*** ANSWERMAN INTERNET EXTRA *** Sunday, May 5, 1996 Welcome! This is a weekly newsletter for the community of new Internet users on America Online. It is brought to you by AnswerMan, AOL's area for those learning to use the Internet. Come see what's new at keyword ANSWERMAN. ||| IN THIS ISSUE || | Purple's Picks - Usenet search services Upcoming AnswerMan Chats - Now on Sundays, wahoo! Answers of the Week - Mommy, Where did the Internet come from? The End ||| PURPLE'S PICKS || | by CJ Purple Hi folks! I hope you all found last week's inaugural column helpful. This week, I'd like to tell you about some Usenet search services. Usenet is the part of the Internet that contains what we all know as Newsgroups. These are online discussion groups about a myriad of topics. They differ from Mailing Lists in that the posts do not get delivered to your e-mail inbox, but rather "sit" on a server somewhere and you have to go look for them. The result is that not all Internet providers (AOL included) carry all available Newsgroups as part of their service. (Please keep this in mind when you want to join a Newsgroup and can't seem to do so using AOL. Of course, if you are sure of the existence and correct name of the group, you can always send e-mail to Newsmaster requesting that AOL add the group to its offerings.) The first service I want to tell you about is an extension of the Liszt Mailing List database we talked about last week. Liszt of Newsgroups, as it is called, is the Newsgroups database companion. It can be found at http://www.liszt.com/cgi-bin/news.cgi. It appears to be updated frequently, so that the information you get should be reliable. This database covers over 15,000 groups and uses an easy, but powerful, search tool to help you find just the group you're looking for. For example, a search on the term "sports" yielded 356 matches. Unfortunately, it seems that Liszt can only display a maximum of 150 groups. The second database for Newsgroups is the Usenet Search Service (http://www.nova.edu/Inter-Links/cgi-bin/news.pl). This is also a thorough search service and will produce matches in which the search term is a part of the description of the group, and not just part of the title of the group. When I did a search on "sports" using this service, 91 matches were returned. In comparison to both these services, the AOL Newsgroups database (Keyword: Newsgroups) will not allow a search on such a general topic as "sports." An attempt to do so will get you the error message "Too many hits. Please try a more restrictive search phrase." In some ways, this can be helpful, since it forces the user to clarify what aspect of a topic he or she is looking for, and helps avoid having to read through a long list of unwanted groups. Here are the numbers of "hits" for 2 other search terms for each of the services: "politics" - Liszt:159 (150 displayed); USS: 57; AOL: "too many hits" "genealogy" - Liszt: 22; USS: 6; AOL: 19 The last way to search Usenet is somewhat different. DejaNews (http://www.dejanews.com) allows you to search Usenet for postings *within* the various Newsgroups on a specific topic -- not just the Newsgroup names or descriptions. A search on the term "sports" using this service yielded 35,000 (!!) matches. This is an excellent service to know about and use if you are doing research on a particular topic, but don't know (or care) what newsgroups to look in. Well, that's it for this week. Remember, if you have any ideas about topics for future columns, please e-mail me at "CJ Purple". ||| UPCOMING ANSWERMAN CHATS || Chat live with AnswerMan in these hour-long sessions || Now four times a week -- check out the new Sunday chats! | Pick an interesting topic & bring your questions | Keyword: AM CHAT Sunday, May 5, 6PM EST "Using Internet Search Tools" The Internet is vast--so vast that there is almost sure to be something of related to topic you can name. The Internet is also home to more than a dozen excellent "search engines" -- tools that will help you find information. Which search engine should you use? Which one is best? Well, that depends on what you're looking for. This hour, AnswerMan will show you how to use the right tool to find what you're after. Wednesday, May 8, 9PM EST "Gopher Nuts & Bolts." Curious about "gopher"? Gopher is a very useful--but often overlooked--Internet tool. During this session, the AnswerMan will tell you everything you need to know about gopher--what it is, how to access it, how to search it, and why you'd want to. Whether you're a seasoned Web surfer or a Internet beginner, you'll learn something this hour. There is life beyond the Web--so come burrow around with the gopher. Thursday, May 9, 9PM EST "20 Internet Sites for Newshounds." Looking for the latest on the war in Bosnia...or the highlights of the Packers game? During this hour, the AnswerMan will show you the way to twenty Internet sites where you can find up-to-the-minute news from around the globe. Find out how to explore CNN's Web site, create custom "agents" to search the newswire services, grab the latest sports stats, and much more. Friday, May 10, 9PM EST "Internet Q&A" Have questions about the Internet? Can't browse the Web? Need to find a site, but don't know where to look? Need to find a friend's e-mail address, but don't know how? Bring your Internet questions--AnswerMan will help you get online, find the information you need and get surfing the 'net in no time. Sunday, May 12, 6PM EST "Making your own Web page" Back by popular demand--itching to create your own home page World Wide Web but don't have the faintest idea how? During this session, the AnswerMan will take you on a crash course on the basics of HyperText Markup Language, creating a web page, uploading it to America Online, and getting the world to beat a path to your homepage. This is the fast track to your own spot in cyberspace. ||| ANSWERS OF THE WEEK || This Week: | Every day, AnswerMan answers a question about the Internet in the Answer of the Day. (You can get this daily dose of Internet info by going to keyword ANSWERMAN and pressing the "Answer of the Day" button.) This column features some of the most popular Q&A from the previous week. This week, AnswerMan answers questions about the origins of the Net and how it works. *** Where did the Internet come from? The Internet was never truly created as an entity of its own. It is an amalgamation of many earlier networks. In 1969, the Advanced Research Projects Agency, a part of the U.S. government's Department of Defense, set up the first parts of the network that would eventually become the Internet. At the time, the network was called the ARPAnet. The ARPAnet would link the military, defense contractors and universities in one seamless computer network. A major problem with computer networks at the time was every machine on a network needed to be operating for the network to function at all. Imagine three computers connected in a row; if the machine in the middle went down (for maintenance, for instance) the first and last computers couldn't communicate. If you were the U.S. government in the middle of a cold war, this was bad. Networks of that type could never be very reliable. The ARPAnet would be the first network of its kind for many reasons--primarily because it was decentralized, with no central computer running the show. Further, if one computer on the network should go down, it was imperative that the others retain the capability of communicating. (You can imagine why this was important to the United States military, which would be more than a little disappointed should their entire network of computers be rendered inoperable by a single well-placed bomb.) The ARPAnet would need to link any number of computers and automatically reroute information should some of those computers go offline. The ARPAnet began by linking four locations: Stanford University, UCLA, UC Santa Barbara, and the University of Utah. The ARPAnet expanded to nonmilitary uses in the '70s when universities and defense-related researchers were permitted to join the network. By the late '70s, the ARPAnet was so large that its original set of standards and communication protocols could not support the growth of the network. After extended bickering and debate, the ARPAnet switched to the TCP/IP communication protocols (still in use today), which would allow further growth in the size of the network. By 1983, all computers on the ARPAnet were using TCP/IP. By 1983, it became clear that most use of the ARPAnet was for nonmilitary purposes, so it was split into two networks: one part became MILNET, a Department of Defense military-only network, and the rest remained ARPAnet, which would resume its job of connecting research sites and other nonmilitary users. The networks continued to grow. In 1987, the National Science Foundation created their own network, called NSFnet. The NSFnet would be a high-speed "backbone" network to support the burgeoning number of networked users as well as new bandwidth-intensive applications. The ARPAnet and the NSFNET, similar in structure and purpose, began to cooperate and merge. By the late 80s, the ARPANet was absorbed by the NSFnet. In the mid 80s, the National Science Foundation began to provide funding for the establishment of research and academic networks throughout the United States. It began linking those networks to the NSFnet. The same sorts of things were happening all over the world--educators, bureaucrats and hobbyists plugging their computers into networks and those networks into other networks. The NSFnet's charter was to support education and research. It was considered inappropriate to use that network for commercial purposes. Although the guidelines of what you could and couldn't do were vague, the NSFnet's appropriate use policies made it clear that for most purposes, commercial activity was forbidden. In many cases, even though it was possible to send business information from two NSFnet-linked networks, it wasn't allowed. In 1991, a group of small commercial networks created a network of their own--the Commercial Internet Exchange (CIX)--which would allow commercial use and be free of those nasty appropriate use policies. Now, commercial users were able to connect with each other quickly and legally by networking with CIX rather than the NSFnet. What this meant was commercial collaboration, technical support by e-mail, pay-for-use databases, you name it. The formation of the CIX gave yet another boost to the growth of the Internet. Now it's today and here we are. NSFnet is no longer exists--major "backbones" of Internet connections in the United States are maintained by large companies like Sprint and AT&T. Commercial activity on the Net is continuing its unprecedented growth, but that certainly hasn't hurt the scientific, educational, and research networks (which are also growing by leaps and bounds.) The Internet will continue to grow and change, meeting the needs of the people who want it, no matter what they use it for. *** Who runs the Internet? No one "runs" the Internet. There is no governing entity or business calling the shots. Remember, the Internet is a decentralized mass of thousands of smaller networks, each running with their own purpose, their own sources of income, and their own rulemakers. The Internet is more or less an anarchy. Every organization that is plugged into the Internet is responsible for its own computers. The fact that no one runs the Internet has its advantages and disadvantages. On the up side, there are no membership fees, no censorship, and no government control. Unfortunately, when something goes wrong (if an important computer goes down or another user begins annoying you), there's no central authority to ask for help. In the absence of "net cops" policing the Internet, users need to rely on their own judgments and the assistance of the system administrators at their site to solve problems or resolve disputes. Most of the time, you're on your own. The InterNIC keeps track of all the Internet addresses -- they make sure that there is only one company with an address of "yahoo.com" for instance. The InterNIC provides Internet registration services including IP address allocation and domain name registration. The Internet is guided in its growth by several organizations (loosely called the Internet technical groups) that manage it. These organizations attempt to structure the Internet while creating a minimum of restrictions. The Internet Technical Groups coordinate the Internet's basic workings--how the protocols should talk to one another, how to plan for the Net's future, and other important (but, if you ask me, dull) details of keeping the network alive. The Internet Engineering Task Force (IETF) coordinates the operation, management, and evolution of the Internet. The IETF develops and maintains the Internet's communications protocols. The IETF is a large open community of network designers, operators, vendors, and researchers concerned with the Internet and the Internet protocols. This group identifies the Internet's technical and operational problems and proposes solutions, specifies the development of protocols to solve those problems, and provides a forum for the exchange of technical information within the Internet community. The Internet Research Task Force (IRTF) examines long-term research problems and technical issues currently affecting the Internet. The task force looks at issues that will become important in five to ten years. Current issues include how the Internet will handle a billion users (a rapidly approaching landmark) and how current users will be affected when 100 million U.S. homes are wired for Internet via cable television by the end of this century. The Internet Architecture Board (IAB) is the master body for technical changes to the Internet. The IAB is concerned with technical and policy issues involving the evolution of the Internet's architecture. AIB members are committed to making the Internet function effectively and to making sure the Net evolves to meet a large-scale, high-speed future. The IAB oversees the IETF and IRTF and ratifies major changes that come from them. *** Hey, wow! I'm sitting in my living room and connected to a computer in Finland. Who's paying for the phone call when I connect to some far-off host? The answer to this question is a mixture of "nobody" and "a lot of people." To understand why this is so, you need to know a little about the way the Internet is organized. First, you have to understand that the Internet transmits the information you send by breaking it up into small pieces called packets and sending those packets to the remote machine. The difference between most networks and the Internet is that for most packet-based networks, the machine you are sending information to must be connected to the same network as the computer you are sending it from. Small networks like this are called local area networks (LAN). Internet is a type of wide area network (WAN). WANs typically consist of several LANs hooked together. When you connect to a host far away, you are not connected by a single phone or data line. The Internet works by connecting lots of little networks with a few big ones. When you communicate with a computer on the other side of the globe, or even just the other end of your state, the information passes through many networks owned and maintained by a variety of organizations. When you want to communicate with a machine that is not plugged in to your own local network, your computer needs to find a way to get the information to the distant machine. This is like trying to get from an airport in Eureka, California, to one in Helsinki, Finland. There are no direct-connecting flights (that is, no direct network connection from Eureka to Finland. Not surprising.) So your local network asks its Internet's travel agent (called a router, the machine that connects your local network to the Internet) whether it knows the way to the remote host and how many "hops" it would require to get the information there. (A "hop" is like a stopover at an airport.) One router might find a path from here to there in five hops (Eureka to San Francisco, San Francisco to New York, and so on). Your network then asks for directions from any other routers that are available. The router that responds with the fewest number of "hops" is given the message to pass along. The network serving as the router does the same thing as your local network, shopping for the shortest route to get your message to its destination. (Your message spends only a few milliseconds at each stopover, a far cry from the endless hours people spend waiting in airports.) So the information is passed from one network and computer to another until it gets where it's going. Back to the question: the only cost you're paying for is the one to connect you to America Online, and AOL is paying for a connection to some other parts of the Internet. Past that, your message uses space on several other networks owned and paid for by many other organizations. You pay a tiny bit, therefore, as does everyone else on the path of your message. Everyone pays, and no one does. Very Zen, don't you think? Using a program called traceroute, I traced the path of a message from San Jose, California, to Finland. It made the journey in 17 "hops" and 200 milliseconds! (This doesn't have to make perfect sense, but it's interesting to look at). 1 rahul.net (192.160.13.203) 3 ms 4 ms 7 ms the packet leaves wy Internet service provider... 2 sl-ana-4-S1/1-T1.sprintlink.net (144.228.74.9) 18 ms 18 ms 22 ms 3 sl-ana-1-F0/0.sprintlink.net (144.228.70.1) 24 ms 17 ms 17 ms 4 sl-fw-6-H2/0-T3.sprintlink.net (144.228.10.29) 41 ms 42 ms 41 ms 5 sl-atl-1-H2/0-T3.sprintlink.net (144.228.10.85) 59 ms 56 ms 58 ms 6 sl-atl-2-F0/0.sprintlink.net (144.228.80.2) 66 ms 58 ms 56 ms 7 sl-dc-1-H3/0-T3.sprintlink.net (144.228.10.65) 97 ms 75 ms 76 ms ...and bounces around SprintLink, a major U.S. backbone... 8 icm-dc-2b-F1/0.icp.net (144.228.20.103) 84 ms 82 ms 75 ms 9 icm-dc-3-F2/0.icp.net (198.67.131.33) 75 ms 77 ms 93 ms 10 icm-pen-1-H1/0-T3.icp.net (198.67.131.18) 493 ms 275 ms 339 ms 11 icm-uk-1-H0/0-T3.icp.net (198.67.131.26) 200 ms 236 ms 231 ms 12 icm-stockholm-1-H0/0-E3.icp.net (198.67.131.42) 170 ms 170 ms 170 ms ...then gets passed to trans-atlantic link... 13 syd-gw.nordu.net (192.36.148.205) 178 ms 185 ms 173 ms 14 fi-gw.nordu.net (192.36.148.54) 180 ms 179 ms 178 ms ...the packet lands in Stockholm, Sweden... 15 lime-gw.funet.fi (193.166.5.6) 186 ms 183 ms 183 ms 16 oliivi-gw.funet.fi (193.166.5.42) 202 ms 201 ms 200 ms ...continues to a Finnish University network... 17 tolsun.oulu.fi (130.231.96.16) 210 ms 205 ms 198 ms .....and ends up at the destination computer in Finland. Then a confirmation packet is sent back to my system. Total round-trip time: about 200 milliseconds! ||| THE END || | That's all for this week. Please send any comments or suggestions about the AnswerMan Internet Extra to "Savetz" (from AOL,) or "savetz@aol.com" (from the Internet.) To unsubscribe frow this mailing list, send an e-mail message -- To: LISTSERV@LISTSERV.AOL.COM Subject: UNSUBSCRIBE Body: UNSUBSCRIBE ANSWERMAN This newsletter copyright 1996 by America Online and Kevin Savetz. All rights reserved. Made from 100% recycled electrons.