jlv@CS.BROWN.EDU (02/14/89)
We are writing a fairly large package using Xlib on a sun3. Occasionally, we get a crash of the X server, which in turn destroys everything and jams the keyboard, forcing us to do a kbd_mode to reset the keyboard. I guess we are in some way abusing X, but we have no way of knowing how. Does anyone know of any documented (of undocumented) reasons for X Server crashes? This would be extremely helpful. Jeff Vogel and Dave Sklar Brown University.
rhl@grendel.Princeton.EDU (Robert Lupton (the Good)) (04/06/91)
I have been having trouble running the MIT Xserver on our IPC's. When you try to move the mouse off the edge of the screen the server hangs (or at least, you get no further response). I can run the openwin server but then I get into all sorts of problems with the infamous widget version number mismatch and so on. I can work around them, but I'd rather use Xsun --- any suggestions? Robert
gcardwel@okazaki.acs.uci.edu (Guy Cardwell) (04/06/91)
In article <7945@idunno.Princeton.EDU>, rhl@grendel.Princeton.EDU (Robert Lupton (the Good)) writes: |> I have been having trouble running the MIT Xserver on our IPC's. When |> you try to move the mouse off the edge of the screen the server hangs |> (or at least, you get no further response). chuckle chuckle.... The IPC has *two* framebuffers, a cgfour (or cgsix) in the s-bus slot and a monochrome frame buffer on the motherboard. When you move the mouse of off the edge of the color screen, it moves to the monochrome screen, which you can't see because you dont have a mono monitor plugged it there! Try specifying the frame buffer when you start the server.. Xsun -dev /dev/cgfour0 or something along those lines. Enjoy,... and no, I dont know of a cheap source for sun mono monitors to have a two headed systems... I wish I did!!1 Guy
adrianho@barkley.berkeley.edu (Adrian J Ho) (04/06/91)
In article <27FD104E.5282@orion.oac.uci.edu> gcardwel@okazaki.acs.uci.edu (Guy Cardwell) writes: >In article <7945@idunno.Princeton.EDU>, rhl@grendel.Princeton.EDU (Robert Lupton (the Good)) writes: >|> I have been having trouble running the MIT Xserver on our IPC's. When >|> you try to move the mouse off the edge of the screen the server hangs >|> (or at least, you get no further response). >chuckle chuckle.... >The IPC has *two* framebuffers, a cgfour (or cgsix) in the s-bus slot >and a monochrome frame buffer on the motherboard. When you move the mouse >of off the edge of the color screen, it moves to the monochrome screen, which >you can't see because you dont have a mono monitor plugged it there! >Try specifying the frame buffer when you start the server.. >Xsun -dev /dev/cgfour0 >or something along those lines. Nope, that's the *default* (which causes the problem). Use: Xsun -zaphod or xinit -- :0 -zaphod See the Xsun(1) man page for full details.
gcardwel@okazaki.acs.uci.edu (Guy Cardwell) (04/06/91)
In article <ADRIANHO.91Apr5183438@barkley.berkeley.edu> adrianho@barkley.berkeley.edu (Adrian J Ho) writes: >In article <27FD104E.5282@orion.oac.uci.edu> gcardwel@okazaki.acs.uci.edu (Guy Cardwell) writes: > >>Try specifying the frame buffer when you start the server.. > >>Xsun -dev /dev/cgfour0 > >>or something along those lines. > >Nope, that's the *default* (which causes the problem). Use: > >Xsun -zaphod or xinit -- :0 -zaphod Sorry, but -zaphod only prevents the mouse from moving off the screen. It does not disable the other screen! My example -dev /dev/cgfour0 is correct with the exception that the frame buffer is /dev/cgthree0. The cgfour is a color display with a mono overlay plane display which can be used as two X displays. It is NOT the framebuffer on the IPC. The IPC uses a cgthree which is a very normal color board, and a bwtwo on the mother board. The default, by the way, is to poll and initialize the available frame bufffers. -dev /dev/cgthree0 does the right thing on an IPC, believe me! Apologies for the confusion! Guy
guy@auspex.auspex.com (Guy Harris) (04/09/91)
>I can run the openwin server but then I get into all sorts of problems >with the infamous widget version number mismatch and so on. Those problems are actually a consequence of using the Open Windows version of "-lXt", *NOT* of using the Open Windows server. The problem is that the MIT patch that fixed the version number came out too late to get that patch into OW 2.0.
steve@lia (Stephen Williams) (04/10/91)
In article <ADRIANHO.91Apr5183438@barkley.berkeley.edu> adrianho@barkley.berkeley.edu (Adrian J Ho) writes: >In article <27FD104E.5282@orion.oac.uci.edu> gcardwel@okazaki.acs.uci.edu (Guy Cardwell) writes: > >>In article <7945@idunno.Princeton.EDU>, rhl@grendel.Princeton.EDU (Robert Lupton (the Good)) writes: >>|> I have been having trouble running the MIT Xserver on our IPC's. When >>|> you try to move the mouse off the edge of the screen the server hangs >>|> (or at least, you get no further response). The problem is that Xsun searches for the frame buffers from a list, as well as /dev/fb. With the IPC, the file /dev/fb and /dev/bwtwo0 exist. One hack would be to remove one or the other of these nodes. There was also another problem where Xsun would not support multiple frame buffers, so in the process of fixing that, I fixed the /dev/fb problem. I would like to send this change to someone official so that it makes it into the rest of the world. Can anybody tell me who where, and how I should send this fix? --Steve Williams steve@lia.com