Creating a crystal clear and pure Internet

michael.dillon at michael.dillon at
Wed Nov 28 11:36:30 UTC 2007

> On a more practical/technical level, I'm interested in how 
> French ISPs that worked on the plan to implement it on their networks?


I couldn't get a good copy from that URL but I did manage to get one
from <>

First, they start out by saying on page 12 that they are persuaded that
stronger disincentives for pirating content have to be organised in a
realistic and pragmatic way. There is no unique solution whose success
is assured. It is an illusion to consider that all forms of piracy on
the Internet can be stopped. Nevertheless it is necessary to communicate
to the younger members of the public that free (as in beer) and illegal
costlessness has a cost. (Note that French has a word meaning
costlessness that is usually translated as free).

They go on to talk about a variety of technical measures which they
recognize pirates can pervert. They want to make it harder to
accidentally pirate stuff. They talk about go after the uploader of
content, not the downloader. 

Elsewhere in the document they get agreement from the industry to change
their behavior as well, such as watermarking content, which presumably
makes it easier to filter pirated content but leave the Linux ISO
torrents alone.

Under filtering of sites and protocols they mention that this can be
legally possible in certain specific circumstances. Under filtering of
files they talk about servers, where ISPs already will delete or block
downloads of pirated files. 

The appendices (annexes) go into more technical details. I didn't read
all 44 pages of this report but it is fairly balanced and they clearly
talked to ISPs as well as content owners. I get the sense that a lot of
this is alread best practices but is probably not all that well
documented in the sense that there isn't a "Piracy Prevention Best
Practices" document that most ISPs try to adhere to. If someone wanted
to produce such a document, extracting the technical bits from this
report would make a reasonable working draft.

Given that operationally, there are no magic bullets for most things, we
have to make do with a set of best practices, each of which deals with
one aspect of the problem. I wonder why we don't see more support for
documenting these best practices through something like this wiki
which was set up by a NANOG member a while back.

--Michael Dillon

More information about the NANOG mailing list