wireless X terms - any ideas?

Jim Carter jimc at math.ucla.edu
Sat Mar 30 05:24:49 EST 2002


On 30 Mar 2002, Aristotle wrote:
> Does anyone know what a comfortable specification would be for a
> wireless X term running on a diskless laptop would be - considering as
> well that the X data would be tunneled through SSH that adds
> compression.

Hmm, the constraints seem uncomfortable to me. Well, maybe not too bad
given 802.11b speeds, but you will need to download 600K or more of kernel,
plus 1800K of X-server, plus a certain amount of additional support files
in /etc, plus any local clients that you are going to provide. Kernels
(including X-server) for HDS and NCD X-terminals range from 1.5 to 2.3 Mb
in size.

I assume you're talking about what kind of CPU should be in the laptop.  A
300 MHz Pentium 2 or 3 should be fine, probably overkill.  The limitation
is going to be what kind of a screen you can get.  I'm not sure whether the
idea is to refurbish used laptops, or to buy the displays, motherboards,
processors, etc. and integrate them yourself.

Where I work, we had a lot of X-terminals, but as the price of a complete
system came way down, we've stopped buying X-terminals. The users can do so
much more, locally, with a real machine on their desks. (Of course it's
Windoze, hiss, boo!  But we have an aggressive program of pushing out
patches and security updates, making sure the users get the patches.)

Also there's a lot of advantage to the "thin client" that some X-terminal
companies are offering, i.e. you run an xterm on the local machine, and do
slogin - rlogin - telnet depending on your security model. It saves a
*whole* lot of network bandwidth and back-and-forth to the X-server across
the net, which is important in the wireless situation even at 802.11b
speeds. That's how I operate; in fact, I'm answering this mail using Pine
executing on the local machine (at home), with a SSH connection executing
remote imapd on my mail server (at work).

I may be throwing cold water on what you're planning to do, but that's the
experience we've had at the UCLA Math Department.

James F. Carter          Voice 310 825 2897    FAX 310 206 6673
UCLA-Mathnet;  6115 MSA; 405 Hilgard Ave.; Los Angeles, CA, USA  90095-1555
Email: jimc at math.ucla.edu    http://www.math.ucla.edu/~jimc (q.v. for PGP key)





More information about the wireless mailing list