From: Digestifier To: Linux-Admin@senator-bedfellow.mit.edu Reply-To: Linux-Admin@senator-bedfellow.mit.edu Date: Sat, 24 Sep 94 14:13:36 EDT Subject: Linux-Admin Digest #100 Linux-Admin Digest #100, Volume #2 Sat, 24 Sep 94 14:13:36 EDT Contents: how much disk space for X (windows) (Dolf Smits) SLIP/CSLIP & E-mail (James Nykiel) Re: /dev/audio: No such file or directory (G Cheng) Pentiums (thanks!) (Frank B. Brokken) Re: Howto duplicate boot floppy? (Frank B. Brokken) Smail on Linux. Was: Smail has intermittent fits (Lee J. Silverman) Re: Linux won't keep correct time (David Kastrup) Re: Need DL/Time Limiting ideas - Linux BBS (Riccardo Pizzi) No Hostname (Ben Pressnall) Can't build new kernel, SLS 1.05 (davidsen@tmr.com) ---------------------------------------------------------------------------- From: dolf@interduct.tudelft.nl (Dolf Smits) Subject: how much disk space for X (windows) Date: Tue, 20 Sep 1994 06:51:00 GMT Hello, Most of this is in the subject. I use linux already (slackware 1.2.0, kernel 1.0.8) but without X I don't have much diskspace so how much is needed to install X? Related question, I recently started using a nfs mounted drive which has space enough. Can I use this disk for X? What directories are needed at booting and cannot be placed on the nfs disk, and which one can be moved to the nfs disk to free up space on my local disk? Can anyone shed some light on this? Thanks Dolf Smits -- Smits@interduct.tudelft.nl D.F. Smits Interduct Delft University Clean Technology Institute Rotterdamseweg 145 2628 AL DELFT tel. (+31) 15 78 72 39 The Netherlands fax. (+31) 15 78 66 82 Linux, The choice of a GNU generation (Anonymous) ------------------------------ From: jnykiel@vortex.tiac.net (James Nykiel) Subject: SLIP/CSLIP & E-mail Date: 20 Sep 1994 07:16:07 GMT This is probably a stupid question but can somebody help me out with this, ok, I call my internet provider and start my SLIP/CSLIP connection as the root user, the connection is established so I switch virtual consoles and log in as jnykiel and start X Windows. I decide to send somebody E-mail which I have no problems but what happens when somebody sends me E-mail and I am offline (no SLIP/CSLIP connection), where does this mail go, does the sender get a message that says I am un- reachable or what, all I know is that I never get any mail, possibly know- body loves me :( or I am doing something wrong. Any help will be greatly appreciated, Thanx... James Nykiel ------------------------------ From: ubacr46@naga-1.uucp (G Cheng) Subject: Re: /dev/audio: No such file or directory Date: Sat, 24 Sep 1994 16:20:27 +0100 (BST) richardk@world.std.com (Richard Keightley) wrote: > >From the Sound HOW TO by Jeff Tranter > > An alternate sound driver is available that requires no additional > sound hardware; it uses the internal PC speaker. It is mostly software > compatible with the sound card driver, but, as might be expected, > provides much lower quality output and has much more CPU overhead. The > results seem to vary, being dependent on the characteristics of the > individual loudspeaker. For more information, see the documentation > provided with the release. > > The current version is 0.6, and can be found at site sunsite.unc.edu > in the file pub/Linux/kernel/misc-patches/pcsndrv-0.6.tar.z. > >For a little more information look at chapter 5 of the Linux Sound User's >Guide also by Tranter.The current version is ALPHA 0.1, and is available on >tsx-11.mit.edu in the directory /pub/linux/ALPHA/LDP. > >-- >Richard Keightley richardk@world.std.com >Scottsdale, AZ fax (602) 443-8196 Thanks Richard and those who answered via email. I got pcsndrv-0.7 at sunsite.unc.edu /pub/Linux/kernel/patches/console, and it works fine now. By the way, it mentions that it is also capable of playing music through parallel ports and external speakers. Haven't got a chance to try though. Have a nice day. -- G Cheng ~{<=3G~} Department of Computer Science G.Cheng@dcs.bbk.ac.uk Birkbeck College, University of London ------------------------------ From: frank@icce.rug.nl (Frank B. Brokken) Subject: Pentiums (thanks!) Date: Sat, 24 Sep 1994 15:03:50 GMT The other day I posted a request about using pentiums with linux (or vv ?) and got about 10 reactions, generally positively answering my question about the use of linux on Pentiums. Everybody who reacted to my request: thanks for taking the time to do so. If anybody is interested in the mail I got, please let me know, and I'll forward the reactions to you. | ---|--- =====+==0==+===== | | Frank E-mail to: f.b.brokken@icce.rug.nl or: frank@icce.rug.nl Phone: (+31) 50 63 36 88 (mo-fr, 9:00-17:00 MET, DST if active) (+31) 59 03 22 23 (otherwise, until Oct. 9, 1995) (+31) 50 403 2223 (starting Oct. 10, 1995). ------------------------------------------------------------------- icmake: obtain it by anonymous ftp from ftp.icce.rug.nl, directory pub/unix The LINUX SYMPOSIUM will be organized december 8/9 1994. finger linux@obelix.icce.rug.nl for the latest details ------------------------------------------------------------------- ------------------------------ Crossposted-To: comp.os.linux.help From: frank@icce.rug.nl (Frank B. Brokken) Subject: Re: Howto duplicate boot floppy? Date: Sat, 24 Sep 1994 15:32:38 GMT jsun@athena.mit.edu (Jim Sun) writes: >How to duplicate more boot floppy from one existing floppy? >I'm facing some serious trouble when installing linux on a Pentium >with 1G HD; linux is within the first 300meg, so I don't think the >1024 cylinder problem should affect me. But the new installation >refused to boot ("partition table error"); the floppy created during >the installation refused to boot too. The only thing got the system >going was a floppy from a previous installation on a 486. Now since >that disk was reated only for backup purpose, I'd prefer to make a >seperate one for booting the Pentium box. Any thoughts on how to >copy a floppy containing kernel image? >Thanks; please cc:jsun@mit.edu in your response >Jim Hi Jim, Karel Kubat did a great job in finding out how to build SAR (Search and rescue) disks. The information and the system he made available might just be the answer to your questions. The SAR-system is in any case available at ftp.icce.rug.nl, directory pub/unix. If you have a www-browser, you can get a lot of information about, e.g., the SAR system from our URL http://www.icce.rug.nl If you have any questions, please yell. Good Luck, | ---|--- =====+==0==+===== | | Frank E-mail to: f.b.brokken@icce.rug.nl or: frank@icce.rug.nl Phone: (+31) 50 63 36 88 (mo-fr, 9:00-17:00 MET, DST if active) (+31) 59 03 22 23 (otherwise, until Oct. 9, 1995) (+31) 50 403 2223 (starting Oct. 10, 1995). ------------------------------------------------------------------- icmake: obtain it by anonymous ftp from ftp.icce.rug.nl, directory pub/unix The LINUX SYMPOSIUM will be organized december 8/9 1994. finger linux@obelix.icce.rug.nl for the latest details ------------------------------------------------------------------- ------------------------------ From: lee@netspace.students.brown.edu (Lee J. Silverman) Crossposted-To: comp.mail.smail Subject: Smail on Linux. Was: Smail has intermittent fits Date: 24 Sep 1994 15:12:16 GMT For all you Linux types: Smail is configured incorrectly in slackware. (Last I checked; I'm running Slackware 1.2 with my own mods). First, and ABSOLUTELY critical: (if you skip the rest of this post, at least read this!!) Put the line: -smtp_debug in your /usr/lib/smail/config file. It turns off SMTP debugging, which can give an unauthorized user access to run commands on your machine (probably as user nobody, but remember that smail runs as root.) Second: smail is set up to be called from inetd. This is startlingly inefficient. Smail must reparse all of its configuration files each time it is started up. If you've looked at the config files, you can imagine that this takes a while. Furthermore, mail that has been deffered or otherwise queued up won't be delivered by this method. So, here's what you do. Remove the call to smail from /etc/inetd.conf. In your /etc/rc.d/rc.inet2 file, add the following: # Start the SMAIL SMTP server. if [ -f /bin/smail ] then echo -n " smail-SMTP" /bin/smail -bd echo -n " smail-Queue" /bin/smail -q20m fi smail -bd starts a daemon that sits and waits for connections on the SMTP port (25), and spawns off a child proccess every time new mail arrives. It only has to procceess its config files when it starts, so it's startlingly more efficient. Smail -q20m sets up a daemon that proccesses the "input" queue every 20 minutes. Mail that smail -bd could not deliver immediately goes into this queue. One of smail's only shortcomings is that it only has one queue. Mailers like MMDF and Zmailer use multiple queues so that if, for example, AOL.com is down when a queue run is performed, Zmailer doesn't have to proccess every message in the queue that's destined for AOL. It simply tries the first one and when that doesn't work it skips the rest. The "retries" files help out here, because it doesn't try to connect for every message, but it still has to proccess the entire queue. Unfortunately, Zmailer doesn't work on Linux boxes yet, and MMDF is a dinosaur that I didn't want to install for fear that it might stomp on something. Good design, though.) Third: If you're on the internet, go to the /usr/lib/smail/ directory and edit the "transports" file. Uncomment the last four lines in order to allow internet mail to be delivered using MX records, not hostnames. Here are the drivers I have in transports, in order (I use procmail for local mail delivery, which is why the first driver is bogus, and the second driver delivers to local users.) bogus: driver = appendfile, # append message to a file local: driver = pipe, # append message to a file pipe: driver = pipe, # pipe message to another program file: driver = appendfile, smtp: driver = smtp, Since I'm not on UUnet uux doesn't do me any good so I got rid of the UUCP drivers. Here's the full smtp driver, since that's going to be the most important one: smtp: driver = smtp, -max_addrs, -from, -max_chars; # For internet use: uncomment the below 4 lines use_bind, # resolve MX and multiple A records defnames, # use standard domain searching defer_no_connect, # try again if the nameserver is down local_mx_okay, # pass on to next router if MX is us It's these last four that make mail deliverable over the internet, especially the use_bind tag. Smail defaults to using gethostbyname, which means that any mail for a machine with an MX record won't be delivered. Next, edit the "routers" file. Once it is decided that mail is going to go to a remote machine, this file is used to determine how to get the message there. Here are some of the key elements in this file: match-inet-addrs: driver=gethostbyaddr, # match user@[aaa.bbb.ccc.ddd] transport=smtp; # delivery is over smtp/tcp forces: # Just in case I need it driver = pathalias, # router to search paths file match_mx_hosts: # Match folks on the internet driver=bind, # Get info from nameserver transport=smtp; # use TCP/IP SMTP for delivery defnames, # use standard domain searching local_mx_okay, # if the MX points to this host, pass # it to the next router and see if # we serve it a different way defer_no_connect, # try again if the nameserver is down gateways="brownvm.brown.edu:bitnet:+:uunet.uu.net:uucp" {This is the REALLY important one. The vast majority of your outgoing mail will use this driver. Notice that it uses bind instead of gethostbyname. VERY important for machines with MX records. Also notice the "gateways" line. Mail to user@XXXXX.bitnet will be sent through Brown's mainframe (substitute in a more local Bitnet gateway in your own files, please) and mail to user@YYYY.uucp will go through uunet.uu.net (again, you may want to substitute a more local UUCP gateway). Defer_no_connect is really important if you're using the bind driver, because it allows the mailer to retry sending the mail if the nameserver is down.} match-inet-hosts: driver=gethostbyname, # match hosts on network transport=smtp; # delivery is over smtp/tcp domain = students.brown.edu # strip trailing domain before lookup {Most of the time the match_mx_hosts will deliver the mail before this driver is reached.} {MAKE SURE THE REROUTE DRIVER IS COMMENTED OUT!!!!!!!!!!!!!!!!!!} smart_host: driver=smarthost, # special-case driver transport=smtp; # by default deliver over SMTP {The smarthost is listed in the config file.} Lastly, the config file. Here's a copy of mine, just to get people started: #-------------------------------------------------------------------------- # # smail configuration for NetSpace # (see smail(5) man page for details and other options) # # Primary hostname hostname=netspace.students.brown.edu # Other hostnames that mail will be accepted for. ALl of your machine's # names should go in this list. Note that there are no returns. more_hostnames=netspace.cis.brown.edu:archive.phish.net:home.eos.brown.edu:ssc.org:netspace.stg.brown.edu:netspace.org # Mail coming out of Netspace has this address on it. visible_name=netspace.students.brown.edu # How to get to the "smart" machine. As if an IBM mainframe were "smart". :-) smart_path=brownvm.brown.edu smart_transport=smtp # When mail is delivered to "postmaster", who does it go to? postmaster=lee # Needed for Listproc to work correctly, among other things. Make sure there # is a user "nobody" in /etc/passwd with a * in the passwd field: # nobody:*:15:12:nobody:/home/nobody:/bin/logout nobody=nobody # If this many messages are being proccesed at once, queue them instead of # attempting immediate delivery. smtp_accept_queue=20 # Don't accept any more than this many SMTP connections at once. Imagine the # load of 60 smail proccesses on a Linux box... cripes! smtp_accept_max=60 # Other schtuff retry_file=retry smail=/bin/smail delivery_mode=foreground # Turn of SMTP debugging. CRITICAL!!!!! -smtp_debug # I don't want to see copies of every bounced message on my machine... -error_copy_postmaster # This is the line that comes up when someone makes an SMTP connection smtp_banner="Welcome to NetSpace! We're running Smail here." # This is the same as the default. received_field="Received: \ ${if def:sender_host\ {from $sender_host by $primary_name\ ${if def:sender_proto: with $sender_proto}\ \n\t(Smail$version #$compile_num) }\ else{by $primary_name ${if def:sender_proto:with $sender_proto }\ (Smail$version #$compile_num)\n\t}}\ id $message_id; $spool_date" #--------------------End of Config File--------------------------------- On a final note: My linux box runs the Listproc mailing list software, with over a dozen lists, 5 of which have over 250 users and see more than 5 messages a day. If you use listproc, append the flag -m 250 after each list definition. This tells Listproc that each "system" SMTP connection, where the initial delivery of mail takes place, delivers mail to up to 250 different people. That means for a list with 900 subscribers, only 4 smail proccesses are started. Although these proccesses take considerably longer to run than a smail proccess with only one message to deliver, they take far less time to run than starting 900 seperate smail proccesses, and the load on the machine is much less! Also, make sure the server line in your listproc config file tells it not to attempt delivery if the load is over "1", unless you're only using your Linux box to deliver mail. A linux box at a load of 1 is slow enough already without Listproc slowing things down more. All told, my machine proccesses about 5000 email messages a day, sometimes more. I have not had a message stay in the "input" queue for more than two days in over a week, even though smail is configured to retry delivery every 20 minutes for 5 days. I haven't had any messages at all in the "error" queue (mostly because of the smart_host driver). The load on my machine due to mail proccessing is SMALL, because Listproc and smail are set up to handle it. I hope this post helps at least a few people out there! Take care! -- Lee Silverman, Brown class of '94, Brown GeoPhysics ScM '95 Email to: Lee_Silverman@brown.edu Phish-Net Archivist: phish-archives@phish.net "Nonsense - you only say it's impossible because nobody's ever done it." ------------------------------ From: dak@messua.informatik.rwth-aachen.de (David Kastrup) Subject: Re: Linux won't keep correct time Date: 24 Sep 1994 15:42:45 GMT tkeidl@fwk103034.res-hall.nwu.edu (Tobias S. Keidl) writes: >It seems that every time I boot up my Linux box, the clock always is 1 >hour behind what it should be. Any time I reset the time, it changes >the next time I reboot. I presume there is some config file somewhere >telling it to do that but I have no idea where or what (and it was also >exhibiting the same behavior before I was networked). Anyone have any >solutions or suggestions? I'd appreciate them! >-- I suppose you have rewritten your hardware clock as well using clock -w after being logged on as superuser? date just sets the system clock, which is maintained using the timer chips. -- David Kastrup dak@pool.informatik.rwth-aachen.de Tel: +49-241-72419 Fax: +49-241-79502 Goethestr. 20, D-52064 Aachen ------------------------------ Crossposted-To: alt.bbs,alt.bbs.unixbbs From: pizzi@nervous.com (Riccardo Pizzi) Subject: Re: Need DL/Time Limiting ideas - Linux BBS Date: Fri, 23 Sep 1994 00:35:48 GMT In article <35r0mv$cu@lvl-sun683.usc.edu> killourh@lvl-sun683.usc.edu (Patrick Killourhy) writes: >|> This is only partially true. UniBoard _does_ check for time expiration >|> after each downloaded file (even during batch download) and will disconnect >|> the user if his time limit is overridden. >Why don't you just modify Uniboard so that it spawns a daemon that does >the online-time checking concurrently with the sz/sb/sx process? People >go out and write things for Unix, and never take full advantage of the >fact that it's a completely multi-user, mutli-tasking OS. Basically, I didn't like the idea of a user file transfer aborted in the middle because the time-for-call expired. This is why it works this way. A UniBoard sysop here in Italy runs a BBS for profit over a `144' line (I think it is 1-900 for you US people: a line which charges the user some money per minute, and some of the $$$ go into the sysop's pockets). He modified his zmodem to do exactly what Patrick suggested: the caller is brute-force disconnected when the timer for the call expires. He did this by forking a daemon from within zmodem (usual rz/sz by Omen Technology). Reason? The telco wouldn't pay to the sysop anything for calls over 30' in lenght :-) Rick -- Riccardo Pizzi @ the Nervous Consulting Headquarters - Rimini, ITALY E-Mail -> pizzi@nervous.com <*> Data -> +39-541-27135 (V.32bis) NERVOUS CONSULTING brought you UniBoard, the great UNIX BBS package for latest UniBoard info, use: echo SEND LIST | mail uniboard@nervous.com ------------------------------ From: pressnal@chem.uidaho.edu (Ben Pressnall) Subject: No Hostname Date: Tue, 20 Sep 1994 23:59:31 GMT At my login prompt the hostname reports (none). I can make a temporary fix by using hostname_notcp (hostname), but defaults back to (none) after a reboot. Can someone help. Thanks Ben Pressnal pressnal@uidaho.edu ------------------------------ From: davidsen@tmr.com Subject: Can't build new kernel, SLS 1.05 Date: 22 Sep 1994 18:33:16 -0400 I just tried to build a kernel on SLS 1.05 for the first time, and everything went as expected but the kernel won't boot. It dies with an error about "can't mount filesystem device 770" 1. I've built kernels with SLS and Slackware before 2. I tried make clean, config, dep, all to get old stuff out. 3. I reran LILO Is there a known problem with this? I can't switch versions, I have to match a client. Oh, I posted to the 'hrlp' group and didn't get any... -- Speaking *from* but never *for* Prodigy "Pain builds moral fiber" -my dad "Pain hurts" -me ------------------------------ ** 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 ******************************