A Beginner's Guide to the Internet

home *** CD-ROM | disk | FTP | other *** search

/ A Beginner's Guide to the Internet / INTERNET.ISO / text / guides / intro / intro.011

Wrap

Text File | 1996-05-06 | 47.6 KB | 917 lines

Introduction to THE INTERNET -by David B. O'Donnell Welcome to the Internet! This handout is intended to complement the brief introduction I am giving you today. Information in this document will be in more depth than our talk and will cover topics I don't have time to discuss with you today. I hope that you will use this handout as a kind of "tourist's guide" to the services and power of the Internet. Throughout this handout you will see instances of the symbol to the left. The sections it appears next to cover topics I consider to be fairly "advanced". They won't be covered in any depth in our talk and may not be discussed at all. Consider them future challenges for you to get out there and plumb the depths of the world's largest collection of networks! The symbol to the left of this paragraph introduces a section of the handout which will be covered during the Internet talk. It is also used to indicate when 'advanced' topics are over. Organization This handout is organized into seven sections as shown below: Section One: Introduction Section Two: E-mail Section Three: Files Section Four: Traveling Section Five: Other Services Section Six: Notes Section Seven: Index to Topics The primary thrust of the talk will be on sections One and Two, with some brief stops in the remaining sections. Section Six is actually a blank page for you to jot down notes. Contacting the Author If you have questions that this handout and the discussion don't cover, or you would simply like to get in touch with me, send RFC 822 compliant e-mail (see Section Two) to one of the addresses below: atropos@netlab.cis.brown.edu el406006@brownvm.brown.edu Be sure to include your name and a subject which is indicative of the mail-I occasionally get so deluged with e-mail that even I have to prioritize, and personal mail ranks highest after my LISTSERV management tasks. Resources Some of the information in this handout comes from !%@:: A Dictionary of Electronic Mail Addressing and Networks, Second Edition, by Donnalyn Frey and Rick Adams. This "dictionary" is published by O'Reilly and Associates, Inc., located at 632 Petaluma Ave, Sebastopol CA 95472. The publishers can be reached via e-mail at: nuts@ora.com Many other sources of information were used; I would like to thank all of them for their indispensable assistance and information. Thanks particularly to the membership of the IETF (Internet Engineering Task Force) LISTSERV forum, located at the Information Sciences Institute of the University of Southern California, Marina del Rey, for their speedy reply to my e-mail inquiries. Thanks also to Peter DiCamillo of Brown University for pointing me in the direction of IETF, and to David Bogartz of Ziff Desktop Information for his insightful suggestions and Macintosh information. Information on gopher was obtained, over gopher, from the University of Illinois at Urbana-Champaign. BITFTP information comes from the BITFTP service at Princeton University. Special thank-yous go to Mark Lottor, Ole Jacobsen, Alex Bochannek, Frank Solensky, Donald Eastlake III and Sheryl Frez for their data and graphics, which made my numbers more accurate and gave me a "bigger picture" on the scope of the Internet. Information on FIDOnet was gleaned primarily from Gene Raymond, to whom I owe a debt larger than can be expressed here. STAR TREK: THE NEXT GENERATION and STAR TREK: DEEP SPACE NINE characters are registered trademarks of Paramount Pictures. I would also like to thank my friends Julie, Carole and Yvonne for providing much-needed editorial comments. Finally, much of this handout is based on personal experience of the author. Any errors are, naturally, the fault of the author only. Publication Specifications This handout was produced in Microsoft Word for Windows*. The cover page, footer graphic and "advanced" and "topical" graphics were created in Corel DRAW!* and imported into Word as 300-DPI TIFF images. Body and footnote text for this handout is set in Monotype Calisto. Paragraph headers and figure/chart text are Univers; section heads, title and drop cap are Monotype Bodoni Bold Condensed and "computer" text is Univers Condensed. Section One: INTRODUCTION Who Am I? You might be asking yourself, "Just who is this David O'Donnell and why is he going on about the Internet?" Well, I have been working and playing with computers since the early 1980's, and I've been traveling the world networks since 1986. I have several accounts which have access to the Internet; own and manage four Internet e-mail conferences while participating in over a dozen more; and maintain a healthy presence on the FIDO and RIME BBS networks. The Internet: 'No Matter Where You Go, There You Are.' Although its name-the Internet-implies one gigantic network, in fact the Internet is a loose collection of networks which literally spans the globe. CREATION. The Internet did not have a specific date of birth: in 1969 the United States' Defense Advanced Research Projects Agency (DARPA) started an experimental packet-switched host-to-host network called ARPANET. This network expanded to include military sites (MILNET) and other non- classified government networks. In 1982, ARPANET joined with MILNET and others and "the Internet" was born. Today the Internet has connections in the United States, Canada, Mexico, Central and South America, Europe, the former Soviet Union and client states, Australia, the Pacific Basin and Asia. Currently the Internet has links in over 40 nations with over sixty thousand sites. Gateways to other networks such as BITNET, JANET and EUnet make connecting all across the world virtually seamless. GROWTH. The Internet is growing at a phenomenal rate. Today technology allows virtually anyone with a personal computer and modem to connect to one of the dozens of "sub" Internets via e-mail. It is becoming common for the more powerful personal computers and workstations to have "direct connections" which allow other services such as FTP and TELNET (see Sections Three and Four). It is estimated that nearly two million computers are connected to the Internet and its subnetworks and that over nineteen million people have accounts on hosts with some kind of Internet linkage. Excerpts from the January 1993 Internet Domain Survey. (Produced quarterly by the Network Information Systems Center at SRI International; contact Mark Lottor mkl@nisc.sri.com for more information.) Jan. 1993 Oct. 1992 Jul. 1992 Apr. 92 Jan. 92 Change Hosts: 1 313 000 1 136 000 992 000 890 000 727 000 80.6%+ Domains: 21 000 18 100 16 300 20 000 17 000 23.5%+ To ping every host on the Internet would take 37 hours assuming a reply time of 100 milliseconds. Given an average of ten users per Internet host, there were approximately thirteen million people "on" the Internet in January 1993. If every user were registered with network name servers, the birthday daemon would have to deliver some 35,600 e-mail announcements per day. For more information, check out the following directories on host ftp.nisc.sri.com: pub/zone pub/zone/iso-country-codes For information on the census program used to create the Domain Survey, see technical report UCSC-CRL-92-34 located on the host ftp.cse.ucsc.edu. Section Two: E-MAIL Basic Information Everyone is familiar with the idea of mail, and e-mail is probably nothing new to you-after all, e-mail is only an extension of the familiar, right? E-mail is more than just an extension, however. It's practically a new medium, only vaguely related to "surface" mail. Briefly put, e-mail is an electronic document transmitted from a sender to one or more recipients. It is composed of two basic pieces: a header and the body. These are characteristics shared by all e-mail in existence today. Beyond this preliminary description, however, similarities can vanish. One of the most interesting features of the Internet is that virtually every personal computer, minicomputer and maiframe can connect to it in one fashion or another. There are easily a dozen operating systems in use on the Internet: VM, VMS, MVS, ULTRIX, DOS, UNIX, NeXTSTEP, OS/2, Macintosh System , and so on. Nearly every operating system has its own e-mail style, as will be covered later in this handout. To overcome the mess of competing "standards", the Internet had adopted a particular format for e-mail. Based on the RFC (Request for Comments) 822, it is called RFC 822-compliant e-mail. Requests for Comments are documents drawn up by scholars, researchers and computer scientists in an attempt to help provide working standards for the Internet. Many file servers on both the Internet and BITNET maintain repositories of some or all of these documents. If you know of a particular RFC you are interested in getting hold of, check out these sites: educom.edu BITNIC Another good way to search is to use the archie service (see Section Three) and search on either RFC or RFCxxxxx, replacing xxxxx with the number of the document you're seeking. RFC 822 Compliant Addressing Don't let the imposing name deter you from exploring further! As explained above, all e-mail consists of two parts: a header and the body. The header contains a series of informative lines which tell the mailing system where to deliver to mail and provide basic memorandum-like information for the sender and recipient(s). The body generally consists of free-form text. However, technological advances are making it possible to imbed graphics, sound, and even full-motion video in the body of e-mail. Some computers, such as the NeXT machine, already have this capability. THE HEADER. If you imagine e-mail as a form of advanced memoranda, it's easy to come up with the basic information needed in an e-mail header. The header lines consist of a tag which identifies the information and the information itself. All tags end in a colon and are followed by white space: one or more spaces or TABs. Tag information is text which may or may not be in a particular format, depending on the tag it applies to. RFC 822 allows for "wrapping" on lines whose contents exceed the informal eighty-character line length standard; wrapped lines are identifiable by the first character on subsequent lines being white space. Date: Friday, 26 March 1993; 22:18:45 EST To: atropos@netlab.cis.brown.edu (David B. O'Donnell), EL406006@brownvm.brown.edu From: dr_babe@ds9.starfleet.net (Dr. Julian Bashir) Subject: Failed mail to user foo@bar.com? Cc: postmaster@bar.com In fact, the text above does constitute a fully legal Internet mail header, even though the originator and carbon copy recipient addresses are fakes. The absolute minimum tags required to address e-mail on the Internet consist of the Date:, To: and From: tags. All others are unnecessary though in practice a great many more are used. Internetters being the industrious folk that we are, the list of frequently used header items is much larger. In fact, it would not be out of the ordinary to receive e-mail where the header lines number in the dozens and the header tags (the words suffixed with colons) are numbered in the tens. Some Additional Common E-mail Header Tags: Received: Return-Path: Full-Name: Mailer: X-tag: Resent-tag: BCC: Message-Content: X-400-Address: Comment: Reply-To: In-Reply-To: Organization: Message-ID: Sender: In general, one can include almost any kind of information in the e-mail header. For example, it is not uncommon to see a tag called "X-Face:" which contains a compressed graphical image some mail systems can decompress and display at the user's terminal. Under RFC 822 any tag which begins with "X-" is considered an extension to the listed set and can safely be ignored by e- mail processing systems. THE BODY. The body of e-mail is separated from the header by exactly one blank line. The RFC 822 specification does not state what format the body information must appear in, but the vast majority of e-mail on the Internet today consists of eighty-character-wide lines of ASCII text. ADDRESSING E-MAIL. To address e-mail it is necessary to know the recipient's e-mail address. Similar to the postal address we all have, an e-mail address is a means of identifying where on the Internet a person can be contacted. For example: I can be reached by several addresses, including atropos@net- lab.cis.brown.edu; I consider it to be my primary e-mail address. Addresses are divided into two parts: the userID and the host domain (or domain name). Unless the recipient of your e-mail resides on the same Internet host as you, it will always be necessary to specify both the userID and host domain when sending a piece of e-mail to another person. The userID is fairly straightforward: it is a collection of letters and possibly numbers which is used by the host to reference the user. In the case of my e-mail address, 'atropos' is my userID. The composition of the host domain, on the other hand, is somewhat more complex. The host domain string is a hierarchical assemblage of components which-if properly constructed-should identify just where in the Internet a user is coming from. The host domain proceeds from left to right in order of increasing generality of location. In other words, if we dissect the domain name from my e-mail address (netlab.cis.brown.edu) we see that I am in the educational domain (edu), am located at Brown University (brown), and that the computer I log in to, netlab, is part of Computing and Information Services (cis). Realistically speaking it is fairly unusual for a host domain string to be that clear to human beings-and I am aided in this dissection by the fact that I already know all this information-but it is generally possible to glean at least some information about the user from the domain name portion of their e-mail address. Because the networks that historically began the Internet were located in the United States, most sites in the US, Canada and US overseas installations follow a breakdown into six generic domains: * EDU: These are educational sites, such as universities (example: brown.edu, Brown University) * COM: These are commercial sites, generally large corporations with research facilities attached to the Internet (example: lotus.com, Lotus Development Corporation) * GOV: These are non-military government sites (example: nasa.gov, the National Aeronautics and Space Administration) * MIL: These are military installations (example: wsmr-simtel20.mil, White Sands Missile Range) * ORG: These are non-commercial, non-network sites and gateways (example: fidonet.org, the FIDOnet*Internet gateway) * NET: These are other networks to which e-mail must traverse a gateway (example: uunet.uu.net, the UUCP*Internet gateway) The "rest of the world" uses a domain-naming standard which makes use of two- character country abbreviations (de for Germany, se for Sweden, uk for the United Kingdom, et cetera). Within the domains the hierarchical schemes can vary but the most common variation is to include a US-style identifier for the type of installation the user is at. For example: j_q_user@captain- kirk.yoyodyne.usydney.edu.au identifies someone at the University of Sydney in Australia. It is becoming more common in the US for sites-particularly small ones-to adopt this naming strategy as well. Therefore someday you may well be able to send e-mail to bill_clinton@white-house.wash.dc.us. Common E-mail Programs and Platforms The following are some of the most common platforms and operating systems which can send and/or receive Internet mail; also given where applicable are the common e-mail system(s) for each entry. UNIX. UNIX systems are particularly common on the Internet. Many "pay-as-you- go" and "free" Internet hosts use UNIX as their operating system. UNIX comes with an e-mail system called, appropriately enough, mail. It can send and receive Internet e-mail and UUCP mail, the original built-in mail network. A more useful e-mail program is available on nearly every UNIX system, elm. Elm uses a menu-driven interface and allows users to set up aliases for frequent e-mail recipients, specify the text editor of their choice, maintain a simple calendar and reminder system and categorize e-mail into 'folders'. VMS. VMS systems are also fairly common on the Internet and there are several sites where access can be purchased. DEC calls its e-mail technology "DECnet"; although the name implies a network, this is not the case; DECnet is only the technology. VMS also comes with a built-in e-mail system called MAIL. While VMS systems can send and receive Internet mail, it is not always easy to do so using the internal mailing system. The VMS mail system can be configured locally for Internet (and BITNET) e-mail delivery. The most common method of addressing mail to a non-DECnet (e.g., non-VMS) system is as follows: network%"userID@host-domain" For example, to send e-mail to my netlab account you would enter in%"atropos@netlab.cis.brown.edu" for the recipient address, while sending e-mail to my BITNET account would entail entering bitnet%"el406006@brownvm" for the recipient address. There is another e-mail package available for VMS called "MM" which uses directly-RFC 822 compliant addressing. The interface for MM is also nicer than that of VMS MAIL, particularly when the user's terminal can support VT-220 or greater emulation. COMPUSERVE. CompuServe Information Service began offering Internet e-mail capabilities in 1989. CompuServe's Internet domain name is compuserve.com. It is important to note that CompuServe's Internet e-mail gateway will only accept messages under 64Kb in size. For personal e-mail this is generally sufficient, but it could cause troubles if you attempt to send or receive UUENCODEd files. To send e-mail to someone who is on the Internet, you will need to know their Internet e-mail address. In Easyplex, you format the recipient address as follows (assuming you were going to send e-mail to me): >Internet:atropos@netlab.cis.brown.edu To send e-mail to someone on CompuServe, convert their CompuServe numeric ID as follows: 72241,544 * 72241.544 and address the e-mail to id@compuserve.com. OTHER SERVICES. There are several other commercial services which have or will be adding Internet connections: Service Name Domain Name Prodigy (unknown) Delphi delphi.com GEnie genie.geis.com America OnLine aol.com MCI Mail mcimail.com AT&T Mail attmail.com While I have heard rumors that Prodigy will be establishing an e-mail gateway to the Internet, I have not received any definite information to confirm or deny this. Also, the domain name for GEnie Information Service may have changed to genie.com. Other Networks There are three networks which are quite solidly established and which do not (by default) use RFC 822 compliant e-mail: UUCP, BITNET and FIDOnet. Sites on BITNET and machines which have UUCP network connections can use RFC 822 compliant addressing, but internally the addressing system is different. BITNET. The BITNET (Because It's Time Network) was formed in 1981 as a consortium of educational sites in the United States, Canada and Western Europe. BITNET is a store-and-forward network (see page 3) whose network technology is based on that of the venerable IBM RSCS (Remote Spooling Communications Subsystem). Internally, host names are limited to eight characters and while non-IBM systems may have userIDs which are in excess of eight characters, the network will generally not understand anything beyond eight. The basic format for e-mail within the BITNET varies with the host operating system. The two most popular are IBM VM/CMS and DEC VMS which use the following formats: username AT sitename (VM/CMS) hostname::username (VMS) Conveniently, it is becoming more popular for BITNET sites to maintain Internet connections as well. However, if you do not know the RFC 822 address for someone on the BITNET to whom you wish to send mail, you can send it via one of the sites which maintains connections to both Internet and BITNET: brownvm.brown.edu, mitvma.mit.edu and cunyvm.cuny.edu being most well-known. For example, to send e-mail to user ENSIGNRO AT NCC1701D, you would address e- mail to: ENSIGNRO%NCC1701D@brownvm.brown.edu Because brownvm.brown.edu resides on both BITNET and Internet, it would know how to route the BITNET address 'encapsulated' in the userID. UUCP. The UUCP network is a haphazard collection of computers running UNIX or UUCP-compatible e-mail systems (many PCs and BBS systems do this). Another store-and-forward network, until fairly recently it was necessary to know virtually the entire network path a piece of e-mail would take in order to route it. Lately, however, UUCP network 'administrators' have been encouraging both the adaptation of Internet-style addressing and a more efficient network pathing structure. UUCP e-mail is somewhat similar to the DEC DECnet addressing convention in that host computers precede the userID in the address: host1!host2!host3! ... !hostn!userID To send e-mail to someone who has a UUCP address, reverse the given address and format it as below (assuming uuhost!futility!borg!locutus): locutus%borg@uunet.uu.net Occasionally uunet.uu.net has problems addressing mail. This can frequently be fixed by adding more of the UUCP address: locutus%borg%futility@uunet.uu.net FIDONET. The FIDOnet is a global network of BBS (bulletin board system) hosts which are connected using modems and both store-and-forward and direct- connection technology. Begun in 1984, it is now becoming vogue for FIDO BBS systems to offer Internet mail, USENET news (see Section Five) and UUCP mail. In fact a number of BBS systems have attached to the Internet and have acquired Internet domain names (for example, the Channel One BBS is known on the Internet as 'channel1.com'). FIDOnet addressing organizes the FIDO network into zones, networks, nodes and points. To address e-mail from the Internet to FIDOnet, take the source address David O'Donnell at 1:323/121 and convert it to an Internet-style address: David_O'Donnell@f121.n323.z1.fidonet.org Note that if you are on a FIDO BBS where the BBS address includes a "point" (e.g., the address ends in a decimal point followed by a number, such as '.15'), you should exclude the "point" when converting to RFC 822 compliant addressing. While FIDO networking software can support this extra "dimension" to an address, not all sites do and therefore, to ensure the greatest compatibility, you should not include points in an RFC 822 compliant address. There are two additional-important-points to make note of: * The addressing scheme listed here works only for sending mail into FIDOnet. Sending mail to a FIDOnet BBS which has an Internet domain name should be done using the Internet format, not the one given above. * Other than for FIDOnet sites which have Internet domain names, I do not know of any fashion for sending mail from FIDOnet to the Internet. The Internet-FIDOnet gateway is located in Arizona and the system operator (or SysOp) there has indicated that large amounts of incoming e-mail are not appreciated, as he must pay for the long distance call to pick up the e-mail. You should check with your local BBS SysOp to see if there are more efficient Internet connections in your area. Section Three: FILES One of the most exciting aspects of the Internet is its huge repository of files-programs, data, graphics, sound and so forth-all of which are accessible to people with Internet connections. While it used to be the case that access was limited to local users and those with FTP, mechanisms are now in place which can locate, index and transfer files from the Internet via e-mail. FTP: File Transfer Protocol FTP is "the mother of all file transfer protocols". Through it a user directly connects to an Internet-connected computer, accesses the file system and gets or stores files. FTP is one of the basic portions of the TCP/IP suite, a collection of protocols which define networking on the Internet. Virtually any kind of file anyone could possibly want can be found on the Internet and retrieved using FTP. Some of the more popular sites and their contents include: Topic Site Name Windows Applications ftp.cica.indiana.edu, wuarchive.wustl.edu, nic.funet.fi, wsmr-simtel20.mil Macintosh Applications sumex-aim.stanford.edu, brownvm.brown.edu, wuarchive.wustl.edu GIFs wuarchive.wustl.edu, sumex-aim.stanford.edu, archive.msdos.umich.edu DOS Applications wuarchive.wustl.edu, archive.msdos.umich.edu, oak.oakland.edu To mangle a metaphor, this is only the tiniest tip of the Internet file iceberg. To help make some organizational sense of the hundreds of thousands of files available on the Internet, archie and gopher were created. To help those without FTP access obtain files, there is the BITFTP service. Archie: the Archive Server Archie is an application which is used to quickly determine what sites on the Internet carry files that are being sought by a user. The current implementation of archie consists of two parts, only one of which is necessary: the archie server and the archie client. It helps to think of archie as "the ultimate Internet file list filter", because that is precisely what it does with the file lists of 'anonymous' FTP sites. ARCHIE SERVER. The "archie server" is an application which is run as the operating environment of a computer account. To use an archie server one must either have access to TELNET (see Section Four) or be able to log in to a computer which has an archie server running. Two of the several available servers are on hosts archie.sura.net and archie.unl.edu At the user login prompt one enters "archie" as the user name. There is no password and the archie server will display a brief introductory banner, frequently containing news of the day. More recent versions of the server system are able to detect the terminal emulation of your computer and will adjust output to match. If you are using a non-standard setup (for example, my VT-320 setup has 32 rows instead of the default 24) you would use the set term command as follows: set term terminal-name screen-rows screen-columns In my case the command is set term vt320 32 80 Other archie commands include the following: help See help topics. pager Break file listings into pages; press SPACEBAR to advance pages. prog argument Search for all entries which contain the string argument. set regex Use UNIX-like 'regular expressions' for entry searches. Release 3.0.1 of the archie server will allow you to specify an e-mail address to which the results of a search can be sent. ARCHIE CLIENT. The archie client is an interface application which is similar to that of the archie server (in fact, the server runs a client application which connects to the database manipulation/search application). There are reputedly versions of the client available for UNIX, VMS and VM/CMS. BITFTP: File Access for the FTP-challenged The BITFTP system was originally created to allow computer users on the BITNET to have access to Internet FTP sites. The application runs at Princeton University and is manipulated via e-mail. To learn how to use BITFTP, send e- mail to bitftp@pucc.princeton.edu. Do not include a Subject: tag in the header of the e-mail; instead, as the body include the single word 'help'. GOPHER: the Internet Distributed Information Delivery Service GOPHER is a sophisticated information retrieval and management system which is becoming popular for CWISes, or Campus-Wide Information Services. Utilizing client-server technology, GOPHERs efficiently divide information retrieval/management tasks between the client (where the user is located) and the server (where the GOPHER 'brains' are located). GOPHER is "the only application that truly makes navigating and using many services on the Internet as natural as choosing an entrie from a dinner menu." UTILITY. GOPHER was first developed at the University of Minnesota as a text- retrieval system. It has since grown into an application of prodigious scope. Text files, binary files, sounds and graphics can be retrieved with GOPHER. With the proper GOPHER client/server arrangement it is possible to browse FTP archive directories and retrieve their contents; use archie to browse the anonymous FTP sites on the Internet and retrieve contents; TELNET and TN3270 to other Internet hosts; and link up with distributed information services like the World-Wide Net (WWN) and other WAISes (Wide-Area Information Services). CLIENTS. GOPHER client software exists for DOS, Macintosh System, Xwindows, VM/CMS and UNIX. Many Internet hosts, particularly educational sites, are installing GOPHER clients as part of their CWIS or WAIS. Each client has differing levels of functionality: some can only retrieve and view text files, some can display graphics and play sound files, and others have intuitive, user-friendly graphical interfaces. To find out if a particular site has a GOPHER, log in to it and try 'gopher' or 'cwis'. You may need to contact system administration for further information. Section Four: TRAVELING While FTP and e-mail are great tools for manipulating the Internet and keeping in touch with users, it is also possible to literally travel the network via the TELNET application. USE. To use TELNET, you must be working on a computer which has full access to the Internet: a TCP/IP connection, IP address and a domain name. Depending on the operating system of your computer it may be necessary to establish access to the TCP/IP software. MODES. TELNET operates in two modes: full-screen and line. Under most circumstances, connections will be established automatically in full-screen mode. In this mode, what you see on your screen is a direct representation of the screen, operating environment and file system of the remote computer. Although you will not have physically moved, for all intents and purposes your computer is now attached to the remote computer is if it were plugged in at the remote site. Line mode acts as though your computer were a teletype terminal attached to the remote system. None of the advantages of full-screen operation are possible. Line mode does have a valuable purpose, however. Due to the simple, streamlined nature of the connection, line mode is particularly suitable for client applications: FTP, IRC, archie, et cetera. Line mode is also activated when your local computer cannot correctly emulate a terminal attached to a remote computer. This most commonly occurs when someone logged into a VM/CMS machine attempts to open a TELNET connection to a UNIX or VMS system. Both latter operating systems support stream-oriented all-points-addressable video terminals while the technology used by IBM mainframes is more primitive. Section Five: OTHER SERVICES This section covers some of the other services available on the Internet as well as list some of the sites where access to the Internet can be obtained for free or for a nominal fee. IRC: Internet Relay Chat IRC is an interactive, real-time conversation system. Originally begun as a hobby by some computer science students, IRC has grown to be a 24-hour-a-day phenomenon. At any given time over a thousand people will be using IRC from sites all across the world. DESCRIPTION. IRC is similar to the "CB" chatting fora on commercial services. Users are organized into channels where conversation is public. Users can also send and receive private messages. Due to the informal nature of IRC (regardless of NSFNET regulations on network resource usage) IRC users, or "ircers" as they are known in the vernacular, are free to adopt nine-character nicknames and can change personal user information to virtually anything. Each channel on IRC is identified by a pound-sign (#) and up to 80 letters or numbers-for example, "#Resistance_is_futile". Channels can have topics which are publicly visible to anyone requesting a list of channels. They can be private, invitation-only, be restricted to a certain number of members and can even be invisible to the public. People who are not "on" a channel are said to be "in Limbo". Channels are organized in whatever fashion suits the members. There can be channel operators, called "chanops", who monitor the members of a channel and can change the attributes of the channel as well as forcibly remove someone from the channel (called "kicking", from the /kick command used). There can be more than one chanop to a channel-or none, if it is so desired. Channels are created when an ircer uses the /join command with the name of a channel not already in existence. Channel creators are automatically made chanop. Above the chanop is the ircop, an individual who runs one of the IRC server applications. All IRC clients must connect to a server to join the "IRCnet", and all servers are interconnected via special line-mode TELNET links. An ircop has what passes for ultimate authority on IRC: he or she can join any channel, irrespective of the 'mode' settings; become a chanop on any channel; kick someone off a channel into Limbo or kill a user's IRC client session; and manipulate the interconnections of IRC servers. In theory only responsible people are given ircop status, but in reality things are often quite different. ACCESS. To use IRC you need to have two things: a TCP/IP connection to the Internet and a client application (usually called "ircii" for "Internet Relay Chat, version two"). In actuality it is possible to use IRC without a client application but writing from personal experience, it is not enjoyable. IRC clients exist for UNIX, ULTRIX, VMS and VM/CMS. Versions for Macintosh System 7.1, MS-DOS and Windows are in the works. When you start IRC you provide a server's Internet name or numeric IP address to the client so it knows where to connect. There are several hundred servers around the world, but normally you will only connect to the closest server. An exception would be when the 'local' server does not support certain features. Some of the most popular servers are: csa.bu.edu The Boston University "hub" server for New England. irc.mit.edu MIT's server is not always available and usually out-of-date. hub.cs.jmu.edu James Madison University's server, which also allows /note spy's. CLIENTS. The latest versions of IRC client software are always maintained on the Internet site at Boston University, cs.bu.edu in the directories irc/clients irc/clients/VMS and irc/clients/VM As other client versions become available they will be added to the FTP site. Many Internet sites already have an IRC client available for users, while others specifically prohibit the use of IRC. Be sure to check with your site administration for policy. USENET News USENET news is one of those slippery, "magical" services which is only now becoming integrated with the Internet. USENET began in 1979 as a bulletin board system between two universities in North Carolina. It has grown to be one of the largest, most comprehensive "news" systems in the world. Several thousand conferences exist, broken down into a loose hierarchy similar to that of the domain naming system. USENET news has the advantage that anyone with either a UUCP-style mailer or access to a NNTP-compliant news reader can receive and contribute to the gestalt. ACCESS. Access to USENET news is primarily through news reader applications. It is possible, however, to participate in a USENET news conference if a computer site offers to redistribute it via e-mail. Keeping current with news is particularly difficult in this situation, however. Many news conferences are extremely prolific and corresponding via the relatively slow system of electronic mail is inefficient. Two of the thousands of conferences available are perfect examples of the profligate nature of USENET news: in a one week period, my threaded news reading application displayed 640 new items in group soc.motss. Read now? [YNyn] 1254 new items in group rec.arts.startrek. Read now? [YNyn] INTERFACES. Every major operating system has a USENET news reading application available. On the Internet, the following four are most common. * Rn, typical UNIX shorthand for "read news," is the archetypal news reading software. Its interface is straightforward and unembellished. It lacks threading, however, and can quickly overrun the user. * Readnews is a somewhat advanced version of rn. Its primary advantage is the ability to easily mark threads of topics as "killed"-the news reader will automatically skip over those topic/threads. Newer releases of rn may have this feature as well. * Trn, or "threaded readnews," is undoubtedly my favorite. The trn user can choose from two formats for newsitem display: the venerable list format of rn, or a visual, threaded format where one can follow the progress of a discussion thread through the newsgroup. * VNEWS is the VMS equivalent of trn. VNEWS has a particularly pleasing visual display, though its built-in assumption that the user's keyboard is the DEC LK102 can be frustrating. VMS' built-in help system also provided inspiration for VNEWS, although VNEWS' is arguably even more difficult to make heads or tails of. NEWSGROUPS. New newsgroups are constantly being created, but the list below should give you some good places to begin exploring USENET news: alt.sex rec.humor.funny comp.os.os2.advocacy alt.ensign.wesley.die.die.die alt.french-captain.borg.borg.borg comp.os.ms-windows.advocacy talk.* soc.* rec.arts.startrek alt.binaries.pictures alt.binaries.sounds.misc Note: in groups like talk.*, the asterisk signifies that there are many newsgroups under the hierarchy, each of which is likely to prove interesting. LISTSERV: The List Server Network LISTSERV is the name of an application written by Eric Thomas of SEARN for the maintenance of e-mail- based redistribution lists. Originally created to service the BITNET, LISTSERV has migrated to UNIX platforms as well and the mailing lists managed by LISTSERVers can be accessed by anyone with Internet e-mail capability. LISTSERV is run as a loosely organized network of "LISTSERVers" spread out across the BITNET. LISTSERVers maintain databases of e-mail distribution lists (commonly referred to as "lists"), their subscribers and any related files or archives of discussion. Several thousand lists currently exist; all that is required to create a new list is to find a LISTSERV manager willing to store the database entry, archives and subscription list. Similar to the organization (or, rather, lack thereof) of USENET, there is no central administrative LISTSERVer, though most are highly integrated into the computing services departments of their organizations and follow both general BITNET regulations and organizational rules. ACCESS. Anyone who can send and receive Internet e-mail can get access to LISTSERV. Users on the BITNET have the additional advantage of being able to command the servers via interactive messaging, but all commands are available through e-mail. LISTS. Every LISTSERVER maintains what is commonly referred to as "the list of lists". This document is several thousand lines of text giving the name and LISTSERV host of every list. To obtain a copy of the "list of lists", address e-mail to any LISTSERVER. In the body of the mail, include the command "list global": Date: Friday, 26 March 1993; 17:38:44 EST To: LISTSERV@brownvm.brown.edu From: Q@q.continuum.org (Guinan is a Pest) list global Sending the command "help" will allow you to obtain LISTSERV's on-line help documentation. How To Get Onboard Now that you've gotten your feet wet, how can you get access to the Internet? If you are content with e-mail access, check out the commercial services, especially CompuServe, America OnLine and Delphi. If you want more, you will need an account on an Internet host with full TCP/IP suite access. The list below includes some of the more popular sites and the services they offer. THE WORLD. The World at Software Tool & Die, Internet host name world.std.com offers a wide range of Internet services, including FTP, TELNET, IRC, USENET, e-mail and LISTSERV. The service is commercial with a reasonable fee structure of $2 per hour with a $5 per month account fee. DRYCAS. The Carneige-Mellon University Computer Club runs a DECstation 5000 with full Internet access. The Internet host name is drycas.club.cc.cmu.edu or DRYCAS on the BITNET. FTP, TELNET, IRC, USENET, LISTSERV and e-mail are accessible from this excellent VMS site. Contact Marc Shannon (accounts@drycas.club.cc.cmu.edu) for the latest information on their annual fee structure. Carneige-Mellon University is not local to the Boston area but DRYCAS is an excellent platform to explore both VMS and the BITNET from another Internet site. GNU/FSF. The GNU Project/Free Software Foundation, Internet host name gnu.ai.mit.edu offers FTP, TELNET, IRC, USENET, LISTSERV, e-mail and a host of GNU Project software for UNIX. This site is free and local to Boston. Unfortunately space is extremely limited and most applications for accounts will be denied without sufficient reason. To apply for an account, TELNET to gate.gnu.ai.mit.edu and login as apply Follow the directions from the application account service. IDS. IDS Data Forum is a BBS system running in Warwick, RI on a DEC minicomputer. Its Internet host name is idsvax.ids.risc.net IDS offers both BBS and Internet services on a fee-related basis. Internet access includes FTP, TELNET, USENET, LISTSERV and e-mail, but you must contact the SysOp to obtain access. You can reach him as green@nic.risc.net or by telephoning (401) 884-9002. LOCAL ACCESS. If you have a Macintosh, you may be able to connect to the Internet right from Ziff-Davis. Contact David Bogartz at x5280 for additional information. Wrapping Everything Up I hope you have found this handout to be useful and entertaining. The Internet provides a wealth of opportunity which is only just beginning to be tapped. If you have any questions about anything contained in this handout or would like more information on the aspects of the Internet which were not covered, please send me e-mail at atropos@netlab.cis.brown.edu Section Six: NOTES (This Page Left Blank) Section Seven: INDEX TO TOPICS [ Note: section removed from text document ] Footnotes: [ Note: regrettably, the footnote numbers did not survive the conversion to text. ] Packet switching networks are the forebears of modern network schemes. In a packet-switched network such as BITNET, data are sent as small packets which are queued through the network in a store-and-forward fashion; that is, to get from one host to another, it may be necessary for the packets to travel through several intermediary hosts. The speed of transmission is greatly dependent on the intermediary systems. Data based on the January 1993 Internet Domain Survey. Extrapolations o f growth figures would indicate that current numbers are between twenty and thirty percent higher than the numbers given. VM and MVS are the realm of IBM mainframes and compatible systems. VMS and ULTRIX run on a variety of DEC computers, and UNIX is running on virtually everything from Intel 80386-based PCs on up. UNIX is a trademark of Bell Labs. This site is completely fictional, but it is a good example of how computers frequently are named on the Internet. Since it's highly unlikely that Buckaroo Banzai's fictitious 'Yoyodyne Propulsion Systems' is really a part of the University of Sydney, this address makes an important exception to the naming style used in my own (atropos@netlab.cis.brown.edu) because all that is certain about the addressee is that he or she is located at the University. Chances are that the computer the user is logged into most frequently is called 'captain-kirk' but there is no insurance that it will always be so. Most DOS- and Windows-based terminal emulation packages can support at least VT-220 emulation. VT-100 and VT-101 are essentially standard and even Windows' TERMINAL application can support VT-100 emulation. Source: !%@:: A Dictionary of Electronic Mail Addressing and Networks, second edition. UUENCODE and UUDECODE are UNIX utilities which were developed to allow the transmission of binary (non-textual) files across networks where 8-bit ASCII was not allowed. Because of the nature of the Internet, it is not safe to assume that binary files transmitted as attachments to e-mail will make it to the destination intact unless they have been UUENCODEd first. Luckily, versions of UUENCODE and UUDECODE exist for virtually every operating system- including DOS and Windows. Unlike other networks, the UUCP network-which is primarily known for USENET news-does not have a central administrative body. Decisions are frequently made by system administrators (sysadmin or sysops) on hosts with many connections or users who are particularly active. UUCP, like much of UNIX, can often be compared to a kind of mystical magic in the way it works (or doesn't work). On FIDOnet it is possible to address e-mail to users which will get distributed via the normal BBS mail cycle (store-and-forward) and via a direct-mailing system where the BBS host will directly dial the recipient's BBS and transmit the e-mail. This address is for GAYtway Communications Network BBS in Providence, RI. TCP/IP stands for "Transmission Control Protocol/Internet Protocol". Th e suite include: FTP (File Transfer Protocol), TELNET, PING, NNTP (Network News Transmission Protocol) and SMTP (Simple Mail Transfer Protocol). Anonymous FTP sites are named thus because anyone can log into the server and retrieve files. To do so, one uses FTP to open a connection to the site. At the username: prompt, enter anonymous or ftp; at the password: prompt, enter your Internet e-mail address. Since not all sites can use 'ftp' as the username, 'anonymous' is best. If you are on the BITNET, you can also manipulate BITFTP through interactive messages to bitftp at pucc. Reprinted from UIUCnet, volume 6 number 1, a University of Illinois publication. An IP address is the 32-bit decimal address used by TCP/IP software; it is divided into four tuples, each of which represents one byte's worth of the 32 bits. For example, the IP address for brownvm.brown.edu is 128.148.128.40. Source: !%@:: A Dictionary of Electronic Mail Addressing and Networks, second edition. The basic USENET hierarchy consists of comp (computer-related topics), sci (scientific topics), soc (social topics), alt (alternative or "on the edge" topics), rec (recreative topics), news and misc (miscellaneous items); in the first newsgroup motss stands for "members of the same sex". And the Internet in general, though the primary thrust remains located on the BITNET. .