[IP] Re: Comcast and network neutrality and all that

[David Farber <dave@farber.net>]

In reaction to David's comment on me and coding. Yes I programmer  
SNOBOL a ,s and a bit of 3 and a lot of other things like most likely  
the first real compression program at BTL and hacked BellSys and SOS  
and various time sharing systems , debugged and moded early Fortan and  
IPL V ...  I can still find bugs in others code BUT as an editor and  
even as a Professor I let my readers and my students do the work and I  
provide the motivation. .. Somewhat in jest please. djf


Begin forwarded message:

From: "David P. Reed" <dpreed@reed.com>
Date: January 14, 2009 3:27:18 PM EST
To: dave@farber.net
Cc: ip <ip@v2.listbox.com>, Brett Glass <brett@lariat.net>
Subject: Re: [IP] Re:   Comcast and network neutrality and all that

Though I have nothing to do with the Switzerland tool, I am a trained  
programmer and I have a lot of detailed knowledge of network  
protocols.  Therefore, before making claims, I actually *read the  
code* when that code is available.

Curious about the basis of Brett's claims below, I reviewed the code  
briefly.   Here is what I learned about the code.  (Dave - you are a  
trained programmer, having coded Snobol, so you could have done this  
as well, rather than taking Brett's word for it).

Does it "detect transparent caching" (I assume Brett's complaint that  
this would be a false positive)?  Maybe. However, the notion of  
transparent caching is discredited.  You cannot transparently cache  
modern dynamic web pages that contain personal content.  The world  
doesn't agree with Brett that transparent caching is legitimate  
practice, and vendors who "count hits" and provide real-time stock  
quotes, etc. have threatened to sue anyone deploying such disruptive  
techniques.

Does it "detect blocking of port 25 against spammers"? No.  It will  
detect the case where a hotel or other "man in the middle" forces mail  
users to post mail through their own service, where it is inspected  
and *discarded without notice* rather than through the hotel guest's  
contractedl mail delivery service.  This detection is detecting an  
*interception* that was not authorized by the user, who unwittingly  
may be putting his mail in jeopardy of interference.

Does it "detect firewalling against open Windows shares"? No.  By  
definition if a switzerland user is validly accessing a shared windows  
file, and a problem is detected, the practice is *preventing* a  valid  
legal and willingly offered service.

Switzerland does one particular thing: detects modifications to the  
packet stream between point A and point B, which is something that the  
IP protocol standard specifies *should not happen* (except for  
fragmentation, dropping, and updates of certain IP flags and header  
options: TTL, ECN, and timing/source route options).

Does it "detect other beneficial practices"? I don't know what Brett  
is referring to here. If it doesn't detect the above, would it detect  
something else Brett claims?  Why should we take a general assertion  
as true?

Can any tool detect "anticompetitive business practices"?  Well, it  
depends on which practice.  It can detect NebuAd's substitution of  
content in a web page between the point of origin (the requested  
server) and the user requesting that.  That may actually be an  
anticompetitive practice, if it (as happened in the early days of  
telephony) redirects business away from one Funeral Home to a Funeral  
home who has paid the ISP for such redirection.

>




-------------------------------------------