From: Digestifier To: Linux-Admin@senator-bedfellow.mit.edu Reply-To: Linux-Admin@senator-bedfellow.mit.edu Date: Tue, 4 Oct 94 00:13:30 EDT Subject: Linux-Admin Digest #140 Linux-Admin Digest #140, Volume #2 Tue, 4 Oct 94 00:13:30 EDT Contents: Buslogic and > 1.1.46 kernels Be CAREFUL (John Spade) how to enable sulog on linux ? (Jacek A. Ponarski) Re: Finger - Is slackware version secure? (Alan Cox) Re: Good dirs to move to sep file systems? (Daniel Quinlan) Re: FTPs PCTCP and Linux lock. (Matti Aarnio) Development Idea: Kernel TOC (zeek) Re: Telnet & ftp freeze! (Seppo Kallio) Re: dip-3.3.7h-uri & ZyXel U1496E (jason a duerstock) Problem with routing table (Hien Pham) Re: DEC's screend port? (Alan Cox) Re: Keyboard mapping under X (Michael James Porter) Ftape works...Not yet ( Jonathan Williams) RPC Problems (xmeter too!) (Charles W. Binko) Re: PPP vs SLIP? (Peter Mutsaers) Re: Can Linux Mount a Mac Floppy (jason a duerstock) [Q] Commercial Software on Linux (Jan Mario Stankovsky) Re: Trn and the remote NNTPserver ("S. O'Connor") Re: Setting up the BaseAddr for Mitumi (Kees de Bruin) Re: Ncurses signals broken? (Mikael Nordqvist) Re: SLIP Configuration (Benjamin John Walter) ---------------------------------------------------------------------------- From: spade@rocko.lab.csuchico.edu (John Spade) Subject: Buslogic and > 1.1.46 kernels Be CAREFUL Date: 2 Oct 1994 00:03:29 GMT Post 1.1.46 kernels seem to override the > 1GB hard drive switch. If I run fdisk with 1.1.38 and install linux, it won't run after a few boots of 1.1.51. And vice versa. After messing up my linux boxes' hard drives several times, I traced it down to the changes around the 1.1.46 kernel. The partition tables seem to be incompatible. I'm out of my league as far as the specifics, but I can say that it changes the cyl/hd/sec specifications. So don't casually upgrade... I had to reformat 2 systems... The were good changes though, so I'm not complaining, just warning people. -- spade@rocko.lab.csuchico.edu "Quality, Service, Price... spade@ecst.csuchico.edu Pick any two..." ------------------------------ From: jacek@eng1.uconn.edu (Jacek A. Ponarski) Subject: how to enable sulog on linux ? Date: Mon, 3 Oct 1994 03:44:05 hi, Is there any way to enable sulog on linux? if i do it from login.defs it does not work. Thanks, --jacek _______________________________________________ _______________________________________________ Jacek A. Ponarski The University of Connecticut tel. (203)427-2854 email: jacek@eng1.uconn.edu (203)877-8655 jacek@micro5.trincoll.edu fax (203)427-2854 URL: http://micro5.trincoll.edu/ GOPHER: mareknt.pc.trincoll.edu FTP: mponarsk.pc.trincoll.edu _______________________________________________ To be or not to be, what was the question again? _______________________________________________ _______________________________________________ ------------------------------ From: iialan@iifeak.swan.ac.uk (Alan Cox) Subject: Re: Finger - Is slackware version secure? Date: Mon, 3 Oct 1994 15:39:21 GMT In article <36al1l$j0j@pluto.njcc.com> bjkramer@pluto.njcc.com (Brian Kramer) writes: >I think I may have had a breaking to my system. Is the finger daemon >from the slackware distribution secure? I think this is not the >current version, but the prior version of slackware. How can I tell? Finger is secure. The very old rlogind/telnetd had a bug as did lpd but thats currently fixed. Oh and someone has just fixed portmap if you are doing loopback nfs mounts. Alan -- ..-----------,,----------------------------,,----------------------------,, // Alan Cox // iialan@www.linux.org.uk // GW4PTS@GB7SWN.#45.GBR.EU // ``----------'`----------------------------'`----------------------------'' ------------------------------ From: quinlan@freya.yggdrasil.com (Daniel Quinlan) Subject: Re: Good dirs to move to sep file systems? Date: 03 Oct 1994 06:43:25 GMT Reply-To: quinlan@yggdrasil.com Timothy Demarest writes: > I just added 2 more drives to our linux system, and am looking to > split the system among several file systems. Right now, only swap > is on its own partition. > > What do you recommend for splitting across several file systems? I > dont want to move anything that I may need on the root partition in > an emergency! If you are using a FSSTND (filesystem standard) compliant (or reasonably close) distribution, then it is relatively easy to split up the system. From my own experience, I might suggest using one or more of the following directories as mount points for partitions: / /home - or the local equivalent(s) /usr /usr/X11 - or the local equivalent(s) /usr/local - or the local equivalent(s) /usr/src /var - or a /var subdirectory if any tend to be large swap I've also found it nice to have a second root partition, but that's because I'm slightly paranoid and I often test Linux distribution installation procedures. (Thus, I don't trash my system every time.) I prefer to avoid the use of extended partitions for aesthetic reasons, as many people do. I'd also advise making certain that the system administrator is able to work comfortably with a (compact) editor available in the root partition. `pico' and `vi' are decent choices. The first, because it is very easy to use, and the second, because it is documented in every Unix book written in the last decade, or so it seems. (This seems to be an area that hampers many people in an emergency.) Also, if root's home directory is situated on a separate /home partition, make certain that your `login' program will still allow the root account to login without a home directory. (This shouldn't be a problem on most Linux systems.) Also be wary of where login shells lie, if root's login shell is in /usr/local/bin, then it might be awkward when /usr/local crashes. (I don't think `login' protects against this, as it does for missing home directories, but I haven't exactly checked either.) Don't try to make the root partition too small or you may regret it later, especially if /tmp is situated on it and not linked to /var or somewhere else. The same advice goes for any volatile partition. Make certain that your root partition, including /boot and the kernel image, is situated completely below cylinder 1024. There is a similar provision that your last partition must begin below 1024, but that may only apply to IDE drives -- I can't recall for certain. My last point is that there some reasoning you should do before you repartition. If you are only a single user system, do whatever works for you. If you are running a large, multiple user system, work to keep the machine running in a capable state with as little downtime as possible. // Dan -- Daniel Quinlan ------------------------------ From: mea@utu.fi (Matti Aarnio) Subject: Re: FTPs PCTCP and Linux lock. Date: Mon, 3 Oct 1994 17:25:52 GMT sheep@news.udel.edu (Brian C. Huffman) writes: > I have Slackware's 2.0 distribution of Linux, and I am having a problem >logging on using FTP Software's PCTCP. It tends to lock the telnet session >(not the linux) and I have to quit and start again. If I log into a unix >machine first and then telnet to the linux, it works fine. Anyone have >any suggestions? > > Brian Recompile the kernel with: "PC/TCP compability mode ON" Actual problem is closely related, but that (intermediate) kludge can help in the mean time. >-- > +---- Brian C. Huffman --------+------------------------------+ > | University of Delaware | | > | 206 New Castle Hall | sheep@bitbytes.clark.net | > | Newark, DE 19717 | sheep@chopin.udel.edu | /Matti Aarnio ------------------------------ From: zeek@sunshine.io.com (zeek) Subject: Development Idea: Kernel TOC Date: 3 Oct 1994 17:04:22 GMT This idea surfaced from my numerous kernel compiles... Please forgive if such a thing does exist, but wouldn't it be handy to include with the kernel source a table of contents (TOC) explaining the drivers, file systems, etc. For example: Within the General Setup of you kernel configuration, you'll find the following: CONFIG_MATH_EMULATION Selecting this option will include a math-emulator in your kernel. If you have a math coprocessor installed on your computer, you probably don't need this. Blah, blah... The source for this option is in the directory /usr/src/linux/drivers/FPU-emu CONFIG_BLK_DEV_HD This option is for something or other. Which is why I'd like to see a kernel TOC. I'll be willing to help put one together. I'll need a fair amount of assistance from someone knowledgable with the kernel contents. Please forgive my ignorance if such a thing already exists. If so, where? -zeek ===== ThesearelessonswhichIlearnedinpartwalkingwithmydogsandwonderinghowtheworld lookswithoutafoveaandveryfewretinalcellsforcolourvisionbutwithahugeneural processingandsensoryareaforsmellsTheeyesmadeavailableinmoderntechnological sciencesshatteranyideaofpassivevision -donnaharaway ------------------------------ From: kallio@network.cc.jyu.fi (Seppo Kallio) Crossposted-To: comp.os.linux.help,comp.os.linux.misc,comp.os.linux.development Subject: Re: Telnet & ftp freeze! Date: 3 Oct 1994 12:58:40 +0200 Seppo Kallio (kallio@network.cc.jyu.fi) wrote: : Usually it works nice, but sometimes the whole net software seems to : do nothing with the new connections. Login opens the window and : connects, but no "login:" prompt. Same problems with smtp : connections, they jammmm similar way. I think all telnet connections : hang somehow or are extremely slow. I have one more problem: looping in.telnetd. Just now I had 3 of them. Someone did mention that Linux seens to slow down when these net errors happen. Have you cheked if it is extra looping in.telnetd ??? Every extra in.telnetd in cpu loop is rising the load by one. Seppo Kallio ------------------------------ From: jduers01@solix.fiu.edu (jason a duerstock) Subject: Re: dip-3.3.7h-uri & ZyXel U1496E Date: 2 Oct 1994 15:44:56 GMT Dirk Hillbrecht (hillbrec@informatik.uni-hannover.de) wrote: : I cannot make the two ones mentioned in the header work together. ZyXEL has : ROM 6.12, kernel is 1.1.51, dip is dated 5.9.94. Script's start is as follows: : # Set the desired serial port and speed. : port ttyS1 Shouldn't this read "port cua1"? Jason ------------------------------ From: hienpham@csulb.edu (Hien Pham) Subject: Problem with routing table Date: 3 Oct 1994 08:42:55 GMT Hi everyone, I have a problem with networking my Linux box. I have Yddrasil Summer 94 Linux 1.1.0 #84 with NET-2 running. I have set up my static route table with local ip address of 128.178.10.1 and 128.178.10.2 for my Linux boxes with netmask 255.255.255.0 and broadcast addr 128.178.10.255. The system is on the same datalink. My local network is 128.178.10.0. My first problem is "arp -a" returns nothing and when I execute "netstat -rn" I got the error "route: unsupported address family 2560 !". Any ideas ? My ifconfig shows that my NIC card is running with arp up as configured. Moreover, my xserver can't seem to run remote x-clients. I keep getting the message "can connect to calvin:0.0". Thanks in advance for any advice. -- % Henry Pham %%% Jet Propulsion Laboratory % % Senior Software%%% 4800 Oak Grove Drive % % Engineer %%% Pasadena, CA 91109 % % %%% E-Mail: hienpham@beach.csulb.edu % ------------------------------ From: iialan@iifeak.swan.ac.uk (Alan Cox) Subject: Re: DEC's screend port? Date: Mon, 3 Oct 1994 10:24:53 GMT In article andrewp@itwhy.bhp.com.au (Andrew PRUSEK) writes: >Has anyone had any luck porting the screening router >screend available from gatekeeper.dec.com? It has two problems 1. Its somewhat slow as it switches all the frames in and out of user mode. 2. The license is extremely restrictive and requires you give dec any changes you make etc. That prevents anyone putting it into the Linux kernel code. There is another as yet more limited but much faster IP firewall codeset in ALPHA test (ipfirewall-x.y.tar.gz on sunsite). This looks promising and I sent the authors a long list of suggested changes as in the end I'd like to put it into the kernel code as yet another config option. Alan -- ..-----------,,----------------------------,,----------------------------,, // Alan Cox // iialan@www.linux.org.uk // GW4PTS@GB7SWN.#45.GBR.EU // ``----------'`----------------------------'`----------------------------'' ------------------------------ From: mike@strauss.udel.edu (Michael James Porter) Subject: Re: Keyboard mapping under X Date: 3 Oct 1994 13:00:56 -0400 In article <36h95e$5uk@bmerha64.bnr.ca>, Adrian Mancini wrote: =>Recently I attempted to remap my function keys to contain =>character strings follow by ^M so that I could execute =>certain shell commands in an xterm by just hitting a function =>key. You need to make an 'Xt' translation table. See the man page on xterm for more info. One thing: The resource that you are going to make is a single, multi-line resource. When I tried to make a rather large table, I ran into a problem with too many lines being concatentated by cpp. cpp Seemed to drop off lines past 20 or so (xrdb -merge adds the resources, and it runs cpp, the C pre-processor). So, when working with large tables, use xrdb -query to make sure what is being stored on the X Server is what is in the file. I got around the problem by making very long lines. On the format for a translation table: I bought a summary reference on X that is very valuable for this type of info. I would write more, but all my references (and system) are at home. The title might be 'X in a Nutshell'? Mike ------------------------------ Subject: Ftape works...Not yet From: williamj@cs.uni.edu ( Jonathan Williams ) Date: 3 Oct 94 09:45:16 -0500 Well, I thought I had ftape working, but I guess I was wrong. I'm running Linux kernel version 1.0.9 and ftape version 1.13b patched for the conner bug and compiled with the -DCONNER_BUG flag. When I finally got done compiling, I used tar to backup my drive. It went through several hundred files, and then suddenly stopped with an I/O error writing to the device. I'm assuming that its a bad sector on the tape, since I was able to verify everything up to that point, but how do I confirm this and mark the sector as bad so I can continue? Jon Williams University of Northern Iowa ------------------------------ From: cwb@mail.cis.ufl.edu (Charles W. Binko) Crossposted-To: comp.os.linux.help Subject: RPC Problems (xmeter too!) Date: 3 Oct 1994 09:01:09 GMT Hello I have compiled xmeter on my system, and it works for the most part (had to get the librpcsvc.a file, but other than that, compile was easy). My problem is that I cannot get xmeter to watch any network stats on my local machine. I am runing SLIP (dip) , but I should still be able to watch the packets in and out (right??!!??) . I mean, they are still packets, and inetd doesn't care about what kind of interface it is talking to. Xmeter watches the network traffic on all of the other machines I look at but not mine :(. What is wrong? I have rstatd running, do I need other rpc daemons as well? The xmeter man page says it gets its info from rstat...doesn't mention any others. Any Help would be appreciated.... Charles ------------------------------ From: plm@atcmp.nl (Peter Mutsaers) Subject: Re: PPP vs SLIP? Date: Mon, 3 Oct 1994 11:31:18 GMT >> On 28 Sep 1994 04:27:20 GMT, pp000547@interramp.com said: p> Also I am finding that my current PPP-vendor seems to like p> to kill my connection if I don't send anything out over the line for p> more than a few minutes. Maybe he hangs up and expects the other end to be 'dial-on-demand'. Alas the Linux implementation doesn't offer that (yet). Many PPP implementations hang up after a few minutes silence, but keep the logical network up; then when you want to send a packet again they dial and make connection again transparently. Especially when ISDN is getting through with its very fast dialing and one second resolution billing dial-on-demand becomes a must for PPP. -- Peter Mutsaers | AT Computing bv, P.O. Box 1428, plm@atcmp.nl | 6501 BK Nijmegen, The Netherlands tel. work: +31 (0)80 527248 | tel. home: +31 (0)3405 71093 | "... En..., doet ie het al?" ------------------------------ From: jduers01@solix.fiu.edu (jason a duerstock) Subject: Re: Can Linux Mount a Mac Floppy Date: 2 Oct 1994 22:46:12 GMT Tim Bass (Network Systems Engineer) (bass@cais2.cais.com) wrote: : Got some ASCII text files on my Powerbook... Would like to write them : to the PB floppy and then mount the floppy on my linux box and : read the ascii text (and do some other stuff). I haven't seen this : in any FAQ or the The Linux Bible. Any clues for the clueless ;-) Write it to a DOS floppy with Apple File Exchange + then mount it as an MSDOS disk on the Linux system. If there are plans to support the Mac HFS, I don't know about them. Jason ------------------------------ From: jan@ifs.univie.ac.at (Jan Mario Stankovsky) Subject: [Q] Commercial Software on Linux Date: 3 Oct 1994 12:14:55 GMT Is there a list of commercial software(i.e. FrameMaker, Smalltalk, a.s.o.) available for Linux? Does anybody knows any details if commercial SW will be available for Linux someday ? thanks ------------------------------ From: "S. O'Connor" Subject: Re: Trn and the remote NNTPserver Date: Mon, 3 Oct 1994 07:14:49 GMT On Sun, 2 Oct 1994, Jeremy J Sugerman wrote: > It's a trick. trn can use a remote nntp server, but you need to tell your environment. Bash enjoys tormenting people by not actually announcing environmnet > variables until you tell it to. Try either: > > NNTPSERVER="news.myhost.edu" > export NNTPSERVER > > or > > export NNTPSERVER="news.myhost.edu" > > ingerneal when you change variable besides the prompt you should export them or > else you should scrutinize the man pages and memorize (yeah right) which bash > exports auto and which you need to tell it. > > BTW, put either of the above in /etc/profile or /etc/cshrc and it should work. > > Good Luck, > Jeremy Sugerman Thank you. The third one worked. Irish ------------------------------ From: bruin@tasking.nl (Kees de Bruin) Subject: Re: Setting up the BaseAddr for Mitumi Reply-To: bruin@tasking.nl (Kees de Bruin) Date: Mon, 3 Oct 1994 08:25:14 GMT Jason Hong writes: > I am using Mitsumi CD with base-addr = 0x340 and irq=11. > I changed "mcd.h" file from 0x300, 10 to 0x340 and 11. Connected to a SoundBlaster MCD card perhaps?? > However, whenever I start Linux, it still try to detect the CD > from old address. I read through CD-ROM HOWTO but I couldn't > find any mention on base-address. Did you do a re-compile of the kernel, and did it actually re-compile the mcd.c file. If not, touch the mcd.c file and re-compile the kernel again. With the next reboot, you should be able to use you CD-ROM player, but be sure to mount any CD's in read-only mode. > Also, I am getting an fsck error message after installation. > Is there anything I have to do after installation to use Linux? What message did you receive from the system. -- Kees de Bruin bruin@tasking.nl Tasking Software B.V. Tel. +31-33-55 85 84 Windows NT: From the makers of Windows 3.0! ------------------------------ From: d91mn@efd.lth.se (Mikael Nordqvist) Subject: Re: Ncurses signals broken? Date: 3 Oct 1994 12:45:10 GMT In article <36nmvl$8li@nkosi.well.com> gonzo@magnet.mednet.net (Patrick J. Volkerding) writes: > >In article , >Zeyd M. Ben-Halim wrote: >> >>It looks like somebody forgot to recompile their code with ncurses 1.8.5. > >That was my diagnosis, too. > >>The terminfo database format changed (to match that in SVR4) in 1.8.3. >>I guess slackware is still using the old one. > >No, Slackware is using 1.8.5 with the new terminfo database, and has been >since version 2.0.0. > >> >>Compile all the programs concerned with ncurses 1.8.5. > >Yes, that should fix the problem. Generally, it does fix the problem. But not for aumix :) Either it relies on bugs in the (old) version of ncurses that the author of aumix used, or new bugs ones were introduced in newer versions of ncurses (I believe the latter). The screen is quite messed up with 1.8.5. I have had much trouble with colors in ncurses 1.8.5. The only way I can get reasonably sure not to trigger a bug is to call refresh() before changing color/attributes (this shouldn't be necessary - one refresh after all changes ought to be enough). Scrolling regions with colorized characters is something I never got to work, there was always some characters that changed colors during the scrolling. I sent a bugreport (with a short program to show the problem) to Zeyd about a month ago, but got no reply (the reason could be a faulty mail-daemon or an overworked Zeyd :-) Slightly off subject: Does the linux-console have support for turning the cursor on and off? I am updating the screen often (several times/second) and it's a bit annoying having the cursor flashing all over the screen. /Mikael -- Mikael Nordqvist, student | d91mn@efd.lth.se | I'm not paraniod, it's just Lund Institute of Technology | mech@df.lth.se | that everyone is out to get me ------------------------------ From: ben@tsunami.demon.co.uk (Benjamin John Walter) Subject: Re: SLIP Configuration Date: Mon, 3 Oct 1994 15:00:32 +0000 : Is there any possibility to declare a timeout (hanging up the phone) and : redial and reinstall if anyone needs a connection (ping,telnet,ftp,etc.) over : the SLIP ?? I know of a program called slip_idle that will disconnect your SLIP connection if its been idle for a while. I don't know about demand dialling with Linux though, sorry. I'm happy to mail you the slip_idle program (its small!) peace, Ben -- __ _ / / (_)__ __ ____ __ / /__/ / _ \/ // /\ \/ / . . . t h e c h o i c e o f a /____/_/_//_/\_,_/ /_/\_\ G N U g e n e r a t i o n . . . ------------------------------ ** FOR YOUR REFERENCE ** The service address, to which questions about the list itself and requests to be added to or deleted from it should be directed, is: Internet: Linux-Admin-Request@NEWS-DIGESTS.MIT.EDU You can send mail to the entire list (and comp.os.linux.admin) via: Internet: Linux-Admin@NEWS-DIGESTS.MIT.EDU Linux may be obtained via one of these FTP sites: nic.funet.fi pub/OS/Linux tsx-11.mit.edu pub/linux sunsite.unc.edu pub/Linux End of Linux-Admin Digest ******************************