Guide to network tools 1993

 EARN Association

                              September 15, 1993

                             Document Number:  2.0

                        Guide to Network Resource Tools


       This  document  has  been  compiled  and  produced  by the   EARN
       Association.   Permission to  copy  all or part  of this document
       without fee  is granted  provided the  copies  are  not used  for
       commercial advantage  and  that the EARN Association  is cited as
       the source of the document.

       This   document  is   available   in   electronic  format   from:

       Send the command: GET filename where the filename is either:

        NETTOOLS PS             (Postscript)
        NETTOOLS MEMO           (plain text)

       Document version: 2.0

       ISBN  2-910286-03-7

       (ISBN  2-910286-01-0  Version 1.0)

       (c) EARN Association, 1993

                         Preface to the Second Edition

       When the first version of the Guide to Network Resource Tools was
       released,  we had  no idea how popular it would  become,  and how
       quickly it would spread beyond the EARN community. While much had
       already been written on these tools,  a booklet with brief expla-
       nations and instructions struck a responsive chord.

       In this second edition,  we have tried to expand the scope,  both
       in terms of  the number of networking tools covered  and the size
       of the audience at which it is aimed.  While the Internet contin-
       ues to grow at an explosive rate, there is still a large communi-
       ty of users,  within EARN and  beyond,  with only electronic mail
       connectivity to the world-wide network. We have tried to show how
       they too can utilize these networking tools.

       The work of many people is reflected here,  but we owe our great-
       est debt of  thanks to the developers and authors  of the network
       tools and documentation.  Their work serves as the basis for this

       We also wish to thank all those who have provided comments, crit-
       icisms and suggestions for the guide. Many of the improvements in
       the second edition  are the result of their  feedback and encour-
       agement.  In particular,  we have benefitted from the cooperation
       of the RARE  Working Group on Information Services  and User Sup-
       port.  We look forward to close collaboration with them on future
       editions of this guide as well as on other projects.

       The Guide  to Network  Resource Tools  has been  produced by  the
       staff of the  EARN Association,  under the direction  of the EARN
       Group on Information Services. Any errors or inaccuracies in this
       guide are the sole responsibility of  the EARN staff.  We welcome
       and encourage your feedback.   Please send comments,  criticisms,
       corrections,  suggestions for future editions  and even praise to
       the electronic mail address:


       If you have any questions or queries on the use of these tools or
       on other  aspects of networking,  and  you are in an  EARN member
       country, you can get help and advice by sending your questions by
       electronic   mail   to   the    EARN   consulting   service   at:

       The EARN Staff


       As the worldwide academic computer  network grows and expands far
       beyond  its previous  confines,  so  the  resources and  services
       available on the network evolve and  multiply at a dizzying rate.
       The typical  user is hardpressed to  keep up with  this explosive
       growth.  Fortunately,  a number of tools are available to facili-
       tate the task  of locating and retrieving  network resources,  so
       that users anywhere can utilize texts, data,  software and infor-
       mation for  public access.  Facilities  to explore  public domain
       software repositories, to consult mailing list archives and data-
       bases,  to retrieve  directory information and to  participate in
       global group discussions are now available to all.

       The key to exploiting these resources is a server,  special soft-
       ware  on  a  computer  somewhere in  the  network  which  accepts
       requests (or queries or commands)  and sends a response automati-
       cally. The requestor does not have to be working on the same com-
       puter (or even in the same part of the world) in order to use the
       server. Many servers accept requests via electronic mail, so that
       often the requestor  needs not even be on the  same computer net-
       work as the server.  In many cases, servers are interconnected so
       that once you have established contact  with one server,  you can
       easily communicate with other servers as well.

       Today,  many users have powerful  computers on the desktop,  with
       advanced graphical,   audio and storage capabilities,   which are
       connected to  the network.  This fact  has given rise to  what is
       known as the client-server model. Users can have special software
       on their  local computer  called a client  which can  utilize the
       capabilities of  that computer  and can  also communicate  with a
       server on  the network.   These clients  provide an  easy-to-use,
       intuitive user interface, allow use of pointing devices such as a
       mouse,  and exploit  other local features.  The  client sends the
       user's requests to a server using a standardized format (called a
       protocol) and the server sends its response in a condensed format
       which the client displays to the user in a more readable way.

       The tools  described in  this guide  have been  divided into  six
       functional areas.  Actually,   several of the tools  have several
       different functions,  and have been classified according to their
       main purpose.  The first section,  Exploring the network,  covers
       two services,  Gopher and World-Wide  Web,  which use the client-
       server model to provide a means of moving through a wide range of
       network sources and resources in a  uniform and intuitive way.  A
       tool for searching in a wide range of different databases located
       throughout  the network,   WAIS,  is  documented  in the  section
       Searching databases.  The problem of  knowing where to find files
       and programs  in the network is  addressed in the  third section,
       Finding network resources,  which deals with archie.  Three tools
       for finding people, computers and their network addresses, WHOIS,
       X.500 and Netfind, are discussed in section four,  Finding people
       and computers. While just about all of these network tools can be
       used to get files of one sort or another, there are a few servers
       available for getting  files easily and efficiently  from various
       repositories in the network.  Two  of these servers,  TRICKLE and
       BITFTP,  are covered in section five on Getting files.  The sixth
       section deals  with what is perhaps  the most popular of  all the
       network resources,  discussion groups  on every imaginable topic.
       The two tools discussed in this  section are LISTSERV and Netnews
       (Usenet). The final section gives brief descriptions and pointers
       for a number of  tools which were not mainstream enough  to get a
       full  description.  Some  are still  in  the developmental  stage
       (Prospero), some are relatively unknown outside a particular net-
       work (ASTRA and Netserv from EARN/Bitnet and Mailbase from JANET)
       and some are meant for chatting rather than work (Relay and IRC).

       The purpose of this guide is to supply the basic information that
       anyone on  the network  needs to  try out  and begin  using these
       tools.   A basic  knowledge of  networking  terminology has  been
       assumed,  as well as familiarity with the basic tools of network-
       ing:  electronic mail (often referred to as e-mail or simply mail
       throughout this guide) and,  for those connected to the Internet,
       FTP (file transfer  protocol)  and Telnet (remote  login).  It is
       beyond the scope of this guide to describe these basic tools. The
       example in the BITFTP section of this guide shows how one can use
       BITFTP to get guides to these tools over the network.

                                    Part 1

                             EXPLORING THE NETWORK


       What is Gopher

       The Internet Gopher, or simply Gopher,  is a distributed document
       delivery service. It allows users to explore, search and retrieve
       information residing on  different locations in a  seamless fash-

       When browsing it, the information appears to the user as a series
       of nested menus. This kind of menu structure resembles the organ-
       ization of a  directory with many subdirectories  and files.  The
       subdirectories and the  files may be located either  on the local
       server site  or on remote sites  served by other  Gopher servers.
       From the user point of view,   all information items presented on
       the menus appear to come from the same place.

       The information can be a text or binary file,  directory informa-
       tion (loosely called phone book),   image or sound.  In addition,
       Gopher offers  gateways to other information  systems (World-Wide
       Web, WAIS,  archie,  WHOIS)  and network services (Telnet,  FTP).
       Gopher is often a more convenient way to navigate in a FTP direc-
       tory and to download files.

       A  Gopher server  holds the  information and  handles the  users'
       queries. In addition, links to other Gopher servers create a net-
       work  wide cooperation  to form  the global  Gopher web  (Gopher-

       Who can use Gopher

       Gopher  uses the  client-server model  to provide  access to  the
       Gopher web.  You must be on the international TCP/IP network (the
       Internet)  in  order to use a  client on your computer  to access

       How to get to Gopher

       Users explore  the Gopher  menus using  various local  clients or
       accessing a remote client via an interactive Telnet session.

       Local clients

       Public domain clients for accessing a Gopher server are available
       for:   Macintosh,   MS-DOS,  OS/2,   VM/CMS,  VMS,   NeXT,  Unix,
       X-Windows.   The  clients are  available  for anonymous  FTP from
       many   FTP sites  (eg,  in the  directory
       /pub/gopher).   See the list of  freely available client software
       in Appendix A.

       Remote clients

       Some sites  allow public access  to a  client.  To access  such a
       remote client, telnet to one of these sites:

       |                                                               |
       |                 Australia (login: info)      |
       |                   Columbia                     |
       |                        Ecuador                      |
       |              Sweden                       |
       |        USA                          |
       |                 USA                          |
       |                 USA (login: panda)           |
       |                                                               |

       At the login: prompt type gopher (unless specified otherwise) and
       the top-level Gopher menu for that site will be displayed.

       Users are requested to use the site closest to them.

       Using Gopher

       The implementations  of the Gopher  clients on  various platforms
       are slightly different to take  advantage of the platforms' capa-
       bilities (mouse,   graphic functions,  X-Windows server)   and to
       offer the popular look and feel.  Even with different implementa-
       tions, the same set of functions and commands is available.

       When issuing the gopher command,  you will be connected automati-
       cally to the default Gopher server specified at the installation.
       The format of the command is:

       |                                                               |
       |  gopher    <hostname>                                         |
       |                                                               |

       where hostname is an optional  alternative Gopher server you want
       to talk to.

       When connected to a Gopher server, it is still possible to access
       another server by exploring the Other  Gopher servers in the rest
       of the world branch.  To locate them more easily, the Gopher ser-
       vers are distributed in geographical regions:

       *   Africa
       *   Europe
       *   Middle East
       *   North America
       *   Pacific
       *   South America

       and then by countries.

       Access to a Gopher server is identical whether using a local or a
       remote  client:  a  simple  menu-driven  interface which  doesn't
       require any special training or knowledge from the user.

       Here is a sample menu:


                        Internet Gopher Information Client v1.1

                             Information About Gopher

           1.  About Gopher.
           2.  Search Gopher News <?>
           3.  Gopher News Archive/
           4.  comp.infosystems.gopher (Usenet newsgroup)/
           5.  Gopher Software Distribution/
           6.  Gopher Protocol Information/
           7.  University of Minnesota Gopher software licensing policy.
           8.  Frequently Asked Questions about Gopher.
           9.  gopher93/
           10. Gopher| example server/
           11. How to get your information into Gopher.
       --> 12. New Stuff in Gopher.
           13. Reporting Problems or Feedback.
           14. big Ann Arbor gopher conference picture.gif <Picture>

       Press ? for Help, q to Quit, u to go up a menu          Page: 1/1


       In the example above, any item can be selected by typing its line
       number or by moving the cursor (-->) next to it.

       An item could be:

       *   a subdirectory
       *   a text file
       *   a binary file
       *   a sound file
       *   an image file
       *   a phone book (directory information)
       *   an index-search
       *   a Telnet session

       Items are displayed with an identifying  symbol next to them.  In
       the example  above,  "<?>" means  a full text  index-search,  "/"
       means a subdirectory, "<Picture>" means an image file and no sym-
       bol means a text file.

       Some Gopher  clients are  not able to  handle certain  file types
       (eg, sound files).  Some clients display only files of types they
       can handle or  files they suppose you are  interested in.  Others
       display all types of files.

       Most Gopher clients  allow you to create,  view  and select book-
       marks.  A bookmark keeps track of  the exact location of a Gopher
       item, regardless of where it resides. It is useful when you often
       need to reach a file or a  service located far from the top-level
       directory.  A collection of bookmarks is like a customized Gopher

       Some capabilities of a local Gopher client are bound to the capa-
       bilities of your own computer.  In fact,  for sound files,  image
       files and Telnet sessions, the Gopher client looks for the appro-
       priate software on your computer and passes control to it to per-
       form the requested task.  When the task is completed,  control is
       returned to the Gopher client.

       At any time,  it is possible  to terminate the session (quit com-
       mand),  to  cancel the current processing  or to get  the on-line
       help (help command).

       An item is processed according to its type:

       a subdirectory
                    its contents are displayed.  To go up one level, use
                    the up command.

       a text file  the  file is  displayed.  Then  you  can browse  it,
                    search for a particular string,  print it on a local
                    printer or copy (save) it onto your local disk space
                    in a user-specified  file (the last 2  functions may
                    not be available to you).

       a binary file
                    the remote  file is  simply copied  onto your  local
                    disk space  in a user-specified file.   Binary files
                    are  binhexed  Macintosh   files,   archives  (.zip,
                    .tar,...), compressed files, programs, etc.

       a sound file the remote file  is played through your  local audio
                    device if  it exists,   as well  as the  appropriate
                    utility.  Only  one sound  file can  be active  at a
                    time;  you will be warned if you try to play a sound
                    before a previous one is done.

       an image file
                    the remote file is displayed on your computer screen
                    if an image viewer exists on your computer.

       a phone book you are prompted for a search string to look up peo-
                    ple  information through  the  selected phone  book.
                    Since different institutions have different directo-
                    ry services,  the  queries are not performed  in the
                    same fashion.

       an index-search
                    you are  prompted for a  search string which  may be
                    one or more words,  plus the special operators  and,
                    or, and not. The search is  case-insensitive. Usual-
                    ly,  an index is  created to help  users  locate the
                    information  in  a set  of  documents quickly. Eg:

                         terminal and setting or tset

                    will find all documents which contain both the words
                    terminal and setting,  or the word tset.  or is non-
                    exclusive so  the documents may  contain all  of the

                    The result of the index-search looks like any Gopher
                    menu, but each menu item is a file that contains the
                    specified search string.

       a Telnet session
                    Telnet sessions are  normally text-based information
                    services, for example, access to library catalogs.


       Veronica was  designed as a solution  to the problem  of resource
       discovery in the rapidly-expanding Gopher  web,  providing a key-
       word search  of more than 500  Gopher menus.  Veronica  helps you
       find Gopher-based information without doing a menu-by-menu, site-
       by-site search.   It is  to the  Gopher information  space,  what
       archie is to the FTP archives

       Veronica is accessible  from most top-level Gopher  menus or from
       the Other Gopher servers... branch.  There is no need for opening
       another connection or another application.

       When you choose a veronica search , you will be prompted to enter
       a keyword or keywords.  The simplest  way to search with veronica
       is to enter  a single word and  hit the RETURN key.   It does not
       matter whether the word is upper-case or lower-case. The veronica
       server will return  a gopher menu composed of  items whose titles
       match your keyword specification.   Items can be accessed as with
       any Gopher menu. Eg:


       will give you a list of menu titles that contain eudora, such as:

            Electronic Mail: Eudora on Macintosh, Micro-08
            Modem Setting Eudora Slip.
            A UNIX-based Eudora reader for those that ...
            Eudora:  Popmail for the Macintosh.


       The search  string may contain  keywords optionally  separated by
       and, or and not.  If there is no operator between 2 keywords, and
       is assumed. Eg:

            eudora and macintosh

       will give you a list of menu  titles that contain both eudora and
       macintosh, such as:

            Eudora:  Popmail for the Macintosh.
            Micro News:  Eudora - A Mailer for the Macintosh.
            Eudora: Electronic Mail on Your Macintosh.
            ACS News - Eudora Mail Reader for Macintosh.


       "*" is the wildcard character. It can replace any other character
       or characters at the end of a keyword. Eg:


       will give you a list of menu titles, such as:

            The Help Desk.
            Keene State College Press Release COMPUTER ON EVERY DESK.
            DESKQview/X... An alternative to Windows???.
            Ethernet at Your Desktop/


       Learning more about Gopher

       The Internet Gopher is developed  by the Computer and Information
       Services Department of the University of Minnesota.  Bug reports,
       comments, suggestions, etc. should be mailed to the Gopher devel-
       opment team at:

       Mailing list:

       To subscribe send a mail to:

       Usenet newsgroup: comp.infosystems.gopher

       A comprehensive description of veronica  search methods is avail-
       able from the veronica menus.

       Veronica is  being developed by Steve  Foster and Fred  Barrie at
       the University of Nevada.   Bug reports,  comments,  suggestions,
       etc. should be addressed to:

                                WORLD-WIDE WEB

       What is World-Wide Web

       World-Wide Web (also called WWW or  W3)  is an information system
       based on hypertext,  which offers a means of moving from document
       to document  (usually called  to navigate)   within a  network of

       Hypertext documents are  linked to each other  through a selected
       set of words.  For example, when a new word, or a new concept, is
       introduced in  a text,  hypertext makes  it possible to  point to
       another document which  gives more details about  it.  The reader
       can open  the second  document by selecting  the unknown  word or
       concept and the relevant section  is displayed.  The second docu-
       ment may also contain links to further details.   The reader need
       not know where the referenced document  is,  and there is no need
       to type  a command to  display it,  or to  browse it to  find the
       right paragraph. Cross-references may be defined in the same doc-
       ument. A collection of documents is a database.

       If you were reading this document on a hypertext system,  instead
       of this all too short explanation about hypertext, you would have
       a selectable pointer to a complete hypertext information web with
       examples and more pointers to other definitions.

       For instance, in the first document you might read:


       The  WorldWideWeb (W3)  is a  wide-area "hypermedia"  information
       retrieval initiative aiming  to give universal access  to a large
       universe of documents.


       Selecting hypermedia  will display the following  explanation for


                              WHAT IS HYPERTEXT
       Hypertext is text which is not constraied to be linear.

       Hypertext is text which contains "links" to other texts. The term
       was coined by "Ted Nelson" around 1965 (see "History").

       HyperMedia is a term used  for hypertext which is not constrained
       to  be text:  it can  include  graphics, video  and "sound",  for
       example. Apparently Ted Nelson was the first to use this term too.


       Then you  can learn more about  links and Nelson.    Indeed,  the
       links in WWW are not confined to text only,  so the term hyperme-
       dia is more accurate. For example, the link to Nelson might point
       to a file containing a picture  of Ted Nelson.  The picture would
       be displayed on your screen if you have a suitable configuration.

       Also,  special documents (indexes)  in  the WWW information space
       can be  search for  given keyword(s).  The  result is  a document
       which contains links to the documents found.

       World-Wide Web uses hypertext over the network:  the linked docu-
       ments may be located at various  sites.  WWW can handle different
       text formats and various information organizations. WWW also pro-
       vides access to many of the other tools described in this guide.

       Who can use World-Wide Web

       WWW uses the client-server model to  provide access to the infor-
       mation universe.  You must be on the international TCP/IP network
       (the Internet)   in order  to use  a client  on your  computer to
       access WWW.   If you  are on the Internet,  but don't  have a WWW
       client on your computer,  you can still enter the World-Wide Web.
       Several sites offer public interactive access to WWW clients (see
       the Remote  clients section  under How to  get to  World-Wide Web

       If you have e-mail access only, or if you are not on the Internet
       then  you  can not  fully  exploit  the  vast potential  of  WWW.
       However,    a   mail-robot   is   available   at   the   address: which gives e-mail access to WWW-accessible
       files.  (see E-mail access section under How to get to World-Wide
       Web below).

       How to get to World-Wide Web

       Users access the World-Wide Web facilities  via a client called a
       browser.  This interface  provides transparent access to  the WWW
       servers. If a local WWW client is not available on your computer,
       you may use a client at a remote site. Thus, an easy way to start
       with WWW is to access a remote client.

       Local clients

       Usage of  a local client is  encouraged since it  provides better
       performance and better response time than a remote client.

       Public domain  clients for  accessing WWW  servers are  available
       for:  Macintosh, MS-DOS, VMS, VM/CMS, MVS, NeXT, Unix, X-Windows.
       The clients are available for  anonymous FTP from in
       the directory /pub/www. All these platforms support a simple line
       mode browser.  In addition,  graphical clients are available for:
       Macintosh, Windows,  X-Windows,  NeXT and Unix.   See the list of
       freely available client software in Appendix A.

       Remote clients

       To access a remote WWW client, telnet to the client site.  If you
       are new  to WWW,  you should  telnet to No  login is
       needed.   You will  immediately enter the WWW  line mode browser.
       Some  publicly  accessible  clients   feature  locally  developed
       clients.  Most remote clients are at  sites with WWW servers with
       information on  specific areas.  After  you telnet to  the client
       site, at the login:prompt enter www, no password is needed.   The
       following remote client sites are available:

       |                                                               |
       |  Site                    Country          Server Specialization
       |                                                               |
       |          Israel            Environment        |
       |            Switzerland (CERN) High-energy physics
       |   USA               Law                |
       |    USA               History            |
       |            USA                                  |
       |                                                               |

       CERN is the entry point to  find information about WWW itself and
       to have an overview of the Web  with a catalogue of the databases
       sorted by subject.

       E-mail access

       In order to get a file, send mail to with a
       SEND command. The SEND command returns the document with the giv-
       en WWW address, subject to certain restrictions.  Hypertext docu-
       ments are formatted to 72 character width, with links numbered. A
       separate list at the end of the file gives the document-addresses
       of the related documents.

       If the document is hypertext, its links will be marked by numbers
       in brackets,  and a list of  document addresses by number will be
       appended to the message.   In this way,  you can navigate through
       the web, more or less.  A good file to start with would be:

       Note that,  despite the name listserv in the address of this mail
       robot, it is not a LISTSERV server.

       A note of caution from the WWW developers and maintainers:

       As the  robot gives potential mail  access to a *vast*  amount of
       information,  we  must emphasise that  the service should  not be
       abused. Examples of appropriate use would be:

       *   Accessing any information about W3 itself;
       *   Accessing any CERN and/or physics-related or network develop-
           ment related information;

       Examples of INappropriate use would be:

       *   Attempting to retrieve binaries or tar files or anything more
           than directory listings or short ASCII files from FTP archive
       *   Reading Usenet newsgroups which your site doesn't receive;
       *   Repeated automatic use.

       There is  currently a 1000 line  limit on any returned  file.  We
       don't want to overload other people's  mail relays or our server.
       We reserve the right to withdraw the service at any time.  We are
       currently monitoring all use of the server,  so your reading will
       not initially enjoy privacy.


       The W3 team at CERN (

       Using World-Wide Web

       When using a graphical interface, you access the WWW functions by
       pressing mouse buttons. In particular, references are highlighted
       or underlined words.  To follow a  link,  click on the associated

       The line mode browser is a more simple user interface: references
       are numbers in square brackets next to words. Type the number and
       hit the RETURN key to follow  a reference.  For example,  here is
       the beginning of  the Subject Catalogue you get on  the CERN ser-


             The World-Wide Web Virtual Library: Subject Catalogue
                             WWW VIRTUAL LIBRARY

       This is  the subject catalogue.  See also arrangement  by service
       type[1]. Mail to  add pointers  to this

       Aeronautics     Mailing list archive index[2]. See also NASA LaRC[3]

       Agriculture[4]  Separate list, see also Almanac mail servers[5].

       Astronomy and Astrophysics
                       Abstract Indexes[6] at NASA, Astrophysics work at
                       FNAL[7],   Princeton's[8]   Sloane  Digital   Sky
                       Survey,  the  STELAR   project,  Space  Telescope
                       Electronic Information System[9], the Southampton
                       University  Astronomy   Group[10],  the  National
                       Solar Observatory[11],  Astrophysics work  at the
                       AHPCRC[12]. See also: space[13].

       Bio Sciences[14] Separate list.

       Computing[15]   Separate list.

       1-81, Back, <RETURN> for more, Quit, or Help:


       The following commands are available  within WWW.  Some are disa-
       bled when not applicable (eg,  Find is enabled only when the cur-
       rent document  is an  index).  Angle  brackets (<>)   indicate an
       optional parameter.

       Help           gives  a list of  available commands depending  on
                      the context, and the hypertext address of the cur-
                      rent document.

       Manual         displays the on-line manual.

       Quit           exits WWW.

       Up, Down       scrolls up or  down one page in  the current docu-

       Top, BOttom    goes to the top or the bottom of the current docu-

       Back           goes back to the document you were reading before.

       HOme           goes back to the first document you were reading.

       Next, Previous goes to the next or previous  document in the list
                      of pointers from the document that led to the cur-
                      rent one.

       List           gives a numbered  list of the links  from the cur-
                      rent document.  To follow a link, type in the num-

       Recall <number>
                      if number is  omitted,  gives  a numbered list  of
                      the documents you have visited.

                      To display  one specific  document,  re-issue  the
                      command with number.

       <Find> keyword queries the  current index with the  supplied key-
                      word(s).  A list of  matching entries is displayed
                      with possibly links to further details.  Find  can
                      be omitted if the first  keyword does not conflict
                      with another  WWW command.  Multiple  keywords are
                      separated by blanks.

       Go docaddress  goes  to the  document  represented  by the  given
                      hypertext address, which is interpreted relatively
                      to the current document.

       Extra command available on Unix versions only:

       Print          prints the current document,  without the numbered
                      document references.  The default print command is
                      lpr,  but it may be  defined in your local working
                      environment by the variable WWW_PRINT_COMMAND.

       To access WWW with the line mode browser, type: www.  The default
       first document will appear on your screen.  From this point,  you
       should be  able to navigate through  the WWW universe  by reading
       the text  and following  the instructions  at the  bottom of  the
       screen. If you want to start with a first document other than the
       default,  or if you want to change some other aspect of the usual
       interaction,  there are  a number of command  line parameters and
       options available.  The full format of  the www command to invoke
       the line mode browser is:

       |                                                               |
       |   www      <options>  <docaddress <keyword>>                  |
       |                                                               |


                 is the hypertext  address of the document  at which you
                 want to start browsing.

       keyword   queries the index specified by docaddress with the sup-
                 plied keyword(s).  A  list of matching entries  is dis-
                 played. Multiple keywords are separated by blanks.

       Options are:

       -n        non-interactive mode.   The document  is formatted  and
                 displayed to the screen.  Pages are delimited with form
                 feed characters (FF).

       -listrefs adds a list of the addresses of all documents referenc-
                 es to the end.  Non-interactive mode only.

       -pn       sets the  page length  to n  lines.  Without  a number,
                 makes the page length infinite. Default is 24.

       -wn       sets the page width to n columns. The default is 78, 79
                 or 80 depending on the system.

       -na       hides references in the text. Useful, when printing out
                 the document.

       -version  displays the version number of the software.

       The commands  listed above  should be  available in  all clients.
       They  may be  abbreviated  (CAPITAL  letters indicate  acceptable
       abbreviation).  Case is not significant.  Special characteristics
       of the line mode browser interface are:

       number    type in a number given in []  and hit the RETURN key to
                 follow the link associated to the reference.

       RETURN    hit the RETURN key to display the next page of the cur-
                 rent document (without a reference number).


       WWW gives you  access to an information universe.   Let's say you
       want to know how many film versions of The Three Musketeers there
       have been.  You browse the WWW  Subject Catalogue and select Mov-


                                        Movie database browser (Cardiff)
       A Hypertext movie database browser

       Sep 2nd... Your help is needed..[1]

       Aug 29th.. Images, sounds, mpegs & reviews[2]

       Select the type of search you'd like to perform:-

          Movie people[3].....(multi Oscar winners)[4] or

          Movie titles[5] .....(multi Oscar winners)[6]

       Searches the "rec.arts.movies" movie database system, maintained
       by Col Needham et-al.

       Here[7] is some information on list maintainers.

       If you have a comment or suggestion, it can be recorded here[8]

       HERE[9] is a pre-1986 movie information gopher server. (at
               Manchester UK)

       1-13, Back, Up, <RETURN> for more, Quit, or Help: 5


       You select Movie titles,  and then  type three musketeers as key-


                                                     Movie title queries
                              MOVIE TITLE QUERY

       Enter a movie title or substring.

       Example,  to search  for movies  with the  word "alien"  in their
       title, type "alien".

       This will return details on several movies, including Aliens[1]

       Note: if the title begins with A  or The, leave it out. If you're
       determined to include it, then put ', A' or ', The' at the end of
       the of the substring e.g.

          Enforcer, The

          Gauntlet, The

       Searching is case insensitive.

        search menu[2] Fun and Games page[3] COMMA home page[4]

       FIND <keywords>, 1-5, Back, Up, <RETURN> for more,
       or Help: three musketeers


       You find that there have been six film versions of the story:


                                                              Movie Info
       Here are the results from the search for "three musketeers"

             Three Musketeers, The (1921)[1]

             Three Musketeers, The (1933)[2]

             Three Musketeers, The (1935)[3]

             Three Musketeers, The (1939)[4]

             Three Musketeers, The (1948)[5]

             Three Musketeers, The (1974)[6]

           search menu[7] Fun and Games page[8] COMMA home page[9]



       FIND <keywords>, 1-10, Back, Up, Quit, or Help: 1


       You decide to look for more information on the 1921 version:


                                                              Movie Info
       Here are the results from the search for "Three Musketeers, The (1921)"

                             THREE MUSKETEERS, THE (1921)


         Cast           Belcher, Charles[1] ......Bernajoux
                        De Brulier, Nigel[2] ......Cardinal Richelieu
                        De La Motte, Marguerite[3] ......Constance Bonacieux
                        Fairbanks, Douglas[4] ......D'Artagnan
                        Irwin, Boyd[5] ......Comte de Rochefort
                        MacLaren, Mary[6] ......Queen Anne of Austria
                        Menjou, Adolphe[7] ......Louis XIII
                        Pallette, Eugene[8] ......Aramis
                        Poff, Lon[9] ......Father Joseph
                        Siegmann, George[10] ......Porthos
                        Stevens, Charles[11] ......Planchet

         Directed by    Niblo, Fred[12]

         Music by       Gottschalk, Louis F.[13]

       1-21, Back, Up, <RETURN> for more, Quit, or Help: 7


       You're hooked! You decide to look for more information on Adolphe
       Menjou, search more titles, find Oscar winners, etc.

       Learning more about World-Wide Web

       World-Wide  Web is  being developed  at  CERN (European  Particle
       Physics Laboratory)   by the  World-Wide Web  team leaded  by Tim
       Berners-Lee.  Bug reports, comments, suggestions, etc.  should be
       mailed to:

       On-line documentation is available from,  for anony-
       mous FTP or using the remote WWW client.

       Mailing lists:
       To subscribe send a mail to

       Usenet newsgroup: comp.infosystems.www

                                    Part 2

                              SEARCHING DATABASES


       What is WAIS

       WAIS, Wide Area Information Server,  is a distributed information
       retrieval system.  It helps users  search databases over networks
       using an  easy-to-use interface.  The databases  (called sources)
       are mostly collections of text-based documents, but they may also
       contain sound,  pictures  or video as well.   Databases on topics
       ranging from Agriculture  to Social Science can  be searched with

       The databases may be organized  in different ways,  using various
       database systems,  but the user isn't required to learn the query
       languages of the different databases.  WAIS uses natural language
       queries to find relevant documents.  The result of the query is a
       set of documents which contain the words of the query:  no seman-
       tic information is extracted from the query.

       Who can use WAIS

       WAIS uses the client-server model to provide access to databases.
       You must be on the international TCP/IP network (the Internet) in
       order to use a client on your computer to access WAIS.

       If you have e-mail access only, or if you are not on the Internet
       you can still  exploit some of the potential of  WAIS.  An e-mail
       interface is available at  the address:
       which gives  e-mail access to  WAIS databases (see  E-mail access
       section under Using WAIS below).

       How to get to WAIS

       There are many WAIS servers throughout the network.  A directory-
       of-servers  database is  available  at  several sites.   You  can
       address a query to it, eg,  to find out what databases are avail-
       able on a particular subject. This database is also available via
       anonymous FTP from in the directory /wais as file wais-

       If you do not have access to a WAIS client, (at least) two demon-
       stration sites are available to allow  you to get acquainted with
       WAIS.  You can telnet to:

       |                                                               |
       |   (login: wais)                              |
       |   (login: swais)                             |
       |                                                               |

       The two demonstration sites above run swais (Screen WAIS), a sim-
       ple WAIS client for Unix.

       Using WAIS

       There are many freely available client software programs for var-
       ious operating systems (Unix, VMS, MVS,  MS-DOS,  OS/2 and Macin-
       tosh) and for specific environments (GNU Emacs, X-Windows,  Open-
       look,  Sunview,  NeXT,  and MS-Windows).   See the list of freely
       available client software in Appendix A.

       The  client interface  differs slightly  on different  platforms.
       However, the queries are performed in the same way,  whatever the

       *   Step 1:  The  user selects a set of databases  to be searched
           from among the available databases.

       *   Step 2:  The user formulates a query by giving keywords to be
           searched for.

       *   Step 3: When the query is run, WAIS asks for information from
           each selected database.

       *   Step 4:  Headlines of documents satisfying the query are dis-
           played.  The  selected documents contain the  requested words
           and phrases.  Selected documents are  ranked according to the
           number of matches.

       *   Step 5:  To retrieve a document,   the user simply selects it
           from the resulting list.

       *   Step 6: If the response is incomplete, the user can state the
           question differently  or feed back to  the system any  one or
           more of the selected documents he finds relevant.

       *   Step 7:  When  the search is run again,  the  results will be
           updated to  include documents which  are similar to  the ones
           selected,  meaning  documents which share  a large  number of
           common words.

       E-mail access:

       You can  query WAIS databases  and retrieve documents  by sending
       commands  in  the  body  part  of  an  e-mail  message  to  wais-
                                                                   wais   The Subject: line is ignored.  The impor-
       tant commands  are (a  vertical bar  (|)  indicates  a choice  of

       help         to get the help file

       maxres number
                    to set the maximum number of results to be returned.

       search source-name | "source-name1 source-name2 ..." keywords

                              is  a   source  name   as  found   in  the
                              directory-of-servers (with or  without the
                              .src ending).   Use double-quotes  (")  to
                              group several sources to be searched.

                    keywords  are the words you would normally type into
                              a query.

                    You may  specify several search  requests in  a mail
                    message.   If  you don't know  what sources  you can
                    search, just try anything. If the source name is not
                    recognised, you'll get a list of sources.

       retrieve docid
                    to retrieve a document from  a database.  docid is a
                    DocID as  returned by a  search above.  You  may put
                    more than one  retrieval request in a  mail message,
                    but you  must leave a  blank line  between requests.
                    The docid must  be written exactly as  returned by a
                    search  request,   including any  spaces.   You  can
                    retrieve non-text documents as well as text.  If the
                    document is  of type TEXT or  WSRC you will  get the
                    result directly. Other types will be UUENCODED.

       DocID: docid same as retrieve. This form is identical to the form
                    which is returned by a  search request.  It makes it
                    easy  to use  the reply  mail  function to  retrieve


       When you log in to the demonstration site at, you
       have immediate  access to  the directory-of-servers  database via
       the swais  client software.  To  find recipes using  papaya,  you
       would select the recipes database and give papaya as the keyword.
       Here are the results of the search:


         #  Score Source                  Title                    Lines
       001: 1000 (recipes) arielle@ta Re: Dawn's Muffins, Pt III     339
       002: 1000 (recipes) arielle@ta Re: Muffins 3                  632
       003: 1000 (recipes) arielle@ta Re: Pineapple                  678
       004:  750 (recipes) arielle@ta Re: Pork and Papaya Salad       33
       005:  750 (recipes) arielle@ta Re: Bread                      681
       006:  500 (recipes) roder@cco. Re: NONFAT BAKERY COLLECTION   423
       007:  500 (recipes) shiva@hoss Re: Juice Recipes               65
       008:  250 (recipes) arielle@ta Re: Prawn Salad                 49
       009:  250 (recipes) arielle@ta Re: COLLECTION: Lots of Avoca  447
       010:  250 (recipes) mecca@acsu Re: REQUEST: blender-made fru   29
       011:  250 (recipes) Ann.Adamci Re: Re: REQUEST: blender-made   38
       012:  250 (recipes) patth@Pani Re: Re: REQUEST: blender-made   49
       013:  250 (recipes) arielle@ta Re: Avocados                   459
       014:  250 (recipes) red_trek@d Re: VEGAN: red beans and rice   78


       You can then select any of  the above documents for viewing,  for
       example, the Pork and Papaya Salad recipe:


       From: (Stephanie da Silva)
       Subject: Pork and Papaya Salad
       Message-ID: <>
       Date: Mon, 29 Mar 1993 06:51:47 GMT
       Lines: 23

       1/4 cup dried currants
       1/2 cup balsamic vinegar
       1/4 cup walnut oil
       1/4 cup chicken broth
       1 tablespoon honey
       1/4 teaspoon ground cinnamon
       1 pound cooked boneless pork loin roast
       1 head Belgian endive
       Bibb lettuce leaves
       2 papayas, seeded, peeled and sliced lengthwise
       2 avocados, seeded, peeled and sliced lengthwise
       1/4 cup broken walnut pieces

       In a small bowl pour enough boiling water over currants to cover.
       Let stand 5 minutes; drain. For dressing, in a screw-top jar
       combine vinegar, oil, chicken broth, honey, and cinnamon. Cover;
       shake well. Trim fat from pork; slice thinly. Separate leaves of
       Belgian endive. Line 6 salad plates with lettuce leaves. Arrange
       pork, endive, papaya, and avocado on plates. Sprinkle with
       currants and walnuts. Drizzle dressing over salads.

       Stephanie da Silva                  


       If you give more than one keyword,  then all documents containing
       any of the keywords will be listed.

       Learning more about WAIS

       A bibliography  of documents,  services  and sources for  WAIS is
       maintained  by   Barbara  Lincoln  Brooks   of  WAIS   Inc.   The
       bibliography  is available  from  in the  directory
       /pub/wais-inc-doc along with many other WAIS documents.

       There are  currently four main  FTP sites for  WAIS documentation
       and software:


       For information on free WAIS software contact

       Mailing list:
       To subscribe send a mail to

       Usenet newsgroup: comp.infosystems.wais

       WAIS was developed at Thinking Machines Corporation.

                                    Part 3

                           FINDING NETWORK RESOURCES


       What is ARCHIE

       Archie is an information system. It offers an electronic directo-
       ry service for  locating information in the  international TCP/IP
       network (the Internet).

       The best known  use of archie is  for scanning a database  of the
       contents of more than 1000 anonymous  FTP sites around the world.
       Currently,  this database contains more than 2,100,000 file names
       from anonymous FTP  sites.  This database is known  as the archie

       The  files made  available at  anonymous FTP  sites are  software
       packages for various systems (Windows, MS-DOS,  Macintosh,  Unix,
       etc.),  utilities,  information or  documentation files,  mailing
       list or Usenet group discussion archives.  At most FTP sites, the
       resources are organized hierarchically  in directories and subdi-
       rectories.  The database  tracks both the directory  path and the
       file names.

       The archie  database is automatically updated,   thereby ensuring
       that the information is accurate.  Using this database, users can
       easily find the  the location of files they  need without logging
       onto several machines.

       Who can use ARCHIE

       Users on any network can access the archie database by electronic
       mail.  Other means of access are available to users on the Inter-
       net (see the section Using ARCHIE below for details).

       You are requested  to respect a few basic rules  when you request
       information from an archie server:

       *   avoid connecting  during working hours;   most of  the archie
           servers are not dedicated machines, they have local functions
           as well.
       *   make your queries as specific as possible;  the response will
           be quicker and shorter.
       *   user  interfaces installed  on  your  computer contribute  to
           reduce the load on the server sites, please use them.
       *   use the archie  server closest to you  and,  in particularly,
           don't overload the transatlantic lines.

       How to get to ARCHIE

       The archie database is maintained in the following locations:

       |                                                               |
       |  Host                       Country        Host               |  
       |                                                               |
       |                  Australia          |  
       | Austria|  
       |        Austria          |  
New Zealand
       |             Canada  |  
       |            Finland     |  
       |     Germany   |  
       |        Great-Britain  |  
       |               Israel     |  
       |            Italy|  
       |  Japan |  
       |          Japan    |  
       |             USA                               |
       |                                                               |

       There are three ways to access  the archie database:  via a local
       client, interactive Telnet session or electronic mail.  Each type
       of access is described below in the Using ARCHIE section.

       Using ARCHIE

       The format of the parameters is given at the end of this section.
       Angle brackets (<>)  indicate an  optional parameter;  a vertical
       bar (|) indicates a choice of parameters.

          Note:   A new version of the archie server (3.0) is now avail-
       able.  Some of the commands for interactive access and the e-mail
       interface are slightly   different from previous versions  of the
       server (2.11 and before).  Command formats marked with a (+)  are
       valid in version 3.0 only, those marked with a (*) are acceptable
       only in previous versions. To find out which version is installed
       at the server you wish to use, issue the version command.

       Using a local client:

       Usage of these clients is encouraged since they provide quick and
       easy non-interactive access to the archie servers, and thus, bet-
       ter performance of  the servers and better response  time for the

       Public domain clients for accessing  archie servers are available
       for:   Macintosh, MS-DOS, OS/2,  VMS,  NeXT,  Unix and X-Windows.
       The clients are available for anonymous FTP from the archie sites
       in the directories /pub/archie/clients  or /archie/clients.   All
       these  platforms  support  a simple  command  line   client.   In
       addition, a graphical interface (called xarchie) is available for

       Archie client command and parameters

       When using a graphical interface, you access the archie functions
       by pressing mousse buttons. The results are displayed with selec-
       table fields for further explorations.

       The basic  archie client  is a command  with parameters  that you
       enter on  your local machine.  With  most versions of  the archie
       client,  if you  type archie with no parameters,  you  will get a
       list of the possible parameters and  a short description of each.
       The format of the command is:

       |                                                               |
       |  archie    <-options> string | pattern                        |
       |                                                               |

       where the options are:

       o         specifies an output file name to store the results (not
                 available with all clients).

       l         lists the result one match per line. This form is suit-
                 able for parsing by programs.

       t         sorts the result inverted by date

       m#        specifies maximum number of matches to return (# within
                 the range 0 to 1000). The default value is 95.

       h archie-server
                 specifies an  archie  server to send the  query to;  if
                 this parameter  is not given,   then the query  will be
                 sent to the default archie server, if one is defined.

       L         lists known servers and current default server.

       The following group of options determines the kind of search per-
       formed on the database. They are mutually exclusive.

       s         a  match occurs  if  the  file/directory name  contains
                 string.  The  search is case insensitive.

       c         as above, but the search is case sensitive.

       e         string must  EXACTLY match (including case)   the file/
                 directory   name   in  the  database.    This   is  the
                 DEFAULT search method.

       r         searches the database using  pattern.  It contains spe-
                 cial characters  which must be interpreted  before per-
                 forming the search.

       There may  be some  slight differences  in the  options available
       with different clients on different platforms.

       The result is a list of FTP site addresses with files or directo-
       ries matching the argument, the size of the file,  its last modi-
       fication date and its directory.  By default,  the list is sorted
       by host address. See the Examples section below for an example of
       archie output.

       Using Telnet:

       To access an  archie server interactively,  telnet to  one of the
       existing servers (see  the list of servers in the  section How to
                                                                  How to
       get to ARCHIE  above).  At the login:  prompt  enter archie,  the
       login procedure ends  leaving the user at a  archie> prompt.  The
       server is  ready for user  requests.  The following  commands are

       exit, quit, bye
                      exits archie.

       help  <command-name>
                      invokes the on-line help.  If issued with command-
                      name, the help request is restricted to the speci-
                      fied topic.    Pressing the RETURN key  exits from
                      the on-line help.

       list  <pattern>
                      provides a list  of the sites in  the database and
                      the time  at which  they were  last updated.   The
                      optional parameter limits the list to sites match-
                      ing pattern.  The result is  a list of site names,
                      sites IP  address and date  of the last  update in
                      the database.   The command  list with  no pattern
                      will list  all sites  in the  database (more  than
                      1000 sites!). Eg,

                           list \.de$

                      will list all German sites

       site(*)  site-name
                      lists the directories and, recursively, the subdi-
                      rectories,   of site-name  in  the database.   The
                      result may be very long.

       whatis  string searches the database of software package descrip-
                      tions for string. The search is case-insensitive.

       prog  string  |  pattern
       find(+)  string  |  pattern
                      searches the database for  string or pattern which
                      represents the name of the resource to be found in
                      the database.  Searches may be performed in a num-
                      ber of  different ways  specified in  the variable
                      search (set command) which also decides the inter-
                      pretation of  the parameter as string  or pattern.
                      The result  is a list  of FTP site  addresses with
                      matching entries,  the size  of the resource,  its
                      last modification  date and the directory  to find
                      it.  The number of hits  is limited by the maxhits
                      variable (set command).  The result of prog can be
                      sorted in different ways,   depending on the value
                      of the sortby variable (set command).  By default,
                      the variables search,  maxhits  and sortby are set
                      to,  respectively,  exact match  search on string,
                      1000 hits and unsorted resulting list.  Typing the
                      keyboard interrupt character during  a search will
                      abort it.   The results up  to that time  are dis-
                      played.   See  the Examples  section below  for an
                      example of the prog command and its results.

       mail  <email>  <,email2...>
                      sends the  result of  the last  command in  a mail
                      message to  the specified e-mail  address(es).  If
                      issued with no argument, the result is sent to the
                      address specified in the variable mailto (set com-

       show  <variable>
                      displays the value of the given variable name.  If
                      issued with  no argument,   it displays  all vari-
                      ables.  See the set command below for the possible

       set  variable  value
                      sets  one of  the archie's  variables.  Values  of
                      these variables  affect how archie  interacts with
                      the user.

       Variables and values are:

       compress(+)  compress-method
                      specifies  the compression  method  (none or  com-
                      press) to be used before mailing a result with the
                      mail command. The default is none.

       encode(+)  encode-method
                      specifies the  encoding method (none  or uuencode)
                      to be used  before mailing a result  with the mail
                      command.   This variable is ignored if compress is
                      not set. The default is none.

       mailto  email  <,email2 ...>
                      specifies  the  e-mail address(es)   to  mail  the
                      result of  the last  command when  mail is  issued
                      with no arguments.

       maxhits  number
                      specifies the maximum number  of matches prog will
                      generate (within the range 0 to 1000). The default
                      value is 1000.

       search  search-value
                      determines  the kind  of search  performed on  the
                      database by  the command:  prog string  | pattern.
                      search-values are:

                      sub    a partial  and case  insensitive search  is
                             performed with string on the database, eg:

                                  "is" will match "islington" and "this"
                                  and "poison"

                             as above but the  search is case sensitive,

                                  "TeX" will match "LaTeX" but not

                      exact  the parameter of prog (string) must EXACTLY
                             match (including  case)  the string  in the
                             database. The fastest search method of all,
                             and the default.

                      regex  pattern is interpreted  before performing a
                             search on the database.

                      sortby  sort-value
                             describes how  to sort the result  of prog.
                             sort-values are:

                             hostname on the FTP site address in lexical

                             time     by  the modification  date,   most
                                      recent first.

                             size     by the size of  the found files or
                                      directories, largest first.

                             filename on file or directory name in lexi-
                                      cal order.

                             none     unsorted (default)

                             The  reverse  sorting   orders  from  those
                             described here are obtained by prepending r
                             to the  sortby value given.   (eg,  reverse
                             hostname order hostname is rhostname).

                      term  terminal-type   <number-of-rows  <number-of-
                             tells the archie server what type of termi-
                             nal you are using,  and optionally its size
                             in rows and columns, eg:

                                  set term xterm 24 100

       Using electronic mail:

       Users  limited to  electronic mail  connectivity  can access  the
       archie servers  by sending mail to  the domain address of  one of
       the  servers listed  in the  section How  to get  to ARCHIE  (eg,  The commands are sent  in the body part of
       the mail.

       The electronic  mail  interface to an archie  server recognizes a
       subset of the  commands described  in Using  Telnet.  Most useful
       commands  and   particularities  to  the  e-mail   interface  are
       described below.  If an empty message, or a message containing no
       valid requests is  received,  it will be considered to  be a help

       Command lines begin  in the first column.  All lines  that do not
       match a valid command are ignored. The Subject: line is processed
       as if it were part of the message body.

       help      sends you the help file. The help command is exclusive,
                 ie, other commands in the same message are ignored.

       path  return-address
       set mailto(+)  return-address
                 specifies a  return e-mail address different  from that
                 which is extracted from the  message header.  If you do
                 not receive a reply from  the archie server within sev-
                 eral hours,   you might need to  add a path  command to
                 your message request.

       list  pattern  <pattern2 ...>
                 provides a list of the sites in the database that match
                 pattern and the  time at which they  were last updated.
                 The result is a list with site names,  sites IP address
                 and date of the last update in the database.

       site(*)  site-name
                 lists the directories and, recursively, the subdirecto-
                 ries, of site-name in the database.

       whatis  string  <string2 ...>
                 searches the database of software packages descriptions
                 for each string. The search is case insensitive.

       prog  pattern  <pattern2 ...>
       find(+)  pattern  <pattern2>
                 searches the  database for each interpretation  of pat-
                 tern  which represents  the name  of a  resource to  be
                 found in the database.  If multiple patterns are placed
                 on one line,  in that case,  the results will be mailed
                 back in  one message.  If  multiple prog  lines appear,
                 then multiple messages will be  returned,  one for each
                 prog line.  Results  are sorted by FTP  site address in
                 lexical order.  If pattern contains spaces,  it must be
                 quoted with single (') or double (") quotes. The search
                 is case insensitive.

                 causes the  result of  the current  request to  be com-
                 pressed and uuencoded.  When you receive the reply, you
                 should run it through uudecode.  This will produce a .Z
                 file.  You can then run uncompress on this file and get
                 the result of your request

       set compress(+)  compress-method
                 specifies the compression method (none or compress)  to
                 be  used  before  mailing the  result  of  the  current
                 request. The default is none.

       set encode(+)  encode-method
                 specifies the encoding method (none or uuencode)  to be
                 used before mailing the result  of the current request.
                 This variable is  ignored if compress is  not set.  The
                 default is none.

                    Note:  set compress compress and set encode uuencode
                 would produce  the same result  as the  former compress

       quit      nothing past this point is  interpreted.  Useful when a
                 signature is automatically appended at  the end of your
                 mail messages.

       Description of pattern

       A pattern describes a character string including characters which
       take a special meaning.  The special  meaning is lost when "\" is
       put before the character. The special characters are:

       .         (period)    this   is  the   wildcard   character  that
                 replaces any other character, eg, "...." will match any
                 4 character string.

       <circ>    (caret)  if  "<circ>"  appears at the beginning  of the
                 pattern,  then the searched string  must start with the
                 substring following the "<circ>". If it occurs anywhere
                 else in the pattern it is regarded as non-special, eg:

                      "<circ>efghi" will match "efghi" or "efghijlk" but
                      not "abcefghi"

       $         (dollar) if "$" appears at the end of the pattern, then
                 the searched string must end with the substring preced-
                 ing the "$". If occurring anywhere else in the pattern,
                 it is regarded as non-special, eg:

                      "efghi$" will match "efghi" or "abcdefghi" but not


       If you are using an archie client, and enter the command:

            archie -s eudora

       or if you send,  by e-mail or  during a Telnet session,  the com-

            prog eudora


            find eudora

       then archie will send you the following results:

       Host    (
       Last updated 03:38  8 Aug 1993

        Location: /pub/MAC
          DIRECTORY  drwxrwxr-x 2048 bytes 00:00  6 May 1992  eudora

       Host    (
       Last updated 03:38  8 Aug 1993

        Location: /pub/MAC/eudora
          FILE  -r--r--r-- 281139 bytes 00:00 21 Oct 1991  eudo-

       Host    (
       Last updated 04:53  9 Aug 1993

        Location: /pub/NetNews/comp.binaries.mac
          FILE  -rw-r--r-- 438 bytes 12:04 10 Jul 1993  Eudora1.3.readme

       Host    (
       Last updated 04:53  9 Aug 1993

        Location: /pub/NetNews/comp.binaries.mac
          FILE  -rw-r--r-- 278912 bytes 12:04 10 Jul 1993  Eudo-


       If you send the command list \.de$  by e-mail or in a Telnet ses-
       sion, then you will get the following results:  12:31  8 Aug 1993  12:25  8 Aug 1993   15:21  6 Aug 1993     00:18 31 Jul 1993  12:10  8 Aug 1993        16:08 31 Jul 1993


       If you send the command whatis compression by e-mail or in a Tel-
       net session, then you will get the following results:

       RFC 468      Braden, R.T. FTP data compression 1973 March 8; 5p.
       arc          PC compression program
       deltac       Image compression using delta modulation
       spl          Splay tree compression routines
       squeeze      A file compression program
       uncrunch     Uncompression program
       unsqueeze    Uncompression programs

       Learning more about ARCHIE

       However you communicate with the  archie server,  on-line help is

       If you  have any  questions about  archie,  write  to the  Archie
       Group, Bunyip Information Systems Inc. at

       Bug reports,  comments,  suggestions,  etc.   should be mailed to  In addition, the database administrator
       at   a   particular   archie   server   can   be   contacted   at
       archie-admin@address.of.archie.server,                        eg:

       Mailing list:
       To subscribe send a mail to:

       Archie was developed by Alan Emtage, Peter Deutsch, and Bill Hee-
       lan from  the McGill University  Computing Center,   Canada.  Now
       archie is supported by Bunyip Information System Inc., Canada.

                                    Part 4

                         FINDING PEOPLE AND COMPUTERS


       What is WHOIS

       WHOIS provides directory service to  network users.  This service
       is a way of finding e-mail addresses,  postal addresses and tele-
       phone numbers.   It may also  deliver information about networks,
       networking organizations, domains and sites.

       The  main database  of  networking-related names  (organizations,
       sites, networks, people, etc.) is maintained by the Internet Reg-
       istration Service (InterNIC). Actually, the names of the adminis-
       trative and technical  contacts for registered domains  are auto-
       matically  entered into  the database  when domain  or IP  number
       applications are processed by  the Internet coordination authori-
       ty.   Each entry of the database has a handle (a unique identifi-
       er), a name, a record type, and various other fields depending on
       the type of record.   This database will be used as an example in
       the descriptions below.

       Before April 1, 1993, the Network Information Center (NIC) of the
       Defense Data Network (DDN) was the Internet coordination authori-
       ty and,   therefore,  maintained the  database (known as  the NIC
       database).  The  NIC database  is now  restricted to  information
       about the .mil domain. Many documents still refer to these names.

       Many academic sites maintain their own database to offer informa-
       tion about their staff members and students.

       In its current implementation,  WHOIS  has some limitations which
       prevent it  from becoming  an efficient  directory service  for a
       large volume of  information and numerous requests:   the various
       WHOIS servers  have no  knowledge of each  other,  a  database is
       maintained at each server site, and, finally, new functionalities
       have been implemented locally at  various sites and not propagat-
       ed.  A  new extended protocol,   WHOIS++,  is being  specified to
       improve the current service.  WHOIS++ will include local enhance-
       ments to  the WHOIS  service,  an improved  query syntax  and its
       architecture will allow a real  distributed directory service for
       the entire Internet.

       This new protocol  for directory services will  be made available

       Who can use WHOIS

       WHOIS is available  to users on the  international TCP/IP network
       (the Internet).

       A WHOIS server is accessible across  the network from a user pro-
       gram running on local machines or  via an interactive Telnet ses-
       sion to the site which hosts the server.

       In addition,  the InterNIC offers an electronic mail interface to
       the database it maintains,  allowing users not on the Internet or
       users with electronic mail only to access this information.  This
       type of access is described below in the  Using WHOIS section.

       In general,  WHOIS servers should only be used for isolated quer-
       ies about specific information.  Typically,  it is not acceptable
       to make an extended series of queries to obtain large sections of
       the directory.   Such a strategy is unfair both because of exces-
       sive consumption of server resources,   and because the directory
       information belongs  to individuals.  In  particular,  extracting
       lists of people for commercial purposes is strictly prohibited.

       How to get to WHOIS

       There are many  WHOIS servers  throughout the network  and a com-
       prehensive list   would be  too long   to  be included   here.  A
       WHOIS server  offers information about  the organization to which
       it belongs:   it doesn't  share a   common directory   with other
       WHOIS servers  and doesn't  know either where to find information
       about other institutions.

       Using WHOIS

       WHOIS  has become  the  familiar name  of  the  user program  for
       accessing a  WHOIS database,   although NICNAME  is the  original

       In the following, angle brackets (<>) indicate an optional param-

       Using a local client:

       Unix computers have a native whois command. On non-Unix machines,
       ask your  system administrator  whether your  computer has  it or
       not. This command searches the database on the specified site for
       entry which contains identifier. The format is:

       |                                                               |
       |  whois     <-h site-name> identifier                          |
       |                                                               |


       site-name is the domain address of the site which hosts the data-
                 base you  want to query (eg,  On
                 some installations,  the default value  is still set to
                 the old NIC database site (

                 is a name (person, host, domain or network), an IP num-
                 ber or a handle.

       Special  characters may  be  used in  identifier  to specify  the

       .         before identifier will cause a name-only search.

       !         before identifier will cause a handle-only search.

       ... or .  after identifier  will cause a partial  search:  every-
                 thing starting with identifier will match.

       @         in  identifier  will  cause  a  search  on  the  e-mail

       *         before  identifier will  return  the entire  membership
                 list of the entry that  matches identifier (eg,  a site
                 and its registered users).

       %         before identifier will return  only the membership list
                 of the entry  that matches identifier (eg,   the regis-
                 tered users of a site).

       The special characters may be used together.

       The results are displayed in one of 2 ways:

       *   a full detailed display for a single match,
       *   a list of summary lines for multiple matches.

       In both cases, the handle is shown in parentheses after the name.

       Using Telnet:

       To access  the InterNIC  database interactively,   telnet to  the
       InterNIC site ( No login is required.

       Other WHOIS databases may have a  Telnet access and offer most of
       the functions  below (eg,  which hosts  the WHOIS
       database of the European IP Networks).

       In the following,  CAPITAL  letters indicate acceptable abbrevia-
       tion; angle brackets (<>) indicate an optional parameter.

       WHOIS        invokes the information retrieval program.

       ?            displays a short on-line help.

       HElp         accesses the full on-line help.

       Q, QUIT, RETURN key
                    exits WHOIS

       <keyword> identifier
                    searches the  database for  an entry  which contains
                    identifier.   The  default action is  to do  a broad
                    search, looking for matches in many fields:  handle,
                    name, nicknames, hostname, IP number, etc, and find-
                    ing all record types.  keyword may be used to narrow
                    the search to a specific record type.

       keyword may be one of:

       PErson       limits the search to persons.

       DOmain       limits the search to domains (eg, DO EARN.NET).

       HOst         limits the search to hosts (eg, HO PRINCETON).

       NEtwork      limits the search to networks (eg, NE EBONE).

       Organization limits the search to organizations (eg, O CREN).

       NAme         same as leading '.' in identifier.

       HAndle       same as '!' in identifier.

       PArtial      same as trailing '.' in identifier.

       Mailbox      same as '@' in identifier.

       EXPand       same as '*' in identifier.

       SUBdisplay   same as '%' in identifier.

       Full or '='  shows detailed display for each match.

       SUMmary or '$'
                    shows summary always, even if just one match.

       Special  characters may  be  used in  identifier  to specify  the

       .            before identifier will cause a name-only search.

       !            before identifier will cause a handle-only search.

       ... or .     after  identifier  will  cause  a   partial  search:
                    everything starting with identifier will match.

       @            in  identifier will  cause a  search  on the  e-mail

       *            before identifier will return  the entire membership
                    list of the entry that match identifier (eg,  a site
                    and its registered users).

       %            before identifier  will return  only the  membership
                    list of the entry that  matches identifier (eg,  the
                    registered users of a site).

       ~            before identifier will return the entry that matches
                    identifier only, no membership list.

       The special characters may be used  together.   Except if Full or
       SUMmary  are  specified,  the results are  displayed in one  of 2

       *   a full detailed display for a single match,
       *   a list of summary lines for multiple matches.

       In all cases, the handle is shown in parenthesis after the name.

       Using electronic mail:

       Users limited to  electronic mail connectivity can  send requests
       to the  database maintained  at the InterNIC  by sending  mail to   The  commands are  sent in  the Subject:
       field.  The body part  of the mail is ignored except  if the Sub-
       ject:   line  is empty.   In that  case,  only the first  line is

       This electronic mail interface  recognizes all commands described
       in Using Telnet. Requests should be prefixed with the word WHOIS.
       Requests are processed automatically once a day.


       If you are using a local client, and enter the command:

            whois \!EARN...       (remark: "\" is an escape character)

       or if you send by e-mail the command:

            whois !EARN...

       then you will get the following results:

            EARN (EARN-HST)   SEINE.EARN.NET      
            European Academic Research Network (EARN-DOM)       EARN.NET

       To obtain detailed information on the second item,  enter or send
       the command:

            whois EARN-DOM

       then you will get the following result:

            European Academic Research Network (EARN-DOM)
               EARN Office
               CIRCE BP 167
               91403 ORSAY CEDEX, France

               Domain Name: EARN.NET

               Administrative Contact:
                  Bovio, Daniele  (DB355)  hi@FRORS12.CIRCE.FR
                  +33 1 6982 3973
               Technical Contact, Zone Contact:
                  Grange, Nadine  (NG4)  grange@FRORS12.CIRCE.FR
                  +33 1 69823973 (FAX) +33 1 69285273

               Record last updated on 19-Aug-93.

               Domain servers in listed order:


       For a partial search, enter:

            whois hi@f...

       then you will get the following result:

            Bovio, Daniele (DB355)  hi@FRORS12.CIRCE.FR
               EARN Office
               CIRCE BP 167
               91403 ORSAY CEDEX, France

               +33 1 6982 3973

               Record last updated on 23-Oct-92.

       Learning more about WHOIS

       The WHOIS service  is documented in an Internet  Request For Com-
       ments (RFC 1400).

       If   you    have   any   questions    about   WHOIS    write   to

       Bug reports,  comments,  suggestions,  etc.   should be mailed to


       What is X.500

       X.500 is an OSI (Open System Information)  based directory servi-
       ces protocol designed  by the CCITT (International  Telegraph and
       Telephone Consultative Committee).

       X.500 provides  distributed directory services to  network users.
       The X.500  directory specifies a  model for  connecting directory
       services to form one distributed global directory. Each directory
       service holds  a part  of the global  database and  the directory
       information is  made available via  a server (called  a Directory
       System Agent - DSA). The database is maintained locally. From the
       user point of view,  the entire  directory is accessible from the
       local server.

       While most of the information available  today via X.500 is about
       people and organizations,   the design of the  X.500 directory is
       also suitable  for storing information  about other  entities (or
       objects),  such as network  resources,  applications or hardware.
       Several projects are underway which utilise these directory capa-
       bilities (eg, the Internet RFCs (Request For Comments) are listed
       in the global directory).

       Each item  (entry)  in the  X.500 directory describes  one object
       (eg,  a person,  a network resource,  an organization)  and has a
       Distinguished Name - DN (a unique  identifier).  It consists of a
       collection  of attributes  (eg,  last  name,  organization  name,
       e-mail,...- for  a person).   The information  held in  the X.500
       directory (or Directory Information Base - DIB) is arranged hier-
       archically. This organization is called the Directory Information
       Tree (DIT).  At the top-level is the root entry (the World), then
       the country level, then the organization level, and,  eventually,
       the people, the resources, etc., at the bottom-level of the hier-

       Who can use X.500

       Although X.500 is part of the OSI standard definition, OSI access
       is not necessary to use the directory services. Many X.500 servi-
       ces are available on the Internet. In addition, users on any net-
       work can access the X.500 directory  by electronic mail.  See the
       section Using X.500 below for details.

       How to get to X.500

       There are three  ways to access the X.500 services:   via a local
       client, interactive session (Telnet or X.25 access) or electronic
       mail.  Each type of access is  described below in the Using X.500

       In addition,  other network tools  (eg,  WWW and Gopher)  provide
       access to X.500 directory services through gateways.

       Accessing a  remote client is an  easy way to start  querying the
       X.500 directory.   Some sites allow  public access via  Telnet or
       X.25 to a client. Public access user interfaces are available at:

       |                                                               |
       |  Telnet (login)               Public X.25 (login)  Country    |
       |                                                               |
       | (fred)                        Australia  |
       | (dua)        222100611            Belgium    |
       | (ds)                               Denmark    |
       | (dua)                                Finland    |
       |                               20800603053201       France     |
       |                    (login: dua, password: ucom.x)  France     |
       |                               26245050230303       Germany    |
       | (de)                               Ireland    |
       | (de or fred) 22225010083212     Italy      |
       | (zoek)                        Netherlands|
       | (de or dish)                Poland     |
       | (directorio)  2142160234013      Spain      |
       | (de)      240374810306        Sweden     |
       | (dua)           22847971014540      Switzerland|
       | (dua)     23421920014853      Paradise   |
       |                                                               |

       Paradise is a  European project to encourage the use  of X.500 in
       European countries.

       To connect to one of these sites, select an access method (Telnet
       or X.25)  and at the login:  prompt type the specified login,  if

       Using X.500

       X.500 supports data management functions (addition,  modification
       and deletion of entries)  and  powerful lookup capabilities.  The
       use of X.500 is primarily for its lookup capabilities, ie, query-
       ing a database for information on a person (postal address, tele-
       phone number, e-mail address, etc.). The basic fields for search-
       ing are the person's name,  the name of the person's organization
       (and department within the organization) and the country.

       In the following, angle brackets (<>) indicate an optional param-
       eter; a vertical (|) indicates a choice of parameters.

       Using a local client:

       In the  X.500 world,  a local  client is called a  Directory User
       Agent (DUA)  Public domain and  commercial DUAs are available for
       numerous platforms ranging from mainframes to personal computers.
       For a comprehensive list of DUAs,  their description and where to
       find them,  consult the Internet document RFC 1292 - A Catalog of
       Available X.500 Implementations.

       Available DUAs range  from simple line commands  to sophisticated
       graphical user interfaces which require a pointing device.

       Using Telnet or X.25:

       3 categories of user interfaces might  be available at the remote

       *   line-oriented: de, dish, fred
       *   menu-driven: sd (formerly known as widget)
       *   X-Windows-based: Xdi, Xlookup (or xlu), pod

       Capabilities of these DUAs range  from basic search facilities to
       full X.500 functionality.

       de (directory enquiries) is recommended for novice users since it
       is a very simple user-interface. It has been designed to run as a
       public access DUA and is accessible from any kind of terminal. It
       supports the basic X.500 functions: read, search, list.  The Sim-
       ple query mode is suitable for those  who are new to querying the
       X.500 directory.

       de        invokes the X.500 interrogation user-interface.

       q         exits de.

       ?<topic>  displays the  on-line help  on the  specified topic  or
                 general help.

       <circ>C   (Ctrl-C) is the interrupt character. It aborts a search
                 in progress or resets the current query specification.

       *         (asterisk)   will list  all  entries  of the  specified
                 field.   It is  also  the  wildcard character  and  can
                 replace any other character in  a name.  It can appears
                 anywhere in the name,  eg:    smit* or *smit* are valid
                 string formats.

       -         resets the default value to a blank string.

       When de is invoked,  the user is requested to fill in 4 fields to
       specify a request.   In all fields,  the value  from the previous
       request is the default value.  Press the RETURN key to accept it,
       or enter a new value. All searches are case insensitive.

       The four fields to be filled in are:

       Person's name
                 Wildcard characters may  be used anywhere in  the name.
                 All matching names will be listed. Typing only "*" will
                 match all people of the specified department or organi-
                 sation.   If this field is blank, the search will be on
                 department or organization only.

       Department name
                 the  name (or  an acronym)   of the  department in  the
                 organization where the person  works.  Wildcard charac-
                 ters may be used anywhere in the name.  Typing only "*"
                 will match  all departments.  If  no person's  name has
                 been entered,  details on the department are displayed,
                 otherwise,  the search is carried out with the selected
                 name.  If no department name is given,  all departments
                 will be searched.  This field could be omitted in small

       Organization name
                 the name (or an acronym)  of the organization where the
                 person works.  Wildcard characters may be used anywhere
                 in the name.  Typing only  "*" will match all organiza-
                 tions.  If no person's name or department name has been
                 entered,  details  on the  organization are  displayed,
                 otherwise,  the search is carried out with the selected

       Country name
                 the name of the country where the person works.  Typing
                 "*" will list all countries.  The country name could be
                 the 2-letter country code (eg,  DK stands for Denmark),
                 the name or a part of it without wildcards (eg,  nether
                 instead of The Netherlands).

       If a large number of matching entries are found,  they are listed
       so that the user can select one entry to get further details.

       Using electronic mail:

       The Norwegian networking organization (UNINETT)  offers an e-mail
       interface to X.500.  To use it, send a mail message to:  Directo-
       ry@UNINETT.NO  with the word find in the Subject: field. The body
       part contains the search request, one per message.

       The format of the search request is:

       |                                                               |
       |  find      <person-name>  <: org-name  <;  country-name>>  |  |
       |            <; country-name>                                   |
       |                                                               |

       If org-name and country-name are omitted,  the sender's organiza-
       tion name and country name are  used as default values.  The mail
       interface guesses these values from the From: field of your mail,
       so the results can be surprising  if your address ends with .bit-

       "*" (asterisk) is the wildcard character and can replace any oth-
       er characters in any name. It can appear anywhere in the name.

       The result  of the  query is sent  back in  a mail  message.  The
       search is case insensitive.

          Note:   To avoid overloading the directory service,  users are
       not allowed to search for a person without selecting an organisa-
       tion.   To receive a help file,  send  the word help instead of a
       find command.


       Using de,  you  can search for the Anthropoloy  department of the
       University College of London in United Kingdom,  with the follow-
       ing request:

            Person's name, q to quit, * to browse, ? for help
            Department name, * to browse, ? for help
            :- a*
            Organisation name, * to browse, ? for help
            :- ucl
            Country name, * to browse, ? for help
            :- uk

       A few entries match the selected  department,  all are listed for
       further selection:

            United Kingdom
              University College London

            Got the following matches.  Please select one from the list
            by typing the number corresponding to the entry you want.

            United Kingdom
              University College London
                  1 A.U.T. Office
                  2 American Institute for Foreign Study
                  3 Anatomy and Developmental Biology
                  4 Anthropology
                  5 Audio Visual Centre
            Department name, * to browse, ? for help
            :- 4
            United Kingdom
              University College London
                    Telephone Number      +44 71-387-7050 x2455
                    fax                   +44 71 380 7728

       If you are  looking for Erik Lawaetz from UNI-C  in Denmark,  you
       can enter the following request (default  values come from a pre-
       vious request):

            Person's name, q to quit, * to browse, ? for help
            :- law*
            Department name, * to browse, <CR> to search all depts, ?
            for help
            Organisation name, * to browse, ? for help
            :- uni-c
            Country name, * to browse, ? for help
            :- dk

       One entry matches the selected criteria, details are displayed:

                  Erik Lawaetz
                    postalAddress         UNI-C
                                          Bygning 305
                                          DK-2800 Lyngby
                    Telephone Number      +45 45 93 83 55
                                          +45 42 88 39 99 x2018
                    fax                   +45 45 93 02 20
                    electronic mail

       If you send mail to   Directory@UNINETT.NO   with the request:

            find geir ped* : *oslo ; no

       you'll get the following result:


       This message is in response to your request to the directory to

                                geir ped* : *oslo ; no

       This is  interpreted as a  request to find  a person with  a name
       matching  "geir  ped*"  in  an organisation  with  name  matching
       "*oslo" in a country with a name matching "no".

       There   were   8  organisations   with   a   name  matching   the
       organizational  name you  specified.  Within those  organisations
       there were 7  persons that had a name matching  the personal name
       you specified.  Directory information for the  located persons is
       shown below.

       Geir Pedersen : Universitetet i Oslo ; Norway

         Alternate        Geir Kenneth Pedersen
         Alternate        Geir K. Pedersen
         E-Mail (RFC)
         E-Mail (X.400)   /G=geir/S=pedersen/OU=usit/O=uio/PRMD=uninett/
                          ADMD= /C=no/
         Postal Address   Postboks 1059 - Blindern
                          0316 Oslo 3
         Phone            +47-22-852478
         Phone            +47-22-852470 (front-office)
         Fax-phone        +47-22-852730
         Description      Project leader for UNINETTs X.500 projects
         User ID          geirp
         Favorite Drink   Farris
         Street Address   Gaustadalleen 23
         Home Address     Gaustadveien 17A
                          0372 Oslo 3
         See also         Geir Pedersen : UNINETT ; Norway
         Entry updated    Tue Jun 15 11:51:31 1993


       and 6 other entries.

       Learning more about X.500

       Several Internet RFC documents deal with X.500:

       RFC 1292  A Catalog of Available X.500 Implementations.

       RFC 1308  Executive Introduction to Directory  Services Using the
                 X.500 Protocol,

       RFC 1309  Technical  Overview  of Directory  Services  Using  the
                 X.500 Protocol,

       The official source  of information on X.500 is  the X.500 recom-
       mendation published by the CCITT (Blue Book, Volume VIII - Fasci-
       cle VIII.8,  Data Communication  Networks Directory,  Recommenda-
       tions  X.500-X.521,  CCITT,   1988,   ISBN 92-61-03731-3).   This
       document is also available electronically:   send the command GET
       ITU-5233  to   This is not intended for the casual


       What is NETFIND

       NETFIND is an Internet user directory tool.  It provides a simple
       Internet white pages directory facility.

       Given the name of  a person on the Internet and  a rough descrip-
       tion of where the person works,  Netfind attempts to locate tele-
       phone and  electronic mailbox information  about the  person.  It
       does so using  a seed database of  domains and hosts in  the net-

       Netfind finds information about people  through the Internet pro-
       tocols SMTP and finger.  If the person  being sought is at a site
       that is not directly connected to the Internet (e.g., the site is
       connected  only  through  a mail  forwarding  gateway),   Netfind
       informs the user that the person can not be found.

       Due to the dynamic nature of Netfind's search procedures and var-
       iations  in  Internet  availability,  different  results  can  be
       obtained for the same search on different occasions.

       The Netfind  software can only run  on Suns running SunOS  4.0 or
       more recent.

       Who can use NETFIND

       You must be on the international TCP/IP network (the Internet) in
       order to use Netfind. Moreover, Netfind can only find information
       on users who are on the Internet.

       There is no e-mail access to Netfind.

       How to get to NETFIND

       You  can access  Netfind  through software  at  your site  (local
       access), or you can use Telnet to access it at one of the follow-
       ing hosts (remote access):

       |                                                               |
       |  Host                   Country    Host                  Country
       |                                                               |
       |              Australia USA  |
       |        Venezuela        USA  |
       |     Singapore     Canada
       |     Chile  England
       | USA        USA  |
       |       Czech Rep. nic.nm.k              Korea|
       |           Slovakia   USA  |
       |                                                               |

       Using NETFIND

       To use Netfind,   you give it the  name of a person  and keywords
       indicating where that person works.   After you specify a search,
       Netfind looks in  its seed database to find  domains matching the
       specified keywords.  If  there is more than  one matching domain,
       Netfind displays the  list of matching domains,  and  asks you to
       select up  to three to search.   If the keys you  specified match
       more than  100 domains,  Netfind will  list some of  the matching
       domains/organizations and ask you to form a more specific search.
       You can use any of the parts of an organization's name (or any of
       the components of  its domain name)  as keys  in searches.  Using
       more than one key implies the logical AND of the keys. Specifying
       too many keys may cause searches to fail.

       When it  completes the search  (or when interrupted  by <circ>C),
       Netfind summarizes the search results. The summary includes prob-
       lems searching remote domains, information about the most promis-
       ing email address for the person being sought (if available), and
       information about when and where the person most recently/is cur-
       rently logged  in (if  available).  If  more than  one person  is
       located by  a search,  the  summary does not  include information
       about email  targets and most  recent/current logins  (since only
       the user can decide which person was the correct one.)

       Local access:

       The format of the Netfind command is:

       |                                                               |
       |  netfind   <options>  name-keyword  place-keywords            |
       |                                                               |

       where the options are:

       -h        tells Netfind  to skip  the domain  search phase,   and
                 immediately begin  searching individual  machines found
                 in the seed database.  This  option exists for measure-
                 ment purposes. It is not of much use to casual users.

       -s        will disable usage of the  SMTP protocol during search-
                 es.  This option is mainly  useful for measurement pur-
                 poses.  Without this protocol, searches will begin pro-
                 ducing finger  output slightly sooner,  but  will often
                 search less  useful machines,   generate more  Internet
                 load,  and fail to find  information for users at sites
                 that do not support finger (such as many companies).

       -t        will report how many timeouts  occurred.  The -T option
                 will set the  timeout interval to the  specified number
                 of seconds.   It may be necessary to use this option to
                 increase the timeout value for intercontinental search-

       -D        sets the  maximum number of  domains that  Netfind will
                 search at once.   The default is  3.  While it may seem
                 convenient to set a high value for this number, we sug-
                 gest you do not do this.  The search will actually pro-
                 ceed faster (and  waste less Internet bandwidth)   if a
                 small number of well chosen domains are searched.

       -H        sets  the  maximum  number of  machines  that  will  be
                 searched by Netfind.   The default value is 50.  Again,
                 we suggest that you do not set this value higher.

       -m        displays  measurement   information.   If  no  filename
                 is  specified,  measurements are output to stderr.  The
                 packet count estimates are usually exaggerated, because
                 they make  pessimistic assumptions  about the  state of
                 the Domain Naming System.

       -d        allows you to turn on various classes of debugging out-
                 put (all of which are output to stderr), using a letter
                 corresponding to each one.  Debugging output is enabled
                 using the -d option with a list of letters, eg,  -dslf.
                 The following classes/letters exist:

                 c:        display control  messages (check if  the pro-
                           gram has reached a specified point)

                 f:        display finger related messages

                 h:        list machine names found in the seed database

                 l:        display lock related  messages (when entering

                 m:        display messages about mail protocol (SMTP)

                 n:        display messages about network failures

                 r:        display hosts matched from  the seed database
                           that were  rejected from searches  because of
                           search scope selection

                 s:        display system call related messages

                 t:        display thread related messages

                 A:        convert the above flags to mean their comple-
                           ment (example  -dAt means  produce all  debug
                           output except that for threads).

       The letters that are most likely to  be of interest to the casual
       user are f,  m,  and n.   By default,  these options are enabled.
       Specifying any of  these three flags with -d on  the command line
       will  disable them  (hence,  the  -d option  toggles the  default
       behavior of each of the flags).

       The  name keyword  specifies the  person being  sought by  first,
       last, or login name (only one name can be specified).

       The place keywords describe where the person works, by either the
       name of the  institution or the city/state/country.   If you know
       the institution's domain name (e.g.,, where there
       are host names like you can specify it as
       keys without the dots (eg,  cs colorado edu).   The host parts of
       domain names (brazil) cannot be used as keywords.   Keys are case
       insensitive and may be specified in  any order,  although using a
       very common key (like university) first will cause internal buff-
       ers to overflow and some domains to be missed.

       Using more  than one  key implies  the logical  and of  the keys.
       Specifying too many keys may cause searches to fail. If this hap-
       pens, try specifying fewer keys.

       Remote access:

       Telnet to  one of  the remote Netfind  sites (see  How to  get to
       NETFIND, above) and log in as netfind.  No password is necessary.
       You will get the following menu:

            Top level choices:
                    1. Help
                    2. Search
                    3. Seed database lookup
                    4. Options
                    5. Quit (exit server)

       If you select Search, you will be given an opportunity to enter a
       name keyword and place keywords.


       To find the address of e-mail address of Nadine Grange, who works
       at the EARN  office at CIRCE in  France,  you could try  the key-

            nadine circe france

       Since there are  more than three domains that fit  the place key-
       words,  you are asked to pick a few.  The search proceeds,  using
       the domains of your choice:


       Please select at most 3 of the following domains to search:
       0. (centre national de la recherche scientifique, orsay ce
       1. (centre inter-regional de calcul electronique, c
       2. (centre national de la recherche scientifique, ors
       3. (centre national de la recherche scientifique,
       4. (centre national de la recherche scientifique, or
       5. (centre national de la recherche scientifique, ors
       6. (centre national de la recherche scientifique, orsa
       7. (centre national de la recherche scientifique, orsa
       8. (centre national de la recherche scientifique,
       9. (centre national de la recherche scientifique,
       10. (centre national de la recherche scientifi

       Enter selection (e.g., 2 0 1) --> 0
       ( 1) check_name: checking domain  Level = 0
       Search of domains completed.  Proceeding to search of hosts.
       ( 3) check_name: checking host  Level = 0
       ( 4) check_name: checking host  Level = 0
       ( 5) check_name: checking host  Level = 0
       ( 1) check_name: checking host  Level = 0
       ( 2) check_name: checking host  Level = 0
       ( 1) do_connect: Finger service not available on host rsovax.circe.
       ( 1) check_name: checking host  Level = 0
       ( 4) check_name: checking host  Level = 0
         Login name: nadine                    In real life: Nadine Grange
         Directory: /home/nadine               Shell: /bin/csh
         On since Sep  7 08:48:05 on ttyp0     4 days 21 hours Idle Time
         New mail received Sun Sep 12 00:00:08 1993;
           unread since Fri Sep 10 11:53:17 1993
         No Plan.

         Login name: nadine                    In real life: Nadine Grange
         Directory: /home/nadine               Shell: /bin/csh
         On since Sep  7 09:17:09 on ttyp6     1 day 12 hours Idle Time

       - "nadine" is currently logged in from, since Sep  7 09:17:09.
       - The most promising email address for "nadine"
         based on the above search is


       Note  that Netfind  found  only an  Internet  address  on a  Unix

       Learning more about NETFIND

       The remote  access version of Netfind  has a large  Help section.
       There is also a set of  frequently asked questions available with
       the software release, in the Doc directory.  These questions cov-
       er Functionality, Methodology, Network and Remote Site Load, Pri-
       vacy, Future Directions, and Related Work.

       A noteworthy article on Netfind is:

       Experience  with a  Semantically Cognizant  Internet White  Pages
       Experience  with a  Semantically Cognizant  Internet White  Pages
       Directory Tool, by M. F. Schwartz and P.  G.  Tsirigotis, Journal
       of  Internetworking Research  and Experience,   March 1991,   pp.

       This publication discusses the research principles,  performance,
       and scope  measurements of Netfind,   and compares it  with other
       white pages facilities.

       There is a  mailing list for Netfind users  (for software updates
       and other discussions).  To be added  to the list,  send an email
       message  to  with the  body
       (not subject line) subscribe netfind-users

                                    Part 5

                                 GETTING FILES


       What is TRICKLE

       TRICKLE is a service  which will send you files on  request or by
       subscription.   TRICKLE works  with various  anonymous FTP  sites
       (computers in the  Internet network that allow  public access and
       retrieval of software and files).  It  provides a  quick and easy
       alternative to FTP,  whether or not you have access to the Inter-

       There are several TRICKLE servers throughout the world that coop-
       erate to distribute the files efficiently.  To request files, the
       user issues commands to the nearest TRICKLE server,  which deliv-
       ers the software either from its local cache disk, from the cache
       of another  TRICKLE server,  or from  an FTP site that  holds the

       Who can use TRICKLE

       There are currently TRICKLE servers at the following addresses:

       |                                                               |
       |  Location    EARN/BITNET         Internet                     |
       |                                                               |
       |  Austria     TRICKLE@AWIWUW11
       |  Belgium     TRICKLE@BANUFS11      |
       |  Colombia    TRICKLE@UNALCOL  |
       |  France      TRICKLE@FRMOP11     |
       |  Germany     TRICKLE@DEARN            |
       |  Israel      TRICKLE@TAUNIVM         |
       |  Italy       TRICKLE@IMIPOLI|
       |  Netherlands TRICKLE@HEARN |
       |  Poland      TRICKLE@PLEARN        |
       |  Sweden      TRICKLE@SEARN       |
       |  Turkey      TRICKLE@TREARN           |
       |  Turkey      TRICKLE@TRMETU  |
       |  UK          TRICKLE@UKACRL          |
       |                                                               |

       When you send a command to  a TRICKLE server,  it either executes
       the  command or  sends  you a  message with  the  address of  the
       TRICKLE server for your area.

       The files which are available from  TRICKLE are organized in main
       directories which contain many subdirectories.  The main directo-
       ries which are currently available are:

       |                                                               |
       |  Directory   Source FTP Site               Contents           |
       |                                                               |
       |  MSDOS        Large MS-DOS software archive
       |  MISC        Software for VM, VMS, Unix
       |  SIGM        SIG/M CP/M archive      |
       |  PC-BLUE        PC-BLUE MS-DOS archive  |
       |  CPM        CP/M Software Archive   |
       |  ARCHIVES        Various discussion group archives
       |  UNIX-C        Unix and C code software archives
       |  MACINTOS        Apple Macintosh software archives
       |  OS2         0tp-os2.nmsu.ed          Large archive of OS/2 software
       |  AMIGA             Large Amiga collection  |
       |  KERMIT   Kermit network software |
       |  TEX   TeX software and fonts
       |  WUARCHIVE      MS-DOS and others       |
       |  EXPO-MIT       Unix and others         |
       |  UUNET               Unix and others         |
       |  SUMEX-AIM   Macintosh and others    |
       |  GARFIELD   Multimedia (pictures and sounds)
       |  X11       X-Windows software distribution
       |  LINUX             Linux system software
       |                                                               |

       Not all directories are available at all servers. If your closest
       server does not provide the directory of your choice, you can use
       any other TRICKLE for the missing directory. If your closest ser-
       ver is  temporarily unavailable,  you  can use any  other TRICKLE

       How to get to TRICKLE

       You send  commands to TRICKLE  by electronic mail.   The commands
       should be in the body of the mail message,  one command per line.
       Any number of  commands (up to your daily command  limit)  may be
       placed in one message.  Users on the EARN/Bitnet network may also
       send commands to TRICKLE by interactive message.

       Using TRICKLE

       All commands begin with a slash (/). Note that the angle brackets
       (<>)  are part of the command,   not an indication of an optional

       Use the /PDDIR command to obtain directory listings.

       |                                                               |
       |  /PDDIR                                                       |
       |                                                               |
       |  /PDDIR    <dirname>                                          |
       |                                                               |
       |  /PDDIR    <dirname.subdirname>pattern                        |
       |                                                               |


       dirname   is the name of a main directory,

                 is the name of a subdirectory,

       pattern   is part of a filename.

       Use /PDDIR  without any parameters to  get a listing of  the main
       directories.  With PDDIR <dirname> you will  get a listing of the
       subdirectories under that directory. If you specify both directo-
       ry and subdirectory, it will list the files that are available in
       that subdirectory.  With pattern,  you will get a listing of only
       those files that match or begin with that pattern.  Wildcards "?"
       and "*"  may be imbedded into subdirname and pattern ("?" matches
       any single character; "*" matches any number of characters).

       Use the /PDGET command to get files.

       |                                                               |
       |  /PDGET    <dirname.subdirname>filename ( delivery-option     |
       |                                                               |


       dirname   is the name of a main directory,

                 is the name of a subdirectory,

       filename  is the name of a file.

                 specifies the  format to which  the file(s)   should be
                 translated before being sent to you.  The possible val-
                 ues are:

                      EBC80  UUE  XXE  HEX  BTOA

                 The option  EBC80 should be used  to get text  files if
                 you work on an IBM mainframe system.  The other options
                 are formats for  translating binary files so  that they
                 can be sent via electronic mail.   You will need a pro-
                 gram to translate  the file back to  its original form.
                 The default for  EARN/Bitnet users is to  send the file
                 as-is. The default for others is UUE.

       Wildcards "?" and "*"  may be  imbedded into subdirname and file-
       name ("?" matches any single character; "*" matches any number of
       characters) to get several files.

       The /SUB command is used to  subscribe to directories or to indi-
       vidual files.

       |                                                               |
       |  /SUB      <dirname>                                          |
       |                                                               |
       |  /SUB      <dirname.subdirname>pattern                        |
       |                                                               |
       |  /SUB      QUERY                                              |
       |                                                               |


       dirname   is the name of a main directory,

                 is the name of a subdirectory,

       pattern   is part of a filename.

       If you subscribe to a directory,  you will get a summary of added
       files about once a week, depending on how active the FTP site is.
       This listing shows the names, sizes and dates of each file added.
       If you subscribe to files,  as soon as TRICKLE is informed that a
       new version of the file has been  stored at its FTP site,  a copy
       of the file will be sent to you.

       Note that since  filenames usually reflect the  version number of
       the file,  it is a good idea to omit the number when specifying a
       pattern. For example, it is better to send the command:


       rather than

            /SUB <MSDOS.VIRUS>SCANV106

       since,  as new  versions of the file are stored,   the name might
       change to SCANV107, SCANV108, etc.

       Wildcards are not allowed with the /SUB command.

       The /SUB QUERY command allows you to  get a list of the files you
       are subscribed to.

       The /UNSUB command may be used to cancel a subscription.

       |                                                               |
       |  /UNSUB    <dirname>                                          |
       |                                                               |
       |  /UNSUB    <dirname.subdirname>pattern                        |
       |                                                               |


       dirname   is the name of a main directory,

                 is the name of a subdirectory,

       pattern   is part of a filename.

       The command /UNSUB * may be  used to terminate all your directory
       and file subscriptions.


       For a listing of the files  in the VIRUS subdirectory of SIMTEL20
       (the MSDOS directory), send the command:

            /PDDIR <MSDOS.VIRUS>

       To get the file 00-INDEX.TXT from <MSDOS.VIRUS> in EBCDIC format,
       send the command:

            /PDGET <MSDOS.VIRUS>00-INDEX.TXT ( EBC80

       If you are not  sure of the exact name of the  file,  you can use

            /PDGET <MSDOS.VIR*>*INDEX* ( EBC80

       To subscribe to automatically get new versions of the SCANV soft-
       ware from <MSDOS.VIRUS>, send the command:


       To unsubscribe from all files in the VIRUS subdirectory, send the

            /UNSUB <MSDOS.VIRUS>*

       Learning more about TRICKLE

       The /HELP command may  be sent to any TRICKLE server  to obtain a
       very detailed help file from the server.

       A brief guide to TRICKLE is available from the EARN documentation
       filelist.  Send mail to LISTSERV@EARNCC.BITNET.    In the body of
       the message, write:



       What is BITFTP

       BITFTP, BITNET FTP Server, allows users of EARN, Bitnet and asso-
       ciated networks to access FTP sites on the Internet.

       The BITFTP server provides a mail interface between a user on the
       EARN/Bitnet network to FTP sites on the Internet.  BITFTP handles
       this task by  passing the commands specified in  the mail message
       from the user to  another server  that actually makes the connec-
       tion to the FTP sites.  When  the server finishes the interaction
       with the FTP site, or fails due to an error,  a transcript of the
       result  is sent  back  to the  user,  as  well  as the  requested
       file(s), if any.

       The format of  the files sent to  the user can be  defined within
       the mail message.  BITFTP can deliver  files in netdata and uuen-
       code formats.

       Who can use BITFTP

       BITFTP is currently available only to  users on EARN,  Bitnet and
       other regional  NJE networks.  BITFTP  does not  support multiple
       file request  via the  mget command nor  does it  support sending
       files to FTP sites (the put command).

       How to get to BITFTP

       There are BITFTP servers currently running at:

       |                                                               |
       |  BITFTP@DEARN.BITNET        Germany                           |
       |  BITFTP@PUCC.BITNET         USA                               |
       |                                                               |

       Users on the  EARN/Bitnet international network only  may use the
       US server.    Only users from EARN  member countries may  use the
       server in Germany.  Servers at other locations are planned.

       BITFTP accepts requests via electronic  mail,  including IBM NOTE
       and PROFS-format messages, as well as by NJE file transfer.

       Using BITFTP

       BITFTP implements a large subset of the FTP commands of the IBM's
       TCP/IP for VM, using the same syntax. This software is documented
       in the IBM manual TCP/IP for VM User's Guide.   In the following,
       angle brackets (<>) indicate an optional parameter.

       Use the  ftp command to specify  which host to connect  to.  This
       command must be the first command in your mail file. You can also
       specify the  file format that you  wish BITFTP to use  to deliver
       files to you.

       |                                                               |
       |  ftp       hostname <fileformat>                              |
       |                                                               |


       hostname  is either the IP address or the domain name of the host
                 to connect to,

                 is the format  in which BITFTP should  deliver files to
                 you. It can be either netdata or uuencode.

       Use the  user command   to tell  the host  the username   and the
       password to be  used for the FTP connection.   Note  that on many
       FTP sites, both username and password are case-sensitive.

       |                                                               |
       |  user      username password                                  |
       |                                                               |


       username  is the user-id to use for the FTP connection,

       password  is the password for the username you specified.  It can
                 be omitted if you specified anonymous for the userid.

       Use the  cd command to select  a particular directory  as current

       |                                                               |
       |  cd        directory-name                                     |
       |                                                               |


                   is the name of the directory to be selected.

       Use the dir  command to display a  list of the files  in the cur-
       rent, or specified, directory.  The file names,  and depending on
       the site,  the file size,  file  creation date and other informa-
       tion will be listed.

       |                                                               |
       |  dir       <pattern>                                          |
       |                                                               |


       pattern   is a  pattern that defines  which file names  should be
                 displayed. Many FTP sites are case-sensitive, thus care
                 must be used with the pattern.  The pattern may contain
                 any number  of characters,  and the  wildcard character
                 "*" (asterisk) may be used to represent any characters.

       The ls command  is similar to the dir command,   except that with
       most FTP sites, it only displays the filenames, without any other
       information. The pattern specification is identical to dir.

       |                                                               |
       |  ls        <pattern>                                          |
       |                                                               |

       Use the get command to obtain a file from the current host.

       |                                                               |
       |  get       foreignname <localname>                            |
       |                                                               |


                 is the filename of the file, as stored at the FTP site.
                 With  many sites,   the case  of the  filename must  be

       localname is an optional local filename.  If you specify a local-
                 name,  it must be in the form:  filename.filetype where
                 neither part is longer than eight characters.

       Use the  binary  command to set  the FTP transfer mode.   If this
       option is set  then no EBCDIC-ASCII translation  will take place.
       This should be used for non-text files.

       |                                                               |
       |  binary                                                       |
       |                                                               |

       Use  the quit  command  to close   the  connection to  the  host,
       and  to terminate the BITFTP session.

       |                                                               |
       |  quit                                                         |
       |                                                               |


       To   get   the   file   from   the   directory
       /pub/nic/network.service.guides   at  the   anonymous  FTP   site,   and  to  get  a listing  of  the  files  in  that
       directory,  you  could send the  following commands by  e-mail to

            user anonymous
            cd pub/nic/network.service.guides

       In response you  will receive an e-mail  containing the following
       lines (some lines have been removed for brevity):

       > ftp
       > user anonymous
       >> USER anonymous

       > cd pub/nic/network.service.guides
       >> CD pub/nic/network.service.guides

       > get
       >>>> "" sent as "HOWTO FTPGUIDE".

       > dir

       total 60
       -rw-rw-r--  1 mtaranto 120   344 Apr 14  1992 README
       -rw-rw-r--  1 mtaranto 120 12759 Oct 30  1992
       -rw-rw-r--  1 mtaranto 120  6327 Mar 24 13:28
       -rw-rw-r--  1 root     120  2818 Mar  4  1992
       -rw-rw-r--  1 mtaranto 120  6136 Oct 30  1992
       -rw-r--r--  1 1086     120     0 Aug 18 21:23 plane.

       The above directory listing is typical  of the format of Unix FTP
       sites.  There is information on file permissions and ownership as
       well as the size of the file in  bytes,  the time and date of its
       last change, and the file name.

       You could request the first two  files by once again sending mail
       to BITFTP with the following commands:

            user anonymous
            cd pub/nic/network.service.guides
            get README

       Learning more about BITFTP

       A four-page instruction to the BITFTP  service can be obtained by
       sending a help  command in the body  of a mail message  to either
       BITFTP server.

       Additional  information  on  BITFTP may  be  obtained  from  EARN
       (NETHELP@EARNCC.BITNET). More information about TCP/IP and FTP in
       general can be obtained from  a wide  variety of sources, such as
       the documents mentioned in the example above.

                                    Part 6

                           NETWORKED INTEREST GROUPS

                            LISTSERV (Version 1.7f)

       What is LISTSERV

       LISTSERV is a  distribution list management package.   It runs on
       IBM  VM/CMS  systems  in the  international  NJE  network  (EARN/
       Bitnet).  It allows groups of computer users with a common inter-
       est to communicate among themselves,   while making efficient use
       of computer and network resources. It makes it easy for even nov-
       ice users to discover,  join,   and participate in these interest
       group mailing lists.  LISTSERV also  provides facilities for log-
       ging and archiving of mail  traffic,  file server functions,  and
       database searches of archives and files.

       There are  LISTSERV lists on every  subject under the  sun,  with
       names ranging  from AARPUB-L (AAR Electronic Publication list) to
       Z3950IW (Z39.50 Implementors Workshop). LISTSERV lists are inter-
       national and eclectic.   You will find lists  in every imaginable
       field,  for every  audience,  in many different  countries and in
       many different languages. See the description of the List command
       in the section Using LISTSERV section for instructions on getting
       an up-to-date list of lists.

       Who can use LISTSERV

       Anyone who can send electronic mail to an EARN/Bitnet address can
       participate in a  mailing list and access  other LISTSERV facili-
       ties,  as  long as the e-mail  format is valid (according  to the
       RFC822 standard),  and  has a usable return  address.  Every day,
       people use LISTSERV from HEPnet,  Internet,  Compuserve,  MCIMail
       and many other networks throughout the world.

       If you don't know how to send electronic mail to EARN/Bitnet, ask
       your local support people for help.

       How to get to LISTSERV

       To  use LISTSERV  facilities,   send  electronic mail  with  your
       LISTSERV commands to:  LISTSERV@host-id where host-id is the host
       computer's  NJE address  (for example,   TAUNIVM.BITNET)  or  its
       Internet domain name (in this case,  VM.TAU.AC.IL).  There may be
       some local variation on the format  needed to send mail to Bitnet
       or Internet addresses. Check with your local support personnel.

       More than one  command can be sent  to LISTSERV in the  same mail
       message.  Each command must be on a separate line.  LISTSERV will
       ignore the Subject:  line of the mail header, so ensure that your
       commands are in the body of the e-mail.

       For EARN/Bitnet users,  interactive messages  are the fastest and
       most convenient  way to send commands  to LISTSERV,  but  bear in
       mind that interactive  messages only work when  the links between
       your computer and LISTSERV are up;  if the message fails, you can
       always send your command via mail.

       To facilitate  communications with  the LISTSERV  servers on  the
       EARN/Bitnet network,  a  special LISTSERV computer node  has been
       defined.   This means that any EARN/Bitnet user can address their
       commands to the userid LISTSERV at the special node LISTSERV.

       All non-EARN/Bitnet users  may address their command  mail to the
       userid LISTSERV at  the special node LISTSERV.NET.   Note that if
       this node is not  yet defined on your network,  you  may also try
       LISTSERV%LISTSERV.BITNET@CUNYVM.CUNY.EDU.  For  example,  if  you
       wanted to join  the 3D-L mailing list that  discusses 3D computer
       graphics but are  not sure of the address of  the LISTSERV server
       hosting  this list.   You could  send a  subscription request  to
       either of the following addresses:

            LISTSERV@LISTSERV       (on BITNET)

       Your subscription request  will be automatically forwarded  on to
       the LISTSERV server hosting the 3D-L  mailing list (in this case,
       at the computer node ARIZVM1).

       You may also utilize the special  LISTSERV address to send e-mail
       to any  LISTSERV mailing list if  you are unsure of  its address.
       For example,  if you wanted to  send mail to the BITFTP-L mailing
       list to request a copy of the BITFTP product,  you could do so by
       addressing your e-mail to BITFTP-L@LISTSERV.NET. It will automat-
       ically be  forwarded to  the list's real  address (in  this case,
       BITFTP-L@EARNCC.BITNET)  when it reaches the LISTSERV node.  Once
       you know the real  address of a list,  it should  be used for all
       future mail sent to that list.

       There are more than 250 sites in over 30 countries throughout the
       world running LISTSERV  as part of the  world-wide network.  Here
       are a few of the LISTSERV sites:

       |                                                               |
       |  Host Computer Site                                    Country|
       |                                                               |
       |  BITNIC        BITNET Network Information Center           USA|
       |  DEARN         GMD, Bonn                               Germany|
       |  EARNCC        EARN Office, Paris                       France|
       |  HEARN         Katholieke Universiteit Nijmegen    Netherlands|
       |  PUCC          Princeton University, New Jersey            USA|
       |  SEARN         Kungliga Tekniska Hoegskolan, Stockholm  Sweden|
       |                                                               |

       Using LISTSERV

       The commands given here are available for everyone and require no
       special LISTSERV command  privileges.  Only the most  common com-
       mands  are described.   For  a  complete list  of  non-privileged
       LISTSERV commands,  retrieve  the LISTSERV User Guide  in the DOC

       Commands are presented  in a particular format:   CAPITAL letters
       indicate acceptable abbreviation, angle brackets (<>) indicate an
       optional parameter,  and vertical bar  (|)  indicates a choice of
       parameters.   All parameters are fully  explained in each command

       There is a standard set of  command keywords available for use in
       some LISTSERV commands.  They may be used with any commands where
       they appear  as optional parameters  in the  command description.
       The important standard keywords are:

       PW= password
                 This keyword  is used to specify  a password in  a com-
                 mand.  If you have registered  a personal password on a
                 LISTSERV server, you will then have to validate certain
                 commands by using  the PW= command keyword  in the com-
                 mand text.  This  feature is designed to  prevent impo-
                 stors  from issuing  commands while  using your  e-mail
                 address.  The use of  passwords is strongly recommended
                 for this reason. See the PW command for details on reg-
                 istering personal passwords. Once you have registered a
                 personal password  at a LISTSERV  server,  you  will be
                 obliged to include the PW=  command keyword in all com-
                 mands where it appears as an option.

       F= format This keyword controls the file format (or internal file
                 structure)  in which files will be sent to you.  If you
                 are  not a  member of  the  EARN/Bitnet network,   then
                 LISTSERV will  always use  the default  file format  of
                 MAIL.   Otherwise, the default file format is dependent
                 upon the  information held  in the  BITEARN NODES  file
                 regarding your computer.   The BITEARN  NODES file is a
                 special network definition file used in the EARN/Bitnet
                 network.  Any user may ask for a file format other than
                 their default by  specifying it in the  F= command key-
                 word in  the commands  where it  appears as  an option.
                 Note that this will not permanently change your default
                 file format,   its effect  is only  for those  commands
                 where it is specified. Following is a list of file for-
                 mats that are valid for all users:

                      XXE  UUe  MIME/text  MIME/Appl  MAIL

                 In addition, EARN/Bitnet users may specify:

                      Netdata  Card  Disk  Punch  LPunch  VMSdump

                 Contact your local support  personnel for more informa-
                 tion about the file formats supported on your computer.

       Commands for LISTS:

       The  primary function  of LISTSERV  is to  operate mailing  lists
       (also referred to as distribution lists).  Mailing lists are used
       to distribute  the e-mail sent  to them  to a list  a recipients.
       They  provide the  means for  a group  of users  to establish  an
       e-mail forum on any topic or  area of common interest.  This ser-
       vice provides an  extremely convenient means for  the exchange of
       ideas and  information between list  members since  LISTSERV (and
       not the mail sender) manages the distribution of e-mail to all of
       its  final recipients.   Users  need only  to  remember one  list
       address to which they send their  mail in order to communicate to
       a potentially large number of users.   Due to the efficiency with
       which LISTSERV distributes e-mail to list members, discussions or
       debates with a world-wide audience may be conducted.

       The following commands are designed for use with LISTSERV mailing
       lists.  With them you may do things such as finding the names and
       addresses of lists,  signing on to or off lists,  reviewing lists
       or changing your  personal options for any mailing  list to which
       you are subscribed.

       |                                                               |
       |  SUBscribe list-name  <full-name>                             |
       |                                                               |

       Use the SUBscribe  command to join a mailing list.   You can also
       use this command  to alter the name (but not  e-mail address)  by
       which you  are known on a  mailing list you have  already joined.
       The list-name parameter is the name of the list to which you want
       to subscribe.  For example,  the EARN  User Group list located at
       the node IRLEARN has a list name of EARN-UG.  Do not confuse this
       with the  list's address (EARN-UG@IRLEARN)   which should  not be
       used.  The optional full-name parameter allows you to give a name
       by which you want to be known on a mailing list. If specified, it
       should be your full, real name (at least your first name and last
       name)  and not your e-mail address.   If you send this command to
       LISTSERV via command mail, the name for the full-name option will
       be taken from the name given in  the From:  mail header should it
       be omitted from the command text. If you send a SUBscribe command
       for a list to which you already belong, then LISTSERV will inter-
       pret the  command as a  request to  change your full-name  on the

       A request to join a mailing list  can be processed in three ways:
       subscription to a list may be OPEN, CLOSED, or BY-OWNER. If it is
       OPEN,  you will be automatically added to the list and sent noti-
       fication. If it is CLOSED, you will not be added to the list, and
       LISTSERV will  send you a message  telling you that  your request
       has been rejected.  If it is BY-OWNER,  your subscription request
       will be forwarded to the list  owner(s),  who will decide whether
       or not to add  you to the list (LISTSERV will  inform you to whom
       your request has been forwarded).  To  see what kind of subscrip-
       tion a list has, use the REView command.

       |                                                               |
       |  UNSubscribe list-name  |  *  <(NETWIDE>                      |
       |                                                               |

       Use the UNSubscribe  command to leave a mailing  list.  The list-
       name parameter is the name of a  mailing list from which you want
       to remove your  subscription.   You can signoff all  the lists to
       which you are  a member at any particular LISTSERV  site by using
       the "*" (asterisk) character in the place of a list name.  If you
       want your  UNSubscribe command to  be propagated to  all LISTSERV
       servers on the network,  include  the (NETWIDE option.   Use this
       option if  you are  changing your e-mail  address or  are leaving
       your computer for an extended period.

       |                                                               |
       |  List      <options>  <F= format>                             |
       |                                                               |

       Use the List command to get  a listing of available mailing lists
       at a  LISTSERV server.  The options  parameter may be any  of the

       Short     This option displays a summary of all the lists managed
                 by a LISTSERV in a brief, one line description. This is
                 the default.

       Long      The Long (or Detailed)  option will send a file (called
                 node-name LISTS)  to you  that contains a comprehensive
                 description of the lists managed by a LISTSERV server.

       Global <pattern>
                 This option gives a complete list of all known LISTSERV
                 mailing lists at all servers at the time the command is
                 issued.  A file (called LISTSERV LISTS) will be sent to
                 you containing the names,   titles and e-mail addresses
                 of these lists. This is a very large file, so make sure
                 you have  the necessary disk  space to  accommodate its
                 size before  you use the  Global option.   The optional
                 pattern parameter  can be used  to match any  string in
                 the list name, list title or list address.

       |                                                               |
       |  REView    list-name  <(>  <options>                          |
       |                                                               |

       Use the REView command to receive a listing of a mailing list. It
       will be sent to you as a file called list-name LIST (or list-name
       node-name for peered lists).  A mailing  list is comprised of two
       parts: a control section and a subscription section.  The control
       section holds the definition parameters for a list which includes
       information such  as who is authorized  to review or join  a list
       and whether or not it is archived. The subscription section holds
       the e-mail addresses  and names of all list  members.  The REView
       command allows  you to  receive a  listing of  either or  both of
       these sections (the default is both)  for any list,  provided you
       are authorized to do so.  Note that at the discretion of the list
       owner(s),  the  REView command can be  restricted in use  to list
       members only.  In this case, you will not be authorized to review
       a mailing  list if you  are not a  member of that  list yourself.
       Also,  individual  list members  can restrict  the appearance  of
       their e-mail address and name in  response to a REView command if
       they have set  the CONCEAL mailing list option (see  the SET com-
       mand for more  details).  The list-name parameter is  the name of
       the LISTSERV list you wish to review. The important options are:

       Short     This option  restricts the  information you  receive to
                 the control  section of a  list (giving  its definition
                 parameters)  and does not  return the subscription sec-
                 tion of a list (giving the list members).

       Countries If you  use this option,  the  list of members  will be
                 organized  by the  nationality  given  in their  e-mail

       LOCal     If the list is peered (that  is,  it is linked to other
                 mailing  lists  of  the  same  name  but  on  different
                 LISTSERV servers),  you will receive listings of all of
                 these mailing  lists in response  to a  REView command.
                 The LOCal option  can be used to  suppress the propaga-
                 tion  of the  REView command  to  the LISTSERV  servers
                 hosting these peered mailing lists.  In this case,  you
                 will receive  a listing only  from the server  to which
                 you send the REView command and not the others.

       |                                                               |
       |  Query     list-name  |  *                                    |
       |                                                               |

       When you join  any mailing list,  you will be  assigned a default
       set of list options that control  such things as way you received
       mail when it is distributed and the type of notification LISTSERV
       will give you when  it distributes mail you have sent  to a list.
       These are  the personal list options  that may be altered  on any
       mailing list  to which you are  subscribed to suite  your circum-
       stances. See the SET command for a complete descriptions of these
       options.  The  Query command can be  used to review  the personal
       list options you  have in effect at any mailing  list.  The list-
       name parameter is the name of a list to which you are subscribed.
       If you use  an "*" (asterisk)  character instead of  a list name,
       you will receive information about  your personal options for all
       lists to which you  belong at the LISTSERV to which  you send the

       |                                                               |
       |  SET       list-name  |  *   options                          |
       |                                                               |

       Use the SET command to change your personal options for a mailing
       list.  These options will remain in effect until they are explic-
       itly changed.  The Query command can be used to display your cur-
       rent optional settings for any list  to which you are subscribed.
       The list-name parameter is the name of the mailing list for which
       you are changing your options.  You may change your options for a
       specific list or for all the lists  you belong to at a particular
       LISTSERV by using the "*" (asterisk) character in place of a list
       name.  After processing your SET command,  LISTSERV will send you
       conformation of  the successful alteration  of your  mailing list
       options via e-mail.  The important options are:

       Mail | DIGests | INDex | NOMail
                 These options of the SET command alter the way in which
                 you receive mail from a  mailing list.  The Mail option
                 means that  you wish to  have list mail  distributed to
                 you as  mail.   This is  the default.  The  DIGests and
                 INDex  options  are available  only if  a list  has had
                 these features  enabled by its owner(s).   Digests hold
                 all the  mail messages  sent to a  list over  a certain
                 period of time. Instead of receiving each mail individ-
                 ually as it  is distributed to list  members,  you will
                 receive all the mail in one batch for a given day, week
                 or month.  You may then  browse through these mail mes-
                 sages at  your leisure.  Note  that mail is  not edited
                 with the DIGests option, you will receive copies of all
                 the mail messages in their  entirety.  The INDex option
                 will provide you  with only the date,   time,  subject,
                 number of lines  and the sender's name  and address for
                 all mail messages sent to a list.  The text of the mail
                 message will not be included.   You may then select and
                 retrieve  any mail  that interests  you  from the  list
                 archive.   Both the DIGests and INDex options provide a
                 means of listening  in to discussions on  mailing lists
                 without having to deal with  large quantities of incom-
                 ing mail messages.   The NOMail option means  that mail
                 sent to the list will not  be distributed to you.  This
                 is useful  when you  are leaving  your computer  for an
                 extended period of  time and you do not  want mail from
                 the list to fill your mailbox. When you return, you can
                 send the  SET command with  the Mail option  to restore
                 mail service.

       SHORThdr | FULLhdr | IETFhdr | DUALhdr
                 All mail messages are comprised of header and body sec-
                 tions.  The header section provides details such as the
                 recipients, the original sender and the date and time a
                 mail message was sent.  The  mail body section contains
                 the text of  a mail message.  These options  of the SET
                 command indicate the  type of mail headers  you want to
                 receive in  the mail distributed  from a  mailing list.
                 SHORThdr means that  the mail header will  include only
                 the essential  informational headers (for  instance the
                 Date:,  To:,  From:,  Subject:,  Sender:  and Reply-to:
                 headers).  This is the default.  You may change this to
                 FULLhdr, which means that all (including non-essential)
                 mail headers  will be present  in e-mail.   The IETFhdr
                 option means that LISTSERV will  not change the headers
                 of a mail  message it distributes to you  other than to
                 add a Received: mail header (and also a Message-id: and
                 Sender:  headers  if either of  these were  not already
                 present). This option is designed specifically for com-
                 patibility with SMTP exploders (as used,  for instance,
                 on the Internet network). Lastly, DUALhdr is very simi-
                 lar to  the SHORThdr option  except that  LISTSERV will
                 also insert mail  headers at the beginning  of the mail
                 body.  Therefore, when the mail is received and read by
                 a recipient using this option,  it will start with this
                 information (for instance,  the first  three lines of a
                 mail message may contain the  To:,  From:  and Subject:
                 mail headers).  This option is useful for users of some
                 PC  based mail  packages  that  will not  display  this
                 information from the real mail headers.

                 Indicates whether  or not you  want your name  and mail
                 address to appear in the  display of list members which
                 is given in response to  a REView command.  The default
                 is NOCONCEAL.   Note that a complete list of members is
                 always given to list owners and LISTSERV administrators
                 regardless of this option.

       |                                                               |
       |  CONFIRM   list-name                                          |
       |                                                               |

       Use the  CONFIRM command  to renew your  subscription to  a list.
       Some mailing lists require subscription renewal at regular inter-
       vals (usually once a year).  A mail message is automatically sent
       to list members indicating that they  must send a CONFIRM command
       within a given  number of days or  they will be removed  from the
       list. This command must be sent from the same e-mail address that
       received the confirmation notice.  The list-name parameter is the
       name of  the mailing list to  which you are confirming  your sub-
       scription.  LISTSERV will  send a message that  your subscription
       has been confirmed.

       Commands for FILES:

       LISTSERV also functions as a file server.  That is,  files can be
       stored at a  LISTSERV server and made available  for retrieval by
       users upon  their request.   These files  are stored  in LISTSERV
       under a hierarchical system of filelists. As the name suggests, a
       filelist is a special file that holds a list of files. Each entry
       in a  filelist describes a file  that is available  for retrieval
       and gives  details such as  the file's name  and size and  also a
       file's access code (also known as  a FAC)  which describes who is
       authorized to retrieve  it.  These files may  themselves be file-
       lists,  in which case they too will hold information about avail-
       able files  or further  filelists.  In  this way,   filelists are
       organized in a hierarchical or tree structure.

       There are two basic types of  filelists on LISTSERV servers.  The
       first type holds  files that have been  specifically placed there
       by file  owners or LISTSERV  administrators.  These files  may be
       documents, maps or diagrams, or even programs. The second type of
       filelist is associated  with a LISTSERV mailing  list.  These are
       referred to  as list  filelists and retain  copies of  all e-mail
       distributed  on a  particular mailing  list.  They  consist of  a
       series of files,  each of which holds a copy of the mail messages
       distributed on  a list  over an interval  of time  (usually,  one
       month).  These  files may then be  retrieved to recover  any list
       mail distributed during a specific period.  Note that e-mail dis-
       tributed on a mailing list may also be retrieved via the database
       functions of LISTSERV.  Not all  mailing lists have an associated
       list filelist; this is enabled at a list owner's discretion.

       A further aspect of the file  server functions of LISTSERV is the
       concept of packages. A package consists of one or more files that
       appear in a  filelist.  These may be,  for example,   a series of
       files  that go  to  make up  a software  package.   A package  is
       declared in a LISTSERV filelist through  a file that has the spe-
       cial name of package-name $PACKAGE.  Within  it will be a list of
       all  the files  that  comprise the  package.   This  list can  be
       reviewed by sending a retrieval  request for that file.  However,
       all the  files that  comprise the package  may also  be retrieved
       through a request for the file:  package-name PACKAGE.  Note that
       in this case,  the "$" (dollar)  symbol is dropped from the name.
       The files that go to make  up the package will then automatically
       be sent  to the  requestor.  This  enables users  to retrieve  an
       entire set of files  that belong to a package with  only one com-
       mand, instead of having to retrieve each of these files individu-
       ally through a series of commands.

       The following LISTSERV  commands enable general users  to manipu-
       late files that are stored at a server. This includes commands to
       search for,  retrieve and subscribe  to files.  When sending file
       server commands to LISTSERV,  you must address them to the server
       and not  to any mailing lists.   Note that where the  PW= keyword
       appears in a command description,  this  need only be included in
       the command text  if you have defined a personal  password at the
       server to which you are sending file server commands.  Otherwise,
       it need not be used. The optional F= command keyword may included
       as desired.

       |                                                               |
       |  INDex     <filelist>  <F= format>                            |
       |                                                               |

       Use the INDex command to get a listing of the files in a particu-
       lar filelist.  The  filelist parameter specifies the  name of the
       filelist you wish to get.  If no  name is specified,  an index of
       the root filelist (called LISTSERV FILELIST) will be sent to you.

       |                                                               |
       |  GET       filename  filetype  <filelist>  <F= format>        |
       |                                                               |

       The GET  command is used to  retrieve a specific file  or package
       from a filelist, provided you are authorized to do so.  The file-
       name and  filetype parameters  identify the  file or  package you
       wish to retrieve.  The optional filelist parameter identifies the
       filelist within which the file or  package resides.  If it is not
       supplied in  the command,  LISTSERV  will determine  the filelist
       through a search of its own internal filelist index.

       |                                                               |
       |  AFD       options                                            |
       |                                                               |

       AFD is an acronym for  Automatic File Distribution.  This command
       enables you  to subscribe  to a  file or  package in  a filelist.
       Every time this file or package  is updated in its filelist,  you
       will then  automatically be sent  of copy  of it by  the LISTSERV
       server.  You  may subscribe to any  number of files  for packages
       from any filelist  with the AFD command.  You may  also review or
       delete your  subscriptions at any  LISTSERV server.   The options
       must be one of the following:

       ADD  filename filetype <filelist> <text> <PW= password> <F=format>
                 The ADD  option allows  you to subscribe  to a  file or
                 package. This means that a copy will henceforth be sent
                 to you automatically by LISTSERV  when it is updated on
                 its  filelist.  The  filename  and filetype  parameters
                 identify a  file or package to  which you want  to sub-
                 scribe.  The filelist option allows  you to specify the
                 name of the filelist in which it is located. If this is
                 not specified in the command,  LISTSERV will locate the
                 filelist through a search of  its own internal filelist
                 index.  Also, if you wish to have an informational text
                 message inserted automatically  at the top of  the file
                 or package when it is sent to you, you may use the text
                 parameter.  Note that if you  omit the filelist option,
                 the  text you  provide in  the text  parameter must  be
                 enclosed in double quotes (").   This parameter is most
                 commonly used by network  server applications that sub-
                 scribe to files at a LISTSERV  on behalf of users.   If
                 you  are subscribing  to files  directly from  LISTSERV
                 yourself, you should not need to use this option.

       DELete  filename  filetype  <filelist>  <PW= password>
                 Remove your subscription to a file or package for which
                 you have an AFD.  The  filename and filetype parameters
                 identify a file or package to  which you wish to remove
                 an AFD. These names may also include the "*" (asterisk)
                 wildcard character  so that multiple files  or packages
                 may be removed  with one DELete command.   The filelist
                 option allows you  to specify the name  of the filelist
                 in which it is located. If this is not specified in the
                 command,  LISTSERV  will locate the filelist  through a
                 search of its own internal filelist index.

       List  <(FORMAT>
                 The List  option shows the  files or packages  to which
                 you are  currently subscribed  at a  LISTSERV.  If  you
                 include the (FORMAT option,  then the file format to be
                 used when a file or package is sent to you will also be

       |                                                               |
       |  FUI       options                                            |
       |                                                               |

       FUI is an acronym for File Update Information.  It enables you to
       subscribe to  a file or package  in a filelist.  Every  time this
       file for package is updated in its filelist,  you will then auto-
       matically be informed of this by the LISTSERV server.   This com-
       mand is similar to the AFD  command, except you will only be sent
       notification of an update rather than the updated file or package
       itself.  You  may subscribe to any  number of files  for packages
       from any filelist  with the FUI command.  You may  also review or
       delete your  subscriptions at any  LISTSERV server.   The options
       parameter must be one of the following:

       ADD  filename  filetype  <filelist>  <PW= password>
                 The ADD  option allows  you to subscribe  to a  file or
                 package by adding an FUI to it.  The filename and file-
                 type parameters identify a file or package to which you
                 want to add  an FUI subscription.  The  filelist option
                 allows you to specify the name of the filelist in which
                 it is  located.  If this is  not given in  the command,
                 LISTSERV will locate  the filelist through a  search of
                 its own internal filelist index.

       DELete  filename  filetype  <filelist>  <PW= password>
                 Remove your  subscription from  a file  or package  for
                 which you have an FUI. The filename and filetype param-
                 eters identify a  file or package to which  you want to
                 delete an FUI. The filelist option allows you to speci-
                 fy the name of the filelist in which it is located.  If
                 this is  not specified in  the command,   LISTSERV will
                 locate the filelist through a  search of its own inter-
                 nal filelist index.

       List      The List option  shows the files or  packages for which
                 you currently  have an FUI  subscription at  a LISTSERV

       |                                                               |
       |  Query File filename  filetype  <filelist>  <(FLags>          |
       |                                                               |

       This command  can be  used to get  update information  on various
       files.  Update information about the file or package specified in
       the filename and filetype parameters will be given.   Optionally,
       you may include a filelist name in the filelist  parameter. If it
       is omitted, LISTSERV will locate the filelist through a search of
       its own internal filelist index.  You may also specify the (FLags
       option to display additional technical data about the file (which
       can be  useful when  reporting problems  to LISTSERV  administra-

       |                                                               |
       |  PW        options                                            |
       |                                                               |

       The PW command  enables you to add,  change or  delete a personal
       password on any LISTSERV server.  A personal password is designed
       to give you  added command security on a LISTSERV  since it helps
       prevent impostors using your e-mail address.  The use of personal
       passwords is strongly  encouraged for this reason.   Every server
       will accept a password registration request from you at any time.
       Passwords are comprised of one  to eight alphanumeric characters.
       You may change or delete your  password at any time.  The options
       parameter must be one of the following:

       ADD  new-password
                 Add a new personal password  on the LISTSERV processing
                 your command.  Once you have registered a password on a
                 LISTSERV server,   you will be  obliged to use  the PW=
                 command keyword in the commands  where it appears as an

       CHange  old-password  new-password
                 Change  your personal  password  on  a LISTSERV  server
                 where you already have one.

       DELete  old-password
                 Remove your personal password from a LISTSERV where you
                 already have one. Once you have removed a password from
                 a LISTSERV server, you will no longer be obliged to use
                 the  PW=  command  keyword in  the  commands  where  it
                 appears as an option.

       LISTSERV DATABASE Functions

       LISTSERV provides  the facility  for users  to retrieve  old mail
       that has been distributed on mailing lists. Each mailing list has
       an associated database  (called a notebook or  list archive data-
       base)  in which list mail is stored.  Note that databases may not
       be maintained for every mailing list, this is done at the discre-
       tion of a list owner. The notebook databases are the most common-
       ly used of the LISTSERV databases. However, every LISTSERV server
       also has a database of all the EARN/Bitnet computer nodes (called
       the BITEARN database).   This is available to all LISTSERV users.
       The backbone   LISTSERV servers also have  a database of  all the
       LISTSERV computer nodes (called the PEERS database).  In addition
       to these databases, a LISTSERV server may have any number of dif-
       ferent databases that have been created locally. To find out what
       databases are accessible at a particular LISTSERV site, send fol-
       lowing command to that server:

            DATABASE LIST

       To perform a database search,  you can send mail to LISTSERV con-
       taining a batch database job  which contains your database query.
       In addition,  EARN/Bitnet  users on VM or VMS  systems can access
       the database facilities interactively via the LDBASE program. For
       more details on  the LISTSERV database facilities,   send an Info
       DATABASE command to  your nearest (or any)   LISTSERV server (see
       the section  Commands for INFORMATION)   or consult  the LISTSERV
       DATABASE Functions chapter in the LISTSERV User Guide.

       Commands for INFORMATION:

       The LISTSERV server can provide a diverse range of information to
       the general user. This includes help files, release levels of the
       server and important configuration files, statistics and informa-
       tion pertaining to  the EARN/Bitnet network.  When  sending these
       commands to LISTSERV,   they must be addressed to  the server and
       not to any mailing lists it  may manage.  For those commands that
       result in files being sent to the requestor (for example the Info
       command), the format of the file can be specified by the optional
       command keyword F= in the command  text (see the section LISTSERV
       Command Keywords for more information).

       |                                                               |
       |  Help                                                         |
       |                                                               |

       Use this command to get a  brief description of the most commonly
       used LISTSERV  commands and also the  name and e-mail  address of
       the server's postmaster.

       |                                                               |
       |  Info      <topic>  <F= format>                               |
       |                                                               |

       Use this command  to get an informational (or help)   file from a
       LISTSERV server.  The  topic option should specify  the topic for
       which an informational file  will be sent to you.  You  may get a
       list of valid topics by sending  the Info command with no parame-
       ters to your nearest (or any) LISTSERV.


       You wish to subscribe to the EARNEWS list which is located at the
       node FRMOP11. Your full name is Mark P. Waugh. Send the following
       command to LISTSERV@FRMOP11.BITNET:

            SUBSCRIBE EARNEWS Mark P. Waugh

       You wish  to leave the INFO-MAC  mailing list (to which  you have
       already subscribed) at the node CEARN. The command:


       should be sent to the LISTSERV  server at CEARN which manages the
       INFO-MAC list.    To leave all the  LISTSERV lists you  belong to
       throughout the network,  send the following command to your near-
       est (or any) LISTSERV:


       You wish to receive a listing of  all mailing lists that have the
       text europe in their name or title. Send the following command to
       your nearest (or any) LISTSERV server:


       You want to  stop receiving mail from  all the lists at  SEARN to
       which you belong. Send the following command to the LISTSERV ser-
       ver at SEARN:

            SET * NOMAIL

       You have received  a message from the LISTSERV  server at IRLEARN
       asking you to confirm your subscription to the EARN-UG list. Send
       the following command to that server:

            CONFIRM EARN-UG

       You wish to  receive a listing of  the files in the  DOC FILELIST
       The command:

            INDEX DOC

       should be sent to the LISTSERV  server at EARNCC where this file-
       list is located. Note that the above command is the same as issu-
       ing a GET DOC FILELIST command.

       You wish to retrieve the file PCPROG ZIP from a filelist and have
       it sent to you in XXE file format.  Send the following command to
       the LISTSERV server that holds this file:

            GET PCPROG ZIP F=XXE

       You want to retrieve all the files  that make up a package called
       PROGRAM (as  listed in  a file called  PROGRAM $PACKAGE)   from a
       filelist called SAMPLE. Send the command:


       You wish to subscribe to a file  called BUGDET MEMO in a filelist
       called EXPENSES with an AFD:


       To subscribe to a  file called VM EMAIL in the  DOC FILELIST with
       an FUI,  you would send the following command to LISTSERV at node

            FUI ADD VM EMAIL DOC

       Learning more about LISTSERV

       A standard set of help files are available upon request from each
       LISTSERV server.  To get a copy of these files, use the INFO com-
       mand (see the section Commands for INFORMATION).

       Detailed  documentation on  LISTSERV (and  related services)   is
       available from the DOC FILELIST at LISTSERV@EARNCC.BITNET.   This
       includes the LISTSERV User Guide which is available in both post-
       script and plain text formats. To obtain a list of available doc-
       uments  use  the INDex  command  (see  the section  Commands  for

       There  are several  mailing  lists  for discussion  of  technical
       LISTSERV issues. They are not intended for casual users, but they
       should be of interest to advanced users. They are:

       LSTSRV-L  Technical forum on LISTSERV

       LSTOWN-L  LISTSERV list owners' forum

       LDBASE-L  Forum on LISTSERV database search capabilities

                               NETNEWS (USENET)

       What is NETNEWS

       Netnews,  or Usenet as it is  more commonly called,  is a message
       sharing system that exchanges  messages electronically around the
       world in  a standard  format.  Messages  exchanged on  Usenet are
       arranged by topic into categories called newsgroups.  Netnews is,
       thus, a huge collection of messages, being passed from machine to
       machine.  The  messages may contain  both plain text  and encoded
       binary information.   The messages also contain header lines that
       define who the  message came from,  when the  message was posted,
       where it was posted,  where it has passed,  and other administra-
       tive information.

       The major, hierarchical categories of Usenet newsgroups which are
       distributed throughout the world are alt, comp, misc, news,  rec,
       sci, soc,  and talk.  There are many other major categories which
       may be topical (eg, bionet, biz, vmsnet) and are usually distrib-
       uted worldwide as well,  or  geographical and even organizational
       (eg, ieee) or commercial (eg, clari).   The latter categories are
       usually distributed only with their area of interest. The messag-
       es of many Bitnet LISTSERV mailing  lists are also distributed in
       Usenet under the major category bit.

       The major categories are further broken  down into more than 1200
       newsgroups on different  subjects which range from  education for
       the disabled to Star Trek and from environmental science to poli-
       tics in the former Soviet Union. The quality of the discussion in
       newsgroups is not guaranteed to be high.   Some newsgroups have a
       moderator who  scans the messages for  the group before  they are
       distributed and decides which ones  are appropriate for distribu-

       Usenet was originally developed for Unix systems in 1979.  Within
       a year, fifty Unix sites were participating. Now, there are thou-
       sands of sites running a number of operating systems on a variety
       of hardware platforms communicating via Usenet around the globe.

       Who can use NETNEWS

       Usenet newsgroups  can be read at  thousands of sites  around the
       world.   In addition, there are several sites that provide public
       dial-up service  so that people who are  not at a Usenet site can
       have access to newsgroups as well. If you don't know if your site
       has Usenet access, check with your local computer support people.

       Protocols and software for the distribution of news are in use in
       several networks,  such as the  Internet,  UUCP,  EARN/Bitnet and

       If you have e-mail service only,  then you can not access Usenet.
       However, many newsgroups are connected to mailing lists which you
       could join.  For a list of  these newsgroups and their associated
       mailing lists,  send mail to LISTSERV@AMERICAN.EDU with the line:
       GET NETGATE GATELIST Moreover, many of the documents which appear
       periodically in  newsgroups are  available by  e-mail from  mail-   For instructions,  send a message with the
       subject HELP

       How to get to NETNEWS

       If your site  provides Usenet access,  then you just  need to use
       one of the many software packages  available for reading news (at
       least one is probably available on your computer). These packages
       either  access a  local  news spool,   or  use  the Network  News
       Transfer Protocol (NNTP)  to access the  news spool on some other
       computer in the network.

       Within EARN,  a  network of Netnews distribution  has been devel-
       oped,  providing  efficient distribution of Usenet  traffic while
       minimising the  load on the  network for the  participating coun-

       If Usenet is not  available to you and you would  like to arrange
       access for  your site,  contact  your system  administrator.  You
       should also read the article How to become a USENET site which is
       posted periodically to  the news.answers newsgroup.   It  is also
       available    by    anonymous    FTP    from    as
       /pub/usenet/news.answers/site-setup     or    by     mail     to:      with     the      line:        send

       Using NETNEWS

       There are many  software packages available for  reading and dis-
       tributing Netnews on a variety  of operating systems (Unix,  VMS,
       VM/CMS, MVS, Macintosh, MS-DOS and OS/2) and environments (Emacs,
       X-Windows,  MS-Windows).  See  the list of freely  available news
       reader software packages in Appendix A.   Note that the number of
       software packages available to run  news,  especially on PCs,  is

       In addition to the software  packages specifically designed to be
       news readers, many other communications programs, particular mail
       interfaces, provide the possibility for Usenet access in addition
       to their main function.

       Most,  if  not all,  of the  news readers provide the  same basic

       *   Subscribing to newsgroups:  This means that your news reading
           software will  make these groups immediately  accessible,  so
           that  you can  choose to  read  the postings  of groups  that
           interest you quickly and easily.

       *   Unsubscribing from  newsgroups:   Removing  groups from  your
           easy access list.

       *   Reading newsgroup postings:   Your news reader presents post-
           ings to  you and keeps track  of which postings you  have and
           have not read.

       *   Threads of  discussion:   You can  follow groups  of postings
           that deal with the same subject easily.

       *   Posting to news groups:  You can participate in group discus-
           sions; your news reader knows where to send your posting.

       *   Responding to  a posting:    You can send  a response  to the
           newsgroup (often  called follow-up)   or to  the author  of a
           posting (often called reply).


       When you  enter the tin  news reader,  you  get a listing  of the
       newsgroups to which you are subscribed:


                                Group Selection (9)              h=help

          1 30637 bit.listserv.novell   local list
          2  1106 comp.mail.misc        General discussions about compu
          3  8031 comp.protocols.tcp-ip TCP and IP network protocols.
          4   840 comp.sys.mac
          5  8789 news.answers          Repository for periodic USENET
       -> 6    29 news.lists            News-related statistics and lis
          7 15056 rec.woodworking       Hobbyists interested in woodwor
          8  7094 sci.psychology        Topics related to psychology.
          9 13093 soc.culture.celtic    Celtic, Irish, & Welsh culture

                                *** End of Groups ***


       In tin, selecting a newsgroup is done with the arrow keys.   When
       you select a group, you get a listing of the articles:


                           comp.mail.misc (41T 64A 0K 0H)         h=help

           1 +   RIPEM Frequently Noted Vulnerabilities  Marc VanHeyningen
           2 +   RIPEM Frequently Asked Questions        Marc VanHeyningen
           3 +   Mail Archive Server software list       Jonathan I. Kamen
           4 + 1 UNIX Email Software Survey FAQ          Chris Lewis
           5 + 2 PC Eudora and Trumpet Winsock problem   Jim Graham
           6 +   X11 mail reader                         Dominique Marant
           7 +   MIME supporting e-mail                  Tim Goodwin
           8 + 1 IBM User name and Address Server        Wes Spears
           9 + 5 Newbie needs MHS/SMTP question answered Chris Pearce
          10 +   FAQ - pine                              Bruce Lilly
          11 +   FAQ: International E-mail accessibility Olivier M.J. Crep
       -> 12 +   PC E-Mail and Dial-in                   Edward Vielmetti
          13 +   Prodigy Mail Manager "01/07"            an33127@anon.pene
          14 +   Prodigy Mail Manager "02/07"            an33127@anon.pene
          15 +   Prodigy Mail Manager "03/07"            an33127@anon.pene


       tins is a threaded news reader:  replies to a posting are grouped
       together with the original posting, so that the reader can follow
       a thread of discussion. Above, you see the threads, the number of
       replies in each thread, the subject and the author. The plus sign
       (+) indicates that not all postings in the thread have been read.
       Other news readers show other details.   When you select an item,
       it appears on your screen:


       Wed, 01 Sep 1993 07:05:49     comp.mail.misc     Thread  13 of 41
       Lines 27          Re: PC E-Mail and Dial-in          No responses  Edward Vielmetti at Msen, Inc. --  Ann Arbor

       Sherry H. Lake ( wrote:

       :    I am looking for an email package that will allow a user to
       : dial-in to his mail machine download any messages to his local
       : PC, delete the messages from the server and then automatically
       : sign him off. The user can then use his client software (local)
       : to read, compose and reply. He then would have to dial-in again
       : to so his outgoing mail will be uploaded to the server.
       Various POP clients for PCs or Windows Sockets will do roughly
       this. You should look at:

       -  NUPOP (MS-DOS)
       -  Eudora for Windows (Windows)
       -  WinQVT/Net (Windows)
       -  various commercial POP clients listed in the 'alt.winsock'
          directory of commercial Windows systems

       You'll want to look particularly for dial up IP software (SLIP or
       PPP) that makes the process of connecting minimally onerous, e.g.
       by scripting the session so that the users don't have to type
       anything, perhaps by automatically dialing for you when you go to
       read or otherwise open a network connection, and offering a
       reasonable way to disconnect.

       Edward Vielmetti, vice president for research, Msen Inc. Msen Inc., 628 Brooks, Ann Arbor MI 48103
       +1 313 998 4562 (fax: 998 4563)


       Learning more about NETNEWS

       News programs communicate  with each other according  to standard
       protocols,  some of  which are described by  Internet Request For
       Comments (RFC).   Copies of RFCs are  often posted to the network
       and obtainable  from archive  sites.   Current  news-related RFCs
       include the following:

       RFC 977   specifies NNTP, the Network News Transfer Protocol,

       RFC 1036  specifies the format of Usenet articles.

       Some  newsgroups carry  articles and  discussions on  the use  of
       Usenet,   notably:   news.announce.newusers,    news.answers  and

       Many of  the articles  which appear  periodically in  these news-
       groups or in others are also available from by anon-
       ymous FTP or by mail to:

                                    Part 7

                            OTHER TOOLS OF INTEREST


       What is ASTRA

       The  ASTRA service   allows users   to  retrieve  documents  from
       databases known by ASTRA  throughout the network.  Users can send
       their queries to the  ASTRA server  which  in  turn forwards  the
       query  to the  related database servers.  This provides an  easy-
       to-use uniform access method to a large number of databases.

       ASTRA provides the same user interface   for all databases it can
       access, even if  the database servers  have different access lan-
       guages,  such as STAIRS, ISIS or SQL.

       Each  database defined   in ASTRA  has an  abstract  which  holds
       information about the database:  title, name  of the maintainers,
       a brief description of the database, the main topics of the data-
       base  and  its language.   Users  are  advised  to  look  at  the
       abstracts before sending  requests,  to avoid sending requests to
       the wrong databases.

       Some databases actually combine  several different databases that
       deal with the same topics. When a user sends a request for such a
       database, the request is forwarded to all related databases.

       How to get to ASTRA

       Anyone who  can send  electronic mail  to EARN/Bitnet  can access
       ASTRA.  Interactive user interfaces (clients) to ASTRA are avail-
       able for VM and VMS systems  on the EARN/Bitnet network.  For all
       other users, there is a batch language that permits batch queries
       using e-mail.

       Currently  there   are five   ASTRA  servers   installed at   the
       following addresses:

       |                                                               |
       |         ASTRADB@ICNUCEVM.BITNET                               |
       |         ASTRASQL@ICNUCEVM.BITNET                              |
       |         ASTRADB@IFIBDP.BITNET                                 |
       |         ASTRADB@IFIIDG.BITNET                                 |
       |         ASTRADB@IRMKANT.BITNET                                |
       |                                                               |

       Learning more about ASTRA

       An extensive help  file is available by sending  the command HELP

       The LISTSERV  list ASTRA-UG  is used  for the  distribution of  a
       newsletter about  new databases  or new  versions of  the current
       databases.  To subscribe, send the command:

            SUB ASTRA-UG Your Name


       An ASTRA newsletter is also available by sending the command NEWS

       The developers of the ASTRA service may be contacted at:


       What is NETSERV

       NETSERV is a server,  which allows  fast access to data files and
       programs of interest to the  EARN/Bitnet community.  NETSERV pro-
       vides a repository consisting of  information files and programs.
       It allows  users to retrieve files,   to store files and  to sub-
       scribe to  the files of their  choice.  The latter  two functions
       however,  require that the user have a password for NETSERV (this
       is called a privileged  user).

       In order to achieve  a balanced load on the network  and a faster
       response time to  users,  NETSERV uses a  distributed server con-
       cept:  this is achieved by the  installation of a large number of
       servers on the network to ensure that the user can locate a near-
       by server.  All servers communicate with each other to distribute
       updated information and  make it available from each  copy of the

       NETSERV's  file  server  functions  include  retrieving any  file
       present in its  filelists,  storing new versions of  a file,  and
       subscribing to files stored on  the server.  Its file directories
       are arranged  in an hierarchical  method,  with  NETSERV FILELIST
       being on top or at the root  of the filelists.  This filelist can
       be obtained  by sending  a GET  NETSERV FILELIST  command to  any
       NETSERV.  Filelists contain short descriptions of the files,  and
       two access codes for each file. These codes represent the get and
       put privileges required for that file.  These codes are explained
       at the beginning of the NETSERV FILELIST file.

       How to get to NETSERV

       The server  is available  in almost every  country.  To  find the
       closest server  for your area send  a QUERY SERVICE command  to a
       server.   The  following   are   examples   of   NETSERV   server

       |                                                               |
       |         NETSERV@FRMOP11.BITNET                                |
       |         NETSERV@HEARN.BITNET                                  |
       |         NETSERV@BITNIC.BITNET                                 |
       |                                                               |

       In EARN,   there is only one  NETSERV permitted for  one country.
       However,  in some limited cases,  such  as when the country has a
       large number of nodes,  additional  servers may be installed.  In
       any case,  the user is not required  to run NETSERV to be able to
       access and use the server.

       NETSERV accepts e-mail  access from users on  any network.   Com-
       mands to NETSERV should be  placed in the body of the  mail file,
       and not in the Subject: line.

       For users in the EARN/Bitnet  network,  NETSERV is accessible via
       interactive message.  Commands from  privileged users requiring a
       password must be sent this way.

       NETSERV  does not have delivery  limitations,  except that a file
       ordered from NETSERV cannot be ordered again on the same day.

       Learning more about NETSERV

       The server   provides a large helpfile  which can be  obtained by
       sending a GET NETSERV HELPFILE command to any NETSERV.

       A    list   for    NETSERV    maintainers    is   available    as

       Additional information  can  be obtained  from the  NETSERV main-
       tainer, Ulrich Giese at U001212@HEARN.BITNET.


       What is MAILBASE

       Mailbase is  an electronic information  service with much  of the
       same functionality as LISTSERV.  It  allows United Kingdom groups
       to manage their own discussion topics (Mailbase lists)  and asso-
       ciated files.  The  Mailbase service is run as part  of the JANET
       Networked Information Services Project (NISP)  based at Newcastle

       How to get to MAILBASE

       Commands should  be sent in an  electronic mail message  to mail-   More than one command may appear in a mes-
       sage to Mailbase.  Commands may be in any order, in UPPER, lower,
       or MiXeD case.

       Learning more about MAILBASE

       For a summary of Mailbase commands,   send the command help in an
       e-mail message to  For a list of on-line
       documentation about Mailbase, send the command:  index mailbase.

       You can  then use  the send command  to retrieve  those documents
       that interest you.   Eg,  to retrieve a file  of frequently asked
       questions, send the following command:  send mailbase user-faq.

       User support  is also available by  sending queries in  an e-mail
       message to:

       Public files on  Mailbase are also available by  anonymous FTP to


       What is PROSPERO

       Prospero is a  distributed file system.   It  differs from tradi-
       tional distributed file systems in several ways.   In traditional
       file systems,  the mapping of names to  files is the same for all
       users.   Prospero supports user centered naming:  users construct
       customized views  of the files  that are accessible.    A virtual
       system defines this  view and controls the mapping  from names to
       files.   Objects may  be organized in multiple ways  and the same
       object may appear in different virtual systems, or even with mul-
       tiple names in the same virtual system.

       In Prospero,  the global file system  consists of a collection of
       virtual file systems.    Virtual file systems usually  start as a
       copy of a prototype.   The root contains links to files or direc-
       tories selected by the user.

       The Prospero  file system provides tools  that make it  easier to
       keep track of  and organize information in  large systems.   When
       first created,   your virtual  file system  is likely  to contain
       links to directories that organize information in different ways.
       As the master copy of each  of these directories is updated,  you
       will see the changes.  You may customize these directories.   The
       changes you  make to  a customized directory  are only  seen from
       within your own  virtual system,  but changes made  to the master
       copy will also be visible to you.

       Users are encouraged to organize their own projects and papers in
       a manner that  will allow them to  be easily added to  the master
       directory.  For example, users should consider creating a virtual
       directory that contains pointers to copies  of each of the papers
       that they want made available to the outside world.  This virtual
       directory may appear anywhere in the user's virtual system.  Once
       set up,  a link may be added to the master author directory.   In
       this manner,  others will be able  to find this directory.   Once
       added to the master directory, any future changes will be immedi-
       ately available to other users.

       How to get to PROSPERO

       In order to use Prospero, you must be on the international TCP/IP
       network (the Internet) and you must have Prospero running on your

       Before you  can begin  using the Prospero  file system  a virtual
       system must be created for you. However, Prospero, as shipped, is
       configured so  that once  you compile the  clients you  can type:
       vfsetup guest  and start  working right  out of  the box  using a
       guest virtual system at the USC Information Sciences Institute.

       The  latest version  of  PROSPERO is  available  as file  prospe-
       ro.tar.Z for anonymous FTP from in the directory

       Learning more about PROSPERO

       Prospero is being developed by Clifford Newman. Several documents
       and articles describing Prospero by  Newman and others are avail-

       The following files are available  via anonymous FTP from prospe-
                                                                 prospe They are also available through Prospero.

       *   Anonymous FTP: /pub/papers/prospero/,
       *   Prospero:        /papers/subjects/operating-systems/prospero/

       This is a useful first paper to read. It gives a good overview of
       Prospero and what it does. It also describes a bit about the Vir-
       tual System model,  of which  Prospero is a prototype implementa-
       tion. It describes what Prospero does, not how it does it.

       *   Anonymous FTP: /pub/papers/prospero/,
       *   Prospero:        /papers/subjects/operating-systems/prospero/

       This paper describes how Prospero can be used to integrate Inter-
       net information services,  including Gopher,  WAIS,  archie,  and
       World Wide Web.


       What is IRC

       IRC,  Internet Relay Chat,  is a real-time conversational system.
       It is  similar to  the talk  command which  is available  on many
       machines in the Internet.  IRC does everything talk does,  but it
       allows more than 2 users to talk at once,  with access throughout
       the global Internet, It also provides many other useful features.

       IRC is networked over much of  North America,  Europe,  and Asia.
       When you are talking in IRC,   everything you type will instantly
       be transmitted around the world to  other users who are connected
       at the  time.  They can then  type something and respond  to your

       Topics of discussion on IRC are varied.   Technical and political
       discussions are  popular,  especially  when world  events are  in
       progress.  IRC is also a way  to expand your horizons,  as people
       from many countries  and cultures are on,  24 hours  a day.  Most
       conversations are in  English,  but there are  always channels in
       German, Japanese, and Finnish, and occasionally other languages.

       How to get to IRC

       Clients and servers for IRC are  available via anonymous FTP from A few sites offer public access to IRC via Telnet. Two
       such  sites are and   At  both
       sites, you should log in as irc.

       The many server hosts of Internet  Relay Chat throughout the net-
       work are  connected via a  tree structure.   The  various servers
       relay control and message data  among themselves to advertise the
       existence of other  servers,  users,  and the  channels and other
       resources being occupied by those users.

       Fundamental to the operation of IRC  is the concept of a channel.
       All users are on a channel while inside IRC.   You enter the null
       channel first.   You  cannot send any messages until  you enter a
       chatting channel,  unless you have  set up a private conversation
       in some way.  The number of channels is essentially unlimited.

       Learning more about IRC

       To get help while in IRC, type /help and follow the instructions.

       If  you  have  problems,   you   can  contact  Christopher  Davis
       ( or Helen Rose ( - known on IRC as ckd
       and Trillian, respectively.  You can also ask for help on some of
       the operator  channels on  IRC,  for  example #twilight_zone  and

       Various documents on IRC, and the archives of IRC-related mailing
       lists, are available via anonymous FTP from


       What is RELAY

       The RELAY server system  is a set of servers in  the global EARN/
       Bitnet network which broadcast interactive messages from one user
       to other users signed on to the same channel of the RELAY system.
       A user signed on to the closest available RELAY is also virtually
       signed on to all RELAYs which are  linked to it.  Most RELAYs are
       closed during peak hours. Only some RELAYs are up 24 hours a day.

       Each RELAY  server provides service  to  a specific collection of
       one or more nodes designated as a service area.  The functions of
       RELAY are available to EARN/Bitnet  users with access to interac-
       tive messages who have not been  expressly excluded from the sys-
       tem by RELAY management.

       RELAY is  a program which  allows several  people to talk  at the
       same time.   In order  to start,  you must sign on  to a RELAY to
       place your  ID in the current  user list.   You  communicate with
       RELAY through   messages just  as you  would send  messages to  a
       user.  RELAY commands start with a slash(/)  character;  anything
       not  beginning with a slash is   considered a message and is sent
       back out to all other current users.

       How to get to RELAY

       RELAY is available  at the following EARN/Bitnet  addresses,  and
       others.  The nickname of the RELAY machine is in parentheses.

       |                                                               |
       |  RELAY@ASUACAD    (Sun_Devils)   RELAY@PURCCVM  (Purdue)      |
       |  RELAY@AUVM       (Wash_DC)      RELAY@SEARN    (Stockholm)  |
       |  RELAY@BEARN      (Belgium)      RELAY@TAMVM1   (Aggieland)  |
       |  RELAY@BNANDP11   (Namur)        RELAY@TAUNIVM  (Israel)      |
       |  RELAY@CEARN      (Geneva)       RELAY@TECMTYVM (Monterrey)  |
       |  RLY@CORNELLC     (Ithaca_NY)    RELAY@TREARN   (EgeRelay    |
       |  RELAY@CZHRZU1A   (Zurich)       MASRELAY@UBVM  (Buffalo)     |
       |  RELAY@DEARN      (Germany)      RELAY@UFRJ     (RioJaneiro) |
       |  RELAY@DKTC11     (Copenhagen)   RELAY@UIUCVMD  (Urbana_IL)   |
       |  RELAY@FINHUTC    (Finland)      RELAY@USCVM   (LosAngeles) |
       |  RELAY@GITVM1     (Atlanta)      RELAY@UTCVM   (Tennessee)  |
       |  RELAY@GREARN     (Hellas)       RELAY@UWAVM    (Seattle)     |
       |  RELAY@HEARN      (Holland)      RELAY@VILLVM  (Philadelph) |
       |  RELAY@ITESMVF1   (Mexico)       RELAY@VMTECQRO (Queretaro)  |
       |  RELAY@JPNSUT00   (Tokyo)        RELAY@VTBIT   (Va_Tech)    |
       |  RELAY@NDSUVM1    (No_Dakota)    RELAY@WATDCS  (Waterloo)   |
       |  RELAY@NYUCCVM    (NYU)          RELAY@YALEVM   (Yale)        |
       |                                                               |

       RELAY is available to users on the EARN/Bitnet network via inter-
       active message (eg, the TELL command of VM or the SEND command of
       VMS/JNET).   All RELAY server machines are on IBM VM/CMS systems,
       but you do not have to be a VM user in order to use RELAY. If you
       are not in the EARN/Bitnet network, you can not use RELAY.

       CHAT,  a full-screen interface to  send and receive TELL messages
       for VM systems,  is particularly useful for users of RELAY.  CHAT
       is available from any NETSERV.

       Learning more about RELAY

       Upon registration,  the files RELAY INFO  and RELAY USERGUIDE are
       sent to the user.  These two  files give a comprehensive descrip-
       tion of RELAY.

       A brief guide  to RELAY is available from  the EARN documentation
       filelist.  Send mail to LISTSERV@EARNCC.BITNET.    In the body of
       the message, write: GET RELAY MEMO.

                                  Appendix A

                     Freely available networking software

       Below you will  find the location of client  software for several
       of the tools described in this guide (Gopher, WWW,  WAIS and Net-
       news).  This is not a complete  listing of available software for
       any of these tools.

       Gopher clients

       Environment   FTP site & directory     Comments



                      XGOPHER_CLIENT.SHARE    for Wollongong or UCX





                      /pub/gopher/Macintosh    requires MacTCP

                      /util/gopher/gopherapp   requires MacTCP


                      /pub/gopher/PC_client    requires packet driver

                      /public/dos/misc         dosgopher, for PC/TCP

                      /nfs/gopher.exe          for PC-NFS

                      /dos/gopher              for LAN Workplace for DOS

                      /pub/micro/pc-stuff/ms-windows/winsock/apps  Gopherbook

                      /pub/gopher/Unix         xgopher (Athena widgets)

                      /pub/gopher/Unix         moog (Motif)

                      /pub/gopher/Unix/xvgopher     or Xview


       World-Wide Web clients

       Environment   FTP site & directory     Comments

                      /pub/www/src            WWW line-mode browser

                      /pub/WWW/lynx           Lynx browser for vt100 terminals

                      /pub/w3browser          tty-based browser written in perl

                      /pub/www/bin/vms        port of NCSA Mosaic for X

                      /pub/www/bin/mac        requires MacTCP

       MS-Windows    Cello


                      /pub/www/src            tkWWW Browser/Editor

                      /pub/www/src            MidasWWW Browser for X/Motif

                      /pub/www/src            ViolaWWW Browser for X11

                      /Web                    NCSA Mosaic Browser for X11/Motif

                      /pub/www/bin/next       Browser and Editor

       WAIS clients

       Environment   FTP site & directory     Comments

                      /pub/freeware/unix-src  swais






                      /pub/tcpip              PCWAIS

                      /pub/freeware/windows   WNWAIS


                      /pub/freeware/unix-src  gwais

                      /pub/freeware/unix-src  xwais


       Netnews - news reader software

       Environment   FTP site & directory     Comments

       Unix              rn   also available via e-mail to:




       VMS        ANU-NEWS


       VM/CMS            NetNews
                                              also available from LISTSERV@PSUVM



       MVS      NNMVS

       Mac             News

       MS-DOS          Trumpet

       MS-Windows          WTrumpet

       X-Windows     many FTP sites           xrn


       Emacs         most GNU sites           GNUS
                                              for use with GNU Emacs editor

                     most GNU sites           Gnews
                                              for use with GNU Emacs editor


       Preface to the Second Edition



             What is Gopher
             Who can use Gopher
             How to get to Gopher
             Using Gopher
             Learning more about Gopher
          WORLD-WIDE WEB
             What is World-Wide Web
             Who can use World-Wide Web
             How to get to World-Wide Web
             Using World-Wide Web
             Learning more about World-Wide Web


             What is WAIS
             Who can use WAIS
             How to get to WAIS
             Using WAIS
                E-mail access
             Learning more about WAIS


             What is ARCHIE
             Who can use ARCHIE
             How to get to ARCHIE
             Using ARCHIE
                Using a local client
                Using Telnet
                Using electronic mail
             Learning more about ARCHIE


             What is WHOIS
             Who can use WHOIS
             How to get to WHOIS
             Using WHOIS
                Using a local client
                Using Telnet
                Using electronic mail
             Learning more about WHOIS
             What is X.500
             Who can use X.500
             How to get to X.500
             Using X.500
                Using a local client
                Using Telnet or X.25
                Using electronic mail
             Learning more about X.500
             What is NETFIND
             Who can use NETFIND
             How to get to NETFIND
             Using NETFIND
                Local access
                Remote access
             Learning more about NETFIND

       Part 5:  GETTING FILES

             What is TRICKLE
             Who can use TRICKLE
             How to get to TRICKLE
             Using TRICKLE
             Learning more about TRICKLE
             What is BITFTP
             Who can use BITFTP
             How to get to BITFTP
             Using BITFTP
             Learning more about BITFTP


          LISTSERV (Version 1.7f)
             What is LISTSERV
             Who can use LISTSERV
             How to get to LISTSERV
             Using LISTSERV
                Commands for LISTS
                Commands for FILES
                LISTSERV DATABASE Functions
                Commands for INFORMATION
             Learning more about LISTSERV
          NETNEWS (USENET)
             What is NETNEWS
             Who can use NETNEWS
             How to get to NETNEWS
             Using NETNEWS
             Learning more about NETNEWS


             What is ASTRA
             How to get to ASTRA
             Learning more about ASTRA
             What is NETSERV
             How to get to NETSERV
             Learning more about NETSERV
             What is MAILBASE
             How to get to MAILBASE
             Learning more about MAILBASE
             What is PROSPERO
             How to get to PROSPERO
             Learning more about PROSPERO
             What is IRC
             How to get to IRC
             Learning more about IRC
             What is RELAY
             How to get to RELAY
             Learning more about RELAY

       Appendix A:

          Gopher clients
          World-Wide Web clients
          WAIS clients
          Netnews - news reader software