Comment by danso | Hacker News Reader

Comment by danso | Hacker News Reader

Parent Full thread

Related thread from 10 months ago, about a MSDN article titled "The Japanese Calendar’s Y2K Moment"

https://news.ycombinator.com/item?id=17430085

From the MSDN article:

> The Japanese Calendar has Japanese Era Names that coincide with the reign of the Emperor. For most of the modern age of computing that has been the Heisei era, however the Emperor is expected to step down on April 30, 2019. Which will bring about the beginning of a new era. Fortunately, this is a rare event, however it means that most software has not been tested to ensure that it will behave with an additional era.

> The magnitude of this event on computing systems using the Japanese Calendar may be similar to the Y2K event with the Gregorian Calendar. For the Y2K event, there was world-wide recognition of the upcoming change, resulting in governments and software vendors beginning to work on solutions for that problem several years before 1 Jan 2000. Even with that preparation many organizations encountered problems due to the millennial transition.

I swear there was an HN thread about a language-specific implementation (Java? Ruby?) of calendar, which included having to deal with the Japanese epoch change.

A big issue is that people couldn't finalize any changes until the new era name had been decided and announced, which happened just a few weeks ago on April 1.

For things like Y2K, you could have done your deployment and testing long before it happened - for this change, everyone has just this one month to wait for any vendors involved to finalize, test and distribute their updates so that they can start deploying them on their own systems. No programming language or calendar library version older than March 31, 2019 could include proper support for Japanese dates of May.

> A big issue is that people couldn't finalize any changes until the new era name had been decided and announced, which happened just a few weeks ago on April 1.

Yeah, I was wondering about this line in the article:

> Officials compounded the problem by keeping the new era’s name secret until April 1, just one month before the transition.

Is there any reason, specific to Japanese tradition/convention/bureaucracy to keep the name of the era secret until one month before the actual change? When is the era name actually decided on, and how soon after the abdication of the current emperor?

The era name was decided by the government a couple of weeks before the announcement. Announcing a month in advance was a concession, conservative politicians wanted it like it always has been done, that the name is revealed at the start of the era.

I had to deal with the previous transition, and that one was going to be the result of the death of the emperor. Therefore, software was not allowed to include any provision for transition to a new date system, because doing so would imply that the emperor might die someday, which was strictly forbidden. I don't know whether it was an actual law, but there was zero tolerance for including any provision for changing the Showa date system.

As a foreign (non-Japanese) company, we just used the western-style date almost everywhere except for published documents to skirt this problem. Doing the wrong Japanese thing would have earned us serious hostility from the Japanese business press, but foreigners doing dates the foreign way would tolerated as mere foreignness.

The one month period was already a major concession by traditionalists: historically, the era name was announced the day it started.

Then again, this transition is also highly unusual because the emperor abdicated, allowing time to prepare. It's the first time in the modern era that this happened, as all previous (recent) emperors have died on the throne.

Does the announcement of a Pope have any advance notice at all?

From the other perspective, the only unique part about Japan is that they use a name for the calendar.

The Pope is conducted by secret election, and the candidates can be virtually any one who is a baptized Catholic man. I was under the impression that Japan's emperor has a line of succession? And which world governments (besides Vatican City) base how their bureaucratic system enumerates dates on the naming of the Pope?

The name of the emperor I believe is decided at the time of coronation though. They don't go by their real name but a new imperial name. That's my understanding at least.

Why only men?

Because for all of history and across cultures and civilizations, women have never been considered equal to men (socially or legally) - with the odd exception/legend here and there, of course.

"Because you're a woman and women can't do that" - "Why?" - "Because they can't." is what it boils down to.

Now for a more interesting stroll through Canon Law:

(1) The pope is the Bishop of Rome, and is (always) already a bishop (but if not, is immediately ordained as a bishop (Can. 332)[1]).

(2) Only men who have had the sacraments of baptism and confirmation (Can. 1033)[2] can receive the sacrament of ordination (Can. 1024)[3] and thus be bishops.

This has been reaffirmed (in 1994-1995)[4][5] as being part of the deposit of faith (i.e. Dogma).

In addition, anyone who tries to confer a sacred order to a woman (and the woman in question) are latae sententiae (implicitly) excommunicated.[6]

[1] http://www.vatican.va/archive/ENG1104/_P16.HTM

[2] http://www.vatican.va/archive/ENG1104/_P3R.HTM

[3] http://www.vatican.va/archive/ENG1104/_P3P.HTM

[4] http://w2.vatican.va/content/john-paul-ii/en/apost_letters/1...

[5] http://www.ourladyswarriors.org/teach/ordisace2.htm

[6] http://www.vatican.va/roman_curia/congregations/cfaith/docum...

Of course, there's still the legend of Pope Joan[7]

[7] https://en.wikipedia.org/wiki/Pope_Joan

Very informative. Thank you for taking the time to write this in addition to citing everything!

"Is there any reason, specific to Japanese tradition/convention/bureaucracy to keep the name of the era secret until one month before the actual change?"

Yes.

> No programming language or calendar library version older than March 31, 2019 could include proper support for Japanese dates of May.

I don't think that's correct. The same way we handle timezone updates via external data source in most languages/libraries, we don't need to update the language itself to handle the new era. This could be done by preparing the code to handle changes and then publishing just the common database update which includes a new "(date) (era)" entry.

> A big issue is that people couldn't finalize any changes until the new era name had been decided and announced, which happened just a few weeks ago on April 1.

How many different ways were dates marked "Heisei" displayed? I suspect it would actually be easy to finalize changes using a placeholder string, and just update to the actual era name when it was announced.

Deploying a one-character change is still a deployment that requires effort and time to do and verify - ensuring that every system that you run, and every library that those systems depend on, and every library those libraries depend on have done and deployed that one-character change.

You think a one-character change is such a major project that there's just no way to get it done inside of a month?

Because every other part of the change can be implemented without reference to what the era name happens to be.

How do placeholders work in languages where you are forced to assign lengths to variables?

Not a Problem in this case, the era is always a single character, a newly formed ligature of two kanjis.

The code point was already reserved in unicode 11.0, so you could even just the unicode code point.

The real world is not so perfect. I never see Heisei or Showa, or earlier reigns as a single character. The Japanese are also not fully onboard with Unicode so there's an extra level of effort needed if you want to get into a normalized form.

For this specific question, I strongly suspect that whatever the length of your Heisei string is, it's safe to assume that the new era string will be the same length.

>The real world is not so perfect. Perfect summation. Curious how this will also impact other consumers of Japanese data.

Then it looks even harder. How do we handle it if our system doesn't support Unicode 11.0? Is platform upgrade only option?

Ruby 2.6.3 added support for the Japanese new era last week. You can see the ruby-lang thread here [0].

[0] https://bugs.ruby-lang.org/issues/15742

a good digital example of the last minute waiting til the name had been announced -- this issue only submitted on april 1

Yeah, probably Java. Here's the relevant info.

https://blogs.oracle.com/java-platform-group/a-new-japanese-...

So they don't use unix time?

Reply on news.ycombinator.com