CSCI 10 Reading Summaries

For Meeting 1

QOS

We don’t need QOS for the Internet. QOS aims to reduce congestion on the Internet. Unfortunately, it helps only when the Internet is moderately congested. When congestion is heavy, having QOS only worsens the traffic. Also, QOS algorithms may be overly complicated, because they must be tailored for each congestion situation and each tefchnology. It is probably that the cost of having QOS outweighs the benefits. On the other hand, we can solve the problem of congestion by simply adding more cables to a network. The benefits of this approach are clearly greater than QOS.

Histories of the Internet

Since its inception, the Internet has gone through two phases: development of the protocol, and growth of the network and its services.

Development of the Protocol:
Protocol is to the Internet as grammar is to a language. It’s the driver application that each computer must run in oprder to communicate with one another. In its infant stage, the Internet, called the ARPANET, used the Network Control Protocol. This protocol allowed computers—very big, mainframe supercomputers—to talk to one another, but it was unable to cope with ARPANET’s growth. For example, it could not deal with transmission loss and transmission error, two problems that must be handled should the ARPANET grow larger. TCP/IP was then designed to cope with growth. In 1982, it replaced NCP as the protocol driving the ARPANET. After personal computers had their implementation of TCP/IP, they too began to join the ARPANET. This marked the beginning of its rapid growth.

Growth of the Internet and its services:
The ARPANET was renamed the Internet. It continued to grow and was then divided into two branches, one for civilian use and one for military. In the US, government agencies built the physical infrastructures connecting within the country and with other countries. Within the country, regional networks (Verizon, AT&T) sought commercial users. Commercialization reached pinnacle in 1995, when the NSFNET Backbone, America’s part of the Internet, was sold to different regional networks. Since then, the Internet not only grew in terms of users, but also in terms of services. There was web page, then there was e-commerce, instant messaging, Internet phone, file sharing, audio, and now video.

Not only is TCP/IP the universal protocol of the Internet, it is also a protocol open to and free for everyone who wants to know about it. This contributes to the rapid growth of the Internet. Researchers look into TCP/IP for flaws and sought ways of improvement. Because TCP/IP is centerl to the Internet, networking vendors cooperate to make their products conform to the protocol to ensure interoperability. Therefore, research prepares the protocol for future needs. In turn, companies use research results to provide more Internet services for more people.

For Meeting 2

The Coming Tug of War Over the Internet

Telecommunication companies own the physical lines of the Internet. Internet companies like Google and Yahoo offer their services via the Internet. Currently, they don't pay telecommunication companies for usage of their lines. Telecommunication companies want Internet companies to pay for line usage, because they don't want to carry the additional digital traffic caused by the services provided by Internet companies without any compensation. If Telecommunication companies' argument wins in Washington, users will probably not be able to use the services from Internet companies that don't pay enough for bandwidth usage. An alliance between Internet companies and bloggers are lobbying against telecommunication companies. They argue that, with this as a precedent, telecommunication companies will be allowed in the future to affect the contents Internet users receive. This is against the "end-to-end" model, on which the Internet was grown. The "end-to-end" design means that content is produced by content provider and received by users without intervention by companies owning the switching equipments and communication lines. When the "end-to-end" model is violated, there will not be neutrality on the Internet, and neutrality is what content providers as well as users cherished and the element that has made the Internet successful.

No Tolls on the Internet

The congress is deciding whether to end net neutrality. Net neutrality means that network providers are not allowed to control the contents of the Internet, and in the past has been protected by law. If net neutrality ends, telecommunication companies will be able stand in between content providers and users, offer express lanes to Internet services of the companies pay them enough, and leave the rest of the content providers in slow lanes. As a result, we will lose the opportunity to freely distribute information. Losing this freedom kills the openness of the Internet, making it no longer an engine that drives innovation in all fields.

Catching the Web in a Net of Neutrality

The Internet is capable of important tasks such as remote disease monitoring and tele-medicine. These services and others like these are beneficial to society as a whole, but depend on the Internet being a reliable place for communication. They would have been already realized, if there had not been Net Neutrality to prevent telecommunication companies from clearing the Internet of junk services. Net neutrality stands in the way, because it allows the Internet companies to pour out their services without regard to bandwidth limits. Each service takes away some of the bandwidth, a negative exeternality imposed on every other service that is on the Internet. However, Internet companies do not care about negative externalities, because they don't pay for it. As a result, they tend to provide more services than are socially desirable, making the Internet unreliable. The way to fix this is to make them aware of the negative externalities by imposing a cost on the extra burden each service imposes on the Internet. When companies have to pay for this cost, they will lower their usage of the Internet by cutting down their services. As a result, they approach a level of Internet usage that is socially desirable.

For Meeting 3

Who will control the Internet

Currently, only the U.S. has the power to grant domain names to anyone. U.S. Department of Commerce issues a statement to foreign governments, saying that the U.S. will continue to hold that power. Foreign governments think it is hegemony in cyberspace. The management of domain names used to be by one person, Professor Jon Postel at USC. In the early 90¡¯s, rapid growth of the Internet convinced the U.S. government that an organization would be needed to manage the domain names. ICANN was founded in 1998 to fill that need. Because Washington had sole control of it, there were resentments from businesses as well as foreign governments. As a result, foreign governments have been trying to set up parallel systems to compete with ICANN-sanction domains.

Read the Letter that Won the Internet Governance Battle

The U.S. will keep its historical role of controlling the Internet. The U.S. asks the European Union to reconsider its position of Internet governance and join the U.S. in its effort of regulating the Internet.

Control the Internet? A futile pursuit, some say

The dispute over the control of domain names on the Internet may lend itself to centralized governmental control and fragmentation of the network itself. ICANN, a U.S. based organization, current controls domain names, and nations have asked to make ICANN¡¯s management international. Yet designers of the Internet argue that all of them have the wrong premise that the Internet can be controlled at all. The Internet is designed as a decentralized system and, unlike telephone networks, has no center point of control. Meanwhile, in an attempt to appease other nations, ICANN is working on bringing other languages into the domain name system.

Don't give UN control over Internet

Although it seems that the Internet, as a platform spanning the whole world, should be controlled by the whole world. This is wrong. The first reason is that international talks on the control of the Internet are sponsored by the UN, and UN is evil and similar talks in the past have given us oil-for-food scandal and a child prostitution ring in Congo. The second reason is that freedom of speech on the Internet may be in danger when nations against freedom of speech join the panel of domain name management. Currently, the Internet is free from the politics of each country, and ICANN in the U.S. guarantees freedom of speech. If other countries have a say in the Internet, then it may be fragmented because of politics.

For Meeting 4

Jail Time in the Digital Age

Dimtri Sklyarov, a Russian programmer, wrote an application that helped users overcome the restrictions in the eBooks read by Adobe's eBook Reader. When he was giving a lecture in Las Vegas, he is arrested for violation of Copyright laws. The charge, from Adobe's perspective, wasthat Sklyarov's program 'could enable a pirate to copy an electronic book otherwise readable only with Adobe's reader technology, and then sell that copy to others without the publisher's permission.' This case led the Congress to enact the Digital Millennium Copyright Act (DMCA), which 'outlaws technologies designed to circumvent other technologies that protect copyrighted material.' DMCA protects copyright protection technology, which in turn protects content. The problem is that copyright laws allow certain fair use of copyrighted material, but if the material is protected by copyright protection, and the protection cannot be cracked because of DMCA, then even fair use will be forbidden. Protecting copyright protection technologies like this is giving code writers the power to make laws. In addition, America could enforce the DMCA on any one in any country. DMCA interferes with legitimate use of copyrighted material, and makes security something that no one should touch, a violation of first amendment rights. Even Adobe, which later dropped the suit on Sklyarov, realized that.

Our Case Against YouTube

Viacom sues Youtube and Google for infringement of Viacome's copyrights. Youtube cites the DMCA as a defense. The DMCA protects service providers who unknowingly store copyrighted materials and delete those materials as soon as they are pointed out, but the DMCA does not protect service providers who are making a profit out of those materials and have control over their distribution. Since youtube attracts people with its contents, and in turn attracts advertising traffic, youtube is a service provider who is making a profit. Further, Youtube is able to screen out pornographic and hate videos, so it should be able to screen out copyrighted materials from its site. This shows Youtube's ability to control the distribution of copyrighted materials. Therefore Youtube is not protected under the DMCA. It is fair to make Youtube responsible for the copyrighted materials it shares, because it is impossible for all copyright owners themselves to patrol the web for violations. Also, forcing Google and Youtube to obey the law does not stifle innovation. Rather, 'protecting intellectual property spurs new investment and the creatio nof new technologies and creative entertainment.'

DRM Protects Downlaods, But Does It Stifle Innovation?

Fritz Attaway is the executive vice president and special policy advisor with the Motion Picture Association of America (MPAA). He talks with Wendy Seltzer, a professor of Law at Brooklyn Law School and a staff attorney with the Electronic Frontier Foundation, specializing in intellectual property and First Amendment issues. Attaway says Digital Rights Management (DRM) is a technology that is key to consumer choice. It allows content owners to tailor their offerings to what consumers want, while protecting the copyrighted contents. Seltzer says DRM is fine, DMCA stifles innovation around DRM-Restricted media, because it is illegal to crack DRM measures. Further, DMCA prohibits fair use of DRM-restricted materials. With DMCA in the way, the true benefits of DRM are never realized. Attaway says there is no evidence that DMCA stifles innovation, and presents a list of new viewing opportunities that have become available in the past few months. They represent new business models that music and movie studios are trying to pursue. Also, fair use is not impacted by DMCA at all. Seltzer says that she is actually not talking about producers, she is talking about users who use copyrighted materials they bought to create new things. In this regard, fair use is stifled because DMCA prevents people from creating new things with copyrighted material. Attaway says that limiting fair use is important to preventing people from making illegal copies of copyrighted contents, and preventing illegal copies is more important than fair use. Therefore, even when we need to limit fair use, we have to do it. Perhaps in the future, DRM technology may be sophisticated enough to allow content owners to protect their work as well as viewers to make transformative uses of the content. Seltzer says that, as DRM gets more sophisticated, old devices won't work. This kind of technology is brittle and fails easily.Attaway says that technology is still needing improvement. DRM technology does not stop pirates, but it keeps 'honest people honest'. Seltzer says that because of DRM, the hardware industry cannot innovate, because it needs to get approved and pay major studios before it can try out new ideas. There won't be iPods for movies and TiVo for radio. DRM is a door that locks out the owner all too often. When people realize that, they will repeal the laws supporting DRM.

Call It The Digital Millennium Censorship Act. Unfair Use

Invoking the DMCA, Microsoft demands that Slashdot remove its posts about Microsoft's Kerberos specification (not public), because people circumvent the end-user-agreement in order to view the specification. DMCA makes it illegal to defeat copyright protection mechanisms, and allows copyright owners to notify service providers of infringing material. Slashdot calls what it has about Microsoft's specification fair use. Microsoft says that even talking about how to circumvent the end-user-agreement is in violation of the DMCA. Another law, the Unfirom Computer and Information Transactions Act (UCITA), copying things that belongs to fair use will still breach copyright law. DMCA and UCITA are a danger to fair use and freedom of speech. Traditionally, copyright law 'forbids authors from controlling the uncopyrightable ideas or functional principles embodied in their work and allows others to make 'fair use' even of copyrightable expression for purposes such as criticism, comment, education, and research.' DMCA bans circumvention of security technology, UCITA bans criticism. UCITA must be unconsittutional. Also, DMCA and UCITA together will allow companies to make lousy products, because they are shielded from criticism. This is not healthy for a competitive market. The two laws also allow companies not to abide by standards.

For Meeting 5

For Meeting 6

Think Before You Share

Many young people nowadays do not think about the risk when they post their identities on websites such as MySpace and Facebook. Websites like these are becoming places where interested individuals can find out details about one's life that should better be private. Taylor Behl, a student in Virginia Commonwealth University who is an active blogger on Mysqpace and LiverJournal, revealed much of her personal details through her writings online. She went missing in August 2005 and her remains were found later. She was murdered by her friend and sex partner. Although her murder was not directly connected to her life on the Internet, after her death people began to be aware of what they put online.

Virgina Commonwealth's students, as it turns out, don't know the dangers of putting their personal information online. For this matter, the school has decided to give tutorials on privacy on the Internet to each incoming class before people have a chance to get their facebook accounts. In a way, websites like Facebook are making the Internet a more dangerous place for young people. With Facebook also opening for high school students, protecting youngsters will become even harder.

To protect users' privacy, Facebook has implemented controls that allow people to show their personal information only to people they normally trust. Some people think that these measures actually defeat the purpose of Facebook, and never use them. There have been incidents in which women are harassed by men who found their information on Facebook. Police officers in some colleges use that to catch people doing unlawful things such as drinking and intoxicating. Employers also go on Facebook to find out things about their candidates. In short, people want to be more cautious about what they put on the Internet from now on.

Facebook's "Privacy Trainwreck": Exposure, Invasion, and Drama

Facebook's News Feed feature keeps everyone on track with his friends by announcing everything his friends do on facebook. Facebook users think it's an invasion of privacy. There are two ways of defining private information. In the digital world, a 0 or 1 can determine whether something is private. In the human world, private information is defined as the information about us that, when others know, we will feel icky. People do not take Mark Zuckerberg's defense that whatever is on News Feed is public information anyway, and that News feed just aggregates that information. This is because News Feed, despite it only displays public information, makes people feel icky.

There are important ramifications that news Feed brings about. By telling you someone's life, News Feed makes you think someone is your friend without any in-person contact. This is bad. By giving you updates on other people's lives, Facebook News Feed confuses online friendship with traditional friendship. It makes you think a Facebook friend is an actual friend. Just as you judge your friend by his actions, you judge your Facebook friend by his edits on Facebook. Yet since the two ideas of friendship are criss-crossed, you also judge that person by his edits on Facebook. If A writes on friend B's wall but not friend C's wall, then C's natural conclusion is that A likes B more than C in real life. If someone joins a Facebook queer group, he maybe thought as a queer. Although there won't be any misunderstanding when one knows well the person he is friends with on Facebook, we are not acquainted with most of our Facebook friends. Therefore our judgment of them is based on the News Feed, and this leads to misconception about people.

Also, the idea of friendship leads us to help friends in need. When one knows that his friend is in trouble, he tries to help, and blames himself when he fails to help. Although friends over cyberspace are less than actual friends, News Feed gets us confused by providing so much of their life details, that we are convinced they are actual friends. Thus when those cyber friends are in trouble, our humanity tells us we should help and blames us if we fail to help. However, nothing really can be done, because we never had in-person contact with those people. All we can do is to mourn their downfall and our inability.

After a Redesign, Facebook Hastily Offers New Privacy Protections

On September 5th, 2006, Facebook offers News Feed. For each user, it compiles a list of changes every friend in his list has made. Users think that feature is "stalkerish". The incident led people to be aware of their privacy online.

Beware of Using Social-networking Sites to Monitor Students, Lawyers Say

Facebook is sometimes used as a tool for law enforcement in universities. (Emails too, but Facebook is so much easier) The Law is unclear on whether people can do that. In similar cases, the Courts have held that students are responsible for what they put on the Web, but college officials should avoid the temptation to snoop. Many colleges have agreed not to look at online postings unless parents contact them about specific incidents.

For Meeting 8

It's the Economy, Stupid

Behind the decisions affecting computing security, there is economics. We need economics to tell us whether we are doing the right amount of work to deter hackers and to police the network. There are examples where economics does not encourage any more security. Encrypted emails are not common because it's not worth it. Insecure operating systems are produced because their behaviors satisfy the customers. Information security such as copyright is also part of an economic argument.

Spyware on My Machine? So What?

Some people spyware is a fair trade-off for free applications. iMesh, a file-sharing application, comes bundled with an adware caled Marketscore, which has the ability to view encrypted contents such as passwords and creditcard numbers. Yet users of iMesh are not particularly concerned by Marketscore, saying that supporting spyware is the only way to support free software. Schools like Columbia and Cornell block access to computers with Marketscore installed in an attempt to protect students, but students think that is annoying.

On the other hand, marketscore strives to satisfy the customers it taps into. It offers an email anti-virus service, which checks emails for viruses before they arrive at the users' computers. Customers like the feature, and don't mind that people may be reading their emails. In addition, marketscore says it only collects information it says it collects, but some worry that it may change its terms in the future. Still, even companies don't mind have spyware on their computers. Computers at a Fortune 500 Firm is infected with Gator, and the Vice President says he wants the spyware is a necessary evil for the eWallt application they want. Security experts say that people like this have given up notions of security by convincing themselves that it is a ok, but those people say that they are tired by security threats and decide that too much security is bad.

Why Spyware Poses Multiple Threats to Security

Spyware can collect information about an individual, install trojans, capture keystrokes, launch DDOs attack, probe system vulnerabilities, all without the user's knowledge. In addition, spyware negatively affects computer performance, and can be used by terrorists to launch DDoS attacks. The solution is to install anti-spyware, make laws to require that, while installing any application, every component to be installed must be made apparent, punish spyware makers, and plan with other countries to minimize the possibility of a DDos attack.

For Meeting 9

Tussle in Cyberspace: Defining Tomorrow's Internet

The Internet today is a place where competing interests collide. This process is called the Tussle. Competing interests are different players in the Internet milieu. For example, music lovers like to share music with one another, but copyright holders want to stop them. ISPs are interconnect, but they are fierce competitors. Competition suggests new requirements for the Internet and motivates new design strategies that accommodate the growing tussle among Internet players.

As a feat of engineering, the Internet was born to be predictable and yield desirable behavior. When it became a popular phenomenon, it became a mirror of the societies in which it operates. It was no longer predictable. The Internet today is like a society. It has several players: users, commercial ISPs, private sector network providers, governments, intellectual property rights holders, providers of contetn and higher level services. The future of the Internet will be defined by the tussles that arise among these players.

The future Internet should modularize tussle space by separating functions that are within the tussles space from functions outside the space. DNS is an example of a tussle space. A domain name is used to express an address on the Web and trademark. There are fights over domain name because of trademarks. To modularize DNS means it should not use as a domain name a name already used for a trademark, separating the correlation between domain name and trademark. This way, if fights over trademarks will not affect domain names, and vice versa. The future Internet should be designed for choice. For example, a user of mail program can select his POP3 and SMTP server. Similarly, an email service provider can control what SMTP server a customer uses by redirecting packets based on the port number. The drawback of choice design is the complexity of configurating and using a service.

A second tussle space is economics. Users want to switch from one service to another. This involves changing IP addresses. The Internet can be designed to facilitate such changes. A third tussle space is pricing. It is not a moral wrong for producers to charge customers on the basis of Internet usage. People who pay more get faster broadband access, people who pay less get slower. The Internet cannot avoid such value pricing. Some people who think this is discrimination want to bypass control by producers. It is not clear what the Internet should do. A fourth tussle space is providing for better residential broadband access. A fifth tussle space is competitive wide area access. A sixth tussle space is the control of security schemes. Who puts up a firewall? Users may do that, but should service providers put up a firewall? A seventh tussle space is whether we want to know to whom we are communicating over the Internet. The future Internet will add more substance to the end to end argument.