On 2018-07-27 09:32, Steve Summit wrote:

...

"Big tech warns of 'Japan's millennium bug' ahead of Akihito's abdication" https://www.theguardian.com/technology/2018/jul/25/big-tech-warns-japan-mill...

Fascinating. From the article:

| Unicode [can't] set the standard for that new character [...] until | it knows what it's called, and it won't know that until late February | at best. Unfortunately, version 12 of Unicode is due to come out | in early March...

Sounds like the problems encountered by some other people I know, trying madly to release their updates in time for the results of last-minute time-related political decisions... :-)

This project has provided an opportunity for responsible orgs to develop and test more fast path release processes, on top of urgent security updates. They can now plan to expand these processes to update and push out Unicode code point, converter, and font changes for common fonts, and additions to LC_TIME category ERA segments for %E[cCxXyY], for those locales. This opportunity is now being made available to Unicode, CLDR, ICU and other locale-related projects and their downstreams and distributors. It should not be too bad, as the last transition was in 1989: a Y2K precursor. Then, a number of alternatives appear to have been discussed and floated, depending on the new emperor's tastes in governance and reading, whatever the era is expected to bring, and what he wants to be called and remembered as after his death, prior to a final decision and announcement of the official name of the "Naruhito" era. Prototype single glyphs for all alternatives may be designed and considered for a reserved Unicode codepoint. Older character sets will have to use two (or more) characters, and conversions take that into account. Usage of era years may be lower nowadays outside of government, and those dealing with goverment documents and forms; although the number of mobiles, tablets, desktops, and servers will be higher; software quality on retail devices may be lower in this respect, with many Android devices, mainly phones, from 2017 or earlier no longer receiving updates by then (more so if Google settles soon with the EU on their demands) although Japan may push to get this update out widely to older devices. [From Wikipedia on Heisei period: "Thus, 1989 corresponds to Shōwa 64 until 7 January, and Heisei 1 (平成元年 Heisei gannen, gannen means "first year") from 8 January. To convert a Gregorian calendar year (after 1989) to Heisei, 1988 needs to be subtracted from the year in question (e.g. 2018−1988 = Heisei 30) (2018 = Heisei 30). The Heisei period will likely end on 30 April 2019 (Heisei 31), the date on which Emperor Akihito is expected to abdicate the Chrysanthemum Throne."] -- Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

On 27 July 2018 at 16:29, Clive D.W. Feather <clive@davros.org> wrote:

Steve Summit said:

...

Fascinating. From the article:

| Unicode [can't] set the standard for that new character [...] until | it knows what it's called, and it won't know that until late February | at best. Unfortunately, version 12 of Unicode is due to come out | in early March...

Why can't they reserve and publish a code point for the character, even though they don't yet know its name or appearance? They know it's coming.

Oh, the codepoint *is* reserved: U+32FF. That's the easy part. Unfortunately, the rest is not so simple. As Ken Whistler's recent message to the organization <https://www.unicode.org/L2/L2018/18220-u121planning.txt>, quoted and linked in the article, states: The characters encoded for these calendrical symbols in Unicode have compatibility decompositions, and those decompositions depend on the actual name chosen for the era. Because the decomposition, once assigned, is immutable, involving Unicode normalization, the UTC cannot afford to make any mistakes here, nor can it just *guess* and release the code point early. -- Tim Parenti

Unless the new emperor messes things up and chooses a single-character name that's already in Unicode. (That'll teach these computer whippersnappers who's boss!...)

I don't envy all the people having to update Japanese calendar or text-processing software. And things will be even more hectic if the current or new emperor dies unexpectedly. Luckily for tzdb, though, Japan does not change its *clocks* when it gets a new emperor, so we will be off the hook on this one.

It have already been announced that the new era name will have 2 characters. Japan specifically asked for a code point in Unicode BMP for legacy systems. That should make thing easier to most people but also not quite sustainable due to the limited space in BMP If era name changes are abrupt due to change of emperor or other reasons, realistically people would just use old era name until updates are ready. For tzdb, iirc the era name info is also stored in some notes comment. That could be updated too.

On 2018-08-04 21:35, Paul Eggert wrote:

Phake Nick wrote:

...

For tzdb, iirc the era name info is also stored in some notes comment. That could be updated too.

I don't see that in the tzdb source: although there are Japanese characters, there is no table of era names. Thanks for the other info, though.

New era name gengo 令和/㋿/Reiwa/R announced Mar 31 goes into effect May 1 R1.5.1 令和元年; H31.5.1 is an R1J bug; see https://en.wikipedia.org/wiki/Reiwa https://en.wikipedia.org/wiki/Japanese_era_name with support in CLDR 35.1, ICU 64.2, Unicode 12.1: http://cldr.unicode.org/index/downloads/cldr-35#TOC-V35.1 http://site.icu-project.org/download/64 https://www.unicode.org/Public/12.1.0/ucd/ -- Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada This email may be disturbing to some readers as it contains too much technical detail. Reader discretion is advised.