I have spent the last 2 weeks hunting for the problem in thefldigi/jpskmail combination.I have changed the server to incorporate lots of debugging routinesto catch the errors...I have found an ultra-seldom race condition in the interface handler for ARQ,and fixed it.I have run the test software on PI4TUE during the weekend, and it looked good.This morning I found both server and fldigi running, but clearly fldigi had stoppedservicing the ARQ socket, and also the display was frozen.The server could still write to fldigi, so the socket was still alive at the fldigi side.I guess the misc/arqio routine had crashed...I am now stopping work on this, it is costing too much of my time. I am nowwaiting for the fldigi people to take action.Here is the backtrace (fldigi still running, but display and ARQ socket not beingserviced):0x0012d422 in __kernel_vsyscall ()(gdb) bt#0 0x0012d422 in __kernel_vsyscall ()#1 0x0022b245 in sem_wait@@GLIBC_2.1 ()from /lib/tls/i686/cmov/libpthread.so.0#2 0x081a2f83 in trx_wait_state () at trx/trx.cxx:640#3 0x0809b67d in init_modem_sync (m=42, f=0) at dialogs/fl_digi.cxx:1242#4 0x080f9f00 in __call<, 0, 1> (this=0xb75baf88, run=true)at /usr/include/c++/4.4/tr1_impl/functional:1137#5 operator()<> (this=0xb75baf88, run=true)at /usr/include/c++/4.4/tr1_impl/functional:1191#6 func_wrap<std::tr1::_Bind<void (*()(int, int))(int, int)> >::destroy(bool)(this=0xb75baf88, run=true) at ./qrunner/fqueue.h:48#7 0x0816756b in fqueue::pop (fd=17, arg=0x82f72f0) at qrunner/fqueue.h:96#8 fqueue::execute (fd=17, arg=0x82f72f0) at qrunner/fqueue.h:102---Type <return> to continue, or q <return> to quit---#9 qrunner::execute (fd=17, arg=0x82f72f0) at qrunner/qrunner.cxx:106#10 0x002c2ad0 in fl_wait(double) () from /usr/lib/libfltk.so.1.1#11 0x00270b6b in Fl::wait(double) () from /usr/lib/libfltk.so.1.1#12 0x00270ce4 in Fl::run() () from /usr/lib/libfltk.so.1.1#13 0x080f2d11 in main (argc=1, argv=0xbffff494) at main.cxx:460(gdb)The event log was still alive, but only output RSID notifications...I am really getiing fed up with this, and will pursue the V2 modem route...Rein PA0R