Quote:
Originally Posted by icsys
If the Phorm system was truly opt-in where no data is passed anywhere near a Phorm server if not opted-in then there is no reason that the above could not be implemented.
|
... and that opt in means provider of SERVER and CLIENT opt in.
Speaking as the operator of several servers, I'd like the chance to decline to opt in too, so I can assume the privacy of my communication is also respected.
Pete
---------- Post added at 15:20 ---------- Previous post was at 15:18 ----------
Quote:
Originally Posted by SimonHickling
To that end does anyone have any links to BT or Phorm material where it states that a copy of the web page is made / kept?
|
Here, in a Webwise webchat. Marc Burgess is Phorm's CTO (Chief Technology Officer).
narcosis
What sort of impact will the scanning & tagging of http traffic have on response times during peak surfing times ?
MBurgess
Pages are not tagged (or modified), and the keyword analysis process is offline so it can't affect response times. The ISPs are very concerned not to interfere with users' browsing experience.
narcosis
If the keyword analysis process is offline then in order to scan for keywords would you not have to have a copy of webpage in order to analyze it offline ?
MBurgess
Yes, a mirrored copy is analyzed.
Also confirmed by
Richard Clayton's analysis
Quote:
1. The basic concept behind the Phorm architecture is that they wish to take a copy of the traffc that passes between an end-user and a website.
2. The actual mechanics of taking the copy differs from ISP to ISP, but one can view it
as a Layer 7 switch
34... Recall from paragraph 2 that the Layer 7 switch mirrors a copy of all of the
browsing activity to an out-of-band machine, it is this machine that inspects the traffc.
|