link button (was: Debian 2.1 NOW! logo)
Carey Evans [EMAIL PROTECTED] writes: With the imminent release of Debian 2.1, and people asking how to make Debian more visible, I've revamped a logo I created a while ago, and put it on my web site. It's less than 2K as a GIF, but even so I won't attach to an email to a public list. I'd just like to clarify this post. There's no way this is the official logo (and I designed it in my capacity as an enthusiastic Debian user, not as [EMAIL PROTECTED]); it's just a little button I threw together for people who want to put something on their web page. I'm looking forward to us getting a real logo soon. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ And so, New York has joined the fraternity of cities whose only admission requirement is to be overrun with evil zombies.http://www.sluggy.com/
Re: help for samba.conf
Dan Pomohaci [EMAIL PROTECTED] writes: I have a Linux server (Debian - slink) which must share files with a bunch of Windows boxes. I've installed Samba on Linux and have made all the modifications (I think) on Windows clients. Now they can see the Linux server but they are not permited to access the files from Linux server. IMHO, you should be using security = user. Give everyone a Unix account, and set the MS password using `smbpasswd'. You may want to consider upgrading to the latest Samba (version 2) from potato too. If it still doesn't work, see what kind of errors you get with smbclient, e.g. $ smbclient '\\hostname\tmp' -U whoever You can add -d 2 or -d 5 to get debugging output from smbclient. You can also increase the debug level, and check (IIRC) /var/log/smb for errors. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ And so, New York has joined the fraternity of cities whose only admission requirement is to be overrun with evil zombies.http://www.sluggy.com/
Re: rm message..what is the answer.
Roddie Rod [EMAIL PROTECTED] writes: I'm still trying to delete broken symlinks and files that are hosing my system. When I try to use rm on files I get the following: rm -r /usr/lib/mc, overiding mode 5265? any answer I give, yes, enter or chmod 666 give me either nothing or operation not permitted. Have a look at my posts in December and February about this: http://www.debian.org/Lists-Archives/debian-user-9902/msg00188.html http://www.debian.org/Lists-Archives/debian-user-9812/msg02771.html and a reply: http://www.debian.org/Lists-Archives/debian-user-9812/msg03298.html -- Carey Evans http://home.clear.net.nz/pages/c.evans/ And so, New York has joined the fraternity of cities whose only admission requirement is to be overrun with evil zombies.http://www.sluggy.com/
Re: Autostart application from console
Jiri Baum [EMAIL PROTECTED] writes: [snip] The showlog script is: date +$0 started %c; /usr/bin/tail -f /var/log/syslog You can also do this by uncommenting some lines in /etc/syslog.conf: daemon,mail.*;\ news.=crit;news.=err;news.=notice;\ *.=debug;*.=info;\ *.=notice;*.=warn /dev/tty12 So that Alt-F12 displays the most recent syslog messages, with very little overhead. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ And so, New York has joined the fraternity of cities whose only admission requirement is to be overrun with evil zombies.http://www.sluggy.com/
Debian 2.1 NOW! logo
With the imminent release of Debian 2.1, and people asking how to make Debian more visible, I've revamped a logo I created a while ago, and put it on my web site. It's less than 2K as a GIF, but even so I won't attach to an email to a public list. The GIMP source for the picture is at: http://home.clear.net.nz/pages/c.evans/pic/debian-now-2.1.xcf or as a GIF at: http://home.clear.net.nz/pages/c.evans/pic/debian-now-2.1.gif If someone decides to announce this on Slashdot or somewhere, *please* let me know so I can put it on a server in the US. This image is my original work, apart from Tux, and I suppose the idea isn't exactly original. There are two restrictions on the use of this logo. First, because it is based on the picture of Tux from the Linux kernel distribution, you (and I) must acknowledge that Larry Ewing created and drew Tux originally. (See linux/Documentation/logo.txt.) Second, you MAY NOT link to the image at the location above as an inline image from HTML. Copy the image itself to your own site, and this restriction disappears. I don't even require credit, although I suppose it would be nice. :-) Despite this lack of restrictions, I really would prefer if it was used as a pointer to URL:http://www.debian.org/ or a suitable page on that site. Obviously, for the release of Hurd I can replace Tux with a Gnu. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ And so, New York has joined the fraternity of cities whose only admission requirement is to be overrun with evil zombies.http://www.sluggy.com/
Re: Debian and Redhat - are most linux users missing the point?
[EMAIL PROTECTED] writes: It seem that dselect and apt-get must use two different Package lists. Right. I would have expected them to use the same one as they both use dpkg. There's various levels to dpkg - apt just uses it to install packages, dselect uses it to manage the available file too. The easiest way to get the package lists in sync is to choose apt as the access method in delect, and use `dselect update' instead of `apt-get update'. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ And so, New York has joined the fraternity of cities whose only admission requirement is to be overrun with evil zombies.http://www.sluggy.com/
Re: Network Not Working!
Roland E. Lipovits [EMAIL PROTECTED] writes: On Thu, Feb 11, 1999 at 12:36:14AM +1300, Carey Evans wrote: Trying to use 192.168.0.xxx is asking for trouble, in my opinion. Try Why? Oops. Because: 1. It was after midnight the first day back at work, and I really shouldn't post in that condition. (At least when I'm drunk the posts are usually amusing to everyone else.) 2. Because I'm very wary of putting too many zeros or ones (bits) in an IP address. For instance, 192.168.1.224 is a bad address for a host when the netmask is 255.255.255.240, but that's not always obvious at first glance. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: [FIXED] Re: Network Not Working!
Peter Ludwig [EMAIL PROTECTED] writes: Umm... why? What IS the /etc/lmhosts file? And if samba wants it, why is it not referred to in the Samba Documentation (the only thing I haven't done is grep the files to find it, perhaps I should...)? $ man lmhosts Reformatting lmhosts(5), please wait... [...] This *is* with Samba 2.2.0beta3 though, so you might not have it in your version. Basically, you use this file if you don't have a WINS server, all your hosts have static names, broadcasting names isn't working properly, or you think it might speed things up. e.g. I have: 192.168.15.7NEPTUNE because broadcast doesn't work over a PPP link and I want to be able to access Neptune at work from my home dialup, without worrying about how WINS will work when I disconnect. More information is in the smb.conf(5) man page under `name resolve order'. If your network works and you don't have an lmhosts file, the only effect is that you get an annoying error, which is why I said you could ignore it. You could also put the following in your smb.conf. name resolve order = bcast host -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: Network Not Working!
Peter Ludwig [EMAIL PROTECTED] writes: I installed Debian telling it that I was on a network (the home network, two machines, nothing fancy), and yet I seem to be unable to talk to the other machine. This is the response I get from SMBCLIENT : Added interface ip=192.168.0.1 bcast=192.168.0.255 nmask=255.255.255.0 startlmhosts: Can't open lmhosts file /etc/lmhosts. Error was No such file or directory This is generally harmless, and you can ignore it for now. error connecting to 192.168.0.11:139 (No route to host) Trying to use 192.168.0.xxx is asking for trouble, in my opinion. Try setting Linux and Windows to 192.168.1.something. The files to modify are /etc/init.d/network, /etc/hostname and /etc/hosts. The hosts file could look something like: - 127.0.0.1 localhost 192.168.1.1 linux linux.kahnnet # I'm just making up the domain. 192.168.1.2 win95 win95.kahnnet - It should look the same in C:\WINDOWS\HOSTS too. Also on Win95, set the default gateway to your Linux box. Things to try now: On Linux: `netstat -rn', `/sbin/ifconfig', `ping 192.168.1.1', `ping 192.168.1.2', `ping win95'. $ telnet win95 Trying 192.168.1.2... telnet: Unable to connect to remote host: Connection refused $ smbclient -L win95 -I 192.168.1.2 [...] $ smbclient -L win95 [... should be the same] On Win95: C:\PING 192.168.1.1 Start-Find-Computer LINUX Start-Run telnet 192.168.1.1 The output of these commands would be useful. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: no permission for rnews
[EMAIL PROTECTED] (Frederick Page) writes: finally got exim up and running with uucp. Everything seemed to work: I received email, even could send it. But when I got news via uucp, the uupoll gave me error-messages and sent the following emails to root: [snip] Return-path: [EMAIL PROTECTED] Received: from uucp by thebetteros with local (Exim 1.92 #1 (Debian)) id 108U1T-3E-00; Thu, 4 Feb 1999 20:01:07 +0100 To: [EMAIL PROTECTED] [snip] UUCP is trying to send mail (the error message) to downtown!news. Exim doesn't understand bang-path addresses, so it thinks that downtown!news must be a username on your system. You could probably fix this easiest by adding a rewriting rule to exim.conf: [EMAIL PROTECTED] [EMAIL PROTECTED] bctrT OTOH, maybe you don't want your ISP getting heaps of mail because your system isn't configured properly yet, so it might be best to leave it for now. [snip] What's wrong? I did uupoll as root, don't understand this. Could someone please tell me which commands to man and/or to issue? Do I need additional software, besides UUCP, exim (kind of sendmail), procmail and pine? You need a news server, such as C News or INN. You're going to have great fun setting one of them up... ;-) You might have to then reconfigure UUCP to recognise the enws server. [snip] So the good thing is: the news are saved, but how do I get them back? Which command do I need to run in order to re-try and deliver the news? If one of these files start with #! rnews ... you can just run rnews on it: # rnews D.downtowdcc58 -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: setserial question
Anthony Campbell [EMAIL PROTECTED] writes: Use of setserial/setrocket to get SPD_* flags is deprecated Are you sure that's get SPD_? The only string like that in the kernel is in drivers/char/tty_io.c line 1903: printk(Use of setserial/setrocket to set SPD_* flags is deprecated\n); [snip] I tried this but it didn't help. I wish I knew what SPD_* flags are and where they are set. As far as I can tell, dip is using tcgetattr() on a TTY, probably a serial line, which at one point leads to tty_get_baud_rate() being called. The latter function is about to return 38400 as the baud rate, but notices that the alt_speed value has been set for that TTY, so returns its value and prints that warning. The value for alt_speed can be set in set_serial_info() in drivers/char/serial.c. I'm guessing that you're specifying spd_something in /etc/rc.boot/0setserial. You shouldn't need to do this if all programs are using the latest POSIX termios code, so Linux 2.2 complains about it. Try removing these flags if they're there, and see if anything stops working. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: Can I block _only_ privileged access with ipchains?
[EMAIL PROTECTED] (Dale E. Martin) writes: I want my firewall to allow access from unprivileged ports to unprivileged ports. So, if I connect to an ftp site on a port above 1023 and my client is using a port above 1023, the packets are allowed to cross. Possible? I can find anything about it in the docs. Any other comments? If you're using Linux 2.2, I've put up a document at http://home.clear.net.nz/pages/c.evans/doc/homefirewall.html that describes the setup I've got to allow this sort of thing. Consider this the pre-release announcement; I've still got a bit of clean-up work to do on it, but it's basically finished. If you're running Linux 2.0, check the ipfwadm(8) man page; my docs might still be of use. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: minicom config
ktb [EMAIL PROTECTED] writes: ~$ minicom d minicom: there is no global configuration file /etc/minirc.d Ask your sysadm to create one (with minicom -s). Why d? Just minicom is all I've needed, except on the dialup server with four modems. Apart from maybe the serial port, you shouldn't have to change anything. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: exim: I get mail but cannot send
[EMAIL PROTECTED] (Frederick Page) writes: My responses are located with cryptic names in /var/spool/exim/input, obviously ready to be delivered but don't seem to get transmitted to my uucp provider. /var/log/exim/* should have info about why your mail is just sitting there. See what sort of messages you've got. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: chmod script
Pere Camps [EMAIL PROTECTED] writes: Hi! Does anybody have a script for changing all the dirs and subdirs into mode 700 and all the files into mode 600 ? Maybe chmod -R u=rwX,go= path Capital X means preserve the exisiting execute permission. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: start-stop-daemon and multiple exim daemons
[EMAIL PROTECTED] (Marc Haber) writes: I am running exim as two daemons. One is listening on port 25, the other does periodic queue runs only when I am online. The queue runner exim is killed off when I go offline. [snip] How can I make start-stop-daemon check for a process' existence before I trying to kill it? I'd really love to get rid of this error message. After some experimentation, it seems that start-stop-daemon doesn't print anything if you pass both the --quite and --exec flags, so this would be one option. You could also just redirect the output with /dev/null at the end. Actually, I found an old version (1.4.0.7) of start-stop-daemon, when it was written in Perl, and it does have the behaviour you want. Maybe this is a bug in start-stop-daemon. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: man/apropos/mandb problem...
Roger Franz [EMAIL PROTECTED] writes: Evidently, fsck lost+found'd the index files for apropos/man/mandb, and when I run mandb -c to recreate them, it tells me it can't. (Permission denied errors, files missing, etc.) I assume you're doing this as root? Try erasing the files with the problems - use `dpkg -S' to find out what packages they're from so that you can reinstall the package and get the file back if necessary. Use `strace -e file' if mandb doesn't explain very well which files are causing the problems. If you can't delete the files as root, first try `chattr -i' on them. If _that_ doesn't work, you may have to resort to black magic: Backup what you need. Copy debugfs onto a floppy. Boot from the rescue disk and mount this floppy (and not the hard disk). Run debugfs on the partition with the problem, then use its `rm' command on the offending file. Save the changes and reboot, and let fsck clean up the mess. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: setserial question
Anthony Campbell [EMAIL PROTECTED] writes: Use of setserial/setrocket to get SPD_* flags is deprecated This seems to be said more in sorrow than in anger, since the process then goes off correctly. But what does it mean, and what can I do about it? It probably means your version of setserial (or maybe dip) is old. Try downloading the one from slink (frozen) and see if it will work without any complaints: # dpkg -i setserial_what-ever.deb It would be a good idea to have the old package around in case this doesn't work. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: Network problem
Hersh, Harry [EMAIL PROTECTED] writes: I switched back to the X11 environment, and find that even though pon connects to my isv (I can ping ), none of my network programs will work: Netscape, ftp, cftp, etc. The all start up, and then just hang. [snip] Things to try: Does ping work with a hostname and an IP address? Do the other programs? Do they work from the console? Do they work as root? Is /etc/resolv.conf OK? Is /etc/nsswitch.conf OK? Where do the programs get hung if you use `strace' on them? Do you see any network traffic with `tcpdump -n -i ppp0' after you install the tcpdump package? Do you get any unusual messages when you start ppp with the `debug' option in /etc/ppp/peers/provider? -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: Kernel 2.2.0 on a 486DX2/66
Ralf G. R. Bergs [EMAIL PROTECTED] writes: This is strange -- I too had a mysterious lockup with the same symptom you describe, but with 2.2.0 (the release version.). I'm running a 5X86-133 (this is a 486 from AMD) on an ASUS PCI/I-SP3. OK, so it might not have been the one-off I thought it was. If it happens again I'll have to report it to linux-kernel. I _did_ use Alt-SysRq-P to find out that it seemed to be looping in schedule(), but didn't record anything else. This is with a Pentium 166 and a no-name motherboard. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: this guy talks to ff01a8c0 from loopback
[EMAIL PROTECTED] writes: /etc/init.d/network (stripped of comments) route add -host 127.0.0.1 lo ifconfig lo 127.0.0.1 netmask 255.0.0.0 ifconfig eth0 192.168.1.1 netmask 255.255.255.0 up route add -net 192.168.1.0 netmask 255.255.255.0 eth0 /usr/local/bin/socks5 -i -t You shouldn't actually need either of the route commands; that gets done automatically now when the interface is configured. (I was about to write varied on. Ugh - I really need this holiday.) I don't quite see where your original message would come from for this anyway. It's probably something for one of the kernel developers to explain. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: How do file transfer with minicom?
Blair Kelly [EMAIL PROTECTED] writes: Would a kind person give me a simple example or point me at documentation with same that describes how to do a file transfer using minicom? I am unable to get kermit to work, and I have no experience with zmodem. I have been trying various combinations of commands with the zmodem download program inside minicom, but I guess I just don't grok it ... It's been a while since I did this; probably with Slackware 3.1 and a few BBSes, before the Internet became popular (and cheap in NZ). First of all, you need to make sure you've installed the `lrzsz' package to enable ZMODEM. I think minicom should recognise ZMODEM downloads automatically, so from the remote end you would tell it to send the file; on a Unix host, type sz filename. If minicom doesn't start downloading when it sees the wierd characters, select ^A R (or Alt-R depending on configuration) and choose zmodem. To send to a Unix host, type rz to start the ZMODEM receiver then choose ^A S in minicom to send the file. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: LPRng - I give up!!
Chuck Stickelman [EMAIL PROTECTED] writes: All I wanted was to setup LPRng to use the nice, new remote printer server I bought (and the one my client bought...). I wanted to use magicfilter so we didn't have to worry about setting-up all of the applications to output in PCL (one can't...) Or have multiple queues for one printer - the client just doesn't seem to get the idea that they can have multiple printers to choose from even though they only have one *physical* printer... With the pretending-to-be-HP printer servers at work, I find lp=printer%9100 works quite well, where printer is the host name of the server. This should work quite well with the cti-ifhp package. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: Minicom installed and ppp working...at l o n g last
[EMAIL PROTECTED] (Larry Fletcher) writes: Do these window managers allow you to cutpaste between windows? I had planned on just using the Bash shell for awhile, but am finding that it is inconvenient not having the ability to cutpaste from one console to another the way I could with DOS and DESQview. It's actually the terminal emulator, i.e. xterm, that lets you cut-and-paste between the windows. On the console, you can use gpm to do this. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: Kernel 2.2.0 on a 486DX2/66
Philip Thiem [EMAIL PROTECTED] writes: ARe you using 2.2.0 with IDE? I found a message on dejanews listing severals bugs on include freezing and corruption of certain IDE drives. Actually, I'm now running 2.2.1 with IDE drives. I had one mysterious lockup with -pre8, where new programs didn't want to start, and with -pre6 (I think) logging funny messages, but aside from that I haven't had any problems with 2.2.x. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: Communicator Package 128 bit encription
Paul Miller [EMAIL PROTECTED] writes: I thought I read here once that there is a package to install to make the Comminucator package installed use 128-bit encription. Is this true, or do I have to stick with a tarball? The Fortify program does this; install the fortify-linux-x86 and fortify packages. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: this guy talks to ff01a8c0 from loopback
[EMAIL PROTECTED] writes: I have upgraded to the 2.2.0 kernel and am getting lots of these messages from the kernel: this guy talks to ff01a8c0 from loopback ff01a8c0 is 192.168.1.255. This message comes from the IP routing code in the kernel. I can't see exactly what conditions lead to you getting it, though. Routing has changed a lot from 2.0.x to 2.2.0. You should check what routes are being set up on what devices, etc., probably in /etc/init.d/network. Post the output of `/sbin/ifconfig -n' and `/sbin/route -n' and I might be able to suggest something else. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: Problem with perl and locale
Gregory Vandenbrouck [EMAIL PROTECTED] writes: I have made a setenv LANG=fr on my system. Which means some programs are displayed in French. It works fine but for perl which prints the following message: [snip] Try adding the country code as well as the language; i.e. LANG=fr_FR. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: Kernel 2.2.0 on a 486DX2/66
Richard Kaszeta [EMAIL PROTECTED] writes: [snip] The kernel builds fine, but when it boots, it stops at 'Starting kswapd v 1.5' and the machine hangs. It might be the next thing _after_ kswapd causing the problem. The next kernel message I get after this line is 'Detected PS/2 Mouse Port.' Do you have a PS/2 mouse (with a little round connector), and should you have it compiled in? -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: http proxy
Jeff Beley [EMAIL PROTECTED] writes: I recently setup an authenticating proxy using squid, however whenever i run dselect it tells me proxy authentication required, how do i set this? I alredy have http_proxy set correctly... I recently got this working for my brother with not too much hassle. Using apt-get, we set the http_proxy to 'http://login:[EMAIL PROTECTED]:80/' and away it went. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: smail. From field
Worik [EMAIL PROTECTED] writes: Please note that the return address in the from field is something like [EMAIL PROTECTED], and smtp.ihug.co.nz is where I send my mail. But I recieve it at [EMAIL PROTECTED] I would like to get [EMAIL PROTECTED] into the from field! How do I do it with smail? Would you consider upgrading to exim? The it's quite simple: two lines at the end of /etc/exim.conf: [EMAIL PROTECTED] ${lookup{$1}lsearch{/etc/email-addresses}\ {$value}fail} fF And a file '/etc/email-addresses' containing: worik: [EMAIL PROTECTED] -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: bash programming question
Michael Meskes [EMAIL PROTECTED] writes: [snip] However, this does not work if there are blanks in the filename as $file would be incomplete. I cannot simply use -exec for find either since I call a function from the same script inside the loop. Finally I need to read some input during this function, so simply piping the find results and reading them via read doesn't work either. If you don't expect any files to have newline characters, you could change the definition of $IFS to what normal find produces. e.g. variable=`find . -print` IFS=' '# No trailing spaces on the previous line. for file in $variable do somestuff $file done You could so some strange stuff with file descriptors to let you read from `find ... |' and stdin. Sorry, no example, I haven't had enough sleep this week to work it out. If you can use zsh, you could do: for file in **/*; do somestuff $file; done which will work even when some files contain a newline character. You could put the find in a separate script: find . -print0 | xargs -r0 somestuff.sh and in somestuff.sh: somestuff () { echo $1 } for file in $@ do somestuff $file done You could rewrite it in Python and use walk() from os.path, or in Perl and use File::Find. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ The risk of U.S. national security resting in the hands of adults who play with children's toys during office hours is left as an exercise to the reader. - Bruce Martin in RISKS
Re: HP 4000 N
Thomas Adams [EMAIL PROTECTED] writes: [snip] This is basically a hack, because I do not know if the LJ4M filter I chose is the best suitable for an HP4000N I'd probably go for the highest model number I could and see if it worked. Although I've found that getting TeX to produce HP4 output looks best (it prints darker than the HP5 setup) on our Lexmarks. and because doing things like ls -l|lpr generates a listing with a staircase effect. if= filters don't get run when you use rm=. Try using lp=192.168.10.20%9100 instead. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: hosts.allow - words of wisdom?
Alexander Kushnirenko [EMAIL PROTECTED] writes: I'm actually using the IP firewall code in Linux 2.2.0-pre5 to provide most of the protection to my system. My ipchains rules are as follows (actually saved in /etc/ipchains.save and read by ipchains-restore in /etc/init.d/network). Interesting, that's quite a new thought to me. I'm not a security expert at all of course. Do you have any web references or other relevant documents telling pro and cons of this technique, as opposed to TCP wrapper? There's the Firewall HOWTO, and the IP Chains HOWTO in /usr/doc/netbase/ipchains-HOWTO.txt.gz. As a summary: Pros: * In the kernel, so it should be faster. * Affects *everything*, including UDP (like the Network Time Protocol server and Samba name server), and even if the application doesn't use hosts.allow (like X11 and the DNS server). Cons: * More complex to configure. * Harder to tell whether it will work right. After thinking about it, I've actually changed my rules slightly, so that the _only_ incoming TCP connections permitted are on the ident port (for IRC and FTP servers), and on the ports from 1024 to 4999 from the ftp-data port, for FTP servers not in passive mode. Using the kernel firewalling code I *know* that a bad application won't leave my system open for abuse. I do have to experiment a little with UDP, to see what's necessary to permit Real Audio to work but keep out other packets. I could also block some kinds of ICMP traffic that I'm not interested in. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: uninstalling smail when a self compiled MTA is present
[EMAIL PROTECTED] (Marc Haber) writes: On 10 Jan 1999 17:51:05 +1300, you wrote: Maybe smail needs to be marked for removal. # echo smail deinstall | dpkg --set-selections This does not give an error message, but does not remove smail either. That just changes the status of smail. See `dpkg --status' for the effect, before and after. ('echo smail install | ...' changes back.) My idea was that maybe dpkg would deconfigure it when using -auto-deconfigure if it was marked for removal later (with `dpkg --remove --pending' or similar). Aha! Testing with `dpkg --no-act ...' shows this to be the case, at least for replacing my current exim with smail. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: uninstalling smail when a self compiled MTA is present
[EMAIL PROTECTED] (Marc Haber) writes: neither dpkg --install packagename --auto-deconfigure nor dpkg --install --auto-deconfigure packagename works. More ideas to get this command to work: Maybe smail needs to be marked for removal. # echo smail deinstall | dpkg --set-selections Maybe equivs needs to `Conflicts:' and `Replaces:' mail-transfer-agent, like exim, smail, sendmail etc. do. (Is anyone reading that actually *understands* dpkg?) -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: HP 4000 N
Jason Gunthorpe [EMAIL PROTECTED] writes: I use lprng not lpr, hplj4000n:HP Laserjet 4000 N:\ - :[EMAIL PROTECTED]:sd=/var/spool/lpd/hplj4000n:\ :[EMAIL PROTECTED]:\ - You can replace these two lines with the following, and it will work just as well, without the bounce queue stuff: :lp=hplj4000n%9100:sd=/var/spool/lpd/hplj4000n:\ The cti-ifhp filters can also help in this case to do real accounting using the printer's page counter, and set things like duplex printing, etc. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: named/BIND 8.1.2-5 won't accept inbound zone xfers
Ian Eure [EMAIL PROTECTED] writes: I'm having a really nasty time getting my BIND zone slave to update from the zone master. the slave keeps saying: --- Jan 9 13:24:21 Phaktory named[145]: Zone zone.domain (class 1) SOA serial# (10719991) rcvd from [xxx.xxx.xxx.xxx] is ours (121219981) --- (names/ips changed to protect he innocent) guilty? The serial number must increase for the zones to be updated. The usual method would make the serial numbers above something like 1999010701 and 1998121201, which means that the slave's old serial number would be less than the new one on the master. For example, using my ISP: $ dig -t soa clear.net.nz @dns1.clear.net.nz [...] ;; ANSWER SECTION: clear.net.nz. 22h47m53s IN SOA dns1.clear.net.nz. hostmaster.clear.net.nz. ( 1998122102 ; serial 2H ; refresh 10M ; retry 3D ; expiry 1D ); minimum [...] If they updated their records tomorrow, the serial number would change to 1999011101, and dns2.clear.net.nz would pick up the changes because 1999011101 (SOA record on dns1.clear.net.nz) 1998122102 (local SOA record on dns2.clear.net.nz). Anyway, the records on the master are *broken*. You should change the serial number to 1999011001 (which is greater than 121219981), or you'll have to try manually getting BIND to forget its cached zone files and get the whole lot from the master again. (And you'll have to do this every year. You *don't* need these problems in Jan 2000.) -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: Search and Replace
[EMAIL PROTECTED] (shaul) writes: Are there www sites from where I can download introductory level texts for these tools ? awk: info gawk (Install gawk package first) sed: man sed, under SEE ALSO Perl: http://language.perl.com/ Python: http://www.python.org/ or /usr/doc/python/html/tut/index.html -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: Printing on win95
Brant Wells [EMAIL PROTECTED] writes: I've got samba set up to where I can print to a shared printer... But the files have to be text files, and I have to do it manually from the smb: prompt :( WP8 will go through the motions, but will not print to the printer... I created a script for this, /usr/bin/netprint it looks like this: * #!/bin/sh smbclient dahouse\\hpdeskje shadowgate.com -U BW07442 -P -c 'print' Shouldn't that be 'print -'? Is shadowgate.com really the password? lp1¦Deskjet:\ ^ That should be an ASCII VERTICAL LINE |, not an ISO-8859-1 BROKEN BAR. Try the script from the command line, i.e. netprint printme.txt Once that works, try it with lpr: lpr -Plp1 printme.txt _Then_ work on WP8. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: hosts.allow - words of wisdom?
Alexander Kushnirenko [EMAIL PROTECTED] writes: [snip] 1. Does it have some striking errors? I bet I forgot some service... I prefer to put the denies in /etc/hosts.allow as well; something like: ALL : ALL : DENY at the end, to catch anything not explicitly allowed. I also prefer to use IP addresses instead of names, in case of spoofed DNS names and DNS servers being down, so, for example, I have a line: ALL : 127.0.0.1 192.168.117. : ALLOW 2. We run xntp3 to set time, but we don't want to be an NTP server. Do one need to allow some ntp service (NTP protocol is quite sophisticated :( The xntp3 access control configuration is documented in /usr/doc/xntp3/html/accopt.html in the xntp3-doc package. I'm actually using the IP firewall code in Linux 2.2.0-pre5 to provide most of the protection to my system. My ipchains rules are as follows (actually saved in /etc/ipchains.save and read by ipchains-restore in /etc/init.d/network). - # Don't accept any incoming packets not explicitly permitted. ipchains -P input DENY # Log any attempts at forwarding. ipchains -P forward DENY ipchains -A forward -j DENY -l # Doesn't matter who we talk _to_. ipchains -P output ALLOW # Keep input rules separate for tidiness. ipchains -N inet_in # Accept anything on loopback interfaces. ipchains -A input -i lo -j ACCEPT ipchains -A input -i dummy0 -j ACCEPT # Check input traffic on PPP links. ipchains -A input -i ppp+ -j inet_in # Accept and log ident lookup connections. ipchains -A inet_in -d 0.0.0.0/0 133:113 -p TCP -j ACCEPT -l -y # Only accept other TCP connections on ports FTP uses (and _not_ X). ipchains -A inet_in -d 0.0.0.0/0 ! 1024:4999 -p TCP -j REJECT -l -y # Accept NTP traffic with truechimer.waikato.ac.nz. ipchains -A inet_in -s 130.217.76.16 123 -d 0.0.0.0/0 123 -p UDP -j ACCEPT # Only accept other UDP traffic on non-privileged ports. ipchains -A inet_in -d 0.0.0.0/0 0:1023 -p UDP -j REJECT -l # Accept other traffic (including ICMP and existing TCP connections). ipchains -A inet_in -j ACCEPT - -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: [Fwd: Installing debian with Win98]
Gary Singleton [EMAIL PROTECTED] writes: OK, so what I did is modify Rob @ slashdot's excellent 'Linux - Don't Fear the Penguins.' artwork (http://slashdot.org/linux/index.shtml) resized/converted it to a 400x320x256 bitmap named logo.bmp. I posted it at http://www.geocities.com/SiliconValley/Pines/2361/logo.bmp if anybody wants to try it. Unfortunately that's the wrong size. :( MS went for a 320 wide by 400 high logo, with rectangular pixels. Also, Geocities transmits .bmp files as text/plain, so Netscape tries to display them. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: Hamm to Kernel 2.2.0
Ruud de Bruin [EMAIL PROTECTED] writes: a) It seems that the initailisation of the COM ports is changed. I saw a message during boot that something was wrong. Install the new version of setserial from slink or potato. b) I could not find a way to access my ZIP drive (parallel). It seems that ppa is not present anymore(?) This is dependent on how the kernel was compiled. You need at least CONFIG_PARPORT, CONFIG_PARPORT_PC, CONFIG_SCSI, CONFIG_BLK_DEV_SD, and CONFIG_SCSI_PPA (all as modules in my case), and you might need to configure the parallel port with the necessary parameters. In /etc/conf.modules, I have alias parport_lowlevel parport_pc options parport_pc io=0x378 irq=7 What changes should I made to a standard Hamm 2.03 system so I am able to use a 2.2.0(Pre) kernel? BTW, 2.2.0-pre5 was out last time I checked. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ $ uname -a Linux psyche 2.2.0-pre5 #1 Thu Jan 7 23:18:29 NZDT 1999 i586 unknown
Re: uninstalling smail when a self compiled MTA is present
[EMAIL PROTECTED] (Marc Haber) writes: Sounds like a bad brute-force approach. Can't dpkg do the remove and the install of the equivs package in a single operation? Yes, AFAIK. dpkg --auto-deconfigure --install new-mta.deb -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: [PERTAMINA/PIMPD/POSTMASTER: Mail failure]
J.H.M. Dassen [EMAIL PROTECTED] writes: Because the mail transfer system of a subscriber is broken (it should not report back to the From: address of a message with Priority: bulk or junk); Rather, it shouldn't ever bounce to the From: address. All bounces should go to the envelope sender, which for debian-user is [EMAIL PROTECTED]. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: need various libs ... ??
[EMAIL PROTECTED] writes: Comes up showing only libf2c.so.0 missing. I found the other libs ... was looking in the wrong directory. I did find a libf2c.so.2 file. Is this a later version of the libf2c.so.0 ??? Should I use that? Or is it something completely different?? I found a 1995 version which fits the bill, but its use would require that I overwrite a file out of my current libf2c.so.2 group. libf2c.so.0 will be an older version that's incompatible with libf2c.so.2. Try taking all the libf2c.so.0* files and putting them in /usr/local/lib then running ldconfig, and see whether 1. ldd finds them, and 2. the program doesn't crash when run. If it _does_ crash, the exact output and the output of ldd might help. I have obtained the source, but am leery about compiling it since I really am unfamiliar with that process. Do you know if the dependencies would remain, or would they change to the versions in my system?? If you install the -dev versions of all the libraries where necessary - probably just libc6-dev - it will depend on your current system, including f2c's current libf2c.so.2. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: Remote boot question
Eric Monson [EMAIL PROTECTED] writes: We are setting up a network with Linux and need some help. We are attempting to remote (network) boot the Win95 systems that we already have as clients for our Linux server. The problem is this: We have been looking into using ROM chips plugged into the network cards on the clients, but commercial solutions are too expensive. We could burn the ROMs ourselves, but we have no program to burn in. A web search (on free boot prom at google.com) turned up this page, which has a lot of useful information: http://www.slug.org.au/etherboot/ Is this the sort of thing you're after? -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: uninstalling smail when a self compiled MTA is present
[EMAIL PROTECTED] (Marc Haber) writes: On 07 Jan 1999 21:33:48 +1300, you wrote: [EMAIL PROTECTED] (Marc Haber) writes: Sounds like a bad brute-force approach. Can't dpkg do the remove and the install of the equivs package in a single operation? Yes, AFAIK. dpkg --auto-deconfigure --install new-mta.deb [Marc proves me wrong.] /usr/lib/dpkg/methods/disk/install says: dpkg --install --auto-deconfigure and uses: dpkg -iB Maybe it matters which order it goes in? -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: [PERTAMINA/PIMPD/POSTMASTER: Mail failure]
[EMAIL PROTECTED] (Marc Haber) writes: I doubt that Pertamina is aware of that - they don't even have postmaster. Had to look them up in the RIPE DB. Which RfC mandates bounces going to the envelope sender? RFC 1123 (aka STD 3), Host Requirements, section 5.3.3. In part: If there is a delivery failure after acceptance of a message, the receiver-SMTP MUST formulate and mail a notification message. This notification MUST be sent using a null () reverse path in the envelope; see Section 3.6 of RFC-821. The recipient of this notification SHOULD be the address from the envelope return path (or the Return-Path: line). So it's not actually required, just encouraged. Software written nearly ten years after the standard was published should be able to do this, though, and it's what mailing list software expects. It's also not using a null envelope sender in its message, which it MUST do. Elsewhere in this RFC is the requirement for a postmaster mailbox. I suppose I should wish them luck... -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: [OFF-TOPIC] qvirtual option in fetchmail
Richard Alhama [EMAIL PROTECTED] writes: I've solved it, fetchmail fetches but, it gets delivered to the ppp-outgoing queue. In ~alias/pppdir/ to be exact. Not on ~/Mailbox :-( You probably haven't put localhost in locals. If this doesn't work, post your rcpthosts, locals and the header from one of the messages after it gets into the maildirsmtp queue. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: How can I set Linux up as a File Server?
Brant Wells [EMAIL PROTECTED] writes: I've got a Linux box up and running, and have no problems with it (Thank God!). I have one more thing I need to do before I will be satisfied... How can I set Linux (Debian2.0) up as a file server? I have three PC's on the network... 1. WinNT4 2. Win98 3. Linux Assuming you've got them networked and TCP/IP working, you just need to install Samba. Otherwise, have a look at the docs in /usr/doc/HOWTO, like the NET-3-HOWTO. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: need various libs ... ??
[EMAIL PROTECTED] writes: Sounds like your program uses the old libc5 libraries that you don't have installed. Try installing the 'libc5' package in the 'oldlibs' section. Thought about that, but I have libc5 installed. The output of `ldd ./executable' would be useful. You could also try `objdump --private-headers executable | grep RPATH' to see if there's a path to the libraries compiled into the program. If there is, carefully change the path using something like Emacs's `hexl-find-file' to something that doesn't exist. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: smbmount-2.1.x: how does it work ?
Jérôme Zago [EMAIL PROTECTED] writes: smbclient //servif3/user4if -U jzago works just fine but I would like to mount this share to /mnt/servif3. Usage: smbmount-2.1.x service password [-p port] [-d debuglevel] [-l log] it seems like smbclient. where does I provide the mount point ? smbmount-2.1.x //servif3/user4if -U jzago -c 'mount /mnt/servif3' There's an example like this right at the top of the smbmount-2.1.x man page here. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: Q: dpkg-ftp and socks
Stephan Witoszynskyj [EMAIL PROTECTED] writes: I' would like to keep my system uptodate using dselect with ftp as installation source. The problem is that I have to use a Socks proxy. Is there a socksified version of dpkg-ftp? Or can anyone tell me on how to socksify dpkg-ftp? The socks4-clients package in unstable includes a runsocks script that lets you socksify most programs automatically, runsocks dselect update and runsocks apt-get install lynx for example. This is what I use at work with our SOCKS firewall. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: Problems with drive usage (Debian 1.3.1)
Torsten Landschoff [EMAIL PROTECTED] writes: - Warning: Unknown PCI device (104c:3d07). This is probably no problem - you just have a pci device which is not known to your ancient kernel. No problem. It's actually a Texas Instruments TVP4020 [Permedia 2], not that I actually know what *that* is. [snip] When I try to add the new partitions to the end of the unused space, or all of them as logical partitions, cfdisk installs all of them. But as soon as I try to initialize the new partitions, the system fails to work properly when I arrive at /dev/sda16 (is the entire device, not just one partition). This might indicate a problem with kernel 2.0.29 - I am not sure if it can address more than 16 partitions per disk. I don't have any SCSI disks, but it looks like that restriction still exists in 2.1.130. In fact, you're limited to 14 because /dev/sda is the entire disk and the extended partition (usually /dev/sda4) doesn't store anything itself. Initializing the partitions in the reversed order yields: Could not stat /dev/sda18 - no such file or directory although cfdisk shows the correct entry in the partition table. This is easy. mke2fs is right - there is probably no /dev/sda18. Just create it yourself: mknod /dev/sda16 b 8 16 mknod /dev/sda17 b 8 17 mknod /dev/sda18 b 8 18 I don't think so. 8,16 is /dev/sdb, 8,17 is /dev/sdb1, etc. Can you please post what your current disk looks like? For 5 DOS and 4 HPFS partitions you only have five partitions left for Linux, or four if you're using Boot Manager. You can probably get by with fewer partitions until you decide you don't need DOS, though, especially if you can use Dosemu and Wine to do everything DOS and Win3.1 do. I also think OS/2 would be able to cope with just one HPFS partition. You can use LOADLIN from DOS if LILO can't boot off an extended partition. So, I would get rid of /httpd and /adabas and put them under /usr, /usr/local, /var and/or /home. I would suggest swap, /, /var, /usr and /home, and at a pinch you could put /usr on the same partiton as /. Plan to back it up and reinstall later when you have a better idea of what size they should be. It might be possible to put the OS/2 and/or DOS partitions on /dev/sda16 etc., if they can cope. I'd back up everything, use OS/2 FDISK to create all the partitions, with the Linux ones as DOS instead, then restore OS/2. Then install Linux and change the DOS partitions to Linux. This is a lot of trouble, especially if it doesn't work. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: slink
Kenneth Scharf [EMAIL PROTECTED] writes: From time to time I have been directed to use a package from slink because the corresponding version from hamm is 'broken', an example is xfstt. However I can't install *ANYTHING* from slink because I get a dependacy error due to lib6c being updated in slink (hamm has an older version). This was actually a bug in slink's libc6. Some packages in slink and/or potato have been recompiled with a new libc6, which doesn't put that restriction in. Can I get the source packages from slink and rebuild under hamm using the existing lib6c lib's in the interm? Unless they also depend on other libraries only in slink, you shouldn't have any problems. In case you or other readers don't know how to use a source package: 1. Download the .dsc, .orig.tar.gz and .diff.gz files into one directory and run dpkg-source -x whatever.dsc. 2. cd into the resulting directory and run debian/rules binary. 3. dpkg -i the resulting .deb package. *Don't* force it if it refuses to install - reinstall the old package and ask for help. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: Network File Server
Austanners Wet Blue Pty Ltd [EMAIL PROTECTED] writes: We are just about to go ahead and upgrade our network. All machines are Wx89 workstations. We were going to use NT Server as a file server but since using Debian i thought about Linux as a file server. I'm sure lots of you out there are doing this so any suggestions (hardware, software packages) and comments (problems etc) would be a appreciated. I was going to get something like a 686 300MMX 128MB RAM 32X CDROM UPS etc. That's probably over-specced even for an NT file server. ;-) Besides what you've listed, you should go for an all-SCSI system, not IDE, and you'll need some kind of tape backup. You should make sure you get a 10/100 Mbs Ethernet card, even if you only have a 10 Mbs network at the moment. For software, you'll want to use Samba as the file server. You can also use this as a print server, and together with LPRng and cti-ifhp it works quite well with HP-compatible laser printers. (Would you have gotten a reply this fast from MS support on a statutory holiday?) -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: Modem trouble, Please help!!
BOHICA [EMAIL PROTECTED] writes: The final resolution was to install the 2.1 base disks with the 2.0 rescue and driver floppies, download the latest kernel and use a linux/bash emulator on my NT desktop to build a custom compilation of the kernel that included the needed Xircom support. Is this mentioned anywhere in the documentation? It seems that it's possible to compile vmlinuz under WinNT or Win9x using CygWin, although 2.0.x won't work properly because of the version of egcs they use. This could help people with strange combinations of hardware that can't get any of the supplied kernel images to work. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: LaTeX and overhead production
Simon Read [EMAIL PROTECTED] writes: I'd like to use LaTeX to produce the overhead slides for my adjunct class. I seem to have enormous problems all along the path, can anyone help me? Have a look at http://home.clear.net.nz/pages/c.evans/scriptintro.html which has a link to LaTeX source for a few slides. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: Xdm on remote servers.
Johan Berglund [EMAIL PROTECTED] writes: I have xdm installed on my server and have configured it to not start a local xserver. I also have a win98 computer with eXceed configured for XDMCP broadcast. I get the login window and it is able to authenticate me as a user and then the login window shows up again and it only loops. You might need to write your own .xsession that _doesn't_ start a window manager. Try: $ cd $ mv .xsession .xsession-old [Might get an error if you don't have one already.] $ cat .xsession #!/bin/sh exec xterm ^D [That's Control-D.] $ chmod +x .xsession And see what happens then. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: Changing(forcing) Alt-Fn to Ctrl-Alt-Fn
[EMAIL PROTECTED] writes: When using the VT's I'd like to block Alt-Fn from being interpreted as a change to antother VT so that the key sequence would be available to the applications. How would that be done? Have a look at /etc/kbd/default.map, and the files under /usr/share/keymaps and /usr/doc/kbd. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: modprobe: Cannot locate module char-major-10
Jeroen N. Witmond [EMAIL PROTECTED] writes: I have searched the mailing list archives for this problem, and I understand it can be solved by aliasing module char-major-10 to off, but I would like to understand what is going on. (It is not really a problem, because in spite of the message everything seems to work.) [snip] In human language: The Debian 2.0.34 configured the serial support as a module, whereas I included it into the 2.0.36 kernel itself. It seems that, even when the serial support is included into the kernel, 'somebody' still wants to access it as a module, but I can't find who, where and why. Serial ports are on char-major-4; see ls -l /dev/ttyS0: crw-rw 1 root dialout4, 64 Feb 10 1998 /dev/ttyS0 major ^ ^^ minor char-major-10 is misc devices such as psaux, watchdog, apm and nvram - cat /proc/misc should list the ones you're using. Presumably something is trying to access one of the misc devices. You might be able to get some idea of what's happening with: $ cd /dev $ ls -ltur [...] crw-rw 1 uucp dialout4, 65 Jan 3 10:32 ttyS1 crw--- 1 root sys 10, 1 Jan 3 11:53 psaux prw-r--r-- 1 root root0 Jan 3 11:55 xconsole| [...] crw-rw-rw- 1 root root 2, 1 Jan 3 11:58 ptyp1 crw--w 1 careytty3, 1 Jan 3 11:58 ttyp1 crw--w--w- 1 root root 4, 7 Jan 3 11:58 tty7 $ -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Larry froze. Was the bag a trap? He could see the way in, but the other end appeared to be sealed.
Re: Cleaning my /usr/lost+found
[EMAIL PROTECTED] writes: Well it's just gone from a weird, bothersome pain to a real problem. I was upgrading some packages when I got some familiar error messages about some file operations - operation not allowed or whatever... Anyway, upon closer inspection I know have this: pr---w-rwt 1 1202946260 Nov 11 1926 /usr/include/lct/unicode.h srS--t 1 55203159901947437670 Dec 17 1909 /usr/lib/picon/usenix/com/hp/hpda/aca So someone ***PLEASE*** tell me how I can force these files to go away. To get rid of the errors I got, I booted from the rescue disk and used debugfs to manually rm the directory entries, then ran fsck again to clean up the resulting mess. If you don't have the backups that would let you reformat and restore, you should start making them or you'll regret it as much as I did when my drive went belly-up a year ago. (Great Christmas present, that.) Note that each of the bad files in lost+found is one (possibly essential) one that has disappeared from where it was supposed to be. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Microsoft is the answer! The question is, Why did my PC crash?
Re: xserver-svga 3.3.2.3 messes virtual terminals
Jean Orloff [EMAIL PROTECTED] writes: 1) what more can I try (outside reboot) to clear the terminals? setfont -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Microsoft is the answer! The question is, Why did my PC crash?
Re: Mouse pointer
Phillip Neumann [EMAIL PROTECTED] writes: # Nifty root cursor xsetroot -cursor $HOME/.rootcursor $HOME/.rootcursor All right.. that worked great... thanks... I have play with bitmap, and im only able to do black points for the cursor... I can see the default `X' cursor, and it has a white thing around the black X. The second $HOME/.rootcursor is actually the background. If you copy it to another file, you can edit that to add the outline, then do xsetroot -cursor $HOME/.rootcursor $HOME/.rootmask How can i paint whites points? Or even better, what about color ones? You _can_ change the foreground and background colours - see the xsetroot man page for details. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: firewall question
Giuseppe Sacco [EMAIL PROTECTED] writes: Weel, we have all our data available in a web server, so we'd like to browse the data in the same way. Maybe we can have something like: CLIENTFW SERVER browser authenticator-daemon firewallweb server when the user outside the firewall ask to connect via https:// then the firewall can ask the client to autenticate himself. HTTPS supports client certificates, which you can use to authenticate the user. A good place to start looking for more information is URL:http://www.verisign.com/. You should be able to configure something with ipportfw or redir to just send the HTTPS connections to the firewall to the web server, and _make sure the web server is secure_, e.g. it only allows connections from listed client certificates. You could also put the web server on two IP addresses, and have virtual servers so that one is used by internal LAN access, and the other is used by forwarded connections from the firewall. That way you can configure each virtual server with different security requirements, e.g. basic authentication for external users if the data isn't very sensitive. See the web server manual for more details. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: emacs xemacs
Kent West [EMAIL PROTECTED] writes: However, last night while running lynx I clicked on a .gif link and low-and-behold, the image appeared! So apparently lynx can handle some types of graphics. All this time when I was asking for a graphical console-based web browser because lynx can't handle graphics, lynx CAN handle (at least some) graphics. That was probably zgv. You can use it from the command line too, e.g. $ zgv b5.jpg You can make Lynx generate links for inline images by pressing *, or turn this on all the time by editing /etc/lynx.cfg and looking for the line #MAKE_LINKS_FOR_ALL_IMAGES:FALSE. You'll still only get a text view of the web page, but you can load each picture separately in zgv. BTW, if you're using Lynx, I really recommend you try upgrading to the latest version, at least 2.8.1. This gives you progressive display of pages, which is *really* good for looking at Slashdot. (But make sure you still have a copy of the old .deb around in case it doesn't work.) -- Carey Evans http://home.clear.net.nz/pages/c.evans/ (I wish I knew why I'm at home on Saturday night, writing brick text.)
Re: auth port security
Pere Camps [EMAIL PROTECTED] writes: Does anybody know if the auth port has any know security compromises? A user has asked me to turn it on (I have almost everything off) and I'm not sure. It will release the names of your users to anyone who is able to connect to it, unless you install a dummy one which claims to know nothing. You should probably ask if your user really needs to use IRC. (This is the only case I can think of when it's _needed_, and other networks than Undernet are usually less strict anyway.) -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: elm email gets through, exmh email doesn't
Mark Phillips [EMAIL PROTECTED] writes: mark at ist.flinders.edu.au: loses; [USER] 550 relaying to [EMAIL PROTECTED] prohibited by administrator post: 1 addressee undeliverable send: message not delivered to anyone What on earth does this mean and why isn't it working??? Try editing /etc/exim.conf and adding ist.flinders.edu.au to local_domains. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: recommendation on presentation software
[EMAIL PROTECTED] writes: i need to make a presentation, and i'd like to do it with my laptop a la Powerpoint. (ie video output of laptop plugged on LCD retroprojector) [snip] I need very basic capacities, as i won't have much time to create the contents of the presentation either. When I needed something like this for a presentation at the local Linux Users' Group I used pdflatex and acroread. You can see my work at URL:http://home.clear.net.nz/pages/c.evans/doc/scriptintro.html. Some of the debian-user readers might also find the content of the presentation interesting. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: how to create locales? [OFF TOPIC]
Eugene Sevinian [EMAIL PROTECTED] writes: I am planing to do some job concering localisation of linux for my native language (Armenian). I would like to know which steps should I do for it. One thing which I need to know is how to create locale databse. Look at `man 5 locale', which is in the manpages package. You might also want to have a look at /usr/doc/gettext/ABOUT-NLS from the gettext package. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: Set-up My Apple Laser Writer on Debian Linux
Bill Miller [EMAIL PROTECTED] writes: I'm trying to setup my Apple Laser Writer printer on my system, but have been unsuccessful in getting it to work. Using the command ls /etc /dev/lp, I'm able to get a light on the printer to blink on and off, but no print follows. [snip] Install the lprng and magicfilter packages. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: surprise! was( USR; dialing wrong number)
Eugene Sevinian [EMAIL PROTECTED] writes: It seems that this modem is very special thing. Might be it was made at April 1. The matter is that during dialing from minicom it performs kind of maping. So 1-9, 2-8, 3-7...9-1, 0-0. The same is under Norton Terminal Emulation. :))) Are you using Pulse dialing? The American standard for Pulse dial phones is different than the New Zealand standard, at least. (I just tried and the digits map like your example above, i.e. ATDP1494050 dials my ISP at 9616050.) I thought tone dial was the same everywhere, but this may be too optimistic. Anyway, the simple solution would be to rewrite the phone number you're calling according to these rules. Maybe your Win95 setup does this automatically - try dialing out manually from HyperTerminal. If this works, display the settings with ATI4 and see if they're the same under Linux. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: w3-el, Howto?
Kent West [EMAIL PROTECTED] writes: 1) Is this what I've been wanting: a graphical web browser that does not require X? Or have I misunderstood its capabilities? Emacs is a HUGE editor, that runs either in text mode (console or xterm) or X11 mode. In X11 mode, W3 runs graphically. In either case, it tries to do frames. (The same applies to XEmacs.) 2) How do I get started with it? Alt-X W 3 Enter Better would be control-h i, and read the info documentation. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: scriptl interacts with telnet port daemon?
zhaoway [EMAIL PROTECTED] writes: I want to do a bot in a mud 10.74.i.forget 1999, want a perl script to do this. Is this the correct thing to do to play with following script and interacts with telnet 10.74.xx.xx 1999? i.e. open (mud, 'telnet localhost 1999') You should probably connect directly from Perl instead of trying to use the telnet program. See `man perlipc'. However, the MUD probably uses Telnet command codes, so something like Net::Telnet would be better. The one that comes with `libnet-perl' isn't the right one - try looking at: http://www.perl.com/CPAN/modules/by-module/Net/Net-Telnet-3.01.tar.gz very few programming experience ;-P I don't think this is the right thing to start with. :-/ -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: Recommendations for Email client?
[EMAIL PROTECTED] (Gary L. Hennigan) writes: [snip] I don't want to get pegged as an X/Emacs fanatic (which I am), but gnus, in conjunction with fetchmail (and maybe even without it) can do all the aforementioned operations. I'll second this. I'd also like to mention the `nnimap' backend that's available for Gnus, that lets you use IMAP folders like they're supposed to be used. It's only slightly beta. The code can be downloaded from: http://www.extundo.com/nnimap/ Un-tar this somewhere (like ~/share/emacs/nnimap/), compile it, and add the following code to ~/.emacs: (setq load-path (cons (directory-file-name (expand-file-name ~/share/emacs/nnimap)) load-path)) (require 'nnimap) Would anyone be interested in this as a Debian package? -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: Re : Re: Re : RE: please help me: problems for users!
[EMAIL PROTECTED] (Patrick RICHARD) writes: Hi, The files /etc/group and etc/passwd are OK. I use the shadow mode (password field = x in the file /etc/passwd). OK then, what are the permissions of the files? Mine are: $ ls -l /etc/passwd /etc/shadow /etc/group -rw-r--r-- 1 root root 577 Nov 26 22:59 /etc/group -rw-r--r-- 1 root root 1434 Aug 9 22:50 /etc/passwd -rw-r- 1 root shadow957 Jan 25 1998 /etc/shadow If this doesn't help, try something like: $ strace -o vi.log -e file vi temp and quit after getting the error? This will list the errors (look for E? symbols in the output) that vi gets accessing the files. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: backup
Richard Black [EMAIL PROTECTED] writes: I have recently converted from nt to linux and have some tapes, made by ntbackup, that I would like to restore under linux. Any help about how I could do this would be much appreciated! The best that I can think of is via wine, but I haven't looked into this yet. I don't think Wine is the right solution. You need to get a/the tape drive recognised and configured under Linux. Then you can see whether the data is actually available. What brand of tape drive is it, and how does it connect? Have a look at some of the HOWTO's in /usr/doc/HOWTO too. If you *have* got the tape working, but just can't get at the tapes, you're beyond my knowledge. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: How to fix Kernel panic
TONY ANG [EMAIL PROTECTED] writes: 1. [ ] text/plain [*] text/html (Please post just in plain text.) Now,I cannot boot from HDD as I did normally. I tried booting from F.Disk but I have this Msg Kernel panic:VFS:Unable to mount root fs on 03:06 Can you please advice me what to fix my problem by : 1. Let me know how this happen and If you're using LILO to boot (quite likely) its parameters have become screwed up. Or worse, your hard drive has. :-( 03:06 is /dev/hda6 (look in /usr/src/linux/Documentation/devices.txt once it's working). The kernel is complaining because it can't find a Linux (ext2) file system on /dev/hda6. 2: How shall I fix this problem If you know which partition the root Linux system is on, you can enter different parameters to boot from a different drive: linux root=/dev/hda1 If not, use your rescue disk and switch to Alt-F2 after it boots but before installing anything. Here you can use fdisk etc. to find out which partition is which, and mount them to see what's on them. Once you get it to boot, edit /etc/lilo.conf to work right, run /sbin/lilo again, and reboot to make sure it worked. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: Virtual Console Messed Up When Switching From X
[EMAIL PROTECTED] (Ruchira Datta) writes: Sometimes when I switch from X to a virtual console with Ctrl-Alt-Fn, the virtual console starts substituting some characters for others. Try the command `setfont'. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: Recommandation for LTT
[EMAIL PROTECTED] writes: But I'm worrying about robots collecting emails. I really don't want that. [snip] Does anyone have other suggestions ? A simple idea I saw is to munge characters in the address as HTML- and URL-encoded text. I don't actually know how effective this would be, but hopefully not many email-collecting robots will decode the following properly: A HREF=mailto:c%2Eevans%40clear.net.n%7A;c.e#118;ans#64;clear#46;net.nz/A (Note especially that there are no @ characters for them to spot.) -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: sendmail check_mail ruleset
[EMAIL PROTECTED] (Ole J. Tetlie) writes: ...or even better; can someone help me fix my headers? I'm using exim and gnus. Since I'm not sure which details are pertinent, I'll wait with config files and such until requested. Please help, I don't want to wreak havoc on the net! :-) Do you get the same if you send the message with the `mail' command? e.g. mail [EMAIL PROTECTED] What does /var/log/exim/mainlog say about a single message sent from Gnus? Have you made any changes to the file generated by eximconfig? There are two places I think Sender: header could be coming from. Gnus might add one, and I can't see how it would ever not have an @ symbol. What do you get if you put your cursor after the following line and type `C-x C-e'? (concat (user-login-name) @ (system-name)) What does the Sender: header in any archived mail look like, if you're using the Gnus message archiving function? Gnus, by default, uses the -f flag on its call to /usr/sbin/sendmail, and exim might include a Sender: header because of this. You can disable this from Customize, or with the second line below, and often get better results. (customize-variable-other-window 'message-sendmail-f-is-evil) (setq message-sendmail-f-is-evil t) If all else fails, you could add the line below to your remote_smtp transport in /etc/exim.conf and see what happens. headers_remove = sender -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: sendmail check_mail ruleset
Oliver Elphick olly@lfix.co.uk writes: Messages from a Debian developer are messing up my combination of sendmail and fetchmail. [snip] I would like to get fetchmail to bounce rejected messages properly and continue, or else stop sendmail rejecting them. I cannot see how to do either. It looks like fetchmail is trying to use the Sender: header as the MAIL FROM: address, but is rewriting it. Try putting norewrite in your .fetchmailrc. Can anyone tell me how, please? Versions: sendmail8.9.1-9 fetchmail 4.3.9-1 That's an old version of fetchmail. Make sure you have your current version as a .deb somewhere, and download the latest version from unstable. *Don't* install the libc6 it wants if dpkg -i won't install it - it should be safe (with 4.6.6-2) to install it with dpkg -i --force-depends fetchmail_4.6.6-2.deb, since it's not an essential part of the system, and you can reinstall the older version if it breaks. ;-) (Anyone else reading: bash and libc6 are essential parts of the system. Don't use --force-depends unless you really know you won't ruin your week by doing so.) In any case, run fetchmail --verbose to get a better idea about what's happening. (I'm not having any problems with Ole's messages because my ISP inserts the Return-Path: header, which fetchmail uses first, and which is the *real* MAIL FROM address.) -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: How to customize fvwm entries ?
[EMAIL PROTECTED] (shaul) writes: It seems to me that you are wrong, since when new packages are installed, update-menus considers the users who run update-menus as well. I am saying that on the basis of the following lines (from /usr/doc/menu/html/ch5.html): [snip] Am I right ? One of us should just try it and see. But the reason I don't think it will update individual users' menus is that when I run it, all the menus get put in ~/.fvwm2/whatever. When root runs it, the menus go into /etc/X11/fvwm2/whatever. If something changes because a new package is installed, the files in /etc/X11/fvwm2/ will be changed, but the old files in ~/.fvwm2/ will still contain the old data, and will override the new data, until that user runs update-menus again. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: virtual hosts
Chris Evans [EMAIL PROTECTED] writes: The ISP suggests mapping the various different host names to the different IP addresses all on the same machine. The tech man there tells me that gets round the problem of older browsers coming in to the numeric address and so hitting the www root rather than the virtual mappings. Sounds good to me and I assume you just have multiple entries in /etc/hosts. No, it's a little more complicated. You need a kernel that supports IP aliasing, then you have to configure the extra addresses as separate interfaces, e.g. ifconfig eth0 192.168.20.1 netmask 255.255.255.0 ifconfig eth0:1 192.168.20.2 netmask 255.255.255.0 This is really a waste of IP addresses though, and almost all browsers will send the required headers so that you don't need to do this. Can I do something similar with Email? In particular with Smartlist? Is there something I can do using sendmail.cf and perhaps MX records to allow the same box to collect for different names and allow procmail, sendmail, smartlist, qpopper and probably some IMAP server to treat the different names appropriately? It's quite common to set up email this way, and should be easy. I don't know how to actually do it with anything besides qmail, though. You don't need different IP addresses - ALL programs sending email to you will include the machine name. Have I got the right idea? What friendly documentation should I be reading? All of it. :-) But start with the friendly Apache documentation. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: Blocking an arbitrary port with ipfwadm
Damon Muller [EMAIL PROTECTED] writes: What I want to do is use ipfwadm to block a single port - namely 31337 (UDP). For anyone who has had their head in the sand for the last few months, that's the port that Back Orrifice listens on by default. ipfwadm -I -a reject -P udp -D 192.168.20.0/24 31337 -o That's: -I = check incoming packets. -a reject = Let the sender know we're rejecting them. You can use -a deny instead to drop them silently. -P udp = UDP packets only. -D 192.168.20.0/24 31337 = destination anywhere on your network (fix this) on port 31337. Maybe 0.0.0.0/0 31337 to stop *your* users connecting to another BO server. -o = Log the address of the offender, so you can complain to their ISP. This will also drop some legitimate UDP comms, if something happens to grab port 31337 itself. I could live with this. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: Problems with Exim with ISP-dialupping
Vesa Kaihlavirta [EMAIL PROTECTED] writes: I'm using an ISP to connect to the net. Exim's autoconfiguration makes a working configuration, except that I cannot mail @kolumbus.fi. Those addresses are considered local, and so if I try to mail some @kolumbus.fi, I get a return receipt: [snip] I'd like @kolumbus.fi understood as a mailname that resides in the net, and @localhost as the _real_ local mailname. As I understand it, you can make this happen by entering _just_ localhost in local_domains, and by commenting out local_domains_include_host as well if your hostname looks like kolumbus.fi to exim. You can also look at the rewriting rules near the bottom, which might be useful. I have the following lines to rewrite local (bogus) addresses. --- /etc/exim.conf --- [EMAIL PROTECTED] ${lookup{$1}lsearch{/etc/mailfrom}{$value}fail} fF -- --- /etc/mailfrom --- carey: [EMAIL PROTECTED] -- -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: ethernet card and ppp modem connection
Colin Telmer [EMAIL PROTECTED] writes: What I ultimately would like to do is to be able to do rlogins through this ppp connection but have all other network connections still using my ethernet connection. I guess I don't completely understand what you mean by Put a script in /etc/ppp/ip-up.d to add a route to your office network via the ppp interface? Suppose your office uses the addresses 192.168.20.0 to 129.168.20.255. Then you want to add a route to 192.168.20.0/255.255.255.0 in /etc/ppp/ip-up.d/local-route, and delete it in a corresponding script in /etc/ppp/ip-down.d. e.g. --- /etc/ppp/ip-up.d/local-route --- #!/bin/sh route add -net 192.168.20.0 netmask 255.255.255.0 gw $PPP_REMOTE -- This route is more specific that the default route, so it will be used first. In ip-down.d, put the same thing but with route del instead. This should work, since it's almost exactly what I use. Could you point me to documentation that I can use to figure this out? Any help is really appreciated. Cheers, Colin. Read and understand all the networking HOWTO's in /usr/doc/HOWTO. I believe it should be possible using fancy firewall stuff, or the new routing in the 2.1.x kernels, to send only rlogin traffic over the PPP interface, but you'll have to work out how to do that yourself. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: How to customize fvwm entries ?
[EMAIL PROTECTED] (shaul) writes: Thank you. How a user can set it for himself ? (It seems to me that menu offers a way to do it, only I couldn't understand how) He can put it in ~/.menu/lyx and run update-menus himself, which will create all the files for just that account. However, I don't think anything will get changed automatically for that user when new packages are installled. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: HELP: best place to place binfmt_misc directives?
Bruno Boettcher [EMAIL PROTECTED] writes: wanted to set up again some of the misc binary modules capabilities on slackware init i had a rc.local where to put such stuff where do i put it in debian? Assuming you're talking about binfmt-misc, create a file in /etc/init.d called something like local-binfmt.sh, and link to it by # ln -s ../init.d/local-binfmt.sh /etc/rcS.d/S90local-binfmt.sh The file would contain something like: - if [ -e /proc/sys/fs/binfmt_misc/register ]; then echo ':rexx:M::/*::/usr/bin/regina:' /proc/sys/fs/binfmt_misc/register fi - BTW wouldn't it be neat if the packages beeing susceptible to propose such services had a switch to set this functionality on when installing the package? You could file a Priority: wishlist bug on the package you think should have this functionality, suggesting they do something like I put above. Ultimately, there should probably be another package that manages this and provides an `update-binfmt' program. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: How to customize fvwm entries ?
[EMAIL PROTECTED] (shaul) writes: My /etc/X11/fvwm2/menudefs.hook has the following entry for lyx: [23:36:37 shaul]$ grep lyx /etc/X11/fvwm2/menudefs.hook + lyx Exec/usr/X11R6/bin/lyx [23:38:51 shaul] I want to change it to + lyx Exec /usr/X11R6/bin/lyx -width 631 -height 461 To change it for all users: $ su Password: # cp /usr/lib/menu/lyx /etc/menu/ # edit /etc/menu/lyx [ Make the change. ] # update-menus # ^D $ -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: procmail....................
Phillip Neumann [EMAIL PROTECTED] writes: [snip] mmm, i just want to make me sure what a colon is.... or , ?? $ ascii colon ASCII 3/10 is decimal 058, hex 3a, octal 072, bits 00111010: prints as `:' Official name: Colon Other names: Double-Dot [snip] ok, now how do i put procmail in the deliver proces?? See `man procmail' near the bottom. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Beginner - Ethernet configuration was: Re: HELP
Alan Mayott Jr [EMAIL PROTECTED] writes: Hello My name is Alan and I need HELP... I am trying to set up a simple network between Win95 and Linux... The problem seems to be with Linux as I've done this in win95... I've checked with the manufacturer about my ethernet card and from what I can tell it takes the tulip.0 driver in Linux, Well during install I could not get it recognized in linux so I went ahead and done the install without it... My question is can someone help explain some of this linux stuff to me... A lot of this sort of stuff is in the Ethernet HOWTO, /usr/doc/Ethernet-HOWTO.gz on your system or URL:http://sunsite.unc.edu/mdw/HOWTO/Ethernet-HOWTO.html. Have a look at this and see if you can get started. A Debian-specific note: You can set up the modules semi-automatically, like you got a chance to do a boot time, with the program modconf. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: How to start program in cron script?
[EMAIL PROTECTED] (Andy Spiegl) writes: 0 8 * * * /usr/local/bin/myprogram 21 | /usr/bin/logger -i -t myprogram /dev/null 21 That's a nice alternative, but I found that logger doesn't quit after myprogram is done. What am I overlooking now? The only thing that comes to mind is that myprogram starts something, and logger won't exit until there's nothing left writing to _it_. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: networking problem with 2.1.x
Ben Jorgensen [EMAIL PROTECTED] writes: When booting with 2.1.128 I can't remove the net entry in the routing table. route del some net address I get SIOCDERT: no such process What's wrong? With 2.0.35 it works fine! 2.1.x drastically changed the way routing works. Usually it looks the same with the route command, but sometimes (like this time) it doesn't. See Documentation/networking/routing.txt and other files in the same directory for the gory details. `iproute' is a Debian package, and `rtmon' is in the same package. Why do you want to remove the net entry? Every interface is usually attached to a network of some kind. If it's not, try configuring it as netmask 255.255.255.255. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: Downloading the Linux file from ftp -Help!
Alan Tam [EMAIL PROTECTED] writes: [ ] text/plain [*] text/html Hi David, Use your browser (Netscape) to download it and save the page as *.txt. But normally Linux is an image file, if you save it as a text file you can't boot your system in text format. Alan David B. Teague wrote: Actually, Stephen wrote this, but because Stephen's web browser pretending to be an email program put an HTML alternative, David's program included the same HTML, and your (Alan's) web browser only saw the HTML, not the very helpful text David wrote. Can anyone can give me some advice on how to get the kernel (linux) file from a ftp site with downloading it as a .txt file?ThanksStephen LavelleAustanners Wet Blue Pty Ltd. Stephen and Alan: Disable HTML posting in your mail programs. Stephen: Try to turn on automatic line-wrapping to 70 columns, or whatever it's called in your program. David: Complain to someone about Pine leaving the HTML alternative in. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: Problem starting X for first time
Kent West [EMAIL PROTECTED] writes: [snip] But to answer your question: I've found that you can generally do a Shift-PgUp to recall the last 2 or 3 screens of info. It's a really handy feature. But it stops working when X starts, because the VGA memory the scrollback buffer uses gets used by X. (This is a gross oversimplification.) Try something like: $ startx X.log 21 If you want to see the messages go flying by, that would be: $ startx 21 | tee -a X.log Another handy program is SuperProbe, to make sure your idea of what the card is matches X's idea. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: SQL Database performance
Jeff Noxon [EMAIL PROTECTED] writes: Postgresql runs like greased lightning. In some tests involving 0.5 million records of a few K in size, it blew away MS SQL Server. I think it was version 6. Postgresql runs well, although it does have some limitations. ODBC was the main weakness I noticed. We're currently deploying Oracle on a Debian server... Oracle is bloatware even by Microsoft standards. I can't comment on speed yet. Oracle's CEO, Larry Ellison, claims that if anyone can show that SQL Server is less than 100 times slower than Oracle, he'll give them a million dollars. See URL:http://slashdot.org/articles/98/11/18/0837247.shtml. I'm also planning on sticking the first page of the PDF at URL:http://www2.software.ibm.com/news/news.nsf/n/cjig425nf2 up at the office somewhere. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: ftp daemon
Robert Rati [EMAIL PROTECTED] writes: I upgraded the proftp package, and now when I boot, I get this error messaage: module: linuxprivs: kernel does not support capabilities, disabling. What capability is it talking about and when do I need to inlcude in the kernel? Tia. capabilities is a new security feature in the development (2.1.x) Linux kernels. proftpd is probably trying to use these for extra security, but will be the same as before if you're running 2.0.35. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird
Re: ftp daemon
Lawrence Walton [EMAIL PROTECTED] writes: Oh? I am running two FTP servers with the 2.1.12* kernels... what is the security feature? [snip] auth.log:Nov 16 12:53:28 ns proftpd[29299]: module linuxprivs: capabilities '= cap_net_bind_service+ep' I'm not especially familiar with what's involved. In this case, I would guess it lets proftpd use reserved ports (like 20 and 21, ftp-data and ftp-control) without needing to be root. This means that if there's any bug somewhere in proftpd, there's no straightforward way to get root access to the box. -- Carey Evans http://home.clear.net.nz/pages/c.evans/ Is there anyone who actually believes that USAicans are so modest or intellectually honest as to be unable to find someone to sue? - Cameron Laird