Pi Day is a fake holiday that exploits a universally beloved mathematical constant to promote US-style date formats.
ISO 8601 - Wikipedia

@RandomDamage @mattblaze
The problem with using ISO 8601 dates for Pi Day is that you have to make 2023 pi's.
@billstewart415 @RandomDamage @mattblaze Some see it as a problem. Others, an opportunity.
@zalasur
I see it as a whole lot of goodness to eat, Dean Winchester style ;)
@billstewart415 @RandomDamage @mattblaze

@billstewart415 @RandomDamage @mattblaze

Um, by my calculations, only 2006.

@ScottAkenhead @billstewart415

Exactly. And the other date formats give even worse results when evaluated. Some people are clearly terrible at basic arithmetic. (-:

https://mastodonapp.uk/@JdeBP/110026211550093604

JdeBP (@[email protected])

#Programming note: 2023-03-14 equals 2006 even in the #CLanguage and #CPlusPlus, because although the leading 0 denotes octal, 03 is luckily just 3. PS C:\> 2023-03-14 2006 PS C:\> In #PowerShell, too. And for the minority date format users: PS C:\> 14/3/2023 0.00230680507497117 PS C:\> 7/22/2023 0.000157282164202579 PS C:\> These are only going to get further away from pi as the years go on. (-: #maths #pi

Mastodon App UK
JdeBP (@[email protected])

#Programming note: 2023-03-14 equals 2006 even in the #CLanguage and #CPlusPlus, because although the leading 0 denotes octal, 03 is luckily just 3. PS C:\> 2023-03-14 2006 PS C:\> In #PowerShell, too. And for the minority date format users: PS C:\> 14/3/2023 0.00230680507497117 PS C:\> 7/22/2023 0.000157282164202579 PS C:\> These are only going to get further away from pi as the years go on. (-: #maths #pi

Mastodon App UK
@RandomDamage @mattblaze This is actually hard to argue with... both non-US orders are pretty valid and the choice between them seems pretty much arbitrary.

@apodoxus @RandomDamage

If you're doing month and day, order is arbitrary. If you add in year the European medium-small-large order makes no sense. I prefer year-month-day (sorts nicely!), but day-month-year is at least logical.

(yes, I fully expect us to resolve this cultural difference in a Mastodon thread)

@dveditz @apodoxus @RandomDamage Perhaps you reversed this? It's the US that does medium-small-large, Europe use small-medium-large (today is 14/3/2023 in Europe).

@not2b @apodoxus @RandomDamage

Ha! You're right. I disqualify myself from this debate (vote YMD!)

@dveditz @apodoxus @RandomDamage
Year month day all the way. That way your files sort chronologically when you alphabetize them

@RnDanger @dveditz @apodoxus @RandomDamage

Can we settle for writing (the hand- kind) DMY and typing YMD in lists, file names, etc?

My employer names our pay stub PDF using long date format (as in "14 mars 2023 - Sylvain Drapeau.pdf") and it tickles my OCD very much the wrong way...

@axnxcamr @dveditz @apodoxus @RandomDamage
I would not like that, no, it isn't good

@axnxcamr @dveditz @apodoxus @RandomDamage

For real. Like this is a perfectly functional listing 🙃

13 May 22
14 Jan 22
14 Oct 22
15 Apr 22
15 Aug 22
15 Dec 22
15 Feb 22
15 Jul 22
15 Jun 22
...

@axnxcamr @RnDanger @apodoxus @RandomDamage

People are going to write numeric dates the way they speak. Americans say "March 14th" so we write "3/14". Someone asks "what year?" so you go out at the end.

At some point we must've said dates the other way ("4th of July" is preserved as a holiday) but I don't know when we switched. If you say "14th of March" now people will think your historical reenactment hobby is leaking into mundane life.

@dveditz @RandomDamage

I meant the choice between YYYY-mm-dd and dd-mm-YYYY is arbitrary.... although I think the ISO-8601 order might have more going for it because you're just getting more and more specific... sorts better on computers anyway

@RandomDamage @mattblaze YYYY-MM-DD:HH:MM:SS is the best time format IMO. A constant tick up into an ever growing number
@lycanmatriarch @RandomDamage @mattblaze An Aztec QR-code LED clock of such would be nifty, whatever base you chose. Wouldn’t be human-readable, but that’s besides the point.

@RandomDamage @mattblaze I'm with Daniel on this. ISO dates are the way to go and it is trivial to accept if Pi day is to be an annual day then you make the year a wild card match. "Fake holiday" problem solved.

IMO the only other valid solution would be to celebrate day 314 of each year which would be November 10th except in a leap year when it's November 9th.

@enmodo @RandomDamage @mattblaze

The correct day is 22 of July. No discussion possible.

@viktorbir @enmodo @RandomDamage @mattblaze well, 22 July is *approximately* the right day...
@kauer @viktorbir @RandomDamage @mattblaze otherwise known as February 27th... You see I can swing both ways when it comes to pi.
@enmodo @RandomDamage @mattblaze Another solution would be to only celebrate Pi Approximation Day (22/7). After all, 3.14 is also only an approximation.

@Piloot @RandomDamage @mattblaze well everyone celebrates their birthday on an entire day which is only an approximation for an infinitely precise moment in time 🙂

So we are discussing fractional vs decimal approximation.

Given 22/7 to six decimal places is 3.142857 that's an error of 0.001265 and 3.14 is an error of -0.001592

So in absolute terms 22/7 is 1.25% more accurate than 3.14. Not much but I guess the fractions have it.

@Piloot @RandomDamage @mattblaze meanwhile, speaking of approximations, may I introduce you in a little saying:

"How I want a drink
alcoholic of course
After the heavy lectures
involving complex functions"

count the letters in each and you get...

3.14159265358979

and that's not even the longest version. See: http://cadaeic.net/naraven.htm

Poe, E.: Near A Raven

@Piloot @RandomDamage @mattblaze by the way, we haven't even gotten into discussions of Tau Day... https://tauday.com/

#tau #tauday #piday

No, really, pi is wrong: The Tau Manifesto

For millennia, the circle has been considered the most perfect of shapes, and the circle constant π captured the geometry of the circle in a single number. But π is wrong, and it’s time to set things right.

Tau Day
@RandomDamage @mattblaze of course by that logic, the TRUE pi day would occur on 3145-11-26!

@tkk13909 @RandomDamage Exactly. ISO dates start with the 4 digit year. 3145 is a LONG way away.

Also, don't get me started on the Y10K problem they're setting us up for.

@RandomDamage @mattblaze @tkk13909 imagine all the cobol developers they’ll be franctically thawing by 9990 🤣