From: Digestifier To: Linux-Admin@senator-bedfellow.mit.edu Reply-To: Linux-Admin@senator-bedfellow.mit.edu Date: Sat, 30 Oct 93 05:29:34 EDT Subject: Linux-Admin Digest #130 Linux-Admin Digest #130, Volume #1 Sat, 30 Oct 93 05:29:34 EDT Contents: PC-NFS Based Printing (Gordon Russell) Re: finger not reporting last login time! (Mike Horwath) Re: finger not reporting last login time! (Jon Thackray) a (unique) kind of boot/root floppy? (Andrew R. Tefft) Re: PLIP Setup Problems (Alan Cox) Re: Sysinstall don't work (Warner Losh) Re: finger not reporting last login time! (Alexander Kourakos) arg! smail: unknown user (John Paul Morrison) Re: UNIX sysadmin FAQ- proposal and volunt (Bennett Todd) comm programs for linux (Mark R. Lindsey) Re: arg! smail: unknown user (Jeremy Bettis) Introduction into Unix (Holger Muenx) Textmode charset (Holger Muenx) [QUESTION}: simple news reader? (Dan Miner) Re: arg! smail: unknown user (John Paul Morrison) Re: finger not reporting last login time! --here's one fix. (Uppie) ---------------------------------------------------------------------------- From: gor@cs.strath.ac.uk (Gordon Russell) Subject: PC-NFS Based Printing Date: 29 Oct 1993 13:06:03 -0000 I was wondering if someone could tell me how to set-up remote printing to use a PC-NFS Printer Demon interface, rather than standard Unix-based demons. My problem is that the network system manager will not give me an entry into the hosts.lpd file, and instead has informed me that, as far as he is concerned, PCs operate using PC-NFS printer interfacing, and a pc running unix is far to big a security loophole to allow us to interface to his systems. Anyway, I can use the network printers from DOS via PC-NFS, but using the rm entry in printcap results in an error message stating that my machine does not have permission to print. I had a hunt through the FAQs for this one, but I must be blind because I could not find anything in the printer or general HOWTOs. Thanks in advance, Gordon +--------------------------------+-------------------------------------------+ | Gordon Russell | EMAIL : gor@cs.strath.ac.uk | | L13.16, Livingstone Tower, | TELEPHONE : 041-552-4400 Ex 3635 | | University Of Strathclyde, | FAX : 041-552-0775 | | 26 Richmond Street, +-------------------------------------------+ | Glasgow, G1 1XH | Spelling mistakes within this document are| | Scotland, UK | caused by internet compaction algorithms. | +--------------------------------+-------------------------------------------+ ------------------------------ From: root@jacobs.mn.org (Mike Horwath) Subject: Re: finger not reporting last login time! Date: 29 Oct 1993 12:39:33 GMT Make sure you have a file called /usr/adm/lastlogin (or /var/adm/lastlogin) created. This is where that information is stored most of the time. (On my system and my friends systems, this is the way it is). Hope this helps. -- Mike Horwath IRC: Drechsau BBS: Drechsau LIFE: lover root@jacobs.mn.org drechsau@jacobs.mn.org Jacob's Ladder 612-588-0201 UUCP, UseNet, Linux files, BBS ------------------------------ From: jrmt@ukc.ac.uk (Jon Thackray) Subject: Re: finger not reporting last login time! Date: Fri, 29 Oct 93 15:03:11 GMT Reply-To: jrmt@ukc.ac.uk (Jon Thackray) root@jacobs.mn.org (Mike Horwath) writes in <2ar2u5$3d7@jacobs.jacobs.mn.org>: >Make sure you have a file called /usr/adm/lastlogin (or /var/adm/lastlogin) >created. This is where that information is stored most of the time. (On >my system and my friends systems, this is the way it is). No. It's /var/adm/lastlog. Even touching this file may not help. Seeing as this question comes up _so_ often, (I'd guess about 20 times in the last 6 months), here are some clues. (Note: obviously it depends what version of finger, login, last, you're runnning. Check out the archive sites) i) if you have an old version of login, it won't even be trying to put records in the 'lastlog' file. You'll need to get a newer version and re-compile. ii) finger may need re-compiling to look in the right place for this file. Look in the relevant pathnames.h config file, and re-compile. iii) /var/adm/lastlog, along with /etc/utmp and /var/adm/wtmp all need to be mode 644. Obvious, but easy to overlook. You may also have problems with the utmp/wtmp files; some programs will think they are SysV format, others in the old BSD format. This may mean you'll get pty's left 'logged in', and other problems. Your version of 'init' may be able to fix this. You could also check out Jeff Uphoff's modified finger. I've also got patches it you prefer to show the location of the user. Hope this helps some people out there; it _can_ all be fixed. Jon. ------------------------------ From: teffta@cs690-3.erie.ge.com (Andrew R. Tefft) Subject: a (unique) kind of boot/root floppy? Reply-To: teffta@cs690-3.erie.ge.com Date: Fri, 29 Oct 1993 15:21:33 GMT I had an idea which I'm currently working on and just wondering if anyone else has done anything similar (I know a lot of people have made bootable root floppies; I'm not looking for ideas on what to put on such a filesystem :-)). First the background: I regularly boot Linux using a floppy which has LILO installed. The kernel itself lives on the hard drive. I much prefer this method to installing LILO on the hard drive -- makes me feel much safer. So here's my idea: 1. Make an ext2 filesystem on a floppy, copy a kernel and other basic utilities to it similar to any other emergency recovery floppy. 2. Install lilo on the floppy, with the default kernel being the one on my hard disk, and a second kernel choice being the one on the floppy, with the root filesystem being set to the floppy. Then I have one standard boot floppy which I can also choose to use as a recovery bootable rootdisk, by holding down shift when booting. Sound good? A waste of time? The only problem I've encountered so far is that LILO loading a kernel off the floppy is WAY slower than booting a kernel dd'd directly to the floppy. But for emergency situations I guess it's ok. And it seems to be a lot easier to maintain than a filesystem sharing space with a raw kernel file (I think I saw a method to make boot disks like this, no?) Well anyway when I come up with something good I'll post the relevant details here. ------------------------------ From: iiitac@swan.pyr (Alan Cox) Subject: Re: PLIP Setup Problems Date: Fri, 29 Oct 1993 15:59:57 GMT In article tim@lorien.demon.co.uk (Tim Towers) writes: >Thomas Uhl posted a query regarding setting up PLIP > >Firstly - PLIP does not work with Net-2d so probably forget it It needs only some tiny changes. Or you can run the pl13 alpha (bear in mind the word alpha's - some of the alphas have some fun bugs). > >If you want the Net-2e code and (supposedly) working PLIP wait for >linux 0.99.14 or join the net mailing list. 0.99.14 will almost certainly not be NET-2E but will have a rougly working PLIP. If anyone wants the tweaks to sort of get plip running in 0.99.13 mail me and I'll forward them. In the meantime if you want to know about NET2E and stuff join the NET channel. Alan iiitac@pyr.swan.ac.uk ------------------------------ From: imp@boulder.parcplace.com (Warner Losh) Subject: Re: Sysinstall don't work Date: Fri, 29 Oct 1993 15:48:18 GMT In article <2afpsg$kic@mtu.edu> sjkangas@major.cs.mtu.edu (STEVEN J. KANGAS) writes: >: parcplace's OI development series. Has anyone successfully installed >: this? > > Yup. Works fine. An extremely impressive package, although >it runs a bit slow since it's pretty big. I have 486DX50 and the lack of >speed is pretty noticeable on mine. With anything less, I'd expect >the speed to be annoying, to say the least. How much memory do you have? During the beta program, my brother was running it on his 386DX-25 and he reported that it was pretty snappy once it had started up. He has 16M. If you have only 8M of memory and are running lots of other things on your system (multiple clocks, xterms, emcas, etc) then you may find that you are swapping. Anytime you swap you take a BIG performance hit. I've had reports from people with 8M who have achieved acceptable performance from uib. This is why we recommend that people have at least 16M of memory. I did most of the initial port of the builder and library on a 486DX33 with 16M and I never the builder was slow. g++ was slow, almost painfully, at times, but the builder was fine. I've since added 16M of memory to make linking faster while I'm running under X. Also, if you are having problems with the compile time, use -O rather than the default -g. There was an article posted a while back that gave instructions on how to get the most performance out of OI/uib. Drop me a line if you'd like a copy. If I get more than 5-10 requests, I'll upload it to tsz-11 in the contrib area. BTW, I've found a couple tricks to reduce the amount of memory required for OI and thus uib by 0.5M. This should help when the next release happens. Warner -- Warner Losh imp@boulder.parcplace.COM ParcPlace Boulder I've almost finished my brute force solution to subtlety. ------------------------------ From: awk@char.vnet.net (Alexander Kourakos) Subject: Re: finger not reporting last login time! Date: 29 Oct 1993 20:38:36 -0400 In article <7132@gos.ukc.ac.uk>, Jon Thackray wrote: >Seeing as this question comes up _so_ often, (I'd guess about 20 times in >the last 6 months), here are some clues. (Note: obviously it depends what >version of finger, login, last, you're runnning. Check out the archive sites) There are as many solutions, it, seems, as there are instances of this question...unfortuntely, none of them seem to work for me! The 3-4 different finger sources I've gotten don't seem to have any place where this file is configured. They seem to get the information from some system calls. So which login/finger/whatever is that I should use to get this to work????????? Will this also fix the screwed up "last" entries too??? One interesting thing, though...fingering "root" always shows the correct last login time. awk ------------------------------ From: jmorriso@rflab.ee.ubc.ca (John Paul Morrison) Subject: arg! smail: unknown user Date: 29 Oct 1993 23:38:38 GMT I'm sick of this! smail consistently bounces incoming mail for only a few users on the system. Smail says the user is unknown but the user *is* in the passwd file. I suspect it is related to the passwd file: I got mail for one user to stop bouncing after shuffled a few things in the passwd file. But after I tried to get it working for another user, it stopped working again. I don't understand why mail to me (jmorriso@rflab.ee.ubc.ca) works fine, yet it always fails for other users. I have combed over /etc/passwd and /etc/shadow with a fine tooth comb. Nobody on the smail users mailing list responded to me, so I'm resorting to this group. Because it seems passwd related, it might be the shadow passwd or standard libs in linux. I'm using the most recent smail binaries uploaded to tsx. -- ___________________________________________________________________________ John Paul Morrison The Liberal Party of Canada: University of British Columbia, Canada proof that those who do not Electrical Engineering remember the past are condemned jmorriso@rflab.ee.ubc.ca VE7JPM to repeat it Day 4 of Canada held hostage: the socialist Liberal tax & spend tyranny ------------------------------ From: bet@std.sbi.com (Bennett Todd) Crossposted-To: comp.unix.admin,comp.sys.sun.admin,comp.sys.sgi.admin,comp.admin.policy Subject: Re: UNIX sysadmin FAQ- proposal and volunt Date: 30 Oct 93 01:29:15 GMT In article <1993Oct20.025007.11821@ntg.com>, Dave Platt wrote: >[ configure sendmail to forward all traffic for local users to mailserver, > and use working locking for email user agents ] >SunOS seems to do both of these things just fine (these days... things >were not so rosy in earlier versions of SunOS). I dunno. I used to run a net with a 3/160 server and loads of diskless 3/60 clients, all running SunOS 3.5. I configured 'em up with about 3M or so of separate root per client (3.5 used ND for root), however much swap they needed (ND again), and then shared /usr via NFS from the server's copy. None of this /var silliness; they just shared /usr (and therefore all of /usr/spool). I ditched Sun's sendmail and built the then-current Berkeley one; I ditched Sun's lpr/lpd and got PLP; I configured everything in the obvious and straightforward way, and it worked beautifully. -Bennett bet@sbi.com ------------------------------ Crossposted-To: comp.os.linux.development From: mlindsey@nyx10.cs.du.edu (Mark R. Lindsey) Subject: comm programs for linux Date: Sat, 30 Oct 93 02:21:30 GMT I'm looking for some decent comm programs that'll compile under linux; my main problem is a lack of vt100 function key support. I'd also like to see if I could find something that'd make zmodem work. Minicom does neither, plus it screws up as text hits the bottom. I tried to get Pcomm, but I think I got a really old release that only seems to want to work on BSD systems. I don't know enough about programming to make it work for POSIX/linux. :( Thanks for any and all help. - Mark ------------------------------ From: jbettis@cse.unl.edu (Jeremy Bettis) Subject: Re: arg! smail: unknown user Date: 30 Oct 1993 02:31:01 GMT jmorriso@rflab.ee.ubc.ca (John Paul Morrison) writes: >I'm sick of this! smail consistently bounces incoming mail for only >a few users on the system. Smail says the user is unknown but the >user *is* in the passwd file. I suspect it is related to the passwd >file: I got mail for one user to stop bouncing after shuffled Do any of these users have capital letters in their names? smail will bounce them if UNIX_LOWERCASE_HACK was defined when compiling or if it is in the transports file. THe solution: Create a /usr/{local?}/lib/smail/transports file and in it set up then without using the unix_lowercase_hack parameter. -- Jeremy Bettis -*- Jerbo Jehoshaphat -*- University of Nebraska INET: jbettis@cse.unl.edu "Those who stand in the middle of the UUCP: jerbo@tddi.UUCP road are often hit by passing cars." Running Linux .99p13 Free Unix for i386/i486/Pentium machines. Ask me how. ------------------------------ From: muenx@heike.informatik.uni-dortmund.de (Holger Muenx) Crossposted-To: comp.os.linux.misc,comp.unix.questions,comp.unix.misc Subject: Introduction into Unix Date: 30 Oct 1993 04:22:44 GMT Guten Tag! Due to the fact that some of my users are no Unix wizards I am looking for some introduction paper for them to learn the basics. It's for a Linux system but needs not to be special written for this OS. It would be fine if it is available as plain ascii file. However, dvi or postscript will do it too. Even references to printed books/papers will be interesting for me. Please tell me the exact locations and names of mentioned files! Thank you in advance! Holger Muenx (muenx@heike.informatik.uni-dortmund.de) ------------------------------ From: muenx@heike.informatik.uni-dortmund.de (Holger Muenx) Crossposted-To: comp.os.linux.misc Subject: Textmode charset Date: 30 Oct 1993 04:33:37 GMT Guten Tag! Printing some files containing characters of the IBM charset (blocks and lines and suchs) yields funny results on a Linux virtual console. The characters seem to be mapped by some scheme to other characters. However, it's not just a 8bit->7bit conversion. On the other hand, if someones dials in over serial line and displays the same files on his terminal (Telemate on MS-DOS in this case) he sees the completely correct results. The files contain no ANSI escape sequences. They are plain 8bit text files with IBM charset. What's going on? How can I fix this without changing the files? Is it possible at all to display the IBM charset on a virtual console? This would be quite important for me because these files are the welcome screens of my former bbs which was replaced by a linux dial in. It was a lot of work to design them, my users like them and I would love to continue to use them, as well on the dial-in screens as on the local consoles where some of my users work sometimes. Any information will be appreciated. Thank you in advance. Holger Muenx (muenx@heike.informatik.uni-dortmund.de) ------------------------------ Crossposted-To: comp.os.linux.misc From: dminer@nyx10.cs.du.edu (Dan Miner) Subject: [QUESTION}: simple news reader? Date: Sat, 30 Oct 93 07:08:43 GMT I'm looking for a news reader like rn (or trn) for home use. I do not have network access except by logging in to nyx.cs.du.edu and getting the articles that way, a simple tar exercise. :) Is there a reader out there that will compile WITHOUT all the network stuff? (Please say trn or rn will do this... :) Thanks, Dan dminer@nyx.cs.du.edu ------------------------------ From: jmorriso@rflab.ee.ubc.ca (John Paul Morrison) Subject: Re: arg! smail: unknown user Date: 30 Oct 1993 07:34:32 GMT In article <2asjl5$bri@crcnis1.unl.edu>, Jeremy Bettis wrote: :jmorriso@rflab.ee.ubc.ca (John Paul Morrison) writes: :>I'm sick of this! smail consistently bounces incoming mail for only :>a few users on the system. Smail says the user is unknown but the :>user *is* in the passwd file. I suspect it is related to the passwd :>file: I got mail for one user to stop bouncing after shuffled : :Do any of these users have capital letters in their names? smail will bounce :them if UNIX_LOWERCASE_HACK was defined when compiling or if it is in the :transports file. THe solution: Create a /usr/{local?}/lib/smail/transports :file and in it set up then without using the unix_lowercase_hack parameter. no, none of the users have upper case. local mail works fine, put for several users incoming mail bounces. I noticed one site where it was doing this, so I kept testing it with mail -s test local_user%rflab.ee.ubc.ca@some_other_host when local_user is my ID, or several others, the mail works perfectly. when I substitute others, it bounces with "user unknown" I went through the passwd file testing each userid, and they worked until I got to one that didnt work. I couldnt find anything weird about the userids that had mail bounce. I found "holes" in the passwd file where the userids weren't consecutive integers, so I filled in those userids, and that didnt work. I checked the /etc/shadow file and found that it was slightly out of sync with /etc/passwd, so I added/deleted accounts and I made /etc/passwd and /etc/shadow correspond with each other line by line. No luck. I briefly corrected the bouncing mail for one two users, but it seems to have reverted when I tried getting it to work for others. :-- :Jeremy Bettis -*- Jerbo Jehoshaphat -*- University of Nebraska :INET: jbettis@cse.unl.edu "Those who stand in the middle of the :UUCP: jerbo@tddi.UUCP road are often hit by passing cars." :Running Linux .99p13 Free Unix for i386/i486/Pentium machines. Ask me how. -- ___________________________________________________________________________ John Paul Morrison The Liberal Party of Canada: University of British Columbia, Canada proof that those who do not Electrical Engineering remember the past are condemned jmorriso@rflab.ee.ubc.ca VE7JPM to repeat it Day 5 of Canada held hostage: the socialist Liberal tax & spend tyranny ------------------------------ From: juphoff@uppieland.async.vt.edu (Uppie) Subject: Re: finger not reporting last login time! --here's one fix. Date: 30 Oct 1993 08:59:36 GMT This works for SLS installs (which don't quite work for last logins, etc.). I wish I could remember who gave this to me, because it works perfectly. Here's the email I received and used to fix my problems... ============================================================ Here are the patches you have been waiting for! In SLS, last login times are not reported by finger. Through email, a few of us have figured out how to fix this, and I have now gotten around to posting it. :-) To make the lastlog file be written so that finger can read it, follow the instructions below: (I have added what is in brackets.) ========================================================================== Two changes have to be made in /usr/src/shadow. You should replace your lastlog.h (probably should change it to point to but I couldn't be bothered) with this: ---cut here------------------------------------------------------- /* * Copyright 1989, 1990, 1992, John F. Haugh II * All rights reserved. * * Permission is granted to copy and create derivative works for any * non-commercial purpose, provided this copyright notice is preserved * in all copies of source code, or included in human readable form * and conspicuously displayed on all copies of object code or * distribution media. * * This software is provided on an AS-IS basis and the author makes * not warrantee of any kind. */ /* * lastlog.h - structure of lastlog file * * @(#)lastlog.h 3.1.1.1 13:02:32 7/27/92 * * This file defines a lastlog file structure which should be sufficient * to hold the information required by login. It should only be used if * there is no real lastlog.h file. */ struct lastlog { time_t ll_time; char ll_line[12]; char ll_host[16]; }; ---cut here----------------------------------------------- and add the line : (void) strncpy (newlog.ll_host, utent.ut_host, sizeof newlog.ll_host); to log.c just after the line : (void) strncpy (newlog.ll_line, utent.ut_line, sizeof newlog.ll_line); then recompile and re-install your shadow passwd suite and reboot. Now login will write the same format lastlog as finger expects. [If /usr/adm/lastlog does not exist, do 'touch /usr/adm/lastlog'] Symbolic link /etc/lastlog to /usr/adm/lastlog. [ by doing 'ln -sf /usr/adm/lastlog /etc/lastlog'] ============================================================================== I don't know why the current version of SLS does this, but the correct set up for last to work requires that the real version of wtmp be in /etc and a symbolic link to it be in /usr/adm. I did the following and now last works: rm /usr/adm/wtmp* rm /etc/wtmp.old mv /etc/wtmp /etc/wtmp.old touch /etc/wtmp ln -sf /etc/wtmp /usr/adm/wtmp ...and now all seems to be well. I deleted the old faulty wtmp so continuous streams of 'still logged in' messages wouldn't be around all the time. Happy linuxing! ============================================================================== *** END OF FIX *** My modified version of finger (I called it 5.23) ignores the invalid utmp entries left over by closed xterms, which showed users still logged in until a user (any user) ran a logout, which seemed to clear up utmp. It also shows mailbox status, as well as gives the proper time zone. -- Jeff Uphoff -- "Uppie" | "The secret to good teaching is sincerity. uppie@vt.edu | As soon as you learn to fake that, you've got juphoff@uppieland.async.vt.edu | it made." ------------------------------ ** 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 ******************************