On Thu, Feb 28, 2008 at 9:27 PM, Brett Glass <[EMAIL PROTECTED]> wrote:
> > And, hence, my suggestions with regard to what (if anything) the > FCC should do with regard to ISPs and P2P. > > First, there have been claims of inadequate disclosure of network > management policies and terms of service by ISPs. While none have > ever been made against our company, I'd love to see all such claims > put to rest once and for all, and would be quite willing to post a > form (much like the ones that come with credit card solicitations) > summarizing our policies prominently on our Web site. These could > include such things as permitted and forbidden activities (e.g. > outbound SMTP), and whether these are allowed under different > service plans (e.g. P2P might be allowed with measured rates but > not flat rates). However, ISPs should not be required to include > technical details (such as lists of blocked ports) that might allow > hackers to bypass security measures or would have to be updated > minute-by-minute as attacks were blocked. > So the users have to guess which ports are blocked, so when we are configuring something like DCC in our IRC clients, we won't know where the problem is when it fails to work. > Third, with disclosure, ISPs should be allowed to do such things as > post messages in the user's browser window, frame pages, or employ > similar mechanisms so as to provide informational messages to them. > (There was a furor a month or two ago when Rogers Cable did this -- > a shame, because the mechanism is extremely useful and unobtrusive.) Great, add yet another way that a user's browser can be hijacked. ISPs have no right to force me to read their messages. In many markets, such as mine, have only ONE provider for broadband access. The government cannot compel me to read speech that I don't want to read, and you shouldn't be allowed to do it either. > Fifth, content and service providers desiring to use P2P should be > required to PROMINENTLY disclose whether their software is capable > of turning the user's machine into a server or consumes any > resources beyond what it takes to transfer the content or provide > the service to that one user. Content and service providers should > also be required to turn off, by default, any features that turn > the user's machine into a server, and only allow them to be turned > on if the user chooses to enable them and the ISP indicates (via a > mechanism such as the "robots.txt" files commonly used to control > indexing) that such use is permissible. Note that, again, this > might vary by service plan or by venue (e.g., a Wi-Fi hotspot might > be more heavily restricted than a private connection). > This is crap. I own my own network up to my router in my home. Why do I need to ask permission to turn on features in software on MY machine in MY network? The ISP is just an upstream service provider. While I'm at it, client-server is how the Internet operates. X11 is a server. So, by your definition, I can't run X? Is your view that only businesses should operate servers? Why don't you just outlaw server software unless you have a license? Then I can wave my certificate and download apache. This is all utter nonsense. You make assumptions that all users are endpoints connecting to your network. It is possible for me to run several p2p apps on machines within my network, and never touch the ISP network. p2p does not equal bittorrent. p2p is a whole subclass of software. It is possible to have well-behaved p2p apps, so these draconian measures are over-reaching and patently unfair. Finally, any rules that are promulgated should make it absolutely > clear that ISPs retain the right to stop abuse of their networks > and to enforce their acceptable use policies and terms of service. > Otherwise, they could not defend against spamming, worms, "bots," > and other scourges of the Internet. > > As long as end users have recourse to sue the pants off the ISP when they "enforce" their AUP against the wrong party or without sufficient cause.
