URLs, IPs and interception

[Roland Perry]

In article <02e001c87c65$69e1b580$1601010a@neos.tv>, Tom Thomson 
<cmt@btinternet.com> writes
>Sgriobh Roland Perry [lists@internetpolicyagency.com]:-
>
>> But the definition above says that interception does not include 
>>traffic  data.
>
>Traffic data is not interception - agreed.  Content is not interception
>either.  Interception is not any sort of data.  Interception is a sort
>of conduct.

And the conduct is defined in terms of what it is that's being 
intercepted (different kinds of data - some of which is content data, 
some of which is traffic data).

>You appear to interpret S.2 (5) (a) as excluding all conduct relating to
>traffic data from being interception.  That's amazing, the words are
>very plain and simple English, and make it absolutely clear that only
>particular conduct relating to traffic data is to be excluded, not all
>conduct relating to traffic data:-
>
>(a) any conduct that takes place in relation only to so much of the
>communication as consists in any traffic data comprised in or attached
>to a communication (whether by the sender or otherwise) for the purposes
>of any postal service or telecommunication  system by means of which it
>is being or may be transmitted; or
>
>So conduct which collects traffic data for a purpose other than the
>purpose of the postal service or telecommunications service by which the
>communication containing it is being transmitted is still interception -
>it is not excluded.

You are making the same fundamental mistake as others.

It's not the *conduct* that must be "for the purpose of a 
telecommunications system", but the *data* which is "comprised in or 
attached .. for the purposes of a telecommunications system".

>The purposes of a telecommunication system are defined elsewhere in the 
>act as being the delibvery of the communication to the intended 
>recipient.

Which in turn defines which of the data in the communication has been 
attached for that purpose.

There is some wriggle room - to give a clear example: I don't think the 
date of an email has much to do with its deliverability. But the "from" 
address does (as well as much more obviously the "to address") because 
if the email is undeliverable there has to be somewhere to send the 
error message back to.

Thus you can examine each tiny part of the data that accompanies what's 
indisputably content (eg the body of an email or an http request), and 
decide which ones are there for the "purposes of the telecommunications 
system" and which are just incidental.

In the case being discussed at the moment, it's clearly a good idea to 
examine each bit of an http request and decide which parts are necessary 
for the request to be delivered, and which aren't. In that way we can 
tell which parts are that sort of traffic data (and incapable of being 
intercepted) and those which aren't.
-- 
Roland Perry