[nycbug-talk] interesting read

[alex at pilosoft.com]

On Mon, 23 May 2005, Jay Savage wrote:

> If that's not IT, then what is?  I wasn't talking about the
> software/firmware that runs EKGs and displays the output over the
> patient's head; I was talking about the cabling, routing hardware and
> software, etc., that gets that information to a remote monitoring
> station--usually over some sort of standard tcp/ip network (usually
> ethernet), and the OS and userland used to display that information at
> the remote monitoring station.  Once the information is turned into
> packets, comes out of a serial or ethernet port, and starts running
> along cat5, coax, or fiber to a different part of the building, that
> certainly fits my definition of IT, especially when the results are
> delivered for local processing and display on a workstation running a
> different OS than the machine that produced the original output.
My definition: If lives depend on it, it is not IT - it is life-critical 
application, which should be managed and treated in a way that is totally 
different from your regular IT management.

> I'd say, "regulating information exchange between workstations,
> servers, and network appliances" was a pretty good start a a
> definition of IT, if such a thing is even possible.
> 
> I can see where there's room for cunfusion, though:  This may be a
> discussion where anecdotes and examples serve to cloud the issue
> rather than clarify it.  So let's put it this way:  If the data can be
> entrusted to any current Windows release, it can certainly be
> entrusted to a *BSD, or possibly even Linux.  If current Windows
> workstations were phased out in favor of opensource OS, there would be
> a significant cost benefit, and probably a significant increase in the
> stability of the infrastructure as a whole.
I certainly hope that life-critical applications are not running on
Windows.

<snip>
> Opensource projects have certainly proven themselves capapble of far
> surpassing the stability and performance of proprietary solutions in
> other venues, particularly mission-critical applications, why not
> healthcare?  Why should I suddenly trust Windows and software designed
> for it with my life when I don't even trust it them with my web and mail
> servers?  That seems backwards.
> 
> Embedded systems, though, I agree, are a place where opensource projects
> are of more dubious value.
Well, there's a point of open source for embedded systems, just not 
life-safety embedded systems.

-alex