Mark Kent wrote:
> High Plains Thumper <highplainsthumper@xxxxxxxxxxxxxxx> espoused:
>> Roy Schestowitz wrote:
>>> High Plains Thumper on Sunday:
>>>
>>>> http://resources.zdnet.co.uk/articles/features/0,1000002000,39291282,00.htm
>>>>
>>>> A Linux thin client for every child
>>
>> I think that thin clients are the wave of the future. Thick
>> clients have always been a liability except for those who require
>> intensive local desktop computing power and data rates, like CAD,
>> laboratory analysis and simulations.
>>
>> By being on a server makes backups a whole lot easier. Patch
>> management and software upgrades are simplified. Client goes
>> dead? Swap out with another, set-up is minimal.
>>
>> This type scenario also can be used to bring new life into older
>> PC's.
>
> Total virtualisation has to be the best way forward, but it does require
> an entirely different economic model than we have at the moment. Who
> will pay for all the CPU cycles on some central machine? Who will pay
> for the bandwidth between? How will billing be performed? How can one
> uniquely identify someone using a Nokia 810 or an Asus Eee and sharing
> processing power?
>
> Unfortunately, I suspect that the shared cycle theory could be too
> utopian to be practical, at least in our present environment.
I was referring to local area networks, not Wide-net services.
The thin clients off a client server makes sense for multi-client
desktops. It holds promise for a home user who uses out-of-date clients
to provide connectivity, let's say for every bedroom and living or den.
Take a corporate work environment. Having to manage 20 servers, say for
400 thin client desktops means reducing infrastructure management to 20
instead of 400+ hosts (clients plus servers). A 4,000 thin client desktop
base then would mean 200 servers, which is definitely easier to manage
than 4,000 thick clients.
--
HPT
|
|