This is a quick utility to check if there are new mails
on any un*x system running qmail with Maildir. It won't work
for someone running mailboxes in mbox format and I can't be
bothered to add that. Do it yourself if you need to support
this old and crappy format.
The responsible for that tool is my friend Stampel who told
me that my system sucked because it didn't display the number
of unread mails when logging in. Oh well, this just took 5
minutes to add.
fetchmail-5.7.2 (Sun Mar 4 19:05:57 EST 2001), 20167 lines:
* NLS build fixed under Linux -- thanks to Nicol<E1>s Lichtmaier.
Now builds with NLS by default, unavoidable side effect of fix.
* Various cleanups and minor bugfixes in OTP support.
* Corrected header error in SSL support.
* Updated Spanish translation by Javier Kohen.
* Incorporated Red Hat's 5.5.0 Kerberos V patch.
* Incorporated Henrique Holschuh's fetchmailconf warning on local names with @.
fetchmail-5.7.1 (Sun Mar 4 00:30:04 EST 2001), 20167 lines:
* Note: NLS build ability is broken!
* FreeBSD port fixes.
* Man page polished and updated.
* Added --sslproto option from Peter Hedwig.
* Fix KPOP support, which got broken in 5.6.8. Closes Debian bug #88288.
getmail is intended as a simple replacement for fetchmail. It retrieves mail
(either all messages, or only unread messages) from one or more POP3 servers
for one or more email accounts, and reliably delivers into a qmail-style
Maildir (or mbox file) specified on a per-account basis. It also has support
for domain (multidrop) mailboxes.
Package submitted by Tomasz Luchowski in PR pkg/11929.
like memory buffer overrun, but the udp send buffer overflow)
1.12: Mar.4,2001:
log.c:
a hack to avoid syslog buffer overflow added.
problem reported by Koji Kawano <kawano@eng.trans-cosmos.co.jp>
allows correct local delivery on multiple hosts while still canonicalizing
visible and sender addresses. (This makes it possible to support vanity
domains and "permanent address" services like iki.fi and pobox.com.)
Changes since 5.6.0:
fetchmail-5.7.0 (Thu Mar 1 18:00:08 EST 2001), 20131 lines:
* Updated Danish translation from Byrial Jensen.
* Fixed bug in NTLM support. Separately, "auth ntlm" now works.
* Imail server and vircom NTLM account added to test list.
* Don't require Tkinter to read --configdump output.
fetchmail-5.6.8 (Thu Feb 22 02:57:31 EST 2001), 20110 lines:
* `preauth' option changed back to `auth'
* IMAP code now bails out if the server forces the mailbox read-only.
* Fixed a core dump in Dave Zarzycki's new plugin code.
* POP3 latency optimization: only do CAPA and set authentication capabilities
from it once at start of run.
fetchmail-5.6.7 (Mon Feb 19 12:31:03 EST 2001), 20082 lines:
* Fixed brown-paper-bag password bug (only showed up if it was necessary
to prompt for a password). This fixes Debian bug #86350: Fetchmail doesn't
ask for password.
* In POP3, query for AUTH methods a la RFC2449.
fetchmail-5.6.6 (Thu Feb 15 20:43:47 EST 2001), 20083 lines:
* Fixed locale setting; this should make i18n actually work.
* Resolved Debian bug #85938: fetchmail asks for a password when using ETRN.
* Resolved Debian bugs #85853 and #86047. POP3: Don't issue AUTH between
USER and PASS, some servers choke on this.
* Resolved Debian bug #85772 re Kerberos compilation.
* Resolved Debian bug #85961: Wrong error message when local connection fails.
* Serious pre-release regression testing begins. This version tested
against 18 different POP3 and IMAP servers.
fetchmail-5.6.5 (Mon Feb 12 04:33:39 EST 2001), 20062 lines:
* CRAM-MD5 authentication of IMAP and POP3 is working. Tested against
IMAP4rev1 2000.287 and v2000.70 POP3 gateway at neo.netnea.com.
* Full support for POP3 AUTH (RFC1734) with KERBEROS_IV, GSSAPI, OTP.
This code has been completely refactored. In the process, it is
possible I have broken GSSAPI, KERBEROS, and OPIE; this needs to be tested.
The old IMAP-LOGIN, IMAP-GSS, and IMAP-K4 protocols are gone; fetchmail
now uses these automatically when it detects the right capabilities.
To prevent having fetchmail look for a password, specify a "preauth"
option other than "password".
* Noted that Debian bugs #78963, #63064, #81312, #78796, #78363, #78149,
#68627, #67559, #63308, #63088, #71428 are fixed.
* Resolved Debian bug #65505: fetchmail now returns a nonzero exit status
when interrupted before a successful fetch.
* configure --ssl works correctly again.
fetchmail-5.6.4 (Sun Feb 11 00:43:14 EST 2001), 20085 lines:
* ODMR port fix for AIX.
* Dave Zarzycki's fix for former FAQ item F5 (%h and %p not being expanded).
* Dave Zarzycki's fix to reap zombie processes when nodetach is set.
* Attempted fix for CRAM-MD5 problem with IMAP 2000.
fetchmail-5.6.3 (Wed Feb 7 10:56:21 EST 2001), 19901 lines:
* VPATH build fixed (thanks to Harry McGavran).
* Danny O'Brien's patch allowing preauth and idle to work together.
* Fixed a bug in configure.in that was resulting in KERBEROS_IV being
set when it should not have been (several reports).
* FAQ change: mailing lists have moved to MailMan.
* Deal with brain-dead netmind mail missing the RFC822 delimiter line.
* ODMR (RFC 2645) support -- untested and probably buggy!
fetchmail-5.6.2 (Fri Jan 5 16:45:47 EST 2001), 19744 lines:
* Dave Bodenstab's fix for the lockfile re-exec problem.
* Fixes for `principal' handling in fetchmailconf.
* Make --with-included-gettext work again (Thanks to Albert Chin-A-Young).
fetchmail-5.6.1 (Mon Dec 11 23:11:59 EST 2000), 19718 lines:
* More on ETRN in the FAQ.
* Horst Klokow's patch to make interface check the remote IP address.
* Roger Luethi's patch to write the UIDL file when you hit a fetchlimit.
* Don Beusee's patch to eliminate wedging on authentication failure.
Instead, fetchmail will now notify the user on the third failure, then
continue polling silently until service is restored (at which time the
user will get a notification).
* Samuel Leo's patch to add LMTP capability to the smtphost option.
* Fix UIDL handling on skip entries.
* Add Don Beusee's `spambounce' option (default off).
(mail/postfix should be overwritten)
20010225
Portability: master sigchld handler based on writing to a
pipe, so that the master wakes up from select(). Based on
code by Erik Forsberg, Linkoping University, Sweden. File:
master/master_sig.c. Disabled until after the major release.
Code cleanup: Postfix should now run with no alias database.
Code cleanup: local_destination_recipient_limit and
local_destination_concurrency_limit have become first-class
configuration parameters. Files: global/mail_params.h,
*qmgr/qmgr.c, postconf/postconf.c.
20010226
Documentation suggestions by Lars Hecking and Richard
Huxton, Matthias Andree and many others.
Code cleanup: some queue/transport operations need to be
moved, after the code cleanup of the recipient/concurrency
limit handling. Patrik Rak. Files: *qmgr/qmgr_message.c.
8.11.3/8.11.3 2001/02/27
Prevent a segmentation fault when a bogus value was used in the
LDAPDefaultSpec option's -r, -s, or -M flags and if a bogus
option was used. Problem noted by Allan E Johannesen of
Worcester Polytechnic Institute.
Prevent "token too long" message by shortening {currHeader} which
could be too long if the last copied character was a quote.
Problem detected by Jan Krueger of digitalanswers
communications consulting gmbh.
Additional IPv6 check for unspecified addresses. Patch from
Jun-ichiro itojun Hagino of the KAME Project.
Do not ignore the ClientPortOptions setting if DaemonPortOptions
Modifier=b (bind to same interface) is set and the
connection came in from the command line.
Do not bind to the loopback address if DaemonPortOptions
Modifier=b (bind to same interface) is set. Patch from
John Beck of Sun Microsystems.
Properly deal with open failures on non-optional maps used in
check_* rulesets by returning a temporary failure.
Buffered file I/O files were not being properly fsync'ed to disk
when they were committed.
Properly encode '=' for the AUTH= parameter of the MAIL command.
Problem noted by Hadmut Danisch.
Under certain circumstances the macro {server_name} could be set
to the wrong hostname (of a previous connection), which may
cause some rulesets to return wrong results. This would
usually cause mail to be queued up and delivered later on.
Ignore F=z (LMTP) mailer flag if $u is given in the mailer A=
equate. Problem noted by Motonori Nakamura of Kyoto
University.
Work around broken accept() implementations which only partially
fill in the peer address if the socket is closed before
accept() completes.
Return an SMTP "421" temporary failure if the data file can't be
opened where the "354" reply would normally be given.
Prevent a CPU loop in trying to expand a macro which doesn't exist
in a queue run. Problem noted by Gordon Lack of Glaxo
Wellcome.
If delivering via a program and that program exits with EX_TEMPFAIL,
note that fact for the mailq display instead of just showing
"Deferred". Problem noted by Motonori Nakamura of Kyoto
University.
If doing canonification via /etc/hosts, try both the fully
qualified hostname as well as the first portion of the
hostname. Problem noted by David Bremner of the
University of New Brunswick.
Portability:
Fix a compilation problem for mail.local and rmail if SFIO
is in use. Problem noted by Auteria Wally
Winzer Jr. of Champion Nutrition.
IPv6 changes for platforms using KAME. Patch from
Jun-ichiro itojun Hagino of the KAME Project.
OpenBSD 2.7 and higher has srandomdev(3). OpenBSD 2.8 and
higher has BSDI-style login classes. Patch from
Todd C. Miller of Courtesan Consulting.
Unixware 7.1.1 doesn't allow h_errno to be set directly if
sendmail is being compiled with -kthread. Problem
noted by Orion Poplawski of CQG, Inc.
CONTRIB: buildvirtuser: Substitute current domain for $DOMAIN and
current left hand side for $LHS in virtuser files.
DEVTOOLS: Do not pass make targets to recursive Build invocations.
Problem noted by Jeff Bronson of J.D. Bronson, Inc.
MAIL.LOCAL: In LMTP mode, do not return errors regarding problems
storing the temporary message file until after the remote
side has sent the final DATA termination dot. Problem
noted by Allan E Johannesen of Worcester Polytechnic
Institute.
MAIL.LOCAL: If LMTP mode is set, give a temporary error if users
are also specified on the command line. Patch from
Motonori Nakamura of Kyoto University.
PRALIASES: Skip over AliasFile specifications which aren't based on
database files (i.e., only show dbm, hash, and btree).
Renamed Files:
devtools/OS/OSF1.V5.0 => devtools/OS/OSF1.V5.x
--- changes (reverse order)
20010204
Laid the ground work for logging of table accesses. This
will give more insight into how Postfix uses its lookup
tables. User interface comes later. File: util/dict_debug.c.
20010216
Bugfix: the pipe delivery agent expanded $size as if it
were a recipient, instead of expanding it as $nexthop or
as $sender. Reported by Michael Tokarev. File: pipe/pipe.c.
20010221
Bugfix: poor LMTP performance for domains that are listed
in $mydestination, because Postfix would send one recipient
at a time, with multiple deliveries of recipients of the
same message in parallel; a similar problem could exist
with virus scanning and with firewall relay hosts that
forward mail for $mydestination to an inside machine. This
behavior is now changed to depend on the transport-specific
xxx_destination_recipient_limit parameter. This also means
that you can now get qmail behavior for SMTP deliveries by
setting smtp_destination_recipient_limit=1. File:
{qmgr,nqmgr}/qmgr_message.c.
Workaround: Solaris socketpair() can fail with EINTR. Added
a sane_socketpair.c module that joins the ranks of the other
sane_whatever workarounds. Reported by Andrew McNamara.
File: util/sane_socketpair.[hc]
20010222
Documentation: the default main.cf file has a prominent
warning that mynetworks should be properly configured in
order to reject unauthorized mail relay requests from
strangers.
Documentation: the INSTALL document, section "mandatory
configuration file edits" has a section that explains that
mynetworks should be properly configured in order to reject
unauthorized mail relay requests from strangers.
20010223
Documentation: the basic.html document has a section that
explains that mynetworks should be properly configured in
order to reject unauthorized mail relay requests from
strangers.
Feature: new "mynetworks_style" parameter that controls
how mynetworks (trusted networks) is derived from the
inet_interfaces (machine interfaces) setting. Specify
"class" for entire class A, B, C networks; "subnet" for
the local subnets only; or "host" for maximal privacy.
Files: util/inet_addr_local.[hc], global/own_inet_addr.[hc],
global/mynetworks.[hc], postconf/postconf.c.
Portability: MACOSX patches by Gerben Wierda.
Portability: Solaris /dev/null is a symlink, which tripped
up the code to safely open a file before local delivery. We now
grudgingly allow symlinks owned by root. File: util/safe_open.c.
20010224
Bugfix: "postconf mynetworks" ignored the inet_interfaces
setting. That was a very old one. File: postconf/postconf.c.
INCOMPATIBLE CHANGE: POSTFIX NO LONGER RELAYS MAIL FOR
CLIENTS IN THE ENTIRE CLASS A/B/C NETWORK. POSTFIX BY
DEFAULT RELAYS MAIL FOR CLIENTS IN THE LOCAL SUBNETWORK.
Specify "mynetworks_style = class" to get the old behavior.
20010225
Portability: master sigchld handler based on writing to a
pipe, so that the master wakes up from select(). Based on
code by Erik Forsberg, Linkoping University, Sweden. File:
master/master_sig.c. Disabled until after the major release.
Code cleanup: Postfix should now run with no alias database.
Code cleanup: local_destination_recipient_limit and
local_destination_concurrency_limit are not first-class
configuration parameters. Files: global/mail_params.h,
*qmgr/qmgr.c, postconf/postconf.c.
INSTALL -d => INSTALL_*_DIR
INSTALL => INSTALL_*
Beware that only invocations are changed that don't set special
values for owner, mode or group!
Cleanup suggested by zuntum@IRC
INSTALL -d => INSTALL_*_DIR
INSTALL => INSTALL_*
Beware that only invocations are changed that don't set special
values for owner, mode or group!
Cleanup suggested by zuntum@IRC
