razdan@chanakya.oakhill.uucp (Anshuman Razdan) (03/06/90)
I recently tried to compile X11R4 on apollos and almost succeded in completely compile and installation procedures. The compilation of the MIT part went ok but when I did the make install I ran in to couple of problems. (a) A Fatal Error was reported by cpp when compiling resource.c. It looked like this : installing in ./clients/xdm... rm -f resource.o cc -c -O -U__STDC__ -D_PROTOTYPES -I../../. -DBINDIR=\"/usr/X11R4/bin/X11\" -DXDMDIR=\"/usr/X11R 4/lib/X11/xdm\" -DSIGNALRETURNSINT -DTCPCONN -DNO_TCP_H -DOSMAJORVERSION=10 -DOSMINORVERSION=1 '-DDEF_SERVER_LINE=":0 secure /usr/X11R4/bin/X11/X :0"' '-DXRDB_PROGRAM="/usr/X11R4/bin/X11/xrdb"' '-DDEF_SESSION="/usr/X11R4/bin/X11/xterm -ls"' '-DDEF_USER_PATH=":/bin:/usr/bin:/usr/X11R4/bin/X11:/ usr/ucb"' '-DDEF_SYSTEM_PATH="/etc:/bin:/usr/bin:/usr/X11R4/bin/X11:/usr/ucb"' '-DDEF_SYSTEM_SHELL= "/bin/sh"' '-DDEF_FAILSAFE_CLIENT="/usr/X11R4/bin/X11/xterm"' '-DDEF_XDM_CONFIG="/usr/X11R4/lib/X11 /xdm/xdm-config"' '-DDEF_AUTH_FILE="/usr/X11R4/lib/X11/xdm/auth-server"' '-DCPP_PROGRAM="/lib/cpp"' Other than this I did not have any problem. I brought up the X server and it performed well except for the following. (a) The menus in the xterm window do not work (b) Some times it takes quite a few seconds to bring(po up) the twm menu. Particulary if it has not been poped up for a while. Sounds like paging. Altho I was trying it on a 32 meg m/c. The performance is certainly better than the R2 server I have been feeding on (Shared X). Any ideas How to get rid of the above problems? Any performance boosting ideas are welcome. I am still on 10.1 Anshuman Razdan Sector CAD Motorola razdan@chanakya.sps.mot.com
dpg@citi.umich.edu (David Gorgen) (03/06/90)
In article <RAZDAN.90Mar5112300@chanakya.oakhill.uucp> razdan@chanakya.oakhill.uucp (Anshuman Razdan) writes: > > I recently tried to compile X11R4 on apollos and almost succeded > in completely compile and installation procedures. The compilation > of the MIT part went ok but when I did the make install I ran in to > couple of problems. > > (a) A Fatal Error was reported by cpp when compiling resource.c. It looked > like this : > > installing in ./clients/xdm... > rm -f resource.o > cc -c -O -U__STDC__ -D_PROTOTYPES -I../../. -DBINDIR=\"/usr/X11R4/bin/X11\" -DXDMDIR=\"/usr/X11R > 4/lib/X11/xdm\" -DSIGNALRETURNSINT -DTCPCONN -DNO_TCP_H -DOSMAJORVERSION=10 -DOSMINORVERSION=1 > '-DDEF_SERVER_LINE=":0 secure /usr/X11R4/bin/X11/X :0"' '-DXRDB_PROGRAM="/usr/X11R4/bin/X11/xrdb"' > '-DDEF_SESSION="/usr/X11R4/bin/X11/xterm -ls"' '-DDEF_USER_PATH=":/bin:/usr/bin:/usr/X11R4/bin/X11:/ > usr/ucb"' '-DDEF_SYSTEM_PATH="/etc:/bin:/usr/bin:/usr/X11R4/bin/X11:/usr/ucb"' '-DDEF_SYSTEM_SHELL= > "/bin/sh"' '-DDEF_FAILSAFE_CLIENT="/usr/X11R4/bin/X11/xterm"' '-DDEF_XDM_CONFIG="/usr/X11R4/lib/X11 > /xdm/xdm-config"' '-DDEF_AUTH_FILE="/usr/X11R4/lib/X11/xdm/auth-server"' '-DCPP_PROGRAM="/lib/cpp"' This is a compilation, not an installation (note that "make install" rebuilds anything not yet build or whose dependencies have changed). The problem is the large number of -D options. The SR10.1 cpp dies if there are too many of them; this is fixed in SR10.2 (at least, any limit there might be is much higher). You don't need xdm anyway, unless you are ready to abolish the use of the DM completely. > > Other than this I did not have any problem. I brought up the X server and it performed well except > for the following. > > (a) The menus in the xterm window do not work You may be assuming that the white control key will work properly with the mouse buttons. This is true for the Apollo product server, and it is true if you build the R4 server explicitly for SR10.2 (see mit/config/apollo.cf) and run it on an SR10.2 system. However, the R4 server built for SR10.1 has the old MIT server shortcomings about not noticing up-downs on any white keys. You have to use black-key synonyms for modifiers with the mouse; see mit/server/ddx/apollo/Xapollo.man for details. > > (b) Some times it takes quite a few seconds to bring(po up) the twm menu. Particulary > if it has not been poped up for a while. Sounds like paging. Altho I was trying > it on a 32 meg m/c. I don't know what it could be other than paging, although 32 Megs of memory should be plenty. It depends what applications you are running and what the paging load from other nodes on the network might be. > > The performance is certainly better than the R2 server I have been feeding on (Shared X). Share mode certainly has a performance cost, although we are working on reducing it. No further details are publicly available yet. -- Dave Gorgen / GTD-East (formerly Apollo Computer), Hewlett-Packard Company located at: University of Michigan, CITI dpg@citi.umich.edu (Center for Information Technology Integration) 313-998-7482 or -7479
tomf@boulder.Colorado.EDU (FREDERICKS THOMAS M) (03/12/90)
Is it possible to ftp the source for X11R4 that supports os 10.2? If so where? Thanks, Tom....
krowitz@RICHTER.MIT.EDU (David Krowitz) (10/20/90)
I've just got up the sources for X11 R4 (uncompressed, untar'd, and patched) and I'm currently compiling the clients and libraries for our Alliant FX/40 using "pcc" (the free portable C compiler which is part of the base OS). The idea is to be able to execute applications of the Alliant with the X output showing up on our Apollo and Sun workstations. The Alliant itself has no graphics screens, so it doesn't need the X11 R4 server built. The Apollo side is a little different ... the X sources come with a predefined option for building both the client and server sides of X11 R4 ... but I'm not certain if this server is a native-X only server. Someone recently sent a message that implied that a share-mode X11 R4 server existed already. Is this the server on the source tapes? -- David Krowitz krowitz@richter.mit.edu (18.83.0.109) krowitz%richter.mit.edu@eddie.mit.edu krowitz%richter.mit.edu@mitvma.bitnet (in order of decreasing preference)