to make sure we can load the font the user selected *and* the font that
would, in theory, be the boldface version of that font; if not, pop up
an error dialog, and don't nuke the font selection dialog (i.e., the
user either has to pick fonts that work, or give up in disgust and click
"Cancel").
Don't set "prefs.gui_font_name" until the "fetch" function is called -
stuff the name in a static variable when the user clicks "OK" in the
font selection dialog.
svn path=/trunk/; revision=2405
may be redrawing a packet window that displays a frame other than the
current frame; give it arguments to specify the raw frame data and
"frame_data" structure for the frame.
This requires that each packet window have, associated with it, a
pointer to the "frame_data" structure; that replaces the "cap_len" and
"encoding" fields in a "PacketWinData" structure, as those are just
copies of fields from the frame's "frame_data" structure.
"packet_hex_print()" needn't be passed both the start and length values
from a "field_info" structure - just pass it a pointer to that
structure, or NULL for "no field is selected in the packet". It also
needn't, any longer, be passed the "cap_len" and "flags.encoding" fields
of a "frame_data" structure - just pass it a pointer to that structure.
In "redraw_hex_dump_all()", don't redraw the hex dump pane of the main
window if there is no current frame.
svn path=/trunk/; revision=2404
"Edit" menu, pending the implementation of cut/copy/paste, so that users
don't think that there's something they can do (other than implementing
cut/copy/paste :-)) to be able to use those functions.
svn path=/trunk/; revision=2402
the text in all "Follow TCP Stream" windows;
the text in the help window if we have one up;
all hex dump windows;
when GUI preference changes are to be applied, so that font changes and
"Follow TCP Stream" color changes show up.
Update both the Roman and bold font when the font is changed.
Don't decrement the reference counts on the old Roman and bold fonts
until that's all done.
svn path=/trunk/; revision=2401
highlighting of the bytes, in the hex dump window, corresponding to a
selected field.
Also, make "remember_ptree_widget()" static, as it's not used outside
"gtk/proto_draw.c".
svn path=/trunk/; revision=2399
on the systems that have the problem we're working around, the
SNMP shared library is "libsnmp.so.0";
on those systems, there's also no "libsnmp.so" unless the user
has installed the UCD SNMP development package or has made a
symlink from "libsnmp.so" to "libsnmp.so.0" by hand, and we
don't want to force users to do that (some of them may be
sufficiently new to UNIX that they don't know how to do that);
the run-time linker, if told to load "libsnmp.so", won't
necessarily realize that it's the same object as "libsnmp.so.0",
and may load it again rather than using the already-loaded
object, which might not be a good idea.
svn path=/trunk/; revision=2398
state so I can dissect things correctly ...
Will have to clean out all the old non-tvbuff cruft as well.
Boy, I will be glad when we re-do Ethereal and have TCP segment re-assembly under control.
svn path=/trunk/; revision=2397
allocated by "g_malloc()", and one should always use "g_free()" to free
stuff allocated with "g_malloc()" (using "free()" works if GLib isn't
compiled with any special memory allocator debugging/profiling options,
but doesn't work if it is compiled with those options).
svn path=/trunk/; revision=2393
a "keep reading" boolean value is returned from the function.
This avoids having to hack around the fact that some file formats truly
do have records that start at offset 0. (i4btrace and csids have no
file header. Neither does the pppdump-style file that I'm looking at right now).
svn path=/trunk/; revision=2392
that these calls work on Win32. I still don't have a good reason as to
why this is necessary, but it fixes the problem. I'll continue looking
for a reason.
svn path=/trunk/; revision=2386
the (too big) buffer for 1024 network cards.
The code comes directly after the ideas in Steven's book (UNIX network
programming).
svn path=/trunk/; revision=2385
Remove tvb_get_stringz from quake dissector and add tvb_get_nstringz,
tvb_get_nstringz0, and tvb_strnlen to tvbuff.c.
Remove multiple definitions of pntohl and friends from various places
(except for wiretap) and put into pint.h
Consolidate duplicate code for turning FT_* enums into strings (ala the
glossary).
svn path=/trunk/; revision=2382
not just addresses. KAME kernel currently uses old format, so ethereal
does not work with current KAME kernel (KAME kernel will be corrected).
svn path=/trunk/; revision=2381
progress, but the beginnings of one exist (it's far from ready for prime
time, but it does read in capture files a lot quicker and consumes a
fair bit less memory).
Get rid of the item about a line-mode capture program - that's
Tethereal.
Get rid of the "Make lines in GTK Tree (proto_tree GUI) user-selectable"
item, as they're now user-selectable.
Put back the item about Kerberos V4 - we hae a V5 dissector, but no V4
dissector yet.
svn path=/trunk/; revision=2377
- add some missing Vines IP header fields in protocol tree
- add protocol type as display filter
- miscellaneous code cleaning
svn path=/trunk/; revision=2374
Add OLD_CHECK_DISPLAY_AS_DATA().
Add display filters.
Check the magic cookie to see if the decoding or frame is correct
(in most cases, we probably need TCP reassembly).
svn path=/trunk/; revision=2371
encapsulation type codes - for those libpcap type codes whose numerical
value is interpreted differently by different versions of libpcap,
include <pcap.h> if you can and, if you can, use what it defines to
control which Wiretap code we map those type codes to.
Also, map the new libpcap type codes introduced by libpcap 0.5.
svn path=/trunk/; revision=2369
examples of errors that generate core dumps, and suggest that a stack
trace from the debugger could be useful for *all* failures that produce
core dumps.
Note that the core dump file may be named "ethereal.core", and note that
"tethereal" rather than "ethereal" should be used in file names if it's
Tethereal that blew up.
svn path=/trunk/; revision=2368
capture, clear the per-frame data pointers of all frames in the capture,
as those pointers now refer to data that's been freed.
Do that to all frames even if the user stops the rescan in the middle -
and clear the "visited" flag for all frames as well.
svn path=/trunk/; revision=2361
- frame numbers are 1-origin, so 0 can be used as an "exception" value.
In the protocol tree for a reply, don't say that the reply is to frame
N, just say that it's to a request starting in frame N - a frame can
contain more than one request, and a request may take more than one
frame.
svn path=/trunk/; revision=2360
redissecting the packets due to some preference changing, as the
preference may affect what state information gets constructed as a
result of some particular frame being dissected.
We don't have to do it when filtering the packets, or colorizing them,
however.
svn path=/trunk/; revision=2359
Richard Sharpe