The example is described as having been run under 32-bit Windows.  
Thus:

3!:3 x         - byte order and word size of the underlying machine
3!:3 x,o.1     - byte order and word size of the underlying machine
2 (3!:3) x,o.1 - standard byte order and 64 bit words

So where is the error?  Note that Intel and AMD use reverse byte
order, not standard byte order.



----- Original Message -----
From: "Miller, Raul D" <[EMAIL PROTECTED]>
Date: Monday, June 19, 2006 1:26 pm
Subject: [Jbeta] j endianness on 64 bit machines

> >From the documentation on 3!:
> 
>   (3!:3 x) ; (3!:3 x,o.1) ; 2 (3!:3) x,o.1 [ x=: 1 2 3 0 _1
> +--------+--------+----------------+
> |e1000000|e1000000|e200000000000000|
> |04000000|08000000|0000000000000008|
> |05000000|06000000|0000000000000006|
> |01000000|01000000|0000000000000001|
> |05000000|06000000|0000000000000006|
> |01000000|00000000|3ff0000000000000|
> |02000000|0000f03f|4000000000000000|
> |03000000|00000000|4008000000000000|
> |00000000|00000040|0000000000000000|
> |ffffffff|00000000|bff0000000000000|
> |        |00000840|400921fb54442d18|
> |        |00000000|                |
> |        |00000000|                |
> |        |00000000|                |
> |        |0000f0bf|                |
> |        |182d4454|                |
> |        |fb210940|                |
> +--------+--------+----------------+
> 
> Why does 2&(3!:3) use a different
> byte order from 3!:3?
> 
> (I was under the impression that amd64 used the same byte order
> as the 32 bit intel machines?)


----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to