In regard to: Re: [Imap-uw] unable to compile imap-2006a on solaris 10,...:
Thanks for calling this issue with alphasort on Solaris to my attention.
I believe that this issue is resolved in the latest imap-2006b development
snapshot.
I will probably be promoting imap-2006b from development to release in the
near future. I'm just trying to wait to see if any other issues come out, in
the hope of avoiding an imap-2006c bugfix release.
I just downloaded the snapshot a few minutes ago. For production builds
in our environment we normally apply a local patch for authorization, but
I've tested with no modifications to the source.
I'm attempting to build on:
$cat /etc/redhat-release
Red Hat Enterprise Linux AS release 4 (Nahant Update 4)
If I just untar the sources and try:
make lrh
or
make lfd
I'm prompted regarding IPv6 support. If I say "n", the build fails:
06:27 PM rh4as-dev BUILD$zcat ../SOURCES/imap-2006b.DEV.tar.Z | tar -xf -
06:27 PM rh4as-dev BUILD$cd imap-2006b.DEV.SNAP-0610031354/
06:28 PM rh4as-dev imap-2006b.DEV.SNAP-0610031354$make lrh
make[1]: Entering directory
`/usr/local/src/RPM/BUILD/imap-2006b.DEV.SNAP-0610031354'
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+ Building in full compliance with RFC 3501 security
+ requirements:
++ TLS/SSL encryption is supported
++ Unencrypted plaintext passwords are prohibited
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
make[1]: Leaving directory
`/usr/local/src/RPM/BUILD/imap-2006b.DEV.SNAP-0610031354'
Applying an process to sources...
tools/an "ln -s" src/c-client c-client
tools/an "ln -s" src/ansilib c-client
tools/an "ln -s" src/charset c-client
tools/an "ln -s" src/osdep/unix c-client
tools/an "ln -s" src/mtest mtest
tools/an "ln -s" src/ipopd ipopd
tools/an "ln -s" src/imapd imapd
tools/an "ln -s" src/mailutil mailutil
tools/an "ln -s" src/mlock mlock
tools/an "ln -s" src/dmail dmail
tools/an "ln -s" src/tmail tmail
ln -s tools/an .
make build EXTRACFLAGS='' EXTRALDFLAGS='' EXTRADRIVERS='mbox'
EXTRAAUTHENTICATORS='' PASSWDTYPE=std SSLTYPE=nopwd IP=4 EXTRASPECIALS=''
BUILDTYPE=lnp IP=6 \
EXTRACFLAGS=" -I/usr/kerberos/include" \
SPECIALS="SSLINCLUDE=/usr/include/openssl SSLLIB=/usr/lib
SSLCERTS=/usr/share/ssl/certs SSLKEYS=/usr/share/ssl/private GSSDIR=/usr/kerberos
LOCKPGM=/usr/sbin/mlock"
make[1]: Entering directory
`/usr/local/src/RPM/BUILD/imap-2006b.DEV.SNAP-0610031354'
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+ Building with IPv6 support
+
+ NOTE: Some versions of glibc have a bug in the getaddrinfo
+ call which does DNS name resolution. This bug causes host
+ names to be canonicalized incorrectly, as well as doing an
+ unnecessary and performance-sapping reverse DNS call. This
+ problem does not affect the IPv4 gethostbyname call.
+
+ getaddrinfo is known to work properly on Mac OS X and
+ Windows. However, the problem has been observed on some
+ Linux systems.
+
+ If you are not building for Mac OS X or Windows, you may
+ want to build with IP=4 unless you are certain that glibc
+ is fixed on your system.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Do you want to build with IPv6 anyway? Type y or n please:
n
make[2]: Entering directory
`/usr/local/src/RPM/BUILD/imap-2006b.DEV.SNAP-0610031354'
Removing old processed sources and binaries...
sh -c 'rm -rf an ua OSTYPE SPECIALS c-client mtest imapd ipopd mailutil mlock
dmail tmail || true'
cd tools;make clean
make[3]: Entering directory
`/usr/local/src/RPM/BUILD/imap-2006b.DEV.SNAP-0610031354/tools'
sh -c 'rm -f *.o uahelper || true'
make[3]: Leaving directory
`/usr/local/src/RPM/BUILD/imap-2006b.DEV.SNAP-0610031354/tools'
make[2]: Leaving directory
`/usr/local/src/RPM/BUILD/imap-2006b.DEV.SNAP-0610031354'
make[1]: *** [ip6] Error 1
make[1]: Leaving directory
`/usr/local/src/RPM/BUILD/imap-2006b.DEV.SNAP-0610031354'
make: *** [lrh] Error 2
06:28 PM rh4as-dev imap-2006b.DEV.SNAP-0610031354$
Even if I try
make lrh IP=4
it still goes through the IPv6 warning and question, because the IP=6
from the lrh target is on the command line later than the IP=4 that I
passed.
It looks like the build proceeds without problems if I say "y" for IPv6,
but we don't want to be testing that at the same time we're testing MIX.
We will be deploying IPv6 in the near future, though, so I'm glad to see
that it's ready when we are.
Tim
--
Tim Mooney [EMAIL PROTECTED]
Information Technology Services (701) 231-1076 (Voice)
Room 242-J6, IACC Building (701) 231-8541 (Fax)
North Dakota State University, Fargo, ND 58105-5164
_______________________________________________
Imap-uw mailing list
Imap-uw@u.washington.edu
https://mailman1.u.washington.edu/mailman/listinfo/imap-uw
