[comp.windows.x] Making X.V11R1 on Sun 3

RWS@ZERMATT.LCS.MIT.EDU (Robert Scheifler) (09/25/87)

    Date: Thu, 24 Sep 87 10:58:26 edt
    From: Alan Schultz <schultz@nrl-aic.ARPA>

    WHAT I'VE ALREADY DONE:
	    0) applied all patches
	    1) made minor change to Sun.macros
	    2) make Makefiles
	    3) make Makefile
	    4) make depend
	    5) per advice from David Stampf, changed servers/ddx/sun/Makefile
	       as follows...
		  change:    all:: $(OFILES) kbd_mode 
		      to:    all:: $(OFILES) kbd_mode $(OBJS)

I don't know why you would want to do this, because making the Makefile
from the Imakefile should add an all:: line for libsun.a, which is what
you really want.  (However, I'm assuming you've applied the patch we
sent out for this Imakefile.)

	    QUESTION! Do I need to do make depend again now?

No.

	    QUESTION: In servers/ddx/sun/README, we are instructed to "make
    sure that the fonts/Makefile uses the pad flag on the font compiler to pad
    the glyphs to 4-byte bounderies." I see nothing in fonts/Makefile
    concerning this. However, in fonts/compiler/fc.h I see...
	
	    #ifdef sun
	    #   define DEFAULTGLPAD 4;
	    ...
	    #endif

    Does this take care of it? If not, what is the exact name of the pad flag
    and how exactly should this be done.

That should take care of it.

					 Also, the #define above ends with a
    semicolon, and that seems wrong!?!?

You are right, that seems wrong.

	    QUESTION: Have I missed anything else????

No, let 'er rip.

schultz@NRL-AIC.ARPA (Alan Schultz) (09/25/87)

I know my knowledge of system programming is not perfect, but I'm really
getting frustrated.

I had no problems installing X.V10R4 on a Sun 3/260 Color under 3.2.
 
First, let me be more specific: the machine is a Sun 3/160 with the COLOR
monitor. Op sys 3.4 is running on it. Device wise, it has bwtwo0 and
cgtwo0 (and console in addition to all the normal things.) We just set
this machine up last week, and the only thing we changed in the
configuation was MAXUSERS to 8 instead of 4 for larger system tables.
 
    WHAT I'VE ALREADY DONE:
	    0) applied all patches
	    1) made minor change to Sun.macros
		Specifically, I changed the man suffix to l instead of n
 		(we use local instead of new)
	    2) make Makefiles
	    3) make Makefile
	    4) make depend
	    5) per advice from David Stampf, changed servers/ddx/sun/Makefile
	       as follows...
		  change:    all:: $(OFILES) kbd_mode 
		      to:    all:: $(OFILES) kbd_mode $(OBJS)
		NOTE: Robert Scheifler has already responded that this is
                      incorrect, but I don't think this will affect the
 		      results. I also tried undoing this change by making clean
                      at that level and remaking: results were the same.
	6) make -k
		Everything was made without error except for GNUplot,
                which was expected.
	7) su root; make -k install
		This completed without error, except again for Gnuplot.
		I also added a link from /usr/bin/X 
		  to /usr/bin/Xsun, so that the server would be named X.
	8) I then logged in to the console and,
		a) added /usr/bin/X11 to my path
		b) added setenv DISPLAY `hostname`:0 to .login
		c) logged out, and logged back in again.
	9) Now, from the console (not in SunWindows) I tried to start X
		I tried:
 			"xinit" (as recommended in ddx/sun/README)
			"Xsun &" ( as recommended in release notes)
			"xinit -- Xsun"
			"xinit -- Xsun 0" (the old way under X.V10R4)
			"X &"
  		
PROBLEM: 
   The screen darkens starting from the top of the screen, moving
slowly down the screen. It takes ~6 - 8 seconds for the screen to be
darkened! With X.V10R4, it happened almost instantaneously.
NEXT, NOTHING HAPPENS! No large X cursor, nothing. A ps from another
terminal shows (depending on how I start it) xinit and X as existing, but
in an "I" state. I wait for 5 minutes, and still nothing, so I then kill
the processes from another terminal.

HELP!


   Alan C. Schultz
   Code 5510
   Navy Center for Applied Research in Artificial Intelligence (NCAR A I)
   Naval Research Laboratory
   Washington, D.C. 20375-5000
   ARPA: schultz@nrl-aic.arpa
   (202) 767-2877

RWS@ZERMATT.LCS.MIT.EDU (Robert Scheifler) (09/26/87)

    Date: Fri, 25 Sep 87 16:40:34 edt
    From: Alan Schultz <schultz@nrl-aic.ARPA>

    First, let me be more specific: the machine is a Sun 3/160 with the COLOR
    monitor. Op sys 3.4 is running on it. Device wise, it has bwtwo0 and
    cgtwo0 (and console in addition to all the normal things.) We just set
    this machine up last week, and the only thing we changed in the
    configuation was MAXUSERS to 8 instead of 4 for larger system tables.
  		
    PROBLEM: 
       The screen darkens starting from the top of the screen, moving
    slowly down the screen. It takes ~6 - 8 seconds for the screen to be
    darkened! With X.V10R4, it happened almost instantaneously.
    NEXT, NOTHING HAPPENS! No large X cursor, nothing. A ps from another
    terminal shows (depending on how I start it) xinit and X as existing, but
    in an "I" state. I wait for 5 minutes, and still nothing, so I then kill
    the processes from another terminal.

    HELP!

Your confusion is covered in the Xsun man page; I guess you didn't read
it.  The solution is to remove the bwtwo0 entry from /dev.  With it
there, the server autoconfig assumes you actually have a screen attached
to that frame buffer, and since the mono screen is defined as screen 0,
the initial xterm and cursor is placed there, where you can't see it.

Note that color is *slow* on the Sun.  You can use the -mono option to
the server to get speed at the expense of color.