[comp.windows.x] Apollo, xterm, & cpp

RJOHNS%BNR.CA@MITVMA.MIT.EDU (Richard Johns, R.) (01/28/89)

I'd appreciate it if someone who's managed to actually get imake working
on an Apollo (SR9.7) would post the necessary steps.  Also, the X11R3
xterm won't insert lines when I type "o" in vi.  I've heard rumors of a
termcap fix for this.  It doesn't seem to be affected by the xterm fixes.

Thanks--
Rick Johns
rjohns@bnr.ca

msy@demon.siemens.com (Marcus S. Yoo) (01/31/89)

In article <8901272331.AA13579@ATHENA.MIT.EDU> RJOHNS%BNR.CA@MITVMA.MIT.EDU (Richard  Johns, R.) writes:
>
>I'd appreciate it if someone who's managed to actually get imake working
>on an Apollo (SR9.7) would post the necessary steps.  Also, the X11R3
>xterm won't insert lines when I type "o" in vi.  I've heard rumors of a
>termcap fix for this.  It doesn't seem to be affected by the xterm fixes.
>
>Thanks--
>Rick Johns
>rjohns@bnr.ca


I've been having the same problems (both problems).
To get imake to work, I tried to use cpp provided with X11R3 distribution,
but that didn't work either. Anyone know if Apollo fixed cpp for SR10??

The xterm problem is also a real pain. Maybe some knowlegeable should
post a fix for this problem. I'd sure appreciate it.

While we're at it, I have another problem with X (on Apollo SR9.7).
For both X11R2 and X11R3, there seem to be some limit on the number of
X client programs that can be running at same time. At most I can
invoke 9 clients (xterm, xclock, etc). I know that this has nothing to do
with /dev/*typ* because it doesn't matter if the client programs are
xterm or xclock. I can't get more then 9 clients running at a time.
(if I use a shell script to invoke Xapollo, the limit is reduced to
7 client programs). Anyone know what's going on?

Also, when I run xdm, I don't get the login prompt. In fact, X server
comes up and doesn't do anything. When I tried to exit Xserver, it
is temporarily kill, and then restarts again (I know, it's supposed to
do that) thus I am stuck with xdm that doesn't do anything (I had to
rlogin from some place else and kill xdm). I've compiled xdm without
any modification and all the default files were installed
(/usr/lib/X11/xdm/... ). What did I do wrong (what didn't I do [right])?

Thanks for any help you can give me....
Marcus S. Yoo	msy@demon.siemens.com
	...{ihnp4|astrovax|princeton|drexel}!siemens!demon!msy
SIEMENS Corporate Research, Inc.		tel: 609) 734 - 3313
755 College Rd East, Princeton, NJ 08540	fax: 609) 734 - 6565

kek@DINORAH.WUSTL.EDU (Ken Krippner) (02/02/89)

 MSY writes:

>In article <8901272331.AA13579@ATHENA.MIT.EDU> RJOHNS%BNR.CA@MITVMA.MIT.EDU (Richard  Johns, R.) writes:
>>
>>I'd appreciate it if someone who's managed to actually get imake working
>>on an Apollo (SR9.7) would post the necessary steps.  Also, the X11R3
>>xterm won't insert lines when I type "o" in vi.  I've heard rumors of a
>>termcap fix for this.  It doesn't seem to be affected by the xterm fixes.
>>
>>Thanks--
>>Rick Johns
>>rjohns@bnr.ca
>
>
>I've been having the same problems (both problems).
>To get imake to work, I tried to use cpp provided with X11R3 distribution,
>but that didn't work either. Anyone know if Apollo fixed cpp for SR10??

I posted, a year or so ago, my handstands doing Makefiles on another
machine to get R2 (server only) built on DN3000.  Well, yes, the long
lines problem in the compiler got fixed in SR10 so it was possible to
unbreak cpp so that imake works (as far as I can tell). (actually cc 6.00)

>I have another problem with X (on Apollo SR9.7).
>For both X11R2 and X11R3, there seem to be some limit on the number of
>X client programs that can be running at same time. At most I can
>invoke 9 clients (xterm, xclock, etc). I know that this has nothing to do
>with /dev/*typ* because it doesn't matter if the client programs are
>xterm or xclock. I can't get more then 9 clients running at a time.
>(if I use a shell script to invoke Xapollo, the limit is reduced to
>7 client programs). Anyone know what's going on?

Nope except to say that the same limit is on SR10 so don't get
hopeful.  I always imagined it to be related to the maximium number of
pseudoterminals, namely 16.  It is running out of some kind of resource,
maybe someone at Apollo will see this and comment or even offer a patch!

While we are all interested in Apollo problems, anybody (Casey?) know
where the /sys/ins/base.ins.c and /sys/ins/time.ins.c included in the 
XTOP/lib/Xaw/Load.c is supposed to be?  I haven't found it yet.  Is this
a work-around hack left over from SR9.7 and/or did I miss one of the patches
which (thanks, Casey!) came over xpert?

On another topic, I have an error message while linking Xapollo:
	ld libapc.a fatal: bad relocation format
I called the HOTLINE and was told that I was the second one today
with this kind of message.  Stay tuned for the outcome.  There was
mention of a patch tape being made for ld(1)...

TIA
Ken Krippner
	{uunet...}!wucs1!dinorah!kek	or
	wucs1.wustl.edu!dinorah!kek	or
	kek@dinorah.wustl.edu		or
	314-362-2950

klt@ntcsd1.UUCP (Kristopher L. Tyra) (02/03/89)

In regards to the file libapc.a, I to had the problem and I found the 
answer.  It seems the the systype and the runtype for the object file
ddx/apollo/apcfont.bin which was supplied by the apollo users group was
compile in sr9.7 and therefore is not supposed to work in sr10.1.  This
can be fixed by changing this in the Imakefile or (Makefile).

apcfont.bin: apcfont.tar
	-$(RM) apcfont.bin
	tar -xvf apcfont.tar
	obj2coff apcfont.bin tmp.o
	ld -A sys,any -A run,any -r -o apcfont.bin tmp.o
	-$(RM) tmp.o

This seems to fix the problem I don't know if it really makes the code
work though.  There is also another problem with the server communications.
It seems that server want's to occasionally send out blocks of greater than
9.1K which cause the server to get an EWOULBLOCK all the time when using the
UNIX protocol and a failer errno=40 (Message to Big or something) when 
using the TCP protocol.  I have a correction for this but it is not so simple
mail me direct if you want the patch.

Now since we are propagating problems.  WHAT THE DAYLIGHTS IS WRONG WITH THE
COLOR on sr10.1 (I does wierd things)  APPOLLO If you are out there please
respond or better yet how about the APOLLO Users Group who is selling a SR10.1
version (R2) but is mum about what changes they have made.  Anybody please
I think there are a lot of APOLLO DN3000 and DN4500 out there that would
like a responce.  Thanks.