en-US/about_Gopher.help.txt

TOPIC
    about_Gopher

SHORT DESCRIPTION
    Gopher is a distributed document search and retrieval protocol that
    was most popular before HTTP and the Web took off. However, it is
    seeing a resurgence among hobbyists and privacy advocates.

RFC 1436: THE INTERNET GOPHER PROTOCOL
  Network Working Group F. Anklesaria
  Request for Comments: 1436 M. McCahill
                                                              P. Lindner
                                                              D. Johnson
                                                               D. Torrey
                                                              B. Alberti
                                                 University of Minnesota
                                                              March 1993


                      The Internet Gopher Protocol
         (a distributed document search and retrieval protocol)

Status of this Memo

   This memo provides information for the Internet community. It does
   not specify an Internet standard. Distribution of this memo is
   unlimited.

Abstract

   The Internet Gopher protocol is designed for distributed document
   search and retrieval. This document describes the protocol, lists
   some of the implementations currently available, and has an overview
   of how to implement new client and server applications. This
   document is adapted from the basic Internet Gopher protocol document
   first issued by the Microcomputer Center at the University of
   Minnesota in 1991.

Introduction

   gopher n. 1. Any of various short tailed, burrowing mammals of the
   family Geomyidae, of North America. 2. (Amer. colloq.) Native or
   inhabitant of Minnesota: the Gopher State. 3. (Amer. colloq.) One
   who runs errands, does odd-jobs, fetches or delivers documents for
   office staff. 4. (computer tech.) software following a simple
   protocol for burrowing through a TCP/IP internet.

   The Internet Gopher protocol and software follow a client-server
   model. This protocol assumes a reliable data stream; TCP is assumed.
   Gopher servers should listen on port 70 (port 70 is assigned to
   Internet Gopher by IANA). Documents reside on many autonomous
   servers on the Internet. Users run client software on their desktop
   systems, connecting to a server and sending the server a selector (a
   line of text, which may be empty) via a TCP connection at a well-
   known port. The server responds with a block of text terminated by a
   period on a line by itself and closes the connection. No state is
   retained by the server.

   While documents (and services) reside on many servers, Gopher client
   software presents users with a hierarchy of items and directories
   much like a file system. The Gopher interface is designed to
   resemble a file system since a file system is a good model for
   organizing documents and services; the user sees what amounts to one
   big networked information system containing primarily document items,
   directory items, and search items (the latter allowing searches for
   documents across subsets of the information base).

   Servers return either directory lists or documents. Each item in a
   directory is identified by a type (the kind of object the item is),
   user-visible name (used to browse and select from listings), an
   opaque selector string (typically containing a pathname used by the
   destination host to locate the desired object), a host name (which
   host to contact to obtain this item), and an IP port number (the port
   at which the server process listens for connections). The user only
   sees the user-visible name. The client software can locate and
   retrieve any item by the trio of selector, hostname, and port.

   To use a search item, the client submits a query to a special kind of
   Gopher server: a search server. In this case, the client sends the
   selector string (if any) and the list of words to be matched. The
   response yields "virtual directory listings" that contain items
   matching the search criteria.

   Gopher servers and clients exist for all popular platforms. Because
   the protocol is so sparse and simple, writing servers or clients is
   quick and straightforward.

1. Introduction

   The Internet Gopher protocol is designed primarily to act as a
   distributed document delivery system. While documents (and services)
   reside on many servers, Gopher client software presents users with a
   hierarchy of items and directories much like a file system. In fact,
   the Gopher interface is designed to resemble a file system since a
   file system is a good model for locating documents and services. Why
   model a campus-wide information system after a file system? Several
   reasons:

      (a) A hierarchical arrangement of information is familiar to many
      users. Hierarchical directories containing items (such as
      documents, servers, and subdirectories) are widely used in
      electronic bulletin boards and other campus-wide information
      systems. People who access a campus-wide information server will
      expect some sort of hierarchical organization to the information
      presented.

      (b) A file-system style hierarchy can be expressed in a simple
      syntax. The syntax used for the internet Gopher protocol is
      easily understandable, and was designed to make debugging servers
      and clients easy. You can use Telnet to simulate an internet
      Gopher client's requests and observe the responses from a server.
      Special purpose software tools are not required. By keeping the
      syntax of the pseudo-file system client/server protocol simple, we
      can also achieve better performance for a very common user
      activity: browsing through the directory hierarchy.

      (c) Since Gopher originated in a University setting, one of the
      goals was for departments to have the option of publishing
      information from their inexpensive desktop machines, and since
      much of the information can be presented as simple text files
      arranged in directories, a protocol modeled after a file system
      has immediate utility. Because there can be a direct mapping from
      the file system on the user's desktop machine to the directory
      structure published via the Gopher protocol, the problem of
      writing server software for slow desktop systems is minimized.

      (d) A file system metaphor is extensible. By giving a "type"
      attribute to items in the pseudo-file system, it is possible to
      accommodate documents other than simple text documents. Complex
      database services can be handled as a separate type of item. A
      file-system metaphor does not rule out search or database-style
      queries for access to documents. A search-server type is also
      defined in this pseudo-file system. Such servers return "virtual
      directories" or list of documents matching user specified
      criteria.

2. The internet Gopher Model

   A detailed BNF rendering of the internet Gopher syntax is available
   in the appendix...but a close reading of the appendix may not be
   necessary to understand the internet Gopher protocol.

   In essence, the Gopher protocol consists of a client connecting to a
   server and sending the server a selector (a line of text, which may
   be empty) via a TCP connection. The server responds with a block of
   text terminated with a period on a line by itself, and closes the
   connection. No state is retained by the server between transactions
   with a client. The simple nature of the protocol stems from the need
   to implement servers and clients for the slow, smaller desktop
   computers (1 MB Macs and DOS machines), quickly, and efficiently.

   Below is a simple example of a client/server interaction; more
   complex interactions are dealt with later. Assume that a "well-
   known" Gopher server (this may be duplicated, details are discussed
   later) listens at a well known port for the campus (much like a
   domain-name server). The only configuration information the client
   software retains is this server's name and port number (in this
   example that machine is rawBits.micro.umn.edu and the port 70). In
   the example below the F character denotes the TAB character.

 Client: {Opens connection to rawBits.micro.umn.edu at port 70}

 Server: {Accepts connection but says nothing}

 Client: <CR><LF> {Sends an empty line: Meaning "list what you have"}

 Server: {Sends a series of lines, each ending with CR LF}
 0About internet GopherFStuff:About usFrawBits.micro.umn.eduF70
 1Around University of MinnesotaFZ,5692,AUMFunderdog.micro.umn.eduF70
 1Microcomputer News & PricesFPrices/Fpserver.bookstore.umn.eduF70
 1Courses, Schedules, CalendarsFFevents.ais.umn.eduF9120
 1Student-Staff DirectoriesFFuinfo.ais.umn.eduF70
 1Departmental PublicationsFStuff:DP:FrawBits.micro.umn.eduF70
                    {.....etc.....}
 . {Period on a line by itself}
                    {Server closes connection}


   The first character on each line tells whether the line describes a
   document, directory, or search service (characters '0', '1', '7';
   there are a handful more of these characters described later). The
   succeeding characters up to the tab form a user display string to be
   shown to the user for use in selecting this document (or directory)
   for retrieval. The first character of the line is really defining
   the type of item described on this line. In nearly every case, the
   Gopher client software will give the users some sort of idea about
   what type of item this is (by displaying an icon, a short text tag,
   or the like).

   The characters following the tab, up to the next tab form a selector
   string that the client software must send to the server to retrieve
   the document (or directory listing). The selector string should mean
   nothing to the client software; it should never be modified by the
   client. In practice, the selector string is often a pathname or
   other file selector used by the server to locate the item desired.
   The next two tab delimited fields denote the domain-name of the host
   that has this document (or directory), and the port at which to
   connect. If there are yet other tab delimited fields, the basic
   Gopher client should ignore them. A CR LF denotes the end of the
   item.

   In the example, line 1 describes a document the user will see as
   "About internet Gopher". To retrieve this document, the client
   software must send the retrieval string: "Stuff:About us" to
   rawBits.micro.umn.edu at port 70. If the client does this, the
   server will respond with the contents of the document, terminated by
   a period on a line by itself. A client might present the user with a
   view of the world something like the following list of items:


      About Internet Gopher
      Around the University of Minnesota...
      Microcomputer News & Prices...
      Courses, Schedules, Calendars...
      Student-Staff Directories...
      Departmental Publications...



   In this case, directories are displayed with an ellipsis and files
   are displayed without any. However, depending on the platform the
   client is written for and the author's taste, item types could be
   denoted by other text tags or by icons. For example, the UNIX
   curses-based client displays directories with a slash (/) following
   the name; Macintosh clients display directories alongside an icon of
   a folder.

   The user does not know or care that the items up for selection may
   reside on many different machines anywhere on the Internet.

   Suppose the user selects the line "Microcomputer News & Prices...".
   This appears to be a directory, and so the user expects to see
   contents of the directory upon request that it be fetched. The
   following lines illustrate the ensuing client-server interaction:


    Client: (Connects to pserver.bookstore.umn.edu at port 70)
    Server: (Accepts connection but says nothing)
    Client: Prices/ (Sends the magic string terminated by CRLF)
    Server: (Sends a series of lines, each ending with CR LF)
    0About PricesFPrices/AboutusFpserver.bookstore.umn.eduF70
    0Macintosh PricesFPrices/MacFpserver.bookstore.umn.eduF70
    0IBM PricesFPrices/IckFpserver.bookstore.umn.eduF70
    0Printer & Peripheral PricesFPrices/PPPFpserver.bookstore.umn.eduF70
                      (.....etc.....)
    . (Period on a line by itself)
                      (Server closes connection)

3. More details

3.1 Locating services

   Documents (or other services that may be viewed ultimately as
   documents, such as a student-staff phonebook) are linked to the
   machine they are on by the trio of selector string, machine domain-
   name, and IP port. It is assumed that there will be one well-known
   top-level or root server for an institution or campus. The
   information on this server may be duplicated by one or more other
   servers to avoid a single point of failure and to spread the load
   over several servers. Departments that wish to put up their own
   departmental servers need to register the machine name and port with
   the administrators of the top-level Gopher server, much the same way
   as they register a machine name with the campus domain-name server.
   An entry which points to the departmental server will then be made at
   the top level server. This ensures that users will be able to
   navigate their way down what amounts to a virtual hierarchical file
   system with a well known root to any campus server if they desire.

   Note that there is no requirement that a department register
   secondary servers with the central top-level server; they may just
   place a link to the secondary servers in their own primary servers.
   They may indeed place links to any servers they desire in their own
   server, thus creating a customized view of thethe Gopher information
   universe; links can of course point back at the top-level server.
   The virtual (networked) file system is therefore an arbitrary graph
   structure and not necessarily a rooted tree. The top-level node is
   merely one convenient, well-known point of entry. A set of Gopher
   servers linked in this manner may function as a campus-wide
   information system.

   Servers may of course point links at other than secondary servers.
   Indeed servers may point at other servers offering useful services
   anywhere on the internet. Viewed in this manner, Gopher can be seen
   as an Internet-wide information system.

3.2 Server portability and naming

   It is recommended that all registered servers have alias names
   (domain name system CNAME) that are used by Gopher clients to locate
   them. Links to these servers should use these alias names rather
   than the primary names. If information needs to be moved from one
   machine to another, a simple change of domain name system alias
   (CNAME) allows this to occur without any reconfiguration of clients
   in the field. In short, the domain name system may be used to re-map
   a server to a new address. There is nothing to prevent secondary
   servers or services from running on otherwise named servers or ports
   other than 70, however these should be reachable via a primary
   server.

3.3 Contacting server administrators

   It is recommended that every server administrator have a document
   called something like: "About Bogus University's Gopher server" as
   the first item in their server's top level directory. In this
   document should be a short description of what the server holds, as
   well as name, address, phone, and an e-mail address of the person who
   administers the server. This provides a way for users to get word to
   the administrator of a server that has inaccurate information or is
   not running correctly. It is also recommended that administrators
   place the date of last update in files for which such information
   matters to the users.

3.4 Modular addition of services

   The first character of each line in a server-supplied directory
   listing indicates whether the item is a file (character '0'), a
   directory (character '1'), or a search (character '7'). This is the
   base set of item types in the Gopher protocol. It is desirable for
   clients to be able to use different services and speak different
   protocols (simple ones such as finger; others such as CSO phonebook
   service, or Telnet, or X.500 directory service) as needs dictate.
   CSO phonebook service is a client/server phonebook system typically
   used at Universities to publish names, e-mail addresses, and so on.
   The CSO phonebook software was developed at the University of
   Illinois and is also sometimes refered to as ph or qi. For example,
   if a server-supplied directory listing marks a certain item with type
   character '2', then it means that to use this item, the client must
   speak the CSO protocol. This removes the need to be able to
   anticipate all future needs and hard-wire them in the basic Internet
   Gopher protocol; it keeps the basic protocol extremely simple. In
   spite of this simplicity, the scheme has the capability to expand and
   change with the times by adding an agreed upon type-character for a
   new service. This also allows the client implementations to evolve
   in a modular fashion, simply by dropping in a module (or launching a
   new process) for some new service. The servers for the new service
   of course have to know nothing about Internet Gopher; they can just
   be off-the shelf CSO, X.500, or other servers. We do not however,
   encourage arbitrary or machine-specific proliferation of service
   types in the basic Gopher protocol.

   On the other hand, subsets of other document retrieval schemes may be
   mapped onto the Gopher protocol by means of "gateway-servers".
   Examples of such servers include Gopher-to-FTP gateways, Gopher-to-
   archie gateways, Gopher-to-WAIS gateways, etc. There are a number of
   advantages of such mechanisms. First, a relatively powerful server
   machine inherits both the intelligence and work, rather than the more
   modest, inexpensive desktop system that typically runs client
   software or basic server software. Equally important, clients do not
   have to be modified to take advantage of a new resource.

3.5 Building clients

   A client simply sends the retrieval string to a server if it wants to
   retrieve a document or view the contents of a directory. Of course,
   each host may have pointers to other hosts, resulting in a "graph"
   (not necessarily a rooted tree) of hosts. The client software may
   save (or rather "stack") the locations that it has visited in search
   of a document. The user could therefore back out of the current
   location by unwinding the stack. Alternatively, a client with
   multiple-window capability might just be able to display more than
   one directory or document at the same time.

   A smart client could cache the contents of visited directories
   (rather than just the directory's item descriptor), thus avoiding
   network transactions if the information has been previously
   retrieved.

   If a client does not understand what a say, type 'B' item (not a core
   item) is, then it may simply ignore the item in the directory
   listing; the user never even has to see it. Alternatively, the item
   could be displayed as an unknown type.

   Top-level or primary servers for a campus are likely to get more
   traffic than secondary servers, and it would be less tolerable for
   such primary servers to be down for any long time. So it makes sense
   to "clone" such important servers and construct clients that can
   randomly choose between two such equivalent primary servers when they
   first connect (to balance server load), moving to one if the other
   seems to be down. In fact, smart client implementations do this
   clone server and load balancing. Alternatively, it may make sense to
   have the domain name system return one of a set of redundant of
   server's IP address to load balance betwen redundant sets of
   important servers.

3.6 Building ordinary internet Gopher servers

   The retrieval string sent to the server might be a path to a file or
   directory. It might be the name of a script, an application or even
   a query that generates the document or directory returned. The basic
   server uses the string it gets up to but not including a CR-LF or a
   TAB, whichever comes first.

   All intelligence is carried by the server implementation rather than
   the protocol. What you build into more exotic servers is up to you.
   Server implementations may grow as needs dictate and time allows.

3.7 Special purpose servers

   There are two special server types (beyond the normal Gopher server)
   also discussed below:

      1. A server directory listing can point at a CSO nameserver (the
      server returns a type character of '2') to allow a campus
      student-staff phonebook lookup service. This may show up on the
      user's list of choices, perhaps preceded by the icon of a phone-
      book. If this item is selected, the client software must resort
      to a pure CSO nameserver protocol when it connects to the
      appropriate host.

      2. A server can also point at a "search server" (returns a first
      character of '7'). Such servers may implement campus network (or
      subnet) wide searching capability. The most common search servers
      maintain full-text indexes on the contents of text documents held
      by some subset of Gopher servers. Such a "full-text search
      server" responds to client requests with a list of all documents
      that contain one or more words (the search criteria). The client
      sends the server the selector string, a tab, and the search string
      (words to search for). If the selector string is empty, the client
      merely sends the search string. The server returns the equivalent
      of a directory listing for documents matching the search criteria.
      Spaces between words are usually implied Boolean ANDs (although in
      different implementations or search types, this may not
      necessarily be true).

   The CSO addition exists for historical reasons: at time of design,
   the campus phone-book servers at the University of Minnesota used the
   CSO protocol and it seemed simplest to just engulf them. The index-
   server is however very much a Gopher in spirit, albeit with a slight
   twist in the meaning of the selector-string. Index servers are a
   natural place to incorperate gateways to WAIS and WHOIS services.

3.7.1 Building CSO-servers

   A CSO Nameserver implementation for UNIX and associated documentation
   is available by anonymous ftp from uxa.cso.uiuc.edu. We do not
   anticipate implementing it on other machines.

3.7.2 Building full-text search servers

   A full-text search server is a special-purpose server that knows
   about the Gopher scheme for retrieving documents. These servers
   maintain a full-text index of the contents of plain text documents on
   Gopher servers in some specified domain. A Gopher full-text search
   server was implemented using several NeXTstations because it was easy
   to take advantage of the full-text index/search engine built into the
   NeXT system software. A search server for generic UNIX systems based
   on the public domain WAIS search engine, is also available and
   currently an optional part of the UNIX gopher server. In addition,
   at least one implementation of the gopher server incorperates a
   gateway to WAIS servers by presenting the WAIS servers to gopherspace
   as full-text search servers. The gopher<->WAIS gateway servers does
   the work of translating from gopher protocol to WAIS so unmodified
   gopher clients can access WAIS servers via the gateway server.

   By using several index servers (rather than a monolithic index
   server) indexes may be searched in parallel (although the client
   software is not aware of this). While maintaining full-text indexes
   of documents distributed over many machines may seem a daunting task,
   the task can be broken into smaller pieces (update only a portion of
   the indexes, search several partial indexes in parallel) so that it
   is manageable. By spreading this task over several small, cheap (and
   fast) workstations it is possible to take advantage of fine-grain
   parallelism. Again, the client software is not aware of this. Client
   software only needs to know that it can send a search string to an
   index server and will receive a list of documents that contain the
   words in the search string.

3.8 Item type characters

   The client software decides what items are available by looking at
   the first character of each line in a directory listing. Augmenting
   this list can extend the protocol. A list of defined item-type
   characters follows:

   0 Item is a file
   1 Item is a directory
   2 Item is a CSO phone-book server
   3 Error
   4 Item is a BinHexed Macintosh file.
   5 Item is DOS binary archive of some sort.
       Client must read until the TCP connection closes. Beware.
   6 Item is a UNIX uuencoded file.
   7 Item is an Index-Search server.
   8 Item points to a text-based telnet session.
   9 Item is a binary file!
       Client must read until the TCP connection closes. Beware.
   + Item is a redundant server
   T Item points to a text-based tn3270 session.
   g Item is a GIF format graphics file.
   I Item is some kind of image file. Client decides how to display.

   Characters '0' through 'Z' are reserved. Local experiments should
   use other characters. Machine-specific extensions are not
   encouraged. Note that for type 5 or type 9 the client must be
   prepared to read until the connection closes. There will be no
   period at the end of the file; the contents of these files are binary
   and the client must decide what to do with them based perhaps on the
   .xxx extension.

3.9 User display strings and server selector strings

   User display strings are intended to be displayed on a line on a
   typical screen for a user's viewing pleasure. While many screens can
   accommodate 80 character lines, some space is needed to display a tag
   of some sort to tell the user what sort of item this is. Because of
   this, the user display string should be kept under 70 characters in
   length. Clients may truncate to a length convenient to them.

4. Simplicity is intentional

   As far as possible we desire any new features to be carried as new
   protocols that will be hidden behind new document-types. The
   internet Gopher philosophy is:

      (a) Intelligence is held by the server. Clients have the option
      of being able to access new document types (different, other types
      of servers) by simply recognizing the document-type character.
      Further intelligence to be borne by the protocol should be
      minimized.

      (b) The well-tempered server ought to send "text" (unless a file
      must be transferred as raw binary). Should this text include
      tabs, formfeeds, frufru? Probably not, but rude servers will
      probably send them anyway. Publishers of documents should be
      given simple tools (filters) that will alert them if there are any
      funny characters in the documents they wish to publish, and give
      them the opportunity to strip the questionable characters out; the
      publisher may well refuse.

      (c) The well-tempered client should do something reasonable with
      funny characters received in text; filter them out, leave them in,
      whatever.

Appendix

   Paul's NQBNF (Not Quite BNF) for the Gopher Protocol.

   Note: This is modified BNF (as used by the Pascal people) with a few
          English modifiers thrown in. Stuff enclosed in '{}' can be
          repeated zero or more times. Stuff in '[]' denotes a set of
          items. The '-' operator denotes set subtraction.


Directory Entity

    CR-LF ::= ASCII Carriage Return Character followed by Line Feed
                character.

    Tab ::= ASCII Tab character.

    NUL ::= ASCII NUL character.

    UNASCII ::= ASCII - [Tab CR-LF NUL].

    Lastline ::= '.'CR-LF.

    TextBlock ::= Block of ASCII text not containing Lastline pattern.

    Type ::= UNASCII.

    RedType ::= '+'.

    User_Name ::= {UNASCII}.

    Selector ::= {UNASCII}.

    Host ::= {{UNASCII - ['.']} '.'} {UNASCII - ['.']}.

    Note: This is a Fully Qualified Domain Name as defined in RFC 1034.
        (e.g., gopher.micro.umn.edu) Hosts that have a CR-LF
        TAB or NUL in their name get what they deserve.

    Digit ::= '0' | '1' | '2' | '3' | '4' | '5' | '6' | '7' | '8' | '9' .

    DigitSeq ::= digit {digit}.

    Port ::= DigitSeq.

    Note: Port corresponds the the TCP Port Number, its value should
        be in the range [0..65535]; port 70 is officially assigned
        to gopher.

Notes:

   It is *highly* recommended that the User_Name field contain only
   printable characters, since many different clients will be using
   it. However if eight bit characters are used, the characters
   should conform with the ISO Latin1 Character Set. The length of
   the User displayable line should be less than 70 Characters; longer
   lines may not fit across some screens.

   The Selector string should be no longer than 255 characters.


Menu Entity

    Menu ::= {DirEntity} Lastline.


Menu Transaction (Type 1 item)

    C: Opens Connection
    S: Accepts Connection
    C: Sends Selector String
    S: Sends Menu Entity

   Connection is closed by either client or server (typically server).


Textfile Entity

    TextFile ::= {TextBlock} Lastline

    Note: Lines beginning with periods must be prepended with an extra
        period to ensure that the transmission is not terminated early.
        The client should strip extra periods at the beginning of the line.


TextFile Transaction (Type 0 item)

    C: Opens Connection.
    S: Accepts connection
    C: Sends Selector String.
    S: Sends TextFile Entity.

   Connection is closed by either client or server (typically server).

   Note: The client should be prepared for the server closing the
          connection without sending the Lastline. This allows the
          client to use fingerd servers.


Full-Text Search Transaction (Type 7 item)

    Word ::= {UNASCII - ' '}
    BoolOp ::= 'and' | 'or' | 'not' | SPACE
    SearchStr ::= Word {{SPACE BoolOp} SPACE Word}

    C: Opens Connection.
    C: Sends Selector String, Tab, Search String.
    S: Sends Menu Entity.

    Note: In absence of 'and', 'or', or 'not' operators, a SPACE is
        regarded as an implied 'and' operator. Expression is evaluated
        left to right. Further, not all search engines or search
        gateways currently implemented have the boolean operators
        implemented.

Binary file Transaction (Type 9 or 5 item)

    C: Opens Connection.
    S: Accepts connection
    C: Sends Selector String.
    S: Sends a binary file and closes connection when done.


Syntactic Meaning for Directory Entities


    The client should interpret the type field as follows:

    0 The item is a TextFile Entity.
        Client should use a TextFile Transaction.

    1 The item is a Menu Entity.
        Client should use a Menu Transaction.

    2 The information applies to a CSO phone book entity.
        Client should talk CSO protocol.

    3 Signals an error condition.

    4 Item is a Macintosh file encoded in BINHEX format

    5 Item is PC-DOS binary file of some sort. Client gets to decide.

    6 Item is a uuencoded file.

    7 The information applies to a Index Server.
        Client should use a FullText Search transaction.

    8 The information applies to a Telnet session.
        Connect to given host at given port. The name to login as at this
        host is in the selector string.

    9 Item is a binary file. Client must decide what to do with it.

    + The information applies to a duplicated server. The information
        contained within is a duplicate of the primary server. The primary
        server is defined as the last DirEntity that is has a non-plus
        "Type" field. The client should use the transaction as defined by
        the primary server Type field.

    g Item is a GIF graphic file.

    I Item is some kind of image file. Client gets to decide.

    T The information applies to a tn3270 based telnet session.
        Connect to given host at given port. The name to login as at this
        host is in the selector string.

Security Considerations

   Security issues are not discussed in this memo.

Authors' Addresses

   Farhad Anklesaria
   Computer and Information Services, University of Minnesota
   Room 152 Shepherd Labs
   100 Union Street SE
   Minneapolis, MN 55455

   Phone: (612) 625 1300
   EMail: fxa@boombox.micro.umn.edu

LINKS
    https://www.rfc-editor.org/rfc/rfc1436

KEYWORDS
    Gopher
    specification
    RFC1436







Anklesari, McCahill, Lindner, Johnson, Torrey & Alberti [Page 15]
 
RFC 1436 Gopher March 1993


   Mark McCahill
   Computer and Information Services, University of Minnesota
   Room 152 Shepherd Labs
   100 Union Street SE
   Minneapolis, MN 55455

   Phone: (612) 625 1300
   EMail: mpm@boombox.micro.umn.edu


   Paul Lindner
   Computer and Information Services, University of Minnesota
   Room 152 Shepherd Labs
   100 Union Street SE
   Minneapolis, MN 55455

   Phone: (612) 625 1300
   EMail: lindner@boombox.micro.umn.edu


   David Johnson
   Computer and Information Services, University of Minnesota
   Room 152 Shepherd Labs
   100 Union Street SE
   Minneapolis, MN 55455

   Phone: (612) 625 1300
   EMail: dmj@boombox.micro.umn.edu


   Daniel Torrey
   Computer and Information Services, University of Minnesota
   Room 152 Shepherd Labs
   100 Union Street SE
   Minneapolis, MN 55455

   Phone: (612) 625 1300
   EMail: daniel@boombox.micro.umn.edu


   Bob Alberti
   Computer and Information Services, University of Minnesota
   Room 152 Shepherd Labs
   100 Union Street SE
   Minneapolis, MN 55455

   Phone: (612) 625 1300
   EMail: alberti@boombox.micro.umn.edu



Anklesari, McCahill, Lindner, Johnson, Torrey & Alberti [Page 16]