Final checkin for 0.10.0.

svn path=/trunk/; revision=9255

1 files changed, 2137 insertions, 2739 deletions

diff --git a/ChangeLog b/ChangeLog
index 71d7be0ecb..02ec0e2a82 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,4100 +1,3498 @@
-$Id: ChangeLog,v 1.32 2003/11/03 23:28:01 gerald Exp $
+$Id: ChangeLog,v 1.33 2003/12/13 00:55:10 gerald Exp $
 
-NOTE:  The format of this file has changed from previous versions.  
 
-2003-09-10 05:35  guy
+2003-12-13 00:38  gerald
 
-	* pcap-util.c, pcap-util.h, tethereal.c, gtk/Makefile.am,
-	  gtk/Makefile.nmake, gtk/capture_combo_utils.c,
-	  gtk/capture_combo_utils.h, gtk/capture_dlg.c,
-	  gtk/capture_prefs.c, gtk/main.c:
+	* Makefile.am:
 
-	  Have "get_interface_list()" return a list of "if_info_t"
-	  structures containing a pointer to an interface name and possibly
-	  a pointer to an interface description (although that pointer
-	  might be null if no description is available), rather than having
-	  the Windows version glue together the name and description into a
-	  single string.
+	  Add packet-isup.h as well.
 
-	  Supply for the Linux "any" device the same description that
-	  libpcap's "pcap_findalldevs()" returns.
+2003-12-13 00:22  gerald
 
-2003-09-10 06:47  guy
+	* Makefile.am:
 
-	* pcap-util.c:
-
-	  Fix a compile error on Windows.
-
-2003-09-10 06:48  guy
-
-	* epan/osi-utils.c:
-
-	  Void functions don't return values.
-
-2003-09-10 06:55  guy
-
-	* gtk/: capture_dlg.c, capture_prefs.c:
-
-	  If we have no interfaces, don't set the popdown list for
-	  interface combo boxes.
-
-2003-09-10 07:02  guy
-
-	* gtk/capture_dlg.c:
-
-	  If there's no saved interface name, set the interface name in the
-	  Capture Options dialog box to the first string in the combo box,
-	  not to the first interface name in the list, so we get the
-	  description.
-
-2003-09-10 07:17  jmayer
-
-	* AUTHORS, Makefile.am, configure.in, gtk/Makefile.am:
-
-	  Albert Chin:	  - AM_PROC_LIBTOOL is just an alias for
-	  AC_PROG_LIBTOOL, which is	     called earlier.	   - Use
-	  AM_CPPFLAGS instead of CFLAGS and CPPFLAGS to add inlude
-	    directories
-
-2003-09-10 09:49  sahlberg
-
-	* packet-dcerpc-netlogon.c:
+	  Add packet-ansi_map.h to the distribution.
 
-	  NETLOGON/DsrDeregisterDNSHostRecords : The dns_host field is a
-	  REF pointer to a unicode string.  NOT a Unique pointer.
+2003-12-12 23:47  obiot
 
-	  Update iethereal so that this packet is dissected properly
-
-2003-09-10 19:23  guy
-
-	* packet-teredo.c:
-
-	  From Albert Chin: fix yet another C++-comment problem.
-
-2003-09-10 19:27  guy
-
-	* packet-gprs-ns.c:
-
-	  Use the *correct* value_string for the IE Type field.
-
-2003-09-10 20:07  guy
-
-	* packet-lwapp.c:
-
-	  Don't use "alloca()", and don't use "match_strval()", just use
-	  "val_to_str()".
-
-2003-09-10 21:19  gerald
-
-	* epan/tvbuff.c:
+	* packet-smpp.c:
 
-	  Make tvb_find_tvb() return -1 if either tvb length is 0.
-	  Otherwise, an assertion happens 'xxxx contains ""'.
+	  Always dissect a sm_submit otherwise reassembly of short messages
+	  does not work.
 
-2003-09-10 22:23  guy
+2003-12-12 23:35  obiot
 
-	* file.c, print.h, tethereal.c, gtk/print_dlg.c:
+	* packet-wbxml.c:
 
-	  When saving packets, we have a "Save only marked packets" option;
-	  label the option to print only marked packets similarly, rather
-	  than as "Suppress unmarked packets" (for consistency, and because
-	  the latter isn't unlike a double negative).
+	  Move info column update code outside of if (tree).
 
-2003-09-10 22:33  guy
+2003-12-12 23:23  guy
 
-	* doc/ethereal.pod.template:
+	* packet-dns.c:
 
-	  Document the "Print only marked packets" option in the "Print"
-	  dialog, and document that you can now search for packets with
-	  particular text in the Info column or protocol tree.
+	  Add some checks to make sure we're not running past the end of
+	  the RR we're dissecting.  (Some RR's still don't do that check,
+	  as they don't maintain an "rr_len" variable, and as
+	  "get_dns_name()" doesn't do any length checking.)
 
-2003-09-10 23:55  guy
+	  That obviates the need for having "compute_key_id()" check the
+	  length, as we won't even get to calling it if the RR is too
+	  short.  Have it still do a check for sanity, but just abort if
+	  the check fails, and have it return -1 if the algorithm is
+	  unsupported.
 
-	* prefs.c, prefs.h:
+	  Don't bother processing T_CERT RRs if we're not building a
+	  protocol tree, don't fetch data from a T_CERT RR until we put it
+	  into the tree, and fix the bug that change found, namely that we
+	  were using the wrong variable in the "val_to_str()" call for the
+	  certificate type.
 
-	  Don't export MAX_VAR_LEN - nobody uses it.
+2003-12-12 22:19  guy
 
-2003-09-11 00:08  guy
+	* packet-ses.c, packet-ses.h:
 
-	* AUTHORS, packet-isup.c, doc/ethereal.pod.template:
+	  From Yuriy Sidelnikov: if the first SPDU in a TSDU has a type of
+	  1 or 2, treat it specially - a type of 1 means GIVE_TOKENS, not
+	  DATA_TRANSFER, in that case.
 
-	  From Anders Broman: add support for BICC dissection.
+2003-12-12 21:17  guy
 
-	  Change his e-mail address.
+	* packet-pim.c:
 
-2003-09-11 10:31  sahlberg
+	  From Hannes Gredler:
 
-	* packet-dcerpc.c:
+		  display the PIM hello option values in brackets after the
+	  option
+		  name;
 
-	  Do not take the pointer to a stack object and pass it to the tap
-	  system.  The tap listener will try to parse this pointer at a
-	  much later stage where the stack frame where this object lived
-	  will have dissapeared and possible got overwritten.
+		  in the LAN prune delay option dissector:
 
-	  best that can happen is that service response times for dcerpc
-	  interfaces is screwed up more probable is that we get a coredump
+			  put the top-level item for the option into the
+	  PIM
+			  options tree rather than the PIM main tree;
 
-2003-09-11 13:24  sahlberg
+			  calculate the correct LAN prune delay and display
+	  it in
+			  units of milliseconds;
 
-	* AUTHORS, packet-dcerpc-netlogon.c:
+			  calculate the correct override interval and
+	  display it
+			  in units of milliseconds;
 
-	  From Jean-Baptiste Marchand
+			  fix the offset and length calculation.
 
-	  update of the type of pointer used for DsrGetSiteName reply
-	  packets
+2003-12-12 19:55  guy
 
-2003-09-11 13:36  sahlberg
+	* AUTHORS, packet-ansi_a.c, packet-ansi_a.h, packet-ansi_map.c,
+	  packet-ansi_map.h, packet-isup.c, packet-isup.h, gtk/Makefile.am,
+	  gtk/Makefile.nmake, gtk/ansi_map_stat.c, gtk/isup_stat.c:
 
-	* AUTHORS, Makefile.am, Makefile.nmake, packet-t38.c:
+	  From Michael Lum:
 
-	  From Hans Viens : T38 dissector
+		  Fix point code display.
 
-2003-09-11 13:37  sahlberg
+		  Add taps for ANSI MAP and ISUP statistics.
 
-	* doc/ethereal.pod.template:
+2003-12-12 16:06  gerald
 
-	  From hans Viens: T38 dissector
+	* README.win32:
 
-2003-09-12 02:48  sahlberg
+	  Switch from gettext 0.13 to 0.12.1.
 
-	* file.c, file.h, packet-frame.c, doc/ethereal.pod.template,
-	  epan/column-utils.c, epan/frame_data.h, gtk/main.c, gtk/main.h,
-	  gtk/menu.c:
+2003-12-12 02:50  gerald
 
-	  Added TimeReference frames.
+	* gtk/: dlg_utils.c, main.c:
 
-	  One can now select a packet and mark it as a TimeReference packet
-	  using the menu.  A TimeReference packet will be indicated by
-	  having all timestamp related column entries replaced by the
-	  string *REF* A TimeReference packet will always be displayed in
-	  the packet pane, and overrides any display filters.
+	  In "dlg_window_new()" check to see if "top_level" has been
+	  initialized, as is the case for the spawned capture process.
 
-	  When a frame is a TimeReference frame, all later frames will
-	  calculate the TimeRelativeToFirstPacket  relative to the
-	  timestamp of the TimeReference frame instead of the first frame
-	  of the capture.
+2003-12-11 21:23  ulfl
 
-	  You can have any number of TimeReference frames you like.
+	* packet-afp.c, packet-bssap.c, packet-cops.c, packet-esis.c,
+	  packet-gsm_sms.c, packet-isakmp.c, packet-osi-options.c,
+	  gtk/dcerpc_stat.c, plugins/docsis/packet-tlv.c,
+	  plugins/docsis/packet-vendor.c, wiretap/erf.c,
+	  wiretap/lanalyzer.c:
 
-2003-09-12 04:52  sahlberg
+	  prevent MSVC warning: "warning C4761: Größenkonflikt im Argument.
+	  Konvertierung vorgenommen" -> size conflict in argument,
+	  conversion done
 
-	* packet-frame.c:
+2003-12-11 18:38  guy
 
-	  Attempt to prettify frame.time_relative and some other fields
+	* packet-dns.c:
 
-2003-09-12 05:52  sahlberg
+	  Have "compute_key_id()" return different values for different
+	  errors, and have its caller put an entry into the tree for "we
+	  don't support the algorithm" (the "RR too short" error is a
+	  "can't happen").
 
-	* packet-tcp.c, packet-tcp.h:
+2003-12-11 18:37  guy
 
-	  Enhancement to the TCP dissector: Track window scaling and
-	  display the window field after it has been scaled to its real
-	  value
+	* epan/configure.in:
 
-	  If we have seen a SYN packet with a WindowScalingOption then if
-	  the option to use RelativeSequence numbers has been enabled, then
-	  ethereal will change the presented window field to be the window
-	  after it has been scaled to the real value.
+	  Add "/usr/local/include" in case the PCRE header file is there.
 
-	  This obviously only works if we have seen the SYN packet and if
-	  the SYN packet contained a window scaling option
+2003-12-11 08:54  ulfl
 
-2003-09-12 10:32  sahlberg
+	* packet-dns.c:
 
-	* AUTHORS, packet-dcerpc-netlogon.c, packet-dcerpc-netlogon.h:
+	  replaced type u_int32_t by glib's guint32 (to satisfy MSVC)
 
-	  upadtes to NETLOGON
+2003-12-11 01:13  guy
 
-2003-09-12 11:13  sahlberg
+	* tethereal.c:
 
-	* packet-dcerpc-netlogon.c:
+	  Fix a typo.
 
-	  Update to netlogon
+2003-12-10 23:21  gerald
 
-	  make it decode DsrGetDCSiteCoverage replies properly
+	* README.win32, packaging/nsis/ethereal.nsi:
 
-2003-09-12 19:40  guy
+	  Update the Windows environment to GLib 2.2.3 and GTK+
+	  1.3.0-20030717.  This requires replacing libintl-1.dll from the
+	  libintl package (which is no longer used) with intl.dll from
+	  intl.dll from the gettext package.
 
-	* packet-cops.c:
+2003-12-10 22:52  gram
 
-	  From Kari Tiirikainen:
+	* print.c:
 
-		  Updated the COPS dissector to print the names of the
-	  Prefix
-		  PRIDs correctly.  The last fix to PRID OID printout
-	  caused this
-		  minor problem for PPRIDs.
+	  Analyzer's PDML will be changed to start the packet offset (the
+	  "pos" field) at 0 instead of 1. This works better for Ethereal,
+	  as our offsets start at 0. Change the "geninfo" pseudo-protocol
+	  to use pos="0" instead of pos="1".
 
-		  Rename a few variables to make the code more logical to
-	  read.
+2003-12-10 22:38  obiot
 
-2003-09-12 22:52  guy
+	* epan/dfilter/Makefile.nmake:
 
-	* packet-aodv.c:
+	  PCRE support requires $(PCRE_CFLAGS) in CFLAGS.
 
-	  According to draft-ietf-manet-aodv-13, "the only changes to the
-	  protocol are that the address fields are enlarged", so presumably
-	  that supercedes the now-expired draft-perkins-aodv6-01 which
-	  gives different type numbers for V6.	We support both - the
-	  draft-ietf-manet-aodv-13 messages are dissected based on the
-	  address type of the source address, while the
-	  draft-perkins-aodv6-01 messages are dissected as V6.
+2003-12-10 22:27  obiot
 
-	  Handle the Route Reply Acknowledgment message.
+	* tap-wspstat.c:
 
-2003-09-13 00:57  guy
+	  WSP status code is hexadecimal.
 
-	* packet-aodv.c:
+2003-12-10 21:50  gerald
 
-	  According to draft-ietf-manet-aodv-13, the only differences
-	  between V4 and V6 AODV messages is the addresses in the message.
+	* config.nmake, configure.in, epan/configure.in:
 
-	  According to the now-expired draft-perkins-aodv6-01, some of the
-	  messages differ in the order of the fields as well.
+	  Bump the version to 0.10.0.  It could be (and has been) argued
+	  that the last release should have been 0.10.0.
 
-	  Dissect the draft-ietf-manet-aodv-13 messages correctly.
+2003-12-10 21:21  guy
 
-2003-09-14 20:59  jmayer
+	* gtk/gsm_a_stat.c:
 
-	* packet-gtp.c:
+	  Oops, I checked the wrong tap into CVS somehow.
 
-	  Fix warning: guint8 is always <= 255
+2003-12-10 21:12  gerald
 
-2003-09-14 21:28  gerald
+	* epan/ftypes/: ftypes-int.h, ftypes.h:
 
-	* packet-gtp.c:
+	  Make the _pcre_tuple_t struct internal.
 
-	  Fix a buffer overflow in msisdn_to_str().  Clean up string
-	  pointer handling in it, imsi_to_str(), and id_to_str() a bit.
+2003-12-10 19:35  guy
 
-2003-09-15 18:31  guy
+	* packet-dns.c:
 
-	* packet-gtp.c:
+	  From David Fort: compute and display the key ID in a DNS KEY RR,
+	  if possible.
 
-	  Use "int"s as array indices - it squelches a
-	  signed-vs-unsigned-comparison complaint in one case, and it's a
-	  generally good habit.
+2003-12-10 19:26  guy
 
-2003-09-15 18:40  guy
+	* packet-isup.c:
 
-	* AUTHORS, packet-ieee80211.c, doc/ethereal.pod.template:
+	  From Anders Broman: add CIC to Info column.
 
-	  From Ritchie: dissect country information tag.
+2003-12-10 19:21  guy
 
-2003-09-15 18:49  guy
+	* AUTHORS, Makefile.am, Makefile.nmake, packet-igap.c,
+	  packet-igap.h, packet-igmp.c:
 
-	* packet-bssgp.c:
+	  From Akira Endoh: IGAP support.
 
-	  From Josef Korelus: call the data dissector on the contents of
-	  the LLC PDU information element.
+2003-12-10 19:21  guy
 
-2003-09-15 19:04  guy
+	* packet-isakmp.c:
 
-	* epan/filesystem.c, gtk/dcerpc_stat.c,
-	  gtk/endpoint_talkers_table.c, gtk/fc_stat.c, gtk/io_stat.c,
-	  gtk/mgcp_stat.c, gtk/rpc_stat.c, gtk/smb_stat.c:
+	  Squelch a compiler warning.
 
-	  From Graeme Hewson:
+2003-12-10 19:05  ulfl
 
-		  Don't crash if "-z" is used on the command line when a
-	  live
-		  capture is being done (e.g., with "-k -S" - in that case,
-		  "cfile.filename" is null when the window is created),
-	  just don't
-		  display the file name in the title.
+	* packaging/nsis/ethereal.nsi:
 
-		  Don't set the title of a non-existent window - create the
-	  window
-		  first.
+	  added some more uninstall info and offering only removal of
+	  program (instead of modify/repair)
 
-2003-09-15 19:08  guy
+2003-12-10 10:47  guy
 
 	* packet-isakmp.c:
 
-	  From Yaniv Kaul: add AES-XCBC-MAC to the list of authentication
-	  methods, per RFC 3566.
+	  Check whether the packet length is less than the length of the
+	  packet header, and give up if it is.
 
-2003-09-15 20:37  guy
+2003-12-09 23:45  obiot
 
-	* cfile.h, file.c, gtk/dcerpc_stat.c, gtk/endpoint_talkers_table.c,
-	  gtk/fc_stat.c, gtk/io_stat.c, gtk/mgcp_stat.c, gtk/rpc_stat.c,
-	  gtk/smb_stat.c:
+	* gtk/tcp_graph.c:
 
-	  Keep a "display name" as part of a capture_file structure; for
-	  live capture temporary files, it's "<capture", and for saved
-	  capture files, it's the last component of the pathname of the
-	  file.  Use that in various places when displaying the file name.
+	  Mark parameter data as unused in callback_create_help().
 
-2003-09-15 20:45  guy
+2003-12-09 23:15  obiot
 
-	* gtk/main.c:
+	* epan/ftypes/Makefile.nmake:
 
-	  Don't allow -k and a capture file to both be specified on the
-	  command line - either we're doing a live capture or we're reading
-	  a saved capture file, we can't do both at the same time.
+	  Internal PCRE field type for efficient RE parsing in dfilters.
 
-2003-09-15 22:16  guy
+2003-12-09 23:02  obiot
 
-	* cfile.h, file.c, file.h, gtk/dcerpc_stat.c,
-	  gtk/endpoint_talkers_table.c, gtk/fc_stat.c, gtk/io_stat.c,
-	  gtk/mgcp_stat.c, gtk/rpc_stat.c, gtk/smb_stat.c:
+	* epan/: ftypes/Makefile.am, ftypes/ftype-pcre.c,
+	  ftypes/ftype-string.c, ftypes/ftypes.c, ftypes/ftypes.h,
+	  dfilter/semcheck.c:
 
-	  Add a routine to return the display name for a "capture_file"
-	  structure, rather than constructing that name when a capture file
-	  is opened and putting a pointer to it in that structure.
+	  Internal PCRE field type for efficient RE parsing in dfilters.
 
-2003-09-15 22:17  guy
+2003-12-09 22:41  ulfl
 
-	* gtk/tcp_graph.c:
+	* file.c, gtk/main.h, gtk/menu.c:
 
-	  Get rid of a no-longer-relevant comment.
+	  print system enhanced, more print ranges and expanded states
 
-2003-09-15 22:32  guy
+2003-12-09 22:27  ulfl
 
-	* gtk/: endpoint_talkers_eth.c, endpoint_talkers_fc.c,
-	  endpoint_talkers_fddi.c, endpoint_talkers_ip.c,
-	  endpoint_talkers_ipx.c, endpoint_talkers_tcpip.c,
-	  endpoint_talkers_tr.c, endpoint_talkers_udpip.c:
+	* print.c:
 
-	  Get rid of a bunch of unnecessary #includes.
+	  print system enhanced, more print ranges and expanded states
 
-2003-09-15 22:48  guy
+2003-12-09 22:21  ulfl
 
-	* capture.c, file.c, file.h, tethereal.c, gtk/file_dlg.c,
-	  gtk/main.c:
-
-	  Rename various capture file routines to have names starting with
-	  "cf_".
+	* print.h, tethereal.c, gtk/print_dlg.c:
 
-2003-09-15 23:15  guy
+	  print system enhanced, more print ranges and expanded states
 
-	* capture.c, capture.h, gtk/main.c:
+2003-12-09 22:04  ulfl
 
-	  Don't pop up tap windows specified with "-z" options until we
-	  have a capture file open.
+	* print.h:
 
-2003-09-15 23:20  guy
-
-	* gtk/main.c:
+	  print system enhanced, more print ranges and expanded states
 
-	  Clean up an error message.
+2003-12-09 18:49  guy
 
-2003-09-15 23:28  guy
+	* Makefile.am, Makefile.nmake, packet-gsm_a.c, packet-gsm_a.h,
+	  tap-ansi_astat.c, tap-gsm_astat.c, gtk/Makefile.am,
+	  gtk/Makefile.nmake, gtk/ansi_a_stat.c, gtk/gsm_a_stat.c:
 
-	* file.c:
-
-	  We no longer pop up tap windows until we have a capture file
-	  open, so we don't have to worry about the file name being null
-	  when "cf_get_display_name()" is called.  (It should never be
-	  called before we have a capture file open; if it is, that's a
-	  bug.)
-
-2003-09-15 23:48  guy
-
-	* capture.c:
+	  From Michael Lum:
 
-	  Make sure we return FALSE in all the failure paths in
-	  "do_capture()".
+		  new taps for GSM A-interface;
 
-2003-09-16 17:36  guy
+		  fixes for ANSI A-interface taps.
 
-	* packet-distcc.c:
+2003-12-09 06:48  guy
 
-	  Re-register the dissector if a preference changes - otherwise,
-	  changing the TCP port number won't have any effect.
+	* column.c, column.h, file.c, ui_util.h, epan/column-utils.c,
+	  epan/column_info.h, gtk/main.c, gtk/packet_list.c:
 
-2003-09-16 17:42  guy
+	  Don't automatically size the columns - that's expensive in large
+	  captures, as it has to compute the width of an auto-resizing
+	  column in every row.	Just pick fixed widths for the columns (and
+	  tune the width of the "Protocol" column so that it's not narrower
+	  than the column title).
 
-	* packet-smtp.c:
+2003-12-09 05:56  gram
 
-	  There's no preference setting for the SMTP TCP port (and SMTP has
-	  a well-known port), so we remove the extra code to handle
-	  changing it - if it ever becomes useful to have the port number
-	  settable, we can put it back.
+	* tethereal.c, doc/README.xml-output, doc/tethereal.pod,
+	  tools/msnchat:
 
-2003-09-16 20:41  guy
+	  -Tpdml and -Tps force -V. Thus, -V is no longer required when
+	  using -Tpdml.  Eventually, -Tps will not force -V, and will print
+	  summaries when -V is not selected. However, work still has to be
+	  done there.
 
-	* gtk/tcp_graph.c:
+2003-12-09 05:10  guy
 
-	  Handle PPP with a pseudo-header.
+	* packet-radius.c:
 
-2003-09-17 15:58  jfoster
+	  Add a new attribute type RADIUS_IP6_ADDRESS for IPv6 addresses.
 
-	* packet-socks.c:
+	  Put in a note explaining the enum for those types.
 
-	  Fix recursive heap overflow problem.	When a socks connection was
-	  used to create a second socks connection the socks dissector
-	  would be called recursively until a heap overflow occurred.
+	  From Rui Carmo: add all 3GPP VSA's.
 
-2003-09-17 19:39  guy
+2003-12-09 05:06  guy
 
-	* gtk/: bootp_stat.c, dcerpc_stat.c, endpoint_talkers_eth.c,
-	  endpoint_talkers_fc.c, endpoint_talkers_fddi.c,
-	  endpoint_talkers_ip.c, endpoint_talkers_ipx.c,
-	  endpoint_talkers_tcpip.c, endpoint_talkers_tr.c,
-	  endpoint_talkers_udpip.c, fc_stat.c, http_stat.c, io_stat.c,
-	  menu.c, mgcp_stat.c, rpc_progs.c, rpc_stat.c, smb_stat.c,
-	  tap_rtp.c, wsp_stat.c:
+	* epan/: resolv.c, to_str.c, to_str.h:
 
-	  Have "register_tap_menu_item()" put the item under Tools, not
-	  Tools/Statistics; change the taps to put things under Statistics
-	  themselves.  That allows taps to go elsewhere if appropriate.
+	  Add "ip6_to_str_buf()", by analogy to "ip_to_str_buf()".
 
-2003-09-17 20:24  guy
+2003-12-09 00:52  guy
 
-	* packet-smpp.c:
+	* packet-ansi_637.c:
 
-	  From Olivier Biot: put a summary into the top-level SMPP item in
-	  the protocol tree.
+	  From Michael Lum: fix incorrect switch values which resulted in
+	  NULL strings being passed to ..._add_text().
 
-2003-09-18 19:19  guy
+2003-12-09 00:12  guy
 
 	* packet-tcp.c:
 
-	  Don't store the setting of the window scale option unless the
-	  "Relative sequence numbers and window scaling" option is set, as
-	  that option says it controls whether we attempt to display the
-	  real post-scaling window size.
+	  As per Didier Gautheron, dissect TCP options regardless of
+	  whether we're building a protocol tree, so the Info column is set
+	  correctly.
 
-	  Also, don't store it unless the "Analyze TCP sequence numbers"
-	  option is set, as "Relative sequence numbers and window scaling"
-	  requires it, because, unless "Analyze TCP sequence numbers" is
-	  set, we don't set up conversations for TCP connections and don't
-	  have a pool of data structures for per-connection information
-	  into which to store the window scale option value.
+2003-12-08 23:40  guy
 
-2003-09-18 21:38  gerald
+	* packet-alcap.c, packet-ansi_637.c, packet-ansi_683.c,
+	  packet-ansi_a.c, packet-ansi_map.c, packet-gsm_a.c,
+	  packet-gsm_map.c, packet-gsm_sms.c, epan/to_str.c, epan/to_str.h:
 
-	* packet-gtp.c:
+	  Get rid of the private "my_decode_bitfield_value()" a number of
+	  dissectors had.  Instead, rename it
+	  "other_decode_bitfield_value()", put it in "epan/to_str.c", and
+	  make "decode_bitfield_value()" use it.
 
-	  From Michal Melerowicz: Properly terminate the return string in
-	  msisdn_to_str().
+2003-12-08 22:28  guy
 
-	  Do the same for imsi_to_str().
+	* Makefile.nmake:
 
-2003-09-18 21:42  gerald
+	  Add a "doc" target, as per Ulf Lamping's comment.
 
-	* packet-gtp.c:
+2003-12-08 22:24  guy
 
-	  Make the string variables in the *_to_str() functions non-static.
-
-2003-09-19 04:08  guy
-
-	* wiretap/erf.c:
-
-	  From Jesper Peterson: add support for AAL5 records in ERF files,
-	  treating them as PDUs with an AAL5 trailer.
+	* doc/ethereal-filter.pod.template:
 
-2003-09-19 04:16  guy
+	  Note the types of fields for which the "contains" operator is
+	  available.
 
-	* AUTHORS, Makefile.am, Makefile.nmake, packet-rtp-events.c,
-	  packet-rtp-events.h, doc/ethereal.pod.template:
+	  Always capitalize the names "Ethereal" and "Tethereal" (we don't
+	  capitalize the command names, however, as they're
+	  all-lower-case).
 
-	  From Kevin A. Noll: RFC 2833 RTP Events support.
+	  Note that you can find out from the GUI whether Ethereal was
+	  built with the PCRE library or not.
 
-2003-09-19 04:27  guy
+	  Fix a typo.
 
-	* packet-rtp-events.c:
+2003-12-08 21:57  guy
 
-	  Use bitfields for the end-of-event, reserved, and volume fields.
+	* print.c:
 
-	  Don't fetch fields until you need them, so you don't throw an
-	  early exception.
+	  "proto_construct_dfilter_string()" can return a null pointer,
+	  e.g. for protocol tree fields that don't have values and that are
+	  zero-length.
 
-	  Give the fields names starting with the name of the protocol.
+	  Combine the "if (fi->hfinfo->type != FT_PROTOCOL && fi->length >
+	  0)" stuff to show the raw value into the previous "if".
 
-2003-09-19 04:52  guy
+2003-12-08 21:36  guy
 
-	* file.c, tethereal.c:
+	* AUTHORS, packet-mtp3.c, packet-mtp3.h, packet-sccp.c,
+	  epan/packet_info.h, epan/to_str.c:
 
-	  From Lars Roland: initialize the flags.ref_time field of a
-	  frame_data structure.
+	  From Michael Lum: support an AT_SS7PC address type for SS7 point
+	  codes, and set the direction in pinfo for SS7 packets based on
+	  source and destination addresses.
 
-2003-09-19 05:30  tpot
+2003-12-08 20:58  guy
 
 	* packet-dcerpc.c:
 
-	  Put the presentation context list and transfer syntax list of a
-	  dcerpc bind request into some subtrees to make things look a bit
-	  nicer.
-
-2003-09-19 07:24  guy
-
-	* gtk/: bootp_stat.c, dcerpc_stat.c, endpoint_talkers_eth.c,
-	  endpoint_talkers_fc.c, endpoint_talkers_fddi.c,
-	  endpoint_talkers_ip.c, endpoint_talkers_ipx.c,
-	  endpoint_talkers_tcpip.c, endpoint_talkers_tr.c,
-	  endpoint_talkers_udpip.c, fc_stat.c, http_stat.c, io_stat.c,
-	  menu.c, menu.h, mgcp_stat.c, rpc_progs.c, rpc_stat.c, smb_stat.c,
-	  tap_rtp.c, wsp_stat.c:
-
-	  Provide a mechanism by which a tap can arrange to have its menu
-	  enabled or disabled based on a currently-selected packet (or lack
-	  of same) or a currently-selected protocol tree item (or lack of
-	  same).  Not currently used, but necessary if we ever make the TCP
-	  graph stuff a tap.  API is tentative and subject to change.
+	  From Jamie Fournier: don't show UNKUUID information for fragments
+	  for known dissectors.
 
-	  Also, enable and disable taps based on whether we have any
-	  packets to process.
+2003-12-08 20:50  guy
 
-2003-09-20 03:20  guy
-
-	* packet-ieee80211.c:
-
-	  Give the tag number a value_string table, and have common code,
-	  rather than tag-specific code, put it, and the tag length, into
-	  the protocol tree.  Make those fields FT_UINT8, as they're one
-	  byte long.
-
-	  Don't pass to "dissect_vendor_specific_ie()" variables whose
-	  names and values are the same as global variables.
-
-2003-09-20 03:31  guy
-
-	* packet-osi.c:
-
-	  From Akira Endoh: handle the NLPID for PPP.
-
-2003-09-20 03:48  guy
-
-	* packet-ieee80211.c:
-
-	  From Richie:
-
-		  the units for output power are dBm, not dbi;
-
-		  decode multiple start channel/number of channels/output
-	  power
-		  combinations;
-
-		  display the output power as a signed value.
-
-	  Additional change from Martijn Schipper to fix the display of the
-	  output power.
-
-	  Pass the tagged data value pointer to
-	  "dissect_vendor_specific_ie()".
-
-	  Don't special-case unknown tag values - use "val_to_str()" even
-	  for unknown tag values, and handle the value in the "default:"
-	  case.
-
-2003-09-20 04:25  guy
+	* packet-srvloc.c:
 
-	* packet-ncp-int.h, packet-ncp2222.inc:
+	  From Greg Morris: ecode the attribute structure instead of
+	  reporting a really long undecoded string.
 
-	  From Greg Morris: update the Info column even if we're not
-	  building a protocol tree.
+2003-12-08 20:43  guy
 
-2003-09-20 04:59  guy
+	* packaging/nsis/ethereal.nsi:
 
-	* gtk/: dlg_utils.c, file_dlg.c:
+	  From Lars Roland: fix a typo.
 
-	  From Giles Scott:
+2003-12-08 20:42  guy
 
-		  in GTK+ 2.x, center dialogs on the parent;
+	* packaging/nsis/Makefile.nmake:
 
-		  make the file selection dialogs transient for the main
-	  window,
-		  just as other dialogs are.
+	  From Lars Roland: fix the NSIS Makefile to handle PCRE.
 
-	  Update Gerald's e-mail address.
+2003-12-08 20:40  guy
 
-2003-09-20 08:56  guy
+	* packet-isis-clv.c, packet-isis-clv.h, packet-isis-hello.c,
+	  packet-isis-hello.h, packet-isis-lsp.c, packet-isis-lsp.h,
+	  packet-isis-snp.c, packet-isis-snp.h:
 
-	* packet-gtp.c:
+	  From Hannes Gredler:
 
-	  You can't safely return from a routine a pointer to an automatic
-	  variable.
+	  - unify the IS-IS CLV space	IS-IS uses a unified CLV space
+	  across all Level and PDU Types   there is no need to define PDU-
+	  and Level- Specific CLV   Types;
 
-2003-09-20 08:56  guy
+	  - clean up Authentication CLV   TLV #10 is the official supported
+	  TLV for carrying authentication   information- todays code
+	  displays TLV #10 as non-standard which   is wrong;   also the
+	  notion of "Domain" "Authentication" and "Link" password   has
+	  disappeared from contemporary routing SW;
 
-	* AUTHORS, Makefile.am, Makefile.nmake, packet-dcerpc-drsuapi.c,
-	  packet-dcerpc-drsuapi.h:
+	  - add IP Authentication CLV dissector   this CLV is depreciated -
+	  however it is using different semantics   than TLV #10 so we need
+	  a dedicated dissector;
 
-	  From Jean-Baptiste Marchand: initial DRSUAPI (Active Directory
-	  directory replication) support.
+	  - add Checksum CLV dissector	 add support for RFC 3358 Checksum
+	  CLVs	 majority of code re-used from the LSP checksum
+	  verification	 dissector.
 
-2003-09-20 09:41  guy
+2003-12-08 20:37  obiot
 
-	* AUTHORS, Makefile.am, packet-tacacs.c, packet-tacacs.h,
-	  doc/ethereal.pod.template:
+	* packet-wsp.c, packet-mmse.c, packet-wbxml.c:
 
-	  From Emanuele Caratti: full TACACS+ dissection.
+	  Get rid of the integer media type dissector table in WSP, WBXML
+	  and MMSE.
 
-2003-09-20 09:54  guy
+2003-12-08 20:36  guy
 
-	* packet-tacacs.c:
+	* packet-dsi.c:
 
-	  Arrange to free the buffer allocated for the decrypted tvbuff.
+	  From Didier Gautheron: check packet format in get_dsi_pdu_len.
 
-2003-09-20 10:06  guy
+2003-12-08 20:36  guy
 
-	* configure.in:
+	* packet-afp.c:
 
-	  Put back the AM_PROG_LIBTOOL call, as some older versions of the
-	  auto* tools and/or libtool require it.
+	  From Didier Gautheron: add offset and size parameters in info
+	  colum for read and write commands.
 
-2003-09-21 20:05  gerald
+2003-12-08 20:25  guy
 
-	* packet-atalk.c, packet-bpdu.c, packet-gtp.c, packet-q931.c,
-	  packet-quake2.c, packet-quake3.c, packet-quakeworld.c,
-	  packet-rx.c, packet-sctp.c, packet-x11.c, gtk/mgcp_stat.c:
+	* packet-sip.c:
 
-	  Remove a bunch of duplicate semicolons.
+	  From Anders Broman:
 
-2003-09-22 00:59  tpot
+		  Call subdissectors regardless of whether we're building a
+		  protocol tree or not.
 
-	* packet-dcerpc-netlogon.c:
+		  Use the "media_type" dissector table to hand off to
+		  subdissectors, rather than wiring in SDP.
 
-	  A bunch of fixups for the NETLOGON sam database synchronisation
-	  dissector:
+		  Put the message body into a subtree of the SIP tree
+	  rather than
+		  at the top level.
 
-	    - some rids marked as user are actually groups
+2003-12-08 19:25  ulfl
 
-	    - decorate some strings higher up in the tree
+	* gtk/menu.c:
 
-	    - dissect the delta id as such and add value string text item
+	  moved "File->Save Highlighted Data" to "File->Export->Raw Packet
+	  Data"
 
-2003-09-22 09:06  sahlberg
+2003-12-08 17:54  ulfl
 
-	* packet-frame.c:
+	* doc/ethereal.pod:
 
-	  Add filterable field for REF TIME frames so it is easy to search
-	  for them
+	  updated to reflect the menu redesign changes
 
-2003-09-22 09:08  sahlberg
+2003-12-08 17:49  ulfl
 
 	* gtk/menu.c:
 
-	  Update the menu text for Time Reference  and give it a hot-key :
-	  ctrl-T
+	  removed seperator between the two print menuitems
 
-2003-09-23 02:35  guy
+2003-12-07 19:09  obiot
 
-	* packet-eapol.c, packet-ieee80211.c, packet-ieee80211.h:
-
-	  Get rid of some unnecessary includes.
+	* AUTHORS:
 
-	  Instead of using "dissect_vendor_specific_ie()" to dissect the
-	  WPA key data, turn the code to dissect a collection of 802.11
-	  tagged parameters into a routine, use that routine when
-	  dissecting 802.11 management frames, and also use it to dissect
-	  the key data.
+	  Update address of Johan Feyaerts. Dfilter matches operator.
 
-2003-09-23 03:18  guy
+2003-12-07 18:32  obiot
 
-	* packet-tacacs.c:
+	* doc/ethereal-filter.pod.template:
 
-	  From Emanuele Caratti: fix a leak in md5_xor.
+	  Clarify dfilter syntax (mainly the implicit exists operator), add
+	  subsections in syntax section, add description of new "contains"
+	  operator (and inform that it is only available if libpcre has
+	  been linked in).
 
-2003-09-23 06:25  oabad
+2003-12-07 18:09  obiot
 
-	* gtk/main.c:
+	* packet-mmse.c, packet-wbxml.c, packet-wsp.c:
 
-	  Temporary patch for the packet_list column resize problem in
-	  gtk2.
+	  Remove the literal WSP content type dissector table, and replace
+	  it with the "media_type" table defined in the HTTP dissector.
 
-	  We should use the same test (event_button->window ==
-	  GTK_CLIST(w)->clist_window) we already use in gtk v1.2 code but
-	  in doesn't seem to work in gtk2.
+2003-12-07 09:43  oabad
 
-2003-09-23 12:06  sahlberg
+	* gtk/main.c:
 
-	* packet-dcerpc-lsa.c, packet-dcerpc-lsa.h, packet-dcerpc-samr.c,
-	  packet-dcerpc-samr.h:
+	  Update various set_menu_object_data() calls to reflect recent
+	  menu hierarchy changes. This avoids a crash when selecting "Folow
+	  TCP stream" from the "Analyze" menu.
 
-	  From JBM Update all names for SAMR and LSA functions so they
-	  match the function names in the API
+2003-12-07 03:46  guy
 
-2003-09-23 18:09  guy
+	* packet-sdp.c:
 
-	* gtk/dfilter_expr_dlg.c:
+	  Register the SDP dissector with the media_type dissector table
+	  with its media type.
 
-	  The relation list is always displayed (that's been true for a
-	  while) - we don't need to support making it visible or not
-	  depending on the selected field.
+2003-12-07 03:34  guy
 
-2003-09-23 18:22  guy
+	* packet-http.c:
 
-	* packet-ncp2222.inc:
+	  Use a cleanup handler to free the Content-Type header value, so
+	  it gets freed even if we throw an exception after processing the
+	  Content-Type header but before we call a subdissector.
 
-	  From Greg Morris: properly initialize request_value->info_string
-	  using "strcpy()", not "strcat()".
+2003-12-07 03:21  guy
 
-2003-09-23 18:36  guy
+	* packet-http.c:
 
-	* packet-per.c:
+	  Change the name of the variable holding the dissector table
+	  pointer to match the previous checkin.
 
-	  From Graeme Reid: modified PER choice dissector to correctly
-	  handle numbered choices starting with non-zero values (eg X.880
-	  ROS).
+2003-12-07 03:17  guy
 
-2003-09-23 18:40  guy
+	* packet-http.c, packet-mmse.c:
 
-	* AUTHORS, Makefile.am, Makefile.nmake, packet-h225.c,
-	  packet-h225.h, doc/ethereal.pod.template:
+	  Rename the "http.content_type" dissector table to "media_type",
+	  as it's for use in any dissector that uses Internet media types.
 
-	  From Graeme Reid: H.450 support.
+2003-12-07 02:42  guy
 
-2003-09-23 18:59  guy
+	* packet-nt-sonmp.c:
 
-	* packet-h450.c:
+	  Only check the destination address if it's AT_ETHER; otherwise,
+	  there's no guarantee that "data[5]" exists or can be interpreted
+	  in that fashion.
 
-	  From Graeme Reid: H.450 support.
+2003-12-07 02:26  guy
 
-2003-09-23 21:37  guy
+	* packet-ssl.c:
 
-	* packet-tacacs.c:
+	  From Devin Heitmueller: include the AES ciphersuites from RFC
+	  3268.
 
-	  As doc/README.developer's "Portability" section says:
+2003-12-07 02:23  guy
 
-		  Don't use "bzero()", "bcopy()", or "bcmp()"; instead, use
-	  the
-		  ANSI C routines
+	* doc/tethereal.pod:
 
-			  [memset, memcpy, memcmp] ...
+	  Put the "-V" flag in boldface, so pod2man doesn't whine.
 
-		  Not all platforms necessarily have
-	  "bzero()"/"bcopy()"/"bcmp()",
-		  and those that do might not declare them in the header
-	  file on
-		  which they're declared on your platform.
+2003-12-07 00:04  gerald
 
-	  Make it so.
+	* Makefile.nmake, README.win32, config.h.win32, config.nmake,
+	  epan/Makefile.nmake, epan/config.h.win32,
+	  epan/ftypes/Makefile.nmake, packaging/nsis/ethereal.nsi:
 
-2003-09-23 22:04  guy
+	  Add PCRE support to the Windows build.
 
-	* packet-ncp2222.inc:
+2003-12-06 19:26  oabad
 
-	  From Greg Morris: get the request data before using it, if we
-	  don't already have it.
+	* packet-nt-sonmp.c:
 
-2003-09-24 00:47  guy
+	  Check if pinfo->dl_dst.data is not NULL before accessing the data
+	  (it was NULL in one of my "linux cooked" captures which contains
+	  no destination address).
 
-	* cfile.h, file.c, file.h, globals.h, gtk/main.c, gtk/menu.c,
-	  gtk/prefs_dlg.c, gtk/proto_draw.c:
+2003-12-06 19:14  jmayer
 
-	  Make "finfo_selected" a member of a "capture_file" structure
-	  rather than an independent global variable.
+	* packet-mtp3mg.c:
 
-2003-09-24 02:36  guy
+	  Declaration included the line above
 
-	* file.c, menu.h, gtk/bootp_stat.c, gtk/dcerpc_stat.c,
-	  gtk/endpoint_talkers_eth.c, gtk/endpoint_talkers_fc.c,
-	  gtk/endpoint_talkers_fddi.c, gtk/endpoint_talkers_ip.c,
-	  gtk/endpoint_talkers_ipx.c, gtk/endpoint_talkers_tcpip.c,
-	  gtk/endpoint_talkers_tr.c, gtk/endpoint_talkers_udpip.c,
-	  gtk/fc_stat.c, gtk/http_stat.c, gtk/io_stat.c, gtk/main.c,
-	  gtk/menu.c, gtk/menu.h, gtk/mgcp_stat.c, gtk/rpc_progs.c,
-	  gtk/rpc_stat.c, gtk/smb_stat.c, gtk/wsp_stat.c:
+2003-12-06 16:35  gram
 
-	  Pass a pointer to a "capture_file" structure to
-	  "set_menus_for_selected_packet()" and
-	  "set_menus_for_selected_tree_row()", and have them decide whether
-	  to enable or disable menu items based on whether that structure
-	  indicates that a packet or field is selected and, if one is, on
-	  its properties.
+	* Makefile.am, acinclude.m4, configure.in, util.c,
+	  epan/acinclude.m4, epan/configure.in, epan/dfilter/dfvm.c,
+	  epan/dfilter/dfvm.h, epan/dfilter/gencode.c,
+	  epan/dfilter/grammar.lemon, epan/dfilter/scanner.l,
+	  epan/dfilter/semcheck.c, epan/dfilter/sttype-test.c,
+	  epan/dfilter/sttype-test.h, epan/ftypes/ftype-bytes.c,
+	  epan/ftypes/ftype-double.c, epan/ftypes/ftype-integer.c,
+	  epan/ftypes/ftype-ipv4.c, epan/ftypes/ftype-none.c,
+	  epan/ftypes/ftype-string.c, epan/ftypes/ftype-time.c,
+	  epan/ftypes/ftype-tvbuff.c, epan/ftypes/ftypes.c,
+	  epan/ftypes/ftypes.h:
 
-	  Pass to the "selected packet enabled" routine for a menu item the
-	  "frame_data" and "edt" members of the "capture_file" structure,
-	  and pass to the "selected tree row enabled" routine the
-	  "field_info" member of that structure.
+	  From Olivier Biot
 
-	  Clear "cf->current_frame" if no packet is selected.
+	  New "matches" operater in display filter language. Uses PCRE.
 
-2003-09-24 03:34  guy
+	  If a "matches" operator is found in a dfilter while libpcre has
+	  not been used to build the binary, then an exception is thrown
+	  after using dfilter_fail() to set an apporporiate error message.
 
-	* packet-ncp-int.h, packet-ncp2222.inc:
+2003-12-06 06:22  gram
 
-	  Create a protocol tree if we don't have one, we're constructing
-	  the Info column, and we need stuff from the protocol tree for the
-	  Info column.
+	* epan/Makefile.am:
 
-	  Go back to the previous scheme for constructing the Info column;
-	  the previous change fixes the problems for which the Info column
-	  changes were fixes.
+	  slab.h needs to be in libethereal_a_SOURCES
 
-2003-09-24 06:15  oabad
+2003-12-06 06:09  gram
 
-	* gtk/: Makefile.am, gtkclist.c, gtkclist.h, gtkclist_v12.c,
-	  gtkclist_v12.h:
+	* Makefile.am, packet-frame.c, packet-frame.h, print.c, print.h,
+	  tethereal.c, doc/README.xml-output, doc/tethereal.pod,
+	  epan/proto.c, epan/proto.h, tools/EtherealXML.py,
+	  tools/Makefile.am, tools/msnchat:
 
-	  Renamed gktclist.[ch] to gtkclist_v12.[ch] to avoid conflicts
-	  with the real gtkclist.h file when building the gtk+ v2 gui.
+	  Add the ability to print packet dissections in PDML (an XML-based
+	  format) to tethereal. It could be added to Ethereal, but the GUI
+	  changes to allow the user to select PDML as a print format have
+	  not been added.
 
-2003-09-24 06:18  oabad
+	  Provide a python module (EtherealXML.py) to help parse PDML.
 
-	* gtk/main.c:
+	  Provide a sample app (msnchat) which uses tethereal and
+	  EtherealXML.py to reconstruct MSN Chat sessions from packet
+	  capture files. It produces a nice HTML report of the chat
+	  sessions.
 
-	  In packet_list_button_pressed_cb (gtk2 version) : - put back the
-	  event_button->window == GTK_CLIST(w)->clist_window test   now
-	  that we use the correct structure definition for GtkCList (from
-	  the right include file, not from our version of gtkclist.h).
+	  Document tethereal's PDML and EtherealXML.py usage in
+	  doc/README.xml-output
 
-2003-09-24 07:48  guy
+	  Update tethereal's manpage to reflect the new [-T pdml|ps|text]
+	  option
 
-	* AUTHORS, doc/ethereal.pod.template, gtk/Makefile.am,
-	  gtk/Makefile.nmake, gtk/rtp_analysis.c, gtk/rtp_analysis.h,
-	  gtk/rtp_stream.c, gtk/rtp_stream.h, gtk/rtp_stream_dlg.c,
-	  gtk/rtp_stream_dlg.h, gtk/tap_rtp.c:
+2003-12-06 04:05  guy
 
-	  From Lars Ruoff: rewritten RTP analysis module.
+	* file.c, ui_util.h, gtk/packet_list.c:
 
-2003-09-24 08:05  guy
+	  From Jeff Morriss: avoid at least some N^2 behavior when changing
+	  the time stamp format.
 
-	* AUTHORS, Makefile.am, packet-dcerpc-dcom.h, packet-dcerpc-oxid.c:
+2003-12-05 22:33  guy
 
-	  From Yaniv Kaul: DCERPC OXID operation #5 dissection.
+	* packet-ccsds.c:
 
-2003-09-24 08:43  guy
+	  "hf_ccsds_secheader" is an FT_BOOLEAN; use
+	  "proto_tree_add_boolean()" for it.
 
-	* gtk/: capture_dlg.c, dlg_utils.c, dlg_utils.h, file_dlg.c,
-	  print_dlg.c:
+2003-12-05 16:00  gerald
 
-	  Add a "file_selection_new()" routine that does all the
-	  positioning (GTK+ 2.x) and transient-for setting that's done for
-	  other dialogs, and use it for dialogs that come from the main
-	  window or from children of the main window.
+	* Makefile.nmake, README.win32, config.nmake:
 
-2003-09-24 18:35  guy
+	  Switch over to Net-SNMP 5.1.	A ZIP file of the Net-SNMP 5.1
+	  sources plus a static Windows library (netsnmp.lib) has been
+	  placed at
 
-	* epan/plugins.c, plugins/plugin_api.h, plugins/plugin_api_list.c,
-	  plugins/Xass-list, plugins/Xplugin_api.c, plugins/Xplugin_api.h,
-	  plugins/Xplugin_api_decls.h, plugins/Xplugin_table.h:
+	      http://www.ethereal.com/distribution/win32/development/
 
-	  From Tomas Kukosa:
+	  The Net-SNMP documentation recommends against using a DLL at the
+	  present time.
 
-		  Remove the internal packet-per.c functions
-		  "dissect_per_length_determinant()" and
-		  "dissect_per_normally_small_nonnegative_whole_number()"
-	  from the
-		  plugin API, as they shouldn't be used outside the PER
-	  dissector.
+2003-12-05 09:34  guy
 
-		  Remove the H.225/H.245 functions
-		  "dissect_h225_TransportAddress()" and
-		  "dissect_h245_NonStandardParameter()" from the plugin API
-	  until
-		  we really need them in plugins.
+	* packet-sdp.c:
 
-		  Add the string dissector table functions to the plugin
-	  API.
+	  We have to parse the SDP packet even if we're not building a
+	  protocol tree, as we might be setting up a conversation.
 
-2003-09-24 18:41  guy
+2003-12-05 09:33  guy
 
-	* plugins/megaco/packet-megaco.c:
+	* packet-isup.c:
 
-	  From Anders Broman: fix a crash, and fix trailing whitespace on
-	  Transaction ID.
+	  Subdissectors must be called regardless of whether a protocol
+	  tree is being built.
 
-2003-09-24 23:35  guy
+2003-12-05 09:25  guy
 
-	* packet-ieee80211.c, packet-lwapp.c:
+	* plugins/mgcp/packet-mgcp.c:
 
-	  From David Frascone: have an 802.11 dissector that byte-swaps the
-	  frame control field, and have a preference in the LWAPP dissector
-	  to specify whether to use it or the regular 802.11 dissector, as
-	  some hardware sends out LWAPP-encapsulated 802.11 packets with a
-	  byte-swapped FC field.
+	  Subdissectors must be called regardless of whether a protocol
+	  tree is being built.
 
-2003-09-24 23:53  guy
+2003-12-04 22:37  guy
 
-	* wiretap/README.developer:
+	* plugins/asn1/Makefile.am:
 
-	  From packet steve: update to give the correct name for
-	  "file_access.c", and add a discussion of how data_offset works.
+	  From some reason, at least on Mac OS X 10.3, the change to have
+	  the protocol tree stuff not use GNodes means that the ASN.1
+	  plugin now needs to be explicitly linked against GLib in order to
+	  get access to the GNode routines, otherwise you get errors from
+	  the run-time linker.
 
-2003-09-25 00:08  guy
+2003-12-04 19:53  guy
 
-	* tethereal.c:
+	* epan/: proto.c, proto.h:
 
-	  From Matthijs Melchior: check whether the ring buffer timeout has
-	  expired *before* writing a packet, rather than *after* writing a
-	  packet, so that if you get no packets for a sufficiently long
-	  period that the timeout expires before you get a new packet, the
-	  new packet is in the beginning of a new file (as you might get
-	  more packets right after that, and want them to be in the new
-	  file, rather than have the first packet at the end of one file
-	  and the rest of the packets in another file).
+	  It turns out the protocol tree parent link was used, so we need
+	  it in our new scheme.
 
-2003-09-25 00:37  guy
+2003-12-04 10:59  guy
 
-	* gtk/tcp_graph.c:
+	* file.c, print.c, proto_hier_stats.c, tap-protohierstat.c,
+	  epan/proto.c, epan/proto.h, gtk/proto_draw.c, gtk/rtp_analysis.c:
 
-	  From Laurent Rabret:
+	  Don't use GNodes for the protocol tree, put the sibling pointer,
+	  and pointers to the first *and* last child, in the "proto_node"
+	  structure itself.  That saves us one level of indirection and
+	  memory allocation, and lets us append to a tree by appending to
+	  the last child directly, rather than having to scan through the
+	  list of siblings of the first child to find the end of that list.
 
-		  handle 802.1Q frames;
+2003-12-04 08:15  sahlberg
 
-		  catch the destroy signal on the main Ethereal window and
-	  destroy
-		  our windows (avoids a crash).
+	* packet-kerberos.c:
 
-	  Get the PPP type value for IP from "ppptypes.h" rather than
-	  defining it ourselves.
+	  update to kerberos
 
-2003-09-25 01:50  tpot
+	  if the kerberos blob contains a krb error, print the error string
+	  to COL_INFO
 
-	* reassemble.c:
+2003-12-04 08:13  sahlberg
 
-	  When showing a fragment subtree, put spaces after colons to make
-	  things look a bit nicer.  Also separate frame and payload data by
-	  a comma.
+	* packet-ldap.c:
 
-2003-09-25 08:20  guy
+	  update to ldap.
 
-	* file.c:
+	  some implementations specify GSSAPI in the bind call.  the
+	  encapsulation seems to be the same as GSS-SPNEGO so handle it the
+	  same way
 
-	  If the currently selected frame doesn't pass the display filter,
-	  select the closest frame to that frame that did pass the display
-	  filter, if any did.
+2003-12-04 06:02  gram
 
-2003-09-25 08:31  guy
+	* packet-wsp.c:
 
-	* file.c:
+	  Add {0, NULL} to the vals_wsp_reason_codes value_string array.
 
-	  Correctly handle the case where the selected frame doesn't pass
-	  the filter and no frames after it pass the filter either.
+2003-12-04 05:59  gram
 
-2003-09-25 11:27  guy
+	* packet-wlancap.c:
 
-	* packet-ncp2222.inc:
+	  Add {0, NULL} to all the value_string arrays.
 
-	  In "nds_defrag()", handle "request_value->ncp_rec" being null.
+2003-12-04 05:57  gram
 
-	  In "dissect_nds_request()", insert the request information into
-	  the hash table the first time we see the packet, regardless of
-	  whether we created a new conversation or not.
+	* packet-socks.c:
 
-2003-09-25 19:35  guy
+	  Add {0, NULL} to the reply_table_v5 value_string array.
 
-	* rtp_pt.h:
+2003-12-04 05:55  gram
 
-	  Add PT_CN_OLD, so the RTP analysis code can get the RTP payload
-	  types it needs from this header.
+	* packet-ldp.c:
 
-2003-09-25 19:35  guy
+	  ldp.msg.tlv.returned.msg.ubit needs to be an FT_BOOLEAN field,
+	  not an FT_UINT8, since it is using a true_false_string struct to
+	  print its values.  Add {0, NULL} to the fec_vc_interfaceparm
+	  value_string array.
 
-	* gtk/: rtp_analysis.c, rtp_stream.c:
+2003-12-04 05:47  gram
 
-	  RTP analysis updates from Lars Ruoff:
+	* packet-isup.c:
 
-		  - can now handle streams with different payload types
-		  - detects payload changes
-		  - detects comfort noise (PT=13 and 19)
-		  - status line now shows: sequence errors, payload
-	  changes,
-		    comfort noise (if any)
-		  - uses colours for lines with status != "Ok"
-		  - new button "next": jumps to next line with status !=
-	  "Ok"
-		    (starting from selected line)
-		  - fixed: wrong jitter calculation (bug from tap_rtp)
-		  - fixed: marker was not shown on first packet or
-	  erroneous
-		    packets (bug from tap_rtp)
-		  - code refactored to improve readability and reuse
+	  Add {0, NULL} to the isup_Pass_on_not_possible_indicator_vals and
+	  ISUP_Broadband_narrowband_interworking_indicator_vals
+	  value_string arrays.
 
-2003-09-26 02:09  guy
+2003-12-04 05:40  gram
 
-	* gtk/: bootp_stat.c, dcerpc_stat.c, fc_stat.c, http_stat.c,
-	  io_stat.c, mgcp_stat.c, rpc_stat.c, rtp_analysis.c,
-	  rtp_stream_dlg.c, smb_stat.c, wsp_stat.c:
+	* packet-fcswils.c:
 
-	  Assorted GUI cleanups.
+	  Add {0, NULL} to end of fc_swils_rscn_addrfmt_val value_string
+	  array.
 
-2003-09-26 04:43  tpot
+2003-12-04 05:38  gram
 
-	* packet-dcerpc.c:
+	* packet-fcip.c:
 
-	  Use zero to mean we haven't seen any authentication level
-	  information in dcerpc_auth_info since auth_level is an unsigned
-	  type.  Zero is not a valid authentication level anyway
-	  (s13.1.2.1, p611 CAE spec).
+	  Add {0, NULL} to end of fcencap_proto_vals value_string array.
 
-	  Remove two inscrutable debugging comments that don't seem to mean
-	  anything.
+2003-12-04 00:45  guy
 
-2003-09-26 06:30  tpot
+	* filters.c, filters.h, gtk/dcerpc_stat.c,
+	  gtk/endpoint_talkers_table.c, gtk/fc_stat.c, gtk/filter_prefs.c,
+	  gtk/find_dlg.c, gtk/h225_counter.c, gtk/h225_ras_srt.c,
+	  gtk/ldap_stat.c, gtk/mgcp_stat.c, gtk/prefs_dlg.c,
+	  gtk/rpc_stat.c, gtk/service_response_time_table.c,
+	  gtk/smb_stat.c:
 
-	* packet-dcerpc.c, packet-dcerpc.h, packet-ntlmssp.c:
+	  In GTK+ 2.x, "gtk_entry_get_text()" returns a "const gchar *";
+	  assign its value to pointer-to-const variables.
 
-	  This commit refactors the dcerpc authentication subdissectors for
-	  handling encrypted request/response PDUs.  Instead of having
-	  dissection function pointers which perform both decryption and
-	  dissection, the function pointers now only decrypt the DCERPC
-	  fragment payload.  Dissection is handled by the
-	  dcerpc_try_handoff() function (with DCERPC fragment reassembly if
-	  necessary).
+2003-12-03 23:54  guy
 
-	  Details:
+	* AUTHORS, Makefile.am, Makefile.nmake, packet-gsm_map.c,
+	  doc/ethereal.pod:
 
-	   - Move the dcerpc_auth_info struct into dcerpc.h as it is now
-	  used in
-	     the function prototype for the decryption function handlers.
+	  From Felix Fei: GSM MAP support.
 
-	   - decode_encrypted_data() was refactored to take a boolean
-	  request
-	     parameter instead of passing the DCERPC PDU packet type.
+	  From Michael Lum:
 
-	   - A tvbuff_t * data field was added to dcerpc_auth to hold the
-	     verifier.	This is passed as an argument to the decryption
-	  function
-	     handlers.
+		  Modified for better TCAP separation, fixed EOC handling
+	  (a la
+		  TCAP).
 
-	   - Dissection of verifiers in request and response PDUs was moved
-	  to
-	     before the payload.
+		  Added parameter parsing (although not dissection or
+	  naming).
 
-	   - The dissect_dcerpc_cn_stub() function was refactored to
-	  perform
-	     the decryption process and hand decrypted data to the
-	  reassembly
-	     code instead of performing the decryption after reassembly.
+2003-12-03 23:46  guy
 
-	   - Removed references to decrypted_info_t as it's not necessary
-	     anymore.
+	* AUTHORS, packet-ansi_a.c, packet-ansi_a.h, tap-ansi_astat.c,
+	  gtk/Makefile.am, gtk/Makefile.nmake, gtk/ansi_a_stat.c:
 
-	  Code was tested using encrypted and unencrypted fragmented PDUs.
-	  Before this commit ethereal could not dissect unencrypted (!)
-	  fragmented PDUs correctly.
+	  From Michael Lum: Ethereal tap for ANSI A-interface.
 
-2003-09-26 08:19  guy
+	  Tweak some Makefile.nmake whitespace.
 
-	* packet-lapb.c:
+2003-12-03 22:50  guy
 
-	  The LAPB dissector can be called from the Ethernet dissector;
-	  don't assume we have an X.25 pseudo-header.
+	* packet-mtp2.c, packet-mtp3.c:
 
-2003-09-26 20:00  guy
+	  From Jeff Morriss: add support for WTAP_ENCAP_MTP2 and
+	  WTAP_ENCAP_MTP3.
 
-	* packet-sip.c:
+2003-12-03 22:40  guy
 
-	  Make the non-heuristic SIP dissector reject the packet if it
-	  doesn't look like a SIP packet, so some other dissector gets a
-	  chance at it.
+	* wiretap/: AUTHORS, libpcap.c, wtap.c, wtap.h:
 
-	  When looking for the blank line separating headers from data, use
-	  "tvb_find_line_end()" so we handle CR/LF and LF as end-of-line
-	  indications (RFC 2543 says "senders MUST terminate lines with a
-	  CRLF", but it also says "but receivers MUSTalso interpret CR and
-	  LF by themselves as line terminators"), and return an offset past
-	  the end of the buffer, rather than -1, if we don't find it (not
-	  all packets have one).
+	  From Jeff Morriss: add support for libpcap MTP2 and MTP3
+	  captures.
 
-	  When checking whether a header is one we know about, do a
-	  case-insensitive comparison (RFC 2543 says header field names are
-	  case-insensitive).
+	  Add a missing entry in "encap_table[]" for Bluetooth H4, and fix
+	  some comments.
 
-2003-09-26 21:32  guy
+2003-12-03 20:01  guy
 
-	* packet-sip.c:
+	* packet-bootp.c:
 
-	  For (non-heuristic) SIP-over-TCP, dissect stuff that's neither a
-	  request nor a response as continuation data.	For
-	  SIP-over-everything-else, reject it.
+	  From Greg Morris: handle the backwards compatibility flag for
+	  options 78 and 79.
 
-	  Parse the headers regardless of whether we're building a protocol
-	  tree or not; if we're not, we just do it to look for a blank line
-	  separating the headers from the body.  Do that instead of
-	  scanning for the message body separately.
+2003-12-03 19:59  guy
 
-	  When scanning for a colon, don't scan past the end of the line.
+	* gtk/menu.c:
 
-2003-09-26 22:20  guy
+	  From Greg Morris: the "File->Save Highlighted" menu item has no
+	  icon, so don't use "ITEM_FACTORY_STOCK_ENTRY()" for it.
 
-	* Makefile.am, Makefile.nmake, packet-h225.c, packet-h225.h,
-	  packet-h245.c, packet-h245.h, packet-h450.c, t35.c, t35.h:
+2003-12-03 19:47  guy
 
-	  From Tomas Kukosa:
+	* wiretap/airopeek9.c:
 
-		  1) string tables for t35CountryCode, t35Extension and
-		     h221ManufacturerCode were moved into the new file
-	  t35.c
-		     because they are common for more dissectors
+	  "strtoul()" returns a "long", not a "long long".
 
-		  2) the
-	  dissect_h245_NonStandardParameter_with_extension_marker()
-		     was moved from h245 to h225 and renamed to
-		     dissect_h225_NonStandardParameter() because the
-		     NonStandardData type is different for H.225.0 and
-	  H.245
+2003-12-03 10:14  sahlberg
 
-		  3) type of the "h245.nsp.object" dissector table was
-	  changed from
-		     FT_UINT32 to FT_STRING, so it can select a dissector
-	  based on
-		     an OID rather than the Adler-32 hash of an OID
+	* epan/tvbuff.c:
 
-		  4) the "h225.nsp.object" and "h225.nsp.h221" dissector
-	  tables
-		     were created
+	  Replace two expensive macros with a much simpler mechanism to
+	  prevent memory leaks due to exceptions.
 
-2003-09-26 22:29  guy
+	  makes my tethereal testcase ~1% faster.
 
-	* gtk/rtp_stream_dlg.c:
+2003-12-03 09:50  sahlberg
 
-	  Get the RTP payload types from rtp_pt.h rather than defining them
-	  ourselves.
+	* epan/: proto.c, tvbuff.c, tvbuff.h:
 
-2003-09-26 23:11  guy
+	  performance update
 
-	* gtk/: capture_combo_utils.c, capture_prefs.c:
+	  replace tvb_raw_offset() which is essentially a simple assignment
+	  and which is called a lot with a macro.
 
-	  From Nathan Jennings:
+	  this makes my tethereal testcase 2-3% faster.
 
-		  update the CList as you enter/modify options;
+2003-12-03 09:28  guy
 
-		  give Windows users OS descriptions in the displayed
-	  devices
-		  list;
+	* packet-ncp2222.inc, print.c, proto_hier_stats.c,
+	  tap-protohierstat.c, epan/proto.c, epan/proto.h,
+	  epan/ftypes/ftypes.c, epan/ftypes/ftypes.h, gtk/main.c,
+	  gtk/menu.c, gtk/prefs_dlg.c, gtk/rtp_analysis.c:
 
-		  display at least 5 rows in the lists;
+	  The "ptr_u" unions no longer have a "next" pointer - they now
+	  just have one member - or have one that's not used, so get rid of
+	  those unions.
 
-		  get rid of the "extra" CList for storing edited values.
+2003-12-03 08:53  guy
 
-2003-09-27 23:34  guy
+	* epan/: proto.c, proto.h, slab.h, ftypes/ftypes.c,
+	  ftypes/ftypes.h:
 
-	* packet-dcerpc-srvsvc.c, packet-dcerpc-srvsvc.h:
+	  Instead of requiring slab-allocated structures to have a "next"
+	  pointer, when adding them to the free list, cast the pointer to
+	  the structure to a pointer to a "freed_item_t" which contains the
+	  "next" pointer.
 
-	  From Jean-Baptiste Marchand: add names of DFS-related operations
-	  in SRVSVC.
+	  This reduces the memory requirement for some of those structures,
+	  and leaves us free to slab-allocate structures that have a "next"
+	  pointer for other reasons.
 
-2003-09-27 23:43  guy
+2003-12-03 08:43  sahlberg
 
-	* packet-dcerpc-wkssvc.c, packet-dcerpc-wkssvc.h:
+	* packet-smb.c:
 
-	  From Jean-Baptiste Marchand: add names of operations in WKSSVC.
+	  Update to SMB ReadAndX replies.
 
-2003-09-27 23:45  sahlberg
+	  There might be a 4 (not 2) byte datalen high field.  If there is
+	  no such field (due to an early dialect being used) these bytes
+	  are often 0xff.  If these four bytes are all 0xff    dont display
+	  them in the tree.
 
-	* packet-dcerpc-messenger.c:
+2003-12-03 07:46  guy
 
-	  From JBM   update the function names for Messenger
+	* README.linux:
 
-2003-09-27 23:48  sahlberg
+	  Bring it up to date.
 
-	* packet-dcerpc-netlogon.c, packet-dcerpc-netlogon.h:
+2003-12-02 23:14  guy
 
-	  From JBM   update some function names in Netlogon
+	* file.c, tethereal.c, epan/column-utils.c, epan/column_info.h,
+	  gtk/main.c:
 
-2003-09-27 23:51  guy
+	  From Didier Gautheron: for each column format type, save the
+	  first and last columns, if any, with that format, and use that to
+	  speed up processing of columns with a particular format and
+	  checking whether we're displaying a column with a particular
+	  format.
 
-	* packet-isup.c:
+2003-12-02 21:15  guy
 
-	  From Anders Broman: further dissect APM messages containing BICC
-	  stuff, and fix a bug in the "upgraded parameter code".
+	* packet-ncp2222.inc, packet-smb-sidsnooping.c, tap-iostat.c,
+	  epan/proto.c, epan/proto.h, epan/dfilter/dfvm.c,
+	  epan/ftypes/ftypes.c, epan/ftypes/ftypes.h, gtk/io_stat.c,
+	  gtk/main.c, gtk/rtp_analysis.c:
 
-2003-09-28 00:00  sahlberg
+	  From Didier Gautheron: put an "fvalue_t" structure into a
+	  "field_info" structure, rather than separately allocating
+	  "fvalue_t"s and having the "field_info" structure point to them -
+	  this appears to speed up protocol tree construction a bit.
 
-	* tap-smbstat.c, gtk/smb_stat.c:
+2003-12-02 20:27  guy
 
-	  Update to SMB service response time stats.  For short packets, we
-	  might not have enough of the payload to decode the transaction
-	  info levels  and thus that data structure is NULL.
+	* wiretap/airopeek9.c:
 
-	  check the pointer to this struct first before we try to
-	  dereference it.
+	  Check for errors and EOF, and handle them appropriately; don't
+	  treat all errors when reading the header as indications that the
+	  file isn't an AiroPeek V9 file.
 
-2003-09-28 00:11  sahlberg
+	  Put in comments nothing some additional checks we should do.
 
-	* packet-smb.c:
+2003-12-02 19:37  guy
 
-	  in ReadAndX when reading what could potentially be the maxcount
-	  high field assume that IF it is 0xFFFFFFFF  that it is not
-	  maxcount high at all but instead just some padding/reserved
-	  bytes.
+	* AUTHORS, doc/ethereal.pod, wiretap/AUTHORS, wiretap/Makefile.am,
+	  wiretap/Makefile.nmake, wiretap/airopeek9.c, wiretap/airopeek9.h,
+	  wiretap/file_access.c, wiretap/wtap.h:
 
-	  If this field is 0xFFFFFFFF just ignore it.
+	  From Martijn Schipper: support for reading AiroPeek files in V9
+	  capture file format (AiroPeek 2.x).
 
-2003-09-28 01:52  sahlberg
+2003-12-02 18:50  guy
 
-	* packet-mount.c, packet-nfs.c, packet-nfs.h, packet-nfsacl.c:
+	* packet-tpkt.c:
 
-	  Prettify NFSv2     decorate COL_INFO and the tree pane as has
-	  already been done for v3
+	  Tomas Kukosa noted that the TPKT dissector spends a significant
+	  amount of time in "find_protocol_by_id()"; the protocol_t pointer
+	  for a protocol doesn't change over time, so get that pointer when
+	  the protocol is registered and save it.
 
-2003-09-28 11:35  tpot
+2003-12-02 18:35  guy
 
-	* packet-dcerpc-reg.c, packet-dcerpc-reg.h:
+	* packet-isakmp.c:
 
-	  Operation name updates for winreg pipe from Jean-Baptiste
-	  Marchand.
+	  From Yaniv Kaul:
 
-2003-09-28 21:39  guy
+		  1) fix description of Check Point version 5004;
 
-	* epan/tvbuff.c:
+		  2) add description of Check Point version 5005.
 
-	  "compute_offset_length()" must, if it returns FALSE, and
-	  "exception" is non-null, set "*exception" to the appropriate
-	  exception - its callers rely on it.
+2003-12-02 10:23  sahlberg
 
-	  Now that it does that, there's no need for
-	  "check_offset_length()" to check for a length of -1, as
-	  "compute_offset_length()" does so, and therefore
-	  "check_offset_length_no_exception()" does so.
+	* epan/: proto.c, tvbuff.c, tvbuff.h:
 
-2003-09-28 23:15  guy
+	  Move the definition of the tvbuff_t structure and friends to
+	  tvbuff.h so that we can change tvb_get_ds_tvb() into a macro.
 
-	* wiretap/netxray.c:
+	  This function was a single line assignment and was called a lot.
 
-	  It appears that, at least for gigabit pod captures, there are
-	  time stamp differences between versions 002.001 and 002.002.
+	  This made tethereal ~2.5% faster in one testcase I use.
 
-2003-09-29 00:01  tpot
+2003-12-02 09:47  sahlberg
 
-	* packet-dcerpc-lsa.c, packet-dcerpc-lsa.h, packet-dcerpc-samr.c,
-	  packet-dcerpc-samr.h:
+	* epan/: proto.c, ftypes/ftypes.c, ftypes/ftypes.h:
 
-	  More operation names updates from Jean-Baptiste Marchand.
+	  Use the LSAB_ALLOC and SLAB_FREE macros to allocate/free fvalue_t
+	  data
 
-2003-09-29 06:41  oabad
+2003-12-02 09:11  sahlberg
 
-	* gtk/dfilter_expr_dlg.c:
+	* epan/: proto.c, slab.h:
 
-	  - in show_relations() : select the first row of the relation_list
-	  when	 using gtk+ v2.  - get rid of some unused variables.
+	  Moving SLAB_ALLOC and SLAB_FREE to its own header file
 
-2003-09-29 18:50  guy
+2003-12-02 09:05  guy
 
-	* packet-tacacs.c, packet-tacacs.h:
+	* AUTHORS, packet-isakmp.c, doc/ethereal.pod:
 
-	  From Emanuele Caratti:
+	  From Arnaud Jacques: fix a typo, and handle another CheckPoint
+	  version number.
 
-		  just use "g_free()" to free the buffer in "md5_xor()", as
-	  it
-		  doesn't throw exceptions;
+2003-12-02 06:06  guy
 
-		  temporarily #ifdef out "tacplus_acct_flags" pending the
-		  arrival of code to dissect the TACACS+ accounting stuff.
+	* packet-fcfzs.c:
 
-2003-09-29 19:17  jmayer
+	  From Dave Sclarsky:
 
-	* packet-smb.c:
+		  Correct several problems with zoneset decodes (e.g.  the
+	  zoneset
+		  and zone name lengths do *not* include the pad bytes, so
+	  the
+		  offset calculation needed to be fixed).
 
-	  Added comment: XXX Argh maxcnt_high is guint16 and thus 16 bit ->
-	  always false
+		  Move the decode of zoneset name into the routine
+		  "dissect_fcfzs_zoneset", so that it is done in only one
+	  place.
 
-2003-09-29 19:18  jmayer
+2003-12-02 05:53  guy
 
-	* tap-httpstat.c, gtk/http_stat.c:
+	* packet-ses.c:
 
-	  Whitespace changes in order to make diff produce more readable
-	  results
+	  Don't call the presentation layer dissector if we don't have a
+	  handle for it.
 
-2003-09-29 19:20  jmayer
+2003-12-02 02:58  guy
 
-	* gtk/rtp_analysis.c:
+	* packet-bssap.c, packet-tcap.c:
 
-	  use mkstemp instead of tmpnam
+	  From Michael Lum: don't register the handoff routine for
+	  dissectors as a preference change callback if the ports they
+	  register on aren't preferences.
 
-2003-09-29 19:32  jmayer
+2003-12-01 23:41  guy
 
-	* plugins/.cvsignore:
+	* epan/value_string.c, epan/value_string.h,
+	  plugins/plugin_api_list.c, plugins/Xass-list,
+	  plugins/Xplugin_api.c, plugins/Xplugin_api.h,
+	  plugins/Xplugin_api_decls.h, plugins/Xplugin_table.h:
 
-	  xyzzy
+	  From Tomas Kukosa: add "decode_enumerated_bitfield_shifted()",
+	  which is like "decode_enumerated_bitfield()" but handles
+	  value_string tables containing values as they appear in the
+	  bitfield rather than as they appear in the item containing the
+	  bitfield.
 
-2003-09-29 21:12  guy
+2003-12-01 23:26  guy
 
-	* packet-isup.c:
+	* packet-m2ua.c:
 
-	  From Michael Lum: fix some val_to_str calls to have a non-null
-	  format string for unknown values.
+	  From Michael Lum: add a preference setting to specify the
+	  parameter tag for "Protocol Data 1", as it changed between draft
+	  7 and RFC 3331.
 
-2003-09-29 21:50  guy
+2003-12-01 23:05  guy
 
-	* packet-isup.c:
+	* Makefile.am, Makefile.nmake, packet-ansi_a.c, packet-ansi_a.h,
+	  packet-bssap.c, tap-ansi_astat.c:
 
-	  Clean up a bunch of length processing - use the reported length
-	  rather than the captured length, and fix up some other stuff.
+	  From Michael Lum:
 
-2003-09-30 20:51  guy
+		  add a message statistics tap for ANSI A interface for
+	  Tethereal;
 
-	* packet-nt-sonmp.c:
+		  fix the BSSAP, BSMAP, and DTAP interface dissectors to
+	  call
+		  subdissectors even if no protocol tree is being built.
 
-	  From Giles Scott: add some new hardware types.
+2003-12-01 22:25  guy
 
-2003-10-01 07:11  guy
+	* packet-ranap.c:
 
-	* packet-eth.c, packet-eth.h, packet-ethertype.c, packet-fw1.c,
-	  packet-ieee8023.c, packet-ieee8023.h, packet-llc.c,
-	  packet-null.c, packet-sll.c, packet-vlan.c, epan/packet.h,
-	  wiretap/5views.c, wiretap/erf.c, wiretap/etherpeek.c,
-	  wiretap/file_access.c, wiretap/iptrace.c, wiretap/lanalyzer.c,
-	  wiretap/libpcap.c, wiretap/netmon.c, wiretap/nettl.c,
-	  wiretap/netxray.c, wiretap/ngsniffer.c, wiretap/radcom.c,
-	  wiretap/snoop.c, wiretap/toshiba.c, wiretap/visual.c,
-	  wiretap/wtap-int.h, wiretap/wtap.h:
+	  From Michael Lum: add a heuristic RANAP dissector, for use atop
+	  SCCP, as the SCCP dissector currently doesn't support
+	  conversations to be given dissectors.
 
-	  Have a pseudo-header for Ethernet packets, giving the size of the
-	  FCS - 0 means "there is no FCS in the packet data", 4 means
-	  "there is an FCS in the packet data", -1 means "I don't know
-	  whether there's an FCS in the packet data, guess based on the
-	  packet size".
+2003-12-01 20:27  gerald
 
-	  Assume that Ethernet encapsulated inside other protocols has no
-	  FCS, by having the "eth" dissector assume that (and not check for
-	  an Ethernet pseudo-header).
+	* util.c:
 
-	  Have "ethertype()" take an argument giving the FCS size; pass 0
-	  when appropriate.
+	  Don't automatically fill in a capture filter if CLIENTNAME is set
+	  to "Console".
 
-	  Fix up Wiretap routines to set the pseudo-header.  This means we
-	  no longer use the "generic" seek-and-read routine, so get rid of
-	  it.
+2003-12-01 17:25  gerald
 
-2003-10-01 08:53  guy
+	* README.win32:
 
-	* packet-smb.c:
+	  Updates for new zlib.
 
-	  The max count high field is 32 bits, and, in order to compare it
-	  against 0xffffffff, it has to be extracted into a 32-bit
-	  variable.
+2003-12-01 17:19  gerald
 
-2003-10-01 14:59  jmayer
+	* packaging/nsis/ethereal.nsi:
 
-	* doc/: .cvsignore, Makefile.am, ethereal-filter.pod.template,
-	  ethereal.pod, ethereal.pod.template, tethereal.pod,
-	  tethereal.pod.template:
+	  Switch from zlib.dll to zlib1.dll.
 
-	  Put the display-filter elements into it's own manpage
-	  (ethereal-filter.4)
+2003-12-01 17:17  gerald
 
-2003-10-01 15:09  jmayer
+	* config.nmake:
 
-	* Makefile.am, configure.in:
+	  Switch to the new (and official) zlib DLL from www.gzip.org.
 
-	  Put the display-filter elements into it's own manpage
-	  (ethereal-filter.4)
+2003-12-01 06:59  sharpe
 
-2003-10-01 18:19  guy
+	* wiretap/file_access.c:
 
-	* wiretap/wtap.def:
+	  It's RedHat, not Red Hat.
 
-	  Have a pseudo-header for Ethernet packets, giving the size of the
-	  FCS - 0 means "there is no FCS in the packet data", 4 means
-	  "there is an FCS in the packet data", -1 means "I don't know
-	  whether there's an FCS in the packet data, guess based on the
-	  packet size".
+2003-12-01 02:01  guy
 
-	  Assume that Ethernet encapsulated inside other protocols has no
-	  FCS, by having the "eth" dissector assume that (and not check for
-	  an Ethernet pseudo-header).
+	* gtk/: file_dlg.c, file_dlg.h, print_dlg.c, proto_draw.c:
 
-	  Have "ethertype()" take an argument giving the FCS size; pass 0
-	  when appropriate.
+	  "select_file_cb()" only needs a title, not a "construct_args_t" -
+	  "construct_args_t" is for use with filter dialogs, and the
+	  members other than the title apply only to filter dialogs.
 
-	  Fix up Wiretap routines to set the pseudo-header.  This means we
-	  no longer use the "generic" seek-and-read routine, so get rid of
-	  it.
+	  Have "select_file_cb()" actually use the title supplied to it.
 
-2003-10-01 19:41  guy
+2003-12-01 01:57  sharpe
 
-	* doc/Makefile.am:
+	* gtk/menu.c:
 
-	  dfilter2pod.pl and ethereal-filter.pod.template are in $(srcdir)
-	  (which defauls to the current directory), not in the parent
-	  directory.
+	  Fix the accelerator issue ...
 
-2003-10-01 19:44  guy
+2003-11-30 23:05  guy
 
-	* doc/Makefile.nmake:
+	* gtk/proto_draw.c:
 
-	  Update for the new ethereal-filter man page.
+	  Not all platforms have <unistd.h> - in particular, Windows
+	  doesn't.
 
-2003-10-01 20:26  guy
+	  If you're using "open()", "close()", "read()", "write()", etc.,
+	  you should include <io.h> on Windows.
 
-	* doc/Makefile.am:
+2003-11-30 18:01  sharpe
 
-	  Put in some missing $(srcdir)/.
+	* gtk/proto_draw.c:
 
-	  Get rid of redundant "../{t}ethereal.1" in CLEANFILES (they were
-	  already there).
+	  Fix some warnings in proto_draw.c
 
-2003-10-01 20:27  guy
+2003-11-30 04:40  sharpe
 
-	* .cvsignore:
+	* gtk/menu.c:
 
-	  Add ethereal-filter.4.
+	  Rename that to Save Highlighted with Ctl-H as the accellerator.
 
-2003-10-01 20:36  guy
+2003-11-30 04:37  sharpe
 
-	* doc/ethereal.pod:
+	* gtk/menu.c:
 
-	  Filters in Ethereal are usually display filters, not read
-	  filters; go back to describing them as such.
+	  Add Save Selected as a menu item in the File Menu.
 
-2003-10-01 21:15  guy
+	  This still has a way to go, though, because the menu item should
+	  be greyed out until something is Highlighted.
 
-	* packet-srvloc.c:
+2003-11-30 04:21  sharpe
 
-	  The PDU length is 3 bytes long in SLPv2.
+	* gtk/: color_dlg.c, file_dlg.c, file_dlg.h, menu.c, print_dlg.c,
+	  proto_draw.c, toolbar.c:
 
-	  The minimum number of bytes of SLP we have to reassemble is 5 -
-	  it's nominally 4 for SLPv1, but we don't have a way of asking for
-	  1 byte (the version) and then saying "I need N bytes of header to
-	  get the PDU length, and an SLPv1 packet less than 12 bytes long
-	  is bogus anyway.
+	  Move print_file_cb out of gtk/print_dlg.c and into gtk/file_dlg.c
+	  and rename it to select_file_cb to reflect its function.
 
-2003-10-01 21:51  guy
+	  While this cleans things up a bit, I am still not happy because
+	  now filter_prefs.h must be included before file_dlg.h just to get
+	  construct_args_t.
 
-	* packet-enip.c:
+2003-11-29 12:34  jmayer
 
-	  Reject frames with no command (too short) or an invalid command.
+	* doc/README.tapping:
 
-	  Clean up white space somewhat.
+	  Fix two small typos
 
-2003-10-02 06:13  guy
+2003-11-29 06:15  sharpe
 
-	* AUTHORS, Makefile.am, Makefile.nmake, asn1.c, asn1.h,
-	  packet-tcap.c, packet-tcap.h, doc/ethereal.pod:
+	* gtk/menu.c:
 
-	  From Samuel Qu, Michael Lum, and Jeff Morriss: TCAP support, and
-	  "asn_id_decode1()" variant of "asn_id_decode()".
+	  Properly capitalize the string for the menu item.
 
-2003-10-02 14:03  gerald
+2003-11-29 06:09  sharpe
 
-	* doc/ethereal-filter.pod.template:
+	* gtk/: file_dlg.h, menu.c, print_dlg.c, proto_draw.c:
 
-	  Fix a malformed "=head1" tag.
+	  Add an entry in the byte_view pane that allows the user to save
+	  the highlighted data to a file. This allows the user to select
+	  some stuff, and analyse it with external tools, and is very
+	  useful for quickly prototying dissectors etc.
 
-2003-10-02 19:22  gerald
+	  This works by retrieving the info that is needed to define where
+	  the selected region is.
 
-	* doc/tethereal.pod:
+	  It puts up a dialog box that asks for the file to save in.
 
-	  Add an example for "-d".
+	  However, it is an ugly hack, because it reuses print_file_cb,
+	  while print_file_cb should be moved into file_dlg.c.
 
-2003-10-02 21:04  guy
+	  It also needs to have some warning dialogs put up in error cases.
 
-	* plugins/rtnet/packet-rtnet.c:
+	  Finally, it would be good to be able to select a region in the
+	  byte_view, which you can do with click and drag, and then have a
+	  menu item to save the selected bytes.
 
-	  RTNET has no preferences, so don't register a preferences module
-	  for it.
+2003-11-29 03:55  guy
 
-2003-10-02 21:06  guy
+	* AUTHORS, packet-gtp.c, doc/ethereal.pod:
 
-	* prefs.c, prefs.h, gtk/prefs_dlg.c:
+	  From melerski [AT] poczta.onet.pl: fix a problem with unknown
+	  IEs.
 
-	  Don't put an entry for a protocol into the Preferences dialog if
-	  it doesn't have any settable preferences (for example, if it has
-	  only obsolete preferences).
+2003-11-28 19:13  ulfl
 
-2003-10-02 21:18  guy
+	* gtk/menu.c:
 
-	* gtk/prefs_dlg.c:
+	  renamed frame to packet in all GUI elements moved "Enabled
+	  Protocols" and such from "View" to "Analyze" toplevel menu
 
-	  From Tomas Kukosa: radio button groups are GSLists, which means
-	  that the radio button group for a button changes when new buttons
-	  are added to it (adding to the beginning of a singly-linked list
-	  takes constant time, adding to the end takes time linear in the
-	  length of the list, and a GSList * points to the beginning of the
-	  list).  Re-fetch the radio button group each time through the
-	  loop that adds new radio buttons to a radio button group for a
-	  preference.
+2003-11-28 19:02  ulfl
 
-2003-10-02 21:37  guy
+	* gtk/goto_dlg.c:
 
-	* packet-dcerpc-dnsserver.c, packet-dcerpc-dnsserver.h:
+	  renamed frame to packet in all GUI elements
 
-	  From Jean-Baptiste Marchand: add names for new dnsserver
-	  operations for W2K3.
+2003-11-28 19:00  ulfl
 
-2003-10-02 21:40  guy
+	* gtk/capture_info_dlg.c:
 
-	* packet-dcerpc-svcctl.c, packet-dcerpc-svcctl.h:
+	  use dlg_window_new instead of gtk_window_new
 
-	  From Jean-Baptiste Marchand: add/update names for svcctl
-	  operations.
+2003-11-28 18:59  ulfl
 
-2003-10-02 21:45  guy
+	* gtk/: toolbar.c, find_dlg.c:
 
-	* packet-dcerpc-dfs.c, packet-dcerpc-dfs.h:
+	  renamed frame to packet in all GUI elements
 
-	  From Jean-Baptiste Marchand: add additional operation names for
-	  dfssvc.
+2003-11-28 00:44  ulfl
 
-2003-10-02 21:48  guy
+	* gtk/capture_info_dlg.c:
 
-	* packet-dcerpc-browser.c, packet-dcerpc-browser.h:
+	  label changed from "Portions" to "% of total" as suggested by Guy
 
-	  From Jean-Baptiste Marchand: add operation names for browser
-	  service.
+2003-11-28 00:13  ulfl
 
-2003-10-02 22:28  gerald
+	* gtk/main.c:
 
-	* doc/Makefile.nmake, packaging/nsis/Makefile.nmake,
-	  packaging/nsis/ethereal.nsi:
+	  using dlg_utils.c functions to catch the ESC key to close the
+	  dialog
 
-	  Fix document creation under Windows, add ethereal-filter.html to
-	  the NSIS package.
+2003-11-28 00:07  ulfl
 
-2003-10-02 22:40  guy
+	* gtk/: supported_protos_dlg.c, plugins_dlg.c:
 
-	* t35.c, t35.h:
+	  using dlg_utils.c functions to catch the ESC key to close the
+	  dialog
 
-	  Add RCS IDs.
+2003-11-27 23:25  ulfl
 
-2003-10-02 22:44  guy
+	* gtk/help_dlg.c:
 
-	* adler32.c, adler32.h, packet-etherip.c, packet-h450.c,
-	  packet-ipsec-udp.c, packet-isns.c:
+	  using dlg_utils.c functions to catch the ESC key to close the
+	  dialog
 
-	  Add RCS IDs.
+2003-11-27 22:29  guy
 
-2003-10-03 04:41  guy
+	* print.c:
 
-	* doc/tethereal.pod:
+	  "representation", in an "item_label_t", is now an array, not a
+	  pointer, so "fi->rep->representation" can't be null.
 
-	  Fix a typo.
+2003-11-27 22:15  guy
 
-2003-10-03 09:09  sahlberg
+	* tap-rpcstat.c, tap-wspstat.c:
 
-	* packet-ipx.c, gtk/endpoint_talkers_table.c:
+	  Fix comments to correctly refer to the "-z" option.
 
-	  Update ipx and conversation list to make it possible to select
-	  and filter for ipx conversations from the conversation list popup
-	  menu
+2003-11-27 22:14  guy
 
-2003-10-03 20:18  tuexen
+	* tap-rpcstat.c:
 
-	* packet-sctp.c:
+	  From Michael Lum: fix the usage message to refer to "-z", not
+	  "-Z".
 
-	  - Fixed a typo.  - Changed the default checksum algorithm from
-	  Adler32 to CRC32C.
+2003-11-27 22:12  guy
 
-2003-10-03 20:58  guy
+	* gtk/main.c:
 
-	* packet-isup.c:
+	  Show the -z flag in the usage message.
 
-	  From Anders Broman:
+2003-11-27 22:04  guy
 
-		  fix some cut and paste errors in "upgraded parameter"
-	  routine;
+	* tethereal.c:
 
-		  more BICC work.
+	  From Michael Lum: fix the usage message to say "-z", not "-Z".
 
-2003-10-03 21:03  guy
+2003-11-27 21:20  guy
 
-	* wiretap/ascend-scanner.l:
+	* AUTHORS, packet-rtp-events.c, doc/ethereal.pod:
 
-	  From Graham Bloice: define YY_NO_UNISTD_H on Win32, so that if
-	  Flex was a UNIX version generating code that, by default, assumes
-	  you have <unistd.h> (as might be the case with recent versions of
-	  Cygwin, which I assume *does* supply <unistd.h>), but you're
-	  building on a platform that lacks <unistd.h> (e.g., building with
-	  MSVC++ or MinGW), you can still compile.
+	  From Arsen Chaloyan: fix offsets in "proto_tree_add" calls.
 
-2003-10-03 21:12  jmayer
+2003-11-27 21:09  guy
 
-	* wiretap/ascend-scanner.l:
+	* etypes.h, packet-ethertype.c, plugins/rtnet/packet-rtnet.c:
 
-	  Use #ifndef HAVE_UNISTD_H instead of #ifdef _WIN32
+	  From Erwin Rol: RTCFG support.
 
-2003-10-03 21:19  jmayer
+	  Add in a URL to the RTNET home page.
 
-	* gtk/rtp_analysis.c:
+2003-11-27 21:02  guy
 
-	  Graham Bloice: Add missing #include mkstemp.h
+	* packet-dns.c:
 
-2003-10-03 22:38  guy
+	  From David Fort: fix to previous patch.
 
-	* gtk/rtp_analysis.c:
+2003-11-27 20:34  ulfl
 
-	  Include "mkstemp.h" only if we're including our own "mkstemp()".
+	* gtk/menu.c:
 
-2003-10-03 23:09  guy
+	  Redesign of the whole menu structure, to be more intuitive to the
+	  user and conforms somewhat better to the Gnome Human Interface
+	  Guideline
 
-	* AUTHORS, Makefile.am, Makefile.nmake, packet-ansi_map.c:
+2003-11-27 17:18  ulfl
 
-	  From Michael Lum: ANSI MAP support.
+	* print.c:
 
-2003-10-03 23:22  guy
+	  bugfix: removed access violation when using "File/Print Packet".
+	  The label_ptr could be NULL, when calling function print_line
 
-	* plugins/gryphon/: NEWS, packet-gryphon.c, packet-gryphon.h:
+2003-11-27 00:22  ulfl
 
-	  From Steve Limkemann:
+	* gtk/help_dlg.c:
 
-	      Added two new "well known destinations": SD_IOPWR AND
-	  SD_UTIL.
+	  using text files from help dir, to display the help texts.  the
+	  "dynamic protocol fields" are already moved to
+	  gtk/supported_protos_dlg.c
 
-	      Added logic to dissect the CMD_SET_TIME command.
+2003-11-27 00:17  ulfl
 
-	      Added an alterate destination for the CMD_PGM_START command.
+	* help/: capture_filters.h, display_filters.h, faq.h, overview.h,
+	  well_known.h:
 
-	      Added logic to dissect the CMD_SCHED_MSG_REPLACE command.
+	  Added this intermediate files, until a Makfile.am is available
+	  for unix systems
 
-	      Added logic to dissect the CMD_USDT_REGISTER command.
+2003-11-26 23:23  guy
 
-	      Added logic to dissect the CMD_USDT_SET_FUNCTIONAL command.
+	* AUTHORS, packet-dns.c, doc/ethereal.pod:
 
-	      Added logic to dissect the following commands
-		 CMD_IOPWR_GETINP
-		 CMD_IOPWR_GETLATCH
-		 CMD_IOPWR_CLRLATCH
-		 CMD_IOPWR_GETOUT
-		 CMD_IOPWR_SETOUT
-		 CMD_IOPWR_SETBIT
-		 CMD_IOPWR_CLRBIT
-		 CMD_IOPWR_GETPOWER
-		 CMD_UTIL_SET_INIT_STRATEGY
-		 CMD_UTIL_GET_INIT_STRATEGY
+	  From David Fort: DNS DSS RR support.
 
-	      Added the ability to recongnize more IOCTLS.  (For the
-	  SJA1000 driver, LIN
-		  and power drivers.)
+2003-11-26 12:22  sahlberg
 
-	      Added the ability to recognize more card types.
+	* epan/: proto.c, proto.h:
 
-	      Added dissection of more fields for CMD_SCHED_TX command.
+	  Speed updates
 
-	      Bug fixes and general updating.
+	  create generic macros for allocating/freeing structures.
 
-2003-10-03 23:31  guy
+	  remove one more slow GMemChunk and replace it with a simple
+	  linked list
 
-	* packet-ranap.c:
+	  ~4% speed improvement in my tests.
 
-	  From Michael Lum:
+	  the allocated data is never freed.   this may be a problem if
+	  ethereal is ever supported on a platform lacking resource
+	  tracking but makes the implementation faster and simpler.
 
-		  support for Global RNC ID;
+2003-11-26 02:58  guy
 
-		  fixed some typos
+	* AUTHORS, packet-giop.c, doc/ethereal.pod:
 
-		  added push of 'NAS PDU' so that a GSM 24.008 (DTAP)
-	  dissector
-		  can be added.
+	  From Patrick Kursawe: rename "getline()" so it doesn't collide
+	  with the "getline()" declared in <stdio.h> with recent versions
+	  of glibc.
 
-2003-10-04 03:10  sahlberg
+	  Fix a typo in the AUTHORS file.
 
-	* gtk/io_stat.c:
+2003-11-26 02:54  guy
 
-	  Fix to IO-Stat.
+	* pcap-util.c:
 
-	  IO-Stat failed to produce Advanced/COUNT(*) statistics for fields
-	  of type FT_NONE.  Fixed.
+	  From Jaime Fournier: fix a typo.
 
-	  Now it is possible to do : Advanced/COUNT(*)
-	  Filter:tcp.analysis.retransmission
-	  Field:tcp.analysis.retransmission Advanced/COUNT(*)
-	  Filter:tcp.analysis.duplicate_ack
-	  Field:tcp.analysis.duplicate_ack
+2003-11-25 20:26  guy
 
-	  And it will plot the number of Retransmissions and Duplicate ACKs
-	  seen in each time interval.
+	* AUTHORS, packet-ucp.c, doc/ethereal.pod:
 
-2003-10-04 16:44  guy
+	  From Reinhard Speyerer: handle 10-digit UCP dates, which lack a
+	  seconds value.
 
-	* packet-isakmp.c:
+2003-11-25 20:02  guy
 
-	  If a payload type doesn't have a dissector function, don't crash
-	  by calling through the null dissector pointer, just dissect it as
-	  "Payload".
+	* epan/proto.c:
 
-2003-10-05 05:04  sharpe
+	  As with fvalue_t, so with field_info and item_label - the
+	  individual items aren't individually g_mallocated and can't be
+	  g_freed, only the chunks can.
 
-	* mkcap.c:
+2003-11-25 19:25  guy
 
-	  Add mkcap.c, a little utility to generate reasonable looking TCP
-	  capture files for pedagogic use.
+	* epan/ftypes/ftypes.c:
 
-2003-10-05 14:58  sharpe
+	  Individual fvalue_t's aren't individually allocated with
+	  "g_malloc()", so they can't be freed with "g_free()"; keep a list
+	  of the chunks of "fvalue_t"s, which are whare are allocated with
+	  "g_malloc()", so we can free them all.
 
-	* mkcap.c:
+2003-11-25 18:02  guy
 
-	  Further updates on mkcap.c
+	* packet-sip.c:
 
-2003-10-05 21:57  jmayer
+	  From Anders Broman: add more SIP headers, and the PUBLISH method.
 
-	* packet-dns.c:
+2003-11-25 14:16  sahlberg
 
-	  Fix warning about strict-aliasing
+	* epan/proto.c:
 
-2003-10-05 22:38  jmayer
+	  replace free_node_field_info with a macro since it is only called
+	  from one other place in the code and this reduces one unnecessary
+	  function call overhead.  and it makes ethereal ~1% faster.
 
-	* plugins/docsis/packet-tlv.c:
+2003-11-25 14:11  sahlberg
 
-	  Get rid of another strict-aliasing warning: verify_tfs is a
-	  true-false-string : dereference it accordingly
+	* epan/proto.c:
 
-2003-10-05 22:44  jmayer
+	  rename FREE_FIELD_INFO to FIELD_INFO_FREE for consistency
 
-	* packet-tzsp.c:
+2003-11-25 14:10  sahlberg
 
-	  Fix 3 strict-aliasing warnings:   Use TFS(&var) instead of
-	  VALS(&var) in case var is a true-false-string
+	* epan/proto.c:
 
-2003-10-05 23:09  jmayer
+	  create a FIELD_INFO_NEW for consistency
 
-	* packet-cops.c:
+2003-11-25 14:07  sahlberg
 
-	  variable.type is of type u_char, so use 0 instead of NULL
+	* file.c, print.c, epan/proto.c, epan/proto.h, gtk/proto_draw.c:
 
-2003-10-05 23:27  jmayer
+	  remove another slow GMemChunk from ethereal making ethereal a
+	  little bit faster
 
-	* packet-bssgp.c:
+2003-11-25 13:23  sahlberg
 
-	  Another strict-aliasing warning fix - I hope I got this right :)
+	* epan/proto.c:
 
-2003-10-06 02:01  jmayer
+	  change free_field_info to FREE_FIELD_INFO to indicate it is a
+	  macro and not a function
 
-	* plugins/gryphon/packet-gryphon.c:
+2003-11-25 13:20  sahlberg
 
-	  Fix 4 warnings in case of strict-aliasing by declaring timestamp
-	  as time_t instead of int.
+	* epan/proto.c, epan/dfilter/dfvm.c, epan/ftypes/ftype-ipv4.c,
+	  epan/ftypes/ftype-string.c, epan/ftypes/ftypes-int.h,
+	  epan/ftypes/ftypes.c, epan/ftypes/ftypes.h,
+	  gtk/dfilter_expr_dlg.c:
 
-2003-10-06 07:26  guy
+	  fvalue_free() is one of the most called functions.  This function
+	  is also very small, so small that teh overhead for the actual
+	  function call and return is likely to be a significant part of
+	  its execution time.
 
-	* packet-lmi.c:
+	  change it into a macro and make it thus slightly faster by
+	  eliminating the function call overhead.
 
-	  Update a URL.
+2003-11-25 08:50  sahlberg
 
-2003-10-06 08:10  guy
+	* epan/ftypes/: ftypes.c, ftypes.h:
 
-	* packet-enip.c:
+	  Change the handling of temporary allocation of fvalue_t
+	  structures to be faster.  Use a similar technique as was used in
+	  epan/proto.c to speed up the field_info allocation/deallocation
+	  routines
 
-	  Fix the handling of padding bytes.
+2003-11-25 08:19  guy
 
-2003-10-06 08:35  guy
+	* packet-dcerpc.c:
 
-	* packet-dcerpc-rpriv.c, packet-dcerpc-spoolss.c, packet-isup.c,
-	  packet-nlsp.c, packet-ranap.c, packet-sua.c:
+	  From Jaime Fournier: show the RPC version of the unknown
+	  interface in the Info column.
 
-	  From packet steve: get rid of some duplicate field definitions
-	  (some aren't exactly duplicates, but they both set the same hf_
-	  variable).
+2003-11-25 08:10  guy
 
-2003-10-06 08:58  guy
+	* AUTHORS, Makefile.am, Makefile.nmake, packet-dcerpc-budb.c,
+	  packet-dcerpc-epm4.c, packet-dcerpc-icl_rpc.c,
+	  packet-dcerpc-rs_prop_acct.c:
 
-	* gtk/rtp_stream_dlg.c:
+	  From Jamie Fournier: DCE RPC EPM version 4 support, and stub
+	  support for BUDB, ICL RPC, and RS_PROP_ACCT DCE RPC-based
+	  protocols.
 
-	  (Based on a patch from Lars Roland.)
+2003-11-25 05:58  guy
 
-	  Use "gtk_dialog_new()" to create the window - that doesn't create
-	  a "dialog box" in the sense of a transient-for window, but it
-	  does create a window with a button vbox that the code expects to
-	  be present.
+	* wiretap/cosine.c:
 
-2003-10-06 09:08  guy
+	  "file_seek()" sets "*err" - we don't have to set it ourselves.
 
-	* packet-ansi_map.c:
+	  The same is true of "parse_cosine_rec_hdr()".
 
-	  From Lars Roland: not all compilers like static const arrays with
-	  unknown size, so don't use them.
+2003-11-25 05:56  guy
 
-2003-10-06 09:18  guy
+	* wiretap/network_instruments.c:
 
-	* AUTHORS, Makefile.am, Makefile.nmake, packet-alcap.c:
+	  "file_seek()" sets "*err" - there's no need to set it ourselves.
+	  It returns -1 on errors, so just check for that.
 
-	  From Michael Lum: ALCAP (Q.2630.1) support.
+2003-11-25 00:21  guy
 
-2003-10-06 14:48  jmayer
+	* ipproto.h:
 
-	* packet-alcap.c:
+	  Fix a typo.
 
-	  Replace the forward declaration of msg_parm_strings by moving the
-	  real declaration from the middle of the file to the place of the
-	  forward declaration. This should make MSVC happy.
+2003-11-24 23:46  guy
 
-	  Spotted and proposed fix by Ulf Lamping.
+	* packet-ccsds.c:
 
-2003-10-06 15:41  jmayer
+	  Eliminate compiler warnings.
 
-	* plugins/gryphon/packet-gryphon.c:
+2003-11-24 22:11  guy
 
-	  check in the following comment:     /* XXX This code is neither
-	  Endianess independent, nor will it work      * on platforms that
-	  do not support the *optional* guin64 type	 */
+	* packet-ncp2222.inc, print.c, proto_hier_stats.c,
+	  tap-protohierstat.c, epan/proto.c, epan/proto.h, gtk/main.c,
+	  gtk/menu.c, gtk/prefs_dlg.c, gtk/rtp_analysis.c:
 
-2003-10-06 19:11  guy
+	  Make the recent epan/proto.{c,h} change compile.
 
-	* doc/README.developer:
+2003-11-24 21:42  guy
 
-	  Add One More Item to the list of things not all compilers can
-	  handle.
+	* ipproto.c, ipproto.h, packet-mip6.c:
 
-2003-10-06 19:25  guy
+	  Preserve 62 as an (old) IP protocol type for mobility headers.
 
-	* AUTHORS, Makefile.am, Makefile.nmake, packet-alcap.c,
-	  packet-ansi_637.c, packet-ansi_683.c, packet-ansi_map.c:
+2003-11-24 21:12  sahlberg
 
-	  From Michael Lum:
+	* epan/: proto.c, proto.h:
 
-		  IS-637-A (SMS) support
-		  IS-683-A (OTA) support
-		  ANSI MAP updates
+	  performance enhancement to proto.c
 
-	  Fix a typo in the previous checkin.
+	  Removed the GMemChunk used to allocate/free field_info structures
+	  and used a free list to store the freed structs until they are
+	  allocated again.
 
-2003-10-06 20:46  guy
+	  Ethereal will allocate more field_info structs as it needs to but
+	  never free them. Instead the are just placed in a cheap and fast
+	  free list so that if we want to use the struct again, this will
+	  be fast.
 
-	* packet-afs-register-info.h, packet-clnp.c, packet-fcdns.c,
-	  packet-isup.c, packet-lmi.c, packet-lmp.c, packet-m3ua.c,
-	  packet-nfsacl.c, packet-rsvp.c, epan/proto.c:
+	  This affects the speed of the two functions alloc_field_info()
+	  that should be slightly faster now free_field_info() that was
+	  replaced with a 2 line macro.
 
-	  When registering a field, make sure its ID is -1 or 0 - if it's
-	  not, that probably means you've registered two fields with the
-	  same field ID variable, which is an error.
+	  All in all  my testing suggests that ethereal is 2-3% faster with
+	  this patch.
 
-	  Fix the bugs doing so found.
+2003-11-24 17:41  ulfl
 
-2003-10-06 22:10  guy
+	* Makefile.nmake:
 
-	* gtk/io_stat.c:
+	  bugfix: corrected clean and distclean targets
 
-	  Make the "needs_redraw" flag a gboolean, as it's a Boolean flag.
+2003-11-24 00:34  guy
 
-	  Rename "max_count_types and "max_calc_types" to
-	  "count_type_names" and "calc_type_names", to make it clearer what
-	  they are.
+	* AUTHORS, Makefile.am, Makefile.nmake, packet-ccsds.c,
+	  doc/ethereal.pod:
 
-	  For the advanced statistics, give different error messages for
-	  the case where no field name was specified and where an invalid
-	  field name was specified.
+	  From Scott Hovis :CCSDS (Consultative Committee for Space Data
+	  Systems) support.
 
-	  Give better error messages for the cases where the calculation
-	  type isn't supported for a particular field.
+2003-11-24 00:05  guy
 
-	  Initialize the calculation type for a given field.
+	* packet-isup.c:
 
-2003-10-06 22:13  guy
+	  From Anders Broman: fix dissection of GRS/GRA messages.
 
-	* gtk/io_stat.c:
+2003-11-23 23:42  guy
 
-	  Rename "max_tick_values" to "tick_interval_values", to make it
-	  clearer what it is.
+	* Makefile.nmake:
 
-2003-10-07 03:09  guy
+	  Include the BFD dissector when building with MSVC++.
 
-	* gtk/print_mswin.c:
+2003-11-23 22:00  gerald
 
-	  From Gisle Vanem: if WIN32_LEAN_AND_MEAN is defined, <windows.h>
-	  doesn't include <commdlg.h>, but "print_mswin.c" needs it, so
-	  include it explicitly.
+	* AUTHORS, Makefile.am, packet-bfd.c:
 
-2003-10-07 03:35  guy
+	  From Hannes Gredler: Add support for Bi-directional Fault
+	  Detection (BFD).
 
-	* packet-ftp.c:
+2003-11-22 12:02  jmayer
 
-	  From Giles Scott: put the IP address and port number in PORT
-	  requests and PASV responses into the protocol tree, and, for PASV
-	  responses, create a conversation for the future FTP data
-	  conversation and make the FTP data dissector be the dissector for
-	  it.
+	* packet-radius.c:
 
-2003-10-07 04:36  guy
+	  - Bugfix: Passworddecoding should only be done for attribute 2,
+	  but		not for VSA 2 - Fixme comment about length handling
+	  added - Cleanup: Setting a variable to the right type removes the
+	  necessity	      for casts.
 
-	* gtk/main.c:
+2003-11-22 11:12  sahlberg
 
-	  Make the window geometry information static - it's not used
-	  outside this file.
+	* packet-h245.c:
 
-	  Fetch the geometry information whenever we get a configure_notify
-	  event, i.e. if it actually changes, rather than doing so when we
-	  get asked to delete the main window or we exit.  Don't save the
-	  geometry if we've never gotten such an event, as it presumably
-	  means the size and position haven't changed.
+	  Change the way signalType is dissected to be compatible with and
+	  dissect the packets from Martin R and Paul H.
 
-2003-10-07 09:30  sahlberg
+	  I still think ethereal was correct before this patch and that
+	  this patch is not compatible with teh standard.  Anyway, that
+	  doesnt matter much. With this patch ethereal will dissect the
+	  packets properly.  Time to study x691 again.
 
-	* gtk/: endpoint_talkers_table.c, find_dlg.c, find_dlg.h:
+2003-11-22 08:35  guy
 
-	  Add Find Next/Find Previous submenu to conversation lists.
+	* ipproto.h, packet-ipv6.h:
 
-	  By using Find Next/Previous you will jump to the next/previous
-	  matching packet in the ethereal main window.
+	  From Martti Kuparinen: use correct values for Mobile IPv6.
 
-	  I could not get CTRL-N / CTRL-B to work and was too lazy to
-	  research.  It would be nice if CTRL-N / CTRL-P would invoke the
-	  same thing as selecting /Find Frame/Find Next/EP1 <-> EP2 or
-	  /Find Frame/Find Previous/EP1 <-> EP2 from the menu.	I could not
-	  figure out how to get gtk to do this.  The person that adds
-	  CTRL-N/CTRL-B here will be a hero.
+2003-11-22 04:41  sahlberg
 
-2003-10-07 09:50  sahlberg
+	* epan/proto.c:
 
-	* gtk/find_dlg.h:
+	  As per Guys suggestion, continue optimizing proto.c to macroize
+	  it and make it even faster
 
-	  Forgot to update a prototype in previous patch
+2003-11-21 22:18  guy
 
-2003-10-07 10:07  sahlberg
+	* packet-smpp.c:
 
-	* gtk/: color_dlg.c, color_dlg.h, endpoint_talkers_table.c:
+	  From Olivier Biot: fix up the description of the
+	  "port_number_udh_means_wsp" preference.
 
-	  Update to endpoint talkers, endpoint talkers now have an extra
-	  submenu on the popup where one can select : Colorize
-	  Conversation.  This opens up the create color filter dialog with
-	  the filter preset to the selected conversation.
+2003-11-21 22:00  guy
 
-2003-10-07 17:21  guy
+	* packet-wsp.c:
 
-	* t35.c:
+	  From Olivier Biot:
 
-	  From Hans Viens: add an H.221 manufacturer code for Mediatrix,
-	  and tweak white space.
+		  * Fix the WSP dissector so a Quoted-string with
+	  extraneous
+		    trailing quote is displayed without duplicated final
+	  quote, and
+		    issue a warning in the protocol tree.
+		  * Remove the Openwave x-up-proxy-client-id header
+	  decoding as it
+		    was incorrectly decoded as a text string, and process
+	  it with
+		    the default Openwave header parser.
+		  * Fix byte highlighting of some WSP parameters.
+		  * Rename the WSP parameter filter names so they don't all
+	  refer to
+		    content_type.
+
+2003-11-21 21:58  guy
+
+	* packet-wbxml.c, epan/packet.c, epan/packet_info.h:
 
-2003-10-07 17:29  guy
+	  From Olivier Biot:
 
-	* gtk/endpoint_talkers_table.c:
+		  * Add a "match_string" field to the "packet_info"
+	  structure,
+		    saving the string value that matched in a string
+	  dissector
+		    lookup, by analogy to "match_port" - this was required
+	  for
+		    dissection with token rendering of WBXML content when
+	  no public
+		    ID was given (e.g. Nokia/Ericsson OTA provisioning
+	  data).
+		  * Add support for textual content type based WBXML token
+		    mapping.
+		  * Add extra WBXML public identifiers.
+		  * Add the Nokia/Ericsson OTA provisioning (version 7)
+	  token
+		    definitions.
+		  * Inform the user when a content-type based token match
+	  is found.
 
-	  Include "gtk/color_dlg.h" to declare
-	  "color_display_with_filter()", and include "color.h" to declare
-	  "color_filter_t", which "gtk/color_dlg.h" requires.
+2003-11-21 21:33  guy
 
-2003-10-07 17:36  guy
+	* AUTHORS, packet-dcerpc-srvsvc.c, doc/ethereal.pod:
 
-	* packet-ospf.c:
+	  From Eric Wedel: fix dissection of NetServerDiskEnum reply.
 
-	  From Taisuke Sasaki: put in missing {} in an "if".
+2003-11-21 19:24  guy
 
-2003-10-07 17:43  guy
+	* gtk/rtp_stream.c:
 
-	* packet-isup.c:
+	  From Lars Ruoff: use the pointer to the RTP payload, rather than
+	  extracting the data from the frame directly.
 
-	  From Anders Broman:
+2003-11-21 14:58  sahlberg
 
-		  change some variable names to conform to abbreviations
-	  used in
-		  the specs;
+	* epan/: proto.c, proto.h:
 
-		  added+dissection of some more BAT ASE elements;
+	  Performance fix
 
-		  fixed some minor bugs.
+	  In the GPROF logs proto_registrar_get_nth() used to take anything
+	  between 2.5 and 5.5% of the time.
 
-2003-10-07 18:19  oabad
+	  Replace the GLIB array with a handroleld one for one of the
+	  private structures.
 
-	* packet-x25.c:
+	  the function should now be virtually zero cost and thus ethereal
+	  should be 2.5-5.5% faster on those traces.
 
-	  Put the "X.264 protocol identifier" and "X.264 sharing strategy"
-	  fields in the "User data" subtree.
+	  anyone that wants to, please rerun GPROF with this fix and see
+	  what has changed.
 
-2003-10-07 21:15  guy
+2003-11-21 08:40  guy
 
-	* epan/to_str.c:
-
-	  Get rid of an out-of-date comment.
+	* packet-dcerpc-epm.c:
 
-2003-10-08 05:36  guy
+	  A "handle_t" as the first argument to a call doesn't seem to be
+	  sent over the wire, at least in the DCE RPC 1.1 code.
 
-	* packet-isakmp.c:
-
-	  Dissect NAT-Discovery and NAT-Original Address payloads, and just
-	  use the version-independent part of the draft ID for "Negotiation
-	  of NAT-Traversal in the IKE" rather than giving a URL for the
-	  draft (the URL in question was out-of-date, and a future one runs
-	  the risk of becoming out of date, given that the -07 draft
-	  expired a few days ago...).
+	  That suggests that the first 24 bytes of a Map request are the
+	  UUID pointer that the EPM IDL says should be there, not a context
+	  handle.
 
-2003-10-08 06:44  guy
+2003-11-21 08:33  guy
 
-	* etypes.h, packet-ethertype.c:
+	* pcap-util.c:
 
-	  0x81fd and 0x81ff appear to be for the Cabletron Interswitch
-	  Message Protocol, although the packets I've seen with 0x81fd
-	  don't look like the ISMP packets described in RFC 2641/2642/2643.
+	  Maybe the problem is that the compilers used on AIX weren't
+	  defining "AIX", and that code was never compiled.  Libpcap uses
+	  "_AIX" to conditionally compile in the extra AIX BPF support;
+	  we'll do the same.
 
-2003-10-08 11:33  jmayer
+2003-11-21 08:30  guy
 
-	* packet-ftp.c:
+	* pcap-util.c:
 
-	  use g_htonl instead of htonl
+	  Pass "strncmp()" the right number of arguments.  (Why hasn't
+	  anybody reported this as a problem when compiling on AIX?
+	  Doesn't any compiler complain that "strncmp()" isn't being passed
+	  enough arguments?  GCC sure did in a test program I built on
+	  FreeBSD 3.4....)
 
-2003-10-08 12:29  sahlberg
+2003-11-21 04:01  guy
 
 	* packet-dcerpc.c:
 
-	  Change dissect_deferred_pointer() from being Ordo(n^2) into being
-	  Ordo(n)
+	  From Jamie Fournier: add the UUID to the Info column for datagram
+	  calls.
 
-	  Makes it slightly faster when n (the number of pointers) is
-	  >10.000
+2003-11-21 02:57  guy
 
-	  The mother of all dcerpc packets (containing one array of >10.000
-	  pointers) was a bit slow.
+	* AUTHORS, Makefile.am, Makefile.nmake, packet-dcerpc-rs_plcy.c:
 
-	  It is still slow but at least completes in out lifetime.
+	  From Jamie Fournier: DCERPC RS_PLCY support.
 
-2003-10-09 18:54  guy
+2003-11-21 02:48  guy
 
-	* packet-mmse.c, packet-wbxml.c, packet-wsp.c:
+	* packet-dcerpc-ndr.c, packet-dcerpc.h:
 
-	  From Biot Olivier:
+	  The 32-bit integer at the beginning of a context_handle is called
+	  "context_handle_attributes" in the DCE RPC 1.1 "nbase.idl".
+	  Rename our structure member accordingly.
 
-		  Rename WSP defines to avoid collisions with Windows
-	  defines.
+	  Note in a comment that we should perhaps display a context_handle
+	  as the attributes and UUID.
 
-		  Rename WSP dissector table for integer-value content
-	  types.
+	  Use "proto_tree_add_item()", not "proto_tree_add_bytes()", to put
+	  the context handle raw data into the protocol tree.
 
-		  Add WSP dissector table for literal content types.
+2003-11-20 23:34  guy
 
-		  Add WSP dissection registration for literal content
-	  types.
+	* packet-rtp.c, packet-rtp.h, gtk/Makefile.am, gtk/Makefile.nmake,
+	  gtk/rtp_stream.c, gtk/rtp_stream.h, gtk/rtp_stream_dlg.c:
 
-		  Register literal and extra WBXML content types to WSP
-	  dissector.
+	  From Lars Roland:
 
-		  Register textual MMS content type to WSP dissector.
+		  - added a pointer to raw rtp data to _rtp_info that can
+	  be used by
+		    taps;
 
-2003-10-09 18:57  guy
+		  - RTP packets are passed to the tap queue only if they
+	  are not
+		    error packets (so that you don't need to filter out
+	  ICMP
+		    packets)
 
-	* doc/README.developer:
+		  - use that pointer in rtp_stream, so it handles packets
+	  with
+		    padding, and should handle RTP packets fragmented
+	  across
+		    lower-level packets
 
-	  Add an item about prefixing #define names and enum names with a
-	  prefix to avoid name collisions.
+		  - moved rtp_stream from tap sources to normal files
+		    (prevents on-start-up registration of the rtp_stream
+	  tap
+		    listener)
 
-2003-10-09 20:52  guy
+		  - rtp_stream tap gets registered/unregistered with the
+	  "RTP
+		    Streams" dialog box
+		    i.e. the tap is registered as long as the dialog box is
+	  open.
+		    Alternatively, it is de-/registered on demand if RTP
+	  Analysis is
+		    called directly on a packet.
 
-	* packet-h225.c, packet-h245.c, packet-h450.c, packet-per.c,
-	  packet-per.h, packet-t38.c:
+		  - rtp_stream tap listener no longer uses a filter in
+	  dissection
+	  `	    and does not need to have a proto tree being built.
+	  (performance increase)
 
-	  Prefix OPTIONAL and NOT_OPTIONAL with ASN1_, to avoid collisions
-	  with other definitions of OPTIONAL (such as the one in Windows).
+		  - fixed: RTP Streams list will get updated in real time
+	  if the
+		    dialog box is open while a redissection takes place.
 
-2003-10-09 21:18  guy
+2003-11-20 23:21  guy
 
-	* doc/README.developer:
+	* packet-dcerpc-srvsvc.c:
 
-	  From Loïc Minier: information on how to do reassembly of PDUs
-	  atop TCP.
+	  #if 0-out some no-longer-used routines.
 
-2003-10-09 22:29  jmayer
+	  Use #if 0/#endif rather than comments to remove the code that
+	  used to refer to those routines - it's a bit more robust in the
+	  presence of comments (there aren't currently any, but, just in
+	  case somebody adds them...).
 
-	* tethereal.c:
+2003-11-20 22:50  guy
 
-	  In case of a parse error in the capture syntax, check whether it
-	  is valid display filter syntax. In case it is, warn about a
-	  possbile confusion of the two filter types.
+	* packet-wbxml.c:
 
-2003-10-09 22:35  guy
+	  From Olivier Biot: add support for media type
+	  "application/x-prov.syncset+wbxml".
 
-	* packet-h225.c, packet-h245.c, packet-h450.c, packet-per.c,
-	  packet-per.h, packet-t38.c:
+2003-11-20 22:38  guy
 
-	  Prefix the EXTENSION #defines with ASN1_ as well.
+	* packet-sip.c:
 
-2003-10-09 22:40  guy
+	  From Martin Mathieson:
 
-	* AUTHORS, packet-isakmp.c, doc/ethereal.pod:
+		  - Add some summary info to REGISTER requests and
+	  responses
+		  - Tidy up whitespace a little
 
-	  From Aki Immonen: add a numbre of vendor IDs.
+2003-11-20 22:24  guy
 
-2003-10-10 03:00  guy
+	* packet-wbxml.c:
 
-	* Makefile.am, acinclude.m4, capture-wpcap.c, configure.in,
-	  pcap-util-unix.c, pcap-util.c:
+	  From Olivier Biot:
 
-	  Use "pcap_findalldevs()" if present.
+		  * Fix an incorrectly #ifdef'ed return introduced by the
+	  previous
+		    patch
+		  * Remove the parsed_length parameter from the function
+	  signature
+		    of the tag and attribute dissectors - have it return
+	  the
+		    parsed length, instead
+		  * Throw ReportedBoundsError exceptions if we run past the
+	  end of
+		    the tvbuff, rather than putting a special entry in
+		  * Produce consistent and extensive debug logging (enabled
+	  by
+		    defining DEBUG_wbxml)
+		  * Get rid of the dozens of "#ifdef DEBUG ... #endif"
+	  constructs
+		    for debug logging
+
+2003-11-20 19:05  guy
 
-2003-10-10 03:04  guy
+	* epan/osi-utils.c:
 
-	* pcap-util-int.h:
+	  From Hannes Gredler: "print_system_id()" should call
+	  "print_system_id_buf()", not "print_nsap_net_buf()".	(What idiot
+	  made *that* mistake? :-))
 
-	  Use "pcap_findalldevs()" if present.
+2003-11-20 14:29  nneul
 
-2003-10-10 06:05  guy
+	* AUTHORS:
 
-	* acinclude.m4, capture-wpcap.c, pcap-util-unix.c, pcap-util.h,
-	  util.c:
+	  update proto list
 
-	  Get the version number of the libpcap/WinPcap with which we're
-	  running with "pcap_lib_version()", if available.
+2003-11-20 06:43  sharpe
 
-2003-10-10 08:39  sahlberg
+	* packet-dcerpc-srvsvc.c:
 
-	* gtk/: main.c, main.h, menu.c:
+	  Clean up some IDL and remove some structures that Windows of all
+	  flavors does not understand. SHARE_INFO_1004, 1005, 1006, 1007,
+	  and 1501 seem not to be understood by any, and it can screw up
+	  dissections.
 
-	  Updates to TimeReference Frames
+2003-11-20 05:20  guy
 
-	  Add a Goto Next/Previous  time reference menu option
+	* epan/proto.c:
 
-2003-10-10 08:52  sahlberg
+	  Use "tvb_get_string()" in "proto_tree_set_string_tvb()" - it
+	  won't leak memory if it throws an exception, as it checks whether
+	  the entire string is in the tvbuff *before* allocating a buffer
+	  for it, and that also means that if the length is absurdly large,
+	  an exception will be thrown, rather than the memory allocation
+	  failing.
 
-	* gtk/service_response_time_table.c:
+2003-11-20 05:04  guy
 
-	  Add a Find Next/Previous menu option to the
-	  ServiceResponseTimeStatistics dialog
+	* epan/resolv.c:
 
-2003-10-10 08:59  sahlberg
+	  On Mac OS X, disable the alarm/longjmp stuff for timing out name
+	  resolution queries - it runs a significant risk of crashing
+	  applications that do that.
 
-	* gtk/service_response_time_table.c:
+2003-11-20 04:28  gerald
 
-	  Add a ColorizeProcedure menu option to the ServiceResponseTime
-	  statistics dialog
+	* AUTHORS, doc/ethereal.pod:
 
-2003-10-10 09:48  guy
+	  Update Jakob Schlyter's address.
 
-	* Makefile.nmake, capture-wpcap.c, config.h.win32, config.nmake:
+2003-11-19 22:55  guy
 
-	  Configure whether we have WinPcap based on whether
-	  WINPCAP_VERSION is set in the config.nmake file.
+	* packet-wbxml.c:
 
-	  Configure whether we have pcap_findalldevs() based on whether
-	  WINPCAP_VERSION is 2.3 (if so, we don't) or 3.0 or 3.1 (if so, we
-	  do).
+	  From Olivier Biot: fix  the WBXML dissector in cases where
+	  dissection is interrupted because of a short frame or packets not
+	  being reassembled, and provide additional debug logging.
 
-	  WinPcap 3.0 has the new libpcap declarations of
-	  "pcap_lookupnet()" and "pcap_open_live()" in which the first
-	  argument is a "const char *" rather than a "char *"; declare the
-	  functions and pointers to them appropriately based on the version
-	  of WinPcap.
+2003-11-19 22:13  nneul
 
-	  If we don't have pcap_findalldevs(), don't declare a pointer to
-	  it, as we don't have a declaration of pcap_if_t.
+	* Makefile.am, Makefile.nmake, packet-sebek.c:
 
-	  We also need to refer to "pcap_freealldevs()", so make a pointer
-	  for it.
+	  Add dissector for the SEBEK kernel read() data capture and/or
+	  rootkit tool used as part of the HoneyNet project.
 
-	  "symbols[]" is a const array; make the pointer to elements in it
-	  a const pointer.
+	  Info at: http://project.honeynet.org/tools/sebek/
 
-	  Fix some typoes.
+2003-11-19 21:24  guy
 
-2003-10-10 10:02  guy
+	* packet-wsp.c:
 
-	* t35.c:
+	  From Olivier Biot:
 
-	  From Tomas Kukosa: updates from Annex to ITU Operational Bulletin
-	  No. 766 - 15.VI.2002), plus other fixes.
+		  correct (and document) the offset computation in the
+	  parameter()
+		  and parameter_value_q() functions.
 
-2003-10-10 10:04  guy
+	  From Loïc Minier:
 
-	* packet-isup.c:
+		  - a copy-paste with a missing replacement;
+		  - content-disposition uses parameters not q-values;
+		  - missing "+" for the calculation of parameter-offsets;
+		  - white-space cleanup;
+		  - comment fixes.
 
-	  Fix a typo.
+2003-11-19 09:58  guy
 
-2003-10-10 10:12  guy
+	* packet-isis-lsp.c, packet-isis-lsp.h:
 
-	* acinclude.m4, configure.in:
+	  Fix the ATT bit definitions.
 
-	  From Brad Hards: fix two problems:
+2003-11-19 09:49  guy
 
-		  1. the --without-adns case wasn't handled correctly;
+	* packet-cpfi.c:
 
-		  2.  the reporting at the end of the configure didn't deal
-	  with
-		  the case were the configure check failed (as in my
-	  example,
-		  where I only had the libadns package installed, not
-		  libadns-devel) - it reported that ADNS would be used.
+	  Get rid of unused variables.
 
-2003-10-10 10:16  guy
+2003-11-19 09:43  guy
 
-	* gtk/rtp_analysis.c:
+	* packet-wsp.c:
 
-	  Use PT_ values rather than raw numbers when checking for u-law
-	  and A-law payload.
+	  From Olivier Biot: get rid of no-longer-used routines.
 
-2003-10-10 10:25  guy
+2003-11-19 09:32  sahlberg
 
-	* packet-gtp.c:
+	* packet-tcp.c:
 
-	  From Michal Melerowicz: remove obsolete GTP'v0 CDR dissection.
+	  Update to TCP,   when we see a SYN|ACK packetm reset base_ack to
+	  the current ACK-1 so that it looks right when doing relative
+	  sequence numbers.  I.e. SEQ :    seq==0  ack==0      SEQ|ACK
+	  seq==0  ack==1      ACK      seq==1  ack==1
 
-2003-10-10 10:54  sahlberg
+	  This looks much more correct.
 
-	* packet-h245.c:
+	  This change also fixes the problem reported to ethereal-dev
+	  recently with "Follow TCP Stream" dropping the first character of
+	  the stream.
 
-	  From Martin Regner,  bugfix for H245.
+2003-11-19 07:13  sharpe
 
-	  The restricted character string for signalType is 1 character,
-	  not 1 to 128 characters.
+	* packet-dcerpc-wkssvc.c:
 
-2003-10-10 11:11  sahlberg
+	  Fix some IDL ...
 
-	* packet-dcerpc.c, packet-dcerpc.h, gtk/dcerpc_stat.c:
+2003-11-19 04:30  guy
 
-	  Updated the DCERPC service response time to also offer the menu
-	  to Filter, Find and Colorize selected procedures in the same way
-	  as SMB and ONC-RPC already does.
+	* packet-fcswils.c:
 
-2003-10-10 11:24  sahlberg
+	  From Dave Sclarsky: fix the offset for the Node WWN.
 
-	* gtk/dcerpc_stat.c:
+2003-11-19 04:26  guy
 
-	  Fix a bug for the DCERPC stats window where it forgot to print
-	  the title inside the window as it were supposed to.
+	* packet-cpfi.c:
 
-2003-10-10 13:33  jmayer
+	  Get rid of some unused arguments and variables.
 
-	* tethereal.c:
+	  Clean up the handling of the footer, so that it works even with
+	  short frames.
 
-	  Part two of different-filter-syntax-hackaround: Warn in case of
-	  capture filter usage where display filter syntax is required.
+2003-11-19 04:05  guy
 
-2003-10-10 21:13  guy
+	* packet-cpfi.c:
 
-	* acinclude.m4, tethereal.c:
+	  "uint" preferences should be "guint"s, not "int"s.
 
-	  Not all versions of libpcap have "pcap_compile_nopcap()"; use it
-	  only if we have it.
+2003-11-19 04:02  guy
 
-	  Not all versions of libpcap have DLT_LINUX_SLL, either; use
-	  DLT_EN10MB instead.
+	* packet-cpfi.c:
 
-2003-10-10 21:16  guy
+	  Fix a typo.
 
-	* AUTHORS, Makefile.am, Makefile.nmake,
-	  packet-diffserv-mpls-common.c, packet-diffserv-mpls-common.h,
-	  packet-ldp.c, packet-rsvp.c:
+2003-11-19 04:00  guy
 
-	  From Akira Endoh: Diffserv MPLS signaling protocol support.
+	* AUTHORS, Makefile.am, Makefile.nmake, packet-cpfi.c,
+	  doc/ethereal.pod:
 
-2003-10-10 21:31  guy
+	  From Dave Sclarsky: CPFI support.
 
-	* Makefile.nmake, doc/Makefile.nmake, epan/Makefile.nmake,
-	  epan/dfilter/Makefile.nmake, epan/ftypes/Makefile.nmake,
-	  gtk/Makefile.nmake, image/Makefile.nmake,
-	  packaging/nsis/Makefile.nmake, plugins/Makefile.nmake,
-	  plugins/artnet/Makefile.nmake, plugins/docsis/Makefile.nmake,
-	  plugins/giop/Makefile.nmake, plugins/gryphon/Makefile.nmake,
-	  plugins/lwres/Makefile.nmake, plugins/megaco/Makefile.nmake,
-	  plugins/mgcp/Makefile.nmake, plugins/pcli/Makefile.nmake,
-	  plugins/rtnet/Makefile.nmake, tools/Makefile.nmake,
-	  tools/lemon/Makefile.nmake, wiretap/Makefile.nmake:
+2003-11-19 03:53  guy
 
-	  Give every Makefile.nmake file a "distclean" rule, and have
-	  "distclean" recurse into subdirectories doing "nmake -f
-	  Makefile.nmake distclean".
+	* packet-smb-browse.c, packet-smb-logon.c, packet-smb-mailslot.c:
 
-	  Have "nmake -f Makefile.nmake clean" not remove stuff that "make
-	  clean" doesn't remove (such as Flex/Bison output and config.h
-	  files) - and have "nmake -f Makefile.nmake distclean" remove
-	  stuff that "make distclean" removes, including
-	  "tethereal-tap-register.c" and "ethereal-tap-register.c".
+	  "dissect_mailslot_browse()", "dissect_mailslot_lanman()", and
+	  "dissect_smb_logon()" always return TRUE, so just get rid of
+	  their return value.
 
-2003-10-10 21:35  guy
+	  "call_dissector()" automatically calls the data dissector if the
+	  protocol for the dissector being called is disabled, so we don't
+	  have to check its result and call the data dissector if it
+	  returns 0.
 
-	* README.win32:
+2003-11-19 02:58  guy
 
-	  Mention gtk/ethereal-tap-register.c.
+	* packet-mmse.c:
 
-2003-10-10 22:52  sahlberg
+	  Register the MMSE dissector with the "http.content_type"
+	  dissector table, rathe than as a heuristic subdissector of HTTP.
 
-	* packet-tcp.c:
+2003-11-19 01:50  guy
 
-	  From Matthijs Melchior Small change to the TCP sequence number
-	  analysis and relative sequence number code so that it plays a bit
-	  nicer with captures generated by text2pcap.
+	* gtk/supported_protos_dlg.c:
 
-	  Change the criterion used to initialize the base sequence and ack
-	  numbers to set these base offsets where it detects that the
-	  bookkeeping structures are NULL (as in no previous packet seen
-	  for this session) instead of using a hardcoded magic number 0,
-	  which might actually occur in normal captures.
+	  Use "OK" for the "OK" button, as we do elsewhere.
 
-2003-10-11 00:10  guy
+2003-11-19 01:45  guy
 
 	* packet-wsp.c:
 
-	  "add_content_type()" returns either a numerical content type and
-	  a null content type pointer or a 0 numerical content type and a
-	  non-null content type pointer; if the content type is numerical,
-	  we cannot match on the string content type (as we'd dereference a
-	  null pointer) and, if the content type is a string, we should not
-	  match on the numerical content type (as the value is bogus).
-
-	  We also have to call "add_content_type()" to get the content type
-	  regardless of whether we're building a protocol tree, as we need
-	  to call subdissectors regardless of whether we're building a
-	  protocol tree.
-
-	  We also need to set the columns regardless of whether we're
-	  building a protocol tree.
-
-2003-10-11 11:23  sahlberg
-
-	* gtk/io_stat.c:
-
-	  Massive update of the _draw() routine in iostat in preparation of
-	  more features.
-
-	  The most visible changes are	 graps drawn as sawtoots instead of
-	  bars.    relative times (mainly used for response times
-	  FT_RELATIVE_TIME) will be plotted as time measured in "s" "ms" or
-	  "us" on the y axis
-
-	  future updates may be  smoothed graphs  and better relative time
-	  support
-
-2003-10-11 21:49  jmayer
-
-	* capture.c:
-
-	  Part 3 of display vs. capture syntax confusion patches: Detect
-	  probable use of display filter syntax where capture filter syntax
-	  need to be used in Ethereal.
-
-2003-10-11 23:17  sahlberg
-
-	* gtk/io_stat.c:
-
-	  minor cleanup and update for printing the y scale as time in
-	  units of s,ms and us
+	  From Olivier Biot:
 
-2003-10-12 04:20  sahlberg
+		  Implement quoted-string parsing by appending the closing
+	  quote
+		  to the string value.	If the quoted-string value does not
+	  start
+		  with a quote, dissection resumes but a warning is
+	  displayed.
+		  Textual headers are assumed to be 0x00 terminated
+	  (otherwise we
+		  have a malformed packet or a short frame).
 
-	* gtk/io_stat.c:
+		  The Trailer header is also dissected now, and the charset
+		  parameter dissection has been rewritten.
 
-	  Further updates/rewrite of iostat
+		  The Content-ID header dissector was registered at the
+	  place of
+		  the Trailer dissector; this has also been fixed.
 
-	  getting closer to add useful features
+2003-11-19 01:39  guy
 
-2003-10-12 12:29  tpot
+	* packet-ansi_map.c:
 
-	* packet-dcerpc-oxid.c:
+	  From Michael Lum: fix param_digits() subtree for "Nature of
+	  Number".
 
-	  Subdissector name updates from Jean-Baptiste Marchand.
+2003-11-19 00:10  ulfl
 
-2003-10-12 16:03  sharpe
+	* gtk/gui_prefs.c:
 
-	* packet-dcerpc-oxid.c:
+	  redraw supported_protos dialog, if GUI prefs changed
 
-	  Fix some C++ comments ...
+2003-11-19 00:04  ulfl
 
-2003-10-13 19:04  guy
+	* gtk/menu.c:
 
-	* epan/Makefile.nmake, tools/Makefile.nmake:
+	  added supported_protos menu item
 
-	  From Ulf Lamping: add some missing "cd .."s.	(I guess the
-	  current directory in nmake files persists across targets;
-	  presumably all commands, or, at least, all "cd" commands, are run
-	  in the same process.)
+2003-11-19 00:02  ulfl
 
-2003-10-13 20:53  guy
+	* gtk/Makefile.am:
 
-	* packet-diffserv-mpls-common.c, packet-diffserv-mpls-common.h:
+	  added supported_protos_dlg
 
-	  Define "phbit_bit{14,15}_vals" in packet-diffserv-mpls-common.c,
-	  not in the header file; just declare it in the header file - that
-	  way, there's one shared version of both tables, and we don't get
-	  compiler warnings because they're not actually used in
-	  packet-diffserv-mpls-common.c.
+2003-11-19 00:01  ulfl
 
-2003-10-14 00:40  guy
+	* gtk/Makefile.nmake:
 
-	* epan/resolv.c:
+	  added supported_protos_dlg.obj
 
-	  From Matthias Melchior: parenthesize check for RESOLV_CONCURRENT.
+2003-11-18 23:56  ulfl
 
-	  Get rid of "!= 0" check - other tests for a bit being on in
-	  g_resolv_flags don't do that.
+	* gtk/: supported_protos_dlg.c, supported_protos_dlg.h:
 
-2003-10-14 00:45  guy
+	  seperated "supported protocols and protocol fields" dialog,
+	  coming from help_dlg.c/.h
 
-	* packet-dcerpc.c:
+2003-11-18 23:40  ulfl
 
-	  "dce_try_handoff()" isn't necessarily passed a non-null
-	  "auth_info" argument - don't dereference it if it's null.
+	* Makefile.nmake:
 
-2003-10-14 01:18  guy
+	  generate "static" help content
 
-	* AUTHORS, Makefile.am, configure.in,
-	  packaging/nsis/Makefile.nmake, packaging/nsis/ethereal.nsi,
-	  plugins/Makefile.am, plugins/Makefile.nmake, plugins/acn/AUTHORS,
-	  plugins/acn/COPYING, plugins/acn/ChangeLog, plugins/acn/INSTALL,
-	  plugins/acn/Makefile.am, plugins/acn/Makefile.nmake,
-	  plugins/acn/NEWS, plugins/acn/acn.h, plugins/acn/moduleinfo.h,
-	  plugins/acn/packet-acn.c:
+2003-11-18 23:11  ulfl
 
-	  From Erwin Rol: ACN support.
+	* help/: capture_filters.txt, convert-include.sh, Makefile.nmake,
+	  display_filters.txt, faq.txt, overview.txt, well_known.txt:
 
-2003-10-14 07:36  guy
+	  "static content" and make environment of redesigned online help
 
-	* acinclude.m4:
+2003-11-18 19:56  guy
 
-	  Do the tests for various extra "-L" and "-l" flags that might be
-	  needed with "-lsnmp" in a loop; this fixes up the configure
-	  script's operation on Solaris.
+	* packet-bootp.c:
 
-2003-10-14 08:56  guy
+	  From Greg Morris:
 
-	* plugins/acn/.cvsignore:
+		  dissect the RFC 2610 DHCP options for SLP;
 
-	  Add a .cvsignore file.
+		  dissect Novell-specific options 85, 86, and 87.
 
-2003-10-14 09:03  sahlberg
+2003-11-18 19:46  guy
 
-	* gtk/io_stat.c:
+	* packet-tcp.c:
 
-	  update to iostat
+	  From Lars Roland: the window size passed to
+	  "tcp_analyze_sequence_number()" is a "guint32", as it might be
+	  scaled - make the arugment a "guint32" as well.
 
-	  add different styles to draw the plot
+2003-11-18 19:36  guy
 
-2003-10-14 09:11  guy
+	* plugins/: Xass-list, Xplugin_api.c, Xplugin_api.h,
+	  Xplugin_api_decls.h, Xplugin_table.h:
 
-	* AUTHORS, packet-eapol.c:
+	  Get rid of the second "col_set_fence()" entry.
 
-	  From Motonori Shindo: Key Information dissection in EAPOL-Key
-	  messages.
+2003-11-18 19:34  guy
 
-2003-10-14 09:15  sahlberg
+	* plugins/plugin_api_list.c:
 
-	* gtk/io_stat.c:
+	  The second of the "col_set_fence()" entries is the one that
+	  should've been removed; that way, the ordinals of existing
+	  entries in the function vector is preserved.
 
-	  Simplify handling of tick val remove unnessecary code and
-	  structures	cleanup
+2003-11-18 19:31  guy
 
-2003-10-14 09:27  sahlberg
+	* plugins/Makefile.nmake:
 
-	* gtk/io_stat.c:
+	  From Lars Roland: put missing "enttec" into "all" list.
 
-	  further cleanup
+2003-11-18 19:29  ulfl
 
-	  remove the redundant pixels_per_tick structure
+	* gtk/capture_dlg.c:
 
-2003-10-14 09:55  sahlberg
+	  removed MSVC compiler warnings
 
-	* gtk/io_stat.c:
+2003-11-18 19:28  guy
 
-	  further cleanup remove redundant structure for yscale_max
+	* packet-ntp.c:
 
-2003-10-14 10:01  sahlberg
+	  From Matthias Drochner:
 
-	* gtk/io_stat.c:
+		  extensions need only be a multiple of 4 bytes in length,
+	  not 8
+		  bytes;
 
-	  remove the redundant count_type structure as part of the iostat
-	  cleanup
+		  put in a missing increment of "offset".
 
-2003-10-14 10:34  sahlberg
+2003-11-18 19:27  ulfl
 
-	* gtk/io_stat.c:
+	* gtk/print_dlg.c:
 
-	  update to iostat, added new plot style: filled bar
+	  removed unused vars
 
-2003-10-14 17:50  guy
+2003-11-18 19:26  ulfl
 
-	* packet-isup.c:
+	* plugins/plugin_api_list.c:
 
-	  From Jeff Morriss: note in the Protocol column that we're
-	  assuming the ITU dialect of ISUP.
+	  col_set_fence occurs twice, removed one appearance
 
-2003-10-14 21:24  jmayer
+2003-11-18 19:20  guy
 
-	* packet-sip.c:
+	* epan/filesystem.h:
 
-	  Update list of SIP RFCs
+	  From Tomas Kukosa: add "extern"s so that the plugin API stuff
+	  works.
 
-2003-10-14 21:26  jmayer
+2003-11-18 19:20  ulfl
 
 	* packet-sdp.c:
 
-	  Add a comment for (maybe) later connection handling
-
-2003-10-14 23:20  guy
-
-	* AUTHORS, prefs.c, prefs.h, doc/ethereal.pod, gtk/file_dlg.c,
-	  gtk/gui_prefs.c, gtk/main.c:
-
-	  From Ian Schorr:
-
-		  Add a preference to control whether the "File > Open"
-	  dialog box
-		  should start out in the last directory in which it looked
-	  - and
-		  save that in the preferences file across invocations - or
-	  should
-		  always start out in a user-specified directory, and add
-	  another
-		  preference to specify that directory.
-
-		  Write out section name comments into the preferences
-	  file.
-
-	  Clean up white space a bit.
-
-2003-10-14 23:42  guy
+	  includes changed, to be working with MSVC also
 
-	* gtk/file_dlg.c:
+2003-11-18 19:18  ulfl
 
-	  Clean up the stuff that fills in the starting directory.
+	* plugins/rdm/Makefile.nmake:
 
-2003-10-15 02:13  gerald
+	  corrected cut and paste errors, makefile was unchanged copied
+	  from rtnet
 
-	* doc/: ethereal-filter.pod.template, ethereal.pod, tethereal.pod:
+2003-11-18 08:59  guy
 
-	  Move the filter syntax description to the ethereal-filter pod
-	  page.
-
-2003-10-15 08:08  guy
-
-	* packet-dcerpc-spoolss.c:
-
-	  We got rid of the wrong hf_devmode_driver_extra; bring the wrong
-	  one back, and get rid of the one that shouldn't be kept around.
+	* plugins/: Xass-list, Xplugin_api.c, Xplugin_api.h,
+	  Xplugin_api_decls.h, Xplugin_table.h:
 
-2003-10-15 08:25  sahlberg
+	  Add "col_get_writable()", "col_set_writable()", and
+	  "col_set_fence()" to the plugin API list.
 
-	* gtk/io_stat.c:
+2003-11-18 08:58  guy
 
-	  remove an extra ';' to keep visual studio happy
+	* plugins/Xplugin_table.h:
 
-2003-10-15 08:41  sahlberg
+	  Export "protocol_t" as an opaque type.
 
-	* gtk/io_stat.c:
+	  Make "proto_is_protocol_enabled()" and
+	  "proto_get_protocol_short_name()" take a "protocol_t *" as an
+	  argument, so they don't have to look up the "protocol_t" - this
+	  will probably speed them up considerably, and they're called on
+	  almost every dissector handoff.
 
-	  Remove the unnecassary HashTable and use
-	  gtk_object_[get|set]_data to keep track of which io structure a
-	  certain draw area is associated with.
+	  Get rid of a number of "proto_is_protocol_enabled()" calls that
+	  aren't necessary (dissectors called through handles, including
+	  those called through dissector tables, or called as heuristic
+	  dissectors, aren't even called if their protocol isn't enabled).
 
-2003-10-15 09:18  guy
+	  Change some direct dissector calls to go through handles.
 
-	* packet-gtp.c:
+2003-11-18 08:55  guy
 
-	  Fix the XXX_to_str routines
+	* plugins/plugin_api_list.c:
 
-		  1) not to rely on their static buffers being initialized;
+	  Add "col_get_writable()", "col_set_writable()", and
+	  "col_set_fence()" to the plugin API list.
 
-		  2) not to bother checking whether a BCD digit is a valid
-	  digit
-		     or not (index into a 16-element array, instead);
+2003-11-18 08:13  guy
 
-		  3) to work correctly, in the case of "id_to_str()".
+	* configure.in, epan/configure.in, wiretap/configure.in:
 
-2003-10-15 13:10  sahlberg
+	  From Olivier Biot: add a "--enable-profile-build" flag to request
+	  that "-pg" binaries be built.
 
-	* doc/ethereal.pod, gtk/io_stat.c:
+2003-11-18 08:04  guy
 
-	  Client LOAD measurement for io-stat
+	* packet-http.c:
 
-	  See manpage (hopefully manpage does not reformat my nice ascii
-	  graph)
+	  Content-Type values are case-insensitive; force them to lower
+	  case (subdissectors must register their Content-Type values as
+	  all-lower-case).
 
-	  While Service Response Times	 and the MIN/MAX/AVG thing in
-	  io-stat are measurements on the server load.	 The new
-	  measurement type LOAD is a measurement of Client LOAD.
+	  Don't include parameters in the value we look up in the
+	  Content-Type subdissector table.
 
-	  Or rather, it is an attempt to measure client LOAD by measuring
-	  how much concurrency in its requests the client generates. It the
-	  client is slow in starting new i/o when a previous i/o has
-	  completed,  this willb e indicated by the concurrency being
-	  lowered.
+2003-11-18 07:49  guy
 
-	  it is an experiment.	i am not aware of any other attempts in
-	  deducing client workload from looking at captures.
+	* packet-http.c:
 
-2003-10-15 19:40  guy
+	  Add an "http.content_type" dissector table,
 
-	* AUTHORS, Makefile.am, gtk/Makefile.am, gtk/Makefile.nmake,
-	  gtk/main.c, gtk/menu.c, gtk/toolbar.c, gtk/toolbar.h,
-	  image/toolbar/capture_24.xpm, image/toolbar/cfilter_24.xpm,
-	  image/toolbar/dfilter_24.xpm, image/toolbar/stock_close_24.xpm,
-	  image/toolbar/stock_colorselector_24.xpm,
-	  image/toolbar/stock_help_24.xpm,
-	  image/toolbar/stock_jump_to_24.xpm,
-	  image/toolbar/stock_open_24.xpm,
-	  image/toolbar/stock_preferences_24.xpm,
-	  image/toolbar/stock_print_24.xpm,
-	  image/toolbar/stock_refresh_24.xpm,
-	  image/toolbar/stock_right_arrow_24.xpm,
-	  image/toolbar/stock_save_24.xpm,
-	  image/toolbar/stock_search_24.xpm,
-	  image/toolbar/stock_stop_24.xpm:
+	  Process the Content-Type header, and, if one is found, look up
+	  its value in the aforementioned dissector table.
 
-	  From Ulf Lamping: toolbar.
+2003-11-18 07:08  guy
 
-2003-10-15 19:57  guy
+	* packet-ses.c:
 
-	* AUTHORS, packet-socks.c, doc/ethereal.pod:
+	  Handle concatenated SPDUs.
 
-	  From David E. Weekly: move the code to put the SOCKS version up,
-	  so it's done in common code.
+	  Handle the Serial Number parameter.
 
-2003-10-15 20:07  guy
+2003-11-18 06:31  guy
 
-	* packet-socks.c:
+	* config.guess, config.sub:
 
-	  Put back a missing comment.
+	  Latest config.guess and config.sub from
 
-	  Get rid of some extra blanks at the end of the text for some
-	  protocol tree items.
+		  http://savannah.gnu.org/projects/config/
 
-	  Use %u, not %d, to format unsigned quantities.
+2003-11-18 06:06  ulfl
 
-	  Make the "Client Authentication Methods" item's length cover all
-	  the authentication methods, not just the count of methods.
+	* gtk/capture_info_dlg.c:
 
-2003-10-15 22:00  guy
+	  removed MSVC warnings
 
-	* packet-ip.c:
+2003-11-18 04:19  gerald
 
-	  From Giles Scott: make some items in the ICMP protocol tree named
-	  fields.
+	* doc/ethereal.pod:
 
-2003-10-15 22:34  guy
+	  Document the capture filter changes that were just checked in.
 
-	* gtk/toolbar.c:
+2003-11-18 04:16  gerald
 
-	  Note that we might want to use the GTK+ 2.x stock icon mechanism
-	  when building for GTK+ 2.x.
+	* util.c, util.h, gtk/main.c:
 
-	  Fix the callback for the "Print" button.
+	  Check the environment variables SSH_CONNECTION, SSH_CLIENT,
+	  REMOTEHOST, DISPLAY, and CLIENTNAME (in that order).	If any of
+	  them are set, create a capture filter that excludes their traffic
+	  and set it as the default.  The longer filters should be
+	  efficient without being overly long; they may need some tweaking.
 
-2003-10-15 22:37  guy
+2003-11-18 00:46  ulfl
 
 	* gtk/toolbar.c:
 
-	  "Find frame" can do more than search with a display filter.
-
-2003-10-16 00:45  guy
-
-	* gtk/: file_dlg.c, main.c:
-
-	  There doesn't seem to be any need to set "cfile.dfilter" to the
-	  contents of the filter text entry when reloading the file, and:
-
-		  1) that doesn't work with the toolbar "reload" button
-	  (the
-		     widget passed in for that button doesn't have a
-		     E_DFILTER_TE_KEY data item pointing to the text
-	  entry);
-
-		  2) that causes the Tools > Summary dialog box to report
-	  what
-		     you've typed in that box, not the filter that's
-	  actually in
-		     effect (i.e., it causes "cfile.dfilter" to reflect
-	  what's
-		     been typed, not what's been applied);
-
-	  so don't bother doing so.  That also means that the
-	  "/File/Reload" menu item doesn't need a E_DFILTER_TE_KEY data
-	  item, so don't give it one.
-
-2003-10-16 06:47  guy
-
-	* packet-gtp.c:
-
-	  Get rid of a value_string table that was used by now-deleted
-	  code.
-
-2003-10-16 09:55  guy
-
-	* packet-gtp.c:
-
-	  From Michal Melerowicz:
-
-		  merge gtpv0 and gtpv1 dissectors into one;
-
-		  fix up XXX_to_str routines.
+	  removed some warnings
 
-2003-10-16 18:14  guy
+2003-11-17 22:56  sahlberg
 
-	* packet-ansi_map.c:
-
-	  From Michael Lum: add missing g_free() calls.
-
-2003-10-16 18:15  guy
-
-	* packet-tcap.c:
-
-	  From Michael Lum: fixed some formatting and removed unused code.
-
-2003-10-16 20:56  oabad
-
-	* gtk/: main.c, toolbar.c, toolbar.h:
-
-	  - use stock icons in the toolbar when using GTK2 - use GTK1/2
-	  compatibility macros [GS]ET_OBJECT_DATA where needed - add a
-	  set_toolbar_object_data() function which associates the display
-	  filter entry (from the main window) with the E_DFILTER_TE_KEY for
-	  the	open and reload buttons (it is needed by the open and
-	  reload   callbacks). The function is called in
-	  create_main_window() - reindent
+	* epan/to_str.c:
 
-2003-10-16 21:04  oabad
+	  updates to ip_to_str[_buf]() to make them faster.
 
-	* gtk/toolbar.c:
+	  This functin was in a recent profiling run the fifth most
+	  expensive function.  It consumed ~3.5% of the total cpu in that
+	  run.
 
-	  Put back Guy's changes from r1.3 (print callback, and "Find
-	  frame..." tooltip).
+	  Using a static table this function is now just over twice as fast
+	  as previously.
 
-2003-10-16 21:19  guy
+	  My testrun calling it 10.000.000 times in a loop changed the
+	  execution time from ~17.3s ro ~8.2s
 
-	* prefs.c, prefs.h, gtk/gui_prefs.c, gtk/toolbar.c, gtk/toolbar.h:
+	  If this test was representative for the speedup and if ~3.5 in
+	  the profile run was representative for normal cases this should
+	  make ethereal ~1.8% faster.  Woohoo.	(low hanging fruit)
 
-	  From Ulf Lamping: add a GUI preference item to control the
-	  toolbar style.
+2003-11-17 22:34  guy
 
-	  Make the style text arrays static, as nobody uses them outside
-	  prefs.c.
+	* plugins/: acn/Makefile.am, artnet/Makefile.am, asn1/Makefile.am,
+	  gryphon/Makefile.am, mgcp/Makefile.am, pcli/Makefile.am,
+	  rtnet/Makefile.am:
 
-	  Use FALSE and TRUE for the values for the Boolean controlling the
-	  highlighting style.
+	  Fix the copyright notices (Gerald has the copyright on Ethereal).
 
-	  Note that we're now using stock icons in the toolbar in GTK+ 2.x.
+2003-11-17 22:31  guy
 
-	  Put back the resizing of elements in the top-level container, at
-	  least for GTK+ 1.2[.x]; otherwise, the toolbar's height never
-	  gets smaller, even if you change the style in such a way as to
-	  reduce the height of the elements (icons+text -> icons or text,
-	  icons -> text).
+	* plugins/giop/Makefile.am:
 
-	  Make some routines and variables not used outside gtk/toolbar.c
-	  static.
+	  Tweak the introductory comment.
 
-2003-10-17 06:06  oabad
+2003-11-17 22:29  guy
 
-	* gtk/toolbar.c:
+	* plugins/: enttec/.cvsignore, rdm/.cvsignore:
 
-	  Change the comments about toolbar elements resizing as it is not
-	  necessary in GTK+ 2.x
+	  Fix the .cvsignore files.
 
-2003-10-17 07:45  sahlberg
+2003-11-17 22:26  guy
 
-	* packet-h225.c:
+	* plugins/rdm/packet-rdm.c:
 
-	  change the h225 emailid from a sequence of bytes to a string
+	  From Erwin Rol: check the checksum in RDM packets.
 
-2003-10-17 08:01  sahlberg
+2003-11-17 22:09  guy
 
-	* packet-h225.c:
-
-	  destinationType is an  EndPointType and NOT an EndPoint
-
-2003-10-17 08:14  guy
-
-	* gtk/toolbar.c:
+	* configure.in:
 
-	  Further expand a comment.
+	  From Loïc Minier: append $(EXEEXT) to the names of binaries, so
+	  that the configure script generates the right executable image
+	  file names on Cygwin.
 
-2003-10-17 17:20  oabad
+2003-11-17 21:52  guy
 
-	* gtk/proto_draw.c:
+	* packet-sdp.c:
 
-	  Use gtk_tree_view_expand_to_path() in gtk2 code to expand a node,
-	  as it also expands parent nodes if necessary.
+	  From Lars Roland: use the c= and m= information to set up
+	  conversations to use RTP/RTCP.
 
-2003-10-17 17:28  oabad
+2003-11-17 20:57  guy
 
-	* gtk/: menu.c, toolbar.c:
+	* AUTHORS, Makefile.am, configure.in,
+	  packaging/nsis/Makefile.nmake, packaging/nsis/ethereal.nsi,
+	  plugins/Makefile.am, plugins/Makefile.nmake,
+	  plugins/artnet/ChangeLog, plugins/artnet/moduleinfo.h,
+	  plugins/artnet/packet-artnet.c, plugins/enttec/.cvsignore,
+	  plugins/enttec/AUTHORS, plugins/enttec/COPYING,
+	  plugins/enttec/ChangeLog, plugins/enttec/INSTALL,
+	  plugins/enttec/Makefile.am, plugins/enttec/Makefile.nmake,
+	  plugins/enttec/NEWS, plugins/enttec/moduleinfo.h,
+	  plugins/enttec/packet-enttec.c, plugins/rdm/.cvsignore,
+	  plugins/rdm/AUTHORS, plugins/rdm/COPYING,
+	  plugins/rdm/Makefile.am, plugins/rdm/Makefile.nmake,
+	  plugins/rdm/moduleinfo.h, plugins/rdm/packet-rdm.c:
+
+	  From Erwin Rol: initial ENTTEC support and RDM support.
+
+2003-11-17 20:00  guy
+
+	* gtk/capture_info_dlg.c:
+
+	  Not all compilers allow array/structure/union members of
+	  automatic variables to be initialized to non-constant values (C89
+	  says that "All the expressions in an initializer for an object
+	  that has static storage duration or in an initializer list for an
+	  object that has aggregate or union type shall be constant
+	  expressions"; presumably the intent of the former was to avoid
+	  run-time initialization and of the latter was to let the
+	  initialization be done by copying from a compile-time-created
+	  blob of memory), so we have to initialize "info->counts" by hand.
+
+2003-11-17 19:40  guy
 
-	  From Olivier Biot : remove packet capture code when compiling
-	  without libpcap.
+	* capture.c:
 
-2003-10-17 21:26  guy
+	  Fix a typo.
 
-	* packet-dhcpv6.c:
+2003-11-17 18:23  ulfl
 
-	  From Shinsuke Suzuki:
+	* etypes.h, packet-ethertype.c:
 
-		  - support RFC3319
-		  - update the RFC/i-d name
-		  - change the DHCPv6 option numbers to catch up with the
-		    latest IANA's assignment status
+	  added ETHERTYPE_PROFINET
 
-2003-10-17 21:27  guy
+2003-11-17 00:27  guy
 
-	* packet-pim.c:
+	* gtk/capture_info_dlg.c:
 
-	  From Shinsuke Suzuki: support Address List Option, newly
-	  introduced in draft-ietf-pim-sm-v2-new-08.txt.
+	  "capture()" takes a "struct pcap_stat *" argument, so you need to
+	  include <pcap.h> before including "capture.h".
 
-2003-10-17 21:28  guy
+	  Include "dlg_utils.h" to declare "dlg_set_cancel()".
 
-	* AUTHORS:
+2003-11-17 00:26  guy
 
-	  Update credits for Shinsuke Suzuki.
+	* capture.c:
 
-2003-10-17 22:59  guy
+	  Get rid of a no-longer-used variable.
 
-	* packet-ndps.c:
+2003-11-16 23:17  guy
 
-	  From Greg Morris: don't add zero-length items to the tree, add
-	  some missing alignment adjustments, fix a call to
-	  "address_item()" to set the offset to its return value rather
-	  than adding its return value to the offset, handle an object type
-	  of 2 (counted list of objects) in an event handling profile,
-	  advance the offset past length values after processing the length
-	  value.
+	* Makefile.am, disabled_protos.c, packet-alcap.c,
+	  packet-ansi_637.c, packet-ansi_683.c, packet-ansi_a.c,
+	  packet-ansi_map.c, packet-atalk.c, packet-clnp.c,
+	  packet-dcerpc.c, packet-dcerpc.h, packet-dvmrp.c, packet-fcip.c,
+	  packet-giop.c, packet-giop.h, packet-gsm_a.c, packet-gsm_sms.c,
+	  packet-gssapi.c, packet-gssapi.h, packet-iscsi.c,
+	  packet-mrdisc.c, packet-msnip.c, packet-pim.c, packet-ppp.c,
+	  packet-rpc.c, packet-rpc.h, packet-smb-browse.c,
+	  packet-smb-browse.h, packet-smb-logon.c, packet-smb-logon.h,
+	  packet-smb-mailslot.c, packet-smb-pipe.c, packet-snmp.c,
+	  packet-tpkt.c, prefs.c, epan/packet.c, epan/proto.c,
+	  epan/proto.h, gtk/dfilter_expr_dlg.c, gtk/help_dlg.c,
+	  gtk/proto_dlg.c, plugins/plugin_api_list.c:
 
-	  Fix some additional "address_item()" calls, and put the
-	  aforementioned length values into the protocol tree.
+	  Export "protocol_t" as an opaque type.
 
-2003-10-17 23:43  guy
+	  Make "proto_is_protocol_enabled()" and
+	  "proto_get_protocol_short_name()" take a "protocol_t *" as an
+	  argument, so they don't have to look up the "protocol_t" - this
+	  will probably speed them up considerably, and they're called on
+	  almost every dissector handoff.
 
-	* packet-fr.c:
+	  Get rid of a number of "proto_is_protocol_enabled()" calls that
+	  aren't necessary (dissectors called through handles, including
+	  those called through dissector tables, or called as heuristic
+	  dissectors, aren't even called if their protocol isn't enabled).
 
-	  Fix up some comments.
+	  Change some direct dissector calls to go through handles.
 
-2003-10-17 23:44  guy
+2003-11-16 23:11  sahlberg
 
-	* packet-lmi.c:
+	* AUTHORS, Makefile.am, Makefile.nmake, h225-persistentdata.c,
+	  h225-persistentdata.h, packet-h225.c, packet-h225.h,
+	  tap-h225rassrt.c, doc/ethereal.pod, doc/tethereal.pod,
+	  gtk/Makefile.am, gtk/Makefile.nmake, gtk/h225_ras_srt.c:
 
-	  Fix a URL.
+	  From Lars Roland   Service Response times for H225 RAS
 
-2003-10-18 18:46  guy
+2003-11-16 22:33  sahlberg
 
-	* packet-rip.c:
+	* packet-h450.c:
 
-	  From Emanuele Caratti:
+	  Bugfix for h450.
 
-		  add the dissection of RIPv2 packets with Keyed Message
-	  Disest
-		  Authentication (RFC2082);
+	  The hf field we give to dissect_per_object_string is a FT_STRING
+	  and not a FT_BYTES.  This caused ethereal to abort with an assert
+	  failure.
 
-		  fix a small bug - the Auth entry must be the first of the
-	  RTE in
-		  a RIP packet.
+	  This fixes the issue that Keith French recently brought up in his
+	  mailing to ethereal-users
 
-2003-10-19 17:30  guy
+2003-11-15 23:58  guy
 
-	* packet-tacacs.c, packet-tacacs.h:
+	* packet-wsp.c:
 
-	  From Emanuele Caratti:
+	  From Olivier Biot: fix the offset for the Encoding-Version header
+	  in the general form.
 
-		  add support for multiple encryption keys, one per
-	  conversation,
-		  in a single capture;
+2003-11-15 23:48  guy
 
-		  add some fields in the accounting dissection.
+	* gtk/menu.c:
 
-2003-10-20 06:06  oabad
+	  From Michael Lum: add <shift><control> hotkeys for Protocols and
+	  Preferences.
 
-	* gtk/gui_prefs.c:
+2003-11-15 11:46  ulfl
 
-	  In fileopen_dir_changed_cb() : - return FALSE immediately if the
-	  text entry is empty ; - return FALSE at the end of the function
-	  so that the signal is sent to   the entry. It avoids a Gtk-ERROR
-	  (and an abort) : file: gtkentry.c: line 4338 (blink_cb):
-	  assertion failed: GTK_WIDGET_HAS_FOCUS(entry))
+	* gtk/menu.c:
 
-2003-10-20 19:07  guy
+	  GTK2 specific: use some Ethereal specific stock icons generated
+	  in toolbar.c
 
-	* gtk/toolbar.c:
+2003-11-15 11:44  ulfl
 
-	  From Ulf Lamping:
+	* gtk/: toolbar.c, toolbar.h:
 
-		  put toolbar separators in the right places;
+	  handlebar removed, GTK2 specific: generate some Ethereal specific
+	  stock icons and use them in the toolbar
 
-		  get rid of the "Capture Stop" button in Win32.
+2003-11-15 11:39  ulfl
 
-2003-10-20 19:13  guy
+	* config.nmake:
 
-	* packet-isup.c:
-
-	  From Anders Broman:
+	  added gdk_pixbuf to GTK2 specific libs to link (needed in
+	  toolbar.c)
 
-		  added decoding of Element IWFA(NSAP address) in IANA ICP
-	  format;
+2003-11-15 10:06  ulfl
 
-		  fixed decoding of CODEC LIST.
+	* file.c:
 
-2003-10-20 19:25  guy
+	  removed no longer used gtk includes
 
-	* packet-laplink.c:
+2003-11-15 08:53  ulfl
 
-	  From Brad Hards: support TCP desegmentation.
+	* gtk/capture_info_dlg.c:
 
-2003-10-20 20:18  guy
+	  seperated capture dialog from capture.c and put into new
+	  gtk/capture_info_dlg.c, added some tiny bars, that will indicate
+	  the percentage, added ESC key handler and use of GTK_STOCK button
 
-	* packet-dcerpc-epm.c:
+2003-11-15 08:50  ulfl
 
-	  From Jim McDonough: add the text sservice name for a UUID/version
-	  string at the top level of a floor, falling back to the
-	  UUID/version strings themselves if the service isn't found.
+	* gtk/capture_dlg.c:
 
-	  Use #defines for protocol IDs.
+	  added "#include <epan/packet.h>" to be able to include
+	  "capture.h"
 
-2003-10-20 22:28  guy
+2003-11-15 08:47  ulfl
 
-	* gtk/: main.c, menu.c:
+	* capture.c, capture.h, gtk/Makefile.am, gtk/Makefile.nmake:
 
-	  If a tap menu item doesn't have a "selected_packet_enabled()" or
-	  "selected_tree_row_enabled()" routine, enable it by default, so
-	  that tap windows can be popped up even if you have no capture
-	  file.
+	  seperated capture dialog from capture.c and put into new
+	  gtk/capture_info_dlg.c
 
-	  Assorted code cleanups.
+2003-11-14 21:18  guy
 
-2003-10-20 23:05  guy
+	* capture.c:
 
-	* packet-laplink.c:
+	  When it comes to whether to use "select()" or not, even if you're
+	  building in Cygwin's pretend-it's-UNIX environment, we need to
+	  treat the platform as Windows.
 
-	  From Jesper Peterson: get rid of GCCism/C++ism - C89 doesn't
-	  allow variables to be declared in the middle of executable code.
+	  Get rid of the BSD #define - just check for the platforms on
+	  which we mustn't use "select()".
 
-2003-10-20 23:07  guy
+2003-11-14 19:20  guy
 
 	* doc/README.developer:
 
-	  Warn about Yet Another GCCism.
-
-2003-10-21 07:17  guy
-
-	* packet-dcerpc.c, epan/packet_info.h:
-
-	  Catch exceptions when dissecting a verifier, so we still dissect
-	  the stub data even if there's a problem dissecting the verifier.
-
-	  Show stub data as "Encrypted stub data" if it's encrypted,
-	  "Decrypted stub data" if it was encrypted but we decrypted it,
-	  and "Stub data" if it wasn't encrypted.
-
-	  Don't attempt to decrypt data unless it was encrypted (i.e., the
-	  authentication level is "Packet privacy".
-
-	  Get rid of "decrypted_data" member of "packet_info" structure -
-	  we don't need it any more.
-
-2003-10-22 01:28  sahlberg
-
-	* packet-h225.c, packet-per.c:
-
-	  Fix to dissection of Mike's problem reported on ethereal-users
-
-	  We did align to byte a bit too frequently inside
-	  dissect_per_octet_string
+	  Fix the stuff that talks about adding to DISSECTOR_SOURCES to
+	  talk about adding to DISSECTOR_SRC instead, and update it to
+	  mention "Makefile.nmake" in the first instance.
 
-	  Also change GroupIP.group from being a FT_BYTES into being
-	  FT_STRING
+2003-11-14 18:59  guy
 
-2003-10-22 01:55  sahlberg
-
-	* AUTHORS, packet-dcerpc-tapi.c, packet-dcerpc-tapi.h:
-
-	  From Jean-Baptiste Marchand add function names for the TAPI
-	  interface
-
-2003-10-22 02:03  sahlberg
-
-	* AUTHORS, Makefile.am, Makefile.nmake, packet-dcerpc-butc.c:
-
-	  From Jaime Fournier
-
-	  New protocol DCERPC BUTC Stub dissector for the BUTC interface
-
-2003-10-22 02:07  sahlberg
-
-	* AUTHORS, Makefile.am, Makefile.nmake, packet-dcerpc-rs_bind.c:
-
-	  From Jaime Fournier New protocol DCERPC/RS_BIND
+	* packet-gprs-ns.c:
 
-	  Added stub dissector for the RS_BIND protocol
+	  From Josef Korelus: update the offset after decoding an IE.
 
-2003-10-22 02:22  guy
+2003-11-14 10:11  guy
 
-	* packet-dcerpc-butc.c, packet-dcerpc-rs_bind.c:
+	* packet-null.c:
 
-	  Get rid of carriage returns.
+	  Fix the heuristics so that they recognize Linux DLT_NULL headers.
 
-2003-10-22 02:24  guy
+2003-11-14 02:07  guy
 
-	* packet-gtp.c:
+	* packet-sip.c:
 
-	  Get rid of C++ comment.
+	  Add support for the compact form of headers.
 
-2003-10-22 20:12  guy
+2003-11-14 01:35  guy
 
-	* packet-sccp.c:
+	* packet-tcap.c:
 
-	  From Michael Lum: add new SSN values for RANAP, IOS, and
-	  BSSAP/BSAP, and add a heuristic subdissector list.
+	  From Michael Lum: fix up handling of ABORT package.
 
-2003-10-22 20:59  guy
+2003-11-14 01:27  guy
 
-	* AUTHORS, Makefile.am, Makefile.nmake, packet-ansi_637.c,
-	  packet-ansi_a.c, packet-bssap.c, packet-bssap.h:
+	* packet-wap.c:
 
-	  From Michael Lum:
+	  From Olivier Biot: fix the debugging output of
+	  "tvb_get_guintvar()".
 
-		  BSSAP (GSM 08.06)/BSAP (IOS 4.0.1) support
-		  IOS 4.0.1 support
-		  Add Transport Layer support to ANSI IS-637-A dissector
+2003-11-13 23:44  guy
 
-2003-10-22 21:21  guy
+	* packet-wsp.c:
 
-	* packet-ymsg.c:
+	  From Olivier Biot: use "g_strdup_printf()" to generate
+	  dynamically-allocated sprintf output.
 
-	  From Yaniv Kaul:
+2003-11-13 23:38  guy
 
-		  remove the check on 3 ports only - since Yahoo! Messenger
-	  can
-		  stream on any port;
+	* packet-eapol.c, packet-isup.c, packet-ldp.c, packet-m2ua.c,
+	  epan/proto.c:
 
-		  remove the check that will not dissect if the packet is
-	  not big
-		  enough - as partial dissection is also helpful sometimes;
+	  Require that field names contain only alphanumerics, "-", "_",
+	  and ".".
 
-		  the version is 2 bytes, not 4 bytes.
+	  Fix the names that contained other characters.
 
-2003-10-22 21:26  guy
+2003-11-13 23:13  guy
 
 	* packet-dcerpc.c:
 
-	  From Yaniv Kaul: if there's more than one context item in a BIND
-	  PDU, show the number of context items before showing the first
-	  one.
-
-2003-10-22 22:13  guy
-
-	* prefs.c, prefs.h, gtk/toolbar.c:
+	  From Jaime Fournier: get rid of space in field name.
 
-	  From Ulf Lamping: add a preference to control whether to show or
-	  hide the main toolbar (currently no GUI to set it; that's in
-	  progress).
+2003-11-13 20:53  sahlberg
 
-2003-10-23 00:16  guy
-
-	* packet-ansi_683.c:
-
-	  From Michael Lum: register with the ANSI A-interface dissector.
-
-2003-10-23 04:57  guy
-
-	* packet-dcerpc-netlogon.c:
-
-	  Use -1 rather than "tvb_length(tvb)" to specify a length that
-	  covers the entire tvbuff for Secure Channel bind credentials.
+	* packet-tcp.c:
 
-	  Use -1 rather than 0 to have the top-level item for Secure
-	  Channel ACK credentials cover the entire tvbuff.
+	  Change the gboolean controlling whether to use TCP Sequence
+	  Number Analysis and TCP Relative Sequence Numbers to default to
+	  ENABLED instead of DISABLED.
 
-2003-10-23 04:59  guy
+	  These features do not consume that much memory or CPU but will
+	  greatly enhance the feature set of ethereal.	Make it enabled by
+	  default so also those that never venture into the preferences
+	  dialog will benefit from it.
 
-	* epan/exceptions.h:
+2003-11-13 08:22  sahlberg
 
-	  Add a new EXCEPT_CODE macro to get the exception code for the
-	  current exception, for use in CATCH_ALL handlers, so you can
-	  catch all exceptions and then pass the exception code on to a
-	  routine that handles different exceptions differently.
+	* file.c, doc/ethereal.pod:
 
-2003-10-23 05:01  guy
+	  For Time Reference frames  reset the counter for
+	  CulmulativeBytes.  This makes the CulmulativeBytes field make
+	  more sense since if we want something to be a TimeReference frame
+	  it is likely that we also want to measure BOTH time and number of
+	  bytes (==culmulative bytes) until the event we are looking at.
 
-	* packet-frame.c, packet-frame.h:
+2003-11-13 08:16  sahlberg
 
-	  Add a "show_exception()" routine that takes an exception code as
-	  an argument, and puts the appropriate exception indication into
-	  the tree.
+	* packet-ip.c:
 
-	  In "dissect_frame()", do a CATCH_ALL for exceptions, and pass the
-	  exception code to "show_exception()".
+	  Change the strings shown for ICMP TTL exceeded to match the
+	  wording in the standard
 
-2003-10-23 05:23  guy
+2003-11-13 05:08  gerald
 
-	* packet-dcerpc.c:
+	* packet-wsp.c:
 
-	  Get rid of the "offset" argument to "dcerpc_try_handoff()" - it's
-	  always 0.
+	  We can't pass NULLs to proto_tree_add_text().  Replace a bunch of
+	  occurrences of match_strval() with val_to_str() so that we don't
+	  throw an assertion.
 
-	  In "dcerpc_try_handoff()", remove the authentication padding from
-	  the stub data handed to the subdissector - that's not really stub
-	  data for the subdissector, and it should throw an exception if
-	  the request or response would go into the authentication padding.
-	   Don't even try to dissect the remaining stub data if the
-	  authentication padding value consumes all the stub data or would
-	  consume even more than that.
+	  These changes fix a problem with a file from Martin Dubovsky;
+	  otherwise they haven't been heavily tested.
 
-	  Show any "Long frame" data before the authentication padding, and
-	  show the authentication padding as the stuff at the very end of
-	  the stub data, after the "Long frame" data.
+2003-11-12 23:17  guy
 
-	  Catch all exceptions when dissecting authentication information,
-	  so that even if it's bad or we don't have all of it, we still
-	  dissect the stub data.
+	* plugins/docsis/packet-docsis.c:
 
-	  Try dissecting authentication trailer information even if we
-	  don't have all of it in the tvbuff - we want an exception to be
-	  thrown if we don't.  Don't try to dissect it if it eats into the
-	  stub data, however.
+	  From Anand Narwani: fix bitmask for Active Grants field.
 
-	  Don't bother catching exceptions in "dissect_auth_verf()" - we
-	  now always catch exceptions in above it in the DCE RPC dissector
-	  call tree.
+2003-11-12 22:44  guy
 
-	  Use CATCH_ALL and "show_exception()" when calling the
-	  sub-dissector for a connection-oriented PDU; that means we won't
-	  have to worry about adding new exception types unless they're
-	  types that we should rethrow.
+	* packet-wsp.c:
 
-2003-10-23 05:58  guy
+	  From Olivier Biot:
 
-	* packet-dcerpc.c:
+		  implement the TE header decoding;
 
-	  Catch exceptions in "dissect_dcerpc_cn_bs()", so that if we get
-	  an exception dissecting stuff past the DCE RPC header, we still
-	  drive on and dissect the next PDU, if any.
+		  fix the exported add_content_type function;
 
-2003-10-23 07:14  guy
+		  reimplement the header parameter dissection so the
+	  parameters
+		  also show up in the related header.
 
-	* packet-dcerpc.c:
+2003-11-12 21:22  guy
 
-	  Fix a call to "dissect_dcerpc_cn()" to handle the new return
-	  value.
+	* packet-tzsp.c:
 
-2003-10-23 07:52  guy
+	  From Chris Waters:
 
-	* packet-ncp.c:
+		  * Better decoding of TZSP header.
+		  * Handle unknown tags in packet.
+		  * Removed encapsulation types which are not actually
+	  supported.
+		  * Added tags for original length and sensor address.
+		  * Fix signedness of some fields.
 
-	  Treat NCP 0x5555 packets as NCP requests, so we store them in the
-	  hash table and can identify replies to them.
+	  Get rid of unused "hf_tzsp_reserved" variable.
 
-	  Clean up white space.
+	  Use -1 rather than "tvb_length(tvb)" when constructing the
+	  top-level protocol tree item, and set the length if we have
+	  payload that we dissect separately.
 
-	  Note that the "is_signed" hack doesn't work with Ethereal.
+	  Have "add_option_info()" take a starting offset, rather than
+	  constructing a subset tvbuff for it.
 
-2003-10-23 08:16  guy
+	  Clean up indentation.
 
-	* packet-ncp2222.inc:
+2003-11-12 20:44  guy
 
-	  Non-NCP_SERVICE_REQUEST packets don't have a subfunction.
+	* AUTHORS, packet-ntp.c, doc/ethereal.pod:
 
-	  The group for a request isn't part of the packet, so give it an
-	  offset and length of 0, so if you select it we doesn't highlight
-	  some part of the packet.
+	  From Matthias Drochner: support for mode 6 and mode 7 control
+	  packets and NTP4 autokey extension data.
 
-2003-10-23 08:40  sahlberg
+2003-11-12 19:04  ulfl
 
-	* packet-smb.c:
+	* gtk/print_dlg.c:
 
-	  The ACL revision field is one byte, not 2 bytes.
+	  completely redesigned print dialog layout, to be more obvious to
+	  the user
 
-2003-10-23 18:49  ashokn
+2003-11-12 18:48  gerald
 
-	* AUTHORS, packet-lmp.c, doc/ethereal.pod:
+	* file.c:
 
-	  Added support for LMP draft version -09. Support for older draft
-	  versions -02 and -03 is still present (since they are in use),
-	  selectable by a protocol preference.
+	  Remove MAX_DECODE_BUFFER_SIZE; it's no longer used.
 
-2003-10-24 00:35  guy
+2003-11-12 09:00  guy
 
-	* packet-dcerpc-reg.c:
+	* packet-dcerpc.c:
 
-	  From Jim McDonough: add dissection of the shutdown, shutdownex,
-	  and abortshutdown commands within the winreg pipe.
+	  Put stub data back where it was in the tree before.
 
-2003-10-24 00:36  guy
+2003-11-12 08:58  guy
 
 	* packet-smb.c:
 
-	  From Jim McDonough: add one more nt status code to
-	  packet-dcerpc-smb.c that is often returned from the winreg
-	  abortshutdown operation.
+	  If what remains of the SMB byte count exceeds the data left in
+	  the tvbuff, show what's left as extra byte parameters - don't act
+	  as if there aren't any extra byte parameters.
 
-2003-10-24 00:38  guy
+2003-11-11 20:49  guy
 
-	* packet-bssap.c:
+	* packet-null.c, wiretap/snoop.c:
 
-	  From Michael Lum: get rid of an unused value_string table.
+	  From Brian Ginsbach: fix handling of IRIX and UNICOS/mp snoop
+	  captures on loopback interfaces.
 
-2003-10-24 00:42  guy
+2003-11-11 20:33  guy
 
-	* AUTHORS, packet-dcerpc-mapi.c, packet-dcerpc-mapi.h:
+	* packet-ses.c:
 
-	  From Jean-Baptiste Marchand: add more MAPI procedure names.
+	  Add parameter length checking.
 
-2003-10-24 00:50  guy
+2003-11-11 20:11  guy
 
-	* packet-sip.c:
+	* packet-rsvp.c:
 
-	  From Anders Broman: make it possible to filter on only address or
-	  tag in the SIP address fields.
+	  From Mohammad Hanif: fix dissection of RSVP DETOUR object code.
 
-2003-10-24 10:46  sahlberg
+2003-11-11 20:08  guy
 
-	* packet-h225.c, packet-per.c:
-
-	  Fix all the issues Michael Oliveras reported
-
-	  1, TunnellingProtocol_id is a CHOICE not a SEQUENCE 2, change
-	  some values to be FT_STRING instead of FT_BYTES 3, update
-	  dissect_per_octet_String to always 0 terminate all FT_STRINGs
-	  (if necessary in a temporary buffer) before passing them on to
-	  proto_tree_add_string() since that function did not want to
-	  handle FT_STRING or FT_STRINGZ othervise.
-
-2003-10-24 10:52  sahlberg
-
-	* wiretap/libpcap.c:
-
-	  The capture tools for DGUX has swapped the included/original
-	  packet lengths of the frame header.  they specify the pcap
-	  version as 543.0
-
-2003-10-24 22:59  guy
-
-	* packet-ospf.c:
-
-	  From Taisuke Sasaki: fix offsets of metrics in
-	  Inter-Area-Prefix-LSAs and Inter-Area-Router-LSAs.
-
-2003-10-24 23:55  guy
-
-	* wiretap/: libpcap.c, wtap-int.h:
+	* image/toolbar/capture_24.xpm, image/toolbar/cfilter_24.xpm,
+	  image/toolbar/dfilter_24.xpm, image/toolbar/stock_close_24.xpm,
+	  image/toolbar/stock_colorselector_24.xpm,
+	  image/toolbar/stock_help_24.xpm,
+	  image/toolbar/stock_jump_to_24.xpm,
+	  image/toolbar/stock_open_24.xpm,
+	  image/toolbar/stock_preferences_24.xpm,
+	  image/toolbar/stock_print_24.xpm,
+	  image/toolbar/stock_refresh_24.xpm,
+	  image/toolbar/stock_right_arrow_24.xpm,
+	  image/toolbar/stock_save_24.xpm,
+	  image/toolbar/stock_search_24.xpm,
+	  image/toolbar/stock_stop_24.xpm, plugins/asn1/moduleinfo.h:
 
-	  Move the version number checking for libpcap files, to see
-	  whether to swap the "captured length" and "length" fields, to the
-	  open-file code; store a tri-state (definitely swapped, definitely
-	  not swapped, maybe swapped) value in the per-capture-file-format
-	  information for libpcap format, and use that when processing
-	  packets.
+	  Get rid of carriage returns in source files.
 
-2003-10-25 00:25  sahlberg
+2003-11-11 19:24  guy
 
 	* packet-tcp.c:
 
-	  Update / cleanup to tcp sequence number analysis and new features
-
-	  moved some variables to the structure where they belonged instead
-	  of where they currently were and reduced the complexity of the
-	  code
-
-	  Fast Retransmission: Ethereal not tries to detect and flag
-	  FastRetransmissions: The heuristics for this check is:   >=3
-	  dupacks in other direction   this semgent is what the dupacks are
-	  asking for   it arrived within 10ms of the last dupack (10ms
-	  should be short enough to not confuse with real RTOs)
-
-	  OutOfOrder segments Previously all segments that did not advance
-	  the right edge of the window was flagged as retransmission   now
-	  ethereal will try to flag segments that are merely reordered as
-	  OutOfOrder segments insteaD
-
-	  tHE HEURISTICS ARE:	 it has not been ACKed yet    we have not
-	  seen it before    it arrived within 4ms of the segment
-	  immediately to the right in the window
-
-2003-10-25 06:07  guy
-
-	* packet-cdp.c:
-
-	  Update comments to give a URL that defines some additional CDP
-	  type values, and to use the names from the URL and to refer to
-	  it.
-
-2003-10-25 06:10  guy
-
-	* packet-cisco-oui.c:
-
-	  Add some additional Cisco Protocol ID values, and URLs for
-	  documents that contain them.
-
-2003-10-25 06:19  guy
-
-	* packet-cisco-oui.c:
-
-	  Add one more Cisco protocol type.
-
-2003-10-25 06:49  guy
+	  Make the declaration of "process_tcp_payload()" match the
+	  definition.
 
-	* packet-per.c:
+2003-11-11 19:23  guy
 
-	  In restricted character strings, deal with character values that
-	  are greater than the alphabet length.
-
-	  Just use "proto_tree_add_item()" if you have a range of bytes, of
-	  known length, that are to be added as an item - that handles both
-	  FT_STRING and FT_BYTES, including null-terminating the string
-	  value.
-
-2003-10-25 07:17  guy
-
-	* packet-chdlc.c, packet-frame.c, wiretap/ngsniffer.c,
-	  wiretap/visual.c, wiretap/wtap.c, wiretap/wtap.h:
-
-	  Add a new WTAP_ENCAP_CHDLC_WITH_PHDR type, to distinguish Cisco
-	  HDLC captures with packet direction information from captures
-	  without it.  Use them appropriately.
-
-2003-10-26 03:09  sahlberg
-
-	* file.c, gtk/io_stat.c:
+	* packet-tacacs.c:
 
-	  Update to cf_get_display_name() return "<no file>" if there is no
-	  file loaded yet instead of crashing in io-stat
+	  "tacplus_pref_cb()" takes no arguments, not an unspecified list
+	  of arguments.
 
-	  io-stat  only print the label for the top tick on the y axis to
-	  make it look less cluttered
+2003-11-11 18:25  guy
 
-2003-10-27 00:54  sharpe
+	* config.nmake:
 
-	* gtk/smb_stat.c:
+	  Get rid of GLib 1.3 support - current versions of GTK+ 1.3
+	  require GLib 2.0, and earlier versions are sufficiently buggy
+	  that we don't want to support them.
 
-	  This changes the Filter: label in the SMB Service Response Time
-	  dialog box to a button that brings up the Create Filter dialog
-	  box. While it works, the problem is that it also acts as an Enter
-	  keypress as far as the start_stat button is concerned.
+2003-11-11 13:59  ulfl
 
-	  Probably needs a small fix.
+	* gtk/print_mswin.c:
 
-2003-10-27 01:09  sharpe
+	  added some print dialog flags, to disable some unsupported print
+	  features (e.g. selection of pages from X to Y)
 
-	* gtk/smb_stat.c:
+2003-11-11 08:29  guy
 
-	  Make sure that Clicking OK in the filter creation dialog box does
-	  not act as an OK to the dialog box it is called from.
+	* AUTHORS, Makefile.am, Makefile.nmake, packet-ses.c, packet-ses.h,
+	  doc/ethereal.pod:
 
-2003-10-27 01:20  sharpe
+	  From Yuriy Sidelnikov: ISO 8327-1 Session Protocol support.
 
-	* gtk/: rpc_stat.c, smb_stat.c:
+2003-11-11 06:10  guy
 
-	  Add Filter Button support for onc-rpc as well ...
+	* plugins/: plugin_api_list.c, Xass-list, Xplugin_api.c,
+	  Xplugin_api.h, Xplugin_api_decls.h, Xplugin_table.h:
 
-2003-10-27 01:35  sharpe
+	  Add "asn1_id_decode1()" to the list of exported functions.
 
-	* gtk/: dcerpc_stat.c, fc_stat.c:
+2003-11-11 05:54  guy
 
-	  Add filter button support to FibreChannel stats and DCERPC stats.
+	* packet-ansi_map.c:
 
-2003-10-27 09:17  guy
+	  From Michael Lum:
 
-	* AUTHORS, packet-http.c, doc/ethereal.pod:
+		  Minor improvements.
 
-	  From Loïc Minier: HTTP header and payload desegmentation.
+		  Added IEI level decoding of cdma2000 Handoff * IOS data
+		  parameters.
 
-2003-10-27 19:30  guy
+2003-11-11 05:51  guy
 
-	* packet-mdshdr.c:
+	* packet-ansi_a.c:
 
-	  From Dinesh Dutt: display the FC CRC for frames encapsulated with
-	  MDSHDR, and show only the low-order 13 bits of the packet length
-	  field.
+	  From Michael Lum:
 
-2003-10-27 19:34  guy
+		  Change P_xxx to ANSI_A_E_xxx.
 
-	* packet-tcp.c:
+		  Fix decoding of some parameters.
 
-	  Squelch some signed vs. unsigned comparison warnings.
+		  Fix initialization of ett array.
 
-	  Get rid of an unused variable.
+2003-11-11 05:23  guy
 
-2003-10-27 22:28  guy
+	* gtk/gui_prefs.c:
 
-	* packet-h225.c, packet-h245.c, packet-h450.c, packet-per.c:
+	  The user interface preference page has only 9 rows in the GTK+
+	  2.x version.
 
-	  From Lars Roland: whitespace cleanups.
+2003-11-11 05:09  guy
 
-2003-10-27 22:29  guy
+	* gtk/gui_prefs.c:
 
-	* packaging/nsis/Makefile.nmake:
+	  Create the last two items in the GUI preferences page's first
+	  column the same way the other items are created - in order, going
+	  down the page, and using "pos++" as the table position.
 
-	  From Lars Roland: fix a typo.
+2003-11-10 22:31  guy
 
-2003-10-27 22:45  guy
+	* packet-ansi_a.c:
 
-	* epan/dfilter/scanner.l:
+	  From Albert Chin: <stdlib.h> on HP-UX 11.00 (and possibly 10.20
+	  and perhaps some versions of Digital/Tru64 UNIX) drags in
+	  <wait.h>, which drags in <sys/procset.h>, which defines P_SID in
+	  ways that cause this not to compile.	Use P_MY_SID instead.
 
-	  Allow + and , in unparsed strings, so that you can use + as a
-	  sign in a number or exponent and so that floating-point numbers
-	  can use , as well as . as a decimal point.
+2003-11-10 21:42  guy
 
-2003-10-27 23:12  guy
+	* packet-h225.c:
 
-	* gtk/io_stat.c:
+	  From Martin Regner:
 
-	  Don't draw a Y-axis label if you haven't constructed the label
-	  string.
+		  "aliasAddress", in an EndPoint sequence, is a sequence of
+		  aliasAddresses.
 
-2003-10-27 23:31  guy
+		  Don't update the Info or Protocol columns in a pdu_item.
 
-	* Makefile.am, Makefile.nmake, packet-dcerpc-initshutdown.c,
-	  packet-dcerpc-initshutdown.h:
+2003-11-10 20:22  guy
 
-	  From a suggestion by Jim McDonough: support for the INITSHUTDOWN
-	  service.
+	* packet-dcerpc-epm.c:
 
-2003-10-28 00:31  guy
+	  From Yaniv Kaul:
 
-	* AUTHORS, Makefile.am, Makefile.nmake, packet-h225.c,
-	  packet-h225.h, tap-h225counter.c, doc/ethereal.pod,
-	  doc/tethereal.pod, gtk/Makefile.am, gtk/Makefile.nmake,
-	  gtk/h225_counter.c:
+		  1. Add dissection of inquiry type (rpc_c_ep_all_elts /
+		     rpc_c_ep_match_by_if / rpc_c_ep_match_by_obj /
+		     rpc_c_ep_match_by_both).
 
-	  From Lars Roland: H.225 message and reason tag counter taps.
+		  2. The version field seems incorrect to me - I believe it
+	  is usually
+		     should be version 2.0, not 0.2.
+		     I suspect that the insertion of version information to
+	  the hash
+		     table is also wrong, (and the get for the version
+	  should have
+		     used _letohl() ) - but I did not change it.
 
-2003-10-28 03:57  guy
+		  3. Clean up some white space.
 
-	* packet-scsi.c:
+2003-11-10 20:15  guy
 
-	  From Dinesh Dutt:
+	* packet-ansi_a.c:
 
-		  Add dissection code for the following additional SCSI
-	  commands:
-		       - Start/Stop Unit
-		       - Write Buffer
-		       - Send Diagnostics
+	  From Albert Chin: dynamically allocate the "ett" array - some
+	  compilers (such as at least one older version of the HP C
+	  compiler) can't cope with constant expressions that involve the
+	  "?" operator, it appears; they think it's not a constant
+	  expression.
 
-		  Don't copy the product serial number to a buffer before
-	  printing
-		  it.
+	  Also, as Albert notes, there's no reason for "ett" to be static
+	  here.
 
-2003-10-28 05:49  guy
+2003-11-10 08:14  sahlberg
 
-	* doc/README.developer:
+	* packet-q931.c:
 
-	  As per a suggestion by Olivier Biot, note that objects pointed to
-	  by pointer arguments to "proto_tree_add_XXX" functions are copied
-	  - if you allocated a buffer for one of them (e.g., a string), and
-	  you don't free that buffer when you're done with it, you'll leak
-	  memory.
+	  The Q.931 dissector never checked whether the codeset variable
+	  was set to a reasonable value or not.
 
-2003-10-28 05:50  guy
+	  This was problematic and caused crashes since this variable is
+	  used as an index into an array where we grab a pointer (which is
+	  later dereferenced).
 
-	* doc/README.developer:
+	  Dereferencing that pointer will have surprising effects. Usually
+	  crashes.
 
-	  Clarify the previous checkin - that applies to the 'value'
-	  argument.
+	  Update Q.931 to verify sanity of the codeset variable everytime
+	  it gets set.
 
-2003-10-28 06:44  guy
+2003-11-10 08:02  sahlberg
 
-	* AUTHORS, Makefile.am, Makefile.nmake, packet-nt-tpcp.c:
+	* packet-smb.c:
 
-	  From Giles Scott: Alteon/Nortel Transparent Proxy Control
-	  Protocol support.
+	  Bugfix for ethereal crashes
 
-2003-10-28 07:02  guy
+	  If the ByteCount field in the SMB PDU spanned beyond the end of
+	  the packet because the packet was short or because the BC field
+	  was corrupted and contained garbade data  then the tree item for
+	  the command (the subtree just after the SMBHeader subtree) would
+	  describe data continuing beyond the end of the packet.
 
-	* AUTHORS, Makefile.am, configure.in,
-	  packaging/nsis/Makefile.nmake, packaging/nsis/ethereal.nsi,
-	  plugins/Makefile.am, plugins/Makefile.nmake,
-	  plugins/asn1/.cvsignore, plugins/asn1/AUTHORS,
-	  plugins/asn1/COPYING, plugins/asn1/ChangeLog,
-	  plugins/asn1/INSTALL, plugins/asn1/Makefile.am,
-	  plugins/asn1/Makefile.nmake, plugins/asn1/NEWS,
-	  plugins/asn1/moduleinfo.h, plugins/asn1/packet-asn1.c:
+	  If we selected one such tree in the dissect pane and used
+	  Prepare/Match Selected  this would cause the filter build thing
+	  to try to access data beyod the end of the packet and ethereal
+	  would dump core.
 
-	  From Matthias Melchior: plugin to decode BER-encoded ASN.1
-	  messages, given a type-table from "snacc" as a protocol
-	  description.
+	  Change the END_OF_SMB macro so that it shrinks bc so that bc
+	  never describes data beyond the end of the packet.
 
-2003-10-28 08:50  sahlberg
+2003-11-10 07:44  sahlberg
 
-	* packet-tcp.c:
+	* packet-ldap.c, gtk/dcerpc_stat.c:
 
-	  Full duplex analyzers that capture each direction of a link with
-	  a separate NIC will lose the time integrity between the two NICs
-	  more often than one might expect.  It is thus relatively common
-	  that a data segment and its ACK being swapped in the capture
-	  file.
+	  Not all LDAP PDUs are aligned to the start of a TCP segment.
 
-	  Therefore, drop the condition that a segment must not have been
-	  acked yet in the detection of OutOfOrder segments.
+	  If we failed to dissect the GSS-SPNEGO blob it probably means
+	  that the segment is somewhere in the middle of an LDAP PDU.
 
-	  Second, fix a bug where we didnt keep track of the ack numbers
-	  properly for relative sequence number analysis.
+	  Just bail out and stop dissecting the PDU instead of aborting
+	  ethereal completely using g_assert() since this is not really a
+	  pathological error, its just something that can and will happen
+	  normally.
 
-2003-10-28 17:27  guy
+2003-11-09 22:57  guy
 
-	* packet-scsi.c:
+	* asn1.c, asn1.h:
 
-	  "%.*" expects an "int" argument giving the precision.
+	  Fix the type of the "integer" argument to
+	  "asn1_uint32_value_decode()", as per a note by Michael Lum.
 
-2003-10-28 17:59  guy
+2003-11-09 22:55  guy
 
-	* packet-wsp.c:
+	* AUTHORS, packet-rtcp.c, packet-rtp.c, doc/ethereal.pod:
 
-	  From Loïc Minier: get rid of the second of a pair of identical
-	  checks.
+	  From Martin Mathieson: display the bitfields for
 
-	  As per a note by Olivier Biot, make the "multipart/XXX" items all
-	  lower case.
+		  - the first 2 bytes of RTP headers
+		  - the first byte of RTCP report.
 
-2003-10-28 18:08  guy
+2003-11-09 22:49  guy
 
-	* packet-ansi_a.c, packet-ansi_map.c, packet-bssap.c:
+	* packet-ansi_map.c:
 
-	  From Michael Lum: fixed BCD decoding of filler nibble and fixed
-	  some minor bugs.
+	  From Michael Lum: add the ability to put parameter data at the
+	  end of the parameter name so that the user doesn't have to open
+	  the parameter subtree to see the most important data.
 
-2003-10-28 19:27  guy
+	  Use "%u", not "%d", to print unsigned quantities.
 
-	* plugins/asn1/Makefile.nmake:
+	  Get rid of includes of <gmodule.h> (we're not loading any code at
+	  run-time in the dissector itself), and <sys/types.h> and
+	  <netinet/in.h> (we shouldn't need either of those).
 
-	  Fix the Makefile to match other plugin Makefiles.
+2003-11-09 22:41  guy
 
-2003-10-28 20:44  guy
+	* packet-gsm_a.c:
 
-	* plugins/asn1/Makefile.nmake:
+	  From Michael Lum get rid of non-portable "%N$" constructs in
+	  printf formats.
 
-	  This dissector uses GTK+ calls (which no dissector should); until
-	  that's fixed, use GTK_CFLAGS and GTK_LIBS rather than GLIB_CFLAGS
-	  and GLIB_LIBS.
+	  Use "%u", not "%d", to print unsigned quantities.
 
-2003-10-28 20:53  guy
+	  Get rid of includes of <gmodule.h> (we're not loading any code at
+	  run-time in the dissector itself), and <sys/types.h> and
+	  <netinet/in.h> (we shouldn't need either of those).
 
-	* plugins/asn1/packet-asn1.c:
+2003-11-09 22:32  guy
 
-	  As we don't supply a default ASN.1 SNACC output file, if we get
-	  an ENOENT when trying to open the SNACC file, and the file we're
-	  trying to open is the default file, don't print a warning.
+	* packet-ansi_a.c:
 
-	  Fix some typoes.
+	  Remove accidentally-inserted letter.
 
-2003-10-28 21:01  guy
+2003-11-09 22:30  guy
 
-	* plugins/: plugin_api_list.c, Xass-list, Xplugin_api.c,
-	  Xplugin_api.h, Xplugin_api_decls.h, Xplugin_table.h:
+	* packet-ansi_a.c:
 
-	  Add "get_datafile_dir()" to the plugin API list, for the benefit
-	  of plugins that might get configuration information from a file.
-
-2003-10-28 21:04  guy
+	  From Michael Lum:
 
-	* plugins/asn1/packet-asn1.c:
+		  Fix some incorrect bit shifts.
 
-	  "get_datafile_dir()" is now in the plugin table.
+		  Get rid of non-portable "%N$" constructs in printf
+	  formats.
 
-2003-10-28 21:45  guy
+	  Use "%u", not "%d", to print unsigned quantities.
 
-	* plugins/: plugin_api_list.c, Xass-list, Xplugin_api.c,
-	  Xplugin_api.h, Xplugin_api_decls.h, Xplugin_table.h:
-
-	  Export "get_datafile_path()", not "get_datafile_dir()" - most if
-	  not all dissectors need only the former, which does the
-	  pathname-construction work for you.
-
-2003-10-28 21:47  guy
+2003-11-09 22:15  guy
 
 	* plugins/asn1/packet-asn1.c:
 
-	  Use "get_datafile_path()", rather than constructing the default
-	  ASN.1 file path by hand.
-
-	  Set "default_asn1_filename", not "asn1_filename", to the default
-	  path.
-
-2003-10-29 10:37  guy
+	  From Albert Chin: the HP-UX 10.20 C compiler doesn't allow
 
-	* plugins/plugin_api.h:
+		  typedef enum _foo foo;
+		  enum _foo {
+			  ...
+		  };
 
-	  There's no need to #define tvb_get_string or tvb_get_stringz -
-	  that's done in Xplugin_api.h.
+	  You have to define the enum before using it in a typedef.  (If I
+	  had my ANSI C89 standard handy, I could say whether that was
+	  legal ANSI C or not.)
 
-	  However, we do need to include "epan/filesystem.h", to declare
-	  "get_datafile_path()".
+2003-11-09 22:13  guy
 
-2003-10-29 10:54  guy
+	* packet-bssgp.c:
 
-	* plugins/asn1/packet-asn1.c:
+	  From Albert Chin: get rid of null statements outside of function
+	  bodies (that's what the ";" in
 
-	  From Matthijs Melchior: #ifdef out the GUI code, for now.
+		  int
+		  foo(...)
+		  {
+			  ...
+		  };
 
-2003-10-29 21:19  guy
+	  is).
 
-	* AUTHORS, packet-diameter.c, doc/ethereal.pod:
+2003-11-09 19:53  guy
 
-	  From Steve Ford: "prefs_register_string_preference()" makes a
-	  copy of any string pointed to by the preference variable - as the
-	  value we set it to is allocated, we should free it after
-	  registering the preference.
+	* packet-dcerpc-dcom.h:
 
-	  The register routine is called only once - don't worry about
-	  whether "gbl_diameterDictionary" is null or not.
+	  From Albert Chin: get rid of C++ comments.
 
-	  Get rid of a duplicate credit entry in the man page.
+2003-11-09 01:36  guy
 
-2003-10-29 21:54  guy
+	* prefs.c:
 
-	* packet-snmp.c:
+	  From Gisle Vanem:
 
-	  Fix some memory leaks found by Steve Ford.
+		  Copy the "gui_fileopen_*" fields in "copy_prefs()".
 
-2003-10-29 22:00  guy
+		  Fix a typo in "free_prefs()".
 
-	* packet-snmp.c:
+2003-11-08 05:47  guy
 
-	  Actually, one of those *isn't* a leak; don't free something we
-	  haven't allocated yet.
+	* packet-tcp.c, packet-tcp.h:
 
-2003-10-29 22:02  guy
+	  Have "decode_tcp_ports()" handle only the handoff to a
+	  subdissector - and have it return TRUE if we succeeded, FALSE
+	  otherwise - and have an internal "process_tcp_payload()" routine
+	  handle the (TCP-specific) PDU tracking and sequence number
+	  analysis, with an argument to indicate whether it should do that
+	  or not (i.e., whether it's being handed a TCP segment or
+	  reassembled data).
 
-	* packet-snmp.c:
+2003-11-08 00:09  guy
 
-	  Another leak that isn't.
+	* AUTHORS, packet-rsvp.c, doc/ethereal.pod:
 
-2003-10-29 22:04  guy
+	  From Mohammad Hanif:
 
-	* packet-snmp.c:
+		  correct and enhance support for RSVP FAST_REROUTE and
+	  DETOUR
+		  objects (source:
+	  draft-ietf-mpls-rsvp-lsp-fastreroute-03.txt);
 
-	  ...and there's one leak we missed.
+		  support an RSVP SESSION_OBJECT object with ctype = 1.
+	  This
+		  object contains resource affinities (source: RFC 3209).
 
-2003-10-29 22:11  guy
+2003-11-08 00:02  guy
 
-	* packet-snmp.c:
+	* packet-msproxy.c, packet-socks.c, packet-tcp.c, packet-tcp.h:
 
-	  Plug that leak with a cleanup handler, so we don't leak the
-	  variable OID if we throw an exception fetching the variable
-	  value.
+	  "decode_tcp_ports()" is for use by protocols that proxy
+	  transport-layer packets/sessions, e.g. MSProxy and SOCKS.  It
+	  should not cause any of the TCP-specific stuff such as sequence
+	  number analysis or PDU tracking to be done.  (Actually, MSProxy
+	  and SOCKS should offer desegmentation services *themselves* and
+	  do their *own* PDU tracking, rather than just passing stuff on to
+	  "decode_tcp_ports()", but that's another matter.)
 
-2003-10-29 22:39  guy
+	  Make "tcp_tree" once again be a local variable to
+	  "dissect_tcp()", and pass it as an argument to those functions
+	  that use it.
 
-	* gtk/toolbar.c:
+2003-11-07 23:27  guy
 
-	  Have the "edit display filters" button do the exact same thing as
-	  the "Edit -> Display Filters..." menu item, as per Ulf Lamping -
-	  there's apparently some problem wherein &args doesn't get passed
-	  properly to the callback.
+	* Makefile.nmake, config.h.win32, config.nmake,
+	  wiretap/Makefile.nmake, wiretap/config.h.win32:
 
-2003-10-29 23:15  guy
+	  Let people configure whether to build with Zlib or not solely by
+	  controlling whether ZLIB_DIR is defined or not in config.nmake.
 
-	* gtk/dfilter_expr_dlg.c:
+2003-11-07 23:26  guy
 
-	  Put quotes around values for FT_ABSOLUTE_TIME variables, and
-	  around values with white space in them.
+	* gtk/Makefile.nmake:
 
-2003-10-29 23:48  guy
+	  Nothing here should require zlib.h, so we shouldn't need to do
+	  "/I$(ZLIB_DIR)".
 
-	* epan/: proto.c, proto.h, ftypes/ftype-bytes.c,
-	  ftypes/ftype-string.c, ftypes/ftype-tvbuff.c:
+2003-11-07 20:23  guy
 
-	  Make the "fvalue_set" methods for types whose value is allocated
-	  free any previously-allocated version first, so that they don't
-	  leak memory.
+	* packet-wsp.c:
 
-	  From Olivier Biot: add a "proto_item_append_string()" routine, to
-	  append to the string value a protocol tree item has.
+	  From Olivier Biot: implement the Range, Content-Range and
+	  Content-Disposition headers.
 
-2003-10-30 00:39  tpot
+2003-11-07 20:07  guy
 
-	* packet-ipmi.c:
+	* packet-wsp.c:
 
-	  Decorate higher level proto_item with NetFn and LUN values in
-	  hex.
+	  From Olivier Biot: add dissection for the Content-Base headeer.
 
-2003-10-30 02:06  guy
+2003-11-07 08:50  guy
 
-	* AUTHORS, Makefile.am, Makefile.nmake, packet-brdwlk.c,
-	  packet-fc.c, packet-fc.h, packet-fcels.c, packet-fcels.h,
-	  packet-fcip.c, packet-fcsb3.c, packet-fcsb3.h, packet-fcsp.c,
-	  packet-fcsp.h, packet-fcswils.c, packet-fcswils.h,
-	  packet-mdshdr.c, epan/packet_info.h:
+	* packet-fcels.c, packet-fcels.h:
 
 	  From Dinesh Dutt:
 
-		  - Dissector for FICON
-		  - Dissector for FC-SP (Security Protocol for Fibre
-	  Channel)
-		  - Patches to correct the reassembly of FC fragments.
-		  - Support for new MDS Port Analyzer Adapters that carry
-	  the
-		    frame length for truncated frames.
+	      - Added support for displaying support for
+	  Multicast/Broadcast supported
+		bits in *LOGI common service params
+	      - Removed restriction for "Random Relative Offset" & its
+	  alternate to be in
+		PLOGI or PDISC only. Some HBAs set it even in FLOGI.
 
-2003-10-30 03:11  guy
+2003-11-07 08:47  guy
 
-	* AUTHORS, doc/ethereal.pod, wiretap/AUTHORS, wiretap/Makefile.am,
-	  wiretap/Makefile.nmake, wiretap/file_access.c, wiretap/hcidump.c,
-	  wiretap/hcidump.h, wiretap/wtap.h:
+	* packet-fcsp.c:
 
-	  From Marcel Holtmann: support for reading Linux Bluez Bluetooth
-	  stack "hcidump -w" traces.
-
-	  Note that Jesper Peterson contributed support for reading Endace
-	  ERF files.
-
-2003-10-30 03:15  guy
-
-	* packet-brdwlk.c:
-
-	  #if 0 out the stuff to set the reported length, as it'd throw an
-	  exception if it would increase the reported length.
-
-2003-10-30 07:00  guy
-
-	* AUTHORS, Makefile.am, Makefile.nmake, packet-ansi_a.c,
-	  packet-ansi_map.c, packet-gsm_a.c, packet-gsm_sms.c:
-
-	  From Michael Lum:
-
-		  GSM BSSMAP (GSM 08.08) support
-		  GSM DTAP (3GPP TS 24.008) support
-		  GSM SMS (3GPP TS 24.011) support
-		  GSM SS (3GPP TS 24.080) support
-		  GSM SMS TPDU (3GPP TS 23.040) support
-
-2003-10-30 07:14  guy
-
-	* packet-gsm_sms.c:
-
-	  From Michael Lum: squelch a couple of compiler warnings.
-
-2003-10-30 08:07  guy
-
-	* packet-q931.c:
-
-	  From Anders Broman:
-
-		  fix the offset when putting the cause code in a Cause IE
-	  into the
-		  protocol tree;
-
-		  in a Number IE, show the number type, numbering plan, and
-		  extension indicator as named-field bitfields.
-
-2003-10-30 08:34  guy
-
-	* packet-q933.c:
-
-	  Apply Anders Broman's changes to the Q.931 dissector to the Q.933
-	  dissector:
-
-		  fix the offset when putting the cause code in a Cause IE
-	  into the
-		  protocol tree;
-
-		  in a Number IE, show the number type, numbering plan, and
-		  extension indicator as named-field bitfields.
-
-2003-10-30 11:21  guy
-
-	* packet-gsm_a.c:
+	  From Dinesh Dutt:
 
-	  Definitions of arrays, with no length specified, are not legal C.
+	      - Added support for decoding related to Security in *LOGI
+	  service params
+	      - Added support for new LS_RJT code for authentication
+	  failure
+	      - Brought packet-fcsp.c up-to-date with FC-SP rev 1.2
 
-2003-10-30 11:53  guy
+2003-11-07 05:26  sahlberg
 
-	* wiretap/hcidump.c:
+	* Makefile.am, Makefile.nmake, packet-kpasswd.c:
 
-	  The quantities in the hcidump header are little-endian, not
-	  big-endian.
+	  New protocol:   MS Kpasswd  (RFC3244)
 
-2003-10-30 11:54  guy
+	  KPasswd is partially dissected for UDP.
 
-	* packet-gsm_sms.c:
+	  It would be very useful if someone added dissection of the asn.1
+	  encoded AP_REQ and the KRB-PRIV blobs. I dont think I will add
+	  those.
 
-	  No zero-length arrays, please.
+2003-11-07 04:03  sahlberg
 
-2003-10-30 11:56  guy
+	* packet-ldap.c:
 
-	* plugins/asn1/packet-asn1.c:
+	  In some captures we might have already established and BOUND LDAP
+	  session where GSS-SPNEGO is used.  If we havent seen the BIND
+	  call ethereal would assume it is vanilla non-GSS-SPNEGO LDAP and
+	  would fail to decode the packet.
 
-	  "strchr()" and the like would be compared against NULL;
-	  "strcmp()", however, is compared against 0.
+	  Add heuristics to the LDAP dissector so that IF the first 4 bytes
+	  of the LDAP PDU looks like ity could be a length field and IF the
+	  fifth byte has the value 0x60 then assume what we have is
+	  GSS-SPNEGO and assume this and all further commands on this
+	  session is GSS-SPNEGO as well.
 
-2003-10-30 19:38  guy
+2003-11-07 03:47  guy
 
-	* packet-ansi_map.c:
+	* packet-http.c:
 
-	  From Michael Lum: step over unknown parameters.
+	  Make a field for the Content-Type entity header, and give it an
+	  EH_ value so it can be handled specially (with a string dissector
+	  table so that subdissectors can register for particular content
+	  types).
 
-2003-10-30 19:39  guy
+2003-11-07 01:29  guy
 
-	* packet-gsm_a.c:
+	* gtk/: menu.c, toolbar.c, toolbar.h:
 
-	  From Michael Lum: dissect DTAP from RANAP NAS PDU fields.
+	  "File->Open" should be available whether or not we have a captue
+	  file open.
 
-2003-10-30 19:43  guy
+	  The toolbar equivalent should *not* be available if we have an
+	  "Update list of packets in real time" capture running.
 
-	* packet-bssgp.c:
-
-	  From Josef Korelus: fix reversed messages in true_false_string
-	  tables.
+	  The toolbar "Save" button should not be available if we don't
+	  have an unsaved capture file.
 
-2003-10-30 19:56  guy
-
-	* tethereal.c:
+2003-11-06 23:02  guy
 
-	  On UNIX, give up set-UID and set-GID privileges before opening
-	  capture files and immediately after opening capture devices, so
-	  we run without privileges as much as possible.
-
-2003-10-30 20:30  guy
-
-	* configure.in:
-
-	  If we don't have pod2man, quit - otherwise, we'd just drive on
-	  and use the empty POD2MAN variable in commands, producing very
-	  confusing errors.
-
-2003-10-30 22:06  guy
+	* gtk/menu.c:
 
-	* configure.in:
+	  Make Shift+Control+S an accelerator for "Save As...", as per the
+	  GNOME HIG.
 
-	  Wrap "x$POD2MAN" in quotes, so if there are spaces in the
-	  pathname, the configure script doesn't fail.	(The Makefile will
-	  probably fail, but that's another matter.)
+2003-11-06 22:45  guy
 
-	  If "$POD2MAN" is empty, it could mean that pod2man is installed
-	  but the user's path doesn't include the directory in which it's
-	  installed; fix up the error message.
+	* wiretap/: file_access.c, netmon.c, network_instruments.c,
+	  network_instruments.h, wtap-int.h:
 
-2003-10-31 00:43  guy
+	  From Scott Emberley: support for writing Network Instruments
+	  Observer files.
 
-	* AUTHORS, doc/editcap.pod, doc/ethereal.pod, doc/mergecap.pod,
-	  doc/tethereal.pod, wiretap/AUTHORS, wiretap/Makefile.am,
-	  wiretap/Makefile.nmake, wiretap/file_access.c,
-	  wiretap/network_instruments.c, wiretap/network_instruments.h,
-	  wiretap/wtap.h:
+2003-11-06 09:52  guy
 
-	  From Scott Emberley: support for reading Network Instruments
-	  version 9 capture files.
+	* doc/README.plugins:
 
-2003-10-31 07:57  guy
+	  Note that not all OSes on which Ethereal runs can support
+	  plugins.
 
-	* util.c:
+	  Note that you have to modify plugins/Makefile.nmake.
 
-	  Include <windows.h> so that OSVERSIONINFO and the like are
-	  defined on Windows.
+	  Fix "plugin/" to "plugins/".
 
-2003-10-31 08:06  guy
+	  Update the sample Makefile.am and Makefile.nmake to match the
+	  current state of affairs.
 
-	* wiretap/: network_instruments.c, network_instruments.h:
+2003-11-06 09:47  guy
 
-	  Temporarily get rid of the "struct tm" in "struct observer_time",
-	  and get rid of the reference to its "tm_gmtoff" member - there
-	  are platforms on which Ethereal runs that don't have "tm_gmtoff"
-	  in "struct tm".  If the time stamp in the packets is nanoseconds
-	  since midnight 2001-01-01 *local* time, we'd need to compute the
-	  offset between that and midnight 2000-01-01 GMT, and adjust the
-	  time with that.
+	* doc/README.developer:
 
-2003-10-31 08:13  guy
+	  Warn people not to use the "numbered argument" feature that many
+	  UNIX printf's implement.
 
-	* doc/: editcap.pod, ethereal.pod, mergecap.pod, tethereal.pod:
+2003-11-06 09:32  guy
 
-	  Fix some warnings from pod2man.
+	* plugins/acn/packet-acn.c:
 
-2003-10-31 08:15  guy
+	  From Erwin Rol: fix some switched fields.
 
-	* README:
+2003-11-06 09:28  guy
 
-	  Update the list of supported capture formats.
+	* packet-sccp.c:
 
-2003-10-31 18:28  guy
+	  From Michael Lum: have the SCCP dissector register itself by
+	  name.
 
-	* plugins/acn/: acn.h, packet-acn.c:
+2003-11-06 09:18  sahlberg
 
-	  From Erwin Rol: update.
+	* packet-ldap.c, packet-tcp.c:
 
-2003-10-31 19:45  guy
+	  Update to LDAP and TCP
 
-	* gtk/rtp_analysis.c:
+	  LDAP messages that span multiple segments will throw an exception
+	  unless we have reassembly enabled.
 
-	  From Lars Ruoff:
+	  Update TCP so that IF an exception was thrown that we still pick
+	  up any hints provided by the subdissector about where the next
+	  PDU starts.
 
-		  - fix missing detection of first packet when writing
-	  payload.
-		  - fix bug of erroneous handling of confort noise when
-	  writing payload.
-		  - fix bug of possible endless silence insertion on first
-	  packet when
-		    writing payload.
+	  Update LDAP so that it will rpovide hints to TCP about where the
+	  next LDAP PDU starts in the sequence number space.
 
-2003-10-31 19:48  guy
+	  Thus now ethereal can find and dissect LDAP PDUs that starts
+	  somewhere in the middle of a TCP segment.
 
-	* packet-h225.c:
+2003-11-06 09:13  guy
 
-	  From Martin Regner: fix a problem with vendorIdentifier in some
-	  H.225 messages.
+	* packet-dcerpc.c:
 
-2003-11-01 02:30  guy
+	  Update a comment.
 
-	* AUTHORS, acinclude.m4, capture.c, capture.h, globals.h,
-	  pcap-util.c, pcap-util.h, tethereal.c, doc/ethereal.pod,
-	  doc/tethereal.pod, gtk/capture_dlg.c, gtk/main.c:
+	  Catch another case where we need to check for a null
+	  decrypted_tvb.
 
-	  Based on a patch from Brian Fundakowski Feldman, add support for
-	  setting link-layer type when capturing, using the
-	  "pcap_set_datalink()" and related APIs.
+2003-11-06 09:05  guy
 
-2003-11-01 03:03  guy
+	* plugins/: acn/Makefile.nmake, artnet/Makefile.nmake,
+	  asn1/Makefile.nmake, docsis/Makefile.nmake,
+	  gryphon/Makefile.nmake, lwres/Makefile.nmake,
+	  megaco/Makefile.nmake, mgcp/Makefile.nmake, pcli/Makefile.nmake,
+	  rtnet/Makefile.nmake:
 
-	* gtk/main.c:
+	  Use "$(OBJECTS)" whenever a complete list of .obj files appears,
+	  rather than repeating the list from the setting of OBJECTS.
 
-	  Return FALSE from the configure event handler, so that anybody
-	  else who's registered for it can get notified about it.
+2003-11-06 08:54  guy
 
-2003-11-01 03:06  guy
+	* packet-http.c:
 
-	* plugins/plugin_api_list.c:
+	  Add a mechanism to handle various entity headers, allowing their
+	  (string) values to be put into the protocol tree as fields and
+	  allowing some headers to get special treatment, and use it for
+	  Authorization, Proxy-Authorization, WWW-Authenticate, and
+	  Proxy-Authenticate.
 
-	  From Charlie Duke: add more functions to the plugin API.
+2003-11-06 08:51  sahlberg
 
-2003-11-01 03:08  guy
+	* packet-tcp.c:
 
-	* plugins/plugin_api_list.c:
+	  Fix for TCP.
 
-	  Move the new functions to the end of the list, so we don't change
-	  the ordinal numbers of the routines already in the list (that
-	  breaks binary compatibility).
+	  If we have short or malformed PDUs in protocols above TCP this
+	  will generate an exception and thus some of the stateful things
+	  such as keeping track of and printing the tcp analysis data will
+	  be shourcutted and not called.
 
-2003-11-01 03:10  guy
+	  Add a wrapper around the call to the subdissectors above TCP so
+	  that if an exception is generated we will still catch it and
+	  explicitely call tcp_print_sequence_number_analysis() so that
+	  also short packets are handled well.
 
-	* plugins/plugin_api_list.c:
+2003-11-06 07:44  sahlberg
 
-	  "End" means end, not "almost the end".
+	* packet-dcerpc.c:
 
-2003-11-01 03:10  guy
+	  Bugfix for dcerpc parsing of encrypted short packets.
 
-	* plugins/: Xass-list, Xplugin_api.c, Xplugin_api.h,
-	  Xplugin_api_decls.h, Xplugin_table.h:
+	  Sometimes if we cant decrypt a DCERPC packet decrypted_tvb is
+	  NULL.
 
-	  From Charlie Duke: add more functions to the plugin API.
+	  do not pass a NULL pointer to show_stub_data() since this will
+	  dump core.
 
-2003-11-01 03:38  guy
+2003-11-05 20:19  guy
 
-	* wiretap/: network_instruments.c, network_instruments.h:
+	* packet-ansi_a.c:
 
-	  The time in Observer files is in nanoseconds since midnight,
-	  January 1, 2000, 00:00:00 *local* time.  The amount to add to
-	  that is just the UNIX time stamp value for that point in time;
-	  get it with "mktime()".
+	  Get rid of includes of <gmodule.h> (we're not loading any code at
+	  run-time in the dissector itself), and <sys/types.h> and
+	  <netinet/in.h> (we shouldn't need either of those, and they might
+	  be responsible for dragging in <sys/procset.h> on Digital UNIX -
+	  that header defines P_SID in a way that conflicts with our
+	  definition).
 
-2003-11-01 04:42  guy
+2003-11-05 20:10  guy
 
-	* ncp2222.py:
+	* plugins/acn/packet-acn.c:
 
-	  From Greg Morris:
+	  Update from Erwin Rol.
 
-		  1. Add several return values
-		  2. Fix the reply structure for NCP 22/48, 23/122
-		  3. Fix the request structure for NCP 23/121
-		  4. ServerID should be displayed as a hex value.
-		  5. ServerStationLong should be LE.
-		  6. JobNumberList should be JobNumberLong.
-		  7. Fix 87/64, 87/65, 87/66, 87/68, 87/69 - These are not
-	  Advanced Audit
-		     Service, they really 64 bit file support NCP's.
+2003-11-05 20:10  guy
 
-2003-11-02 03:55  tpot
+	* packet-ldap.c:
 
-	* packet-dcerpc-netlogon.c:
+	  A "GHashFunc()" returns a "guint", not a "gint".
 
-	  In a secure channel verifier, sometimes the nonce isn't present;
-	  not sure why this is so.
+2003-11-05 09:04  sahlberg
 
-2003-11-02 19:31  gerald
+	* gtk/ldap_stat.c:
 
-	* gtk/main.c:
+	  the LDAP service response time dialog
 
-	  The recent link-layer type additions broke "-r".  Make it work
-	  again.
+2003-11-05 09:04  sahlberg
 
-2003-11-02 22:12  gerald
+	* packet-ldap.c, packet-ldap.h, doc/ethereal.pod, gtk/Makefile.am,
+	  gtk/Makefile.nmake, gtk/service_response_time_table.c:
 
-	* ipproto.c, ipproto.h:
+	  update for LDAP measure the response time for some LDAP commands
+	  and add a service response time dialog for it
 
-	  Add an entry for Novell NCS heartbeats (ID 0xE0).
+2003-11-05 04:23  guy
 
-2003-11-02 23:12  gerald
+	* packet-wtp.c:
 
-	* packet-snmp.c, util.c, epan/filesystem.c, epan/filesystem.h,
-	  plugins/plugin_api_list.c, plugins/asn1/packet-asn1.c:
+	  From Olivier Biot: have a separate subtree ett_ value for
+	  concatenated PDUs.
 
-	  From Gisle Vanem:
+2003-11-04 22:14  guy
 
-	  * Added a new function get_file_in_temp() to	 epan/filesystem.c.
-	  This because of asn1.dll plugin which   had code to write to a
-	  log-file "c:\temp\ethereal.log". I feel   this patch makes this
-	  safer; I don't even have a c:\temp dir.
+	* wiretap/snoop.c:
 
-	  * Patched packet-asn1.c to use get_file_in_temp().
+	  Sigh.  Some snoop files have 6 bytes of padding; assume the file
+	  is a snoop file unless it has enoguh padding to hold a Shomiti
+	  trailer record.  (DEAR SUN MICROSYSTEMS: PLEASE DO NOT STUFF 16
+	  OR MORE BYTES OF PADDING INTO A SNOOP PACKET.  THANK YOU.  HAVE A
+	  NICE DAY.)
 
-	  * Added some #undef to packet-snmp.c to silence gcc.
+	  Add a little paranoia about the record and captured data lengths.
 
-	  * Changed "%u" -> "%lu" formats in util.c
+2003-11-04 21:44  guy
 
-	  Rename get_file_in_temp() to get_tempfile_path() to match other
-	  function names.
+	* packet-isakmp.c:
 
-2003-11-02 23:24  gerald
+	  Get rid of all the UDP-encapsulated IPSec stuff.
 
-	* packet-gsm_a.c:
+	  draft-ietf-ipsec-udp-encaps-06 says "The UDP port numbers are the
+	  same as used by IKE traffic, as defined in [Kiv05]", and
+	  "[Kiv05]" is draft-ietf-ipsec-nat-t-ike-05, which has been
+	  superseded by draft-ietf-ipsec-nat-t-ike-07; the latter appears
+	  to imply that it goes over port 4500, not port 500.
 
-	  From Michael Lum: Add decode of Bearer Capability parameter.
+	  We already have a draft-ietf-ipsec-udp-encaps-06 dissector for
+	  port 4500 (packet-ipsec-udp.c), and it's not entirely clear to me
+	  how, if draft-ietf-ipsec-udp-encaps-06 traffic *did* go over port
+	  500, you'd distinguish it from regular ISAKMP traffic.
 
-2003-11-03 02:41  gerald
+2003-11-04 20:33  guy
 
-	* epan/filesystem.c:
+	* gtk/prefs_dlg.c:
 
-	  Make get_tempfile_path() work like get_datafile_path(), which
-	  removes a dependency on PATH_MAX (which apparently isn't defined
-	  under Windows).
+	  Skip the check for the module if a page doesn't have a module.
 
-2003-11-03 02:46  gerald
+2003-11-04 20:22  guy
 
-	* plugins/: Xass-list, Xplugin_api.c, Xplugin_api.h,
-	  Xplugin_api_decls.h, Xplugin_table.h:
+	* gtk/prefs_dlg.c:
 
-	  Export get_tempfile_path() to the plugins.
+	  Clean up the way we find the properties page for a given protocol
+	  somewhat; this fixes a bug wherein we were assuming all modules
+	  necessarily had preference pages (they won't, if the only
+	  preferences for the module are obsolete preferences).
 
-2003-11-03 10:15  guy
+2003-11-04 19:47  guy
 
-	* AUTHORS, packet-wsp.c, packet-wsp.h:
+	* packet-wsp.c:
 
 	  From Olivier Biot:
 
-		  * Use function reference arrays for header parse function
-	  lookup
-		    (avoids switch())
-
-		  * Use macros for the common parts of the header parse
-	  functions
-
-		  * Use macros for identical header dissection functions
-
-		  * Implement many missing header parse functions
-
-		  * Automatic header parse error notification
+		  * Extra headers: Encoding-Version, WWW-Authenticate,
+		    Proxy-Authenticate
 
-		  * Some minor edits
+		  * Fix memory allocate/free in header parsing functions &
+	  macros
 
-		  * Add HTTP equivalents to WSP status codes
-
-2003-11-03 10:36  guy
-
-	* gtk/capture_combo_utils.c:
-
-	  "g_list_first()" takes an arbitrary list entry as an argument;
-	  you don't need to call it if you already have a pointer to the
-	  first entry in the list, which is what "a pointer to the list"
-	  is.
-
-2003-11-03 10:40  guy
-
-	* gtk/capture_dlg.c:
-
-	  Don't try to get the list of link-layer header types for an
-	  interface if the interface isn't in the list of known (local)
-	  interfaces - that way we don't try to get entries if the user's
-	  in the middle of changing the "Interface:" text, or if the
-	  interface is an rpcap: URL (getting the link-layer header type
-	  list for a remote interface is currently impossible, and even
-	  just getting its *default* interface type could hang for a long
-	  time if the remote machine isn't responding).
-
-	  Free the link-layer header type list when we're done with it.
-
-	  Label the option menu for that list "Link-layer header type", as
-	  it doesn't control the data link type of the interface in the
-	  only case I know of where it's settable, namely 802.11 interfaces
-	  - it just controls whether the packets you get from the interface
-	  have a fake Ethernet header or a real 802.11 header.
-
-2003-11-03 20:45  guy
-
-	* gtk/rtp_analysis.c:
-
-	  From Lars Ruoff: fix a problem with wrong maximum delay
-	  calculation - first or marked packets are now ignored when
-	  calculating maximum delay.
-
-2003-11-03 20:49  guy
+		  * Add missing "ok = TRUE" for some header parsing
+	  functions
 
-	* packet-q931.c:
+		  * Header subfields all share the same subtree identifier
+	  (more
+		    intuitive)
 
-	  Fro Anders Broman: fix the meaning of the extension bit, and add
-	  some more bitfields.
+		  * Get rid of unused header fields and subtrees
 
-2003-11-03 20:57  guy
+		  * Some whitespace changes
 
-	* packet-q933.c:
+2003-11-04 18:37  guy
 
-	  Apply Anders Broman's changes to the Q.931 dissector to the Q.933
-	  dissector:
+	* plugins/asn1/packet-asn1.c, plugins/gryphon/packet-gryphon.c,
+	  AUTHORS, doc/ethereal.pod:
 
-		  fix the meaning of the extension bit, and add some more
-		  bitfields.
+	  From Masaki Chikama: fix static compilation of "generic ASN.1"
+	  and Gryphon plugins.
 
-2003-11-03 21:00  guy
+2003-11-04 16:44  gerald
 
-	* gtk/proto_draw.c:
+	* plugins/asn1/packet-asn1.c:
 
-	  From Olivier Abad: Insert the GTK+ 2.2.4
-	  "gtk_tree_view_expand_to_path()" when building with GTK+ 2.0[.x],
-	  so that we can build with 2.0[.x].
+	  #ifdef out "#include <gtk/gtk.h>", so that --disable-ethereal
+	  works again.
 
-2003-11-03 22:32  guy
+2003-11-04 08:16  guy
 
-	* epan/filesystem.c:
+	* packet-http.c:
 
-	  Use "g_get_tmp_dir()" to get the pathname of the temporary file
-	  directory.
+	  If we've already constructed a subset tvbuff, don't construct an
+	  identical one and throw away the first one.