Like any other engineering project, the first logical step is to compile a
complete system requirement. Once you really understand what the project
expectations are, it's pretty easy to decide on an architecture to support
those expectations.
To arbitrarily decide what's important to the customer, will certainly lead
to failure. For instance: You state that one of the reasons that you would
recommend VMS is that there are no known viruses. If viruses are really not
a concern for the customer, that should NOT be a factor in the decision
making process. After all, they may already have a very secure LAN and
experienced administrators. To make architectural decisions based on areas
that you perceive as important is a path to disaster. Don't assume that you
know what the customer wants and just as importantly, don't assume that the
customer knows what they want!
On the other hand, if "usability" is one of the primary factors, they may be
stuck with NT. As much as we may detest the OS for it's poor reliability,
far more people are comfortable with the WIN interface than any other OS. In
some environments, it may actually be more acceptible to reboot the system
every day or have unexpected failures than to retrain the people that will
use the system. Actually, NT used only for data presentation is not all that
bad. Once again, that's a business decision that should be addressed in the
requirements.
My guess: NT for the presentation layer, LINUX or a derivitive with strong
device support in the backend.
Just my $.02
Steve Robertson
>One of the Radiologist's I work with asked me the following question.
>
>Which computer would you buy if you were only going to do file transfers
>and
>archiving? I answered a VAX/Alpha running VMS since there were no known
>viruses and that system security was good. My answer was predicated on the
>fact that he needed storage for medical images and that he needed to
>maintain complete medical confidentiality. He is totally frustrated with
>the lack of reliability on Windows NT and is favoring SUN based systems.
>The trend in radiology is to move toward Windows NT platforms because the
>companies can't get support staff that can handle anything else. Our
>biggest problem, since we are a children's hospital is that we must
>maintain
>the data for 25 years. Actually 7 years once the child turns 18. In
>practice we plan never to get rid of any data.
>
>Examples of systems in Radiology
>
>CT scanner using SGI computer
>CT workstation using SUN running Solaris
>Vitrea 3D workstation using Windows NT on HP Kayak computer
>Laser film printer network running MS-DOS
>Laser film printer network running LINUX
>MRI scanner using SUN
>Film digitizer running on PC using Windows NT
>Nuclear medicine system with Windows NT, Mac G4, and Linux on PC
>VME systems embedded in film printers
>
>I have 8" floppies from old CT scanners, QIC tapes, DAT tapes, MO disks, 5
>1/4 floppies, 3 1/2 floppies, 9 track tapes
>
>I was offered from a major US company a radiology image management system
>the other day that had the following systems.
>Windows NT display stations
>Apple Macintosh image archive server
>Sun webserver and image server
>I don't think anybody could support and troubleshoot such a system. Their
>answer is that they support the system today and will provide an upgrade
>path when they change every 3-4 years.
>
>woe-is-me
>Mike
>mmcfadden_at_cmh.edu
>
_________________________________________________________________
Get your FREE download of MSN Explorer at
http://explorer.msn.com
Received on Wed Mar 07 2001 - 12:39:14 GMT