho_Chi_Minh after 1970 and should not be added, per the
Theory document, as Robert Elz alluded to. Whether a city is a capital or
official legal basis for a time zone should not affect our nomenclature.

The Wikipedia page, <https://en.wikipedia.org/wiki/Time_in_Vietnam>, says
that they were different.

I would note that that page only had one author and does not cite any
sources relevant to the data it presents. The second and fourth sources
provided in the proposed patch (in Vietnamese) don't seem to support its
data; however, the third reference provided,
http://www.thoigian.com.vn/?mPage=P80D01, seems somewhat promising,
although I don't think it provides nearly as much specificity as the
proposed patch.

For example, from 1968 to 1975, North Vietnam (Hanoi) was on UTC+07:00, and

from 1960 to 1975, South Vietnam (Saigon, or Ho Chi Minh City) was on
UTC+08:00.

If this is true, then Asia/Ho_Chi_Minh needs to be updated, in which case a
new zone for North Vietnam is *indeed* needed to accurately reflect the
pre-unification differences. Per Theory, this zone would probably be
called Asia/Hanoi as "Hanoi" is more common in English than "Ha Noi".

--
Tim Parenti
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mm.icann.org/pipermail/tz/attachments/20141003/ba0f4618/attachment.html>

[Paul Eggert]

Thank you for this new information. Since Hanoi and Ho Chi Minh City
disagree for time stamps as late as 1975, we need to add a Zone for Hanoi.

The real jewel here is the indirect reference to the authoritative
source"L?ch Vi?t Nam: th? k? XX?XXI (1901?2100) & Ni?n Bi?u L?ch S? Vi?t
Nam" (2014)
<http://bookaholic.vn/lich-viet-nam-the-ki-xx-xxi-1901-2100-va-bien-bieu-lich-su-viet-nam/>.
I'd like to verify the information from this source if possible. For
example, if I'm reading the website correctly the 1906 transition was on
July 1, not June 9, but I'd like to check that against the book. I have
requested a copy of the book from my library; it has a 2005 edition,
which should suffice for our needs.

This underscores the unreliability of our old source Shanks. We already
knew Shanks was wrong for large swaths of the world in the 20th
century. Even so, it's surprisingly bad for a U.S. source like Shanks
to get South Vietnam wrong in 1973. Wasn't Shanks reading the newspapers?

[Lester Caine]

Post by Paul Eggert
This underscores the unreliability of our old source Shanks. We already
knew Shanks was wrong for large swaths of the world in the 20th
century. Even so, it's surprisingly bad for a U.S. source like Shanks
to get South Vietnam wrong in 1973. Wasn't Shanks reading the newspapers?

What it does demonstrate is that now we are getting more people
involved, the better documented provenance is starting to be uncovered?

[Guy Harris]

Post by Lester Caine

Post by Paul Eggert
This underscores the unreliability of our old source Shanks. We already
knew Shanks was wrong for large swaths of the world in the 20th
century. Even so, it's surprisingly bad for a U.S. source like Shanks
to get South Vietnam wrong in 1973. Wasn't Shanks reading the newspapers?

What it does demonstrate is that now we are getting more people
involved, the better documented provenance is starting to be uncovered?

...and that, as Paul has repeatedly said, much of the current historical data in the tz database is bogus.

[Lester Caine]

Post by Guy Harris

Post by Paul Eggert
This underscores the unreliability of our old source Shanks. We already

Post by Paul Eggert
knew Shanks was wrong for large swaths of the world in the 20th
century. Even so, it's surprisingly bad for a U.S. source like Shanks
to get South Vietnam wrong in 1973. Wasn't Shanks reading the newspapers?

What it does demonstrate is that now we are getting more people
involved, the better documented provenance is starting to be uncovered?

...and that, as Paul has repeatedly said, much of the current historical data in the tz database is bogus.

I'd not disagree with that ... but hiding proven valid data away in
backzone is not the way to improve the quality. There is as much
uncorroborated data still in main database as has been ring fenced in
backzone as suspect. If we spend time researching material only for it
to be a lottery if it is included or not what is the point of bothering?

[Trần Ngọc Quân]

Post by Paul Eggert
The real jewel here is the indirect reference to the authoritative
source"L?ch Vi?t Nam: th? k? XX?XXI (1901?2100) & Ni?n Bi?u L?ch S?
Vi?t Nam" (2014)
<http://bookaholic.vn/lich-viet-nam-the-ki-xx-xxi-1901-2100-va-bien-bieu-lich-su-viet-nam/>.
I'd like to verify the information from this source if possible. For
example, if I'm reading the website correctly the 1906 transition was
on July 1, not June 9,

Yes, you right

Post by Paul Eggert
but I'd like to check that against the book. I have requested a copy
of the book from my library; it has a 2005 edition, which should
suffice for our needs.

You just compare date on this book with
<http://www.thoigian.com.vn/?mPage=P80D01>.
Unfortunately, the library near me don't have this book. "National
Library of Vietnam" has three book of 2005 edition, but I'm not in Hanoi
now. So you have to verify it by yourself. If every thing is OK, I will
change both Hanoi and Ho Chi Minh city. If you can't read it, I will try
to find people who can help us.

[Brian Inglis]

Post by Trần Ngọc Quân

Post by Paul Eggert
The real jewel here is the indirect reference to the authoritative
source"L?ch Vi?t Nam: th? k? XX?XXI (1901?2100) & Ni?n Bi?u L?ch S?
Vi?t Nam" (2014)
<http://bookaholic.vn/lich-viet-nam-the-ki-xx-xxi-1901-2100-va-bien-bieu-lich-su-viet-nam/>.
I'd like to verify the information from this source if possible. For
example, if I'm reading the website correctly the 1906 transition was
on July 1, not June 9,

Yes, you right

Post by Paul Eggert
but I'd like to check that against the book. I have requested a copy
of the book from my library; it has a 2005 edition, which should
suffice for our needs.

You just compare date on this book with
<http://www.thoigian.com.vn/?mPage=P80D01>.
Unfortunately, the library near me don't have this book. "National
Library of Vietnam" has three book of 2005 edition, but I'm not in Hanoi
now. So you have to verify it by yourself. If every thing is OK, I will
change both Hanoi and Ho Chi Minh city. If you can't read it, I will try
to find people who can help us.

Rough facts in the linked passage via Google translate cleaned up (especially random month/day switches):

On 01.07.1906
When completed construction Observatory Phu Lien, Indochina government decree dated 09.06.1906 (Official Gazette dated 18.06.1906 Indochina) fixed now legal for all countries of Indochina under the meridian passing through Phu Lien (104?17'17" east of Paris) from 0 hours on 01.07.1906.

On 01.05.1911
After France signed the Treaty of international time zones, according to a decree dated 06.04.1911 (Indo Gazette on 13.4.1911-page 803) taken under the provisions of the new time zone 7 hours (from meridian line passing through Greenwich) for all countries of Indochina starting at 0 hours on 1.5.1911.

On 01.01.1943
The French government issued a decree dated 23.12.1942 (Official Gazette dated 30.12.1942 Indochina) link Indochina in 8 time zone and the clock is so advanced by 60 minutes at 23 hours on 31.12.1942.

On 14.03.1945
On 09.03.1945 Japanese invaded and forced French Indochina to the time zone of Tokyo (Japan) ie ninth time zone so time officially was advanced quickly on again 1 hour at 23 hours on 14.03.1945 .

On 02.09.1945
After the August Revolution Provisional Government of Vietnam Democratic Republic claims made ??7 hour time zone official (Ordinance No. year / SL of the Ministry of Internal Affairs).

On 01.04.1947
According to the decree dated 28.3.1947 of the colonial government (Official Gazette dated 14.10.1947 Indochina), then in the occupied areas in Vietnam, Laos and Cambodia is now officially eighth time zone from date 1.4.1947. But in the liberated areas remains seventh time zone and after the Geneva Agreement for liberated areas in the north also time zone 7 (Hanoi and Hai Phong from the end of months 10.1954 5.1955); Laos separately return to seventh time zone on 15.4.1955.

On 01.07.1955
South Vietnam back to seventh time zone from 0 hours on 01.07.1955.

On 01.01.1960
Saigon government decree dated 30.12.1959 No. 362-TTP rules now official time zone South Vietnam is eighth, promptly advancing the clock by 1 hour from 23hours night of 31.12.1959 (ie, 0 hours on 1.1.1960)

On 31.12.1967
The Government of the Democratic Republic of Vietnam in Decision 121 / CP dated 8.8.1967 confirmed the country's official time is seventh time zone from 0 hours on 1.1.1968.

On 13.06.1975
After the South was completely liberated, the provisional revolutionary government issued a formal decision and went back to the seventh hour time zone in Saigon returning again to 1 hour later.

(According Calendar Vietnam XX-XXI century, author Master Tran Tien Binh, state board schedule)

[Paul Eggert]

Post by Brian Inglis
On 01.07.1906
When completed construction Observatory Phu Lien, Indochina government decree
dated 09.06.1906 (Official Gazette dated 18.06.1906 Indochina) fixed now legal
for all countries of Indochina under the meridian passing through Phu Lien
(104?17'17" east of Paris) from 0 hours on 01.07.1906.

Yes, thanks, I'm working my way through that web page. For that paragraph, I
assume "east of Paris" means east of Paris Mean Time, which was legally 00:09:21
east of Greenwich, and since 104?17?17? works out to 06:57:09.1333..., that
would make Ph? Li?n Mean Time legally 07:06:30.1333.... An alternate
interpretation would be that it refers to Fran?ois Arago's definition of the
Paris Meridian; in that case Ph? Li?n Mean Time would be the temporal equivalent
of 2?20?14.03? + 104?17?17? east, which would be 07:06:29.33333.... Luckily
both values round to 07:06:30, so we can use this rounded value under either
interpretation.

By the way, if Google Maps and <http://kml.inovmapping.com/Ph-Lin-Observatory>
are to believed the observatory is a bit east of both values, at roughly
07:06:31; but perhaps they moved the observatory since 1906, and anyway the
legal definition is close enough.

I assume the abbreviation should be PLMT for Ph? Li?n Mean Time.

[Tim Parenti]

Attached are two patches which are a WIP "starting point" for this
change. Obviously, we need to check the rest of the dates more
thoroughly and include translated commentary where necessary, and so
they will certainly still need some further iteration.

There are still a few things that bother me about this. Specifically,
ICT referring to different UT offsets for several periods. Perhaps,
prior to reunification, NICT and SICT should be used for "North/South
Indochina Time"?

In any case, I think this WIP patchset gets us closer to where we'll
ultimately be headed with this change.

These patches are on my 2014-10-vietnam-wip-1 branch on Github and can
be reviewed in the browser at
https://github.com/timparenti/tz-experimental/compare/eggert:master...timparenti:2014-10-vietnam-wip-1

--
Tim Parenti

Post by Paul Eggert

Post by Brian Inglis
On 01.07.1906
When completed construction Observatory Phu Lien, Indochina
government decree
dated 09.06.1906 (Official Gazette dated 18.06.1906 Indochina) fixed now legal
for all countries of Indochina under the meridian passing through Phu Lien
(104?17'17" east of Paris) from 0 hours on 01.07.1906.

Yes, thanks, I'm working my way through that web page. For that
paragraph, I assume "east of Paris" means east of Paris Mean Time,
which was legally 00:09:21 east of Greenwich, and since 104?17?17?
works out to 06:57:09.1333..., that would make Ph? Li?n Mean Time
legally 07:06:30.1333.... An alternate interpretation would be that
it refers to Fran?ois Arago's definition of the Paris Meridian; in
that case Ph? Li?n Mean Time would be the temporal equivalent of
2?20?14.03? + 104?17?17? east, which would be 07:06:29.33333....
Luckily both values round to 07:06:30, so we can use this rounded
value under either interpretation.
By the way, if Google Maps and
<http://kml.inovmapping.com/Ph-Lin-Observatory> are to believed the
observatory is a bit east of both values, at roughly 07:06:31; but
perhaps they moved the observatory since 1906, and anyway the legal
definition is close enough.
I assume the abbreviation should be PLMT for Ph? Li?n Mean Time.

-------------- next part --------------

[Paul Eggert]

... ICT referring to different UT offsets for several periods. Perhaps, prior to
reunification, NICT and SICT should be used for "North/South Indochina Time"?

It wasn't North vs South. The new data say that French Indochina mostly used
UT+7, but sometimes all of French Indochina used UT+8, sometimes just South
Vietnam, sometimes more of a hodgepodge. I had independently run into this
problem and thought of using the abbreviation "IDT" (short for InDochina Time)
for UT+8 in Indochina to help clear this up.

There's a bigger problem, though. The new data also tell us that our entries
for Laos and Cambodia are almost entirely bogus. We already knew that, and I
had slated them for 'backzone' anyway, so now's a good time to do that. As a
corollary, although we clearly need to fix Asia/Ho_Chi_Minh, we don't need a
separate Zone for Asia/Hanoi as it is identical to Asia/Bangkok since 1970.
This will give us a simpler fix and an easier-to-maintain result.

A proposed pair of patches is attached. The first moves the bogus data for
Cambodia and Laos to 'backzone'. The second fixes Asia/Ho_Chi_Minh and adds a
line to zone1970.tab so that users in Vietnam now have two choices, one for each
half of the country.

-------------- next part --------------

[Trần Ngọc Quân]

Post by Paul Eggert

... ICT referring to different UT offsets for several periods.
Perhaps, prior to
reunification, NICT and SICT should be used for "North/South
Indochina Time"?

It wasn't North vs South. The new data say that French Indochina
mostly used UT+7, but sometimes all of French Indochina used UT+8,
sometimes just South Vietnam, sometimes more of a hodgepodge. I had
independently run into this problem and thought of using the
abbreviation "IDT" (short for InDochina Time) for UT+8 in Indochina to
help clear this up.

I agree with you.

Post by Paul Eggert
There's a bigger problem, though. The new data also tell us that our
entries for Laos and Cambodia are almost entirely bogus. We already
knew that, and I had slated them for 'backzone' anyway, so now's a
good time to do that. As a corollary, although we clearly need to fix
Asia/Ho_Chi_Minh, we don't need a separate Zone for Asia/Hanoi as it
is identical to Asia/Bangkok since 1970. This will give us a simpler
fix and an easier-to-maintain result.
A proposed pair of patches is attached. The first moves the bogus
data for Cambodia and Laos to 'backzone'. The second fixes
Asia/Ho_Chi_Minh and adds a line to zone1970.tab so that users in
Vietnam now have two choices, one for each half of the country.

About second patch:

+Zone Asia/Ho_Chi_Minh 7:06:40 - LMT 1906 Jul 1
+ 7:06:30 - PLMT 1911 May 1
+ 7:00 - ICT 1942 Dec 31 23:00
+ 8:00 - IDT 1945 Mar 14 23:00
+ 9:00 - JST 1945 Sep 2
+ 7:00 - ICT 1947 Apr 1
+ 8:00 - IDT 1955 Jul 1
+ 7:00 - ICT 1959 Dec 31 23:00
+ 8:00 - IDT 1975 Jun 13
7:00 - ICT

This is Hanoi, not Ho Chi Minh City.

I find an other book about this topic:

Name: L?ch ?m d??ng Vi?t Nam 1900-2010
Author: Nguy?n V?n Chung
Publisher: Nh? xu?t b?n v?n h?a d?n t?c
Edition: 2001

- It confirms the date `1906 Jul 1'. So I think this date is correct.
- It said Ph? Li?n is 104?17?17? East. But I think `east of Paris' is
correct. In Google map, it is: 20?47'05.1"N+106?38'22.0"E
Unfortunately, It don't show any other information that we need.

Some photos about Ph? Li?n:

* <http://www.panoramio.com/photo/22547402>
* <http://www.panoramio.com/photo/57154621>

[Paul Eggert]

Yes, there are a lot of books on Vietnamese dates and times! All in Vietnamese,
unfortunately for me. The book by Tr?n Ng?c Qu?n appears to be one of the most
authoritative. Dershowitz and Reingold cites its 2005 edition in their book
Calendrical Calculations (3rd ed.)
<http://emr.cs.iit.edu/home/reingold/calendar-book/third-edition/>. They also
cite the 2000 edition of a calendrical book by L T L?n and personal
communications from H? Ng?c ??c -- none of which agree exactly, unfortunately.

Post by Trần Ngọc Quân
This is Hanoi, not Ho Chi Minh City.

We need to correct the Asia/Ho_Chi_Minh entry regardless of what we do about
Hanoi, and that is what the patch does. As I pointed out in my most recent
email, it appears to be unnecessary to add an entry for Hanoi, since Hanoi does
not have a unique time zone history since 1970. In contrast, Ho Chi Minh City
(now that its data are corrected) does have a unique time zone history since
1970 and therefore needs a distinct Zone.

I had planned to replace the Ho Chi Minh City zone with a link on the grounds
that it was redundant, but this new information means we need to keep it.

[Trần Ngọc Quân]

Post by Paul Eggert
Yes, there are a lot of books on Vietnamese dates and times! All in
Vietnamese, unfortunately for me. The book by Tr?n Ng?c Qu?n appears
to be one of the most authoritative. Dershowitz and Reingold cites
its 2005 edition in their book Calendrical Calculations (3rd ed.)
<http://emr.cs.iit.edu/home/reingold/calendar-book/third-edition/>.
They also cite the 2000 edition of a calendrical book by L T L?n

Dr. L? th?nh L?n
I find his email, but it I'm not sure he still use it. But I will try to
contact.
lethanhlan at yahoo.de <mailto:lethanhlan at yahoo.de>

Post by Paul Eggert
and personal communications from H? Ng?c ??c -- none of which agree
exactly, unfortunately.

I will try to contact H? Ng?c ??c.

Post by Paul Eggert

Post by Trần Ngọc Quân
This is Hanoi, not Ho Chi Minh City.

We need to correct the Asia/Ho_Chi_Minh entry regardless of what we do
about Hanoi, and that is what the patch does. As I pointed out in my
most recent email, it appears to be unnecessary to add an entry for
Hanoi, since Hanoi does not have a unique time zone history since
1970. In contrast, Ho Chi Minh City (now that its data are corrected)
does have a unique time zone history since 1970 and therefore needs a
distinct Zone.
I had planned to replace the Ho Chi Minh City zone with a link on the
grounds that it was redundant, but this new information means we need
to keep it.

I find other document about time zone in Vietnam (an archive on
website). Note: date format is dd/mm/yyyy. It is translated by google
with some fix:

Previously,withthenationalstageorapart ofthe countryusingdifferent time
zones:

- From01/01/1943:Vietnamaccording toGMT +8(1hourearlierthannormal).
-From01/04/1945:VietnamGMT+9,2 hoursearlierthanstandard time.
-From01/04/1947,theFrench colonialistsoccupiedGMT+8.
-In the South,from01/07/1955accordingto GMT+7.To01/01/1960,thetime
zoneofsouthernGMT+8.
-In the north,according toDecision No.121/CPsigned on08/08/1967byCouncil
ofGovernment,from08/08/1967accordingto GMT+7.
-From02/08/1976,the countrystandardtime zoneGMT+7.

We seem have to find more document for both Vietnam and Lao, Cambodia. I
hope I reply soon.

[Trần Ngọc Quân]

Hello Paul Eggert,

Post by Trần Ngọc Quân
Dr. L? th?nh L?n
I find his email, but it I'm not sure he still use it. But I will try
to contact.
lethanhlan at yahoo.de <mailto:lethanhlan at yahoo.de> I will try to contact
H? Ng?c ??c.

Unfortunately, I can't contact both of them.

Post by Trần Ngọc Quân

Post by Paul Eggert
We need to correct the Asia/Ho_Chi_Minh entry regardless of what we
do about Hanoi, and that is what the patch does.

Of cause, Asia/Ho_Chi_Minh properly need to fix.

Post by Trần Ngọc Quân

Post by Paul Eggert
As I pointed out in my most recent email, it appears to be
unnecessary to add an entry for Hanoi, since Hanoi does not have a
unique time zone history since 1970. In contrast, Ho Chi Minh City
(now that its data are corrected) does have a unique time zone
history since 1970 and therefore needs a distinct Zone.

Wiki [1] point out Hanoi and Ho Chi Minh timezone diff from 1970 - 1975:
Hanoi used GMT +7 while Ho Chi Minh used GMT +8. So I sure timezone in
Hanoi is unique. You decide add or not by yourself.
I read carefully both [1] and [2]. They are the same content. I think
they are the best document. So you can use [1].

Paul Eggert, please make new patch. I will review.
About Cambodia and Lao, I recommend remove them. They can use
Asia/Bangkok instead. Unique timezone for them will create when we have
enough document.

[1] http://en.wikipedia.org/wiki/Time_in_Vietnam
[2] http://www.thoigian.com.vn/?mPage=P80D01

[Paul Eggert]

Post by Trần Ngọc Quân
About Cambodia and Lao, I recommend remove them. They can use
Asia/Bangkok instead.

As can north Vietnam, which was the point of the 2nd patch in:

http://mm.icann.org/pipermail/tz/2014-October/021669.html

See its two lines with the string "VN". With the patch installed, you can run
the command "tzselect" to see these two lines in action.

[Tim Parenti]

Even though we believe that north Vietnam can be handled by Asia/Bangkok,
given this new source of reliable historical data, shouldn't we put that
somewhere, e.g. in backzone? We should keep these data around in case we
discover changes to one of the other merged zones that means the
assumptions under which they were merged is no longer accurate.

I realize that including this as an actual Zone in backzone would chip away
a bit at the firmness of the 1970 cutoff, and there has already been much
debate over whether or not that would be a Good Thing, which I do not wish
to rehash. But I feel that having a second-class area like backzone makes
loosening this cutoff a bit more necessary in cases like this, so that good
data like these can be preserved in case we need them again, whether within
our current scope or the scope this project takes on in the future.

So maybe we should keep this "extra" information about north Vietnam in
comments?

--
Tim Parenti

Post by Trần Ngọc Quân
About Cambodia and Lao, I recommend remove them. They can use

Post by Trần Ngọc Quân
Asia/Bangkok instead.

http://mm.icann.org/pipermail/tz/2014-October/021669.html
See its two lines with the string "VN". With the patch installed, you can
run the command "tzselect" to see these two lines in action.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mm.icann.org/pipermail/tz/attachments/20141011/64fec936/attachment.html>

[Lester Caine]

Copied to tzdist list as I feel this starts to fill in the gaps!

Post by Tim Parenti
I realize that including this as an actual Zone in backzone would chip
away a bit at the firmness of the 1970 cutoff, and there has already
been much debate over whether or not that would be a Good Thing, which I
do not wish to rehash. But I feel that having a second-class area like
backzone makes loosening this cutoff a bit more necessary in cases like
this, so that good data like these can be preserved in case we need them
again, whether within our current scope or the scope this project takes
on in the future.

That the 'second class' area has 'first class' material simply because
the date of change pre-dates 1970 is the whole problem. There is
substantial pre-1970 data in the 'first class' area but no indication if
that actually applies to the linked aliases! The discussion on the
tzdist list is about 'truncation' and I've only just realised that it is
just for every alias that the TZ data is only valid from 1970. Of cause
the 'primary' set of data is perfectly valid back to it's start date.
Personally I feel where a current alias has got first class data back to
an earlier time then it should always be included in the base set. Since
in most cases it IS just an extra start rule prior to another generic
data set then there is little extra data. Where it becomes a problem is
when the data set for an ID is always fully expanded such as on the
tzdist current protocol.

'Secondary' ID's which just have a small data set prior to a more
generic one post some later date are the problem. In many cases it is
the start of 'DST' which forms a smaller set of 'variable' data and
around 2/3rds of ID's never hit that point. I was originally looking for
a flag between LMT and the start of a timezone, but the real point here
is using the 'common' event time as the end marker of 'aliases' so that
the handling of say Europe/UK/Oxford would have fine detail prior to
Europe/UK and in this case Europe/UK/London is the 'master' for
Europe/UK. Europe/Isle_of_Man/Castle_Rushen then folds down to
Europe/Isle_of_Man which picks up Europe/UK at the start of DST changes
rather than the start of London/GMT or 1970, both of which are wrong. It
*IS* the use of an arbitrary location to identify a set of timezone data
which is the whole problem of making this scale well back in time? A
'DST' data set SHOULD be a region rather than a location and a location
simply has a date on which they started using that data set.

In tzdist the conversion of an alias to a set of data actually needs to
be a two stage process of which only the second stage is currently being
developed. The TZID for Europe/London provides a set of timezone data
which is accurate from 1916 and while not fully verified, all other UK
locations adopted that and used that from 1916. (There are some gaps in
later provenance on DST changes). Prior to 1916 there are a number of
other common points such as the adoption on GMT in different areas
culminating in the eventual standardisation by around 1884. Some
locations like counties in Idaho are a list of changes between two of
the 5 American main data sets. Most locations are a lot easier, either
only having a single date for the start of the international GMT
adoption, or perhaps two or three dates with a local pre GMT standard
followed by the GMT change all well before 1970. 'GMT' is essentially
just a fixed set of 24 offsets which is all that the large majority of
locations reference and which is fine tuned by UTC in 1972, so stage one
is to identify the history for a location, which can be simple if all
one is interested in is post 1970 or so. This will either just give a
fixed offset, or identify a dataset to use. Where I don't think the
current tzdist model scales is where a large number of historic ID's
only vary by a single start date. The system does not need to duplicate
all the following data.

[Paul Eggert]

Post by Tim Parenti
given this new source of reliable historical data, shouldn't we put that
somewhere, e.g. in backzone?

Sure, why not? Proposed patch attached.
-------------- next part --------------

[Trần Ngọc Quân]

Post by Paul Eggert

Post by Tim Parenti
given this new source of reliable historical data, shouldn't we put that
somewhere, e.g. in backzone?

Sure, why not? Proposed patch attached.

There is an incorrect date here. Please see below:

$ git diff
diff --git a/backzone b/backzone
index 70a526f..44f010c 100644
--- a/backzone
+++ b/backzone
@@ -333,7 +333,7 @@ Link Asia/Chongqing Asia/Chungking
# Vietnam
# From Paul Eggert (2014-10-11):
# See Asia/Ho_Chi_Minh for the source for this data.
-# Tr?n's book is quoting as saying the 1954-55 transition to 07:00
+# Tr?n's book is quoting as saying the 1954-1955 transition to 07:00
# was at the end of October 1954 in Hanoi.
Zone Asia/Hanoi 7:03:24 - LMT 1906 Jul 1
7:06:30 - PLMT 1911 May 1
@@ -341,7 +341,7 @@ Zone Asia/Hanoi 7:03:24 - LMT
1906 Jul 1
8:00 - IDT 1945 Mar 14 23:00
9:00 - JST 1945 Sep 2
7:00 - ICT 1947 Apr 1
- 8:00 - IDT 1954 Nov 1
+ 8:00 - IDT 1954 Nov 10
7:00 - ICT

# China

[Paul Eggert]

Post by Trần Ngọc Quân
- 8:00 - IDT 1954 Nov 1
+ 8:00 - IDT 1954 Nov 10

Thanks, but where did the "10" come from? Is there a source that we can cite?
The cited web page <http://www.thoigian.com.vn/?mPage=P80D01> says this:

"Ng?y 1/4/1947. Theo ngh? ??nh ng?y 28/3/1947 c?a ch?nh quy?n th?c d?n (C?ng b?o
??ng D??ng ng?y 14/10/1947) th? trong c?c v?ng b? t?m chi?m ? Vi?t Nam, ? L?o v?
Campuchia gi? ch?nh th?c l? m?i gi? 8 k? t? ng?y 1/4/1947. Tuy nhi?n trong c?c
v?ng gi?i ph?ng v?n gi? m?i gi? 7 v? sau Hi?p ??nh gi?nev? c?c v?ng gi?i ph?ng ?
mi?n b?c c?ng theo m?i gi? 7 (H? n?i t? 10/1954 v? H?i ph?ng cu?i th?ng 5/1955);
ri?ng L?o tr? l?i m?i gi? 7 v?o ng?y 15/4/1955."

I don't read Vietnamese, but I've read Brian Inglis's translation
<http://mm.icann.org/pipermail/tz/2014-October/021654.html>, and if I understand
it correctly the parenthesized text can be translated as "Hanoi from the end of
October 1954 and Haiphong from the end of May 1955", which would mean a
transition from UT+8 to UT+7 at 1954-10-31 24:00 or (equivalently) at 1954-11-01
00:00. Perhaps the translation is incorrect, or perhaps I'm reading it
incorrectly, but I'd like to know the details.

[Brian Inglis]

Post by Trần Ngọc Quân
- 8:00 - IDT 1954 Nov 1
+ 8:00 - IDT 1954 Nov 10

"Ng?y 1/4/1947. Theo ngh? ??nh ng?y 28/3/1947 c?a ch?nh quy?n th?c d?n (C?ng b?o ??ng D??ng ng?y 14/10/1947) th? trong c?c v?ng b? t?m chi?m ? Vi?t Nam, ? L?o v? Campuchia gi? ch?nh th?c l? m?i gi? 8 k? t? ng?y 1/4/1947. Tuy nhi?n trong c?c v?ng gi?i ph?ng v?n gi? m?i gi? 7 v? sau Hi?p ??nh gi?nev? c?c v?ng gi?i ph?ng ? mi?n b?c c?ng theo m?i gi? 7 (H? n?i t? 10/1954 v? H?i ph?ng cu?i th?ng 5/1955); ri?ng L?o tr? l?i m?i gi? 7 v?o ng?y 15/4/1955."
I don't read Vietnamese, but I've read Brian Inglis's translation <http://mm.icann.org/pipermail/tz/2014-October/021654.html>, and if I understand it correctly the parenthesized text can be translated as "Hanoi from the end of October 1954 and Haiphong from the end of May 1955", which would mean a transition from UT+8 to UT+7 at 1954-10-31 24:00 or (equivalently) at 1954-11-01 00:00. Perhaps the translation is incorrect, or perhaps I'm reading it incorrectly, but I'd like to know the details.

Perhaps confusion of the OP by the tzdata use of the English month 11 abbreviation Nov for November,
in representing "the end of the month of 10/1954" i.e 31.10.1954 24.00 as the equivalent 01.11.1954 00.00.

[Trần Ngọc Quân]

Post by Paul Eggert

Post by Trần Ngọc Quân
- 8:00 - IDT 1954 Nov 1
+ 8:00 - IDT 1954 Nov 10

Thanks, but where did the "10" come from?

"1954 Nov 10" is the day that "Government of the Democratic Republic of
Vietnam" returned Hanoi.

Post by Paul Eggert
Is there a source that we can cite? The cited web page
"Ng?y 1/4/1947. Theo ngh? ??nh ng?y 28/3/1947 c?a ch?nh quy?n th?c d?n
(C?ng b?o ??ng D??ng ng?y 14/10/1947) th? trong c?c v?ng b? t?m chi?m
? Vi?t Nam, ? L?o v? Campuchia gi? ch?nh th?c l? m?i gi? 8 k? t? ng?y
1/4/1947. Tuy nhi?n trong c?c v?ng gi?i ph?ng v?n gi? m?i gi? 7 v? sau
Hi?p ??nh gi?nev? c?c v?ng gi?i ph?ng ? mi?n b?c c?ng theo m?i gi? 7
(H? n?i t? 10/1954 v? H?i ph?ng cu?i th?ng 5/1955); ri?ng L?o tr? l?i
m?i gi? 7 v?o ng?y 15/4/1955."
"Hanoi from the end of October 1954 and Haiphong from the end of May
1955",

Yes, It don't specify day.
Must translate "since" or "from begin of" not "from the end of".
I've contacted Dr. L?n yesterday. And I will confirm all information
with him.
Thanks,

[Brian Inglis]

Post by Trần Ngọc Quân

Post by Paul Eggert

Post by Trần Ngọc Quân
- 8:00 - IDT 1954 Nov 1
+ 8:00 - IDT 1954 Nov 10

Thanks, but where did the "10" come from?

"1954 Nov 10" is the day that "Government of the Democratic Republic of
Vietnam" returned Hanoi.

Post by Paul Eggert
Is there a source that we can cite? The cited web page
"Ng?y 1/4/1947. Theo ngh? ??nh ng?y 28/3/1947 c?a ch?nh quy?n th?c d?n
(C?ng b?o ??ng D??ng ng?y 14/10/1947) th? trong c?c v?ng b? t?m chi?m
? Vi?t Nam, ? L?o v? Campuchia gi? ch?nh th?c l? m?i gi? 8 k? t? ng?y
1/4/1947. Tuy nhi?n trong c?c v?ng gi?i ph?ng v?n gi? m?i gi? 7 v? sau
Hi?p ??nh gi?nev? c?c v?ng gi?i ph?ng ? mi?n b?c c?ng theo m?i gi? 7
(H? n?i t? 10/1954 v? H?i ph?ng cu?i th?ng 5/1955); ri?ng L?o tr? l?i
m?i gi? 7 v?o ng?y 15/4/1955."
"Hanoi from the end of October 1954 and Haiphong from the end of May
1955",

Yes, It don't specify day.
Must translate "since" or "from begin of" not "from the end of".
I've contacted Dr. L?n yesterday. And I will confirm all information
with him.

The translation would need to be:
(Hanoi after 10/1954 and Hai Phong from the end of the month 5/1955.)
but would that not require "sau" (after), rather than "t?" (from or "since"),
or else "t? ??u" ("from the beginning of" or 'from the start of'),
which would exclude 10.11.1954 (10 November 1954).

[Trần Ngọc Quân]

Post by Brian Inglis

Post by Trần Ngọc Quân

Post by Paul Eggert

Post by Trần Ngọc Quân
- 8:00 - IDT 1954 Nov 1
+ 8:00 - IDT 1954 Nov 10

Thanks, but where did the "10" come from?

"1954 Nov 10" is the day that "Government of the Democratic Republic of
Vietnam" returned Hanoi.

Post by Paul Eggert
Is there a source that we can cite? The cited web page
"Ng?y 1/4/1947. Theo ngh? ??nh ng?y 28/3/1947 c?a ch?nh quy?n th?c d?n
(C?ng b?o ??ng D??ng ng?y 14/10/1947) th? trong c?c v?ng b? t?m chi?m
? Vi?t Nam, ? L?o v? Campuchia gi? ch?nh th?c l? m?i gi? 8 k? t? ng?y
1/4/1947. Tuy nhi?n trong c?c v?ng gi?i ph?ng v?n gi? m?i gi? 7 v? sau
Hi?p ??nh gi?nev? c?c v?ng gi?i ph?ng ? mi?n b?c c?ng theo m?i gi? 7
(H? n?i t? 10/1954 v? H?i ph?ng cu?i th?ng 5/1955); ri?ng L?o tr? l?i
m?i gi? 7 v?o ng?y 15/4/1955."
"Hanoi from the end of October 1954 and Haiphong from the end of May
1955",

Yes, It don't specify day.
Must translate "since" or "from begin of" not "from the end of".
I've contacted Dr. L?n yesterday. And I will confirm all information
with him.

(Hanoi after 10/1954 and Hai Phong from the end of the month 5/1955.)
but would that not require "sau" (after), rather than "t?" (from or "since"),
or else "t? ??u" ("from the beginning of" or 'from the start of'),
which would exclude 10.11.1954 (10 November 1954).

Not "after", You translate Hai Phong ok

Post by Brian Inglis
From origin document, It mean "Hanoi apply GMT +7 since October 1954"

(Maybe 1st of October 1954, or one of day of this month).
About 10 or 11, I will confirm later. And also need to verify 2nd or 3th
of September 1945.
I have a question:
# Zone NAME GMTOFF RULES FORMAT [UNTIL]
So if time not specify, like:
8:00 - IDT 1954 Nov 10
So IDT timezone end in 23:59 1954 Nov 10 or end of the previous day (9)?
I read zic(1), and think it mean 1954 Nov 10 00:01? Is it true?

[Brian Inglis]

Post by Trần Ngọc Quân

Post by Brian Inglis

Post by Trần Ngọc Quân

Post by Paul Eggert

Post by Trần Ngọc Quân
- 8:00 - IDT 1954 Nov 1
+ 8:00 - IDT 1954 Nov 10

Thanks, but where did the "10" come from?

"1954 Nov 10" is the day that "Government of the Democratic Republic of
Vietnam" returned Hanoi.

Post by Paul Eggert
Is there a source that we can cite? The cited web page
"Ng?y 1/4/1947. Theo ngh? ??nh ng?y 28/3/1947 c?a ch?nh quy?n th?c d?n
(C?ng b?o ??ng D??ng ng?y 14/10/1947) th? trong c?c v?ng b? t?m chi?m
? Vi?t Nam, ? L?o v? Campuchia gi? ch?nh th?c l? m?i gi? 8 k? t? ng?y
1/4/1947. Tuy nhi?n trong c?c v?ng gi?i ph?ng v?n gi? m?i gi? 7 v? sau
Hi?p ??nh gi?nev? c?c v?ng gi?i ph?ng ? mi?n b?c c?ng theo m?i gi? 7
(H? n?i t? 10/1954 v? H?i ph?ng cu?i th?ng 5/1955); ri?ng L?o tr? l?i
m?i gi? 7 v?o ng?y 15/4/1955."
"Hanoi from the end of October 1954 and Haiphong from the end of May
1955",

Yes, It don't specify day.
Must translate "since" or "from begin of" not "from the end of".
I've contacted Dr. L?n yesterday. And I will confirm all information
with him.

(Hanoi after 10/1954 and Hai Phong from the end of the month 5/1955.)
but would that not require "sau" (after), rather than "t?" (from or "since"),
or else "t? ??u" ("from the beginning of" or 'from the start of'),
which would exclude 10.11.1954 (10 November 1954).

Not "after", You translate Hai Phong ok

Post by Brian Inglis
From origin document, It mean "Hanoi apply GMT +7 since October 1954"

(Maybe 1st of October 1954, or one of day of this month).
About 10 or 11, I will confirm later. And also need to verify 2nd or 3th
of September 1945.
# Zone NAME GMTOFF RULES FORMAT [UNTIL]
8:00 - IDT 1954 Nov 10
So IDT timezone end in 23:59 1954 Nov 10 or end of the previous day (9)?
I read zic(1), and think it mean 1954 Nov 10 00:01? Is it true?

If it was "one day of this month" then that would imply some date
in October 1954, and exclude 10th November 1954.

In original rule 1954-10-31 24:00:00 local time == 1954-11-01 00:00:00 local time;
in revised rule 1954-11-09 24:00:00 local time == 1954-11-10 00:00:00 local time.

[Tim Parenti]

On 13 October 2014 10:56, Brian Inglis <Brian.Inglis at systematicsw.ab.ca>

Post by Brian Inglis
If it was "one day of this month" then that would imply some date
in October 1954, and exclude 10th November 1954.

The way I'm reading the translations that have been going around, it seems
Hanoi had UT+7 since *some time* in October 1954, so the only thing we know
somewhat definitively is that it at least happened before the start of 1
November 1954. I think the current rule reflects that, but lacking other
sources, perhaps the uncertainty should be noted.

--
Tim Parenti
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mm.icann.org/pipermail/tz/attachments/20141013/a8bcf271/attachment.html>

[Paul Eggert]

Post by Tim Parenti
perhaps the uncertainty should be noted.

Proposed patch attached. I'm hoping to put this to bed soon, as we're spending
a lot of time on out-of-scope data.
-------------- next part --------------

[Paul Eggert]

Post by Trần Ngọc Quân
8:00 - IDT 1954 Nov 10
So IDT timezone end in 23:59 1954 Nov 10 or end of the previous day (9)?
I read zic(1), and think it mean 1954 Nov 10 00:01?

Omitted times default to 00:00:00. The style is to specify the time as 00:00 if
the time is known to be 00:00, and to omit the time if it's not known.