"gtk_toggle_button_set_active()" to set the state of radio buttions;
"gtk_toggle_button_set_active()" doesn't exist in GTK+ 1.0[.x], and
"gtk_toggle_button_set_state()" is an alias for it in GTK+ 1.2[.x].
Compute the column widths in the summary display based on the longest
string in the column; recompute it whenever we update the columns.
svn path=/trunk/; revision=319
"Display/Options" dialog box.
"get_column_width()" should check "timestamp_type" against ABSOLUTE, not
against COL_ABS_TIME.
Clean up the code to check the state of the time stamp radio buttons in
the "Display/Options" dialog box.
Check in "timestamp.h", which we failed to check in on the previous
commit.
svn path=/trunk/; revision=318
is the same as "Tools/Capture", and "Display" has an "Options" item,
which pops up a dialog box to let you change the "default" time-stamp
column display format on the fly (the "default" is what the "-t"
command-line option sets), and have the display change when you do that.
Made infrastructure changes to make the immediate display update work.
Removed some unused functions, declared some functions used only in the
file in which they're defined "static", and removed some unnecessary
#includes.
svn path=/trunk/; revision=317
(standard convention in many GUIs).
Make "Save as" be "Save As", and make "A" be the menu mnemonic for it in
the "File" menu, with GTK 1.2.
svn path=/trunk/; revision=315
already appears in the list; the extra line is probably leftover from
cutting-and-pasting the author list from the AUTHORS file).
svn path=/trunk/; revision=314
"gtk_widget_show()" of the main window; that way:
1) any alert box popped up because a capture file specified with
the "-r" flag can't be read shows up after the main window
shows up, and is more likely to end up on top of the main
window, rather than behind it;
2) as the main window is up before we start reading a capture
file specified with the "-r" flag, the progress bar can show
our progress reading that file.
If the "-k" flag was specified, make sure "-i" and "-w" were specified
as well, rather than just dropping core trying to dereference NULL
"cf.iface" or "cf.save_file" pointers.
svn path=/trunk/; revision=313
to use the decode_*_bitfield() routines. This needed to happen anyway, but
I finally made the change so that I can figure out how I'm going to handle
bitfield fields in my experiment of changing the implementation of the
protocol tree in ethereal.
svn path=/trunk/; revision=311
influence came from
http://developer.apple.com/techpubs/mac/HIGuidelines/HIGuidelines-232.html
which has a section on dialog box and alert box messages. However,
we're largely dealing with technoids, not with The Rest Of Us, so I
didn't go as far as one perhaps should.)
Unfortunately, it looks like it's a bit more work to arrange that, if
you give a bad file name to the "-r" flag, the dialog box pop up only
*after* the main window pops up - it has the annoying habit of popping
up *before* the main window pops up, and sometimes getting *obscured* by
it, when I do that. The removal of the dialog box stuff from
"load_cap_file()" was intended to facilitate that work. (It might also
be nice if, when an open from the "File/Open" menu item fails, we keep
the file selection box open, and give the user a chance to correct
typos, choose another file name, etc.)
svn path=/trunk/; revision=310
we've shown the top-level window, because we want it to pop up on *top*
of the top-level window, rather than being hidden behind it.
Put a "strerror()"-based reason *why* the open failed in that dialog.
svn path=/trunk/; revision=308
GtkMenuEntry table, and add the comment from the equivalent part of the
GTK 1.2 tutorial before the GTK 1.2 GtkItemFactoryEntry table.
svn path=/trunk/; revision=307
doesn't modify anything pointed to by that argument, and that keeps us
from getting complaints if we pass a "const gchar *" to it.
svn path=/trunk/; revision=306
versions, PDU types, error statuses, and trap types to strings.
Set the protocol column to "SNMP", and set the info column to an error
message if the parse fails (rather than printing a message to the
standard error) and to the PDU type if it succeeds.
svn path=/trunk/; revision=305
of the packet count combo box; there's no need to do so (we don't
remember the string, just its value when converted to a number), and, as
we don't free what "g_strdup()" returns, and don't remember it to save
it later, we leak memory.
svn path=/trunk/; revision=302
source and destination port numbers, check both port numbers against the
specified port, rather than checking the lower of the two port numbers
against the specified port, just in case you happen to either have
1) the port number for that type being high enough that you can
get client sockets using it
or
2) client sockets using it for some other reason.
svn path=/trunk/; revision=301
dialog box, put "0 (Infinite)" first, so that we default to that rather
than to the number of packets in the last capture we read.
svn path=/trunk/; revision=300
"get_nbns_name()", make sure the array into which you put the name is
"big enough" - MAXDNAME+4 is more than big enough for those error
messages.
svn path=/trunk/; revision=299
"fd->cap_len - offset", i.e. END_OF_FRAME, is unsigned, so
while (END_OF_FRAME > 0) {
doesn't keep you out of the loop if "offset" is already beyond
"fd->cap_len", so you can try processing stuff past the end of the
captured data in the packet.
svn path=/trunk/; revision=298
request or reply. (Redid "get_dns_name()" along the lines of the code
in the BSD resolver.)
Add code to dissect SOA RRs.
svn path=/trunk/; revision=297
the NCP field table structure so that it is easier to add new request/reply
types. I even added a new type myself. There's still more work to be done;
so for now ignore the warnings that gcc emits.
svn path=/trunk/; revision=287
types in "packet-tftp.c", and grab the opcode directly from the first 2
bytes of the packet, using "pntohs()" (which will also keep us from
crashing, on processors that require strict alignment, if the UDP
payload isn't aligned on a 2-byte boundary).
Also use "pntohs()" to grab the block number in data and acknowledgment
packets, for the same reason we use it to grab the opcode.
svn path=/trunk/; revision=286
supplied by Tim Farley.
Tim also indicated that the Network Monitor network types may be NDIS
network types+1. It also appears that NetXRay/Windows Sniffer network
types may be NDIS network types as well.
svn path=/trunk/; revision=284
if you have the UCD or CMU SNMP library available. If you have the SNMP
library but do not with to have SNMP support, use the
./configure --disable-snmp
option. Otherwise 'configure' finds the SNMP library and uses it.
svn path=/trunk/; revision=281
can be used by dissectors other than ARP to display byte arrays as
strings of hex digits.
Add a routine to extract a null-terminated Unicode string and turn it
into an ISO 8859-1 string for display. (Ultimately, we should determine
what character sets the X server or printer or whatever can handle, and
turn it into the appropriate character set.)
Display the challenge in "core-to-LANMAN-2.1" Negotiate Protocol
responses as a string of hex digits - but only if the length is
non-zero. (It's a counted array, not a null-terminated string.)
Display some additional security mode bits in an NT LM 0.12 Negotiate
Protocol response.
Display some additional bits in the "capabilities" field of the
Negotiate Protocol response.
Display the challenge in an NT LM 0.12 Negotiate Protocol response as a
string of hex digits (it's a counted array, not a null-terminated
string).
Display the domain name as Unicode in an NT LM 0.12 Negotiate Protocol
response if the capabilities field has the "supports Unicode" bit set
(no, not the "Unicode" bit in the "flags2" field - NT doesn't set that
in the response, even though it sends the domain name over in Unicode!).
Display some additional bits in the "flags2" field of an SMB.
svn path=/trunk/; revision=275
Fix the "proto_tree_add_item()" used when the server didn't support any
of the dialects the client offered not to print the dialect index.
In the "Core to LANMAN 2.1" "negotiate protocol" decode, show the time
zone as minutes from UTC, just as is done for the "post-LANMAN 2.1"
decode.
svn path=/trunk/; revision=272
fixes, as I had to manually apply his fixes. It is a long story,
but my development environment is, umm, interesting.
Thanks to Guy for spotting the problems in my Core Proto to LanMan 2.0
decode probs.
I did not spot them and I do not have any such traces to check
against.
svn path=/trunk/; revision=271
line, so that they show up as SMB packets, rather than NBDS packets, in
the summary display.
Put SMB at the top level of the decode tree for NBDS packets, as is done
for NBSS packets.
svn path=/trunk/; revision=270
