Hacker News new | past | comments | ask | show | jobs | submit login
New emperor, new era: How a single word defines Japan (bbc.co.uk)
88 points by rwmj on March 30, 2019 | hide | past | favorite | 56 comments



In the article they mention the "leaked" era name "ankyu". This leak I expect was more of a political joke and never a true candidate.

Ankyu would be 安久 and thus mean "peace forever". Except the current prime minister Shinzo Abe's Abe is 安倍, notice the same first charcter. Thus the double meaning of Ankyu would be "Abe Forever".


strongmen stubbornly refusing to abdicate their office would however be very much in the spirit of modern times.


So that's why there was this Java update a few weeks ago...

    --------------------- java-8-openjdk/calendars.properties ---------------------
    index 49f68ac..8ffee4c 100644
    @@ -29,12 +29,14 @@
     #   Taisho since 1912-07-30 00:00:00 local time (Gregorian)
     #   Showa  since 1926-12-25 00:00:00 local time (Gregorian)
     #   Heisei since 1989-01-08 00:00:00 local time (Gregorian)
    +#   NewEra since 2019-05-01 00:00:00 local time (Gregorian)
     calendar.japanese.type: LocalGregorianCalendar
     calendar.japanese.eras: \
      name=Meiji,abbr=M,since=-3218832000000;  \
      name=Taisho,abbr=T,since=-1812153600000; \
      name=Showa,abbr=S,since=-1357603200000;  \
    - name=Heisei,abbr=H,since=600220800000
    + name=Heisei,abbr=H,since=600220800000;   \
    + name=NewEra,abbr=N,since=1556668800000


Disclaimer: I am a Japanese national who was born and raised in this country.

For god sake, I really hope fellow Japanese people stop using this inefficient and confusing year counting system.

Every time I’m asked what Heisei this year is, I don’t remember, so I have to open safari! It is really annoying.

Besides, who cares about who will become next emeperor of Japan. I would not care if current emperor suddenly die tomorrow. This might sound very disgraceful, but I rather hope the royal family of Japan vanish along with old-fashioned tradition.


Even worse is time keeping based on the combination of Chinese and Japanese zodiacal years which repeat in a 60 year sequence in conjunction with the imperial era. Showa ended up too long so there were duplicates.


Almost everything in life boils down to some kind of social convention which is essentially 'tradition'.

I don't know if you have kids, but as soon as they start to ask questions, it becomes clear.

Like why you can't just wear your thong to work? Or nothing at all? Pull at the thread and it comes apart pretty quick. Answers lie more often lie in identity and culture, then they do some kind hard logic.

Obviously Monarchies represent a distinct issue as there's the essential problem of class, but I'd also argue there's something much more than intangible there: so many of the most successful countries have monarchies. Popular ones to boot. Even countries which had powerful socialist leaders (and of course socialism/communism would be utterly opposed to monarchy in theory because that's literally the top of the class system they want to destroy). Sweden, for example is a very socialist country with a monarchy that's more popular than the UK monarchy in the UK.

Of course, you can dump all of that and just use Western time, and mark your birthday in 'days since Jesus was born' ... but I'm not sure how relevant that would be in bucking tradition :)

I also think there is value in the Japanese 'era' system in some ways.

While we see it through the lens of 'Monarchy', maybe we consider just seeing it as 'era' which uses the Monarchy a specific demarcation point.

After all, a new 'era' would have represented entirely new forms of governance, structure, international/feudal relationships. An opportunity to do things like debt forgiveness, heck even build some public works like big libraries or museums, such as happens for each US President.

In many ways, it makes (or perhaps did) make more sense to mark 'eras' in relation to the geopolitical context as it would have in hard dates.

In the grand scheme of history we use the Geological Time Scale [1] (i.e. eras, epochs) as it's quite a bit more convenient that just citing dates.

[1] https://en.wikipedia.org/wiki/Geologic_time_scale


Thank you so much for your thoughtful reply. I understand your idea.

> It boils down to “tradition”

Yeah, I think you are 100% correct, but I don’t think everything can be justified for that reason. For some countries, say, Sodomy Law is still tradition. In the U.S., people still use weird unit system, refusing adopting Metric system. I know they are extreme examples.

Celebrating new era is wonderful, but I hope people reconsider the use of this peculiar year counting system. It is a good time, I guess. Everything boils down to “tradition”, like you said, but I don’t think we have to stick to the same tradition.


Yes, I should have added that a lot of tradition/convention is arbitrary and dumb as well! Of course.


とにかく、昭和時代の音楽は一番。


It's just "subtract 1988".


Which makes it very easy for all people who were born in 1988, but inconvenient for everyone else :)


Makes it easier for people born in 1987.


Or just add 12 to the last two digits of the current year. Not complicated for current dates.

Not so helpful for non-Hesei dates.


The Emperor is used as a nationalist symbol to tie the people together (to promote us vs them).


What is "this country?"


In a thread about Japan, with a reply from a Japanese national talking only about Japan.

Hmm, I wonder which country it could possibly be.


One issue is that most programming languages links their version to a particular unicode version, by example, .NET 4 supports Unicode 5.1 while .NET 4.5 supports Unicode 6, etc.

The new Era will be soon supported by Unicode 12.1 [1]. For older version of .Net (or Java) you also need to support the new Era without supporting the whole Unicode 12.1 spec, introducing a non backward compatibility behavior because the new code point U+32FF has now a different meaning.

Currently we have already added tests for the new code point in all our supported applications and are ready to update them using the same channel as a security issue.

[1] http://blog.unicode.org/2018/09/new-japanese-era.html


Maybe era kanjis should be encoded in a systematic manner similar to flags. The EU flag is encoded as <Flag letter E><Flag letter U>, the new era would be encoded like <JP era number 2><JP era number 0><JP era number 1><JP era number 9>.


Your proposal does not handle the edge case of an emperor dying in the year of his ascension.


Easily solved by adding another digit after the year. That leaves the edge case of Japan having more than 10 emperors in a year, but thats unlikely.


Or they could just have era1=Meiji, era2=Taisho, era3=Showa, era4=Heisei etc. since Japan only had 4 eras named that way so far (and past era are pretty much not changing).

Then era1-era0 for the 10th era. Much simpler than building from the Gregorian year.


By building from the Gregorian year, you can convert the Japanese year to the Gregorian one without needing updates when a new era begins. I dont know how useful that is in practice though.

Otherwise you can just reserve a range of codepoints for era names.


Except that Japanese eras actually go back to approximately 645 AD, and some emperors reigned through multiple eras and others shared eras (Emperor Meiji himself began his rule in Keio year 3, and it wasn't until Keio year 4 that that era ended and the Meiji era began). Your scheme only works for modern eras rather than the full span of them.


Those kanjis are probably already encoded anyway.


It's not even edge case. 2019 is Heisei 31st until April 30, and whatever the new era will be from May 1.


Oh, interesting, didn't know that they "create" a new Kanji with the sum of all the possible kanjis in the Era name. Two questions:

- Should I be safe in the way that, even in my software doesn't support that unicode version, or my font doesn't have that symbol, I'll at least get a "notdef" or whatever symbol and not break everything? - what if the new name is composed by just one Kanji? Is it getting an unicode symbol anyway? Is it considered a different Kanji?


> what if the new name is composed by just one Kanji?

It will be two existing kanji. That's pretty much a given.

Don't worry about it. These compound characters are compatibility characters for older software, and for use as a compact notation in print (i.e., it is mostly a display issue). Any new code point assigned will not be in use now (U+32FF), so you won't break anything. Your software may display a box with the code point address in it, or just an empty box, instead of a glyph from one of the installed fonts.

If you deal with years written with the era name, you tend to get them written as two characters. This year (until the abdication) is 平成31年 (year 31 of Heisei).

Abbreviations do occur; you will then get only the first character of the name era, so 平31. This is why the new name will not start with 平, 昭, 大, or 明 (the first characters of the past four era names).


I believe newspapers are the most likely to use compounds, like they do for katakana, e.g. ㌔ for キロ.


The compound characters are also very useful for compact typography in tables, including public transport timetables.


To be clear there's no new kanji being created here, just a new unicode character with two existing kanji compacted into one glyph. So most typical use cases won't be affected either way, since era names are normally just written with two regular characters rather than with the compacted one.


So what's the point then? This sounds like a job for font ligatures.


I don't know much about unicode but presumably it's analogous to characters like "㎏" or "㎒". There are quite a lot of code points for several characters compressed into one glyph.


Those only exist for compatibility reasons, so you can losslessly round-trip convert from character sets that have those characters.


And you can bet the new compound will be in Shift-JIS. Most system in Japan still amazingly run in Shift-JIS.


> Oh, interesting, didn't know that they "create" a new Kanji with the sum of all the possible kanjis in the Era name.

They don't really, they create a new codepoint for a special (<square> in this case) composition of the era name.

> Should I be safe in the way that, even in my software doesn't support that unicode version, or my font doesn't have that symbol

It might fallback on the compatibility decomposition (the two combined kanji, so instead of "㍻" you'd have "平成".


It can fallback if the font does not have the new kanji, but not if the Unicode db is outdated.


Indeed, sorry for the confusing way I answered it.


Fun fact: Taiwan (ROC) also has an unusual year counting scheme, where they count from the founding of the Republic of China.

I once bought a Japanese product in Taiwan which had an expiration date using Japanese years, included a ROC year translation of said date and the receipt was printed with the gregorian year. Three different counting schemes in just one transaction, quite amusing.


The Democratic People's Republic of Korea (aka North Korea) uses what is basically the same scheme, and calls it the Juche calendar. That it counts from the same year is coincidental.


Even more coincidence: Japanese era Taisho also counts from the same year (1912).


America uses this as well, but only in very formal situations


I wonder if anyone here will be working hard in April to implement the new era in software?


Shouldn't be much work at all to add another name and offset. This isn't the first era change in my lifetime, and date libraries should already support the basic concept, e.g., https://developer.apple.com/documentation/foundation/nsdatec...


Well this is the first era change in the life on many software and libraries, so while everyone knew it was coming sooner or later most software have never been tested in "real life".

I just hope no Japanese developer stored dates as "era+month+day" instead of Gregorian or epoch (or relying on a library that does).


You bet there are. Many old Japanese system still store 2019 as Showa 94.


It's a huge annoyance for OS vendors as they have to rush to add the new character (yes, it has its own codepoint) to Unicode and make software start using it, and don't know what it will be in advance.


> Era names typically come from Chinese classical texts, and are usually only revealed when one emperor has died and another has taken the Chrysanthemum Throne.

Chinese? I'm guessing that's a typo. Not only would Japanese obviously fit better, but considering the bad blood between Japan and China I think they'd sooner not choose a word than choose one from China.


All kanji were originally Chinese, and still have an original Chinese reading along with Japanese.


Oh, they're reading very old texts then. That wasn't the time frame I thought of when it said "classic".


"According to news outlet Nikkei, the name-picking process has been rehearsed at least once a year for over 30 years, so officials are always primed for "The Day"."

I don't even care how inefficient this seems to Western sensibilities ... I love this.


See https://www.japanfs.org/en/news/archives/news_id034293.html https://www.smithsonianmag.com/smart-news/this-japanese-shri...

In many western countries many important traditional skills has been lost by the advent of technology. Either new materials and methods have been employed or many parts has been automated away or outsourced, therefore I love this instance from Japan where they rebuild a wooden temple once every 20 years so that many generations of artisans can work on the project together and pass on the knowledge:

“I saw one elderly person who probably has experienced these events three or four times saying to young people who perhaps participated in the event as children last time, ‘I will leave these duties to you next time,’”



My undergraduate school that I won't name was located on a 'land locked ship' in Canada. It was literally land, but commissioned and christened in the Royal Navy as a ship in the 19th century.

When they closed it, they had no idea what to do, and the Canadian Navy had to call up the Royal Navy to figure out the protocol.

They didn't know either.

So they had to dig through their own records to figure out the right protocol.

And then we did it: a 'ship decomisioning ceremony' for a land-locked ship, which I don't think has been done in a very long time in the UK/Empire, and probably never in Canada.

Pretty much the entire W. Coast Canadian Navy 'paraded by' ship by ship, and a few other neat things. A sight to behold.

Someone, somewhere is keeping track of all that.

The interesting paradox is how from one perspective, it makes no sense, and yet from another, it makes perfect sense.


>My undergraduate school that I won't name was located on a 'land locked ship' in Canada. It was literally land, but commissioned and christened in the Royal Navy as a ship in the 19th century.

Interesting. Near Pune in India, there is a training institute of the Indian Navy called INS Shivaji (INS = Indian Naval Ship). I was surprised by the name when I first heard about it. Seems like the same custom as the one you mention.

https://www.indiannavy.nic.in/insshivaji/


That sounds odd to me - more than one stone frigate in the RCN has been paid off in living memory, more than one during the unification process (which is within my memory). Yes, it's a bit of a ceremony, and necessarily a kerfuffle, but there is no reason why the protocol should have been a mystery.


My guess is that with the restoration of name and insignia of the Royal Canadian Navy, various traditions that had been dropped or elided from unification onwards were also brought back. Paying off of stone frigates probably changed in ceremony like so many other things did with unification of the forces, and so the modern RCN had to look up the traditional ways again.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: