Netizens-Digest Tuesday, April 15 2003 Volume 01 : Number 498 Netizens Association Discussion List Digest In this issue: Re: [netz] Bad Things to Avoid, whether done "commercially" or not Re: [netz] Bad Things to Avoid, whether done "commercially" or not [netz] No tragedy but good model Re: [netz] No tragedy but good model Re: [netz] No tragedy but good model Re: [netz] No tragedy but good model Re: [netz] No tragedy but good model ---------------------------------------------------------------------- Date: Mon, 14 Apr 2003 20:18:32 -0400 (EDT) From: lindeman@bard.edu Subject: Re: [netz] Bad Things to Avoid, whether done "commercially" or not Howard, > Seriously, do consider how business-to-consumer sales can empower the > small business. Oh, I didn't intend to dismiss that point (although I don't have much to say about it). It's other points about sustainable ag that I'm setting aside. OK, things we are avoiding: > > > 1. price gouging on "basic" access that perpetuates a digital divide. ... > Sorry. I was thinking of what might or might not be an analogy, > depending on the level of integration between data and telephony. > Calling the cops is a basic _telephony_ function. Let's just assume for the sake of argument that I'm analogy-impaired (unless the analogy has something to do with terraforming Mars). Can we characterize basic Net access? [Come to think of it, one reason the cop-calling analogy confused me is that as I understand it, out-of-contract cell phones can be used _only_ to call emergency services. A nice feature, but not really "basic access."] > At the residential level, however, integrating voice, video, and > telephony to reduce costs makes sense. OK, plausible. > >> 2. restrictions on information offered by private individuals, unless > >> that information violates law.[...] Come to think of it, I'm not sure whether you mean to contrast "private individuals" here with "public" below. > >> 2.1 The provider is entitled to characterize statistical user > patterns, > >> and distinguish the traffic volume associated with a public web > >> site, > >> for example, from a web/mail user. The provider may charge > >> additional > >> fees for the resources consumed by a public facility. OK, if the point is to charge high-bandwidth users more, shouldn't you frame it that way, instead of specifying resources consumed by "a public facility"? Or am I misunderstanding your meaning of "public" here? Maybe you mean "public facility" = "public web site" = any web site available to Net users at large? And if that's what you meant, then your reference in #2 perhaps refers to information that is _not_ available on public web sites, such as the contents of e-mails? I'm not sure I see a rationale for the private/public distinction even so. Mark ------------------------------ Date: Mon, 14 Apr 2003 21:56:29 -0400 From: "Howard C. Berkowitz" Subject: Re: [netz] Bad Things to Avoid, whether done "commercially" or not Mark wote, >r >... >> Sorry. I was thinking of what might or might not be an analogy, >> depending on the level of integration between data and telephony. >> Calling the cops is a basic _telephony_ function. > >Let's just assume for the sake of argument that I'm analogy-impaired (unless >the analogy has something to do with terraforming Mars). Can we characterize >basic Net access? > >[Come to think of it, one reason the cop-calling analogy confused me >is that as >I understand it, out-of-contract cell phones can be used _only_ to call >emergency services. A nice feature, but not really "basic access."] It's worth keeping in mind though -- emergency services access is one of the things that usually is considered a "lifeline" telephony function. I don't have a good parallel for data use, but that doesn't mean there isn't one. Perhaps access to some (undefined) community server, where the user participates in the grassroots political process, is basic access. Depending on whether or not there are significant bandwidth costs, the access to that server, hypothetically, might be minimal text in "basic" mode and full Flash-style graphics only in "enhanced payment" mode. > >> At the residential level, however, integrating voice, video, and >> telephony to reduce costs makes sense. > >OK, plausible. > >> >> 2. restrictions on information offered by private individuals, unless >> >> that information violates law.[...] > >Come to think of it, I'm not sure whether you mean to contrast "private >individuals" here with "public" below. Maybe we need to start over here. Let's start with the definition of a public server, web or not. It has a name and associated address propagated through the Domain Name System, and it is "always on" as opposed to being operational only when its human operator is generating information. The "always on" property will cause additional traffic, in the form of queries from web-crawling bots and other indexing tools, to be attracted to it. Always on servers follow the "pull" model of information access. If they have any level of visibility, they will tend to have multiple concurrent accesses, as to the access pattern of a client to an email server. So, they tend to require more bandwidth than clients, even before considering the particular information they present. Their total bandwidth is, before getting into some exotic technical tweaks, the product of the number of users times the amount of information transferred per user. >And if that's what you meant, then your reference in #2 perhaps refers to >information that is _not_ available on public web sites, such as the contents >of e-mails? I'm not sure I see a rationale for the private/public distinction >even so. You're confused? I'm confused. Trying again, email is basically an exception-driven (i.e., the user writes something) client to server interaction. Ignoring mailing lists for the moment, the user's communication is for a specific destination -- a push model. Mailing lists are also a push model, but have greater bandwidth -- the bandwidth may be identical to a single message between the user and his post office equivalent, but have greater overall net impact. Mailing lists are not completely "public" in that you have to locate their postmaster and register before using them. Historically, there have been "open relays" for mailing lists that do not require preregistration, but these have become blatant spam magnets and are routinely blocked by service providers. ------------------------------ Date: Tue, 15 Apr 2003 00:24:28 -0400 (EDT) From: Jay Hauben Subject: [netz] No tragedy but good model Hi, Howard asks: >>How do you avoid the tragedy of the commons? This myth of the tragedy of the commons flies in the face of all historical evidence. For thousands of years and still today one of the most stable economic forms is the commons. People who benefit from a common holding take responsibility to insure its continuity. Throughout history, commoners have collectively generated and modified and enforced the rules they expect each other to follow and the penalties for deviations from these rules. In general and for most of history, the integrety of each commoner has lead to the stability and success of the commons. But in addition each commons has found formal and informal means of enforcing those rules. When Garret Hardin coined the phrase 'tragedy of the commons' he was not basing it on historical facts but instead on the mistaken assumption that humans are basically calculating, economic entities. He and those who have echoed him since see people as basically greedy or selfish. The advoctes of the basic selfishness of humans are found in all fields. But this selfishness as a main aspect of human behavior is ideology not scientific fact. The commons is a proper model for the net as Michael realized in the early 1990s. The strength of the netizens list is that in general the people who gather here share a belief in the basic goodness of people. That conviction helps strenthen us against falling for the arguements of those who assume people are basically selfish. Take care. Jay ------------------------------ Date: Tue, 15 Apr 2003 00:55:26 -0400 From: "Howard C. Berkowitz" Subject: Re: [netz] No tragedy but good model >Jay wrote, >Howard asks: > >>>How do you avoid the tragedy of the commons? > >This myth of the tragedy of the commons flies in the face of all >historical evidence. For thousands of years and still today one of the >most stable economic forms is the commons. I agree, and agree this was the original model for the Internet. But how do you explain the very real problem of those who release viruses and create denial of service attacks? Distributed denial of service attacks (http://www.cert.org/tech_tips/denial_of_service.html) are hardly something that promoted the commons. There is a distinct subculture of hackers who gain social status in their peer group by acquiring resources or denying them to others. For example, see Bruce Sterling, _The Hacker Crackdown: Law and Disorder on the Electronic Frontier_, or the archives of network operator forums such as www.nanog.org. Also, the design of the main Internet transport protocols, TCP and UDP, do not operate on a "fairness" model. TCP automatically tries to acquire all the bandwidth it can until it encounters packet loss due to congestion, at which point it slows its transmission rate. One might say, "if it slows down, where is the problem?" The reality is that in a resource-sharing network, the earliest-to-start TCP instances, and those that are closest to their destinations, tend to gain an unfair share of bandwidth. The reason they do so is that when additional instances start, they immediately encounter congestion and keep their transmission rates slow. The problem of TCP bandwidth seizing is unsolved, and probably can't be solved within the protocol itself. Current thinking is more to restrict the number of sessions that can be started at the network edge (connection admission control, analogous to not letting phone calls start when the PSTN has no capacity for them). There is some research on information-sharing between TCP sessions, but this is a niche that does not scale to the size of the global Internet. TCP, at least, does have a negative feedback mechanism to slow its transmission rate. Some vendors ignore this feedback mechanism to make their implementation "look better" in tests, but the majority of implementations do honor the feedback. UDP, however, has no mechanism for feedback about congestion. The protocol itself has no restraint on how fast it sends. In full protocol stacks, it _may_ get information from higher-level protocols that could slow the rate, but this is by no means a given. UDP-like mechanisms are the fundamental technology for one-to-many multicast content distribution on the Internet, because UDP has much lower overhead than TCP. It is technically impossible to maintain connections to every recipient in a large multicast domain; there's too much overhead for the sender. >People who benefit from a >common holding take responsibility to insure its continuity. Throughout >history, commoners have collectively generated and modified and enforced >the rules they expect each other to follow and the penalties for >deviations from these rules. In general and for most of history, the >integrety of each commoner has lead to the stability and success of the >commons. But in addition each commons has found formal and informal means >of enforcing those rules. We have very real problems with spammers abusing the Internet commons, as well as malicious hackers. Spammers often steal others' resources to amplify their effective transmission rates. Some countries have no law or custom restricting the activities of spammers and crackers. It would be excellent if the Internet had an enforcement arm to deal with the threats to the common. I invite proposals as to how that enforcement could work. > >When Garret Hardin coined the phrase 'tragedy of the commons' he was not >basing it on historical facts but instead on the mistaken assumption that >humans are basically calculating, economic entities. He and those who have >echoed him since see people as basically greedy or selfish. The advoctes >of the basic selfishness of humans are found in all fields. But this >selfishness as a main aspect of human behavior is ideology not scientific >fact. > >The commons is a proper model for the net as Michael realized in the early >1990s. > >The strength of the netizens list is that in general the people who >gather here share a belief in the basic goodness of people. Jay, I would suggest you attend a NANOG, RIPE, or other forum where real-world network operations people meet. On the one hand, you will find a group of people who generally love their work and how it benefits world communication. But you will also find -- examine, for example, the meeting archives at www.nanog.org -- the people that RUN networks believe that a significant minority of people have demonstrated they will try to destroy the Internet given the chance. This isn't theory. This is direct observation. >That >conviction helps strenthen us against falling for the arguements of those >who assume people are basically selfish. Conviction is fine. Technical rate-limiting, blacklisting sites with dangerous security holes making them spam havens, validating routing updates, balancing privacy vs. accountability (so attackers can be identified) denial of service detection, and sanctions against violators are realistic. ------------------------------ Date: Tue, 15 Apr 2003 08:38:44 -0400 (EDT) From: lindeman@bard.edu Subject: Re: [netz] No tragedy but good model Just so we're analytically clear, the idea behind "tragedy of the commons" is that each individual is ineluctably driven by calculations of interest to increase his herd (and thus his impact on the commons) without limit. Defending the commons against a minority is a different issue. I think that the popularity of SUVs in the United States represents a sort of tragedy of the commons with respect to greenhouse gas emissions, although obviously "increase without limit" doesn't apply. There's also some sort of game-theoretic tragedy, a multiple-player security dilemma, at work in SUVs' effect on safety: everyone would be safer if everyone drove well-built smaller cars, but many people believe that they are safer in a larger rather than a smaller vehicle -- even though the outcome is that most new car sales are SUVs and other light trucks. (That argument would take some sorting out, but I'm not trying to hijack the thread -- just to indicate some other dimensions.) Jay wrote, > >When Garret Hardin coined the phrase 'tragedy of the commons' he was not > >basing it on historical facts but instead on the mistaken assumption that > >humans are basically calculating, economic entities. He and those who have > >echoed him since see people as basically greedy or selfish. The advoctes > >of the basic selfishness of humans are found in all fields. But this > >selfishness as a main aspect of human behavior is ideology not scientific > >fact. I think that's basically right, although "basically greedy or selfish" is kinda vague, since it can cover both myopia and savage rapacity. > >The strength of the netizens list is that in general the people who > >gather here share a belief in the basic goodness of people. Jay, I spend much of my time studying all the way that people mess each other and the world up, intentionally and unintentionally. Now, there are also countless ways in which people _don't_ mess each other and the world up. But as a political scientist and a preacher's kid, I'm a bit unnerved by talk of "basic goodness." Perhaps, differently put, we share a fundamental optimism that, with careful tending, the Net will enable more good than harm. ------------------------------ Date: Tue, 15 Apr 2003 08:58:30 -0400 From: Luis De Quesada Subject: Re: [netz] No tragedy but good model Hello Mark: Just want to put my 2 cents in about SUV's and people. About SUV's its not only gas emissions we have to worry about since all other vehicles produce them but also safety, the driver better slow down on curves or the vehicle will tip over easier than lower vehicles, also because of their height they block your lateral view in parking lots. I think their safety hazards are more than meet the eye. Like Jay I think there's good, well intentioned people everywhere and then as you go along and also find there's also bad people around, you select the good ones but never giving up on the bad ones, trying to convert them to the "good side of the force" just like Darth Vader finally did. Take care, Lou D. lindeman@bard.edu wrote: > Just so we're analytically clear, the idea behind "tragedy of the commons" is > that each individual is ineluctably driven by calculations of interest to > increase his herd (and thus his impact on the commons) without limit. > Defending the commons against a minority is a different issue. > > I think that the popularity of SUVs in the United States represents a sort of > tragedy of the commons with respect to greenhouse gas emissions, although > obviously "increase without limit" doesn't apply. There's also some sort of > game-theoretic tragedy, a multiple-player security dilemma, at work in SUVs' > effect on safety: everyone would be safer if everyone drove well-built smaller > cars, but many people believe that they are safer in a larger rather than a > smaller vehicle -- even though the outcome is that most new car sales are SUVs > and other light trucks. (That argument would take some sorting out, but I'm > not trying to hijack the thread -- just to indicate some other dimensions.) > > Jay wrote, > > > >When Garret Hardin coined the phrase 'tragedy of the commons' he was not > > >basing it on historical facts but instead on the mistaken assumption that > > >humans are basically calculating, economic entities. He and those who have > > >echoed him since see people as basically greedy or selfish. The advoctes > > >of the basic selfishness of humans are found in all fields. But this > > >selfishness as a main aspect of human behavior is ideology not scientific > > >fact. > > I think that's basically right, although "basically greedy or selfish" is kinda > vague, since it can cover both myopia and savage rapacity. > > > >The strength of the netizens list is that in general the people who > > >gather here share a belief in the basic goodness of people. > > Jay, I spend much of my time studying all the way that people mess each other > and the world up, intentionally and unintentionally. Now, there are also > countless ways in which people _don't_ mess each other and the world up. But > as a political scientist and a preacher's kid, I'm a bit unnerved by talk > of "basic goodness." Perhaps, differently put, we share a fundamental optimism > that, with careful tending, the Net will enable more good than harm. ------------------------------ Date: Tue, 15 Apr 2003 11:22:11 -0400 From: "Howard C. Berkowitz" Subject: Re: [netz] No tragedy but good model At 8:38 AM -0400 4/15/03, lindeman@bard.edu wrote: >Just so we're analytically clear, the idea behind "tragedy of the commons" is >that each individual is ineluctably driven by calculations of interest to >increase his herd (and thus his impact on the commons) without limit. >Defending the commons against a minority is a different issue. Thank you, Mark, for that clarification. The second definition is my most immediate concern. The first applies in the sense that users of a service may desire greater service, without knowing the resources required to meet those greater demands. Malicious hackers, spammers, virus developers, and script kiddies fall into the "defense" category. My concern about the second category remains--let me give a reason why it applies in today's Internet. The Web and related technologies were originally developed as hypertext -- note that "text" is part of the name of the technology. The innovation was dynamic linking and cross-referencing between documents, so entering one document could carry the user to a completely unexpected and useful destination. As the Web becomes more common, people -- and individual users are as guilty as corporate advertisers -- tend to add lots and lots of multimedia content, "because they can." This content often doesn't improve understanding but it looks pretty -- and vastly increases the bandwidth requirements for web access, enough to preclude the use of inexpensive technologies (e.g., modems) to provide interim bridges across the digital divide. I say "interim" because widely available broadband service is a trend across most new technologies, and is also implicit in delivery of converged voice/video/data. Obviously, some graphics are essential. I work with some healthcare applications where small communities use the web to deliver cardiac ultrasound imagery to university specialists for review. But the widespread use of graphics in web pages is a drain on the bandwidth resource. Incidentally, I also find HTML-enabled email to be pernicious, both in the resource requirements and as a very real security risk -- it can be a portal to admit viruses and worms. > >Jay wrote, > >> >When Garret Hardin coined the phrase 'tragedy of the commons' he was not >> >basing it on historical facts but instead on the mistaken assumption that >> >humans are basically calculating, economic entities. He and those who have >> >echoed him since see people as basically greedy or selfish. The advoctes >> >of the basic selfishness of humans are found in all fields. But this >> >selfishness as a main aspect of human behavior is ideology not scientific >> >fact. > >I think that's basically right, although "basically greedy or >selfish" is kinda >vague, since it can cover both myopia and savage rapacity. > >> >The strength of the netizens list is that in general the people who >> >gather here share a belief in the basic goodness of people. > >Jay, I spend much of my time studying all the way that people mess each other >and the world up, intentionally and unintentionally. Now, there are also >countless ways in which people _don't_ mess each other and the world up. But >as a political scientist and a preacher's kid, I'm a bit unnerved by talk >of "basic goodness." Perhaps, differently put, we share a >fundamental optimism >that, with careful tending, the Net will enable more good than harm. I might put it a third way. If I remember things correctly, it was George Bernard Shaw who said "Some look at things the way they are and say, 'why'. I look at things that have never been and say, 'why not'". My vision of the Internet is somewhat different. I look at things that are real problems in the real Internet, and accept that stupid or evil people are causing them. I recognize that the average person has little malice, but to get, reliably, to where most people want to be, we have to maintain vigilance and safeguards against those that would prevent the good people from gaining their goals. There's an irony that there's been controversy about war-related posts. I'll simply say that if we could rely on the basic goodness of people, why would we have wars? True, it may be a minority that causes wars, but we can't go defenseless and assume that goodness will always win -- without proactive approaches. This is why I introduced specific suggestions about looking at real problems caused by real threats. It has been the fairly universal experience of network operators that if you give users the right to prioritize traffic without economic penalties, they tend to mark all of their traffic at high priority. The trivial argument is that marking email as high priority may improve its delivery by a few minutes, not really justifying the priority. If voice, real-time collaborative text applications, and video can't rely on low-delay networks, they simply won't operate. ------------------------------ End of Netizens-Digest V1 #498 ******************************