I get it, people. You've all got raging hard-ons for π. But this whole π Day thing really bugs me. All it is, is a date whose numerals happen to correspond with the first three digits of a decimal approximation of π. The thing is, π is an irrational number. It is not 3.14, nor is it 22/7. Hell, it's not even this

A rational approximation, no matter what level of precision it's computed to, will never be any more π than 10 is 11. Approximations are fine for science and industry, i.e. practical applications, as no known phyisical process requires precision beyond what can be offered by the available approximations. But in purely mathematical terms, to say that the million digit approximation linked above is equal to π is as wrong as saying 7 is equal to 98.

I realize this sounds pretty pedantic, but mathematics is pedantic. And I imagine that anyone who isn't a bandwagoner and actually enjoys mathematics enough to celebrate a holiday for π isn't bothered by pedantry. So I propose that if you want a π Day with any mathematical substance to it, consider celebrating on the 183rd day of the year. That's the day that contains the halfway point of the solar year. (Remember your junior high geometry? Circumference is 2π radians.)

Or maybe it's just me.

Addendum:

Since Earth's orbit is elliptical, Euler's number could also have a meaningfully-timed holiday based on Earth's location in orbit. Granted, deciding where in the orbit this should be might be a bit more contentious as it will propably require a more complex method of finding the appropriate day for it to have any real meaning. Given e's changing powers in the computation of elliptical circumference, one can't just divide it out and be left with a nice number like 2 to divide the year by. One of these weekends I might just get bored enough to figure out a good e-Day.