Also, the Liberation day of Hong Kong after WWII which British rule over the territory resumed was August 30, 1945, which I think should be the termination date for the use of JST in the territory, instead if the day currently stated in the tz database. 在 2018年10月26日週五 23:16，Phake Nick <c933103@gmail.com> 寫道：

Additional online electronic records I found: 1941 June 14, Hong Kong Daily Press, Daylight Saving from 3am Tomorrow: https://i.imgur.com/05KkvtC.png 1941 September 30, Hong Kong Daily Press, Winter Time Warning: https://i.imgur.com/dge4kFJ.png

Phake Nick <c933103@gmail.com> 於 2018年10月26日週五 下午10:34寫道：

...

Today I went to check microfilm records stored at Hong Kong Public Library and found following situation that are inconsistent with tz database record. * Around April 1 1941, there're no news about DST on various newspapers that I checked. Time given for sunrise/sunset around the period are also continuous (as in they're minutes within each others on those days, indicating there are no change to time standard around 1941 April 1. * I have subsequently checked Hong Kong Observatory's site and the site now list the 1941 DST start date for Hong Kong as June 15. (url is same as the one in tz file but content is different from when the 2009 comment was made) I have not tried validating the info from those newspapers as I wasn't aware of the table's current content until I stopped using the library. * And then, on September 30 1941, according to Ta Kung Pao (Hong Kong edition), it was stated that fallback would occur on the next day (the 1st)'s "03:00 am (Hong Kong Time 04:00 am)" and the clock will fall back for a half hour. (03:00 probably refer to the time commonly used in mainland China at the time given the paper's background), after it was spring forward for an hour months ago. I cannot find out record of English newspaper from that day, but the sunrise/sunset time given by South China Morning Post for October 1st was indeed moved by half an hour compares to before. * After that, in December, the battle to capture Hong Kong started and the library doesn't seems to have any record stored about press during that period of time. Some media resumed publication soon after that within the same month, but there were not much information about time there. Later they started including a radio program guide when they restored radio service, explicitly mentioning it use Tokyo standard time, and later added a note saying it's half an hour ahead of the old Hong Kong standard time, and it also seems to indicate that Hong Kong was not using GMT+8 when it was captured by Japan. * Image of related sections on newspaper: 1941 September 30, Ta Kung Pao (Hing Kong) , "Winter Time starr tomorrow": https://i.imgur.com/6waY51Z.jpg (Chinese) 1941 September 29, South China Morning Post, Information on sunrise/sunset time and other things for September 30 and October 1: https://i.imgur.com/kCiUR78.jpg 1942 February 5, The Hong Kong News, Radio Program Guide: https://i.imgur.com/eVvDMzS.jpg

Phake Nick wrote:

According to Singaporean newspaper http://eresources.nlb.gov.sg/newspapers/Digitised/Article/singfreepresswk190... , the day that Hong Kong start using GMT+8 should be Oct 30, 1904.

Thanks for doing all that research on Hong Kong. I also see that sometime between 2009 and 2014 the Hong Kong Observatory changed the 1952 DST fallback date from "25 Oct" (which agrees with Shanks) to "2 Nov"; we might as well stay in sync with the HKO unless we can find newspapers to the contrary. Also, a British astronomer H.P.H. (I don't know who that is) reported in the February 1905 Monthly Notices of the Royal Astronomical Society that the 1904-10-30 time ball drop in Hong Kong occurred at 17:00 GMT (presumably this was GMT the previous day). I attempted to distill the above into tzdata and installed the attached proposed patch to the development version. Please let me know if you see any problems with it. And thanks again for these fixes.

Phake Nick wrote:

Somewhere in the message I have mistakenly typed "Hing Kong" instead of "Hong Kong"

Thanks for catching that. Fixed as per attached.

Michael H Deckers via tz wrote:

  On 2018-11-17 19:25, Paul Eggert proposed a change to Hong Kong time:

...

+# Assume the 1904 switch occurred when the time ball was dropped.

  1904-10-30T16:00Z probably is a better assumption for the switch in   Hong Kong to UT + 08 h; it is a local midnight, and Macao switched   at the same instant.

How do we know that they switched at the same instant? Although (unlike Hong Kong) for Macao we have a copy of the government order (it's cited in tzdb), that order doesn't specify even the date of transition, much less a time of day.

The source quoted gives just an ante quem datetime for the switch. It suggests that the time ball regularly was dropped at 01 h local time,

I'm thinking it happened something like this: * The Governor of Hong Kong asked the Colonial Office for permission to make the switch. (See <http://adsbit.harvard.edu/cgi-bin/t2png?bg=%23FFFFFF&/seri/Obs../0027/600/00...>.) * The Governor was given permission, and put out an announcement that the change would take effect 1904-10-30, without specifying a time of day. Possibly the official announcement didn't even specify a date. * The Hong Kong Observatory changed the time when it dropped the time ball near what was then the Marine Police Station, Tsim Sha Tsui <https://commons.wikimedia.org/wiki/File:Timeballs_near_Marine_Poice_Headquar...>. Local ships and people in the port could pick up the new time from the time ball. * Most people in Hong Kong didn't care about the time that accurately, and probably got the change indirectly via the minority who did care. Now, a complication. A reliable source says that the time ball was regularly dropped at 1pm, not 1am, and suggests that it wasn't dropped on 1904-10-30, a Sunday. "The ball was raised manually each day and dropped at exactly 1pm (except on Sundays and Government holidays)." See page 26 of: Dyson AD. From Time Ball to Atomic Clock. Hong Kong Government. 1983. https://www.hko.gov.hk/publica/gen_pub/timeball_atomic_clock.pdf So the scenario of the ball being dropped at 17:00 GMT daily appears to be wrong - most likely the astronomer in London got a.m. and p.m. mixed up, and since the transition was on a Sunday there was no time ball to observe anyway. This suggests that we should go back to an unspecified time for that transition. Proposed further patch attached and installed into the development version. And thanks for bringing this up.

On Sun 2018-11-18T02:58:04+0000 Joseph Myers hath writ:

...

So the scenario of the ball being dropped at 17:00 GMT daily appears to be wrong - most likely the astronomer in London got a.m. and p.m. mixed up, and

An astronomer before 1925 referring to GMT would have been using the old astronomical convention where the day started at noon, not midnight.

Yes. On Sat 2018-11-17T21:50:11+0000 Michael H Deckers via tz hath writ:

The source quoted gives just an ante quem datetime for the switch. It suggests that the time ball regularly was dropped at 01 h local time, and that local time had been advanced by 00:23:18.14 h at some time on or before the ball was dropped on 1904-10-30. The source also treats the switch to UT + 08 h as a different event, the "final step" in the adoption of UT + 08 h in Hong Kong. On would not expect such a wording if the switch had happened at the same time.

According to the observations made at Hong Kong Observatory by the folks whose clocks controlled the drop of the time ball Meteorological Observations made at the HongKong Observatory in the year 1904 page 4 https://books.google.com/books?id=kgw5AQAAMAAJ&lpg=RA7-PA31&dq=hong%20kong%2... the time ball is not dropped on Sundays or Government holidays, but the log of drop times in Table 2 shows that on Sunday 1904-10-30 the ball was dropped. So that looks like a special case drop for the sake of broadcasting the new local time. Michael Deckers is probably right that the change was effective in the middle of the night between Saturday and Sunday (but a lot of clocks would not have been reset until observation of the usual and expected time ball drop on Monday). -- Steve Allen <sla@ucolick.org> WGS-84 (GPS) UCO/Lick Observatory--ISB 260 Natural Sciences II, Room 165 Lat +36.99855 1156 High Street Voice: +1 831 459 3046 Lng -122.06015 Santa Cruz, CA 95064 http://www.ucolick.org/~sla/ Hgt +250 m

Thanks to all. I applied the attached proposed patch to the development version. I had some trouble reading that Hong Kong newspaper: the URL is enormous and apparently it requires Flash to read, which these days is problematic for security reasons (my browser refuses to run Flash). So I just put in the root of the URL.

   On 2018-11-18 02:05, Paul Eggert asked:

How do we know that they switched at the same instant?

    Because Decree No 204 of 1904 for Macau says that the switch is     done to keep Macau time equal to the time in "Cantão" (Guangdong).

Although (unlike Hong Kong) for Macao we have a copy of the government order (it's cited in tzdb), that order doesn't specify even the date of transition, much less a time of day.

    Incorrect. The decree does give the date ("30 do corriente")     for the switch and it is quite clear about the time of day     of the switch: it states that the new offset applies since     Oct 30, and for the full day of Oct 30. As the offset is     increased, this implies that the switch happened at the     instant when UT was 1904-10-29T16:00.     Actually, the switch described in decree 204 is the one done     in "Cantão", but it is obviously also intended to be applied     in Macau.     Michael Deckers.

    On 2018-11-18 20:04, Paul Eggert wrote:

However, Hong Kong switched at 17:00 UT so if the intent was for Macau to keep in lock-step with Hong Kong, the Macau transition should be at 17:00 UT.

  But we simply do not have any evidence that "Hong Kong switched at 17:00 UT".   Steve Allen has shown that the documents available for Hong Kong only show   that the time ball was flown (and a cannon was shot) when UT was 1904-10-30T05   (and GMT was 1904-10-29T17). That happened 12 h after your proposed switch   of Hong Kong time to UT + 08 h. We just do not seem to have any new   evidence for the switch of Hong Kong time in 1904.   But we do have evidence that Macau switched to UT + 08 h when UT was   1904-10-29T16 (tzdb currently says, 1904-10-29T16:23:18), and   that they did that in order to achieve a common time scale with the   neighboring parts of China.   I do not want to pester the list with my quibbles about your proposed   changes, and this will be my last post on this issue -- I just want   to make sure that tzdb makes the most proficient use of the aggregated   knowledge of the people in this list.   Portaria N° 204 of 1904-10-21 for Macau was announced by P Chan on 2018-05-10   and it was rightly noted in tzdb that this indicated the switch to a common   time scale with Hong Kong. The new documents for Hong Kong do not   change this view (otherwise the text of tzdb would have to be corrected   anyway). For all we know, it is likely that Macau and Hong Kong switched   to UT + 08 h at the same instant (and that instant most likely was when   UT was 1904-10-29T16).   Michael Deckers.

Steve Allen wrote:

almost certainly Henry Park Hollis http://www.royalobservatorygreenwich.org/articles.php?article=1099

Thanks, I incorporated that into the patch I just circulated <https://mm.icann.org/pipermail/tz/2018-November/027237.html>.