[LinuxISP-BR] FreeBSD Security Advisory: SA-96:21 - talkd (fwd)
Just got this cert advisory... it's for talkd, and although from the FreeBSD advisory, applies to most talkd's... Will debian fix its talkd so we can reinstall a package? Ricardo -- Forwarded message -- Date: Tue, 21 Jan 1997 19:00:05 -0200 From: [EMAIL PROTECTED] Subject: [LinuxISP-BR] FreeBSD Security Advisory: SA-96:21 - talkd -- Forwarded message -- Date: Sat, 18 Jan 1997 10:38:57 -0600 From: FreeBSD Security Officer <[EMAIL PROTECTED]> To: [EMAIL PROTECTED] Subject: BoS: FreeBSD Security Advisory: SA-96:21 - talkd Resent-Date: Tue, 21 Jan 1997 21:04:26 +1100 (EST) Resent-From: [EMAIL PROTECTED] -BEGIN PGP SIGNED MESSAGE- = FreeBSD-SA-96:21Security Advisory FreeBSD, Inc. Topic: unauthorized access via buffer overrun in talkd Category: core Module: talkd Announced: 1997-01-18 Affects:1.0, 1.1, 2.1.0, 2.1.5, 2.1.6, 2.1.6.1 Corrected: 2.2-current as of 1997-01-18 2.1-stable as of 1197-01-18 FreeBSD only: no Patches:ftp://freebsd.org/pub/CERT/patches/SA-96:21/ References: AUSCERT AA-97.01 (Australian CERT organization), SEI CERT VU#5942 (internal tracking reference only) = I. Background Buffer overrun (aka stack overflow) exploits in system supplied and locally installed utilities are commonly used by individuals wishing to obtain unauthorized access to computer systems. The FreeBSD team has been reviewing and fixing the source code pool to eliminate potential exploits based on this technique. Recently, the Australian CERT organization received information of a buffer-overrun vulnerability in the talkd daemon shipped in most modern BSD based systems. II. Problem Description To quote AUSCERT: talk is a communication program which copies text from one users terminal to that of another, possibly remote, user. talkd is the daemon that notifies a user that someone else wishes to initiate a conversation. As part of the talk connection, talkd does a DNS lookup for the hostname of the host where the connection is being initiating from. Due to insufficient bounds checking on the buffer where the hostname is stored, it is possible to overwrite the internal stack space of talkd. By carefully manipulating the hostname information, it is possible to force talkd to execute arbitrary commands. As talkd runs with root privileges, this may allow intruders to remotely execute arbitrary commands with these privileges. This attack requires an intruder to be able to make a network connection to a vulnerable talkd program and provide corrupt DNS information to that host. This type of attack is a particular instance of the problem described in CERT advisory CA-96.04 "Corrupt Information from Network Servers". This advisory is available from: ftp://info.cert.org/pub/cert_advisories/ Recent versions of FreeBSD 2.2 -current may not be affected with this vulnerability due to improved security in new versions of BIND, which sanity-check the results of reverse name lookups performed by the DNS system. III. Impact Intruders may be able to remotely execute arbitrary commands with root privileges. Access to a valid user account on the local system is not required. IV. Workaround Disable the ntalkd program found in /etc/inetd.conf by commenting the appropriate line out and reconfiguring inetd. # grep -i ntalk /etc/inetd.conf ntalk dgram udp waitroot/usr/libexec/ntalkd ntalkd After editing /etc/inetd.conf, reconfigure inetd by sending it a HUP signal. # kill -HUP `cat /var/run/inetd.pid` V. Solution The patches found at the following URL fix this vulnerability. Patches are available for FreeBSD 2.1.x (-stable) and -current. Acknowledgment: These patches were based off of published work provided by BSDI, Inc. After applying these patches, recompile and re-install the affected utilities. For FreeBSD -current (2.2 prerelease and 3.0 prerelease) systems: Index: announce.c === RCS file: /cvs/freebsd/src/libexec/talkd/announce.c,v retrieving revision 1.6 diff -u -r1.6 announce.c --- announce.c 1997/01/14 06:20:58 1.6 +++ announce.c 1997/01/18 08:27:04 @@ -34,7 +34,7 @@ */ #ifndef lint -static char sccsid[] = "@(#)announce.c
weird su behavior
Hi, recently I've been having problems going into root... When I do a 'su' and type in the passwd, it kicks me out with a "Broken pipe" When this is happening, I also can't telnet into the machine; the telnet session gets closed. What could be causing this? Where could I look to diagnose the problem? (nothing unusual is showing up in /var/log/messages or /var/log/syslog). TIA Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: how to install PGP?
Hi! I couldn't find the pinepgp package under contrib... :( Anyone else know where it is? On Tue, 21 Jan 1997, Hakan Ardo wrote: > > Another way is to install the pinepgp package, resently uploaded to the > contrib section. It contains script to will, decrypt and check signatures > on incommimg mail, and crypt or sign outgoing. > > - --- > Name:Hakan Ardo > E-Mail: [EMAIL PROTECTED] > WWW: http://www.ub2.lu.se/~hakan/sig.html > Public Key: Try "finger [EMAIL PROTECTED]" > Fingerprint: E9 81 FD 90 53 5C E9 3E 3D ED 57 15 1B 7E 29 F3 > Interests: WWW, Programming, 3D graphics > > Thought for the day: As long as one understands, the > spelling does not matter :-) > - --- > > -BEGIN PGP SIGNATURE- > Version: 2.6.3i > Charset: noconv > > iQCVAwUBMuR92N6dx9igIm71AQGCIwQA39iB7+iObGy2n5qZPLs+boPVTGVaWqNK > nofp8U3syOwnN6pXpj08l8t8ujlJ/wAtoPMdf5BiPQGupAE2jTAIWciPeEJcqQ5C > htjJ5fqVtz7G5ikYFvxmDKLocGE+xO0YRdfj6Ac1pJ+CFrdsSZTZFKtK8LhnYEWd > fD8Q6oIHraI= > =5BtL > -END PGP SIGNATURE- > -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
how to install PGP?
Hi, I tried installing the mailpgp package but it complains about not having pgp-i or pgp-us. How do I install these, and how can I "automate" pgp into something like pine, for example? Thanks Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
SECURITY: Important bug fix for /sbin/login (fwd)
Guys, has this been fixed in debian? If not, can anyone explain how to install an rpm package so I can try out rpm within debian? ;-) -- Their is a buffer overrun in /bin/login which has the potential to allow any user of your system to gain root access. util-linux-2.5-29 contains a fix for this and is available for Red Hat Linux 4.0 on all four platforms. We strongly recommend that all of Red Hat 4.0 usres apply this fix. Users of Red Hat Linux versions earlier then 4.0 should upgrade to 4.0 and then apply all available security pacakges. Users whose computers have direct internet connections may apply this update by using one of the following commands: Intel: rpm -Uvh ftp://ftp.redhat.com/updates/4.0/i386/util-linux-2.5-29.i386.rpm Alpha: rpm -Uvh ftp://ftp.redhat.com/updates/4.0/axp/util-linux-2.5-29.axp.rpm SPARC: rpm -Uvh ftp://ftp.redhat.com/updates/4.0/sparc/util-linux-2.5-29.sparc.rpm All of these packages have been signed with Red Hat's PGP key. Erik [mod: Forwarded by Richard Jones, Mangled by me to make this appear to have been sent by Erik himself... -- REW] - - - -- | I told you I'm not very bright -- Sugar in "Some Like It Hot" | | "RPM is the greatest thing since swap-space" - Bryan C. Andregg | | | Erik Troan = [EMAIL PROTECTED] =[EMAIL PROTECTED] | -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: using fork in cgi & perl
Hi...
Thanks for the help.
I've tried different combinations of using wait. And they all work fine
for making the child exit. But now the problem is that the parent for some
reason is not going to completion. It seems the very first statement in
the parent (after the child) executes, but the rest doesn't :(
Basically I have:
# cgi initialization ...
...
# output a list of users to a file
unless (fork) {
# output html code
}
else {
wait;
# debug
system ("cp /tmp/email.list /tmp/email.list2");
while () {
# read in file above and send mail to each recipient
}
# debug
system ("cp /tmp/email.list /tmp/email.list3");
}
That's pretty much what I have. I also put in a "close STDOUT" to force
the html output to go out. But what's happening now is that the parent
process is not doing its entire task. For debug, I put in some file copy
statements as seen above. The first cp goes thru, but I never get an
email.list3 file...
? help! ;)
Ricardo
On Fri, 3 Jan 1997, Jens B. Jorgensen wrote:
> Well, the child is a zombie because the process which called fork
> did not wait() on the child process. This can be accomplished
> asynchronously by handling the SIGCHLD signal. The following (somewhat
> modified to remove unecessary SysV stuff) example comes from the
> perlipc manpage:
>
> sub REAPER {
> $waitedpid = wait;
> }
>
> $SIG{CHLD} = \&REAPER;
> # now do something that forks...
>
> That code will make sure the child doesn't sit there as a zombie.
> As for why the result doesn't come up quickly I can't speculate
> since I don't know what you're doing.
>
> --
> Jens B. Jorgensen
> [EMAIL PROTECTED]
>
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
[EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
using fork in cgi & perl
Hi,
I'm having a hard time getting fork to work correctly in a cgi script.
Basically I have a time consuming process which I want to be done in the
background.
I have something like this:
unless (fork) {
# print out result html page
exit 0;
}
else {
# do time-intensive stuff
}
What's happening is that the child process executes (and becomes zombie)
and the parent process continues to chug along. The cgi result html only
comes up after the long process finishes...
I've tried switching the parent & child to see if it makes a difference,
but it doesn't. In every case, the html output goes into zombie state, and
the result page is only displayed after the entire process is complete.
Almost as if the fork isn't working as intended...
Any ideas?
Thanks,
Ricardo
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
[EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
status of hylafax?
Hi, I remember someone was working on a hylafax pkg for debian... What's the status? AFAIK no distributions for linux have a hylafax pkg... (does redhat?) but freebsd does have one... Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
problems with st0 - HELP! :)
Hi, I'm having problems using my DAT tape drive under 2.0.25 This drive used to be in a 1.2.13 system and worked fine. Ever since I moved it to a newer server, with 2.0.23 and then 2.0.25 I've had problems using the tape drive. Right now I'm getting errors like these: Dec 27 09:17:00 irvine kernel: st0: Error with sense data: extra data not valid Current error st09:00: sns = 70 3 Dec 27 09:17:00 irvine kernel: ASC= 3 ASCQ= 2 Dec 27 09:17:00 irvine kernel: Raw sense data:0x70 0x00 0x03 0x00 0x00 0x00 0x00 0x0a 0x00 0x00 0x00 0x00 0x03 0x02 0x00 0x2a Any clues out there? TIA Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
stubborn innd problem wont go away... HELP!
Hi guys, I'm getting frustrated with this... :( Every night my news server will basically stop because it gets throttled. I keep getting the error innd: SERVER throttle File exists writing symlinking article file -- throttling I've upgraded to inn 1.5 in the hopes of clearing some of this stuff up, but nothing... I'm running renumber every night, and I've run makehistory a couple of times in the hopes that the makehistory/renumber sequence could clear it up. I've cleaned up some of the articles to make sure I have plenty of disk left. Please, does anyone know how to fix this throttling problem? Why is the server complaining about symlinking article and why does that force a throttle? My crontab runs news.daily around 3am and it seems like everyday after news.daily runs, the server stops, because the log stays at length 0 after news.daily runs... TIA, Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
success with inn 1.5?
Hi, Anyone have success in building inn 1.5? I haven't found any binary distributions and having a rough time getting the distribution to compile properly... TIA Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: problems with innd
Hmmm thanks for the pointer! I went to check, and my news.daily is already renumbering every day... I had already tried the stop-makehistory-start-renumber sequence, but for some reason I'm still getting throttled... What are the reasons for throttling and the system staying in that mode? Thanks again, Ricardo On Tue, 17 Dec 1996, Tim Sailer wrote: > In your email to me, Ricardo Kleemann, you wrote: > > > > Why does the server keep throttling? :( I can't find any error messages > > that point me to my problems... But I've been finding that everynight the > > server is stopping (I assume throttling). Each morning I look at the > > logfile and it is of size zero. Only after I do a ctlinnd go "" does it > > start back up... > > I use innd built from scratch, and I don't have these problems. However, > on some other boxes that I admin, I see these from time to time. > Did you install debian's innd on a running news system? Try shutting > down the news system, and run makehistory. Then start up and renumber. > If you can spare the cpu cycles, have news.daily do the renumber > for you at night. > > Tim > > -- > (work) [EMAIL PROTECTED] / (home) [EMAIL PROTECTED] - > http://www.buoy.com/~tps > "It takes more hot water to make cold water hot > than cold water to make hot water cold." >Jon Blummer > ** Disclaimer: My views/comments/beliefs, as strange as they are, are my > own.** > -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: problems with innd
Why does the server keep throttling? :( I can't find any error messages that point me to my problems... But I've been finding that everynight the server is stopping (I assume throttling). Each morning I look at the logfile and it is of size zero. Only after I do a ctlinnd go "" does it start back up... Help! ;) Ricardo On Tue, 17 Dec 1996, Tim Sailer wrote: > In your email to me, Ricardo Kleemann, you wrote: > > > > I ran > > ctlinnd renumber "" > > > > and got the message "Must be running". > > > > This is the second time I've seen this message... what does it mean? > > innd is up and running and receiving articles. Why would ctlinnd say that? > > > > The server throttled with the error, so you have to run > 'ctlinnd go "" ' to get it running first. > > Tim > > -- > (work) [EMAIL PROTECTED] / (home) [EMAIL PROTECTED] - > http://www.buoy.com/~tps > "It takes more hot water to make cold water hot > than cold water to make hot water cold." >Jon Blummer > ** Disclaimer: My views/comments/beliefs, as strange as they are, are my > own.** > -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: problems with innd
I ran ctlinnd renumber "" and got the message "Must be running". This is the second time I've seen this message... what does it mean? innd is up and running and receiving articles. Why would ctlinnd say that? On Mon, 16 Dec 1996, Robert Stone wrote: > > preforming a: > ctlinnd renumber "" > > nightly from cron fixed this for me... i don't believe it is a > problem with mmap, but I have yet to figure out where it is. > -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
problems with innd
Hi, I'm using Debian 1.1, with the supplied innd package. I'm having problems posting articles, getting the message Can't send article to the server: 441 400 File exists writing symlinking article file -- throttling (Article not posted.) According to the INN-FAQ, Linux must not use MMAP for innd; supposedly that's the cause of the problem. Is that true? I assumed that the debian-distributed innd would work fine out of the box... What is the problem with mmap? How do I fix this problem? Please help... Thanks! Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
more problems with newsfeeds
Hello, In my testing of feeding a site, it is apparent that the articles are not being fed in their entirety, rather going in 16K chunks. Why would that be? Where is this configured? thanks! Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: checksecurity in cron.daily/standard HELP!
Thanks! I went to try to upgrade it to the frozen version (and also to get the man page), but couldn't find any packages named check* when I did a find... What's the package that I should install? Ricardo On 6 Dec 1996, Guy Maor wrote: > Ricardo Kleemann <[EMAIL PROTECTED]> writes: > > > I see no man page for checksecurity... > > There is a man page in version in frozen, 3.0pl1-34, and it has an > answer for you: > >The checksecurity command scans the mounted files systems >(subject to the filter defined in /etc/checksecurity.conf) > > > Guy > > > -- > TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to > [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] > -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: newsfeeds help!
Thanks... I'm trying to figure out if I'm really feeding the site... where else can I look to confirm? After many attempts, the user at the other end informed me that they had received one article. That's it... and there's a lot of articles here. What is the criteria for sending and when articles get sent? I've run nntpsend and the send-nntp script and it basically starts and stops right away. Where would I look to see if anything actually got sent? Ricardo On Fri, 6 Dec 1996, Tim Sailer wrote: > In your email to me, Ricardo Kleemann, you wrote: > > > > Hi, > > > > I've set up a test feed to another provider... > > > > I configured my newsfeeds file like this: > > > > testfeed:!*,comp.*:Tf,Wnm:feed.domain.com > > > > (of course "feed.domain.com" is replaced with the FQDN of the site) > > > > Anyway, when I look at the .outgoing directory, it lists all these > > groups which are not configured in the feed. > > > > Does that mean that innxmit is offering all those groups/articles even > > though I've told it to only send the comp.* groups? > > > > If so, how do I make sure that only the groups I want get sent? > > > > I'm testing the feed by running the send-nntp script with the feed site > > host name as the argument: > > > > send-nntp feed.domain.com > > My first reaction is that you are seeing the results of crossposting. > Take a look at the actual headers of the article and see. > > Tim > > -- > (work) [EMAIL PROTECTED] / (home) [EMAIL PROTECTED] - > http://www.buoy.com/~tps > "A desk is a dangerous place from which to view the world." >-- John le Carre' > ** Disclaimer: My views/comments/beliefs, as strange as they are, are my > own.** > -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
newsfeeds help!
Hi, I've set up a test feed to another provider... I configured my newsfeeds file like this: testfeed:!*,comp.*:Tf,Wnm:feed.domain.com (of course "feed.domain.com" is replaced with the FQDN of the site) Anyway, when I look at the .outgoing directory, it lists all these groups which are not configured in the feed. Does that mean that innxmit is offering all those groups/articles even though I've told it to only send the comp.* groups? If so, how do I make sure that only the groups I want get sent? I'm testing the feed by running the send-nntp script with the feed site host name as the argument: send-nntp feed.domain.com thanks for any help! Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
checksecurity in cron.daily/standard HELP!
I see no man page for checksecurity... I see that this program goes thru all my file systems using a find... Is there a way to prevent the find from going to certain directories/filesystems? I have news running on my system, and basically any find running thru my news spool simply just skyrockets my cpu utilization... I want to prevent checksecurity from going into my news spool... Anyone help? thanks. Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: making disk bootable
Thanks! How do I install the mbr package? Is it a normal .deb package? Right now my /dev/hda does NOT have an MBR... I need to install it in order to use lilo. Will liloconf help me out? Ricardo On Sat, 23 Nov 1996, Bernd Eckenfels wrote: > Hi, > > > How do I go about inserting an mbr if my current setup isn't using LILO > > (and thus hasn't made the hd bootable)? > > You should install lilo in the root-fs partition. Then it will be loaded if > there is a normal MBR and the partition is activated, or you can install the > debians mbr package into the MBR. Try to boot from disk and use the liloconf > for this (both). > > Greetings > Bernd > -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
making disk bootable
Hi! How do I go about inserting an mbr if my current setup isn't using LILO (and thus hasn't made the hd bootable)? Does toggling the bootable flag in fdisk take care of that? I'd like to install lilo on my system and up to now I haven't used it, so my hd is not bootable. TIA! Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
libc 4.7 available?
Hi, I'm having trouble compiling the virtual ftpd, but I do have a working binary, execpt that it is expecting to see libc-4.7 and my debian system has libc-4.6.27. I do have libc-5.2.18 as well, but I guess that it only looks at a minor version of the major version (meaning that even though I have libc-5.x, it is only looking for a libc-4.x compatible, which mine is not) How can I get a libc-4.7.x to install? Anyone been successfull in compiling the virtual ftpd under 2.0.x? It compiles fine for me under 1.2.13 but not under 2.0.x. Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: libc
Thanks! But I installed the debian ELF system and what it put in as default is 4.6.27. I have 5.2.18 in there (as the ldconfig -v shows) but it is not being used as default. How do I go about upgrading/changing the default? Ricardo On Fri, 22 Nov 1996, Dimitri Maziuk wrote: > Ricardo Kleemann wrote: > > > > Hi, > > > > I've installed the stable debian 1.1 release and noticed that my libc is > > defaulting to libc.so.4.6.27 > > > > Is there any reason why it's not using a later libc? Are there bugs with > > libc versions > 4.6 (either 5.x or 4.7)? > > > > Hi. > > I suspect libc4 never got upgraded to libc-4.7.5 because it's an a.out > library needed by only a few obscure packages which only a few people > use. > > Debian is ELF and is based on libc5. The latest libc5 package (in rex) > is libc-5.4.7, your system might have a libc-5.3.x. I suggest you > re-check your /lib or just run ldd on something -- say, 'ldd /bin/bash'. > > Rgds > Dimitri > -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
libc
Hi, I've installed the stable debian 1.1 release and noticed that my libc is defaulting to libc.so.4.6.27 Is there any reason why it's not using a later libc? Are there bugs with libc versions > 4.6 (either 5.x or 4.7)? I have another system running 4.7.x and it seems fine. Should I upgrade libraries, and if so, what's the recommended method in Debian to do so? Thanks, Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: rdist as root?
I just did this. You'd have to set up root with permission for rsh by putting a .rhosts in the root home directory. Beware of security issues though... On Tue, 19 Nov 1996, Dirk Luetjens wrote: > > How can I invoke a rdist command as root to distribute the /usr/local/ > tree to the other computers in the network? > > Dirk > > -- > TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to > [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] > -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
radiusd
Is there a HOWTO or detailed info on how to setup radiusd? I tried a web search but it returns basically 1,000's of ASCEND and portmaster realted stuff... nothing really on how to setup and configure it. TIA Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Using users/passwords from freebsd
Hi, Anyone know of a tool/library that allows linux to use freebsd passwords? If I'm not mistaken, linux uses DES encryption and freebsd uses an MD5 encryption, right? Is there a way that linux can make use of the freebsd passwords or maybe the other way around? Thanks! Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
using NIS
Hi guys, I've set up a Debian linux system to be NIS master and have a FreeBSD system trying to act as client. How do I test/debug this? I tried logging into the FreeBSD system with a user account from the master system, but it didn't work... First time with NIS for me, so I'm looking for pointers. Thanks! Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: too many open files?
Thanks!!! I did su to root! That must be the reason! I'll give it a try! Ricardo On Mon, 11 Nov 1996, Heiko Schlittermann wrote: > Ricardo Kleemann wrote: > : > : Hi guys, > : > : I tried running apache and got back the "too many open files" error. > : > : How can I check/monitor the number of file descriptors being used?? > : > : Where do I change it if it needs to be increased? > > Installed lshell? > Became root via su? > > > Heiko > -- > email : [EMAIL PROTECTED] [EMAIL PROTECTED] [EMAIL PROTECTED] > pgp : A1 7D F6 7B 69 73 48 35 E1 DE 21 A7 A8 9A 77 92 > finger: [EMAIL PROTECTED] [EMAIL PROTECTED] > -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: too many open files?
Thanks for the help!!! So now I'm puzzled! :( here are my values: irvine:/$ cat /proc/sys/kernel/file-nr 192 irvine:/$ cat /proc/sys/kernel/inode-nr 16801502 irvine:/$ cat /proc/sys/kernel/file-max 1024 irvine:/$ cat /proc/sys/kernel/inode-max 3072 I'm only running about 15 virtual WWW domains from apache, but when I run httpd, it reports the "too many open files" error! :( How can it be if before running it, there are only 192 open? Thanks again, Ricardo On Mon, 11 Nov 1996, Oskar Pearson wrote: > Hi > > > I tried running apache and got back the "too many open files" error. > > There are generally two possible errors: > You can have too many files open on your system (say you have lots of programs > that are opening a few files each) > or a single process (like apache) can open 256 files (this is the default > value) > > The one you can increase on the fly, the other not. > > To increase the total number that you can open on the system, you can > "echo" values to various files in /proc (assuming you are running 2.0.latest) > > > How can I check/monitor the number of file descriptors being used?? > cat /proc/sys/kernel/file-nr (maximum Overall open files on system opened) > cat /proc/sys/kernel/inode-nr (maximum Overall open inodes on system opened) > > cat /proc/sys/kernel/file-max (maximum files on the system that can be opened > concurrently) > cat /proc/sys/kernel/inode-max (maximum inodes on the system that can be > opened > concurrently) > > > Where do I change it if it needs to be increased? > > If you want to increase the files per process, have a look at > http://www.linux.org.za/tweak.html (Yes, I know that it is almost the only > thing on the server :( > > If you want to increase the number of maximum open files (ie the value > in file-max is the same as the value in file-nr) you can echo values as > follows: > > echo "4096" >/proc/sys/kernel/file-max > echo "12288" >/proc/sys/kernel/inode-max > > (inode-max's value is almost allways 3 times the size of file-max, keep it > that > way!) > > Oskar > > -- > TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to > [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] > -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
too many open files?
Hi guys, I tried running apache and got back the "too many open files" error. How can I check/monitor the number of file descriptors being used?? Where do I change it if it needs to be increased? thanks. Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
SMC cards
Hi guys, Which SMC PCI card is the one mostly used? (Is it the tulip model?) Which model SMC, Etherpower, EtherEZ? Thanks ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: [Linux-ISP] NCR 825 ctrler
Thanks! On Thu, 10 Oct 1996, Jon Lewis wrote: > > > Anyone use it? Is it worth to buy it? > > I've not gotten an 825 yet, but I have around a dozen 810's in use. > They've become the standard FDT SCSI card. At about $59 each, they're > hard to pass up. I'd get the 825 if you want fast-wide on a budget, or a > BT-958 if you're not. I'd pass on the Adaptec since they seem to be less > cooperative about releasing programming info and the 29xx driver seems > to have been slow to stabilize because of this (I assume). > Would you say the Buslogic is better due to higher cost? I've always heard good things about the NCR controllers relative to Linux, but does the fact that they are _much_ cheaper than the BT or Adaptec mean they are poor performers? Has anyone made any comparisons? Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
NCR 825 ctrler
Hi guys, I saw a price list which had the NCR 825 controller, and it said it's a Fast & Wide controller. Is that true? Why is it so much cheaper than the Adaptec, for example? Is it a poor performer? The price was about $100 cheaper than the adaptec. Anyone use it? Is it worth to buy it? Thanks, Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
O'clocking P6-180 to P6-200... safe?
Hi guys, Many of us are out trying to buy/put together a Pro200 system... However, I've run into a bit of trouble I suspect many others also have. The P6-200 is OUT OF STOCK! Back-ordered everywhere I've checked. I've had some decent price quotes, but then they say they don't have it in stock! However, it seems the Pro-180 is available (and is cheaper). So, the question, should/could I overclock it to 200 (or more)? What would be the steps to overclocking a P6? I plan on getting the Intel Venus M/B, so how would I set it up to use a P180 but run it at 200mhz? Thanks, Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
analog log analyzer?
I heard something about a program called "analog" in debian which does httpd log file analysis? Does it exist? I tried to look for it in the debian ftp area but couldn't find anything (at least in the packages file). Thanks Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
ftp statistics
Is there a program out there to compute FTP statistics, similar to what is done for www statistics? Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
shadow for debian
Does debian have a shadow package which would simplify installation of shadow? Or do I have to install it manually? Thanks, Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
1.1.9?
What's the difference between Debian 1.1.8 and 1.1.9? I see the i-connect page still refers to 1.1.8, so is 1.1.9 not "official"? -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Buslogic SCSI
Thanks! So what would you say is the best performing SCSI controller (in wide or ultra flavor)? Is the 2940UW the best? Ricardo On Tue, 1 Oct 1996, Bruce Perens wrote: > Don't buy a FlashPoint for Linux. If you do buy one, BusLogic will > upgrade it to one of the MultiMaster controllers for a nominal fee. I have > the details. The FlashPoint isn't a good controller for multi-tasking > systems, and there's no Linux driver. > > Bruce > -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
BusLogic SCSI
Hi, Anyone know if the Buslogic FlashPoint (wide scsi) controllers are supported? What is the price compared with Adaptec 2940UW? Thanks! Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Linux SMP
Hi guys... I'm wondering if attempting to use Linux-SMP is any good. Afterall, for real efficiency, don't the programs/daemons/utilities have to be designed in such a way that they'll make use of SMP capabilities? For example, if I have sendmail, pop3d, httpd daemons running, on a dual processor system with SMP kernel, as compared to a single-processor system, would I gain much advantage? Let's say, for example, a dual-133 and a single-PPro200. My feeling is that for raw speed and efficiency, the PPro200 would easily win out... Now, due to the fact that most ISP usage is dependent on bandwidth limitations (meaning that a lot of the time the raw processor power won't do much good) what is your opinion on choosing a more effective system? Between say dual-133 or single-Pro200? I definitely feel that on "internal" processing, like compiling, using msql, etc, the Pro200 would by far be the better choice. Regards, Ricardo -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Using reserved addresses
Hi, I have a client who would like to use the reserved addresses internally throughout his LAN... My question is, will the workstations which use reserved IPs be able to go out into the net? If those reserved IPs are not routed "outside", then how would a workstation be able to properly communicate? What are the limitations/issues with using the reserved IPs? Thanks, Ricardo
Version 2.0
Hi, Some are saying that 2.0.20 may be the final 2.0 version... Does debian already have 2.0.20 in its stable tree? If I buy a Debian CD, will it come with 2.0.20? Thx. Ricardo
RAID in Linux (was Re: SCSI and EIDE)
Hi guys, starting a new thread about the subject... ;-) I'm still trying to get a grip on this... Few questions: 1. can raid0/raid1 be done on either scsi or ide or both? 2. what's the difference between raid0 and raid1? 3. what exactly does it do? Does it mirror data accross multiple devices, and if so, how does it maintain the data in all devices? 4. How does one go about "creating" an md device? Would it automatically mirror a "non-md" drive into the multiple devices? Thanks! Ricardo
Re: SCSI and EIDE
That's wonderful! Now will Linux implement anything greater than RAID0? Would you say your performance is significantly increased with striping? How many drives can be striped? On 15 Aug 1996, Rob Browning wrote: > Ricardo Kleemann <[EMAIL PROTECTED]> writes: > > > Good point about RAID0. > > > > I know linux's implementation of RAID0 is with the MD program. Do you > > know if that's fully functional, and if so, is there a Debian package for > > it? How is it used? > > I'm using it on two machines in the lab. One has two identical > Quantum Fireballs, and the other has a Fireball and a Western Digital > Caveat (ahem Caviar). On each machine most of the partitions are > striped, but at least one is not. It works fine. > > You need to compile a kernel with bultiple device/RAID0 support and > you need to install the debian mdutils package. Then read the stuff > in /usr/doc/mdutils. That's it. The md package doesn't care if the > drives are the same type, or even on the same kind of bus. That's the > advantage to a software implementation. Granted I would expect > performance to be best when you have 2 identical drives on the same > kind of bus... > > -- > Rob >
RE: SCSI and EIDE
Good point about RAID0. I know linux's implementation of RAID0 is with the MD program. Do you know if that's fully functional, and if so, is there a Debian package for it? How is it used? On Wed, 14 Aug 1996, Al Youngwerth wrote: > Given equal drive parameters, the IO throughput (what matters for > multi-tasking OSs) for a SCSI-2 system should be better than EIDE. Where SCSI > really begins to shine in IO throughput is using disk block striping (RAID 0) > using two or more SCSI disks. You can place 7 disks on a single SCSI bus and > come no where near saturating the SCSI bus bandwidth with the random IOs that > Linux generates. If one SCSI drive gives you 60 IOs/sec of throughput, seven > of those drives gives you 420 IOs/sec of throughput. I only wish it was that > easy to increase my CPU performance! > > Al Youngwerth > [EMAIL PROTECTED] >
dump for a.out?
Hi guys, I'm really needing to use the dump utility to schedule my backups. My problem is that dump requires libc5.x, and I'm still running libc4.x. I tried upgrading libc, but it runs into other dependency conflicts, and it seems to me that unless I start upgrading *everything*, which I can't afford to do now (due to downtime), I'm SOL. So, question: Is the source for dump available so I can compile it myself with the old libs? Help! ;-) ricardo
More questions on upgrading...
Hi! I've mirrored most of the debian distribution, and I've gone thru the select process in dselect. Basically I'm ready to perform a full upgrade to Debian-1.1 now. But I'm still weary of doing it because I don't want to have much downtime since my server is used for others dialing in. My main question is: Debian-1.1 is all ELF. My current system is a.out, including the kernel. If I go ahead with the upgrade from dselect, will my current 1.2.13-a.out kernel still be useable once the upgrade is complete? Or will I have to upgrade the kernel as well (which means I'd have to immediately reboot the server). I want to be able to go thru with the upgrade but not reboot immediately. I want to keep the server running while I go thru the kernel upgrade process (like compiling a new 2.0.x kernel). Also, will the current mgetty and pppd I have (a.out) work with the upgraded Debian? I will still upgrade those, but, similar to the kernel issue, I'd like to be able to continue using my current mgetty/pppd setup and upgrade those in a timely fasion. Suggestions are welcome! :) Ricardo
pppd2.2.0f Patch (fwd)
Guys... saw this in linux-ppp mailing list... Has this patch been
incorporated into the debian pppd??
-- Forwarded message --
Date: Mon, 05 Aug 1996 19:58:33 +0200
From: Michael Lausch <[EMAIL PROTECTED]>
To: linux-ppp@vger.rutgers.edu
Subject: pppd2.2.0f Patch
This is a patch which removes an error when using Linux as a server
and Windows for Workgroups as a client with PAP authorization.
Forwarded: Mon, 05 Aug 1996 17:57:13 +0200
Forwarded: [EMAIL PROTECTED]
Forwarded: [EMAIL PROTECTED]
Replied: Thu, 01 Aug 1996 10:53:49 +0200
Replied: [EMAIL PROTECTED]
Replied: Wed, 24 Jul 1996 08:18:57 +0200
Replied: [EMAIL PROTECTED]
>>From [EMAIL PROTECTED] Wed Jul 24 02:36:19 1996
Received: from mario.gams.at (mario.gams.at [194.42.96.10]) by
bononunu.gams.co.at (8.7.1/8.7.3) with ESMTP id CAA15680 for <[EMAIL
PROTECTED]>; Wed, 24 Jul 1996 02:36:19 +0200
Received: from avago.anu.edu.au (avago.anu.edu.au [150.203.162.34]) by
mario.gams.at (8.7.5/8.7.3) with SMTP id CAA28750 for <[EMAIL PROTECTED]>; Wed,
24 Jul 1996 02:36:15 +0200
Received: (from [EMAIL PROTECTED]) by avago.anu.edu.au (8.6.12/8.6.9) id
KAA00860; Wed, 24 Jul 1996 10:35:29 +1000
Date: Wed, 24 Jul 1996 10:35:29 +1000
Message-Id: <[EMAIL PROTECTED]>
From: Paul Mackerras <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]
In-reply-to: <[EMAIL PROTECTED]> (message from Michael
Lausch on Tue, 23 Jul 1996 12:09:15 +0200)
Subject: Re: FW: pppd2.0.2f and Trumpet Winsock
Reply-to: [EMAIL PROTECTED]
> You get more then requested :-). I'm including the whole debug trace.
> It's MIME encoded to preserve formatting.
Oops, looks like the NAK from Trumpet was perfectly OK, and the bug is
in pppd. This is something I fixed a little while ago in ppp-2.3,
which is only in alpha release as yet, and its linux port needs some
more work. So I've tried to isolate the necessary changes to
pppd/lcp.c and put them into the ppp-2.2.0f source. I've included the
patch below. Let me know if it works.
Paul.
--- lcp.c.orig Wed Dec 27 13:25:51 1995
+++ lcp.c Wed Jul 24 10:30:17 1996
@@ -792,19 +792,22 @@
*/
if ((go->neg_chap || go->neg_upap)
&& len >= CILEN_SHORT
- && p[0] == CI_AUTHTYPE && p[1] >= CILEN_SHORT) {
+ && p[0] == CI_AUTHTYPE && p[1] >= CILEN_SHORT && p[1] <= len) {
cilen = p[1];
+ len -= cilen;
+ no.neg_chap = go->neg_chap;
+ no.neg_upap = go->neg_upap;
INCPTR(2, p);
GETSHORT(cishort, p);
if (cishort == PPP_PAP && cilen == CILEN_SHORT) {
/*
-* If they are asking for PAP, then they don't want to do CHAP.
+* If we were asking for CHAP, they obviously don't want to do it.
* If we weren't asking for CHAP, then we were asking for PAP,
* in which case this Nak is bad.
*/
if (!go->neg_chap)
goto bad;
- go->neg_chap = 0;
+ try.neg_chap = 0;
} else if (cishort == PPP_CHAP && cilen == CILEN_CHAP) {
GETCHAR(cichar, p);
@@ -822,7 +825,7 @@
/*
* Stop asking for PAP if we were asking for it.
*/
- go->neg_upap = 0;
+ try.neg_upap = 0;
}
} else {
@@ -831,25 +834,14 @@
* Stop asking for what we were asking for.
*/
if (go->neg_chap)
- go->neg_chap = 0;
+ try.neg_chap = 0;
else
- go->neg_upap = 0;
+ try.neg_upap = 0;
p += cilen - CILEN_SHORT;
}
}
/*
- * Peer shouldn't send Nak for protocol compression or
- * address/control compression requests; they should send
- * a Reject instead. If they send a Nak, treat it as a Reject.
- */
-if (!go->neg_chap ){
- NAKCISHORT(CI_AUTHTYPE, neg_upap,
- try.neg_upap = 0;
- );
-}
-
-/*
* If they can't cope with our link quality protocol, we'll have
* to stop asking for LQR. We haven't got any other protocol.
* If they Nak the reporting period, take their value XXX ?
@@ -869,6 +861,11 @@
looped_back = 1;
);
+/*
+ * Peer shouldn't send Nak for protocol compression or
+ * address/control compression requests; they should send
+ * a Reject instead. If they send a Nak, treat it as a Reject.
+ */
NAKCIVOID(CI_PCOMPRESSION, neg_pcompression,
try.neg_pcompression = 0;
);
@@ -895,7 +892,7 @@
while (len > CILEN_VOID) {
GETCHAR(citype, p);
GETCHAR(cilen, p);
- if ((len -= cilen) < 0)
+ if (cilen < CILEN_VOID || (len -= cilen) < 0)
goto bad;
next = p + cilen - 2;
@@ -903,6 +900,9 @@
case CI_MRU:
if (go->neg_mru || no.neg_mru || cilen != CILEN_SHORT)
goto bad;
+ GETSHORT(cishort, p);
+ if (cishort
Re: dpkg-ftp or installing via FTP
Thanks! I've actually tried running dselect with ftp, but it didn't work. It looks like my version of perl may have some problems as can be seen in the error messages below. Since I'm still using Debian-0.93 and am trying to upgrade, I'm not sure how to upgrade perl because all the packages in the ftp site are newer packages dependent on newer libcs, etc. I choose ftp under dselect, this is what I get: Socket 1.3 required--this is only version (undef) at /usr/lib/dpkg/methods/ftp/perl/Net/FTP.pm line 41 BEGIN failed--compilation aborted at /usr/lib/dpkg/methods/ftp/perl/Net/FTP.pm line 41. BEGIN failed--compilation aborted at /usr/lib/dpkg/methods/ftp/setup line 8. query/setup script returned error exit status 2. Any ideas/suggestions?? Ricardo On Fri, 2 Aug 1996, Michael Karafotis wrote: > I always take the easy way out and run dselect .. then select my method > from there. It is quite nice .. set to ftp then select the server and > packages you want. > > Enjoy, > Michael >
Re: mirror
Hi! First, thanks for the pointers... But I'm still having problems. I've done an unpack and install on mirror-2.8-0.deb (that's the version that I have here for 0.93), but there's nothing under /etc/mirror (as a matter of fact, that directory didn't even exist). There's also no README.debian under /usr/doc/examples/mirror. I've run into problems unpacking newer version packages into 0.93 because my system is a.out and has old libraries. As a matter of fact, I've actually killed some of my apps (like emacs, for example), because the unpack changes things before checking dependencies, it looks like. So I've killed some stuff by attempting to upgrade packages... :( Where can I get sample copies of the mirror package "ftp.debian.org" and also the README for mirror? Thanks, Ricardo On Fri, 2 Aug 1996 [EMAIL PROTECTED] wrote: > > Ricardo> Here's the output: > Ricardo> TS1:/packages/debian# mirror -n > -gsite:ftp.debian.org:/pub/linux/distributions/debian/stable/ > > The Debian package is set up to use the 2nd method of mirror invocation, ie > mirror /etc/mirror/packages/ftp.debian.org > > Ricardo> not loading mirror.defaults >^^^ > > This happens when you call it with the -gsite: method. See the manpage. > > Ricardo> Scanning local directory . > Ricardo> Connecting to site:ftp.debian.org > Ricardo> Failed to connect > Ricardo> Pausing between retries > Ricardo> Cannot connect, skipping package > > Check out some of the doc I provide with the package, for example > /usr/doc/examples/mirror/README.debian > /etc/mirror/packages/ftp.debian.org > This example package file was created in December for the 0.93R6 structure on > the Debian site, you will have to change it a bit. Read the manpage, and you > will understand it all, I hope. > > Best regards, Dirk > > > -- > Dirk Eddelb"uttel http://qed.econ.queensu.ca/~edd >
dpkg-ftp or installing via FTP
Hi guys, I grabbed and installed dpkg-ftp. But I see no mention of any ftp usage when I run dpkg --help. Is there supposed to be a separate binary called dpkg-ftp? Because I can't find any... I'm trying to download the latest Debian distribution via FTP, but I've had problems both with ncftp and with dftp. I was hoping dpkg-ftp would be a better solution... Another suggestion was to mirror debian. How much space is required to mirror? Can a mirror be partial? Lastly, once I've gotten all that's required from FTP, is dselect a good method to use for performing the install? Thanks, Ricardo
