Re: 64-bit

["Leif Svalgaard" <leif@xxxxxxxx>]

From: <thomas@inorbit.com>
> > > IBM's first 64-bit computer, the 7030, was delivered to "Los Alamos
> > > Scientific Laboratory" on April 16th, 1961, so 64-bit computing goes
> > > way back.
> >
> > Please keep in mind that we're discussing bus widths, etc., for
addressing
> bits; not systems that had 64-bits of total memory.

the hyphen in "64-bits" prevented me from seeing the joke ("64 bit").
After 35 years in this business I have developed an occupational
disease: "I take things too literally".

>
> you are quite correct, the 7030 only had an address space of 24 bits,
> quite the same as a segment on the AS/400. How do you in RPG create
> and directly access an array with, say, 1000 Terabytes of data? as
> you should be able to do in "true" 64-bit addressing.


Although I was obviously (I hope) joking the first time around, this is now a
significant technical issue. My take on it is simple -- it's irrelevant.

That is, a particular language implementation, limitation, expression or
whatever, has no necessary connection to a particular hardware architecture
feature. A major point of HLLs is that such hardware details are hidden and
even untouchable.

==> true, but the hype about 64-bit computing is that "freed from the
limitations
of the smaller address space (4GB!) the programmers are much more
productive".
Since most of these programmers work in some HLL, if the HLL does not give
them access to the bigger address space, because "it is hidden", the issue
gains meaning.

-----------------

OTOH, more often than not, you have takes on such issues that are
instructive; so I gotta ask... Do you think that issue is meaningful? If so,
can you say why?