Before you read this, you'll need to understand the notation of Graham's number, which can be found here. You'll also need to need to read up a little on the up-arrow notation done by Don Knuth Here .

In order to understand the world's largest finite
number, *Ed*, you'll need to understand the notation involved at arriving
at that number. If, indeed, Graham's
number is the largest defined number at the moment, it only seems fitting to
create yet another notation to encompass and surpass it:

Don Knuth constructed Graham's number as follows:

63. Graham's number ** G** = 3^^...^^3,
where there are G

Similarly, I will define a function called Ed_{1},
which uses the same method of construction:

Ed_{1}(x) is defined as G_{x} =
3^^...^^3, where there are G_{x-1} up-arrows, or in other words,
defined as the number of times you follow the process.
Graham's number is Ed_{1}(63).

Now imagine x was Graham's number. This produces a number of significant size,
but why stop there? That number is
simply Ed_{1}(Ed_{1}(63)).

why not go for Ed_{1}(Ed_{1}(Ed_{1}(63)))?
or Ed_{1}(Ed_{1}(Ed_{1}(Ed_{1}(63))))?
Well, now we have to come up with another
function, Ed_{2}(x), where x is the number of Ed_{1}s in the
number, as well as the x of the last Ed_{1}, i.e. Ed_{2}(3) =
Ed_{1}(Ed_{1}(Ed_{1}(3))), and Ed_{2}(63) = Ed_{1}(Ed_{1}(Ed_{1...}(Ed_{1}(Ed_{1}(63)))...))).
Naturally we'll want Ed_{2}(Ed_{2}(63))
and Ed_{2}(Ed_{2}(Ed_{2}(63))), which means we'll have
to have Ed_{3}, then Ed_{4}.
Ed_{x}(x) is defined, then, as the number of Ed_{x-1}s
in the number, as well as the x of the last Ed_{x-1}.
To simplify things, we're going to define ED_{1}(x)
as Ed_{x}(x). This makes ED_{1}(1)
3^^^^3, but ED_{1}(2) is already too large to express in the older
notations.

We have already written the definition of Ed_{1}(x).
If we wrote it out at the Ed_{1}(x)
level, those same words would be in the definition of ED_{1}(x), but
there would be other words added. Those
"other words" separating any two logical functions are now defined as
the *Difference in Logic Structure* (DLS).

We're going to define __Ed___{1}(1) as the
definition of Ed_{1}(1) plus the DLS of Ed_{1}(1) and ED_{1}(1).
This makes __Ed___{1}(2) the
definition of Ed_{2}(2) plus the DLS of Ed_{2}(2) and ED_{2}(2),
*plus* the DLS of ED_{2}(2). and **Ed**_{2}(2), where **Ed**_{1}(1)
is defined as ED_{x}(x). Since
we started with Ed_{x}, Ed_{x} is defined as logic level
1. ED_{x} is logic level 2, and
**Ed**_{x} is logic level 3.

__
Ed___{1}(x), then, is the definition of Ed_{x}(x)
plus the DLS of every logic level from 1 (Ed_{ x}(x) to x.
This makes __Ed___{1}(4) the
definition of Ed_{1}(4) plus the DLS of every logic level from 1 (Ed_{4}(4))
to 4 [**ED**_{4}(4), where (**ED**_{1}(x) is defined as **Ed**_{x}(x).].

__ED___{x}(y)
is defined as the number whose definition is __Ed___{x}(y), which
makes __ED___{1}(1) equal to 3^^^^3, and __ED___{1}(2)
equal to a far greater number.

Now, of course, there'll be __ED___{1}(__ED___{1}(__ED___{1}(__ED___{1}(4)))),
which can be rewritten as __ED___{2}(4), by essentially the same
rules that Ed, ED, **ED**, and all the other logic levels follow in their
subscripts. Now, we're going to define
the DLS of Ed and ED as logiclevel(1) and the DLS of Ed and __ED__ as
logiclevel(2). By defining **ED**_{1}(x)
as __ED___{x}(x), their DLS is logiclevel(1), and the DLS of Ed and **ED**
is logiclevel(3). Now, we'll define
ED_{
1}(x) as
the function whose DLS from Ed is logiclevel(x), and whose subscript and
independent variable is x. thus
ED_{
1}(1) is
ED_{1}(1), and ED_{
1}(3) is __ED___{3}(3).

Since I followed basically the same process to get
from Ed_{1} to Ed_{2} to ED_{x} to __ED___{x}
to ED_{
x}, by
making a new definition each time, we're going to define Edward(x) as the
following this process x times. Our
last function, Tivursky(x), is defined as the function arrived at by Edward(x),
with subscript and independent variable of x.
The number of bytes manipulated by every computer in history (an
ever-growing and yet finite number) will be defined here as *allbytes*.

*Drumroll please*

My number *Ed* is defined as

Tivursky(**ED**_{allbytes}(allbytes)) * 729 + 1.