one-to-many messaging

[Peter Fairbrother]

James Firth wrote:
> Peter Fairbrother wrote:
>> Roland Perry wrote:
>>
>>> But it's so easy to persuade them it is. Assuming of course you can get
>>> the point over that port numbers identify which destination machine is
>>> required to service the request - after which it easily follows that the
>>> port number is part of the information required to deliver the message
>>> to the right server.
>> But suppose that port numbers are _not_ needed to identify the actual
>> machine, and all that is needed is the IP address - which will
>> frequently, if not usually, be the case.
>>
>> Then the section 2(5) requirement, the "purposes of a telecommunication
>> system by means of which it is being or may be transmitted" is not met -
>> there is no such system, and any "perhaps it might be needed for this
>> system system" is entirely imaginary.
>>
> 
> But the argument others do make is that, for several reasons including
> network resilience e.g. to DDOS attacks, performance e.g. low-latency
> protocols (Q/priority), then the network operator has to look at port
> numbers is order to decide on the best "path".
> 
> I use the term "path" loosely because usually it means how to shape the
> traffic and what storm pattern/characteristics to look for.
> 
> Also transparent caching relies on reading port numbers.

Fine, and that's (probably) legal - but it is interception as it's not 
excluded under 2(5). Roland is arguing that it isn't interception.

-- Peter Fairbrother