+20040114
+ - (dtucker) [auth-pam.c] Have monitor die if PAM authentication thread exits
+ unexpectedly. with & ok djm@
+ - (dtucker) [auth-pam.c] Reset signal handler in pthread_cancel too, add
+ test for case where cleanup has already run.
+ - (dtucker) [auth-pam.c] Add minor debugging.
+
+20040113
+ - (dtucker) [auth-pam.c] Relocate struct pam_ctxt and prototypes. No
+ functional changes.
+
+20040108
+ - (dtucker) [auth-pam.c defines.h] Bug #783: move __unused to defines.h and
+ only define if not already. From des at freebsd.org.
+ - (dtucker) [configure.ac] Remove extra (typo) comma.
+
+20040105
+ - (dtucker) [contrib/ssh-copy-id] Bug #781: exit if ssh fails. Patch from
+ cjwatson at debian.org.
+ - (dtucker) [acconfig.h configure.ac includes.h servconf.c session.c]
+ Only enable KerberosGetAFSToken if Heimdal's libkafs is found. with jakob@
+
+20040102
+ - (djm) OSX/Darwin needs BIND_8_COMPAT to build getrrsetbyname. Report from
+ jakob@
+ - (djm) Remove useless DNS support configure summary message. from jakob@
+ - (djm) OSX/Darwin put the PAM headers in a different place, detect this.
+ Report from jakob@
+
+20031231
+ - (dtucker) OpenBSD CVS Sync
+ - djm@cvs.openbsd.org 2003/12/22 09:16:58
+ [moduli.c ssh-keygen.1 ssh-keygen.c]
+ tidy up moduli generation debugging, add -v (verbose/debug) option to
+ ssh-keygen; ok markus@
+ - markus@cvs.openbsd.org 2003/12/22 20:29:55
+ [cipher-3des1.c]
+ EVP_CIPHER_CTX_cleanup() for the des contexts; pruiksma@freesurf.fr
+ - jakob@cvs.openbsd.org 2003/12/23 16:12:10
+ [servconf.c servconf.h session.c sshd_config]
+ implement KerberosGetAFSToken server option. ok markus@, beck@
+ - millert@cvs.openbsd.org 2003/12/29 16:39:50
+ [sshd_config]
+ KeepAlive has been obsoleted, use TCPKeepAlive instead; markus@ OK
+ - dtucker@cvs.openbsd.org 2003/12/31 00:24:50
+ [auth2-passwd.c]
+ Ignore password change request during password auth (which we currently
+ don't support) and discard proposed new password. corrections/ok markus@
+ - (dtucker) [configure.ac] Only test setresuid and setresgid if they exist.
+
+20031219
+ - (dtucker) [defines.h] Bug #458: Define SIZE_T_MAX as UINT_MAX if we
+ typedef size_t ourselves.
+
+20031218
+ - (dtucker) [configure.ac] Don't use setre[ug]id on DG-UX, from Tom Orban.
+ - (dtucker) [auth-pam.c] Do PAM chauthtok during SSH2 keyboard-interactive
+ authentication. Partially fixes bug #423. Feedback & ok djm@
+
+20031217
+ - (djm) OpenBSD CVS Sync
+ - markus@cvs.openbsd.org 2003/12/09 15:28:43
+ [serverloop.c]
+ make ClientKeepAlive work for ssh -N, too (no login shell requested).
+ 1) send a bogus channel request if we find a channel
+ 2) send a bogus global request if we don't have a channel
+ ok + test beck@
+ - markus@cvs.openbsd.org 2003/12/09 17:29:04
+ [sshd.c]
+ fix -o and HUP; ok henning@
+ - markus@cvs.openbsd.org 2003/12/09 17:30:05
+ [ssh.c]
+ don't modify argv for ssh -o; similar to sshd.c 1.283
+ - markus@cvs.openbsd.org 2003/12/09 21:53:37
+ [readconf.c readconf.h scp.1 servconf.c servconf.h sftp.1 ssh.1]
+ [ssh_config.5 sshconnect.c sshd.c sshd_config.5]
+ rename keepalive to tcpkeepalive; the old name causes too much
+ confusion; ok djm, dtucker; with help from jmc@
+ - dtucker@cvs.openbsd.org 2003/12/09 23:45:32
+ [clientloop.c]
+ Clear exit code when ssh -N is terminated with a SIGTERM. ok markus@
+ - markus@cvs.openbsd.org 2003/12/14 12:37:21
+ [ssh_config.5]
+ we don't support GSS KEX; from Simon Wilkinson
+ - markus@cvs.openbsd.org 2003/12/16 15:49:51
+ [clientloop.c clientloop.h readconf.c readconf.h scp.1 sftp.1 ssh.1]
+ [ssh.c ssh_config.5]
+ application layer keep alive (ServerAliveInterval ServerAliveCountMax)
+ for ssh(1), similar to the sshd(8) option; ok beck@; with help from
+ jmc and dtucker@
+ - markus@cvs.openbsd.org 2003/12/16 15:51:54
+ [dh.c]
+ use <= instead of < in dh_estimate; ok provos/hshoexer;
+ do not return < DH_GRP_MIN
+ - (dtucker) [acconfig.h configure.ac uidswap.c] Bug #645: Check for
+ setres[ug]id() present but not implemented (eg some Linux/glibc
+ combinations).
+ - (bal) [openbsd-compat/bsd-misc.c] unset 'signal' defined if we are
+ using a real 'signal()' (Noticed by a NeXT Compile)
+