I wasn't actually complaining about the exit, just noting the behaviour as a good tester :)

Testing this branch:

mstevens@mstevens-desktop:~/elim % ./elim-client </dev/urandom
51 NULs in input, abort.

During build:


cc -Wall -g -O2 -std=c99 -D_GNU_SOURCE -DHAS_ADD_SECONDS -DHAS_QUEUE_INIT -DHAS_GET_HASH_KEYS  -I/usr/include/glib-2.0 -I/usr/lib/glib-2.0/include    -I/usr/include/libxml2    -I/usr/include/libpurple -I/usr/include/glib-2.0 -I/usr/lib/glib-2.0/include      -c -o xnode/xnode.o xnode/xnode.c
xnode/xnode.c: In function ‘xnode_first_child_tag’:
xnode/xnode.c:37: warning: value computed is not used

Well… it's not at the top of my list or anything, but
I'll have another look and see if there are any easy
wins I hadn't thought of before.

mstevens@mstevens-desktop:~/elim % ./elim-client </dev/urandom                 
zsh: segmentation fault (core dumped)  ./elim-client < /dev/urandom            
                                                                               
Core was generated by `./elim-client'.                                         
Program terminated with signal 11, Segmentation fault.                         
#0  0x0804e1cd in xnode_first_child_tag ()                                     
(gdb) bt                                                                       
#0  0x0804e1cd in xnode_first_child_tag ()                                     
#1  0x0804d5cf in sexp_handler ()                                              
#2  0x0024ce5b in ?? () from /lib/libglib-2.0.so.0                             
#3  0x00205855 in g_main_context_dispatch () from /lib/libglib-2.0.so.0        
#4  0x00209668 in ?? () from /lib/libglib-2.0.so.0                             
#5  0x00209ba7 in g_main_loop_run () from /lib/libglib-2.0.so.0                
#6  0x0804d965 in main ()