From: Paul Miller
To: Damon Hart-Davis
Subject: OpenTRV ETV progress
Date: Thu, 15 Sep 2016 14:32:24 +0000

Hi Damon,

Thanks for making time to talk to my today.

We agreed that the comments you had made on the protocol (18/08/2016) are f=
ine. We will add our parts regarding quality and auditing and then finalise=
 our internal review with guidance from one of our mathematical colleagues.=
 As soon as we have feedback regarding any improved statistical method, I'l=
l let you know.

The test plan is good, though as agreed it needs tidying and completing. My=
 recommendations are for the test body (i.e. OpenTRV) details and performan=
ce parameters and tests to be described at the start of the document. I als=
o recommend that the use of passive statements (e.g. possibly, should) is m=
inimised.

The draft report has enough information to follow the tests and analysis, a=
nd you will be performing more analyses and validations in the near future.=
  I'll send an example test report later.

The conclusion that your dataset does not meet the requirements of the prot=
ocol and test plan is not a disaster, but I will again ask for guidance fro=
m the commission regarding publishing a verification that does not meet the=
 performance parameters. The benefit of publishing a verification report (r=
ather than a statement of verification) is that it will provide a methodolo=
gy for future verifications.

...