Neil Sloane <njasloane@gmail.com> wrote:

The proposed definition mentioned "Nth term is the year when pi was first calcuated to 10^N"

There are some obvious worries:

- why 10^n? why not 6^n?

Because people mostly use base 10. Similarly, I track decimal digits of pi, not base-11 digits. (Sorry, Carl Sagan.)

- who can be really sure which year something happened in?

Perhaps I should have said "published" rather than "calculated." I gave references. And yes, I have the Beckmann book, I'm not just taking Wikipedia's word for what it says.

- why year? what calendar? why not "minute"?

The calendar that's almost universally used, even in non-Christian countries, though they prefer to label it "CE" (Common Era) rather than "AD" (Anno Domini). The time of publication isn't known to the nearest minute, or even the nearest day. The best known non-year "calendars" are obscure indeed, whether they're Julian dates or seconds since the Unix epoch. Does JD 2457523 mean anything to anyone here without doing a conversion? Or Unix time 1464144038? Anyhow, Unix time is negative before 1970, which is awkward.

Is the instant when a computation was completed all that important?

The instant, no. The year -- I for one find it a lot more interesting than New York City subway stops or primes whose decimal representation contains "666," both of which are in OEIS. Why not Washington DC subway stops or primes whose base-6 representation contains "444" while you're at it?

- do you really think 1400 is a(2) or a)3) or whatever?

I could leave that one out, as I did with Archimedes, and start with the ones that are definitely known to the exact year. Though I have no reason to doubt that that one is not the exact year. Some years do just happen to end with 00. To avoid years that look too round, I could express all the years in base 6 if you like. :-)