Re: matlab 5.3 seg fault (can it use both libc.so.6 libc.so.5 ?)
On Wed, 14 Apr 1999 [EMAIL PROTECTED] wrote: Hi all, Mathworks tech support has been unable to help get matlab 5.3 running on my Debian system. I was wondering if anybody else has got it working and perhaps would know how to fix my problem. When I start matlab is says, Segmentation fault ... This is what I see for shared library linking. ldd /usr/local/matlab5.3/bin/lnx86/matlab libut.so = /usr/local/matlab5.3/extern/lib/lnx86/libut.so (0x4000b000) libmwhardcopy.so = /usr/local/matlab5.3/bin/lnx86/libmwhardcopy.so (0x4002b000) libmwhg.so = /usr/local/matlab5.3/bin/lnx86/libmwhg.so (0x4004f000) libmwsimulink.so = /usr/local/matlab5.3/bin/lnx86/libmwsimulink.so (0x40149000) libmwgui.so = /usr/local/matlab5.3/bin/lnx86/libmwgui.so (0x404ae000) libmwnumerics.so = /usr/local/matlab5.3/bin/lnx86/libmwnumerics.so (0x40545000) libmwmpath.so = /usr/local/matlab5.3/bin/lnx86/libmwmpath.so (0x405b3000) libmwuix.so = /usr/local/matlab5.3/bin/lnx86/libmwuix.so (0x405c2000) libmwcompiler.so = /usr/local/matlab5.3/bin/lnx86/libmwcompiler.so (0x40854000) libmatlbmx.so = /usr/local/matlab5.3/bin/lnx86/libmatlbmx.so (0x408fa000) libdl.so.1 = /lib/libdl.so.1 (0x40959000) libg++.so.27 = /usr/local/matlab5.3/sys/os/lnx86/libg++.so.27 (0x4095c000) libstdc++.so.27 = /usr/local/matlab5.3/sys/os/lnx86/libstdc++.so.27 (0x4099) libm.so.5 = /usr/local/matlab5.3/sys/os/lnx86/libm.so.5 (0x409bf000) - libc.so.5 = /usr/local/matlab5.3/sys/os/lnx86/libc.so.5 (0x409c8000) libXt.so.6 = /usr/X11R6/lib/libXt.so.6 (0x40a84000) libX11.so.6 = /usr/X11R6/lib/libX11.so.6 (0x40ac7000) libXpm.so.4 = /usr/X11R6/lib/libXpm.so.4 (0x40b5f000) libSM.so.6 = /usr/X11R6/lib/libSM.so.6 (0x40b6d000) libICE.so.6 = /usr/X11R6/lib/libICE.so.6 (0x40b75000) - libc.so.6 = /lib/libc.so.6 (0x40b89000) ld-linux.so.2 = /lib/ld-linux.so.2 (0x40c2e000) it seems you haven't got the libc5-compatible X11-related libraries installed... have a look at the packages in the 'oldlibs' section, I think 'xlib6' and 'xpm' are the ones you need. BTW, is there a FAQ entry on this topic somewhere? Just about every Debian user can face this problem when trying to run libc5-based applications... Andras
Re: what exactly is a segmentation fault
rich wrote: Every item in the list should have a left and right side. Here is what that tells me - it seems like everything is okay? libvga.so.1 = /usr/lib/libvga.so.1 (0x4000c000) libvgagl.so.1 = /usr/lib/libvgagl.so.1 (0x40058000) libX11.so.6 = /usr/lib/libc5-compat/libX11.so.6 (0x40067000) libm.so.5 = /lib/libm.so.5 (0x40105000) libc.so.5 = /lib/libc.so.5 (0x4010e000) libc.so.6 = /lib/libc.so.6 (0x401cc000) libm.so.6 = /lib/libm.so.6 (0x40271000) ld-linux.so.2 = /lib/ld-linux.so.2 (0x4028a000) I think depending both on libc5 and libc6 shouldn't be a healthy situation. This one seems to be a libc5 program depending on svgalib, but maybe you haven't installed the libc5-compatible svgalib libraries (package 'svgalib1' in section 'oldlibs') so your program tries to use the libc6 ones, resulting in this conflict. Does this tell me anything? This also seems similar to an error that I get when I try to run emusic - nothing happens until I ^C, then I get the next message... Gdk-Message: ** ERROR **: sigint caught I haven't tried 'emusic' so I have no idea why nothing happens, but the SIGINT is due to the ^C itself (in general, pressing ^C sends a SIGINT to the process on the corresponding terminal). András
Re: Using powersave Features on monitor
Jeff Katcher wrote: Hi all, I would like to use the PowerSave mode on my monitor (in win9x when i am idle for X minutes the screen goes into powersave mode) is there a way to do this in Linux? On a Linux console, you can enable this via setterm(1). For X, xset(1x) can do the job. András
Re: not booting into Xwindows
Sean wrote: Remove the xdm script from your runlevel directory. If you're using runlevel 2 (most probably), then go to /etc/rc2.d, and delete the S**xdm package where ** will be some number. Sean Lev Lvovsky wrote: can anyone tell me how to keep my computer from booting into Xwindows? This is a rather annoying automation... thanks :) -lev Alternately, if your machine is connected to a network and you need xdm in order to enable remote X logins to your workstation, remove the entry from /etc/X11/xdm/Xservers that starts the X server on your machine. András
Re: multi-volume tar?
On Sun, 28 Mar 1999, Vincent Murphy wrote: i would like to create a (big) tar archive which spans multiple zip disks. is this possible? It should be just a matter of the -M switch of 'tar' and writing the archive directly to the disk (not to a file system). I often use this with floppy disks. Note: you can't make a multi-volume compressed archive. Andras
Re: Sound Board
On Mon, 22 Mar 1999, Ed Cogburn wrote: With 2.2, the config of sound drivers occurs outside the kernel config. You can no longer build the sound modules into the kernel, they must be built as modules. With sb16, for example I need a config line like: options sb io=0x220 irq=5 dma=1 dma16=5 mpu_io=0x330 in /etc/modules.conf. The above doesn't apply to 2.0 kernels, so your problem might not be related to configuration of the sound module. If you are using a 2.2 kernel then see /usr/src/linux/Documentation/Introduction. Isn't it just the opposite? Under 2.0, I've always configured and compiled the sound module separately, but when I used Slackware 3.0, the kernel (v2.0.27) shipped with it had the SB driver compiled in, although I never succeeded in configuring the sound driver in the kernel config. Under 2.2, there's no opportunity for configuring and/or compiling the sound module on its own, it's just like any other standard kernel driver. Too bad that I couldn't get it working with my AD1848-based Aztech card, nor could I compile the old driver for 2.2, so I'm still back on 2.0.37. Andras
Re: Help with PAP!
Helllo, I am now connecting to a different server from my ISP for X2 V.90 dialup. Previosly, to dial in all I had to do was change /etc/chatscripts/provider with my username and password and the number to phone (there was no PAP/CHAP authentication). Now allegedly there is some sort of PAP authentification required, and no script is run (so no expect: ogin stuff). Under NT 4.0, the settings I use are don't run a script, dynamic DNS server (but have their IP adresses) and accept any authentication including clear text. This works. I'm confused about what to do with Debian (ham) however - shoudl I not be running my chatscript? If I just change the dialup number,name, etc in /etc/chatscripts/provider when I dialin I get garbage after expect: ogin {}{#}}{*^### type stuff. How should I connect now there's PAP authentication and no script being run at the providers end? Confused, Martin Your 'connect' chat script should end at CONNECT. PAP means that the authentication is done during the PPP negotiation. All you have to do is to add a line like this to your /etc/ppp/pap-secrets file: your-name provider-name your-password and to /etc/ppp/peers/provider: user your-name remotename provider-name The latter line can be omitted if this is your only ISP connection - in that case, substitute 'provider-name' with an * (asterisk). Andras
Re: How to get Debian and (argh!) Windows 95 machines linked through null modem?
Pablo Longhi Lorenzzoni wrote: Hi Debianers!! I have a big problem and little time to solve. Can anyone help me? I have to get a Debian (hamm) machine and a (argh!) Windows 95 one linked through a null modem cable using PPP protocol. The Debian machine can be able to connect Internet through ppp0. I already tryed everything I know. Does anyone have any ideas? I solved this problem as follows: 1. I configured the corresponding serial port as a standard modem on the W machine. 2. I observed the modem control commands sent by the W machine, and answered them as expected (OK, CONNECT, ...) so the W thought it was talking to a working modem. 3. Now I run pppd from inittab (respawn) with the appropriate connect and disconnect scripts (using chat) to automate the previous step. The W machine starts up the PPP link after this fake modem reports a successful connection. After that, everything should work as a normal PPP link. Note: if you want to use both this and the Internet connection on the Debian machine, you should write the correct ip-up and ip-down scripts to update your routing table, and you should enable IP forwarding in the kernel in order to enable Internet connection for the W machine (I'm not sure it works, 'cause I use this whole stuff only for connecting the two machines). I've also got a question according to this. I'd like to get Samba working over this nullmodem PPP link to share the disk space and the printer on the Linux machine. I've tried a lot of fine-tunings concerning interface and routing table settings and Samba configuration, but the result is always the same: it sometimes works correctly (browsing, sharing and the like), but sometimes it doesn't work at all (the W machine doesn't even start up M$ Networking), with exactly the same setup on both machines. Now I'm not sure of even which machine is to be blamed for this. András
debian-cd for slink ?
Hi there, do you know about any tool for creating the slink cd-images from a mirror of the distribution (like the debian-cd package in hamm) ? Thanks, Andras