Development in the Landis case

In article <[email hidden]>, Tony Raven
[email hidden] says...

Quoted message said:

Clive George wrote on 18/03/2007 15:24 +0100:

Quoted message said:

And of course there's nothing wrong with running instrumentation on
"obsolete" software. If the machine was supplied with a computer and
software, there's no need to update it to get the latest shiny software
- in fact, it's a good idea to not do this, as that sort of thing
typically requires hardware upgrades at the same time.

There is if you are an accredited lab and the upgrades are to fix
deficiencies in the original software (as opposed to adding new features).

Is there any evidence that this is the case?

Simon Brooke wrote on 20/03/2007 09:43 +0100:

Quoted message said:

But is this the case?

It's not 100% clear because the lab is not releasing data on the
instrumentation and methodology which is the subject of a discovery
request by the legal team. However there is circumstantial evidence
that the equipment they have was a pre-production unit and possibly the
first of its type. If that is correct, and we won't know until and
unless it is forced to be disclosed, then there is at least a
requirement to bring it up to the commercial instrument standard.

Regardless the fact that they don't even have an operator's manual for
it despite the manual being downloadable from the web, is indicative
that the software has not been updated because of laziness/sloppiness
rather than a positive decision not to update based on a considered
assessment of the update changes.

--
Tony

"...has many omissions and contains much that is apocryphal, or at least
wildly inaccurate..."
Douglas Adams; The Hitchhiker's Guide to the Galaxy

Rob Morley said:

OS/2 used to be very popular with banks. ISTR reading that some
banks started running ATMs on Windows last year - people noticed
because they were crashing.

Nice example here. <http://daimyo.org/node/25>

Lot's more, including a number of blue screens of death, at <http://
daimyo.org/bsod>

--
Dave...

dkahn400 said:

Rob Morley said:

OS/2 used to be very popular with banks. ISTR reading that some
banks started running ATMs on Windows last year - people noticed
because they were crashing.

Nice example here. <http://daimyo.org/node/25>

Lot's more, including a number of blue screens of death, at <http://
daimyo.org/bsod>

Much, much scarier is the move to Windows on medical scanners. My head
of dept. has been given the reason from Philips Medical that they're
farmed out all their software to India, and the programmers there know
Windows much better than UNIX, so all the scanner consoles are coming
off UNIX and onto Windows.
Windows /is/ capable enough to run a scanner console, easily so. More
worrying, however, is scanners have to be connected to networks to be of
any use, and if you change the basic software from what is supplied
(including adding patches, virus scanners etc.) then the suppliers are
Not Happy and won't support it...

So sensationally stooopid you couldn't make it up :-(

Pete.
--
Peter Clinch Medical Physics IT Officer
Tel 44 1382 660111 ext. 33637 Univ. of Dundee, Ninewells Hospital
Fax 44 1382 640177 Dundee DD1 9SY Scotland UK
net [email hidden] http://www.dundee.ac.uk/~pjclinch/

Peter Clinch wrote on 20/03/2007 13:38 +0100:

Quoted message said:

Much, much scarier is the move to Windows on medical scanners. My head
of dept. has been given the reason from Philips Medical that they're
farmed out all their software to India, and the programmers there know
Windows much better than UNIX, so all the scanner consoles are coming
off UNIX and onto Windows.
Windows /is/ capable enough to run a scanner console, easily so. More
worrying, however, is scanners have to be connected to networks to be of
any use, and if you change the basic software from what is supplied
(including adding patches, virus scanners etc.) then the suppliers are
Not Happy and won't support it...

So sensationally stooopid you couldn't make it up :-(

I can understand where they are coming from having been there myself on
the supplier side. One of the big nightmares of software based medical
systems is validation and verification of the software, especially as
the operating system almost certainly cannot be either verified or
validated. As a supplier you will be in deep regulatory poo, especially
with the FDA is the US if they allow users to make unauthorised changes
to that software. The only way to protect yourself and discourage
unauthorised changes is to disown them and say you are on your own if
you do.

--
Tony

"...has many omissions and contains much that is apocryphal, or at least
wildly inaccurate..."
Douglas Adams; The Hitchhiker's Guide to the Galaxy

"Tony Raven" <[email hidden]> wrote in message
news:[email hidden]...

Quoted message said:

Peter Clinch wrote on 20/03/2007 13:38 +0100:

Quoted message said:

Much, much scarier is the move to Windows on medical scanners. My head
of dept. has been given the reason from Philips Medical that they're
farmed out all their software to India, and the programmers there know
Windows much better than UNIX, so all the scanner consoles are coming off
UNIX and onto Windows.
Windows /is/ capable enough to run a scanner console, easily so. More
worrying, however, is scanners have to be connected to networks to be of
any use, and if you change the basic software from what is supplied
(including adding patches, virus scanners etc.) then the suppliers are
Not Happy and won't support it...

So sensationally stooopid you couldn't make it up :-(

I can understand where they are coming from having been there myself on
the supplier side. One of the big nightmares of software based medical
systems is validation and verification of the software, especially as the
operating system almost certainly cannot be either verified or validated.
As a supplier you will be in deep regulatory poo, especially with the FDA
is the US if they allow users to make unauthorised changes to that
software. The only way to protect yourself and discourage unauthorised
changes is to disown them and say you are on your own if you do.

Which means the suppliers need to issue the patches, etc. As long as they do
that on a timely basis, then all is fine. (shouldn't really need a virus
scanner though - the machine should be sufficiently locked down that the
routes in shouldn't be open).

cheers,
clive

Clive George wrote on 20/03/2007 14:04 +0100:

Quoted message said:

Which means the suppliers need to issue the patches, etc. As long as
they do that on a timely basis, then all is fine. (shouldn't really need
a virus scanner though - the machine should be sufficiently locked down
that the routes in shouldn't be open).

The issue being timely. Verifying and validating a change in software,
even a trivial change, for a medical device can be a massive job of
work. I know of several medical devices that are locked in time warps
because the regulatory burden of making changes is simply untenable.

--
Tony

"...has many omissions and contains much that is apocryphal, or at least
wildly inaccurate..."
Douglas Adams; The Hitchhiker's Guide to the Galaxy

"Tony Raven" <[email hidden]> wrote in message
news:[email hidden]...

Quoted message said:

Clive George wrote on 20/03/2007 14:04 +0100:

Quoted message said:

Which means the suppliers need to issue the patches, etc. As long as they
do that on a timely basis, then all is fine. (shouldn't really need a
virus scanner though - the machine should be sufficiently locked down
that the routes in shouldn't be open).

The issue being timely. Verifying and validating a change in software,
even a trivial change, for a medical device can be a massive job of work.
I know of several medical devices that are locked in time warps because
the regulatory burden of making changes is simply untenable.

TBH I'd be tempted to have a separate firewall device between the scanner
box and the network - that way the scanner box doesn't need to be kept up to
date, only the firewall. Which isn't part of the machine, so doesn't need
the same level of validation etc.

cheers,
clive

Tony Raven said:

I can understand where they are coming from having been there myself on
the supplier side. One of the big nightmares of software based medical
systems is validation and verification of the software, especially as
the operating system almost certainly cannot be either verified or
validated. As a supplier you will be in deep regulatory poo, especially
with the FDA is the US if they allow users to make unauthorised changes
to that software. The only way to protect yourself and discourage
unauthorised changes is to disown them and say you are on your own if
you do.

Quite right, it's not *that* that's dumb, but moving from a more to a
less secure and stable platform to run your software from simply to let
you pay lower wages to the developers.

Pete.
--
Peter Clinch Medical Physics IT Officer
Tel 44 1382 660111 ext. 33637 Univ. of Dundee, Ninewells Hospital
Fax 44 1382 640177 Dundee DD1 9SY Scotland UK
net [email hidden] http://www.dundee.ac.uk/~pjclinch/

Clive George said:

Which means the suppliers need to issue the patches, etc. As long as
they do that on a timely basis, then all is fine. (shouldn't really need
a virus scanner though - the machine should be sufficiently locked down
that the routes in shouldn't be open).

In NHS Tayside we're using external CISCO firewalling boxes immediately
between the scanner and the network, and yes, that's a Good Thing, but
it would be a much /better/ thing if the scanner was based on an OS
that's far less prone to malware and attacks in the first place!

One point about locking them down is that the suppliers need remote
access to them... not impossible, certainly, but it is another hole to
guard, and another hole over a more fragile than necessary system
underneath.

Pete.
--
Peter Clinch Medical Physics IT Officer
Tel 44 1382 660111 ext. 33637 Univ. of Dundee, Ninewells Hospital
Fax 44 1382 640177 Dundee DD1 9SY Scotland UK
net [email hidden] http://www.dundee.ac.uk/~pjclinch/

Peter Clinch said:

Tony Raven wrote:

[snip]

Quoted message said:

Quite right, it's not *that* that's dumb, but moving from a more to a
less secure and stable platform to run your software from simply to let
you pay lower wages to the developers.

Same decision has been made for all future Royal Navy ship systems.
"Windows for Warships".

Peter

--
www.amey.org.uk

Peter Amey wrote on 20/03/2007 16:34 +0100:

Quoted message said:

Peter Clinch said:

Tony Raven wrote:

[snip]

Quoted message said:

Quite right, it's not *that* that's dumb, but moving from a more to a
less secure and stable platform to run your software from simply to
let you pay lower wages to the developers.

Same decision has been made for all future Royal Navy ship systems.
"Windows for Warships".

OTOH I have seen plenty of organisations spend a lot of money writing
their own software only to discover the hard way it is non-standard,
poorly supported and needs constant effort to maintain it.

--
Tony

"...has many omissions and contains much that is apocryphal, or at least
wildly inaccurate..."
Douglas Adams; The Hitchhiker's Guide to the Galaxy

In article <[email hidden]>, Peter Amey
[email hidden] says...

Quoted message said:

Same decision has been made for all future Royal Navy ship systems.
"Windows for Warships".

I was so concerned about that I emailed my MP about it a year or two
back (for all the good it did).

In article <[email hidden]>, Tony Raven
[email hidden] says...

Quoted message said:

Peter Amey wrote on 20/03/2007 16:34 +0100:

Quoted message said:

Peter Clinch said:

Tony Raven wrote:

[snip]

Quoted message said:

Quite right, it's not *that* that's dumb, but moving from a more to a
less secure and stable platform to run your software from simply to
let you pay lower wages to the developers.

Same decision has been made for all future Royal Navy ship systems.
"Windows for Warships".

OTOH I have seen plenty of organisations spend a lot of money writing
their own software only to discover the hard way it is non-standard,
poorly supported and needs constant effort to maintain it.

Just like Windows, you mean? :-)

in message <[email hidden]>, Tony Raven

(') said:

Peter Amey wrote on 20/03/2007 16:34 +0100:

Quoted message said:

Peter Clinch said:

Tony Raven wrote:

[snip]

Quoted message said:

Quite right, it's not *that* that's dumb, but moving from a more to a
less secure and stable platform to run your software from simply to
let you pay lower wages to the developers.

Same decision has been made for all future Royal Navy ship systems.
"Windows for Warships".

OTOH I have seen plenty of organisations spend a lot of money writing
their own software only to discover the hard way it is non-standard,
poorly supported and needs constant effort to maintain it.

The software which runs the systems of a warship are by definition non
standard. All warships use highly specialised equipment; different types
of warship have different types of specialised equipment; and there are
very few warships of each type. So whether the basic platform is Windows,
OS/2, UN*X, VME or bloody George III, all the software to run the ships
systems has to be written from scratch.

Given that, it's a bit of a no-brainer to choose the most stable and
reliable platform available. I mean, we wouldn't like to be like the Yanks
who have to have their warships towed back into port whenever the
operating system crashes....

Oh, whoops.

--
[email hidden] (Simon Brooke) http://www.jasmine.org.uk/~simon/

to err is human, to lisp divine
;; attributed to Kim Philby, oddly enough.

Rob Morley wrote on 20/03/2007 16:40 +0100:

Quoted message said:

In article <[email hidden]>, Peter Amey
[email hidden] says...

Quoted message said:

Same decision has been made for all future Royal Navy ship systems.
"Windows for Warships".

I was so concerned about that I emailed my MP about it a year or two
back (for all the good it did).

I hope it was to support the move. The less chance there is of weapons
working the happier I am. If Skynet had been based on Windows the whole
Terminator thing would never have happened. As it was they built it on
my favourite operating system, Hollywood OS* ;-)

*http://en.wikipedia.org/wiki/Hollywood_os

--
Tony

"...has many omissions and contains much that is apocryphal, or at least
wildly inaccurate..."
Douglas Adams; The Hitchhiker's Guide to the Galaxy

Simon Brooke wrote on 20/03/2007 17:26 +0100:

Quoted message said:

The software which runs the systems of a warship are by definition non
standard. All warships use highly specialised equipment; different types
of warship have different types of specialised equipment; and there are
very few warships of each type. So whether the basic platform is Windows,
OS/2, UN*X, VME or bloody George III, all the software to run the ships
systems has to be written from scratch.

And most of it is but as The Register said:

"However, the sad fact is that Windows will probably be a big step
forward for the Royal Navy (RN). Anyone who has spent time in an RN
warship is entirely accustomed to seeing equipment on which he may
depend for his life occasionally throw a double six for no good reason.
Windows may be unreliable, but it's hard to imagine it being as
failure-prone as the kit which is out there already."
http://www.theregister.co.uk/2007/02/26/windows_boxes_at_sea/

In fact IIRC one of the major losses in the Falklands occurred because
they were having to reboot the computer for inbound tracking at a
critical time

But I have seen organisations write their own e-mail clients, their own
AV software, their own firewalls etc etc. None of which needed to be
done with all the choices out there and none of which performed better
than the worst of what is out there but was done because their techies
though they could do it better.

--
Tony

"...has many omissions and contains much that is apocryphal, or at least
wildly inaccurate..."
Douglas Adams; The Hitchhiker's Guide to the Galaxy

Tony Raven said:

Simon Brooke wrote on 20/03/2007 17:26 +0100:

Quoted message said:

The software which runs the systems of a warship are by definition non
standard. All warships use highly specialised equipment; different types
of warship have different types of specialised equipment; and there are
very few warships of each type. So whether the basic platform is Windows,
OS/2, UN*X, VME or bloody George III, all the software to run the ships
systems has to be written from scratch.

And most of it is but as The Register said:

"However, the sad fact is that Windows will probably be a big step
forward for the Royal Navy (RN). Anyone who has spent time in an RN
warship is entirely accustomed to seeing equipment on which he may
depend for his life occasionally throw a double six for no good reason.
Windows may be unreliable, but it's hard to imagine it being as
failure-prone as the kit which is out there already."
http://www.theregister.co.uk/2007/02/26/windows_boxes_at_sea/

If you strip an NT class version of 'doze down to it's bare escentials it's
actually a pretty solid piece of work.

If we assume that you are not going to connect your warship's main control
computer to the internet, nor load it up with media-players and other consumer
gew-gaws, it should be fine.

Tony Raven twisted the electrons to say:

Quoted message said:

I hope it was to support the move. The less chance there is of weapons
working the happier I am.

The problem with that is, what if your opponent's military is running on
a (more) stable OS?
--
These opinions might not even be mine ...
Let alone connected with my employer ...

Tony Raven said:

I hope it was to support the move. The less chance there is of
weapons working the happier I am.

It rather depends on the mode of failure, Shirley?