Re: [lojban] (from lojban-beginners) pi'e

[Rob Speer <rob@twcny.rr.com>]

On Fri, Sep 14, 2001 at 09:06:37AM -0400, pycyn@aol.com wrote:
> Well, aside from being lousy at math  and history, what else can you tell us
> about yourself?

Nice flame.

> What new has been discovered about PA and where are these 
> discoveries published?

A while ago Xorxes made an informal grammar of PA cmavo, which was part of a
thread about what PA in various combinations meant. It seemed to meet with
general approval. (http://groups.yahoo.com/group/lojban/message/5817)

Of course, I suppose it might not _really_ be part of Lojban history, because
you didn't get to bless it with one of your sacred Records.

In the first go-round of the date argument, I saw nobody mentioning {no'o}
or {tu'o} or any such number, except in a digression about specifying
centuries. They just weren't used then. There weren't enough examples of
using them, and so using them was scary.

> What is this if not {pi'e}? -- it looks like {pi'e}, 
> acts like {pi'e}, and indeed is a paradigm case of {pi'e} in the Book.  As 
> for the math, you carried wrong, as I assume you know (among other things, 
> {pi'e} warns that something weird is likely to happen).  You also go the 
> bases wrong, since they are way over 2 -- 31 (give or take) and 12, though 
> the numerals don't keep up with the numbers.

You seem to be unfamiliar with the concept of a mixed base. I refer you to
http://groups.yahoo.com/group/lojban/message/2716.

You say I got the bases wrong and I carried wrong. I did neither; what I did
was to write the date wrong (day pi'e month pi'e year) and try to take it to
its logical conclusion. I apologize for being unclear in doing so.

Here's how this backwards date system works if you try to pretend it's any sort
of mixed base:

day month year = A B C

C is the units digit. It goes from negative infinity to positive infinity -
which is a strange thing for a units digit to do.

B is the 1/12ths place. Its digit ranges from 1 to 12.

A is the 1/365ths place. Its digit ranges from 1 to 28-31.

This is not good.

Consider an actual instance of mixed bases, such as a time including days. (If
I didn't include days the bases wouldn't be mixed.) When you express an elapsed
time such as 1d 2h 35m 6s, these are digits in the mixed base (24, 60, 60). In
this case, they're even nice enough to do what digits expect to do - the minute
and second places range from 0 to 59, and the hour place ranges from 0 to 23.

To add two times, you add the seconds and if the total is more than 60 seconds
you carry it over into a minute. If the total minutes are more than 60 you
carry over into an hour. If the total hours are more than 24 you carry over
into a day.

This could be done the same way with dates in year-month-day form, except with
less convenient digits. The units digit (the day) would range from 1 to 30 or
so, and the months would range from 1 to 12. The years would not have a set
range because they would not be carried.

I assume this concept is what is meant by "variable base" in the ma'oste.

What I was trying to demonstrate was not, in fact, that I am bad at math, but
that operations which apply to other {pi'e} numbers, such as addition or
subtraction, do not apply to day-month-year dates.

So, if the places have no mathematical relation to each other, why use pi'e? It
would have just as much meaning to express the day-month-year date 9/14/2001 as
{li 9 pe li 14 pe li 2001}.

-----

To get this argument out of the abstract:

In day-month-year, how do you refer to an event happening during a certain
year? It seems you don't. The lessons avoid this by naming years. So this
year is {la renonopananc.} and the next year is {la renonorenanc.} and the
next year is {la djimbab.}, or might as well be, because cmene are not
analyzable.

How do you tell me that {la kristoBAL. koLON. pu falnu litru le xasmi co blanu}
in 1492, and not during the lifespan of some guy named Pavoso Renanc?
-- 
Rob Speer