diff options
Diffstat (limited to 'mail/smail/files/EDITME')
| -rw-r--r-- | mail/smail/files/EDITME | 1227 |
1 files changed, 1227 insertions, 0 deletions
diff --git a/mail/smail/files/EDITME b/mail/smail/files/EDITME new file mode 100644 index 000000000000..a6321c9898a0 --- /dev/null +++ b/mail/smail/files/EDITME @@ -0,0 +1,1227 @@ +# @(#)conf/EDITME-dist 1.50 9/10/92 01:07:17 +# +# EDITME - edit me to perform high level configuration of smail +# +# The EDITME file is used by makefiles to build individual defs.h, +# defs.sh and defs.sed files, which in turn control the build/install +# process of smail. +# +# The EDITME-dist file servers as a reference copy only. Patches to +# smail will refer to EDITME-dist rather than EDITME. Only the EDITME +# file is used in the build/install process. Any editing should be done +# to the file EDITME, which should be located in the same directory as +# EDITME-dist. +# +# The EDITME is a /bin/sh shell script. That is, comments begin with +# the first non-quoted/non-escaped '#' character. Values are set +# using NAME=VALUE or NAME=, with no spaces around the equal sign. +# Values that have spaces or special shell characters should be +# qouted and/or escaped. Consult the sh(1) man page for further details. +# +# Values that are indicated as optional may, in some cases, not be set. +# All other values MUST be set to some value. The NAME= form is +# equivalent to not setting the value, with the exception of TEST_BASE. +# +# An alternate path to the EDITME file can be specified through setting +# the environment variable SMAIL_EDITME. This can either be a full +# pathname or a relateive pathname, which will be referenced relative +# to the smail conf directory. Care must be taken with dependencies in +# makefiles when the path to the EDITME file has changed. If this is +# a permanent change, then type "make depend" at the top of the smail +# source tree. Otherwise the conf/lib/mkdefs.sh shell script can be +# run in affected directories to create new defs files. + + +# CFLAGS - the base set of flags for the cc command +# +# optional +# +# When debugging smail, it is useful to set this to "-g". Otherwise +# "-O" is normally used to optimize the code produced. +# +# NOTE: I get core dumps on SCO MPX 2.0 when using the bind router (I +# haven't tried other SCO releases). Whatever the problem is, it +# goes away when I compile without -O. If anybody finds the cause, +# please let me know. For now, if you get core dumps when using the +# bind router on SCO, try compiling without -O. + +CFLAGS=-O # common + +# CC - the C compiler to use for compiling smail +# +# optional (defaults to cc) +# +# CAUTION: gcc does not work with smail on all architecures. In +# particular, problems have been reported on the SunOS platform. The +# typical problem is that gcc uses different structure-passing +# conventions for function calls. This is a problem for a few +# networking calls which pass around in_addr structures. +# +# Some conf/os files set an alternate default value for CC. In +# particular, the SunOS CC defaults to /bin/cc, to ensure that +# /usr/sys5/cc is not used. + +#CC=gcc # use the GNU C Compiler + + +# LDFLAGS - flags to pass to the loader +# +# optional +# +# These flags are passed to the cc command before any object +# modules or libraries. + +#LDFLAGS=-L/usr/ucblib # use an alternate directory for libs +#LDFLAGS="-L/usr/ucblib -Bstatic" # and don't use shared libraries + + +# LIBS - libraries to include in binaries +# +# optional +# +# These libraries are passed to the cc command before any libraries +# mentioned in the conf/os file. It can be used to link in additional +# libraries before the system libraries. +# +# If you wish to use the GNU dbm library, then add -lgdbm to LIBS +# (perhaps with a -L to indicate the directory containing gdbm), +# and then add NDBM to the HAVE variable. + +#LIBS=-lgdbm # use the GNU dbm library +#LIBS="-L/usr/local/lib -lgdbm" # if gdbm is in /usr/local/lib +#LIBS="-L/usr/local/lib -lgdbm -lalloca" # load alloca function from somewhere + + +# OSLIBS - ending libraries to include in binaries +# +# optional +# +# These libraries are passed to the cc command after all other +# libraries from any other source. For example, this can be used +# to force the use of an alternate C library, or a shared version +# of the C library. +# +# Some conf/os files already specify use of the shared C library, and +# some necessary system libraries. + +#OSLIBS=-lc_s # SVR3 shared C library +#OSLIBS="-lc_s -lcposix" # POSIX and shared C libs + + +# STRIP - define this if you wish to strip binaries when installing +# +# optional +# +# The smail binaries will only be striped on installation if STRIP=yes. +# This may save some disk space at the expence of the ability to do +# symbolic debugging. + +STRIP=yes + + +# USE_SYMLINKS - define this if you wish to use symlinks when installing +# +# optional +# +# The smail binaries are normally installed using hardlinks between +# the various copies of the smail binary. Where hardlinks cannot be +# made, real copies are made. To use symbolic links instead of +# hard links and copies, set USE_SYMLINKS=yes. This will cause only +# one copy of the smail binary to be installed. Do not define this +# if your system does not have symbolic links. + +USE_SYMLINKS=yes + + +# OS_TYPE - define the operating system type for your machine +# +# required +# +# Consult the conf/os directory for a complete list of recognized +# operating system types. The names given below may not represent +# all of the operating systems for which an OS description file +# is available. + +#OS_TYPE=aix3.1 # IBM AIX 3.1 +#OS_TYPE=aix3.2 # IBM AIX 3.2 +#OS_TYPE=bsd4.2 # 4.2 BSD systems +#OS_TYPE=bsd4.3 # 4.3 BSD systems +OS_TYPE=freebsd2.0 # FreeBSD 2.0 (BSD 4.4Lite) +#OS_TYPE=cpc3.0 # Counterpoint C-XIX Release 3.0 +#OS_TYPE=forpro # fortune FOR:Pro 2.1 to 3.1 +#OS_TYPE=sco3.2 # SCO UNIX 3.2 +#OS_TYPE=sco3.2+tcp # SCO UNIX 3.2 with SCO TCP/IP +#OS_TYPE=sun_os3 # Sun/OS version 3.x +#OS_TYPE=sun_os4 # Sun/OS version 4.x +#OS_TYPE=sys5.2 # System V release 2 +#OS_TYPE=sys5.3 # System V release 3 +#OS_TYPE=sys5.4 # System V release 4 +#OS_TYPE=ultrix1.2 # DEC Ultrix release 1.2 +#OS_TYPE=unixpc3.0 # AT&T Unix PC (3b1) release 3.0 +#OS_TYPE=unixpc3.5 # AT&T Unix PC (3b1) release 3.5 +#OS_TYPE=uts1.2 # UTS/580 release 1.2 +#OS_TYPE=uts2.0 # UTS/580 release 2.0 +#OS_TYPE=xenix5 # SCO Xenix System V +#OS_TYPE=xenix5+tcp # SCO Xenix System V with TCP/IP +#OS_TYPE=isc2.2.1 # Interactive release 2.2.1 +#OS_TYPE=delta # Motorola delta +#OS_TYPE=mips-bsd4.3 # MIPS Risc/OS in bsd universe +#OS_TYPE=3b2-s5.3+win # 3b2 with SVR3.2 and WIN TCP/IP + + +# ARCH_TYPE - architecture type +# +# required +# +# The ARCH_TYPE of 16bit refers to machines with 16 bit words. This +# mode is untested. It is assumed that 16bit machines have extended +# address spaces as smail is more than 64k bytes long. + +ARCH_TYPE=32bit # common +#ARCH_TYPE=small32bit # 32 bit without reasonable VM +#ARCH_TYPE=16bit # untested + + +# DRIVER_CONFIGURATION - configuration file describing smail drivers +# +# optional +# +# By default, the os configuration file (conf/os/OS_TYPE) refers to +# the file that describes the available director, router and transport +# drivers. +# +# For example on BSD-based OS_TYPEs, the os configuration file refers to +# a driver configuration that uses BSD networking. A BSD based system +# could not set DRIVER_CONFIGURATION and get the BSD netwokring based +# drivers; or a system could set DRIVER_CONFIGURATION=unix-generic and +# disable BSD networking drivers. +# +# Currently few conf/os files specify the arpa-network driver +# configuration (which includes support for the BIND server). +# If you want to use this, you must usually set DRIVER_CONFIGURATION +# explicitly. You may also have to add BIND to the HAVE list (later +# in this file) to use the arpa-network driver configuration. +# +# IMPORTANT NOTE FOR INTERNET HOSTS +# +# NOTE: The bind router driver is not configured into the compiled-in +# routers, so you will need to install a routers file to use it. Get +# a routers file from samples/generic/routers and copy it to +# /usr/lib/smail. Comment out the first version of inet_hosts from +# that file, and uncomment the second (the bind version) of the +# router. You may also want to uncomment the forcepaths router, if +# you forward for any hosts in the UUCP zone. + +#DRIVER_CONFIGURATION=unix-generic # no BSD networking +#DRIVER_CONFIGURATION=$ROOT/mydriver.cf # customized configuration +#DRIVER_CONFIGURATION=arpa-network # include bind router + + +# LMAIL - where the real /bin/mail resides +# +# optional +# +# On many System V machines, the /bin/mail program will deliver what +# it thinks is local mail directly into user mailboxes, and may also +# call uucp, or some other remote delivery mechanism. The /bin/mail +# program will therefore can bypass smail. +# +# A solution to this program is to move the /bin/mail supplied with +# your system to another name, LMAIL, and to then install the binmail +# program, in pd/binmail as /bin/mail. If binmail is invoked to read +# mail, the LMAIL (old /bin/mail) is called, otherwise SMAIL is called +# to deliver mail. +# +# If LMAIL is defined below and the specified file does not exist, +# then the binmail makefile install rule will attempt to move +# /bin/mail to LMAIL and then copy binmail into /bin/mail. If LMAIL +# is defined and the file LMAIL does exist, then the binmail +# makefile will refuse to install binmail. +# +# If LMAIL is not defined then /bin/mail is not touched. +# +# NOTE: due to potential problems or confusion with binmail +# installation, binmail is not automatically built and installed +# from higher level makefiles. One must go to the pd/binmail +# directory and "make install" explicitly. + +#LMAIL=/bin/lmail # standard System V + + +# MISC_DEFINES - miscellaneous definitions +# +# optional +# +# A colon-separated list that names miscellaneous macros to define +# for C programs and shell scripts. Each entry in the list can be +# a simple name, or the name can be followed by an equal sign to +# provide a value. For example: +# +# MISC_DEFINES="void=int:DECLARE_UTIMBUF:NO_FORWARDTO_FILE" +# +# Defines void to be int for compilers that do not correctly support +# the void type, and declares the macro DECLARE_UTIMBUF, so that smail +# will not try to get the utimbuf structure from a header file. +# +# +# IMPORTANT FOR INTERACTIVE UNIX USERS: +# +# Some versions of Interactive UNIX have a <sys/socket.h> header +# file that directly includes the <time.h> header file, which is +# not protected against multiple inclusion. If your system has +# this problem, you will encounter a multiple declaration problem +# when compiling src/modes.o. To repair the situation, use: +# +# MISC_DEFINES=ISC_SOCKET_TIME_BUG +# +# +# IMPORTANT FOR BSD AND SUNOS USERS: +# +# Some mail readers compare access and modify times to determine +# whether mail has been delivered which has not been read. The local +# mail transport (using the appendfile driver) preserves access time +# for files so that this will work. However, if smail is configured +# to use the System V convention of reading forwarding information +# from user mailbox files, then this will be defeated (the forwardto +# director does not preserve access times). +# +# Since BSD systems (and System V systems running Smail) have little +# reason to use the System V forwarding convention, it is sufficient +# to disable it. To disable the System V forwarding convention, add +# NO_FORWARDTO_FILE to MISC_DEFINES. +# +# +# IMPORTANT FOR USERS OF OLD BIND RELEASES +# +# The bind resolver can make use of features in newer releases of +# BIND (named). In particular, it will use the res_search function, +# rather than older res_mkquery function. If you do not have the +# new BIND release (e.g., res_search is listed as an undefined +# variable when you link smail), then add OBSOLETE_RESOLVER to +# MISC_DEFINES. Also, the res_search function does not appear to +# interract correctly with the 'defnames' attribute to the bind +# driver, on some versions of the resolv library, so beware. If bind +# doesn't work for you, try defining OBSOLETE_RESOLVER. +# +# +# IMPORTANT FOR USERS OF SOME SEQUENT OS RELEASES +# +# Some releases of the Sequent C compiler recognize private as a keyword. +# However, smail uses this as a variable or structure name in several +# places. To get around this problem use: +# +# MISC_DEFINES=private=smail_private_tag +# +# +# IMPORTANT FOR NON-POSIX OPERATING SYSTEMS +# +# Smail uses the utime() system call to set file times in a few places. +# All U*IX releases appear to support this call, although not all +# systems define the structure (utimbuf) used as an argument to the +# call. In particular, true 4.2 and 4.3BSD do not declare this +# structure (at least not the original, non-POSIX, 4.3BSD release). +# +# The bsd4.2 and bsd4.3 conf/os files define DECLARE_UTIMBUF to cause +# smail itself to define this structure. However, this may fail if +# your system is extended to define the utimbuf structure in an +# incompatible fashion. This can be disabled by defining +# NO_DECLARE_UTIMBUF in MISC_DEFINES. +# +# For other operating systems, you may need to add DECLARE_UTIMBUF +# in MISC_DEFINES. Some releases of Interactive U*IX require this. +# Other releases don't (I don't have a list). +# +# +# IMPORTANT FOR USERS OF GCC ON NON-ANSI C SYSTEMS +# +# If you are using a C compiler that defines __STDC__ (such as gcc) but +# you are running into problems due to non-ANSI #include files or missing +# ANSI definitions, then you can define NO_ANSI_C. + +#MISC_DEFINES=ISC_SOCKET_TIME_BUG +#MISC_DEFINES=NO_FORWARDTO_FILE # recommended for BSD systems +#MISC_DEFINES=ISC_SOCKET_TIME_BUG:NO_FORWARDTO_FILE +#MISC_DEFINES=OBSOLETE_RESOLVER +#MISC_DEFINES=DECLARE_UTIMBUF # if smail must define utimbuf struct +#MISC_DEFINES=NO_DECLARE_UTIMBUF # if smail must _not_ define utimbuf +#MISC_DEFINES=private=smail_private_tag # for sequent +#MISC_DEFINES=NO_ANSI_C # don't believe __STDC__ + + +# HAVE - miscellaneous supported features +# +# optional +# +# A colon-separated list of features that the local system supports +# in addition to those defined in the OS configuration file. Some +# interesting features you might want to list are: +# +# NDBM - this system supports the new DBM library +# functions introduced in 4.3BSD. +# DBM - this system supports the old DBM library +# distributed with older BSD systems and +# many System V systems. This *requires* the +# dbmclose() function, which does not exist +# in most historic dbm implementations. +# HDB_UUCP - this system uses HoneyDanBer UUCP, rather +# than a traditional version of UUCP. +# NIALIAS - aliases with NeXT NetInfo databases. +# BIND - the system supports the bind resolver library +# +# Defining NDBM or DBM may require the specification of additional +# libraries, in LIBS or in OSLIBS. See conf/os/template for a more +# complete list of recognized features, and for more detailed +# descriptions. If neither NDBM or DBM is specified, Ozan Yigit's +# sdbm (an ndbm workalike) will be used. +# +# CAUTION: The HoneyDanBer UUCP in System V Release 4.0 uses multi- +# letter message grade values, rather than the traditional +# single-letter message grades in earlier versions. Smail +# presumes that message grades are single letters, so Smail and +# the default grades for SVR4.0 are incompatible. +# +# As a result of this incompatiblity, either HDB_UUCP must not +# be defined with SVR4.0, or the message grades that smail +# uses will have to be added to the file /etc/uucp/Grades. +# The list of grades that smail will use is specified in the +# grades variable (whose default value can be specified by +# setting GRADES in this file). Grades are specified by +# providing a longer string in the Precedence: field. The +# grade letters for the default value of GRADES are: +# +# 9 - special-delivery +# A - air-mail +# C - first-class (also the default message grade) +# a - bulk +# n - junk +# +# To add the default letters, you might add the following lines +# to /etc/uucp/Grades: +# +# 9 9 Any User Any +# A A Any User Any +# C C Any User Any +# a a Any User Any +# n n Any User Any + +#HAVE=HDB_UUCP # have HoneyDanBer UUCP +#HAVE=NDBM # have the new DBM functions +#HAVE=DBM # have the old DBM functions +#HAVE=HDB_UUCP:NDBM # new DBM and HoneyDanBer UUCP +HAVE=BIND + + +# UUCP_ZONE - this host is within the UUCP zone +# +# optional (defaults to undefined) +# +# Smail supports both strictly compliant SMTP transfers, and a +# modification the the SMTP protocol for use within the UUCP zone. +# The modified protocol transmits sender and recipient addresses in +# accordance with the RFC976 specification, and provides behavior that +# is more likely to work with other mailers in the UUCP zone. +# Transferring mail using SMTP over private IP networks can work well +# within the UUCP zone, except for the fact that the SMTP +# specification doesn't work as well in the presence of generated +# UUCP-zone routes form path files. +# +# The built-in smail transport definitions include transport +# definitions that provide either behavior: inet_zone_smtp and +# uucp_zone_smtp. The first provides conformant behavior and should +# be used when transferring mail to sites on the Internet. The +# second, uucp_zone_smtp, provides the modified protocol. +# +# An additional builtin transport protocol, smtp, is defined as having +# the same behavior as inet_zone_smtp or uucp_zone_smtp, depending +# upon whether UUCP_ZONE is set. To make the the basic smtp transport +# use the modified behavior for the UUCP zone, set UUCP_ZONE to true. +# +# In the presence of generated routes from paths files, the +# Internet-zone version of the SMTP transport will generate route-addr +# addresses. The introduction of RFC1123 removes the requirement that +# route-addrs work as routes, preferring instead the use of the % +# operator for routing. Intermixed % and ! operators, particularly in +# the presence of an @ operator (e.g., veritas!mitsu%tron@apple.com), +# are sufficiently ambiguous that it is undesirable to use with ! or % +# operators for routing when strict SMTP compliance is needed. This +# is one of the motivations for introducing the modified UUCP-zone +# protocol. However, generating route-addr addresses works well for +# transferring mail between nodes running Smail 3.1. +# +# NOTE: the UUCP_ZONE flag does not affect the batched SMTP transports +# that operate through UUCP. The uusmtp and demand_uusmtp transports +# always operate with the modified UUCP-zone protocol. Two additional +# transports, inet_uusmtp and inet_demand_uusmtp use compliant +# protocols. + +#UUCP_ZONE=true # this site is in the UUCP zone +#UUCP_ZONE= # this site is on the Internet + + +# HOSTNAMES - names for the local host +# +# optional +# +# The HOSTNAMES value is a colon-separated list of names your system. +# This does not need to include the name given by UUCP_NAME. If +# HOSTNAMES is set, then the first name in the list will be used as the +# canonical name of your system. That is, the name by which your +# machine can be uniquely named across all networks. +# +# Normally HOSTNAMES is not set. In this case the HOSTNAMES value will +# be computed at run-time from the DOMAINS value and the system-dependently +# computed name of your system. +# +# If you are registered in a domain, you might consider seting HOSTNAMES +# to a list such as: sitename.subdom.dom:sitename.dom:sitename.uucp +# +# NOTE: It is preferable, in most circumstances, to set the hostnames +# variable in /usr/lib/smail/config, rather than setting HOSTNAMES +# in this file. The value of the "hostnames" variable in the config +# file can be changed without recompiling smail. + +#HOSTNAMES=kgbvax.ussr.comm:kgbvax.uucp # example only + + +# DOMAINS - domains under which the local host resides +# +# optional (except when HOSTNAMES is not set) +# +# If HOSTNAMES is not defined, then HOSTNAMES is computed by concatenating +# local host's actual name with each name in this colon-separated list of +# domain names. This computation is done at run-time. +# +# Example: if DOMAINS="uts.amdahl.com:uucp", then the HOSTNAMES value +# for the site "amdahl" would become "amdahl.uts.amdahl.com:amdahl.uucp". +# +# NOTE: The uucp domain is not meaningful in any standard way, though +# many sites treat this as implying any site that is registered +# in the UUCP zone through the USENET maps. Sites that really +# wish to be able to receive mail from anywhere in the world +# should get a registered domain. +# +# NOTE: It is preferable, in most circumstances, to set the domains +# variable in /usr/lib/smail/config, rather than setting DOMAINS +# in this file. The value of the "domains" variable in the config +# file can be changed without recompiling smail. + +DOMAINS=uucp # common +#DOMAINS=ussr.comm:uucp # example only + + +# UUCP_NAME - name for the local host on the UUCP network +# +# optional +# +# This should be set to the name of the local host as known by the +# UUCP software. +# +# Normally, no value is given for this variable, in which case the +# value will be computed by smail at run-time from the actual name for +# local host, as computed in a system-dependent fashion. +# +# NOTE: It is preferable, in most circumstances, to set the uucp_name +# variable in /usr/lib/smail/config, rather than setting UUCP_NAME +# in this file. The value of the "uucp_name" variable in the config +# file can be changed without recompiling smail. + +#UUCP_NAME=kgbvax # example only + + +# VISIBLE_NAME - local host name used in outgoing addresses +# +# optional +# +# Some installations may wish to hide a group of machines under a common +# name. For example, at Amdahl, we may wish to hide the machines on our +# ethernet under the domain "uts.amdahl.com". If we set the value +# VISIBLE_NAME="uts.amdahl.com" on the machine "futatsu", then mail +# from "futatsu" sent as if it came from "user@uts.amdahl.com" rather than +# "user@futatsu.uts.amdahl.com". Of course, the "uts.amdahl.com" gateways +# will need to forward mail to "user", however the fact that a user +# does not have check for mail on all machines on the network is useful. +# +# The VISIBLE_NAME value will not be recognized as a name for the local +# host unless it is also included in one of the other lists of names for +# the local host. +# +# NOTE: It is preferable, in most circumstances, to set the visible_name +# variable in /usr/lib/smail/config, rather than setting VISIBLE_NAME +# in this file. The value of the "visible_name" variable in the config +# file can be changed without recompiling smail. + +#VISIBLE_NAME=kgbvax.uucp # example only + + +# GATEWAY_NAMES - more names for the local host +# +# optional +# +# It is often important that hosts which are domain gateways recognize +# the domain names as alternate names for the local host. These names +# should be defined in addition to those computed automatically from +# the actual name for the local host by putting them in GATEWAY_NAMES. +# +# This variable is a colon-separated list of names for the local host +# which is not computed from the actual name for the local host. +# +# NOTE: It is preferable, in most circumstances, to set the gateway_names +# variable in /usr/lib/smail/config, rather than setting GATEWAY_NAMES +# in this file. The value of the "gateway_names" variable in the config +# file can be changed without recompiling smail. + +#GATEWAY_NAMES=moscow.ussr.comm:ussr.comm:comm # example only + + +# AUTH_DOMAINS - domains for which this host is authoritative +# +# optional (defaults to none) +# +# If this host contains all routing information needed for a particular +# set of domains, then smail should generate an error if it is given a +# hostname that is not resolved, but is within one of those domains. +# Normally, this isn't a problem because mail that is not matched by +# any routers generates a "host not found" error. However, the +# smarthost router can upset this by matching any hostname. Without +# something to generate an error, a mail loop can occur where mail +# is sent to a host that controls a domain, and that host sends mail +# to another host listed as the "smarthost", which sends it back. +# +# A host which contains all routing information for a domain is called +# "authoritative" for that domain. A host can be authoritative for +# any number of domains. To declare this host to be authoritative +# for this domain, set AUTH_DOMAINS here (or auth_domains in the +# /usr/lib/smail/config file) to a colon-separated list of domains. +# This will prevent the smarthost router from matching any hosts within +# the indicated domains. +# +# NOTE: It is preferable, in most circumstances, to set the auth_domains +# variable in /usr/lib/smail/config, rather than setting AUTH_DOMAINS +# in this file. The value of the "auth_domains" variable in the config +# file can be changed without recompiling smail. + +#AUTH_DOMAINS=ussr.com +#AUTH_DOMAINS=veritas.com:tolerant.com + + +# POSTMASTER - the default address for the mail administrator +# +# optional (defaults to root) +# +# This address is used by smail as a last resort in trying to deliver +# to the address "postmaster". Normally, "postmaster" is aliased to a +# responsible person, or persons through the alias file. In the event +# that no alias file exists, or that "postmaster" was not found in +# that file or by any other directors, the default address will be +# used. + +#POSTMASTER=root + + +# NOBODY - a user with few access capabilities +# +# optional +# +# The user named here will be used by smail whenever a user ID is +# desired that cannot do any more damange than any unpriveledged user +# on the system. Under 4.3BSD and SunOS, this would be the user named +# "nobody". Under other operating systems, it may be reasonable to +# create a "nobody" entry in the passwd file. Some systems have a +# user such as "unknown" which will suffice. +# +# If this is not defined, then a default will be chosen. This default +# is os-type dependent, and is commonly nobody for BSD- and sun-derived +# systems. + +NOBODY=nobody +#NOBODY=unknown # some sites have this in their passwd file + + +# TEST_BASE - directory where smail test files are kept +# +# optional (special when defined to an empty string) +# +# When testing smail, it is convenient to put binaries and configuration +# files in an area separated from the actual installation areas. The +# TEST_BASE directory defines this alternate area. Smail will assume +# assume a fixed hierarchy below this directory, with a "bin" subdirectory +# containing the smail binary and utilities, a "lib" directory containing +# smail configuration files, and a "spool" directory containing smail +# spool files. +# +# If no TEST_BASE is defined, then this facility is turned off. If a +# TEST_BASE is defined to be an empty string, then all programs will be +# used in the area where they are compiled, LIB_DIR will be set to the +# "lib" directory under the root of the smail source tree, and SPOOL_DIRS +# will be set to the "spool" directory under the root of the smail source +# tree. In this case, a "make install" is not required. +# +# If a TEST_BASE is defined as a relative path (e.g. "."), then it is +# defined relative to the root of the smail source directory. + +#TEST_BASE=/usr/project/smail +#TEST_BASE=test +#TEST_BASE= # use progs in source area + + +# SMAIL_BIN_DIR - directory where copies of the smail binary are kept +# +# optional +# +# The Smail program comes in user callable names: smail, uupath, +# pathto, optto, and so on. A copy of smail will be linked to files +# under SMAIL_BIN_DIR. The SMAIL_BIN_DIR should be a directory +# that is commonly in users search path (i.e., $PATH). + +SMAIL_BIN_DIR=!!PREFIX!!/bin # BSD local convention + +# SMAIL_NAME - file where the primary working copy of smail is located +# +# optional +# +# Any program that needs to call smail, including smail itself will +# attempt to execute the program named by SMAIL_NAME. +# +# Often the primary working copy of smail is /usr/lib/sendmail. This +# should be used for systems that used to run sendmail. Programs such +# as Berkeley Mail, System V mailx or /bin/mail can be made to, or do +# call /usr/lib/sendmail for mailer activity. It is common for public +# domain programs to expect a mailer to exist under this name, also. +# +# If SMAIL_NAME is not defined here, or is set to a null string, then +# $SMAIL_BIN_DIR/smail is used instead. If this is the name that you +# want to use as the primary binary pathname, then do set SMAIL_NAME +# to the null string. This will prevent the smail src/Makefile from +# installing it as $SMAIL_BIN_DIR/smail twice. + +#SMAIL_NAME=/usr/lib/sendmail # common convention +#SMAIL_NAME= # use smail in bin directory + + +# OTHER_SMAIL_NAMES - other names under which to install smail +# +# optional +# +# Many systems will wish to install smail as /bin/rmail to catch mail +# coming in over UUCP directly with smail. To install under this +# name set OTHER_SMAIL_NAMES to /bin/rmail. As implied by the name, +# other pathnames can be specified as well, if a system has other +# potential rendezvous points for mail. This should be a colon or +# white-space separated list of full pathnames. +# +# Users of XENIX may wish to use smail as their execmail interface. +# To do this, add /usr/lib/mail/execmail to this list. +# +# This is assumed to be empty when TEST_DIR is defined. +# +# NOTE: A pathname CANNOT be in both SMAIL_NAME and OTHER_SMAIL_NAMES. + +OTHER_SMAIL_NAMES=!!PREFIX!!/bin/rmail + +# LIB_DIR - directory where various smail files are found +# +# required +# +# The LIB_DIR is where various static smail files reside, by default. +# Files which may reside under this directory are: the primary config +# file, the directors, routers and transports files, an aliases file, +# pathalias database, uuwho database, and the COPYING file. +# +# The common subdirectories under LIB_DIR are: "methods", where method +# files are by default found; "maps", where local pathalias files, +# mkmap configuration files and the getmap batch file are located; +# and "lists" where mailing lists are commonly located. +# +# It should be noted that none of these files, except for COPYING, +# is required by the smail binary as it is released. See smail(5) +# and smail(8) for more details on this and related topics. + +LIB_DIR=!!PREFIX!!/lib/smail + + +# UTIL_BIN_DIR - directory where smail utilities are located +# +# required +# +# The smail system has a number of programs such as pathalias, mkline, +# mksort, mkdbm that users normally need not execute directly. Such +# utilities will be placed under the UTIL_BIN_DIR directory. +# +# UTIL_BIN_DIR is often the same as LIB_DIR. + +UTIL_BIN_DIR=$LIB_DIR/util + + +# NEWALIASES - alternate pathname for mkaliases program +# +# optional +# +# For compatibility with sendmail, mkaliases can be installed under +# a name such as /usr/ucb/newaliases, or /usr/lib/newaliases. To setup +# smail to perform this installation, set NEWALIASES to the desired full +# pathname. When testing, this is set to the empty string, signifying +# that mkaliases will be installed only under the name mkaliases. + +NEWALIASES=$UTIL_BIN_DIR/newaliases + +# SECOND_CONFIG_FILE - pathname of an alternate config file +# +# optional +# +# Smail allows one to have a second set of configuration files that can +# override both the compiled in and the optional primary config file. +# Normally, the path to this secondary configuration file is defined +# in the primary smail configuration file. However the value +# SECOND_CONFIG_FILE can be the filename of the secondary (alternate) +# config file. In the case where both this value, and where the +# LIB_DIR/config file specifies a "second_config_file" value, the +# LIB_DIR/config file takes presidence. +# +# A secondary configuration file is useful in an environment with +# distributed filesystems. In such an environment, the regular config +# directory can be maintained centrally on a remotely mounted +# filesystem. Individual machines, can then setup a secondary +# configuration file which overrides these centrally maintained files +# for specific purposes. + +# For example, in a network with one gateway to the outside world, +# all could share the same smail binaries and standard configuration. +# However the the gateway machine could have a secondary configuration +# file which so that the gateway's the smail setup has greater intelligence +# and thus act as a smart_user and smart_path host. + +SECOND_CONFIG_FILE= # common for non-NFS systems +#SECOND_CONFIG_FILE=/private/usr/lib/smail # convention for NFS systems + + +# PATHS_FILE, PATHS_TYPE - location and type of the paths database +# +# optional +# +# If PATH_FILE begins with a '/', then it is the name of the optional +# pathalias database, otherwise it is assumed that PATH_FILE refers to +# a file under the LIB_DIR directory. +# +# For use with YP, it may be more convenient to use the "yp" type, +# rather than "aliasyp", because this way a paths file can be maintained +# in a format compatible with other YP maps. +# +# If PATHS_FILE is left undefined, no paths file is defined in the +# preloaded smail configuration. + +PATHS_FILE=paths # for file under LIB_DIR +#PATHS_FILE=mail.paths # YP (NIS) + +#PATHS_TYPE=dbm # indexed +#PATHS_TYPE=bsearch # sorted +PATHS_TYPE=lsearch # linear search +#PATHS_TYPE=yp # YP (NIS) + +# ALIASES_FILE, ALIASES_TYPE - location and type of the aliases database +# +# optional +# +# If ALIASES_FILE begins with a '/', then it is the full pathname of the +# optional aliases database, otherwise it is assumed that ALIASES_FILE +# refers to a file under the LIB_DIR directory. +# +# For use with YP, it may be more convenient to use the "yp" type, +# rather than "aliasyp", because this way an alias file can be maintained +# in a format compatible with other YP maps. +# +# NOTE: If ALIASES_TYPE is specified as bsearch, then the suffix .sort +# will be appended to the filename given in ALIASES_FILE. The +# mkaliases command will then treat a file without that suffix +# as the unsorted source file for the database. +# +# If ALIASES_FILE is left undefined, no aliases file is defined in the +# preloaded smail configuration. +# +# NOTE FOR AIX SYSTEMS: +# I have received reports that AIX3.1 requires use of "yp", +# rather than "aliasyp". If this is true, I don't quite see +# how Suns and RS/6000s running sendmail can share the same +# YP maps. In any case, use whatever works in your +# environment. + +ALIASES_FILE=$LIB_DIR/aliases # compatible with sendmail +#ALIASES_FILE=aliases # for file under LIB_DIR +#ALIASES_FILE=mail.aliases # YP (NIS) +#ALIASES_FILE=NetInfo # for NeXT NetInfo aliases + +ALIASES_TYPE=dbm # compatible with sendmail +#ALIASES_TYPE=lsearch # for sites without dbm(3x) +#ALIASES_TYPE=bsearch # for sorted files, faster +#ALIASES_TYPE=aliasyp # YP (NIS) +#ALIASES_TYPE=nialias # for NeXT NetInfo aliases + + +# ALIASES_REMOVE_SENDER - remove the sender (by default) from alias expansions +# LISTS_REMOVE_SENDER - remove the sender (by default) from list expansions +# +# optional +# +# For compatibility with sendmail, smail can be configured to remove the +# sender of a message (if the sender is a local user) from alias +# expansions. Users can disable this feature by invoking smail with +# the -m flag, typically by setting "metoo" in their .mailrc file. +# +# However, many people find this behavior confusing and some administrators +# would prefer to disable this "sender removal" feature altogether. +# Also, it has recently been noticed that smail's sender removal +# behavior is not quite sendmail-compatible with regards to nested +# aliases and have been surprised by the results of aliasing. +# +# Smail supports two forms of aliasing to which sender removal can be +# applied: the regular aliases file (defined by ALIASES_FILE above), and +# a directory of mailing lists (typically /usr/lib/smail/lists) where +# all files in that directory define a list. A site that has converted +# entirely over to smail3 can use alias files strictly for username +# aliasing, while using the lists directory for real mailing lists. With +# such a usage model, sender removal should be disabled for the alias +# file, but can be enabled (depending upon preference) for the lists +# directory. +# +# To enable sender removal, set ALIASES_REMOVE_SENDER or LISTS_REMOVE_SENDER +# (depending upon preference) to "yes". If these values are empty, or are +# note defined, then sender removal is disabled. +# +# This behavior can be configured using the run-time directors configuration +# file by setting the "sender_okay" flag, to indicate that sender removal +# should NOT be done. + +#ALIASES_REMOVE_SENDER=yes +LISTS_REMOVE_SENDER=yes + + +# RETRY_INTERVAL - minimum per-host retry interval for TCP/IP delivery +# RETRY_DURATION - maximum time to keep messages in the input queue +# +# optional (RETRY_INTERVAL defaults to 10 minutes, RETRY_DURATION to 5 days) +# +# Messages to hosts on TCP/IP networks may block for quite some time, if +# the recipient host is down, or some intervening network is inoperative. +# Connections to such unreachable or downed hosts can block for quite +# some time until a timeout indicates indicates that the condition +# exists. +# +# To prevent the mail queue from blocking repeatedly on the same set of +# unreachable hosts, smail keeps a directory of files that store the +# last connection failure encountered for each host. Message deliveries +# to a host that has failed recently will be deferred until a certain +# period of time (the retry interval) has elapsed, preventing queue runs +# from hanging again immediately. +# +# If a host is down for an extended period of time (such as forever), +# then sending a message to the message sender is preferable to keeping +# the message in our mail queue. If message deliveries to a particular +# host have not succeeded within a particular length of time (the retry +# duration), smail considers delivery to have failed, and sends a bounce +# notification to the sender or address owner. +# +# Retry durations and intervals for specific hosts or domains can be +# defined by creating a retry file in $LIB_DIR and adding lines of +# the form: +# +# hostname duration/interval +# *.domain duration/interval +# +# A default can be added at the end of this file with: +# +# * duration/interval +# +# Alternately, a default can be specified by setting RETRY_INTERVAL +# and RETRY_DURATION in this file. Both values are in seconds. To +# help in calculating, 1 day is 86400 seconds. You can actually +# use C constant expressions here, if you want to use addition or +# multiplication. +# +# These values can also be set using the retry_interval and +# retry_duration variables in $LIB_DIR/config. In the config file +# and in the retry file, times can be specified with suffixes, such +# as 'm', 'h', or 'd' to indicate minutes, hours, or days. For +# example: +# +# *.uu.net 3d/10m + +RETRY_INTERVAL=600 # 10 minutes +#RETRY_INTERVAL='20*60' # 20 minutes +RETRY_DURATION='86400*5' # 5 days +#RETRY_DURATION='86400*14' # 2 weeks + + +# UUCP_SYSTEM_FILE - path to UUCP file containing remote systems +# +# optional +# +# The normal smail configuration defines a router that scans the +# output of the /usr/bin/uuname command for neighboring UUCP sites. +# When smail is running as a daemon, it can cache the output of uuname +# so that the uuname command need not be executed for each mail +# message. In order to be able to detect when the output of uuname +# will change, smail daemons will stat the UUCP configuration file +# which contains the names of neighboring hosts. This EDITME variable +# defines the full pathname to this file. If this variable is not +# defined, then the output of the uuname will not be cached. +# +# NOTE: Many OS files specify the correct file for this, though some +# don't. Also, if you add HoneyDanBer UUCP to a system that +# does not normally have it, then you will need to set this. + +#UUCP_SYSTEM_FILE=/usr/lib/uucp/L.sys # use this for normal UUCP +#UUCP_SYSTEM_FILE=/usr/lib/uucp/Systems # use this for HoneyDanBer UUCP +#UUCP_SYSTEM_FILE=/etc/uucp/Systems # for System V Release 4 + + +# SPOOL_DIRS - smail spooling directories +# +# required +# +# Smail can use one or more spooling directories, where spool directories +# other than the first are used if earlier spool directories were +# inaccessible or were on file systems which filled up. The list of spool +# directories should be colon-separated and may contain only one spool +# directory if desired. + +SPOOL_DIRS=/var/spool/smail # common +#SPOOL_DIRS=/var/spool/smail # for System V Release 4 +#SPOOL_DIRS=/usr/spool/smail:/usr2/spool/smail # use of alternate filesystem +#SPOOL_DIRS=/usr/smail/spool # I use this right now +#SPOOL_DIRS=/var/smail/spool # Another alternative + + +# LOG_DIR - smail logging directory +# +# optional (defaults to (more or less) $SPOOL_DIRS/log) +# +# Smail creates two log files in this directory: logfile and paniclog. +# The first is a file that logs all incoming messages and deliveries, +# plus many errors. The second file (paniclog) logs important system +# errors that smail can manage to write into the log file. + +#LOG_DIR=/usr/spool/smail/log # common +LOG_DIR=/var/spool/smail/log # for System V Release 4 +#LOG_DIR=/usr/smail/log # I use this right now +#LOG_DIR=/var/smail/log # Another alternative + + +# UNSHAR_MAP_DIR - where unshared USENET map files are to be placed +# +# optional +# +# The getmap utility will read a list of files on the file +# LIB_DIR/map/batch, and using unsharmap unshar these maps into the +# UNSHAR_MAP_DIR directory. +# +# A common way that maps are distributed is through the USENET news group +# "comp.mail.maps". A sys file line of: +# +# usenet-maps:comp.mail.maps,world:F:<UNSHAR_MAP_DIR>/work/batch +# +# will add names into the UNSHAR_MAP_DIR/work/batch file for getmap to +# process. +# +# If UNSHAR_MAP_DIR is not defined, then the getmap and uuwho utilities +# will not function. + +UNSHAR_MAP_DIR=/var/spool/uumaps +#UNSHAR_MAP_DIR=/var/spool/uumaps # for System V Release 4 + + +# NEWS_SPOOL_DIR - where news articles are stored +# +# optional (defaults to /usr/spool/news) +# +# C News stores pathnames in the batch file that are relative to +# the top-level of the news article directory hierarchy. Smail needs +# to know the name of this directory in order to be able to extract +# the news articles from comp.mail.maps. + +#NEWS_SPOOL_DIR=/usr/spool/news +NEWS_SPOOL_DIR=/var/spool/news + + +# UUWHO_FILE - file used to store the uuwho file +# +# optional (defaults to $LIB_DIR/uuwho) +# +# This specified the file used to store the host database generated by +# and used by the uuwho command. If the uuwho database is stored in a +# DBM file, then UUWHO_FILE is the file prefix before adding the .pag +# and .dir suffixes to form the actual files. +# +# If UUWHO_FILE does not begin with /, then the name is relative to +# the $LIB_DIR directory. + +#UUWHO_FILE=$UNSHAR_MAP_DIR/uuwho # to store database with the map files + + +# UUWHO_USE_DBM - configure the uuwho command to use the DBM library +# +# optional +# +# If you would like the uuwho utility to use the DBM library, set +# UUWHO_USE_DBM to true. If this is not set, then the uuwho utility +# will use a sorted database. Earlier versions of the uuwho utility +# worked only with the DBM library. +# +# Not using DBM does not result in any real measurable speed-up, since +# binary searches of even large databases happen fast enough to +# prevent users from noticing. However, generation of the DBM uuwho +# database can take significantly longer. The uuwho utility creates +# sorted databases by calling the sort utility, which is usually quite +# fast. + +UUWHO_USE_DBM=true # use DBM +#UUWHO_USE_DBM= # default, don't use DBM + + +# TMP_DIR - secure temp directory used by smail utilities +# +# optional (defaults to /tmp) +# +# Some smail utilities use TMP_DIR to define a directory in which +# their temporary files can be created. Utilities that use this +# directory are those that wish to use a secure temporary area. +# +# The only smail utilities that need to use TMP_DIR are utilities that +# should be run as root, or as some other user with appropriate privledges. +# It is therefore recommended that TMP_DIR not be globally writable. + +TMP_DIR=/var/tmp + + +# SECURE_PATH - path for smail utilities +# +# optional (defaults to system-specific path or /bin:/usr/bin) +# +# The smail utilities often set their PATH to begin with the SECURE_PATH. +# The SECURE_PATH should be a path of directories where standard commands +# are located. These directories and commands should not be writeable by +# normal users. A common directory for smail utilities to append to +# SECURE_PATH is UTIL_BIN_DIR. + +#SECURE_PATH=/bin:/usr/bin:/usr/ucb # common for BSD systems +#SECURE_PATH=/bin:/usr/bin:/usr/amdahl/bin # common for UTS/580 systems +#SECURE_PATH=/bin:/usr/bin # common for System V system + + +# MAN1 - where smail user command man pages are to be installed +# MAN5 - where smail file format man pages are to be installed +# MAN8 - where smail administrator man pages should be installed +# MAN1_EXT - file extension for user command man pages +# MAN5_EXT - file extension for file format man pages +# MAN8_EXT - file extension for adminstrator man pages +# +# optional +# +# The man page for "foo.an" in the smail man/man5 source directory +# will be installed in MAN5/foo.MAN5_EXT, if MAN5 is defined. +# +# For some reason, BSD and System V differ on where file format man +# pages are kept, hence the unusual values of MAN5 for System V. + +MAN1=!!PREFIX!!/man/man1 +MAN1_EXT=1 + +MAN5=!!PREFIX!!/man/man5 +MAN5_EXT=5 + +MAN8=!!PREFIX!!/man/man8 +MAN8_EXT=8 + +# COMPRESS - file compression +# +# optional (default is system dependent or uses cat) +# +# The savelog utility can save space by compressing old log files. +# Normally when a file is compressed, the suffix is added to the +# filename. +# +# The COMPRESS symbol is the name of a program such that: +# +# $COMPRESS $COMP_FLAG filename ... +# +# the file: filename is replaced by: filename$DOT_Z +# regardless of the size of filename. The command: +# +# $UNCOMPRESS filename$DOT_Z +# +# will replace filename$DOT_Z with the original filename. +# The command: +# $ZCAT filename$DOT_Z +# +# will read the compressed file: filename$DOT_Z and write the +# plain text to standard output while leaving the file compressed. + +COMPRESS=gzip +COMP_FLAG="-f" +UNCOMPRESS=gunzip +ZCAT=gzcat +DOT_Z=".gz" + +#COMPRESS=pack # System V standard +#COMP_FLAG="-f" +#UNCOMPRESS=unpack +#ZCAT=pcat +#DOT_Z=".z" + +#COMPRESS=true # if no compress is used +#COMP_FLAG="" +#UNCOMPRESS=true +#ZCAT=cat +#DOT_Z="" + + +# MISC_C_DEFINES - miscellaneous #defines for C programs +# +# optional +# +# The value of MISC_C_DEFINES is included directly into the file defs.h +# in each source directory. It is useful as a central place for +# miscellaneous #defines not otherwise setable above. In particular, +# it is useful for overriding other more values given in the file +# src/config.h. It was felt that it was appropriate to keep the +# EDITME file fairly small, with only a reasonable subset of the +# configurable aspects of smail being explicitly described. As a +# result, there may be some other things in config.h that you may +# wish to change. Use #define's here to accomplish this. +# +# A suitable collection of #undef's and #define's here can also +# change the behavior defined in the file under conf/os or conf/arch +# for your operating system and architecture. However, use of +# MISC_C_DEFINES for this is discouraged. +# +# As with /bin/sh scripts, a logical line beginning with a : causes +# that line to be parsed but ignored. If it were simply commented out +# then the second line would be processed. + +# MISC_SH_DEFINES - miscellaneous script to include in defs.sh +# +# optional +# +# The value of MISC_SH_DEFINES is included directly into the file +# defs.sh, in each source directory. It is a useful place to store +# script lines to override values set by the mkdefs.sh shell script +# or in a conf/os or conf/arch file. +# +# As with /bin/sh scripts, a logical line beginning with a : causes +# that line to be parsed but ignored. If it were simply commented out +# then the second line would be processed. + + + +# MISC_SED_DEFINES - miscellaneous lines to include in defs.sed +# +# optional +# +# The value of MISC_SED_DEFINES is included directly into the file +# defs.sed, in each source directory. It is a useful place to store +# sed lines to override lines set by the mkdefs.sh shell script from +# information in a conf/os or conf/arch file. +# +# As with /bin/sh scripts, a logical line beginning with a : causes +# that line to be parsed but ignored. If it were simply commented out +# then the second line would be processed. + +# We can't readily think of good examples for this one. |