--- cygnus/src/include/ChangeLog	Tue Jul  1 14:06:54 1997
+++ kerbnet-1.2/include/ChangeLog	Tue May 27 21:57:41 1997
@@ -1,7 +1,94 @@
+Thu Aug 15 16:31:20 1996  Theodore Ts'o  <tytso@rsts-11.mit.edu>
+
+	* k5-int.h: Since we are only supporting the db in the util/db2
+		directory, just use db-ndbm.h, instead of hoping that the
+		system ndbm.h will be appropriate.
+
+Fri Aug  2 14:15:26 1996  Ezra Peisach  <epeisach@kangaroo.mit.edu>
+
+	* krb5.hin: Add prototype for krb5_free_keyblock_contents
+
+Wed Jul 24 00:38:40 1996  Sam Hartman  <hartmans@tertius.mit.edu>
+
+	* k5-int.h: Prototype krb5_setenv and krb5_unsetenv
+
+Sun Jul  7 12:27:39 1996  Ezra Peisach  <epeisach@kangaroo.mit.edu>
+
+	* krb5.hin: Add prototypes for krb5_get_credentials_renew() and
+		krb5_get_cred_from_kdc_renew() 
+
+Wed Jun 12 01:32:33 1996  Theodore Ts'o  <tytso@rsts-11.mit.edu>
+
+	* syslog.h: Add #ifdef _WIN32 in places where we had #ifdef _MSDOS
+
+Mon Jun 10 16:52:35 1996  Theodore Ts'o  <tytso@rsts-11.mit.edu>
+
+	* krb5.hin		
+	* k5-int.h: For Win-32: always #include windows.h.  
+
+		Change INTERFACE to be KRB5_CALLCONV, which is where
+		the calling convention is defined.  Add KRB5_DLLIMP
+		which is where the DLL import/export should be put for
+		Win32.  (Win16 is just different.)  The correct way to
+		declare a function which will be used in a DLL is now:
+		KRB5_DLLIMP func_return_t KRB5_CALLCONV func(long)
+
+		Change function delcarations to use the new
+		convention.  Actually, it doesn't hurt to use the old
+		convention as long as func_return_t doesn't contain a
+		'*'.  But in the long run we should be exterminating
+		all uses of INTERFACE in favor of KRB5_CALLCONV and
+		KRB5_DLLIMP.
+
+Fri Jun  7 18:13:33 1996  Theodore Y. Ts'o  <tytso@mit.edu>
+
+	* krb5.hin: 
+	* k5-int.h: Beginnings of Win-32 support.
+
+Thu Jun  6 14:14:28 1996  Theodore Y. Ts'o  <tytso@mit.edu>
+
+	* krb5.hin (krb5_cc_get_type): Remove extraneous '*' from definition.
+
+Mon May 20 02:51:06 1996  Sam Hartman  <hartmans@mit.edu>
+
+	* k5-int.h: Ultrix CC wants krb5_seteuid to take uid_t so that the
+ 	type can be promoted from a short.
+
+Sat May 18 16:53:06 1996  Sam Hartman  <hartmans@tertius.mit.edu>
+
+	* k5-int.h: Add krb5_seteuid
+
+Tue May 14 20:29:09 1996  Richard Basch  <basch@lehman.com>
+
+	* krb5.hin: replaced CKSUMTYPE_SHA_DES3 with CKSUMTYPE_HMAC_SHA
+
+Thu May  9 19:28:48 1996  Theodore Y. Ts'o  <tytso@mit.edu>
+
+	* k5-int.h: Added new field values for ap_req_sumtype and
+	 	safe_sumtype.  Added new convenience function for
+		verifying magic numbers: KRB5_VERIFY_MAGIC.
+
+	* krb5.hin (krb5_auth_con_set_req_cksumtype,
+	 	krb5_auth_con_set_safe_cksumtype): Added prototypes of
+		new functions, to replace old krb5_auth_con_setcksumtype.
+
+Fri May 10 01:26:24 1996  Richard Basch  <basch@lehman.com>
+
+	* k5-int.h krb5.hin: Replaced 3des-md5 with 3des-sha.
+
+Tue May  7 17:19:17 1996  Richard Basch  <basch@lehman.com>
+
+	* k5-int.h: We are providing 3DES routines in libcrypto, too...
+
+Thu May  2 18:41:00 1996  Richard Basch  <basch@lehman.com>
+
+	* krb5.hin: krb5_init_random_key & krb5_finish_random_key now
+		pass the eblock to the underlying crypto routine.
+
 Fri May  3 00:26:47 1996  Mark Eichin  <eichin@cygnus.com>
 
-	* krb5.hin (krb5_get_credentials_validate): declaration for new
-	function.
+	* krb5.hin
+	(krb5_get_credentials_validate): declaration for new function.
 	(krb5_get_cred_from_kdc_validate): ditto.
 
 Thu May  2 22:52:14 1996  Mark Eichin  <eichin@cygnus.com>
@@ -15,6 +102,11 @@
 		the raw DES cryptosystem is properly included.  Removed
 		Macintosh definitions of PROVIDE_* since that is done in
 		mac/libaries/KerberosHeaders.h.
+
+Wed Apr 17 20:56:51 1996  Marc Horowitz  <marc@mit.edu>
+
+	* k5-int.h, port-sockets.h: moved socket stuff into a separate
+	file so that gssapi doesn't have to include k5-int.h
 
 Thu Apr 11 23:50:24 1996  Theodore Y. Ts'o  <tytso@dcl>