mmh
6 years agoWhoops, user & port were in the wrong order so msgchk didn't work for POP mmh-start
Ken Hornstein [Wed, 13 Apr 2011 14:50:49 +0000 (10:50 -0400)]
Whoops, user & port were in the wrong order so msgchk didn't work for POP
connections.

6 years agoUpdated the developer README with a few notes about the git conversion.
Ken Hornstein [Tue, 22 Feb 2011 01:44:30 +0000 (20:44 -0500)]
Updated the developer README with a few notes about the git conversion.

Signed-off-by: Ken Hornstein <kenh@pobox.com>

6 years agoSigned-off-by: Ken Hornstein <kenh@pobox.com>
Paul Vixie [Mon, 7 Feb 2011 19:13:54 +0000 (14:13 -0500)]
Signed-off-by: Ken Hornstein <kenh@pobox.com>

More documentation about the hooks support.

6 years agoClean up code in inc.c
Peter Maydell [Tue, 21 Dec 2010 16:50:24 +0000 (16:50 +0000)]
Clean up code in inc.c

Clean up the confusing use of a variable 'i' in inc.c.
The case where it was being used as 'an error occurred
which means we should not truncate the mailspool' is now
called 'incerr' and merged with the equivalent 'p' variable
used in the POP-specific code. Cases where it was being
used as a random temporary are still called 'i' but now
declared with a much more restricted scope.
This fixes a compilation warning for some compilers.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

6 years agoSuppress tsort diagnostic waffle
Peter Maydell [Mon, 20 Dec 2010 21:13:32 +0000 (21:13 +0000)]
Suppress tsort diagnostic waffle

tsort produces waffle to standard error if it detects loops in its input.
For our purposes in ordering object files this is harmless, so just send
it all to /dev/null.

Signed-off-by: Peter Maydell <pmaydell@chiark.greenend.org.uk>

6 years agoFix unused-function warnings in lex output
Peter Maydell [Mon, 20 Dec 2010 21:13:32 +0000 (21:13 +0000)]
Fix unused-function warnings in lex output

Fix the unused-function warnings in the output of lex. In particular, we
silence a warning about yyunput being static but unused by exposing it as
a global function. This costs us a few hundred bytes in the binary but
the other available solutions are flex-specific or more complicated.
YY_NO_UNPUT used to silence this but no longer works in recent flex.
On the other hand YY_NO_INPUT does work to silence warnings about input().

Signed-off-by: Peter Maydell <pmaydell@chiark.greenend.org.uk>

6 years agoFix used-uninitialised warning in fmt_rfc2047.c
Peter Maydell [Mon, 20 Dec 2010 21:13:32 +0000 (21:13 +0000)]
Fix used-uninitialised warning in fmt_rfc2047.c

Fix a used-uninitialised warning in fmt_rfc2047.c.
I don't think it can actually be used uninitialised, but the
code is too complicated for the compiler to realise this.

Signed-off-by: Peter Maydell <pmaydell@chiark.greenend.org.uk>

6 years agoFix used-uninitialised warning in new.c
Peter Maydell [Mon, 20 Dec 2010 21:13:32 +0000 (21:13 +0000)]
Fix used-uninitialised warning in new.c

Fix a used-uninitialised warning in new.c by having the check_folders()
routine return first=last=NULL for empty folder lists. This is a more
sensible API anyway. (It would not actually have been possible for
the code to try to use last while it was uninitialised, as it happens.)

Signed-off-by: Peter Maydell <pmaydell@chiark.greenend.org.uk>

6 years agoFix crash if fnext/fprev are given an empty folder list
Peter Maydell [Mon, 20 Dec 2010 21:13:32 +0000 (21:13 +0000)]
Fix crash if fnext/fprev are given an empty folder list

Add a check for a NULL pointer, which stops fnext/fprev
from segfaulting if given an empty folder list (for example
via 'fnext -folders emptyfile').

Signed-off-by: Peter Maydell <pmaydell@chiark.greenend.org.uk>

6 years agoFix warning in sbr/folder_addmsg.c by code cleanup
Peter Maydell [Mon, 20 Dec 2010 19:54:43 +0000 (19:54 +0000)]
Fix warning in sbr/folder_addmsg.c by code cleanup

Clean up some rather oddly written code in folder_addmsg()
so it is easier to understand for both humans and for the
compiler (and the compiler thus stops complaining about an
uninitialised variable).

Signed-off-by: Peter Maydell <pmaydell@chiark.greenend.org.uk>

7 years agoFix so this compiles without SASL support.
Ken Hornstein [Thu, 16 Dec 2010 18:20:31 +0000 (13:20 -0500)]
Fix so this compiles without SASL support.

Damn you, Oliver Kiddle, for finding all of the bugs I add to nmh!

7 years agoAll of the arguments to AC_ARG_ENABLE need to be quoted, otherwise the
Ken Hornstein [Thu, 16 Dec 2010 18:11:17 +0000 (13:11 -0500)]
All of the arguments to AC_ARG_ENABLE need to be quoted, otherwise the
last action doesn't make it into the configure script.

7 years agoMake sure we ignore config.h.in~
Ken Hornstein [Thu, 16 Dec 2010 18:02:32 +0000 (13:02 -0500)]
Make sure we ignore config.h.in~

7 years agofix typo and clarify two sentences in mhpath man page
Markus Schnalke [Thu, 16 Dec 2010 00:36:43 +0000 (01:36 +0100)]
fix typo and clarify two sentences in mhpath man page

7 years agoimprove mhparam man page
Markus Schnalke [Thu, 16 Dec 2010 00:25:11 +0000 (01:25 +0100)]
improve mhparam man page

7 years agofix badly formatted code found by Markus Schnalke
Oliver Kiddle [Thu, 16 Dec 2010 00:23:16 +0000 (01:23 +0100)]
fix badly formatted code found by Markus Schnalke

7 years agopropagate and check return status of close() in slocal
Scott Snyder [Wed, 15 Dec 2010 21:17:52 +0000 (22:17 +0100)]
propagate and check return status of close() in slocal

7 years agodocument msg parameter to folder
Markus Schnalke [Wed, 15 Dec 2010 21:02:30 +0000 (22:02 +0100)]
document msg parameter to folder

7 years agoswitches need to be renumbered following removal of some switches so,
Oliver Kiddle [Tue, 14 Dec 2010 17:04:01 +0000 (18:04 +0100)]
switches need to be renumbered following removal of some switches so,
e.g. inc -version was not working

7 years agoMassive garbage collection. Remove functionality for APOP, RPOP,
Ken Hornstein [Thu, 9 Dec 2010 20:29:29 +0000 (15:29 -0500)]
Massive garbage collection.  Remove functionality for APOP, RPOP,
all NNTP support (including bboards) and MPOP.

7 years ago * uip/whatnowsbr.c: enforce requirement of at least one file
David Levine [Wed, 8 Dec 2010 20:36:36 +0000 (14:36 -0600)]
* uip/whatnowsbr.c: enforce requirement of at least one file
  argument to attach command.

7 years agoman/send.man: added name= portion to Content-Type
David Levine [Wed, 8 Dec 2010 16:47:03 +0000 (10:47 -0600)]
man/send.man: added name= portion to Content-Type
headers to reflect 2006-10-24 change to uip/sendsbr.c.
That change allowed mhstore to use the (file)name when
extracting attachments.

7 years agoRemove RCS keywords, since they no longer work after git migration.
Ken Hornstein [Fri, 3 Dec 2010 20:23:11 +0000 (15:23 -0500)]
Remove RCS keywords, since they no longer work after git migration.

7 years agoWhoops, forgot to bracket the calls for TLS shutdown with a #ifdef TLS_SUPPORT
Ken Hornstein [Fri, 3 Dec 2010 02:04:30 +0000 (21:04 -0500)]
Whoops, forgot to bracket the calls for TLS shutdown with a #ifdef TLS_SUPPORT

7 years agoUpdate for git conversions (and cleanup of autogen.sh to use autoreconf)
Ken Hornstein [Fri, 3 Dec 2010 01:37:54 +0000 (20:37 -0500)]
Update for git conversions (and cleanup of autogen.sh to use autoreconf)

7 years agoChange configure.in to emit a warning when an argument is given to
Ken Hornstein [Thu, 2 Dec 2010 23:03:22 +0000 (18:03 -0500)]
Change configure.in to emit a warning when an argument is given to
--with-cyrus-sasl.

7 years agoFix the test for sasl.h to look in the correct place. cvs-repository-moved-to-git
Ken Hornstein [Sun, 21 Nov 2010 22:19:51 +0000 (22:19 +0000)]
Fix the test for sasl.h to look in the correct place.

7 years ago(Minor) cleanup of some of the autoconf code with respect to the handling
Ken Hornstein [Fri, 19 Nov 2010 20:13:39 +0000 (20:13 +0000)]
(Minor) cleanup of some of the autoconf code with respect to the handling
of compiler flags for SASL, and also rototill (and simplify) the autoconf
handling for the same.

Add TLS support for the SMTP MTA.

7 years agoman/slocal.man: fix formatting error that meant a subsection heading
Peter Maydell [Fri, 12 Nov 2010 23:28:26 +0000 (23:28 +0000)]
man/slocal.man: fix formatting error that meant a subsection heading
wasn't displayed as such. Followon cleanup: since the new heading is too
wide to fit in the margin, tidy the margin to 4 characters wide so we
are consistent with how all the headings are displayed.

7 years agoconfigure.in: add a level of expansion for mandir when printing the
Peter Maydell [Wed, 10 Nov 2010 22:26:07 +0000 (22:26 +0000)]
configure.in: add a level of expansion for mandir when printing the
end-of-configure install path summary.

7 years agoman/rmm.man: fix error in synopsis: the option is [no]unlink, not [no]link.
Peter Maydell [Tue, 9 Nov 2010 21:33:22 +0000 (21:33 +0000)]
man/rmm.man: fix error in synopsis: the option is [no]unlink, not [no]link.

7 years agoSilence a (harmless) uninitialized variable warning.
Peter Maydell [Fri, 5 Nov 2010 22:06:10 +0000 (22:06 +0000)]
Silence a (harmless) uninitialized variable warning.

7 years ago * uip/.cvsignore: added "new". PRE_POSIX_CONVERSION
David Levine [Tue, 31 Aug 2010 15:00:53 +0000 (15:00 +0000)]
* uip/.cvsignore:  added "new".

7 years ago#ifdef AI_ADDRCONFIG since older IPv4-only systems may not define this flag,
Earl Hood [Sat, 6 Feb 2010 07:47:38 +0000 (07:47 +0000)]
#ifdef AI_ADDRCONFIG since older IPv4-only systems may not define this flag,
causing files that use to not compile.

7 years agoFixed variable declaration so older compilers will not complain.
Earl Hood [Sat, 6 Feb 2010 07:45:47 +0000 (07:45 +0000)]
Fixed variable declaration so older compilers will not complain.
I.e. Older compilers may not support C99, so variables must be
declared at the beginning of a block.

7 years ago* Bug #15213, #18635: The use of the insecure m_scratch() and
Earl Hood [Thu, 4 Feb 2010 01:39:21 +0000 (01:39 +0000)]
* Bug #15213, #18635: The use of the insecure m_scratch() and
  m_tmpfil() functions have been replaced by m_mktemp()
  or m_mktemp2() functions (defined in sbr/m_mktemp.c).
  The new functions use mkstemp() to securely create
  temporary files to avoid the numerous race conditions
  that exist with the old functions.  This does assume
  that mkstemp() is available.  Unsure if we need to
  create an alternative implementation if mkstemp() is
  not available.  More information about new temp file
  functions in m_mktemp.c, including the support for
  MHTMPDIR, TMPDIR, and TMP envvars.

NOTE: The files sbr/m_scratch.c and sbr/m_tmpfil.c should be removed.
Until more testing verifies changes to support new functions have not
broken functionality, the files are be left in the tree, but comments
have been added stating the functions in them should not be called.

7 years ago* mts/smtp/smtp.c: added SASL support if mts configuration
Earl Hood [Wed, 3 Feb 2010 05:56:56 +0000 (05:56 +0000)]
* mts/smtp/smtp.c: added SASL support if mts configuration
  setting is set to "sendmail".  This is useful if sendmail
  conf option is to a custom script that creates a proxy
  connection to an smtp server.

* sbr/mts.c: added support for MHMTSCONF and MHMTSUSERCONF
  envvars.  The former specifies an alternative system
  mts.conf to use.  The later specifies a user-specific
  mts.conf to use.  This one will be read after the system
  conf, so the user's conf only needs to set options they
  want to override.  The MHMTSUSERCONF allows a user to set
  personal alternative mail submission methods w/o affecting
  other users on the system.

* uip/whom.c: added SASL-based options so address checking
  can work against a server that requires SASL.

7 years ago * uip/mhlistsbr.c, uip/mhlsbr.c, uip/picksbr.c: cast
David Levine [Wed, 30 Dec 2009 01:42:45 +0000 (01:42 +0000)]
* uip/mhlistsbr.c, uip/mhlsbr.c, uip/picksbr.c: cast
pointers to unsigned long before casting to unsigned int
to avoid warnings on 64-bit about casting from pointer
to int of different size.

7 years ago * uip/new.c: cast folder_len to int to avoid warning on
David Levine [Wed, 30 Dec 2009 01:41:47 +0000 (01:41 +0000)]
* uip/new.c: cast folder_len to int to avoid warning on
64-bit about passing a size_t to an int conversion specifier.

7 years ago * uip/inc.c: gcc 4.4.1 noticed that maildir could have been
David Levine [Fri, 25 Dec 2009 14:47:39 +0000 (14:47 +0000)]
* uip/inc.c: gcc 4.4.1 noticed that maildir could have been
used before initialization.  In fact, if that code was ever
reached, maildir would never have been initialized.  It's in
an adios call, so replaced with NULL.

7 years ago* mts/smtp/smtp.c: In sm_auth_sasl (), removed zeroing of
David Levine [Mon, 21 Dec 2009 17:18:04 +0000 (17:18 +0000)]
* mts/smtp/smtp.c: In sm_auth_sasl (), removed zeroing of
sasl_inbuffer because it could be used in sm_fget ().  Also,
removed allocation of sasl_inbuffer because that was done in
either smtp_init () or sendmail_init ().

8 years agoRemove duplicate sm_fgetc() declaration
Josh Bressers [Mon, 26 Jan 2009 13:38:13 +0000 (13:38 +0000)]
Remove duplicate sm_fgetc() declaration

8 years agoFix warning in getaddrinfo() call.
Ken Hornstein [Wed, 21 Jan 2009 19:38:36 +0000 (19:38 +0000)]
Fix warning in getaddrinfo() call.
Add missing support for -port in whatnow.
Make sure the SASL input buffer is allocated when using the sendmail mts.

8 years ago * etc/mhn.defaults.sh: Never try to make mhshow start xterm.
Eric Gillespie [Sun, 18 Jan 2009 04:48:42 +0000 (04:48 +0000)]
* etc/mhn.defaults.sh: Never try to make mhshow start xterm.

8 years agoNote 'new' commands.
Eric Gillespie [Sat, 17 Jan 2009 16:53:07 +0000 (16:53 +0000)]
Note 'new' commands.

8 years ago * test/runtest, test/tests/inc/test-deb359167,
Eric Gillespie [Sat, 17 Jan 2009 16:47:29 +0000 (16:47 +0000)]
* test/runtest, test/tests/inc/test-deb359167,
test/tests/inc/test-eom-align, test/tests/manpages/test-manpages:
Load common.sh via absolute path, otherwise some /bin/sh
(e.g. dash) can't load it.

* uip/Makefile.in, uip/new.c, test/tests/new/test-basic: Add new
program, and fn/fp/unseen symlinks.

* test/{runtest,setup-test}: Move MH profile under Mail directory
so each test script will have its own to muck with, if needed.

* h/Makefile.in, h/crawl_folders.h, sbr/Makefile.in,
sbr/crawl_folders.c, uip/folder.c: Extract the folder crawling
code from folder.c into new crawl_folders function, using a
callback to assemble the folder info in folder.c.  Drop compare
function and use strcmp instead.  Rename addfold and addir to
add_folder and add_children (add dir vs. add folder?
confusing names).

8 years agoMove freeaddrinfo() to inside of our loop; this prevents a memory leak
Ken Hornstein [Fri, 16 Jan 2009 15:48:09 +0000 (15:48 +0000)]
Move freeaddrinfo() to inside of our loop; this prevents a memory leak
and also prevents it from being called with a NULL pointer.

8 years agoMassive overhaul of networking code. Changes:
Ken Hornstein [Fri, 16 Jan 2009 02:28:54 +0000 (02:28 +0000)]
Massive overhaul of networking code.  Changes:

- All networking interfaces now converted to new APIs (getaddrinfo() and
  the like).  All networking now supports IPv6 (with the exception of
  the built-in FTP client).
- SASL security layers are now supported for SMTP.
- Command-line and profile-based selection of the SMTP server and port.

8 years ago* sbr/m_getfld.c: fix two bugs which could cause us to walk off
Peter Maydell [Fri, 26 Dec 2008 21:54:19 +0000 (21:54 +0000)]
* sbr/m_getfld.c: fix two bugs which could cause us to walk off
the beginning of the stdio or prefix-string buffer when checking
for presence of the end-of-message delimiter in some situations.
This might cause inc to dump core if you were unlucky. (This was
Debian bug 359167.)
* test/tests/inc/{md5sums,msgheader.txt,test-deb359167,
test-eom-align,fromline.txt,filler.txt,deb359167.mbox}: new tests
both for the specific problem and to try to check various alignments
of the eom string with the stdio buffer ending.
* test/common.sh: Added some functions for doing progress indicators.

8 years agotest/common.sh: new file for common utility functions for the test scripts.
Peter Maydell [Fri, 26 Dec 2008 16:32:07 +0000 (16:32 +0000)]
test/common.sh: new file for common utility functions for the test scripts.
Moved findprog out of manpage test script into this new file.

8 years agotest/runtest: automatically run setup-test if it hasn't already
Peter Maydell [Thu, 25 Dec 2008 23:30:33 +0000 (23:30 +0000)]
test/runtest: automatically run setup-test if it hasn't already
been done. Improve error handling in test script with 'set -e'.

9 years ago * h/mh.h: raised NAMESZ from 128 to 999. RFC 2822 limits line
David Levine [Thu, 23 Oct 2008 18:38:22 +0000 (18:38 +0000)]
* h/mh.h: raised NAMESZ from 128 to 999.  RFC 2822 limits line
lengths to 998 characters, so a header name can be at most that long.
m_getfld limits header names to 2 less than NAMESZ, which is fine,
because header names must be followed by a colon.  Add one for
terminating NULL.

9 years ago * sbr/m_getfld.c: in warning message, use NAMESZ-2 instead of
David Levine [Thu, 23 Oct 2008 18:14:52 +0000 (18:14 +0000)]
* sbr/m_getfld.c: in warning message, use NAMESZ-2 instead of
NAMESZ-1 bytes because the message says "exceeds", not
"is greater than or equal to".

9 years ago * docs/MAIL.FILTERING: added note on removing procmail -f or
David Levine [Thu, 23 Oct 2008 16:45:36 +0000 (16:45 +0000)]
* docs/MAIL.FILTERING: added note on removing procmail -f or
-r options to try to suppress generation of a "From " line on
incoming messages.

9 years ago+mhn.defaults.sh.cfg
Eric Gillespie [Tue, 19 Aug 2008 04:36:42 +0000 (04:36 +0000)]
+mhn.defaults.sh.cfg

9 years ago * test/setup-test: Don't run autogen.sh unless needed. Build in a
Eric Gillespie [Thu, 14 Aug 2008 19:47:29 +0000 (19:47 +0000)]
* test/setup-test: Don't run autogen.sh unless needed.  Build in a
blddir under the temporary directory so as not to interfere with
whatever the developer may be doing in the source directory.
If the temporary directory already exists, rebuild and install it.

9 years agoThis is what happens with last-minute "safe" improvements: actually point
Eric Gillespie [Thu, 14 Aug 2008 06:19:08 +0000 (06:19 +0000)]
This is what happens with last-minute "safe" improvements: actually point
chset to something in the new merged InitText!

9 years ago * sbr/ambigsw.c: Send print_sw output to stderr. This avoids
Eric Gillespie [Thu, 14 Aug 2008 01:50:45 +0000 (01:50 +0000)]
* sbr/ambigsw.c: Send print_sw output to stderr.  This avoids
strange problems with scan `pick -bogus`, not to mention being
nicer anyway.

* sbr/getans.c, sbr/print_help.c, uip/msh.c: Send print_sw output
to stdout.

* h/prototypes.h, sbr/print_sw.c (print_sw): Add FILE * argument
and send output there rather than to stdout.

* test/tests/pick/test-stderr: Add test that error messages don't
end up going to stdout.

9 years ago * uip/pick.c: Ensure valid sequence name as soon as we parse it from
Eric Gillespie [Thu, 14 Aug 2008 01:31:11 +0000 (01:31 +0000)]
* uip/pick.c: Ensure valid sequence name as soon as we parse it from
command, instead of searching the messages first.

9 years ago * uip/pick.c: Print matching messages immediately, instead of
Eric Gillespie [Thu, 14 Aug 2008 01:28:56 +0000 (01:28 +0000)]
* uip/pick.c: Print matching messages immediately, instead of
waiting until the very end to print anything.

9 years ago * uip/Makefile.in, uip/popi.c: Remove popi.c, which has been dead
Eric Gillespie [Thu, 14 Aug 2008 01:11:12 +0000 (01:11 +0000)]
* uip/Makefile.in, uip/popi.c: Remove popi.c, which has been dead
since PatchSet 257 (2000/06/12 11:07:34 UTC).

9 years ago * h/mhparse.h: Move interfaces also used by mhbuildsbr.c here
Eric Gillespie [Thu, 14 Aug 2008 00:56:38 +0000 (00:56 +0000)]
* h/mhparse.h: Move interfaces also used by mhbuildsbr.c here
from mhparse.c.

* uip/Makefile.in (mhbuild): Build with mhparse.o.

* uip/mhbuildsbr.c: Drop the duplication of mhparse.c.

* uip/mhbuild.c, uip/mhlist.c, uip/mhn.c, uip/mhshow.c,
uip/mhstore.c, uip/mhtest.c: Drop check swdeclaration, which is
now in mhparse.h.

* uip/mhparse.c: Move some declarations to mhparse.h, and remove
'static' from the definitions here.
(get_ctinfo): Take magic mhbuild support from mhbuildsbr.c's copy,
and adapt callers.
(incl_name_value, extract_name_value): Move from mhbuildsbr.c to
support get_ctinfo change).
(get_content, open7Bit): Add Content-Disposition support, added to
mhbuildsbr.c's copy in mhbuildsbr.c r1.12 (PatchSet 549).
(InitText): Default to CHARSET_UNSPECIFIED instead of "US-ASCII",
as mhbuildsbr.c's copy did.  Explain termproc.
(openBase64, openQuoted, open7Bit): Fix comment from r1.5:
suffixes, not prefixes.
(openMail): Explain showproc.

* test/tests/mhshow/test-cte-binary: Add test for
Content-Transfer-Encoding: binary (Debian #136976).

* test/tests/mhbuild/test-forw: Add an mhbuild test, testing forw
with and without RFC-934 mode.  More mhbuild tests would be nice.

9 years agoOops, meant to use $failed (number of failed checks) as the exit code.
Eric Gillespie [Wed, 13 Aug 2008 23:59:56 +0000 (23:59 +0000)]
Oops, meant to use $failed (number of failed checks) as the exit code.

9 years agoAdd descriptive headers, as other tests have.
Eric Gillespie [Wed, 13 Aug 2008 23:32:03 +0000 (23:32 +0000)]
Add descriptive headers, as other tests have.

9 years ago * test/tests/bad-input/test-header: Add test for it.
Eric Gillespie [Wed, 13 Aug 2008 18:27:36 +0000 (18:27 +0000)]
* test/tests/bad-input/test-header: Add test for it.

* sbr/m_getfld.c: If we reach the end of the line without finding
a ':' when parsing a header field, treat that line as the
beginning of the body rather than blowing up.  These messages are
usually spam, but it's nice to be able to at least scan them.

9 years ago * test/tests/mhshow/test-qp: Test various valid and invalid
Eric Gillespie [Wed, 13 Aug 2008 01:01:12 +0000 (01:01 +0000)]
* test/tests/mhshow/test-qp: Test various valid and invalid
escape sequences.

* uip/mhparse.c (openQuoted): Simplify the decode-or-show for loop by
peeking ahead to the next byte(s) when encountering '=', and just let
invalid escape sequences through as literals (fixes bug #15245).

9 years agoIgnore test-temp-dir.
Eric Gillespie [Wed, 13 Aug 2008 01:00:20 +0000 (01:00 +0000)]
Ignore test-temp-dir.

9 years agoadd one more hint for developers, to explicitly point at docs/README.developers
Paul Fox [Tue, 12 Aug 2008 21:01:30 +0000 (21:01 +0000)]
add one more hint for developers, to explicitly point at docs/README.developers

9 years agoautogen.sh (new file): add script for running the GNU autotools in the
Peter Maydell [Tue, 12 Aug 2008 18:04:29 +0000 (18:04 +0000)]
autogen.sh (new file): add script for running the GNU autotools in the
right order. Mention and use new script in the appropriate places.

9 years ago * sbr/discard.c, sbr/m_getfld.c, uip/scansbr.c: Add support for
Eric Gillespie [Fri, 8 Aug 2008 23:45:24 +0000 (23:45 +0000)]
* sbr/discard.c, sbr/m_getfld.c, uip/scansbr.c: Add support for
DragonFlyBSD stdio (from NetBSD pkgsrc).

9 years agos/ls/cd/ in comment (probably pasted from test-ls)
Eric Gillespie [Fri, 8 Aug 2008 04:45:05 +0000 (04:45 +0000)]
s/ls/cd/ in comment (probably pasted from test-ls)

9 years agouip/folder.c: chdir(nmhdir) in main rather than in dodir, which
Peter Maydell [Tue, 5 Aug 2008 21:06:00 +0000 (21:06 +0000)]
uip/folder.c: chdir(nmhdir) in main rather than in dodir, which
is called many times.  Drop the chdir in get_folder_info, which
served no purpose at all.  If struct stat has d_type (BSD), use
that to skip processing (strcmp, stat, and so on) of message
files, mostly undoing the slow-down from the last change.
Plus autoconf machinery to detect d_type field. Patch from Eric Gillespie.

9 years agoDump hacky overoptimisation in addir -- it doesn't actually get the case
Peter Maydell [Tue, 5 Aug 2008 19:09:03 +0000 (19:09 +0000)]
Dump hacky overoptimisation in addir -- it doesn't actually get the case
of symlinks to directories right. Patch from Eric Gillespie.

9 years agoSimplify dodir/addir/addfold (patch from Eric Gillespie)
Peter Maydell [Tue, 5 Aug 2008 19:05:23 +0000 (19:05 +0000)]
Simplify dodir/addir/addfold (patch from Eric Gillespie)

9 years agobug #23319: rework the way we handle the working directory when invoking
Peter Maydell [Sun, 3 Aug 2008 18:47:56 +0000 (18:47 +0000)]
bug #23319: rework the way we handle the working directory when invoking
the user's shell, so that we don't have issues when the cwd contains a
space or shell special character.

9 years agotest/setup-test: use 'set -e' so we stop on compile failure.
Peter Maydell [Sun, 3 Aug 2008 18:45:34 +0000 (18:45 +0000)]
test/setup-test: use 'set -e' so we stop on compile failure.
Configure --enable-debug for convenience in debugging.

9 years agoAdd protective 'do { ... } while (0)' wrappers to multistatement macros
Peter Maydell [Sun, 3 Aug 2008 15:19:53 +0000 (15:19 +0000)]
Add protective 'do { ... } while (0)' wrappers to multistatement macros

9 years agoBug #23436: fix some minor manpage syntax errors and add a test which runs
Peter Maydell [Sun, 3 Aug 2008 15:14:23 +0000 (15:14 +0000)]
Bug #23436: fix some minor manpage syntax errors and add a test which runs
groff over them and checks that no warnings are emitted.

9 years agoAllow tests to indicate that they have been skipped by printing their
Peter Maydell [Sun, 3 Aug 2008 15:13:34 +0000 (15:13 +0000)]
Allow tests to indicate that they have been skipped by printing their
own message and returning with exit status 120.

9 years agoIgnore editor backup files
Peter Maydell [Sun, 3 Aug 2008 13:59:19 +0000 (13:59 +0000)]
Ignore editor backup files

9 years agoetc/replcomps, etc/replgroupcomps: Trim trailing whitespace from
Peter Maydell [Sun, 3 Aug 2008 13:54:37 +0000 (13:54 +0000)]
etc/replcomps, etc/replgroupcomps: Trim trailing whitespace from
Subject field (patch from Eric Gillespie)

9 years agoWhen doing an if-test on the result of a function which returns a string
Peter Maydell [Sun, 3 Aug 2008 13:32:17 +0000 (13:32 +0000)]
When doing an if-test on the result of a function which returns a string
result, check whether the string is non-empty (as the documentation says we
do). Previously we were always testing the integer 'value'.

9 years ago* test/setup-test: Run 'make clean' before building
Josh Bressers [Fri, 25 Jul 2008 01:07:39 +0000 (01:07 +0000)]
* test/setup-test: Run 'make clean' before building

9 years ago* test/runtest: cat test-temp-dir only once. Destroy and create
Josh Bressers [Fri, 25 Jul 2008 00:11:53 +0000 (00:11 +0000)]
* test/runtest: cat test-temp-dir only once.  Destroy and create
test Mail hierarchy here, so each test has its own independent
test data.

* test/setup-test: Drop needless pushd/popd, which don't exist in
sh anyway.  Don't use aclocal, which is part of automake.
Don't create Mail hierarchy here (see test/runtest).

* test/tests/folder/test-create, tests/folder/test-total: Drop
unused BINDIR.

* tests/pick/test-pick: Remove failing no-op test.

* test/tests/scan/test-scan: Write expected and actual output to
files and diff them.  This way, if the output differs, diff
provides the FAILing exit code and shows the differences.

9 years ago* man/repl.man, etc/mhl.reply:
Josh Bressers [Fri, 13 Jun 2008 00:11:49 +0000 (00:11 +0000)]
* man/repl.man, etc/mhl.reply:
Add an attribution line when replying to messages.

9 years agoRemove an accidentally committed file
Josh Bressers [Thu, 12 Jun 2008 23:29:16 +0000 (23:29 +0000)]
Remove an accidentally committed file

9 years agoWhen building on gcc, use noreturn attribute on adios and the various done
Peter Maydell [Mon, 2 Jun 2008 22:37:01 +0000 (22:37 +0000)]
When building on gcc, use noreturn attribute on adios and the various done
functions.

9 years agoBack out previous, completely broken, change (due to a foulup I was
Peter Maydell [Mon, 2 Jun 2008 22:05:11 +0000 (22:05 +0000)]
Back out previous, completely broken, change (due to a foulup I was
compiling a different tree to the one I was committing to CVS).

9 years agoIf we're compiling with gcc, mark adios() as never returning (better code
Peter Maydell [Mon, 2 Jun 2008 21:49:07 +0000 (21:49 +0000)]
If we're compiling with gcc, mark adios() as never returning (better code
and fewer spurious 'uninitialized variable) warnings)

9 years agoUpdate bit of docs about release announcements.
Peter Maydell [Sun, 1 Jun 2008 16:50:22 +0000 (16:50 +0000)]
Update bit of docs about release announcements.

9 years agoDon't use MB_CUR_MAX if we aren't compiled with multibyte support.
Peter Maydell [Thu, 22 May 2008 10:37:55 +0000 (10:37 +0000)]
Don't use MB_CUR_MAX if we aren't compiled with multibyte support.

9 years agoFactor out common code for writing ls shell command. Rework its length
Peter Maydell [Thu, 22 May 2008 10:21:48 +0000 (10:21 +0000)]
Factor out common code for writing ls shell command. Rework its length
checking so it doesn't rely on the return value of sprintf(), for the
benefit of SunOS 4.

9 years agoDon't rely on realloc() to follow POSIX in handling a NULL pointer
Peter Maydell [Wed, 21 May 2008 18:05:49 +0000 (18:05 +0000)]
Don't rely on realloc() to follow POSIX in handling a NULL pointer

9 years agoAdd some table size declarations for the benefit of elderly lexes with
Peter Maydell [Wed, 21 May 2008 17:16:44 +0000 (17:16 +0000)]
Add some table size declarations for the benefit of elderly lexes with
small defaults (eg SunOS 4)

9 years agoUpdate comment now we don't use lex -o.
Peter Maydell [Tue, 20 May 2008 19:04:57 +0000 (19:04 +0000)]
Update comment now we don't use lex -o.

9 years ago* sbr/addrsbr.c: removed "err" from conditional, because gcc
David Levine [Sun, 18 May 2008 13:56:58 +0000 (13:56 +0000)]
* sbr/addrsbr.c: removed "err" from conditional, because gcc
correctly warned that it would always evaluate as true.  It's
the address of a static array.

9 years agodon't use $< in target rules in makefiles, as POSIX says it's only defined
Oliver Kiddle [Fri, 16 May 2008 15:31:26 +0000 (15:31 +0000)]
don't use $< in target rules in makefiles, as POSIX says it's only defined
in inference rules

9 years ago* configure.in, INSTALL: if --enable-masquerade is not
David Levine [Tue, 13 May 2008 02:39:04 +0000 (02:39 +0000)]
* configure.in, INSTALL: if --enable-masquerade is not
specified to configure, enable all supported masquerade forms.
This allows users to masquerade with the default
configuration. That seems to be more worthwhile now than
trying to prevent users from using it, especially with
single-user installations or those where a user can edit
etc/mts.conf.

9 years agobug #23167: sbr/ruserpass.c (ruserpass): make bad permissions on .netrc be
Peter Maydell [Sun, 4 May 2008 20:58:35 +0000 (20:58 +0000)]
bug #23167: sbr/ruserpass.c (ruserpass): make bad permissions on .netrc be
an instantly fatal error. Previously we returned an error value; however, no
caller was checking it. So now ruserpass() has a void return type.

9 years ago* bug #23163: various minor fixes for the benefit of older Unixes
Peter Maydell [Sun, 4 May 2008 16:09:08 +0000 (16:09 +0000)]
* bug #23163: various minor fixes for the benefit of older Unixes
(specifically SunOS 4): reintroduce strerror() substitute implementation;
provide memmove() substitute implementation
* bug #23163: fix broken 'build outside source directory' feature
* bug #23162: sbr/dtime.c: fix stray HAVE_TM_GMTOFF that wasn't updated
to the new macro name.