PS: If I use another host (not where Xvnc is) to run vncviewer, the locking is prevented. There must be some signaling on localhost that does not happen on Ethernet, or the timing is just slightly slower, that Solaris does not need to send zero length messages that taste like EOF. Maybe I'll route it through TCP_relay and see if that makes it slow without entangling another host in the reliability model. Best regards, David --------------------- In a message dated 6/11/2007 10:57:52 A.M. Eastern Daylight Time, DGPickett writes:
I have a variety of Solaris 8 boxes where I run VNC 4.1.1 with no problem, but now I have a Solaris 10 box with new 1593 MHz CPUs, and when vi sets the visual bell off on my xterm, vncviewer locks. I ran a truss, and saw that vncviewer got a zero return from read on the socket. Since it is just a different box, I suspect some subtle timing issue. There is a similar open thread on vncviewer using SSH, which might be the same problem. In the past on Solaris, I have occasionally seen sockets returning zero to a read() without being at EOF (connection closed or disconnected). We wrote a loop with poll(0,0,1) to insist on 50 consecutive return zero read() returns across 50 Milliseconds before declaring EOF to avoid this! Best regards, David ____________________________________ AOL now offers free email to everyone. Find out more about what's free from AOL at _AOL.com_ (http://www.aol.com/?ncid=AOLAOF00020000000437) . ************************************** Get a sneak peek of the all-new AOL at http://discover.aol.com/memed/aolcom30tour _______________________________________________ VNC-List mailing list [email protected] To remove yourself from the list visit: http://www.realvnc.com/mailman/listinfo/vnc-list
