Hello Andrew,

Wednesday, February 8, 2006, 8:55:30 AM, you wrote:

>> I am pleased that NEHTA have finally given us something to go on, and I
>> am relieved that they are pushing us to adopt what I think is a great
>> technology.

I think we need some more specifics however, otherwise there will be
as many brands of eSOAP as there are of people SOAP. Its like saying
we are going to message in "ASCII" - you need PIT and HL7 to define
structure within ASCII and we need wsdl to define structure within
SOAP.

AP> Agreed, this is a really useful, concrete plan for how things
AP> should work, using some excellent technology.

>> I have only two questions about their strategy:
>> 1. Not using SSL...?

AP> As I understand it, SSL has a very subtle problem
AP> which is that the PKI certificates
AP> are consumed at lower layers in the establishment of
AP> the HTTPS connection ,and are hence not available for
AP> non-repudiation of the actual payload. So whilst we
AP> can connect and mutually authenticate to each
AP> other, once the connection is finished, the only proof
AP> we had of each others identity, or proof of the content
AP> of the message is whatever we have logged in our
AP> systems. And that may be fine, and is certainly as
AP> much as most systems are doing today. But the

The certificates are used for the transport and the messages come out
without any proof of who it was from when you are asked the question
down the track. If you use client certificates you know at the time
but have no record of it other than logs, and logs do not provide
legal proof.


AP> WS Security standard actually allows the payload
AP> to be signed and encrypted, allowing both ends
AP> to mutually authenticate, but also keep a signed
AP> record of the message payload.
AP> I think this is generally considered the way to go
AP> (especially in health where non-repudiation of
AP> messages may be important)

But you do not need SOAP to do that, you can encrypt the message and
sign it and send it via SOAP, HTTP, EMail, floppy disc.

Our SOAP spec uses a digitally signed and encrypted payload.

We use three layers, an optional inner "in message" digital signature
that lasts forever and encryption and digital signatures for EACH
message in the transport layer. The inner signature need not use the
same technology as the transport layers (as only HESA PKI makes it
legal we use that but can do it with PGP and GNUPG). Its the inner
layer that provides non-repudiation (as its optional as specialists do
not need this to send to GPs). As the inner layer is vaid HL7 there is
no difficulty in keeping the signature and it can be evaluated at any
time in the future, in fact we do it every time we display it.

The outer layers are for transport security and authentication.

We (ie Medical-Objects) can also apply this scheme recursively, where
there are multiple transports in a chain of responsibility pattern. ie
we keep adding digital signature and encryption envelopes to the
original message or what we call our "Tunnelling protocol". As the
message is routed to the destination the layers are stripped off until
it gets to the final recipient.


AP> Andrew
AP> _______________________________________________
AP> Gpcg_talk mailing list
AP> [email protected]
AP> http://ozdocit.org/cgi-bin/mailman/listinfo/gpcg_talk



-- 
Best regards,
 Andrew                            mailto:[EMAIL PROTECTED]

Andrew McIntyre
Buderim Gastroenterology Centre
www.buderimgastro.com.au
PH: 07 54455055 FAX: 54455047


_______________________________________________
Gpcg_talk mailing list
[email protected]
http://ozdocit.org/cgi-bin/mailman/listinfo/gpcg_talk

Reply via email to