From: Digestifier To: Linux-Admin@senator-bedfellow.mit.edu Reply-To: Linux-Admin@senator-bedfellow.mit.edu Date: Wed, 21 Sep 94 03:13:39 EDT Subject: Linux-Admin Digest #86 Linux-Admin Digest #86, Volume #2 Wed, 21 Sep 94 03:13:39 EDT Contents: Term-rlogin setuid-root: security hole? (Alexandra Griffin) Re: SEARCH: Clock setting Programm for 24X (Greg Robertson) Re: GNU finger on Linux ? (Juha Virtanen) Re: AutoMount For Linux (Mitchum DSouza) Re: Disappearing Keyboard (Kai Petzke) Re: who/finger shoing hole net (benny@bigfoot) Tape problems, (Emerald 9000) (Wim ten Have) problems with connecting.. (Alan Donald) Dosemu in X (Dan Wold) Re: Clean shutdown from X (Christoph Best) Re: Printers on the parallel port (Corey Brenner) Re: What user interface to use??? (Corey Brenner) Re: Modem 14400 and uugetty (Kevin Cummings) Re: ftp login message (John M Hansen) Some Networking Problems... (Charles W. Binko) Re: Printers on the parallel port (George Photakis) Re: Routing A<-slip->B<-ether->C (Kevin Cummings) ---------------------------------------------------------------------------- From: acg@kzin.cen.ufl.edu (Alexandra Griffin) Subject: Term-rlogin setuid-root: security hole? Date: 20 Sep 1994 19:30:37 GMT I grabbed the pre-compiled RLOGIN binary off the bohr.physics... site, and found that it did not work properly in "shared" mode (i.e. it wouldn't find the TERM socket except when run from the account that started TERM, or by root). Making rlogin setuid-root fixes this, and things seem to work perfectly: when rlogin'ing to a remote system, the user name that gets passed through is that of whoever runs TERM, not root. So, it seems to work great, but am I opening myself up to any security holes by doing this? I've checked for the obvious ones already (it won't let you rlogin as root to localhost without a password, etc.), but is there something I'm overlooking? Thanks, -- alex ------------------------------ From: grober1@abacus.tis.tandy.com (Greg Robertson) Subject: Re: SEARCH: Clock setting Programm for 24X Date: 20 Sep 1994 16:39:39 GMT : NO. freq is for the Diamond proprietary programmable clock generator. : Diamond doesn't use it on their 24X cards (WD chipset). I thought that : I read once that the 24X should work just fine once the chipset was : supported by XFree. Disclaimer: I don't own/use a 24X, so everything : I'm saying is from memory. I DO own/run a Diamond Stealth VRAM, so I'm : not totally clueless here! : -- : Kevin J. Cummings Peritus Software Services, Inc. : cummings@kjc386.framingham.ma.us cummings@peritus.com WRONG.....freq is for the 24X cards....I own one and use freq to set my clocks and my card does have a WD chipset. Greg -- +------------------------------+-----------------------------------------+ | Greg Robertson | Tandy Information Services | | Production Control | Tandy Technology Square, Suite 1431 | | grober1@tis.tandy.com | 200 Taylor Street, Fort Worth, TX 76102 | | Voice: (817) 870-0879 +-----------------------------------------+ | Fax: (817) 390-2132 | It doesn't hurt to ask! | +------------------------------+-----------------------------------------+ ------------------------------ From: Juha.Virtanen@hut.fi (Juha Virtanen) Subject: Re: GNU finger on Linux ? Date: 20 Sep 1994 14:03:55 GMT Reply-To: jiivee@hut.fi >>>>> On 17 Sep 1994 11:10:30 GMT, phil@rivendell.apana.org.au (Phil Homewood) said: :> Compiles straight out of the box once you install process accounting, :> I am told. Then you have been told wrong information! GNU finger and process accounting performed by kernel have nothing to do together. Rather, there are several different patches for GNU finger-1.37 to make in compilable. I don't remember which one I used, but it works fine for me. Check diffs mentioned in previous article. Juha -- Plääh. En mä käytä .signaturea. ------------------------------ From: Mitchum.DSouza@mrc-apu.cam.ac.uk (Mitchum DSouza) Subject: Re: AutoMount For Linux Date: 20 Sep 1994 14:20:00 GMT In article , jcolman@lehman.com (Jake Colman) writes: |> SethMeister G. (consp05@bingsuns.cc.binghamton.edu) wrote: |> |> : GHi There, |> |> : Does anyone know if there is an AutoMount Utility, Like Volume |> : Manager on Solaris, available for Linux. What this does is |> : automatically mounts a diskette when it is inserted in the disk drive |> : and mounts it to a directory you specify. Anyone know about this? |> : -Thanks |> |> : -- |> : ---------------------------------------------- |> : O.G. SethMeister <<<<<<<<<<>>>>>>>>>> D.F.W.M. |> : ---------------------------------------------- |> |> There is a version of amd (AutoMounter) for Linus that works like a charm |> for autmounting remote directories. It does not automatically mount |> diskettes. I would be interested in such a beastie as well. If you take a look at the "program" option you will be able to see how floppies may be (u)mounted relatively eaisly. Robert Sanders kindly sent me some scripts to do just this with the automounter, and I promised him I would add this to any doccument I wrote on AMD. However lack of time has prevented me from doing so. If you would like a copy please email me, otherwise just wait until such time that I am able to doccument such things for pubilc consumption. Mitch ------------------------------ From: wpp@marie.physik.tu-berlin.de (Kai Petzke) Subject: Re: Disappearing Keyboard Date: 20 Sep 94 09:10:52 GMT gburk@netcom.com (Gene R Burk) writes: >Does anyone have an idea why after boot up my keyboard would suddenly act >as if it's not getting any power? Everything worked fine as I set things >up using the root and boot disks. Did you accidently press Ctrl-Q? That stops any further screen output, including, what you type. Press Ctrl-S for remedy. -- Kai Petzke | How fast can computers get? Technical University of Berlin | Berlin, Germany | Sol 9, of course, on Star Trek. wpp@marie.physik.tu-berlin.de | ------------------------------ From: benny@bigfoot Subject: Re: who/finger shoing hole net Date: 20 Sep 1994 14:21:44 GMT Hans Petter Fasteng wrote in article <1994Sep20.122634.5720@kfdata.no> : > >Is it possible to configure w or finger (bouth) to show all hosts on our >local ethernet? > >-Hans > > Try GNU finger. It's supposed to compile out of the box if you got accounting installed. If not, like me, you can apply a patch that can be found on sunsite. sunsite.unc.edu:/pub/Linux/distributions/debian/misc/finger-1.37.debian.diff.gz GNU finger is found at your favourite ftp site for GNUware. / Benny ------------------------------ From: wimth@gouldnl.encore.nl (Wim ten Have) Subject: Tape problems, (Emerald 9000) Date: Mon, 19 Sep 1994 08:06:47 GMT Is there anybody who knows about support under linux for an Emerald Systems 9000 Tape Streamer? The Drive and interface card look like an ordinary QIC-02 tape subsystem (card and tape streamer). Additional information: - According to the documentation the drive is an Emerald Series 150-9000 - The card address can be set by 4 dip, where the irq is fixed to 5 and the dma to 3. - The drive works perfectly under DOS with the supplied software named asd and asdiags. - The tape subsystem does always report "tpqic02: Reset failed!" If any of the available tpqic02 drives are selected. (Wangtek, Archive or Mountain). - The Drive head does reposition, together with a short blink of the drive LED, every time /sbin/qic02conf is given to install the drive parameters to the drive port, irq and dma. Thanks for your/any reply that can help us, -- Wimth. ------------------------------ From: cpamd@ingers105.halls.colostate.edu (Alan Donald) Subject: problems with connecting.. Date: 20 Sep 1994 11:44:14 GMT I have this small problem.. it seems that i can't connect to my own pc with any consistancy. I can connect out everytime, but when i try to connect via ethernet, either when i use telnet it says connection established, but no login appears. It also appears that any other service that is run via inetd suffers this problem, it connects but does not run the appropriate daemon. The twist is that when i type inetd from the prompt _everything_ from then on works fine until a random time later when it decides to not work anymore :(. I am running slackware 2.0 with a 1.1.50 kernel with a ne2100 compatible ethernet card, also the /etc/rc.d/* file are pretty much as they were when they were installed (except for ip's,gateways,etc) One last thing, would recompiling inetd and accompanying daemons fix this problem? I have heard that recompiling can sometimes fix problems with software made with later kernels. Any help/patches/fixes would come in most handy...thanks.. Alan cpamd@ingers105.halls.colostate.edu -- ======================================================== - Boy, your soul better belong to jesus, - - 'Cause your ass belongs to me.... Megadeth - ======================================================== ------------------------------ From: danw@panix.com (Dan Wold) Subject: Dosemu in X Date: 20 Sep 1994 19:57:49 -0400 I just setup pre53_20. It's working pretty well. When I do "dos -AX" the dosemulator starts up in a nice "Dos in a Box" window. I was able to run Telix (a dos telecom program) in the window. I can't seem to figure out how to get backspace & delete to work in this window. When I logged onto a remote system with Telix the backspace and arrow keys worked normally. Does anyone have a hint on how to fix this? The other problem is booting from the hdimage file. When I do "dos -CX" then the emulator starts up in the original xterm rather than opening "Dos in a Box". I could use a clue on this one too. Thanks for any help! -Dan danw@panix.com -- danw@panix.com Daniel Wold 239 City Island Ave, Bronx, NY, 10464 finger danw@danw.dialup.access.net Sysop: WORLD CITIZEN BBS 718-885-2346 14.4 24 hrs FREE Fidonet BahaiNet ------------------------------ From: cbest@rs502.th.physik.uni-frankfurt.de (Christoph Best) Subject: Re: Clean shutdown from X Date: 20 Sep 1994 11:56:37 GMT Tim Brailsford (Tim_Brailsford@vme.nott.ac.uk) wrote: : I am having a problem shutting down a Linux system from X (Linux 1.0.9, : XFree86 2.1.1). When X is permanently loaded with xdm then the usual : 'shutdown now -h' quits X and then hangs the system - causing a screen full : of : errors when the machine is restarted. This works OK when X is loaded using : startx, but I need to use xdm. : Any ideas would be gratefully recieved. We use the following script to shutdown Linux PCs: echo -e "\033[H\033[2J\n\n\rPLEASE WAIT 20 SECONDS\n\r" >/dev/tty1 echo -e "UNTIL IT SAYS 'The system is halted'\n\r" >/dev/tty1 /sbin/halt >/dev/tty1 2>&1 X dies, you return to console 1, and the text is on the screen. Then it says 'Sending SIGTERM...' and so on, and finally 'The system is halted.' I know I shouldn't call halt directly, but it works fine that way. Another nice trick is to put this is xdm/Xstartup: if [ "$USER" = "halt" ]; then /usr/local/bin/xhalt # this is above script exit 1 fi Add a user halt to /etc/passwd with no shell. Then anybody sitting at the X11 console can login as halt with a password and shut down. We use this to let users shut down their desktop machines without getting root access. Shouldn't shutdown make sure that it does not kill itself? -Christoph Best | Mathematicians are like Frenchmen: | Whatever you say to them they Institute for Theoretical Physics | translate into their own language, Johann Wolfgang Goethe University | and fortwith it is something entirely Frankfurt, Germany | different. - J. W. Goethe ------------------------------ Crossposted-To: comp.os.linux.help From: brennerc@saucer.cc.umr.edu (Corey Brenner) Subject: Re: Printers on the parallel port Date: Tue, 20 Sep 1994 23:19:15 GMT Kevin Cummings (cummings@hammer.westboro-ma.peritus.com) wrote: : In article , georgep@sabre.com (George Photakis) writes: : > LPT1 in DOS is /dev/lp0 under Linux.... : > George Photakis : Generally yes, but not necessarily. Your BIOS looks down a list of : three possible parallel devices, and assigns the first one it finds to LPT1:, : The second to LPT2:, and the third to LPT3:. If you have all three, then : the parallel port on your monochrome video card is LPT1, and the ports : with hardware address 378 is LPT2, and 278 becomes LPT3. If you don't have : the video card parallel port, then 378 and 278 are LPT1 and LPT2. If you : only have port 278, then it becomes LPT1. : Linux uses a 1-to-1 correspondance with the IO ports to the Linux devices. : lp0 is always the video port. lp1 is the parallel port at 378, and lp2 is : always the port at 278. Regardless of whether or not you have them them : in your system. : I have one parallel port at home: 378. It is LPT1: under DOS, and lp1 : under Linux. : A similar correspondance holds true for the serial ports at 3F8, 2F8, 3E8, 2E8 : and COM1-4 under DOS and /dev/ttya-d under Linux. : Your mileage will vary depending on your hardware. : -- : Kevin J. Cummings Peritus Software Services, Inc. : cummings@kjc386.framingham.ma.us cummings@peritus.com I believe he said he was using Slackware 1.0... what kernel? and also, if the parallel printer driver is not compiled in, he won't be able to use it anyway. I believe the solution here is to have a look at your kernel config, compile in parallel printer support if it's not already there, and hope for the best. BTW, are the plip ports laid out the same? ie. lp0 => plip0 lp1 => plip1 lp2 => plip2 Corey Brenner ------------------------------ Crossposted-To: comp.os.linux.development From: brennerc@saucer.cc.umr.edu (Corey Brenner) Subject: Re: What user interface to use??? Date: Tue, 20 Sep 1994 23:23:57 GMT Tony Schwartz (tony@teleport.com) wrote: : When do dial into your local ISP using a standard terminal connection, what : software is used to providet the menuing, ability to do internet functions : like telnet, ftp, gopher, etc??? : Recommendations please and locations on obtaining these.... : Thanks : Tony Schwartz I would go the dip route. It provides SLIP which is, in my experience, a bit more stable than ppp in the later kernels. This will work only if your Internet Service Provider allows slip connections. Menuing is handled by some funky DOS client ( :) ). Using your system as just another node on the 'net will make you very happy. Corey Brenner ------------------------------ From: cummings@hammer.westboro-ma.peritus.com (Kevin Cummings) Subject: Re: Modem 14400 and uugetty Date: Tue, 20 Sep 1994 19:24:10 GMT In article <35du97$53n@clarknet.clark.net>, csamsi@clark.net (Caesar M Samsi) writes: > Michael Zill (mzill@saturn.RoBIN.de) wrote: > : First of all you should configure your modem that it returns > : the DTE speed not the DCE speed in the CONNECT string. > > : Than you should enable hardware flow control for the modem and > : the serial line. > > : Than you can lock the modem at 38400. So the modem is using > : the handshake lines to prevent overruns. > > What is the AT command string to do this ? What is your modem type? It varies. RTFM. -- Kevin J. Cummings Peritus Software Services, Inc. cummings@kjc386.framingham.ma.us cummings@peritus.com ------------------------------ From: m8412hj1@mimi.mbar.dtu.dk (John M Hansen) Subject: Re: ftp login message Date: 20 Sep 1994 14:44:18 GMT Srini Seetharam (srini@runabout.igt.com) wrote: : I am trying to provide my users with a message when they try to : ftp. Especially when they log in and also when the CWD command : is issued. : Can someone tell me where these message files are stored ? : I looked in the spool directories but was unable to find any. : -- : srini ------------------------------ From: cwb@cis.ufl.edu (Charles W. Binko) Subject: Some Networking Problems... Date: 21 Sep 1994 00:57:34 GMT Howdy. I have recently set up Linux (for the first time) and have actually had pretty good luck. I have one major problem (I think it's only one) that has me stumped. First, here is the setup: Kernel 1.0.9 (from slackware CD-Rom, recompiled with everything I need) running on a 486 DX50 8MB mem 16MB swap partition Using Dip version 3.3.7-uri to connect using a USR Sportster 28.8 V.FC (waiting for the V34 upgrade :) X-Free86 Ver 2.1 in standard VGA mode (while monitor gets fixed) Ok, here is the problem...I CAN'T See myself! I am using a dynamic CSLIP interface to the UF netork, and that works fine (I can ping any host out there and get my stats, and xterms, netrek, and _almost_ all of the apps I try work fine). My problem is that if I do a 'ping localhost' or 'ping DittoHead' (my host name - go ahead and flame, I don't care) or 'ping slip-c12.cis.ufl.edu', I get this: DittoHead:/usr/lib/X11/app-defaults> ping slip-c11 PING slip-c11 (128.227.224.231): 56 data bytes ping: sendto: Network is unreachable ping: wrote slip-c11 64 chars, ret=-1 ping: sendto: Network is unreachable ping: wrote slip-c11 64 chars, ret=-1 --- slip-c11 ping statistics --- 2 packets transmitted, 0 packets received, 100% packet loss I THINK I set all of the information in the hosts, networks, etc correctly, but something is missing. The main reason I want this is that xmosaic will not run. I get an error saying that it can't get my local address. Any ideas? I thought about the netmask, but it is set in hosts as 255.255.255.0 ! Thanks for any help... Charles cwb@cis.ufl.edu ------------------------------ From: georgep@sabre.com (George Photakis) Crossposted-To: comp.os.linux.help Subject: Re: Printers on the parallel port Date: Tue, 20 Sep 1994 10:50:30 UNDEFINED In article <35mg9d$ob3@crl4.crl.com> sefarlow@crl.com (Stephen E. Farlow) writes: >George Photakis (georgep@sabre.com) wrote: >: In article <350th4$dtj@mozo.cc.purdue.edu> webblab@koala.ecn.purdue.edu (K >Webb/Lab Acct) writes: >: >I am trying to connect a laser printer to the parallel port. >: >The printer HOWTO file specifies to use /dev/lp1 for this port, but it >: >doesn't work. I even tried ls -l > /dev/lp1 and I got the following error. >: LPT1 in DOS is /dev/lp0 under Linux.... >: George Photakis >***************************************************************** >WRONG!!! A laser printer in Linux is /dev/lp1 !!! Becareful here! >***************************************************************** >-- That's funny, my HP LaserJet IIIP is LPT1 under DOS and is accessed by lp0 under Linux. My HP DeskJet 550C is LPT2 under DOS and is lp1 under Linux. I've been printing to lp0 for the past month now and the paper continues to come out on the LaserJet. George ------------------------------ From: cummings@hammer.westboro-ma.peritus.com (Kevin Cummings) Subject: Re: Routing A<-slip->B<-ether->C Date: Tue, 20 Sep 1994 20:10:19 GMT In article , jshiffle@netcom.com (John Shifflett) writes: > We have 3 linux computers here: A & B are connected via SLIP, B & C are > connected via ethernet. Ping, telnet, etc work fine between A & B, and > between B & C, but not even ping works between A & C. I presume this is > because the routing is setup wrong. I have made an effort to read about > the subject, and to do some trial and error fooling around, but have had > no success. Now's the time to ask all you folks for help! First, a few > questions: > > 1) I've assigned 2 IP numbers to B - one for SLIP and one for the > ethercard. Is this the correct thing to do? Are A, B, and C all on the same IP network? If so there is one answer. If not, there is another. > 2) Are the two halves - SLIP & ethernet - considered to be two separate > networks, subnets of one network, or one network? In other words, > do I need a subnetting setup? The simplest setup is to do routing and subnetting between a SL/IP network and your ethernet network. You will need two IP networks for this (they needn't have the same sized netmasks, nor need they distinct, but the case of overlapping networks gets a little more complicated to setup right). If you are trying to do this with one IP network, then there are two solutions: a specific host route, or proxy arp. You can add a specific host route to your ethernet network that basically says route all packets to A to machine B. Then machine B can know that A is at the other end of its SL/IP line when SL/IP is running, and transfer the packets appropriately. Proxy arp will esentially do the same thing, except that machine B will "proxy" itself as machine A by using its own ethernet address for packets destined machine A. When B gets those packets, it then routes them over the SL/IP line to machine A. The cost for the two different methods: Static host routing, some machine must publish the static route to B (it could be machine B itself if it is allowed to publish routes, otherwise it is the router for the ethernet network that B & C are on). Proxy Arp, needs some machine to publish the artificial arp address for machine A so that it points to machine B. Again, it could be machine B, or some other machine on the network. > 3) Do all 3 machines need a different routing table (not counting the > different IP numbers, of course)? Or would A's and C's be more or less > identical? > 3) I'm running 1.1.50, and do NOT have 'IP forwarding/gatewaying' enabled. > Is this required in my case? If yes, does only B need it? IP forwarding is needed (I think) if you are using different IP networks. > 4) Is B considered to be a 'gateway' to C from A (& vice versa), or is > A through B to C considered to be a 'direct' route? B is a gateway only if different IP networks are being used. When proxy arp is being used, then A is considered part of a direct route (only B knows that it isn't, and B handles that case right for you). Disclaimer: I am not a network expert, but I have (at different times) configured parts of my network at work using different IP networks, and using proxy arp. However, I have never actually made a linux box be the gateway machine (at work, I do have that setup working at home, however, I no longer have a real IP network at home, so I am now using 192.0.2.0, and we all know that you can't (shouldn't) route packets to/from that network elsewhere). I have used a DOS 286 PC running KA9Q sucessfully, and an annex terminal server as well. (The reason I haven't used Linux has to do with a lack of the right kind of PC at work, not with my reluctance to do it.) -- Kevin J. Cummings Peritus Software Services, Inc. cummings@kjc386.framingham.ma.us cummings@peritus.com ------------------------------ ** 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 ******************************