In your special report “Filtering & Beyond” (March 2002), N2H2 is found to be lacking, supposedly because of scalablility. Volera and Inktomi, both of whom provide content delivery management technologies for N2H2 and who specialized in using the Internet Caching Protocol (ICP), would disagree.

There is a hint from one school administrator in the article that caching closer to the requestor can speed things along, which is the first half of web acceleration. But it doesn’t go on to say that with ICP, filtering would scale, be it N2H2, X-Stop, or whomever. The other half of scaling is to deploy more caching servers upstream from the local cache, and restrict the local cache to the upstream one. This can tier from two to many caches, the final one filtering.

N2H2 comes out smelling bad because of the way the article glossses over ICP. Just thought you should know.

–Steve Scarbrough
Technology Coordinator
Storm Lake Community School District
Storm Lake, Iowa

I just finished reading your front page article titled “Report: Some web filters might reflect bias” (April 2002), about a report released by the Responsible Netizen Project of the University of Oregon’s Center for Advanced Technology in Education, and I wondered why this article was even in your paper.

Internet filtering is mandated for schools by the government. Its very nature reflects bias, judgement, censorship, discrimination, etc. This means that some access will be blocked according to criteria by persons or groups in responsible positions. To question why conservative agencies are interested in web filtering is to miss that basic concept. To carry the thought further, I would submit that religious or conservative groups would be more interested in web blocking than non-religious or liberal groups. To link this with “excessive entanglements of religion with schools” is absurd.

Students are in school to be educated according to the criteria established by the community and the government. The government has mandated that schools provide internet filtering, and responsible schools are providing this filtering through a variety of methods. To suggest a conspiracy by filtering companies to block only non-conservative information is to be ignorant of how filtering software works. To suggest that the management of web filtering software is beyond the expertise of school district information technology staff is to be ignorant of how easily the filters are managed. If liberals (or conservatives) want to see what is available to students, they simply have to contact their local schools.

The real news and information regarding web filtering is “Who is making these filtering decisions, and how do school administrators provide reasonable guidelines to their IT departments?” It would have been informative if the article focused on that topic and had not been so slanted toward a fear of the religious right blocking access to the internet.

I question the degree of responsibility of the Responsible Netizen Project of the University of Oregon. To publish a report that “raises questions” of bias but has no factual proof or evidence other than guilt-by-association reminds me of the McCarthy era. I would hope that our well-paid, well-educated university professors would have shown a little more common sense in compiling this report.

–Terry Stephan
Aliso Viejo, Calif.

Your recent article “Report: Some web filters might reflect bias” caught my attention, because it goes to the heart of an issue that is of great concern to many of us in the educational community who are attempting to protect adolescents and young adults from sexually explicit and predatory web sites.

The article raises two issues that deserve attention. The first is that filtering algorithms are complex and could carry a developmental bias that can block sites that should be accessible. The Responsible Netizen Project may be a little ahead of itself in implying that because a vendor has a relationship with a religious organization, the religious organization’s point of view may be found throughout the product line. The entire point of block lists and filtering algorithms is to create customized blocking lists designed to meet the needs of each customer, not to develop a one-size-fits-all solution. However, this does not preclude the creation of programs that could effectively reflect a bias.

Secondly, just because a site may concern sex does not mean that it is pornographic. As noted in the article, there are many sites concerning sex that are not pornographic, and internet blocking software is often incapable of making the distinction between these two types of sites.

The Juvenile Justice Foundation ( is working on a project that would require ICANN (The Internet Corporation for Assigned Names and Numbers) to classify all commercial pornographic web sites with the extention “.xxx.” This extension would at least provide a filtering blanket that would permit schools and libraries to more easily filter out commercial pornography, which educators and parents find most objectionable when compared to other types of sites. Such a naming scheme would also make internet filtering programs more effective, because the largest block of objectionable sites could be easily identified, thereby permitting more resources to be used for selective site blocking.

Finding the correct balance between access and restriction is a very delicate and complex process—especially in the educational arena, where intellectual exploration and growth are the goal. Because there are so many commercial pornograhpy sites, a filtering extension that identifes them and permits easier filtering would allow us to be more selective about the other types of sites we filter in our schools and libraries.

–George Abruzzese
Network and Systems Specialist
Connetquot Central School District
Bohemia, N.Y.