Re: Two computers in one: two users each with their own accounts, monitor, and keyboard?
Albretch Mueller put forth on 1/12/2010 1:14 AM:
> Multiseat environments would be the next great
> thing, not only regarding the savings on hardware, but also on the
> utility bill ;-)
But here is where you diverge from reality. The total cost in hardware,
OS/application setup man-hours, and electricity is *higher* if you attempt the
multi-seat route with one big powerful box vs many small power efficient boxen.
And, as you point at, users end up with less capability wrt freedom of external
device use. Please take the following into account:
1. If/when this one machine breaks, all your seats (users) are down. In the PC
model, one machine breaks, only one user is down.
2. Because the market has evolved to the "PC everywhere" model, PCs are dirt
cheap, approaching $100 USD for the box (remember you have to buy all those
monitors, keyboards, and mice for the multi-seat setup so the only differential
factor is the price of the CPU box). The hardware cost for specialty cards,
signal repeater boxes and cables for the multi-seat configuration is actually
higher than buying a bunch of cheap low end power efficient PCs with standard
cables.
3. This is exactly why the Citirx et al "dumb Windows terminal" model has
failed. The dumb graphics terminals cost more than a PC by a factor of over
2:1, so people just bought PCs and installed the Citrix client. Then they
realized "why am I paying this Citrix license when I have a fully functional PC
sitting here for which I've already paid a Windows license?" Thus Citirix has
become basically nothing more than a remote access solution or a niche "security
minded" application solution for banks, hospitals, and such. Few, if any, are
running a Citrix desktop on every screen in their organization.
Linux is not Citrix, so there are no licenses involved. But to go the route you
are suggesting increases overall costs in hardware and man-time significantly
over a PC at every seat. I can't count the number of studies I've read over the
years that make the argument that centralized computing is the better model.
Thousands of organizations have tried it, and quickly dumped it, because it just
doesn't work in the real world as well as the PC model, and the overall costs
end up being much higher.
For the hobbyist at home who just wants the wife and kids to have a screen, KB,
and mouse to surf and check email, this multi-seat solution might work ok. But
again, given that you can buy a fully functional new laptop for $300 USD, or a
very decent used laptop for $150, the cost of extra vid cards, cables, KBs,
mice, a CRT/LCD screen and time time and hassle of trying to make multi-seat
work, it just isn't really worth the effort, is it?
The only reason I can see someone actually doing this is the satisfaction one
would receive from pulling off the hack and crowing about it. No offense
intended. It would be a fairly neat hack. But given the factors I've outlined
above, it wouldn't be worth duplicating by the larger community.
--
Stan
Reply to: