-$Cambridge: exim/doc/doc-txt/NewStuff,v 1.151 2007/06/20 14:13:39 ph10 Exp $
+$Cambridge: exim/doc/doc-txt/NewStuff,v 1.163 2010/06/01 11:13:54 pdp Exp $
New Features in Exim
--------------------
test from the snapshots or the CVS before the documentation is updated. Once
the documentation is updated, this file is reduced to a short list.
+
+Version 4.72
+------------
+
+ 1. TWO SECURITY FIXES: one relating to mail-spools which are globally
+ writable, the other to locking of MBX folders (not mbox).
+
+ 2. MySQL stored procedures are now supported.
+
+ 3. The dkim_domain transport option is now a list, not a single string, and
+ messages will be signed for each element in the list (discarding
+ duplicates).
+
+ 4. The 4.70 release unexpectedly changed the behaviour of dnsdb TXT lookups
+ in the presence of multiple character strings within the RR. Prior to 4.70,
+ only the first string would be returned. The dnsdb lookup now, by default,
+ preserves the pre-4.70 semantics, but also now takes an extended output
+ separator specification. The first output separator is used to join
+ multiple TXT records together; use a second separator character, followed
+ by a colon, to join the strings within a TXT record on that second
+ character, or use a semicolon to concatenate strings within a TXT record
+ with no separator. Administrators are reminded that DNS provides no
+ ordering guarantees between multiple records in an RRset. For example:
+
+ foo.example. IN TXT "a" "b" "c"
+ foo.example. IN TXT "d" "e" "f"
+
+ ${lookup dnsdb{>/ txt=foo.example}} -> "a/d"
+ ${lookup dnsdb{>/; txt=foo.example}} -> "def/abc"
+ ${lookup dnsdb{>/+: txt=foo.example}} -> "a+b+c/d+e+f"
+
+ Some character combinations are currently unsupported.
+
+Version 4.70 / 4.71
+-------------------
+
+ 1. Native DKIM support without an external library.
+
+ 2. Experimental DCC support via dccifd (contributed by Wolfgang Breyha).
+
+ 3. There is now a bool{} expansion condition which maps certain strings to
+ true/false condition values (most likely of use in conjuction with the
+ and{} expansion operator).
+
+ 4. The $spam_score, $spam_bar and $spam_report variables are now available
+ at delivery time.
+
+ 5. exim -bP now supports "macros", "macro_list" or "macro MACRO_NAME" as
+ options, provided that Exim is invoked by an admin_user.
+
+ 6. There is a new option gnutls_compat_mode, when linked against GnuTLS,
+ which increases compatibility with older clients at the cost of decreased
+ security. Don't set this unless you need to support such clients.
+
+ 7. There is a new expansion operator, ${randint:...} which will produce a
+ "random" number less than the supplied integer. This randomness is
+ not guaranteed to be cryptographically strong, but depending upon how
+ Exim was built may be better than the most naive schemes.
+
+ 8. Exim now explicitly ensures that SHA256 is available when linked against
+ OpenSSL.
+
+ 9. The transport_filter_timeout option now applies to SMTP transports too.
+
+
Version 4.68
------------
accept
+ 6. The variable $max_received_linelength contains the number of bytes in the
+ longest line that was received as part of the message, not counting the
+ line termination character(s).
+
+ 7. Host lists can now include +ignore_defer and +include_defer, analagous to
+ +ignore_unknown and +include_unknown. These options should be used with
+ care, probably only in non-critical host lists such as whitelists.
+
+ 8. There's a new option called queue_only_load_latch, which defaults true.
+ If set false when queue_only_load is greater than zero, Exim re-evaluates
+ the load for each incoming message in an SMTP session. Otherwise, once one
+ message is queued, the remainder are also.
+
+ 9. There is a new ACL, specified by acl_smtp_notquit, which is run in most
+ cases when an SMTP session ends without sending QUIT. However, when Exim
+ itself is is bad trouble, such as being unable to write to its log files,
+ this ACL is not run, because it might try to do things (such as write to
+ log files) that make the situation even worse.
+
+ Like the QUIT ACL, this new ACL is provided to make it possible to gather
+ statistics. Whatever it returns (accept or deny) is immaterial. The "delay"
+ modifier is forbidden in this ACL.
+
+ When the NOTQUIT ACL is running, the variable $smtp_notquit_reason is set
+ to a string that indicates the reason for the termination of the SMTP
+ connection. The possible values are:
+
+ acl-drop Another ACL issued a "drop" command
+ bad-commands Too many unknown or non-mail commands
+ command-timeout Timeout while reading SMTP commands
+ connection-lost The SMTP connection has been lost
+ data-timeout Timeout while reading message data
+ local-scan-error The local_scan() function crashed
+ local-scan-timeout The local_scan() function timed out
+ signal-exit SIGTERM or SIGINT
+ synchronization-error SMTP synchronization error
+ tls-failed TLS failed to start
+
+ In most cases when an SMTP connection is closed without having received
+ QUIT, Exim sends an SMTP response message before actually closing the
+ connection. With the exception of acl-drop, the default message can be
+ overridden by the "message" modifier in the NOTQUIT ACL. In the case of a
+ "drop" verb in another ACL, it is the message from the other ACL that is
+ used.
+
+10. For MySQL and PostgreSQL lookups, it is now possible to specify a list of
+ servers with individual queries. This is done by starting the query with
+ "servers=x:y:z;", where each item in the list may take one of two forms:
+
+ (1) If it is just a host name, the appropriate global option (mysql_servers
+ or pgsql_servers) is searched for a host of the same name, and the
+ remaining parameters (database, user, password) are taken from there.
+
+ (2) If it contains any slashes, it is taken as a complete parameter set.
+
+ The list of servers is used in exactly the same was as the global list.
+ Once a connection to a server has happened and a query has been
+ successfully executed, processing of the lookup ceases.
+
+ This feature is intended for use in master/slave situations where updates
+ are occurring, and one wants to update a master rather than a slave. If the
+ masters are in the list for reading, you might have:
+
+ mysql_servers = slave1/db/name/pw:slave2/db/name/pw:master/db/name/pw
+
+ In an updating lookup, you could then write
+
+ ${lookup mysql{servers=master; UPDATE ...}
+
+ If, on the other hand, the master is not to be used for reading lookups:
+
+ pgsql_servers = slave1/db/name/pw:slave2/db/name/pw
+
+ you can still update the master by
+
+ ${lookup pgsql{servers=master/db/name/pw; UPDATE ...}
+
+11. The message_body_newlines option (default FALSE, for backwards
+ compatibility) can be used to control whether newlines are present in
+ $message_body and $message_body_end. If it is FALSE, they are replaced by
+ spaces.
+
Version 4.67
------------
git://git.exim.org / users / jgh / exim.git / blobdiff