NCs (network computers)

[jim bell]

At 12:53 PM 7/21/96 -0700, Vladimir Z. Nuri wrote:
>I thought this was a great message on the future "network computers"
>that may be coming out soon. a balanced view that shows how there
>may be a niche, and that there are also places where they will
>not be appropriate. the NCs could really potentially change the
>computer and cyberspace as we know it in a very significant way.


I think that the concept of "network computers" as presented is nearly a 
joke.   (It's a rehash of the common portrait of terminal/modem computers 
that was commonly promoted in the late 60's and early 70's.)   Their main 
advantage was supposed to be cost:  The "$500" figure is the one which is 
commonly presented.  However, in the middle of 1995 the components of 
computers which were NOT bargains were memory, which was kept artificially 
high by what I can't help concluding was price fixing, and over-priced CPU's 
from Intel.  The memory-price problem has now been solved after an extreme 
price decrease, and the CPU-price problem can be avoided to a great extent 
by staying with 486's or lower-end Pentiums.  What, then, are the remaining 
advantages of a "NC"?

The one thing that these network computers were supposed to save on, in 
addition to this, was a hard disk, but when I keep seeing those ads for $170 
1-gigabyte hard disks, it's hard to imagine how anybody would WANT to save 
this amount.  

Let's do a comparison:  Even a 28.8 kbps modem can't transmit much above 
3500 bytes per second after decompression, which is about 3 million seconds 
of data to fill a 1 gigabyte hard disk, or about 800 hours.  You'd have to 
be buying Internet access time for $170/800, or 20 cents per hour, to 
justify re-loading anything twice from the Internet as opposed to storing it 
locally.  Hard disks are a bargain, and it isn't worth NOT having one.



Jim Bell
jimbell at pacifier.com