Re: Shell scripts: variable assignment within read loops
On Sun, 17 Aug 2008, David Wolfskill wrote: [snipped] > will assign to foo the value of the bar variable form the last record > read (in FreeBSD 6.3-STABLE, at least), the following fails to do so: > > foo="" > cat $filename | while read bar ... ; do >... > foo=$bar >... > done > echo $foo > > Well, that's not *quite* accurate:the assignment is done all right, but > in the latter case, it appears to be done in a subshell, so by the time > we get to the "echo" statement, any variable assignments from within the > read loop have vanished. You've already defined the reason behind this. Since subshells for parts of pipelines aren't guaranteed, you need something more clever. Assuming you're only interested in the _last_ value of foo, you could try this: foo=$( cat $filename | ( while read bar; do ... foo=$bar ... done echo $foo ) ) Cheers, jan -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ ( echo "ouroboros"; cat ) > /dev/fd/0 # it's like talking to yourself sometimes ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: sh script difficulties (running parallel functions)
On Thu, 1 Nov 2007, David Naylor wrote: > Hi, > > I am having a hard time getting (very complex script for me) to work. The > basic idea is that this script runs a bunch of tarkets, many of which are > time consuming but low on resources (such as downloading files). Now if I > run the tarkets all at once (given some dependancy issues) it greatly speeds > up the process (about 5 time speed increase). However I do not know how to > do this using sh... > > Example > > #!/bin/sh > > worker1() { > # Copy some files > } > > worker2() { > # Download some files > } > > worker3() { > # Do something else > } > > . # and so on > > run_jobs() { > worker1 & > worker2 & > worker3 & > # !!! Somehow wait for over workers to finish before continuing !!! > } > > #Finished > > Furthermore, how can signals be handled such that the signals get > accumulated and once all the other workers have finished the signals get > passed on (appropriately) The "wait" shell builtin is part of what you're after. You probably will need to trap the signals you're interested in catching. Just a trap 'int=1' INT wait trap - INT if [ "x$int" = x1 ]; then ... ; fi should do it. -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ There's no convincing English-language argument that this sentence is true. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: is this for OO-2 for FBSD?
On Mon, 29 Oct 2007, Aryeh M. Friedman wrote: > OOo-2 from ports is really odd anyways... some > people seem to be able to get it made right out of the box and others > can't as far I can tell there is no rhyme or reason as to why it fails > or does not fail Seem to recall that OOo won't complete a build on a filesystem with noatime; there's a perl(? it's a long time since I did this) script that can't tell the difference between an fstat successfully returning 0 (for midnight, Jan 1 1970) and failing. jan -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ OORDBMSs make me feel old; I remember when this was all fields. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Repeated PXE jumpstart
On Fri, 21 Sep 2007, Erik Cederstrand wrote: > Found the answer elsewhere, so this is just for the records... > > # sysctl kern.geom.debugflags=16 > # dd if=/dev/zero of=/dev/ad0 bs=512 count=1 > > wipes the MBR and solves my problem. The alternative is to put PXE ahead of the HD in the boot order, and call back to the deployment host at the end of installation (prior to reboot) to signal a DHCP reconfiguration. It adds a PXE timeout to each boot; the upside is that replacing a wedging or otherwise broken install is just a matter of reconfiguring a DHCP server. -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Random act of violence against bread: whole pint. -- extract from the "Hawk the Slayer" drinking game ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: best programming language for console/sql application?
On Sun, 22 Apr 2007, Mike Jeays wrote: > On Sunday 22 April 2007 06:32, Zhang Weiwu wrote: > > Dear list > > > > This is OT. I am a 4 year php developer and is very familiar with > > javascript and awk (familiar = knows and used all functions and features > > of the language itself) and I am a 5 year FreeBSD user. Being frustrated > > for the lack of a good console-based issue tracking tool (like mantis or > > bugzilla), I think I should start to write my own. I'll either start > > from scratch or (better) write a frontend for mantis which I used for > > years. > > > >1. If someone has already started, I should try join him/her rather > > than reinventing the wheel. So if someone knows any person who is > > starting to work on a slim console-based issue tracker, please let > > me know. I already did quit a few searches. I know someone is > > working on a console front-end of G-Forge, but a big software like > > G-Forge is not what I am thinking of. Not sure it's quite what you're after, but have a look at request tracker. It is primarily web based, with an email gateway, but also supports command-line operation via its bin/rt. -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Whenever I see a dog salivate I get an insatiable urge to ring a bell. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Corrupted C Compiler
On Mon, 4 Dec 2006, Rachel Florentine wrote: From: Jan Grant <[EMAIL PROTECTED]> >>I don't think it makes much sense, no. Zope is python-based and unless >>you're building products that rely on native libraries, what you >>describe doesn't sound like an accurate diagnosis. It's more likely >>(this is a stab in the dark) that you're running a script to regerate >>.pyc files; these are precompiled python bytecode files that are built >>from the corresponding .py files. > >I don't have my references in front of me, but I ran into an error when >I tried to runzope that I didn't understand, so I yahoo'd it and >discovered that my c libraries were corrupted. It was recommended I run >a certain command to clean them up. I ran that command and everything >went smoothly. I repeated this process a few times. So yes, I'm sure >that's what I did. When you have a chance, it'd be nice to get a detailed error report (including the command you were running to clean up the libraries). -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ ...and then three milkmaids turned up (to the delight and delactation of the crowd). ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Corrupted C Compiler
On Mon, 4 Dec 2006, Rachel Florentine wrote: > Hi; > I entered the following stupid command: cp -R /* /ad2 thinking that > would copy the contents of my working HD to my new 1/2 teraflop HD > (ad2). What it did was manage to wipe out some very important files > (thank goodness I had up-to-date backups) and it appears to have > corrupted gcc...my C compiler. I deduce this because when I go to build > Zope (as an example) from source I have to run a script afterwards that > repairs the broken C files. (This, strangely, is not the case if I build > Zope from port.) So, my questions for you programmers more experienced > than I, are: > > 1) Does my assessment make sense? Is gcc corrupt? I don't think it makes much sense, no. Zope is python-based and unless you're building products that rely on native libraries, what you describe doesn't sound like an accurate diagnosis. It's more likely (this is a stab in the dark) that you're running a script to regerate .pyc files; these are precompiled python bytecode files that are built from the corresponding .py files. That's a part and parcel of readying Zope for production - however, Zope will run without those .pyc files (the .py files are compiled on first load instead). I suspect that that's what's going on. Cheers, jan -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Usenet: The separation of content AND presentation - simultaneously. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Real-time command history sharing between interactive shells
On Thu, 30 Nov 2006, Andrew Pantyukhin wrote: > I think, I'll follow your advice. It's high time I forgot about > csh, but I wonder if you tried to change root's shell to zsh? You can; I haven't. ("exec zsh" is simple to type.) sudo works well for single commands. I don't tend to spend much time as root, but that's a question of personal taste. Cheers, jan -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Spreadsheet through network. Oh yeah. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Real-time command history sharing between interactive shells
On Wed, 29 Nov 2006, Andrew Pantyukhin wrote: > I want to be able to define groups of interactive > shells (preferably even across different users) > so they have one single shared command history. > Any command executed in one of them should be > available through all history mechanisms in the > other ones. > > I imagine some ways to do it in tcsh. I'm sure > many users would like this kind of functionality, > maybe some of them have already implemented it? zsh is a pretty good interactive shell (it finally weaned me off tcsh), as well as supporting a full range of redirection and control constructs. You should look at that, in particular the set -o sharehistory option (which does half of what you're after). Combine this with a shared .history file and you should get the effect you're after. -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Whenever I see a dog salivate I get an insatiable urge to ring a bell. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: What Is the PATH to cyrus-sasl2?
On Wed, 15 Nov 2006, Rachel Florentine wrote: > Hi; > I'm building openldap from source since I can't figure out how to pass > arguments to the port. I need to build with cyrus-sasl2, which is > built. > However, I don't know what the path is, and my build can't find it by > itself. Please help. TIA, Rachel The generic approach to determining what a port has installed and where is this: pkg_info -L {package name} jan ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Questions on first-time installation
On Fri, 10 Nov 2006, Jerry McAllister wrote: > On Wed, Nov 08, 2006 at 08:42:04AM +0000, Jan Grant wrote: > > > On Tue, 7 Nov 2006, Bob Schwartz wrote: > > > > > ... at least as it is desribed in the DELL > > > docs and bios... > > > > Be really careful. The windows boot loader should be able to boot a > > freebsd install for you. > > I have never had a Windows MBR work to boot FreeBSD. On the other > hand I have never had a FreeBSD MBR fail to boot any MS-Win system. > > So, I wonder at your comment here. That's a case of your mileage varying, I guess. > I have never had to correct any partition/slice alignment when using > the standard FreeBSD install. I ignore that Diagnostic slice that > Dell puts on - just leave it there as slice 1, leave MS there as > slice 2 and put FreeBSD on slice 3 - or sometimes I put a FAT32 on > as slice 3 to use for communication if the MS is NTFS, and then put > FreeBSD on slice 4. There has never been any problem with any > alignment or offset. [...] > I have been doing this the way I describe since 1998 with no problem. So to the OP, there you go: use of partition magic to manage the fdisk label apparently works. Cheers, jan -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Leverage that synergy! Ooh yeah, looking good! Now stretch - and relax. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
RE: Questions on first-time installation
On Wed, 8 Nov 2006, Bob Schwartz wrote: > >>Be really careful. The windows boot loader should be able to boot a > freebsd install for you.<, > > Thank you. > > Either one works for me as long as I get full choice. If the windows loader > simply picks up bsd and offers it in addition to the two windows installs > that I need, no problem. Ditto if the bsd boot manager hands me off to the > windows boot manager and gives me both choices. > > My concern...and I was going to write the group and ask...is that bsd's > manager only returns one of the windows systems. > > What do you think, please? The dell box I have at home boots using the NT/XP boot manager, that Just Works. > >> Dell usually ship their machines with a small partition at the front of > the drive (at least, they did last time I received one). That partition is > at an odd offset and didn't agree with the FBSD partition editor: I found > this out the hard way. Things may have changed since then; I usually nuke > the dell partition anyway.<< > > They still do...and while I could nuke it, now that I have windows installs > that I don't want to disturb, I am afraid to. > > If I leave them there, what, in your experience, happens? Like I said, this experience is a few years old, but the fbsd partition manager wound up having problems and screwed the partition table. On the other hand, if you create an fdisk partition inside windows and don't do any fdisk editing from the freebsd install, just label and install into the precreated slice, you should be ok. Usual provisos apply: back up, and if at all possible try it first on some scratch kit. G'luck, jan -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Leverage that synergy! Ooh yeah, looking good! Now stretch - and relax. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
RE: Questions on first-time installation
On Tue, 7 Nov 2006, Bob Schwartz wrote: > ... at least as it is desribed in the DELL > docs and bios... Be really careful. The windows boot loader should be able to boot a freebsd install for you. Dell usually ship their machines with a small partition at the front of the drive (at least, they did last time I received one). That partition is at an odd offset and didn't agree with the FBSD partition editor: I found this out the hard way. Things may have changed since then; I usually nuke the dell partition anyway. jan PS. I should say that my bad experience with the partition editor is a few years old; I've not tried again recently with new Dell kit. -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Axioms speak louder than words. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: file redirect is destroying the file ? Help!
On Mon, 30 Oct 2006, Ensel Sharon wrote: > I have a script that, among other things, removes a line from > /etc/ftpchroot. > > I do this with this method: > > > cat /etc/ftpchroot | grep -v $remove > /etc/ftpchroot > > > Easy. You cat all of the file except the line you want to remove, and > redirect it back to itself. > > The problem is, about 50% of the time, I end up with an empty ftpchroot > file. It is zero bytes. This obviously has nothing to do with a bad > variable, since if it wasn't there, the starting file and ending file > would just be identical. > > Instead, I get an empty file. I have reproduced this with other files in > other places - works some of the time, other times gives me an empty file. > > What gives ? > > (note, I know a lot of ways to work around this - so I'm not so much > asking how to fix this, as I am asking "why does this happen" ?) It happens because the two processes either side of the pipe are competing for CPU resource. If the "cat" gets to run first then it'll slurp in the contents of the existing file (which will probably fit into one stdio buffer) and everything's hunky-dory. If the other half gets to go first, then the shell will open the file for output, truncating it in the process, and exec grep. At that point the "cat" will have another go and find nothing there. Basically, what you've written is buggy and you need to use one of your workarounds. Cheers, jan -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ stty intr ^m ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: tao.thought.org is back.....
On Wed, 18 Oct 2006, Gary Kline wrote: > On Thu, Oct 19, 2006 at 12:57:08AM +0300, Giorgos Keramidas wrote: > > By setting the "MX" records for `thought.org' up so that the one with > > the lowest value of them all points to that specific server. > > > > This should be configured in the name-server which hosts the DNS zone > > for `thought.org'. > > This would work [or ought to!]; right now, all my mail echange > entries are equi-valued at 50. But this seems like a back door > way of dealing with sendmail. I'm the first to admit that it's a > less tha optimal suite, obscure beyond words, (etc). But I'd > like to understand how to resolve this problem with senmail I would have suggested the DNS approach except that when I looked, tao.thought.org didn't exist in your DNS (as an A record) and since ns1 looked like a dialup account, I assumed you're using NAT. Your mailertable entry on ns1 should probably just read thought.org relay:tao.thought.org or maybe thought.org smtp:tao.thought.org and tao.thought.org should have "thought.org" in it's w class: that is, sendmail.cw or local-whatever-the-file-is-called. You might need "makemap" to recreate the database from your plain file. I suspect that you have an alias set up on ns1 that you're fiddling with. In order to get it to stick, you probably need to run "newaliases". Since that's run on reboot if required, that'd explain your reboot-to-reconfigure behaviour. Cheers, jan PS. If sendmail config is such an arcane art (and it is, unless you do it regularly) you might find there's mileage in installing an alternative MTA. -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Goth isn't dead, it's just lying very still and sucking its cheeks in. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: tao.thought.org is back.....
On Wed, 18 Oct 2006, Gary Kline wrote: > My one question is given that mail defaults to my > ns1.thought.org, HOW can I get it to go to (say) zen.thought.org, > or to tao.thought.org? or to ethos.thought.org? OR what > re-initialization do I have to do? other than a shutdown -r > now?? Sounds like you need a mailertable entry that maps your incoming domain name to smtp:whatever.thought.org jan -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ New Freedom of Information Act: theirs, to yours. Happy now? ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: How do I give 2 parameters to programs in an unix enviroment?
On Mon, 11 Sep 2006, hackmiester (Hunter Fuller) wrote: > On 8 September 2006, at 08:10, Lasse Edlund wrote: > > > If I have two files "foo" and "bar" and try to run diff on them I write: > > $diff foo bar > > I can also write > > $cat foo | diff - bar > > But how do I give a program two (2) commands? not only to diff > > but to any program that wants double input... > > I wanna do > > $cat foo | cat bar | diff - - > > The entire purpose of cat is to concatenate files (make them output one after > another). So, do: > > cat foo bar | diff - - This advice is wrong. To answer the original question: the shell pipe connects the stdout of the first process to the stdin of the second process using a pipe. The stock shells don't have a way of doing what you're after. If you have fdescfs mounted, ksh can do something like what you're after using the syntax: diff <(cat foo) <(cat bar) zsh supports something similar and can work around the lack of fdescfs. -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ ( echo "ouroboros"; cat ) > /dev/fd/0 # it's like talking to yourself sometimes ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: man page bug in mv(1) ?
On Sun, 3 Sep 2006, James Long wrote: > The man page mv(1) states: > > "It is an error for either the source operand or the destination path > to specify a directory unless both do." > > > However: > > mv file /tmp/ > > works. Am I reading things wrong, or is the man page incorrect? The man page is correct, but a little misleading unless you read it carefully. The first two paragraphs of the description section define the term "destination path". -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Solution: (n) a watered-down version of something neat. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Cluster mail system using FreeBSD
On Wed, 23 Aug 2006, bsd wrote: > The idea is to offer a simple and very efficient solution in order for the > server to have a 100% uptime under any circumstances. That goal is not realistic. > I was first thinking about using Linux-HA aka. "heartbeat" and syncing > the two boxes using rsync ? One question is whether your uptime definition means "someone can connect via IMAP" or "someone can connect via IMAP and IMAP acknowledgements of state-changing operations are guaranteed to be replicated across the cluster". If you actually want your cluster's semantics to preserve email client operations in the face of the loss of one member of the cluster then rsync is not enough. > What other solution would you think of ? You might want to chase down the Cambridge patches to cyrus, which added an application-level transaction log to that particular imap server. -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Leverage that synergy! Ooh yeah, looking good! Now stretch - and relax. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: /tmp permissions
On Mon, 7 Aug 2006, dick hoogendijk wrote: > Today I read that /tmp always is "noexec". > That should probably be on linux, because on my fbsd-6.1 box it's "rw" > and that's it. > > Question: should I change /tmp to "rw,noexec" to be safer? Not necessaily, but you might want to consider adding "nosymfollow" -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Whenever I see a dog salivate I get an insatiable urge to ring a bell. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Need /bin/sh script help
On Tue, 11 Apr 2006, [EMAIL PROTECTED] wrote: > On Mon, 10 Apr 2006 22:30:32 -0700 Garrett Cooper wrote (my brief response > follows all of his text): > > Just making a series of sh scripts to help automate updating and > whatnot of my fileserver (since I am trying to avoid having mistakes > occur with my system, and maybe help the community out a bit by > providing some decent means of updating their own machines), and I was > wondering if anyone could help me out with the following script I've > developing (the grep if statements are incorrect..): > I see a problem in the line > if [ -n `grep -e s/KERNCONF=/ /etc/make.conf` ] # want to look for > you should have double-quotes around the `grep ... conf` > because it is likely to produce more than one token and so the > [ -n ... ] statement violates the syntax (there should be exactly 1 > token between the -n and the ] , even no token there is an error, the > way that is handled is to quote it. I am writing this quickly without > bringing up my FreeBSD system to check it. Good luck. Or simply use the error status of grep, no "[" invocation: if grep -q ... then ... fi Note that if you're looking at automating the update process you should probably pay careful attention to the world/kernel update process described in the handbook; there are steps (like an initial mergemaster -p) that you will want to include. -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ If you have received this email in error, do whatever the hell you want with it. It's not like I can stop you anyway. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: UFS extended attributes
On Mon, 10 Apr 2006, Robert Watson wrote: > > On Sun, 9 Apr 2006, Duane Whitty wrote: > > > Started doing a little reading on the UFS and UFS2 file systems. I'm just > > wondering if all types of files have extended attribute blocks available > > including named pipes, sockets, and device files? > > > > Is it still the case that there are three unused extended attribute blocks > > available? > > Extended attribute storage is available for all objects in UFS, including > files, directories, named pipes, UNIX domain sockets, and device nodes. I'm > probably not th eright person to answer questions about the layout itself. Looks like it is: see /usr/include/ufs/ufs/dinode.h -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Theoremhood is positively decidable. It just takes time at least exponential in the length of the proof. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Shell scripting question [newby]
On Mon, 10 Apr 2006, Malcolm Fitzgerald wrote: > > On 10/04/2006, at 12:39 AM, Jan Grant wrote: > > > On Sun, 9 Apr 2006, Malcolm Fitzgerald wrote: > > > > > I'm trying to follow the instructions at > > > <http://www.daemonology.net/freebsd-upgrade-5.4-to-6.0/> > Your advice got me to step 7 where the need to pass a control structure to the > loop stopped me again. > > I got a bash shell and I write: > > for dist in base dict doc games info manpages ports; do > cat /mnt/6.0-RELEASE/${dist}/${dist}.?? > /usr/${dist}.tgz > done > > I put it onto three lines by typing "\" at the end of each line to achieve the > layout and I get the prompt ">". When I get to the end, ie, "done" I press > Enter and get another prompt. > > How can I get the multi-line command executed? What you're doing is roughly this: (note, I supplied a separate "done", you'll see why) [[[ $ for i in one two three; do \ > echo $i \ > done > done one done two done three done $ ]]] the first "done" is counted as part of the "echo" argument list. If you terminate a line with a "\" character, then the intervening newline is treated as "just whitespace". Consequently, were you to use this syntax, you'd need to punctuate your script properly: for i in one two three; do \ echo $i; \ done Having said that, you don't need the "\" marks, because bash (and sh) are smart enough to parse as you go and keep prompting until the command is complete. Thus you can just type: for i in one two three; do echo $i done and it'll do what you're asking. Cheers, jan -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Strive to live every day as though it was last Wednesday. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: OT - Scalable email server solution needed
On Sun, 9 Apr 2006, Bill Moran wrote: > I have only a small amount of experience with Cyrus. However: > http://cyrusimap.web.cmu.edu/twiki/bin/view/Cyrus/Backup > > Based on that document, it appears as if you're dodging the bullet with > backups. My interpretation is that Cyrus keeps mailboxes in some sort > of db file. If a db file is being modified while you're backing it up, > the backed up version will be inconsistent, thus the entire mailbox > unusable. Not quite. Cyrus maps the imap hierarchy onto directories, with one (written once only) file per message. As flat-file as you can get. On top of this are per-folder index files (which can be recreated using reconstruct) holding stuff like preformatted IMAP responses, header indexes etc. That's where cyrus gets its speed from. Additionally there's a per-server "mailboxes" DB holding folder information (including ACLs), per-user seen/subscription databases and a deliverdb for duplicate suppression. The latter can all be stored in multiple formats, including bdb; hot backups for these work much the same way as for anything other bdb. It's not perfect, but there's minimal bullet-dodging. jan -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ ioctl(2): probably the coolest Unix system call in the world ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Shell scripting question [newby]
On Sun, 9 Apr 2006, Malcolm Fitzgerald wrote: > I'm trying to follow the instructions at > <http://www.daemonology.net/freebsd-upgrade-5.4-to-6.0/> > > At point four it offers this shell script. > > cut -f 1 -d '$' /usr/local/freebsd-update/work/md5all | uniq | > while read X; do > if [ -f $X ]; then echo $X; fi; > done | sort > /root/base-old > > Running this from root shell in konsole (bash) I get "while: Expression > Syntax". The various hints and clues I get from the shell, the web and man > bash haven't helped me. Would someone provide the correct syntax for me? That syntax is correct for sh and bash; you're not running it, however. Double-check that after you su to root, you're really running bash. That error is what csh will tell you. -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Whenever I see a dog salivate I get an insatiable urge to ring a bell. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: /proc/loadavg?
On Fri, 7 Apr 2006, Andy Greenwood wrote: > Does FreeBSD have a location that stores the load average information, > similarly to /proc/loadavg in Linux? I've got a php site that displays this > info, but I'm not sure where to point it to. There's a "linuxprocfs" which you might want to mount on /compat/linux/proc, that should have waht you're after. -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ "No generalised law is without exception." A self-demonstrating axiom. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Why are so many people using 4.x?
On Tue, 28 Mar 2006, Tom Grove wrote: > I would certainly recommend going with 6.x. The reason that many of our > servers still run 4.x is that 5.x got a bad reputation and there really is no > upgrade path from 4.x to 6.x. 5 and 6 default to using UFS2 and 4 uses UFS > so, IMHO it's better to rebuild and taking a few hundred users offline for a > couple of hours whilst this happens isn't fun. > > That's my scenario...I'm sure others have totally different reasons. In addition to Kris' comments about UFS being perfectly viable for 5.x and 6.x: there is an upgrade path, but it's 4.x -> 5.x -> 6.x. FWIW I've done this successfully without a hitch*. jan * Having said that, I use a liveupgrade-a-like setup with a primary / and /usr (that I'm running from) and a secondary (that I rebuild into and reboot into). It means I have something solid to fall back to if the upgrade fails. -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ That which does not kill us goes straight to our thighs. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Shellscript syntax question
On Mon, 9 Jan 2006, Frank Staals wrote: > About the asteriks : No they weren't but for some reason Thunderbird had > problems with the color remaining from the KATE Syntax-hilighting. Anyway: > thanks for the solusion, but what would be the expressions for ">" "less or > equal" and "greater or equal" ? As others have pointed out: [ is a synonym for test(1), which uses -gt, -ge, -eq, -ne, -lt, -le for numeric comparisons. For string comparisons, it uses <, !=, =, and >. Your original expression didn't work because "<" is also a shell meta-character used to indicate file redirection. To get something like that to work, you'd need to quote it somehow: if [ "$a" \< "$b" ]; then ... ; fi or if [ "$a" '<' "$b" ]; then ... ; fi -- jan grant, ISYS, University of Bristol. http://www.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ I'm the dandy information superhighwayman. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Fast diff command for large files?
On Fri, 4 Nov 2005, Kirk Strauser wrote: > > I wonder if rsync could be modified to output its patches rather than > silently applying them to a target file. It seems to be pretty good at > comparing large files quickly... > More thinking out loud: since these are database dumps, they're order-independent. So, sort the files into a predictable lexical order, then a diff is a linear operation over tonight's file versus last night's. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Rereleasing dolphins into the wild since 1998. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: portupgrade -ar (why?)
On Sat, 15 Oct 2005, Peter Matulis wrote: > What is the use of specifying the 'r' switch when using the 'a' > switch? > > # portupgrade -ar > > This says to upgrade all ports plus the ones that depend on all > those ports. Am I missing something? Wouldn't "the ones that > depend" be upgraded anyway? Not necessarily. For instance: package P might use library L. A change in L might alter the size and layout of structures exposed to P. The source-level API of L is unchanged; the binary-level ABI is altered. So whilst the source code of P might not have changed, it might (for instance) be using a macro defined by a header in L that will look at the wrong offset in the new structure. These kinds of ABI compatibility problems can be fixed by recompilihng P. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Strive to live every day as though it was last Wednesday. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Help Understanding While Loop
On Fri, 14 Oct 2005, Will Maier wrote: > > count = $(( count + 1 )) > ^^^ > > You're missing something here ;) Actually, you don't need the leading "$". With it, you'll get variable interpolation, which'll normally do the same thing. Once the expression has been interpolated, it's evaluated, using the values of any variables named in the expression. That is, count=1 count=$((count + 1)) will do what Drew intended. I think his problem is elsewhere (the while loop is a subshell). -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ "I think therefore I am." -- Ronnie Descartes ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: sysctl or system tweak for symbolic links?
On Fri, 9 Sep 2005, Forrest Aldrich wrote: > Hi, > > Using FreeBSD_6.0_Beta4 (applies to Beta2, also). > > I'm trying to track down a problem I've been having with apache-2.0.54 not > following symbolic links. > > It's basically come down to my being able to follow the link if it's in the > same directory structure (ie: .. or /path/to/..), but fails if the symbolic > link is located elsewhere (ie: /usr/local/path/directory) or on another disk. > > I wonder if there's a sysctl or other system variable that handles the > behavior of or access to symbolic links in this fashion that I may have > missed. Seemed like a reasonable conclusion after these tests have been > failing, though it could be something else, too. Stock apache configuration. Look at the documentation for the "Options" directive, particularly "SymLinksIfOwnerMatch". -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Just because I have nothing to hide doesn't mean I have nothing to fear. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: What is fsck trying to tell me?
On Fri, 9 Sep 2005, Leonard Zettel wrote: > When I issue the followinf command: > > mount /dev/ad1s1c /mnt > > I get the response > WARNING: R/W mount of /mnt denied. filesystem is not clean - run fsck > mount: /dev/ad1s1c: Operation not premitted > > Then when I try > > fsck /dev/ad1s1c > > I get > fsck: exec fsck_unused for /dev/ad1s1c in sbin: /usr/sbin: No such > file or directory > > BTW, mount -f /dev/ad1s1c /mnt > > gets me what I expect, but the hassle leading up to it has > me scared to death. Now what? punt? You're using the default "whole slice" partition, ad1s1c. My guess is, you're using the default disklabel for that slice. If you look at that disklabel, # disklabel ad1s1 you'll see a line like this: c: 1563014250unused0 0 # "raw" part, don't edit Now, fsck uses external helper utilities to check the consistency of various types of filesystem. If the filesystem has an entry in /etc/fstab, it'll pull the type from there if you specify the mount point. If you specify the device, it looks like fsck is using the disklabel rather than actually "tasting" the partition to determine what fsck to use. You can fix this by disklabelling your device and fixing the type of partition "c": this should be ok. You can probably also tell fsck explicitly what type of filesystem to check, or just invoke the appropriate fsck_ufs directly. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44 (0)117 3317661 http://ioctl.org/jan/ Usenet: The separation of content AND presentation - simultaneously. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
RE: BSD legal question
On Thu, 19 May 2005, Ted Mittelstaedt wrote: > Suppose I distribute a library that is under my own copyright, > yet carries a BSD-like license. > > Suppose you then come along and take my library, and a GPLed > library, link both of them together into a new program of yours. > > The FSF says that the entire code now becomes GPL. That's not true. The GPL requires you to license any distributed code derived from GPLed code under the GPL. Since, as you point out... > The problem here is that since you never owned copyright on > my library, you do not have legal rights to modify the copyright > and license on it. Thus, you cannot legally apply GPL to it. > Nor can the FSF or anyone else apply GPL to it. ... the conclusion is that you cannot *distribute* the derived program; NOT that it magically relicenses code you've used to build it. > Naturally, the parts of the program you are distributing that > YOU wrote are under GPL. But, suppose another guy comes along, > takes your program, and pulls my library out of it and uses it > for his program. According to the GPL, if he does this that > library is GPLed now, and his program must be GPL. According > to the law, though, it's not GPL, it's mine. The problem is this: "the program you are distributing". The GPL prevents you from distributing such derived code for exactly the reasons you point out. You're free to do whatever you like with it, except distribute. In other words, by distributing the derived code, you're in breach of the GPL. (The exceptions arise depending on whether the non-GPLed license form a core part of functionality or whether the resulting program is functional without the hypothetical TedLib.) jan PS. This is basic implication: you claim P = "you distribute the code" Q = "all of it is under the GPL" and P -> Q. I'm not taking issue with that; merely pointing out that (as you describe), Q is false, therefore since -Q -> -P the conclusion is that you can't distribute. Incoherently yours, etc. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44 (0)117 9287088 or 3317661 http://ioctl.org/jan/ Theoremhood is positively decidable. It just takes time at least exponential in the length of the proof. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: PF RULES! But mine doesn't ...
On Tue, 10 May 2005, Fafa Hafiz Krantz wrote: > Ok, after having added that it seems that my DNS works. > The same goes for my WWW and mail server. > > SSH servers are all OK to connect to. > > I have to wait like 5 minutes after booting my computer > before I can connect to those certain FTP sites. What's > that all about? > > > If you add the "query-source address * port 53;" to your named.conf > > "options" section, that'll suffice; additionally, since your DNS query > > source port is then predictable, you can drop it from the DNS and NTP > > rule. > > What do you mean by that? The rules I suggested are so that external machines can talk to your DNS server (querying about the domain it is authoritative for), and so that responses can get back to those machines. Your nameserver, however, may also be trying to get requests out. When it does this, by default, it will use a random source-port. By specifying options { query-source address * port 53; } in your named.conf, your nameserver will _also_ use port 53 as the source port on any requests _that it originates_. (That's the distinction). If you do this, then you won't need port 53 mentioned in your other "keep state" rule. I suspect that this might actually be the cause of your transient FTP concern; you should try modifying your nameserver config before you go any further. (This assumes that your resolv.conf is configured to use the local machine as a nameserver in the first instance. If that is not the case, then you will still need the port 53 clause in your "DNS and NTP" section, because other programs will use random ports in an attempt to get DNS queries out into the wild.) > Anyway, it's pretty close to perfection now :) > > Jan, any idea how I can simplify my ruleset? > Also, I'm wondering if I can move the NAT part down below the Outgoing > so I can combine it with the Active FTP ruleset so they don't have to be > spread troughout the conf. Thanks! Your ruleset looks pretty simple, to be honest. I'm afraid that where the specifics of PF are concerned, I know nothing: the advice I've given you is just generic firewall stuff :-/ It looks to me like your PF config is set up to use some kind of FTP proxy running on localhost:8021. On the other hand, I could be barking up the wrong tree completely; I've pretty much run out of useful things to say about this config. Cheers, jan -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44 (0)117 9287088 (with luck) http://ioctl.org/jan/ Prolog in JavaScript: http://ioctl.org/logic/prolog-latest ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: PF RULES! But mine doesn't ...
On Sun, 8 May 2005, Fafa Hafiz Krantz wrote: > Hello. > > My ruleset is all twisted. > Unless I disable the default deny policy, this is what happens: > > * My nameserver setup goes disfunctional. > * My web, mail and fileserver goes disfunctional. > * I cannot SSH and FTP into certain servers. > * I cannot ping my IP from the outside. > > Can anyone tell what's wrong? > And maybe also how I can simplify my ruleset? It's a question of letting DNS traffic _in_ to your nameserver: > int_if="ep0" > ext_if="lnc0" > > # *** Options > # > set block-policy drop > > # *** Scrub incoming packets > # > scrub in all > > # *** NAT > # > nat on $ext_if from $int_if:network to any -> ($ext_if) > rdr on $int_if proto tcp from any to any \ > port 21 -> 127.0.0.1 port 8021 > > # *** Default deny policy > # > # block drop log all > > # *** Pass loopback traffic > # > passquick on { lo0 $int_if } > > # *** Outgoing > # > passout on $ext_if inet proto { tcp, udp, icmp } \ > from ($ext_if) to any keep state > > # *** Bootstrap > # > passout on $ext_if inet proto udp \ > from any port 68 to any port 67 keep state > > # *** DNS and NTP > # > passout on $ext_if inet proto udp \ > from ($ext_if) to any port { 53, 123 } keep state > > # *** SSH, HTTP and Ident > # > passin on $ext_if inet proto tcp \ > from any to ($ext_if) port { 22, 80, 113 } flags S/SA keep state pass in on $ext_if inet proto { tcp, udp } \ from any to ($ext_if) port 53 ^^^ that lets the traffic in pass out on $ext_if inet proto { tcp, udp } \ from ($ext_if) port 53 to any ^^^ and that lets it back out. If you add the "query-source address * port 53;" to your named.conf "options" section, that'll suffice; additionally, since your DNS query source port is then predictable, you can drop it from the DNS and NTP rule. > # *** Active FTP > # > passin on $ext_if inet proto tcp \ > from port 20 to ($ext_if) user proxy flags S/SA keep state -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44 (0)117 9287088 (with luck) http://ioctl.org/jan/ Usenet: The separation of content AND presentation - simultaneously. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: What is --- WRONG --- with my network?
On Fri, 6 May 2005, Fafa Hafiz Krantz wrote: > > Hello boys! > > I just spent a few days doing a make world and kernel. > My machine is terribly slow. Yet, my network problem hasn't > gone away. This shows that it wasn't an asynchronisation > between my world and kernel. I've also compiled io and mem > into my new kernel. > > Here is my problem description: > > * My nameserver setup is disfunctional. > * My web, mail and fileserver is disfunctional. > * I cannot SSH and FTP into certain servers. > * I cannot ping my IP from the outside. > * My ISP controls the PTR of my reverse DNS lookup. >This fails to resolve too. > > PF is disabled. > > My configuration has been running flawlessly for the past > few months before this strange happening occured. I know for > a fact that this is not related to a misconfigured rc.conf > or named.conf. Maybe it is my ISP? Unfortunately, you appear to be preemptively rejecting the most obvious advice, and implicitly asking that people start troubleshooting "from the middle". Was your machine up without reboot for months? If so, there's no guarantee that the state of named.conf actually reflects the state of the previously running named prior to a reboot (alas, I've seen this all too often). Can you begin by posting your fully-functional named.conf and resolv.conf? And possibly describing exactly what you mean by "my nameserver is dysfunctional"? Do you mean that you cannot resolve addresses from your host? Does dig work against your local nameserver instance? Can you see any of the root servers with dig? Is named just refusing to start? ... and so on. You'll probably have to be more explicit about "certain servers", too. Are they on-site? Off-site? If the latter, and the issue only appears with a subset of ssh servers, this may well be indicative of DNS problems again, since sshds can be configured to be more or less picky about the name resolution of their clients. Have you tried to resolve the PTR record for your IP address from offsite? If this is failing, it's possibly the root cause of a lot of your problems, and you'd need to raise it with your ISP. Cheers, jan ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: KDE not starting. Cannot start kdeinit
On Sat, 16 Apr 2005, Danny Pansters wrote: > IIRC exec used to be needed for some reason, but that's been a while, merely > startkde (which is just a shell script) will do. The reasoning is in the man page for startx(1): [[ The .xinitrc is typically a shell script which starts many clients according to the user's preference. When this shell script exits, startx kills the server and performs any other session shutdown needed. Most of the clients started by .xinitrc should be run in the back- ground. The last client should run in the foreground; when it exits, the session will exit. People often choose a session manager, window manager, or xterm as the ''magic'' client. ]] Since the session stops when the .xinitrc process dies, you can either launch startkde from within the shell script (in which case the .xinitrc script waits until startkde exits, then exits), or you can use "exec startkde" to replace the .xinitrc process with the startkde one. In either case the only difference is that leaving out the exec means you've got one additional shell process hanging around until you quit X. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44 (0)117 9287088 (with luck) http://ioctl.org/jan/ "I like oranges more than apples!?" - that's like comparing apples and oranges! ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Is this a good or a bad idea to delete all off this ?
On Tue, 15 Mar 2005, Michael C. Shultz wrote: > On Tuesday 15 March 2005 06:17 am, Tom Trelvik wrote: > > Gert Cuykens wrote: > > > can i do controle C when i want to go to sleep while upgrading ? > > > > A cleaner solution, regardless of whatever you're running that you'd > > like to ^c, is to just let it run inside a screen session and > > reattach to the screen session later to see how it went. > > > > Screen should be every sysadmin's best friend. =) > > > > Tom > > Tom, mind providing a link that tells us non-sysadmins how to do this? He's referring to the misc/screen port. The man page is pretty comprehensive. It basically gives you "virtual" terminal sessions that you can attach, detach and reattach a client to without disturbing the underlying session. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287864 or +44 (0)117 9287088 http://ioctl.org/jan/ Random act of violence against bread: whole pint. -- extract from the "Hawk the Slayer" drinking game ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: How do I set the source address on a multi-homed host?
On Tue, 15 Feb 2005, Daniela wrote: > Well, if the OS selects the source IP, can't I just modify the code that > selects it? Will this work all the time, or just when the application lets > the OS select an address for it? It should "just work" - that is, the source address for internal packes should already _be_ that of the internal interface. Your routing table looked good, too, IIRC. What does your NAT / ipfw configuration look like? -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287864 or +44 (0)117 9287088 http://ioctl.org/jan/ "NOP" is a trivial implementation of an executable Z subset. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: How do I set the source address on a multi-homed host?
On Sun, 13 Feb 2005, Jan Grant wrote: > On Sat, 12 Feb 2005, Daniela wrote: > > > Yes, this happens when I connect from my machine (which functions as a > > router > > with NAT to allow the other LAN machines connect to the internet) to > > another > > LAN machine. When the router establishes a connection to another point in > > the > > intranet, the source address used is my official IP, and not 10.0.0.1, > > which > > is the intranet IP of the router. > > In other words, I want the source address to be 10.0.0.1 on every outgoing > > connection where the destination is inside my intranet. > > Assuming you haven't munged the internal IP address to hide it, and with > all due deference to the FreeBSD "mechanism, not policy" mantra: no, you > don't want to do this. Excuse my misinformation. Misread "inside" for "outside". -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287864 or +44 (0)117 9287088 http://ioctl.org/jan/ Solution: (n) a watered-down version of something neat. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: How do I set the source address on a multi-homed host?
On Sat, 12 Feb 2005, Daniela wrote: > Yes, this happens when I connect from my machine (which functions as a router > with NAT to allow the other LAN machines connect to the internet) to another > LAN machine. When the router establishes a connection to another point in the > intranet, the source address used is my official IP, and not 10.0.0.1, which > is the intranet IP of the router. > In other words, I want the source address to be 10.0.0.1 on every outgoing > connection where the destination is inside my intranet. Assuming you haven't munged the internal IP address to hide it, and with all due deference to the FreeBSD "mechanism, not policy" mantra: no, you don't want to do this. The 10.0.0.0/8 block of addresses is explicitly for private use and is not routable on the internet. If your firewall is causing problems with this setup, you might need to re-examine your firewall settings. Having said that: technically, you specify source addresses for connections by calling bind(2) prior to calling connect(2). If you fail to do this, the operating system will select a source IP address for you. This'll often be the IP of the outgoing interface. Unless your particular application explicitly supports the selection of source addresses, you're mostly out of luck. For instance, ping(8) supports this (see the -S option). Cheers, jan -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287864 or +44 (0)117 9287088 http://ioctl.org/jan/ Bolstered by my success with vi, I proceeded to learn C with 'learn c'. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: what is /entrophy ?
On Sun, 6 Feb 2005, Jay Moore wrote: > On Wednesday 02 February 2005 05:52 am, Gert Cuykens wrote: > > > what is /entrophy ? can i delete it ? > > I believe it is a mis-spelled version of /entropy Your computer attempts to collect "randomness" by sampling the timings of various physical events. That's what the /dev/random device provides: this kernel-harvested randomness. Various cryptographic systems require a supply of "good" random numbers in order to operate. When the machine first boots, the kernel's entropy pool is empty. It would consequently take potentially quite a few minutes to harvest sufficient randomness from interrupts in order to satisfy the needs of such things as sshd. The solution is the /entropy file: when the machine shuts down, it saves "spare" random bits that have not yet been used into this file. On reboot, the kernel's random pool is reinitialised using these "spare" bits. Assuming nobody's sneaked a peek at them in the itme the machine's been turned off, this is a reasonable way to quickly satisfy the startup requirements for randomness. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287864 or +44 (0)117 9287088 http://ioctl.org/jan/ You see what happens when you have fun with a stranger in the Alps? ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: 300Gb hard drive formatting to 249Gb - boo.
On Mon, 6 Dec 2004, Alex Teslik wrote: > [root]/home/alex# df -h > FilesystemSize Used Avail Capacity Mounted on > /dev/ad1s1e 271G 1.0K 249G 0%/1 This is an FAQ: see http://www.freebsd.org/doc/en_US.ISO8859-1/books/faq/disks.html#DISK-MORE-THAN-FULL and the -m option of tunefs(8): http://www.freebsd.org/cgi/man.cgi?query=tunefs&sektion=8 Summary: the filesystem needs "elbow room" to operate under typical load; this 8% space is reserved. (It may be consumed by the root user, if required.) Tuning down the reserved percentage is not recommended unless you know what you're doing (that is: benchmark your typical usage to see if the performance is still within acceptable limits). -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287864 or +44 (0)117 9287088 http://ioctl.org/jan/ You know something's gone badly wrong when your algorithm takes O(n^2) time but uses O(2^n) space. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Is my Apache server running as the root user or not?
On Sat, 4 Dec 2004, csnyder wrote: > On Sat, 4 Dec 2004 04:47:49 -0500, Peter C. Lai <[EMAIL PROTECTED]> wrote: > > This isn't on-topic for the list, but I'll answer it anyway. The Apache > > parent runs as root so that it can attach to port 80. > > [...] > > > > On Sat, Dec 04, 2004 at 07:24:27AM +0100, Jesper Wallin wrote: > > > [...] > > > If I got a lot of users connecting to my server on port 80, will > > > thier requests ever be answered by the root fork or the www:www forks? > > >[...] > > If this is really a nagging question, enable the mod_status in > httpd.conf, then watch http://yourserver/server-status while you pound > the server with apache/bin/ab -- you'll see exactly which PIDs are > answering queries. There's a description of what you're seeing here: http://httpd.apache.org/docs-2.0/mod/prefork.html which, although it's apache 2 documentation, also describes the apache 1.3 preforking behaviour. The root process creates sockets and log files, then forks child worker processes which retain the open file handles. However, it is only the child processes that ever call "accept": ps axlw | grep httpd -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287864 or +44 (0)117 9287088 http://ioctl.org/jan/ "I think therefore I am." -- Ronnie Descartes ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: shell programming challenge
On Thu, 25 Nov 2004, Don Wilde wrote: > J65nko BSD wrote: > > On Thu, 25 Nov 2004 10:26:38 -0700, Don Wilde <[EMAIL PROTECTED]> wrote: > > > > > Hey, folks - > > > > > > I need to find a way to kick off an xterm running BASH and then execute > > > a program within that xterm, but NOT close the new xterm after the > > > program finishes. > > [snip] > > > > > xterm -hold -e sh -c ./ticktock & > > > > "-hold" is explained in the xterm man page ;) > > > Yes and no. I don't get as command prompt back in the new xterm, either with > sh or bash. ALso, the --rcfile arcument does not appear to work in conjunction > with -c. > > This is what the man for hold specifies, but I need the prompt. If you have the option to modify it, ensure that your script exits via "exec sh". Alternatively a wrapper that does this is straightforward to build. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287864 or +44 (0)117 9287088 http://ioctl.org/jan/ Political talk? / What is said can be unsaid / with good old BS -- ASCII haiku ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Backup/Restore
On Fri, 1 Oct 2004, Brian McCann wrote: > On Thu, 30 Sep 2004 13:59:05 -0700 (PDT), Richard Lynch <[EMAIL PROTECTED]> wrote: > > > > Brian McCann wrote: > > > Hi all...I'm having a conceptual problem I can't get around and > > > was hoping someone can change my focus here. I've been backing up > > > roughly 6-8 million small files (roughly 2-4k each) using dump, but > > > restores take forever due to the huge number of files and directories. > > > Luckily, I haven't had to restore for an emergency yet...but if I > > > need to, I'm kinda stuck. I've looked at distributed file systems > > > like CODA, but the number of files I have to deal with will make it > > > choke. Can anyone offer any suggestions? I've pondered running > > > rsync, but am very worried about how long that will take... > > > > Do the files change a lot, or is it more like a few files added/changed > > every day, and the bulk don't change? > > > > If it's the latter, you could maybe get best performance from something > > like Subversion (a CVS derivative). > > > > Though I suspect rsync would also do well in that case. > > > > If a ton of those files are changing all the time, try doing a test on > > creating a tarball and then backing up the tarball. That may be a simple > > managable solution. There are probably other more complex solutions of > > which I am ignorant :-) > > I have the case where a new file is created about every second or two, > nothing gets changed, but files get deleted occasionally (it's a mail > server). I thought of using tar, but it would be just as slow as dump > I would think. I've thought of breaking it up into chunks, but that > still doesn't solve my speed issue...i'm beginning to consider using > dd since it reads the actual disk bits, and just hope that a)I don't > ever need one file and b) the system I restore to has at least or more > space then the original server. Any other thoughts anyone? You might want to experiment with something like rsync to maintain a "live" (ie, on a FS) second copy. If you do this don't be put off by the initial rsync time (which may well take ages - tar or dump/restore may be faster to get the second copy in place initially). Rsync over such a large filesystem may take quite a while but the best bet is to actually try it to see if it meets your needs. Obviously a restore of a mail repository is a pretty awful thing to have to do. Amongst other things, users can find the "ressurrection" of deleted mails to be a real pain. You might want to see if your mail repo can generate some kind of replay log - if so, this might be the best route for minimising the amount of time needed to synchronise mailstores and to get the closest fidelity out of the copy. Breaking your mailstore into separate chunks may well help. Yes, the total time for a dump/restore may be close to your current state of play, but if you can split the partitions between machines then you have the option to perform these in parallel. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ "...perl has been dead for more than 4 years." - Abigail in the Monastery ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: White Paper on routing & freebsd
On Tue, 21 Sep 2004 [EMAIL PROTECTED] wrote: > Hello, > > I am looking for a good white paper or tutorial that will explain how to > configure routing for freebsd. > > A good pointer for understanding linux & static routing will be apreciated > too. > > I have already read the HandBook, but there are still couple of things that I > haven't catch. Are you after the basic theory behind routing? If you have specific questions then this list is as good a place as any to ask them. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Whenever I see a dog salivate I get an insatiable urge to ring a bell. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: what are the pros and cons of running in single user?
On Thu, 16 Sep 2004, Ed Budd wrote: > mailing lists at MacTutor wrote: > > I have a machine running 4.10-STABLE that will be a dedicated gateway with a > > router/firewall combo and web server plus mysql server (maybe). What would > > be the pros and cons of running this system single user? Processes that run > > under their own uid, would they be able to run? > > > > Just curious. Any extra thoughts welcome. > > > > er...doesn't "single user" mode mean no networking? My understanding is that > this is really only for maintenance (ie. make installworld, etc.), not regular > operations. Perhaps you meant something else or I just haven't had enough > caffeine yet... "Single-user mode" refers to the point in the boot process prior to running the startup scripts that make multi-user services available: for instance, mounting all drives, turning on swap, configuring network interfaces, starting daemons etc. and potentially most importantly, setting the securelevel. Unlike the sysV init, there is no real "magic" about single-user mode apart from the fact that you can get init to stop the boot process and drop you directly into a shell in "single-user" mode. Running "shutdown" drops you into much the same state - that is, it kills off daemon processes etc. so that the machine can be administered* without unexpected interference from spurious processes. However, there's nothing (in principle) stopping you from kicking off those processes again, providing their environmental needs are satisfied. So to answer the question: you can certainly tune the scripts and services available that launch you into multi-user mode to get a minimum profile on the machine. However if you modify rc to the extent that it turns on everything you need in order to set up bridging, run a few daemons etc then you're effectively duplicating the multiuser startup anyway. jan * modulo securelevel changes which can only be reverted via reboot. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ The Java disclaimer: values of 'anywhere' may vary between regions. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: gcc34 Question - Request for help
On Wed, 11 Aug 2004, SP Network Solutions wrote: > Hi there, > > I have just install the gcc 3.4 into my FreeBSD 4.10 via the ports/lang > > after make, make check and make install it shows that all install... > > but when I type : gcc -v it still show my old version 2.95.4 > > My question is: How do I go about in making my machine to use the new gcc34? > > What will be the next step after "make install" ? > > Anyone can help??...I'm confuse and this is not the first time I come > accross as I have the same problem with my gcc33. The appropriate binaries are installed as gcc34, &c. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Goedel would be proud - I'm both inconsistent _and_ incomplete. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: implementing spf
On Mon, 26 Jul 2004, Robert Storey wrote: [on spf] 1) Is the technology useful? 2) How does one implement spf on the server side? 3) How does one implement spf on the client side? I most interested in No. 3 above - specifically, is there anything that I must do as an end-user to make use of spf? The answer to (3) directly is, "no"* - which is why SPF is so useful, since it doesn't require large (or any) changes to a user's clients (assuming you're not doing mailing list expansion via your MUA, which I believe some packages do). The SPF pages at pobox have all the details you could want, including pointers to various implementations. It's pretty simple and straightforward; furthermore, the license situation is positive. If the MARID stuff ends up encumbered with what RMS calls a non-free-software-compatible license then it's quite likely that it simply won't fly, regardless. SPF itself is in use right now, however. jan * if you're originating email from "offsite" (dialup lines etc.) then you may need to coordinate with your mail admin to ensure you're targeting the correct MSA. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Donate a signature: http://ioctl.org/jan/sig-submit ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: setuid diffs...
On Fri, 23 Jul 2004, Steve Bertrand wrote: Hi all, Late yesterday, I ``cloned'' my single, primary IDE FreeBSD hard disk onto a larger one. Then, using a Promise ATA IDE RAID controller I built a RAID-1 array. Everything went as planned, the box is now back up using the 'ar' driver for the array. However, in the security run output last night, I got this: [diffs snipped] Did this happen because the files were transferred from one disk to another and the system knew it? Yes. When you copied the files across, the copies were allocated different inode numbers on the new filesystem. It's that difference that the nightly check is picking up on. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Strive to live every day as though it was last Wednesday. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: "sanitizing" disks: wiping swap, non-allocated space, and file-tails
On Sat, 17 Jul 2004, David Kreil wrote: I wonder, in particular, how "system" directories like /var would be kept on a gdbe partition. Much like any other, but the major issue is that, unlike /tmp/ and swap (which can be wiped clean when a machine boots with no ill effects), other partitions need to persist. That means you need to do one of two things: 1. Be available when the machine boots to enter the keys to mount the persistent partitions; or 2. Store those keys somewhere so the machine can do it for you. If you choose (2) then you might as well not use an encrypted partition; secure use needs human intervention. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ If it's broken really badly - don't fix it either. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: closer, no cigar.
On Tue, 13 Jul 2004, Eric Crist wrote: > On Tuesday 13 July 2004 22:19, Giorgos Keramidas wrote: > > The first of these processes is sendmail started in submit mode by the > > > > `rc.sendmail' startup script. The relevant rc.conf options are: > > : sendmail_enable="NO" > > : sendmail_submit_enable="YES" > > : sendmail_submit_flags="-L smtpd -bd -q30m > > : -ODaemonPortOptions=Addr=localhost" > > > > - Giorgos > > Can you exlpain exactly what submit mode is for? Is it something you want > running on a production mail server? Eric, the first three paragraphs of section 1 of RFC 2476 explain this. In a nutshell: an MTA is not supposed to "munge" email (apart from adding Received: headers and the like); however, many local clients submit via SMTP and the mail server needs to do lots more work: rewriting email addresses, and so on. The split of sendmail's operation into MTA (Transmission) and MSA (Submission) is to support this. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Whose kung-fu is the best? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: HP Visualize workstation
On Tue, 29 Jun 2004, Ade Falusi wrote: > Can FreeBSD be installed on an HP visualize B132L workstation, if so how and what > type i.e > alpha, amd64, i38, ia64, pc98 or sparc64. > Pls help. I believe that machine is based around the PA-RISC. You might want to see if NetBSD will run on it. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ On modesty: whoever said "it's hard being perfect" obviously wasn't me. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: md5 of a filesystem / verifying filesystem integrity after dump/restore operation
On Wed, 23 Jun 2004, Ruben Bloemgarten wrote: > Hi all, > > Does someone know how to reliably run a checksum of sorts on a filesystem, > to be able > > to verify filesystem integrity after a restore from dump level 0 has > occurred? Tripwire and its ilk live in the ports system. The base system utility, mtree, also has this capability, although you'll have to fiddle with its options to get it to work. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ You see what happens when you have fun with a stranger in the Alps? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Utility to guess a remote hosts operating system?
On Wed, 23 Jun 2004, Edd wrote: > My question is: > > Does such a utility exist? I know nmap can guess os, but it takes a few > seconds and a port scan is needed first. Is there just a simply util > that can tell me without the port scan? How would that operate? Some kind of network fingerprinting is required. If you can narrow down the parameters of your question (eg: I have a network of windows machines and I'd like to figure out exact versions on each one) then you might have more luck. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ __/\/\_/\/|_ flatline ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: [OT] Re: What's the best possible email failover solution
On Tue, 22 Jun 2004, Bill Moran wrote: > The other option is to take what appears to be the best IMAP server out > there (Cyrus) and figure out a way to do real-time mirroring of the > mailboxes. I was wondering if it could be done with Coda, but I don't > know anything about Coda, and it doesn't look like I'll have time to > experiment in the near future. A previous responder to this thread has already pointed at the cam.ac.uk work which offers transactional replication for fast fail-over. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ "Impact of vulnerability: Run code of an attacker's choice Maximum Severity Rating: Moderate" -- M$ security bulletin ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: [OT] Re: What's the best possible email failover solution
On Mon, 21 Jun 2004, Bill Moran wrote: > During my research of the IMAP protocol, I determined that _the_best_ > way to store email for high-performance would be to put them in a > database. This is because IMAP doesn't see email as a big blob of > text like POP does. It sees the headers as one thing, and the > different MIME parts of the email each as a seperate thing that can > be fetched independently of the other MIME parts. This is a pretty > good layout for a one -> many relationship in a database. Fact is, > every current IMAP server that I'm aware of has to break emails > apart on the fly in order to server IMAP. Have a closer look at the cyrus layout. Each message is in a single file, true, but they are also preparsed to extract the data required for common IMAP operations. The index files contain things like preformed bodystructure responses and offsets to each mime piece. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ HP-unix: Open Sauce product, available in 57 distributions. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: java not running .jar files without absolute path. (take II)
On Wed, 9 Jun 2004 [EMAIL PROTECTED] wrote: > hello jan, > > > echo $JAVA_HOME > JAVA_HOME: Undefined variable. > > hmmm. i don't recall ever setting that variable. yet, i do recall > this not being an issue with previous installs of java. > > is JAVA_HOME a variable that is required? shouldn't the $PATH entry be > sufficient? Well, the path entry only suffices on my installation; I merely hypothesised that an erroneous setting of that variable might be the cause. If I were you I'd take this to the freebsd-java list, there's more of a concentration of expertise there :-) -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ New Freedom of Information Act: theirs, to yours. Happy now? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: java not running .jar files without absolute path. (take II)
On Wed, 9 Jun 2004 [EMAIL PROTECTED] wrote: > > which java > /usr/local/jdk1.4.2/bin/java > > > java -jar Jreepad-1.0.jar > Error: could not find libjava.so > Error: could not find Java 2 Runtime Environment. > > however, the program will run with an absolute path: > > /usr/local/jdk1.4.2/bin/java -jar Jreepad-1.0.jar > > > i don't see any typos in my $PATH (witness 'which' result above). > > the same occurs with linux-sun-jdk (when i put it earlier in the $PATH) > > am i missing something? is this standard behaviour for java? Works for me; is JAVA_HOME set to something broken in your environment? -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ ( echo "ouroboros"; cat ) > /dev/fd/0 # it's like talking to yourself sometimes ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: wine
On Wed, 12 May 2004, Radu MOLNAR wrote: > The second problem that i had is more wine related. In all of the 3 games > i mentioned i was unable to connect to any network games or host any games > that other could see. I posted these problems on the wine mailing list too > but it seems that not many freebsd people use wine or play games. This is > odd as from what i heard freebsd's power is in it's speed, something that > i noticed for my self by using it for desktop aplications. I've also experienced a loss of network connectivity for Wine apps since the version around 2004-04. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Leverage that synergy! Ooh yeah, looking good! Now stretch - and relax. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: looks like i've dug my hole too deep, plz help quick.
On Wed, 12 May 2004, carvin5string wrote: > You have only those two files in root? What about the kernal files I > see on my system? You're running a -STABLE branch, the responder is running -CURRENT. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ "Sufficiently large"="infinite" for sufficiently large values of "sufficiently" ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: private subsubdomains
On Wed, 7 Apr 2004, Sebastian Kutsch wrote: > I want to setup subsubdomains to reach individuell PC's in my LAN from > the Internet without portmapping. > I habe got a ADSL line using ppoe to conect to my ISP getting a public > IP on the tun0 interface. Using the no-ip service I have got a subdomain > of the form my.domain.com. ppoe and the no-ip daemon are running on > a saperate machin with FreeBSD 4.7 functioning as the network > gateway/router wich is conected with dc0 to the ADSL-modem and with rl0 > (192.168.33.1) to the LAN. What I want is to do is reach an individual > PC from the Internet by using a subsubdomain of the form > pc1.my.domain.com. > I have configured bind 8 running on my gateway so that I can reach the > PC's in the LAN from within the LAN by using i.e. pc1.my.domain.com. > But trying to reach a PC from the Internet conects me to the gateway. > By reading the Handbook and man-pages I have not got a clue how to > resolve this problem. In general, you're out of luck. Pretty much every service uses IP addressing only; thus, unless you have sufficient routable IP addresses (ie, your ISP is providing you with a subnet) then you simply can't do this. The major exception is HTTP (and the various things that layer on top of it) which carries the target hostname as part of the protocol. Unfortunately, HTTPS leaves you back at square one since it uses SSL directly rather than via some form of STARTTLS negotiation that might be able to send the desired hostname. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Bolstered by my success with vi, I proceeded to learn C with 'learn c'. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: How to get memory usage for process?
On Tue, 6 Apr 2004, Artem Koutchine wrote: > Hi! > > I need to figure out how much memory process really takes. > For example, i am running 100 perl scripts, they are all the > same source and i guess some memory is shared among them > (mostly perl interperter i guess). So, i need to know how much > memory is shared and how much memory is used for each new > running script (including buffers, e.t.c.). What command shoud > do the trick and with what options? You're probably after the sysutils/pmap utility, in the ports. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ "...perl has been dead for more than 4 years." - Abigail in the Monastery ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: secure cvs server, urgent
On Sun, 4 Apr 2004, dave wrote: > Hello, > I'm wanting to offer cvs services to a select group of users from my > internal server. I need this to be as secure as possible using ssh. I've > tried cvsd, and although it starts when i try to log in, i'm using :pserver: > at the moment, i get the message, premature end of file from server, consult > above messages if any. There are none and nothing in the logs. When i tried > to use the :ext method i got the error: > login can only be used with the pserver method. > If anyone has secure cvs services going behind a firewall to users on > the net please let me know. > Thanks. > Dave. Depends what you mean by "secure". Typical use with ssh is to set CVSROOT=:ext:[EMAIL PROTECTED]:/path/to/cvs/repo CVS_RSH=ssh in which case the user will need an account on your machine. If you set up public/private key pairs then cvs access from the command line is seamless without requiring "cvs login" (you use ssh-agent to keep your passphrases for you). You can "force" a command line by configuring the user's .ssh/authorized_keys file; you might want to consider setting up a jailed environment for your users with sshd and cvs in it. Even after all that, cvs still has enough mechanisms to permit the execution of arbitrary programs on the server with the user's credentials; so your cvs server still effectively has a high level of trust in your users. If you want something "secure" like this you might want to consider an alternative source control system. Or maybe, if licensing permits, offload that risk to sourceforge, who've got a lot of practice at this sort of thing. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Personal responsibility for corporate decisions: if they've nothing to hide, they've nothing to lobby against. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Java Requires X Windows?
On Sat, 27 Mar 2004, Matthew Seaman wrote: > On Fri, Mar 26, 2004 at 06:01:51PM -0900, Joe Pokupec wrote: > > > Do I need to have X Windows installed for Java to 1.4 to run on FreeBSD 5.1? To run, no. To build, what Matthew said. > You don't need a full installation of XFree86: all you need are the > devel/imake4 and the x11/XFree86-4-libraries ports. Although if you're doing any AWT stuff server-side (eg, rendering graphics on the fly for delivery by servlet) you may find you need this: JAVA_OPTS=-Djava.awt.headless=true I believe the 1.4 release notes for Java include details about its somewhat improved (ie, it works) "headless" support. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ ...and then three milkmaids turned up (to the delight and delactation of the crowd). ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: sleeping for 30 seconds
On Fri, 26 Mar 2004, Richard P. Williamson wrote: > What's a good sh algorithm for the above? (assume < 10 > seconds indicates error condition and a 50 second sleep > between iterations when necessary, and I don't need a > 'five attempts then sleep' thing). Also, assume I've > simplified for the home audience. The sh script is > lots more involved then the example above implies. You might, if the possibility is there, want to use the exit value of the main process to indicate errors, which is perhaps a little more reliable. If you want the time in seconds between two points, you might try something like this: start_time=`date +%s` # ... do something end_time=`date +%s` time_taken=$(($end_time - $start_time)) -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ New Freedom of Information Act: theirs, to yours. Happy now? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: The chown command
On Mon, 22 Mar 2004, Gerald S. Stoller wrote: > This gives the system owner the flexibility to leave > it this way, or to restrict this ability to root as it is now by > seting chown's permissions to 500 , it is already owned by root. The "chown" command merely uses the "chown" system call. It is perfectly possible for a user to write and compile their own version of the chown command; so setting permissions on a particular executable do not, in and of themselves, prevent users from effectively duplicating the effect of the command. This is broadly true across all unixalikes. > This is all that a single actual user (as most home systems are) > system needs, but for a true multi-user system one may want to restrict > the change to cases where the new owner and the current owner are members > of one group (and the system administrater should be careful about adding > users to the group wheel ). If the system has some groups that contain > all users, we may want to allow them to be excluded from consideration, > though we shouldn't worry about this now. > I would like to push for such a change and wish others would > join me; if anyone knows of any possible problems from this change, or > has any objections to it, please let me know. This seems overly complicated. The reasons chown is generally limited are security-motivated: for example, one can subvert a quota system by "giving away" files. Rather than present your solution first, perhaps you could indicate the use cases that motivate your suggestion. There may be other ways to achieve the goals you have. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ "...perl has been dead for more than 4 years." - Abigail in the Monastery ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Why does `df` lie about free space (it doesn't)
On Thu, 18 Mar 2004, Kyryll A Mirnenko wrote: > >Using "tunefs -m". You need to be really careful doing this, and read > >the man page for tunefs again, particularly the warning about how > >lowering this number can trash your filesystem's performance. > > I don't want that, I need to allow using preserved 8% of disk space > to a little group of non-root users (for ex. postgres & rootty, my > unprivileged user), but noone more. How do I do this? You don't, without hacking filesystem code. The suggestion of another poster to buy more disk is a good one. > >PS. You keep on appearing to confuse the notion of free data >blocks with > >free inodes. They're not the same thing: they are two distinct >resources > >and your filesystem can run out of either pretty much >independently. > > inode(5) descrbes inodes as a table of block addresses kinda FAT but > with variable block sizes inodes point to. That is. It's not really like FAT operation at all; but another responder has given some detail along these lines. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Theory and practice _are_ the same thing. In theory. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Why does `df` lie about free space
On Wed, 17 Mar 2004, Kyryll A Mirnenko wrote: > Thanks, thats what I want. So that means nobody but root can write > to that preserved (with `tunefs -m`) space? How can I allow more users > to do that? Using "tunefs -m". You need to be really careful doing this, and read the man page for tunefs again, particularly the warning about how lowering this number can trash your filesystem's performance. > (my mail server crashed on friday, so I didn't receive freebsd > digest about this) jan PS. You keep on appearing to confuse the notion of free data blocks with free inodes. They're not the same thing: they are two distinct resources and your filesystem can run out of either pretty much independently. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Don't annihilate, assimilate: MacDonalds, not missiles. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: RAID1 vs RAID5 [ was Re: 1 processor vs. 2]
On Thu, 4 Mar 2004, Chuck Swiger wrote: > Also, RAID-5 performance degrades horribly if a drive is down, whereas RAID-1 > does fine... Using the algorithm you indicate below, RAID-5 performance would not degrade on the loss of a drive, it's start out that badly. > A five-disk RAID-5 array has to read 4 sectors and write five sectors if you > change one byte. Wrong; see previous response. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Q: What's yellow and equivalent to the axiom of choice? A: Zorn's lemon. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: 1 processor vs. 2
On Thu, 4 Mar 2004, Chuck Swiger wrote: > The parity calculations for RAID-5 are a lot of work and that work scales > linearly with the number of drives in the array. The longer you make the > array, the worse the performance becomes for small writes in particular. How did you come to this conclusion? For a RAID 5 with a single parity drive, the reason you zero the disks out completely on initialisation is to set up the integrity of the parity check. Then any update to any RAID5 with single parity requires a read of two drives (the target sector and the corresponding parity drive), an in-memory exclusive or against the new data, and two writes. Reads and writes can be in parallel. The "work" for parity updates only scales linearly with number of disks if you use a naive parity algorithm. Or, obviously, if a drive fails. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ There's no convincing English-language argument that this sentence is true. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Problem with sed and awk
On Thu, 26 Feb 2004, [iso-8859-2] RoubĂ?ek Zden?k (T-Systems PragoNet) wrote: > > Hello questions > > Any idea what I am missing? > > >cat test > 1;1 > 2;2 > >awk -F ';' '{print $1}' > 1 > 2 > >awk -F ' FS=";" {print $1}' > 1;1 > 2 > > The FS=";" is a pattern expression that is used to match the first line of input, after it has already been split into fields. It evaluates true so the block it guards is always run. After the first line has been dealt with, future lines will be split using the new FS setting. As another poster supplied, slap the FS setting in a BEGIN-guarded block. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ If it's broken really badly - don't fix it either. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Shell scripting woes.
On Wed, 25 Feb 2004, Mathias Haas wrote: > Oh my! What stupidity! Of course. I'm afraid sometimes my DOS-roots are > revealed... Thanks! In general, you ought to be aware that cron sets up only a minimal environment. If you want your scripts to run predictably then you probably ought to ensure the first thing they do is to recreate the environment you're after: setting $PATH, the working directory, umask, and so on. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Boycott Arabic numerals! What have they ever done for us? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Move to mail server? WAS Re: Sending mail gets 'Relaying Denied' from off network.
On Wed, 18 Feb 2004, Eric F Crist wrote: > OK, I'm tired of trying to configure sendmail. I think I give up. I've hear > postfix and qmail recommendations the most. I need a mail server that can do > a couple of things for me: > > 1) Host multiple domains on the same server > 2) Easy user management and control (quotas?) > 3) I NEED MY SPAM ASSASSIN > 3) Webmail recommendation for this mail server. > 4) Your love and support with what I choose... ;) > > Sendmail, since I've used it heavily, has become too much of a burden for me. > Please reply with a 'vote' and possibly and howtos or 'change-over' > instructions. If you all recall, I've failed this once before. They'll all do this. Personally: exim (actually, multiple exims, but I've a rather unusual setup) as MTA, cyrus as mailstore, MailScanner in queue-to-queue mode, and silkymail as webmail client. Unfortunately, depending on how you want to manage your users, cyrus is one of the "heavier" imap servers out there. I love it to bits but it can be tricky to get exactly right. You will need some scripting to tie your directory/authentication to cyrus quotas. You'll probably get the same story with most roll-your-own setups. There are a number of products that claim to be email "appliances" and you might look at those, depending on how much preexisting infrastructure you've got for managing virtual domains and users. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Axioms speak louder than words. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Lock order reversal
On Thu, 29 Jan 2004, Piotr Gnyp wrote: > What is exactly lock order reversal? I`ve encountered something like > this on my console: > lock order reversal > 1st 0xc720dce4 vm object (vm object) @ vm/swap_pager.c:1323 > 2nd 0xc06c01a0 swap_pager swhash (swap_pager swhash) @ > vm/swap_pager.c:1838 > 3rd 0xc0c358c4 vm object (vm object) @ vm/uma_core.c:873 A lock is a software object that is attached to a piece of data in order to prevent multiple concurrent accesses/updates to that data getting an inconsistent picture of the data. FreeBSD "names" classes of lock depending on the objects they protect. There are situations that can arise called "deadlock" where process A holds lock x and waits for lock y; process B holds lock y and waits for lock x. Various schemes for ordering lock acquisition can be used to guarantee that deadlocks can arise. A "lock order reversal" means that the debugging watchdog in your kernel has detected that locks of two classes were acquired in different orders. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ If you have received this email in error, do whatever the hell you want with it. It's not like I can stop you anyway. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: portupgrade
On Fri, 23 Jan 2004, Jeff Elkins wrote: > Ouch! > > I'm located somewhere deep within a portupgrade cycle. I misunderstood the > command (portupgrade -a), and "assumed" that it would only upgrade ports that > I had installed, not pkgs installed with pkg_add -r. > > How can I avoid portupgrade touching big packages such as KDE, Mozilla and > OpenOffice which were installed from binary? Look for HOLD_PKGS in /usr/local/etc/pkgtools.conf -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Axioms speak louder than words. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Problems using portupgrade to recompile all ports
On Thu, 15 Jan 2004, Scott I. Remick wrote: > So I'm upgrading my 5.1R desktop to 5.2R. Used cvsup, followed the > instructions in UPGRADING, did a custom kernel, etc etc. That part went > fine, no probs. > > I noticed some of my daemons (from ports) seemed a bit annoyed though upon > booting up 5.2. I tried using portupgrade -Rf on them individually, and > then all was well. I decided then that it'd be best to do everything (-Raf) > to play it safe. I've done this before. > > So it finally finished last night, but not really... about 132 ports were > failed/skipped. My problem is figuring out the most efficient way to deal > with it from here. LAST time I did a portupgrade -Raf I had a much smaller > number failed/skipped, and what I did was work out the dependency tree for > the remaining ones by hand using pkg_info -R and -r, figure out the order, > and do a portupgrade -f on each in the proper order. This was to avoid > rebuilding stuff already built on the first -Raf pass, and multiple times > over (since I was taking care of each remaining one individually). Seems to > me that if 50 of those 132 are X apps and I do a portupgrade -Rf on each, > I'll be rebuilding XFree86 50 times. Hence the need to work out the install > order by-hand based upon dependencies and only use -f. But I don't see that > as practical this time around with so many left to do. > > So... my ultimate question is: how do you pros handle situations like this? > Is there a trick I'm missing? Do you know why the failure happened? The most frequent cause of this when I've encountered the problem is that a distfile could not be fetched. I tend to try to avoid that these days by prefetching the distfiles prior to a build (ie, while I'm around to sort out problems manually rather than overnight). -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ If it's broken really badly - don't fix it either. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: NOQUEUE: SYSERR (root): host "localhost" unknown ?
On Thu, 15 Jan 2004, Rommel B. Ikeda wrote: > Thank you very much for the reply... > This was what was in my /etc/hosts before your response... > > ::1 IBM-R40e [EMAIL PROTECTED] > 127.0.0.1 IBM-R40e [EMAIL PROTECTED] > 192.168.1.35IBM-R40e [EMAIL PROTECTED] > > This is my new /etc/hosts now: > ::1 localhost > 127.0.0.1 localhost > 10.0.0.1[EMAIL PROTECTED] Domain names should not contain "@" characters. > I had to supply 10.0.0.1 [EMAIL PROTECTED] because Gnome 2 complains > that it can not find it...So, I just invented it...I hope that it is > right... > One thing I do not I understand is that when I do: > %host localhost. > It will be give a "Host not found" response...But, my system is working > fine...What do you think is wrong with it... The "host" command doesn't use the full resolver; it uses DNS queries only. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Theory and practice _are_ the same thing. In theory. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: /usr/home directory
On Wed, 24 Dec 2003, flux wrote: > Maybe kinda strange question, but... > Why users' home directory located in /usr by default, not in > root directory unlike Linux? > Any ideas? Hysterical raisins. Amongst other things, /home was often given to the automounter so people's home directories could rove. If you want to change this, pw useradd -D -b /home will probably do the trick (you will need to move existing directories). -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ "Impact of vulnerability: Run code of an attacker's choice Maximum Severity Rating: Moderate" -- M$ security bulletin ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: use vi style in all line-edit environment
On Tue, 23 Dec 2003, Zhang Weiwu wrote: > I have been using vi all the time, even in csh I set bindkey -v to use vi > style command-line edit. > > But when the context is switched to python, mysql, or scilab I have to warn > myself not to type in vi style. Very often in mysql I mistyped something > and try to type a "x" to correct it, or I press "A" and thought I would go > to the end of the line. > > Is it possible to make a global configuration to let all command-line > processor accept vi style key? Most of the tools you're referring to use the GNU readline library. The manpage (readline(3)) _claims_ that it offers a vi-style editing mode. You'll have to suck it and see: try echo set editing-mode vi > ~/.inputrc and check the man page for more information. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Solution: (n) a watered-down version of something neat. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: How to count the lines of code in a project?
On Tue, 23 Dec 2003, Zhang Weiwu wrote: > > Hello. How to count the number of lines in all *.c file in a directory? > > I can think of this on csh: > > > grep -c "" `find . -name "*.c"` | sed "s/.*:/e=e+/" > /tmp/countlines.py > > And edit the py file, and "e=0" as first line, "print e" as the last > line, and execute the python script. wooo pretty cool for a newbie like > me:) > > So is there a better method? Not bad; but when it comes to basic operations on text files, unix has a surfeit of tools that probably already do what you want. In this case: find . -type f -name \*.c -print0 | xargs -0 wc -l -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ (ECHOY GRUNTING) (EERIE WHISPERS) aren't subtitles great? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: FreeBSD CVS for DOS/Windows sources?
On Sun, 21 Dec 2003, Doug Lee wrote: > On Sun, Dec 21, 2003 at 08:58:30PM +, Daniela wrote: > > On Sunday 21 December 2003 17:25, Doug Lee wrote: > > > I'm trying to use FreeBSD's CVS (v1.11.5) to manage DOS/Windows > > > projects, which of course means files with CR/LF line endings. The > > > docs claim the repository will internally store everything (text files > > > anyway) with only LF endings, but in fact I'm seeing CR/LF endings in > > > repository text files. This is still fine until I pull a project out > > > with a DOS/Windows CVS client (the standard cvs.exe), which gives me > > > lines ending in CR/CR/LF. If I pull the project under Unix, I get the > > > CR/LF endings just fine; but my coworkers will not be telneting to > > > Unix just to pull code... > > > > I don't know anything about Micro$oft Winblows eXPensive, but why don't you > > just strip the CR on the server? So the Windoze client can add the CR, and > > you always get the native format, respectively. > > This is definitely the fault of the client, so don't blame the server :-) > > Actually I wasn't sure whose "fault" it was, since the docs say the > CR's should not be stored by the server. Given that, the client did > what it should under the circumstances. If the file is stored in text mode, the _client_ does line-ending conversions according to local conventions. The server expects the files to be in the unix \n-terminated format. So the problem is actually the "fault" of the unix client since it "knows" that it doesn't need to do any CRLF translation on upload. You see this a lot when using Unix CVS to deal with windows-originated files. Your options are either to do the first import using a windows cvs client, or to correct the line ends (eg, with dos2unix) on the unix cvs client prior to upload. Once the files are correctly formatted on the server, both unix and windows clients will be able to deal with them correctly. > I didn't edit the repo because I don't consider myself savvy enough > about CVS yet to start playing with repo files directly, except to > look at them with a curious eye. For text files, they're line-oriented. The server itself will not insert any '\r' characters, so any that are in your repo files definitely came from a borked upload. You should be able to strip them out and re-checkout the files: diffs will be unaffected. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ There's no convincing English-language argument that this sentence is true. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: spamassassin
On Sun, 21 Dec 2003, Lowell Gilbert wrote: > "Doug Reynolds" <[EMAIL PROTECTED]> writes: > > > I've been trying to setup spamassassin on my freebsd box with postfix. > > I was reading the other thread about spamassassin. What i couldn't > > figure out, was how it got the mail from the mail system. I looked at > > spamd but it didn't look like the program. I googled and still > > couldn't find a good tutorial. If someone knows of a good tutorial and > > could give me a quick explaination, i'd be greatful. Oh, I am running > > 5.1-RELEASE. thanx > > My preferred method is by filtering through procmail. It's a simple > procmail recipe: > # pass through spamassassin > :0fw: spamassassin.lock > | /usr/local/bin/spamassassin > > (where the "w" flag tells it to continue procmail processing after > running the message through spamassassin) This is ok on workstations but it's hard to manage the load on a multiuser system, since procmail is effectively a final delivery. There's also a port for MailScanner, which operates prior to final delivery on mail spools. It can integrate SpamAssassin and the virus scanner(s) of your choice; as a plus, the load generated is far more predictable. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Rereleasing dolphins into the wild since 1998. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Release Tag and Branch Tag and Upgread ..
On Tue, 16 Dec 2003, Vahric MUHTARYAN wrote: > Hi Everybody , > > I know that all FreeBSD version have Branch on code tree . Now > I'm using to keep up-to-date FreeBSD witch CVSup and buildworld method . > > I'm using 4.9 or 5.1 versions . in stable-supfile I setted > > *default release=cvs tag=RELENG_4_9 > > I know that, this method for only security advisories and other > seriously critical fixes ... > > What is going on if I set it RELENG_4 ?! Do I have to use it ? You instead track the -STABLE branch. As well as vital security fixes, new functionality may also appear on this branch as it is tested under -CURRENT and backported. Despite the name, occasionally instabilities have been known to crop up in -STABLE from time to time. They're usually dealt with promptly, but if you're going to track -STABLE then you should subscribe to the freebsd-stable mailing list to stay abreast of announcements (it's pretty low-traffic). > What about RELEASE Tag ?! When Do I have to use this type of > update for example Can I use it when I want to upgread my FreeBSD 4.9 > RELESE to FreeBSD 5.1 or FreeBSD 5.0 RELEASE to Freebsd 5.1 RELEASE ... The RELEASE tags just mark a single point in time. You generally do not want to "track" these using cvsup because the tags never move; nothing ever changes. > I wonder about CURRENT Branch , How FreeBSD CURRENT users can > keep-up-to date their src tree and system because all of those updates > looks like for STABLE or RELEASE ?!! cvsup knows about cvs branches and will only apply the patches from a branch you tell it to track. In any production environment it's common to have a test/build machine which you can try the latest -STABLE or -CURRENT on. If you're happy that no regressions have been introduced, then you can share the /usr/src and /usr/obj directories out via NFS to your other machines and use those to drive an installation. > Last question is ; Some body said that you I did not need to > make buildworld every time ?! I want to know When need to make > buildworld process ?! "make buildworld" will rebuild (but NOT install, that's a separate step and you should read the section in the handbook about the right way to do this) all the userland bits in a base FreeBSD system. "make buildkernel" will rebuild the kernel. [ If you are installing third-party software via the ports system, that will need separate maintenance. A common solution is to install the "sysutils/portupgrade" port and use that to manage your ports. Once you've got that set up it pretty much takes care of the rest of the leg work involved in maintaining a system. ] Although it is possible on older machines to selectively build parts of the system that you wish to update, in general I've found it to be a false economy on modern hardware (unless you're actually a freebsd developer). A full build can run over a lunchbreak and be ready for testing on your return :-) -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ "...perl has been dead for more than 4 years." - Abigail in the Monastery ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: last question about up-to-date ( I hope )
On Thu, 11 Dec 2003, Vahric MUHTARYAN wrote: > Hi , > > For keep up to date FreeBSD I think all people are using source update > method ( When I sent a message to list almost everybody adviced this ) Only > one person said that binary update but this is not recommanded because > compiled version always work better and I saw that compile update program is > not working quickly because Colin Percival waiting lest version > > > I'm just wonder Why patching is not used instead of source update.. > it's patching source tree too for security bugs ... I checking output of the > cvsup -g -L 2 stable-supfile command . it's only download openssh , bind and > like this almost what writen in security advisories . > > if you said soruce-update method more then security update Thats Okey . > But I want to know or understand if I don't want to use new features and > only interest with security updates ( patch updates ) Why patches does not > enough ?! You _are_ downloading patches when you use cvsup. However, the tool provides a handy level of automation and therefore can prevent simple pilot error compared to hand-application of patches. You are not required to track the -STABLE branch. Every (recent) release also has a maintenance branch, which merely receives security updates. Cvsup can track these just as easily for you. The handbook has more information on this. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Political talk? / What is said can be unsaid / with good old BS -- ASCII haiku ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: When FreeBSD have higer performans than Linux and When Linux have higher performans than FreeBSD
On Thu, 11 Dec 2003, Vahric MUHTARYAN wrote: > I red Explaining BSD documents and on 4.7 Which should I use , BSD or > Linux section it said that " BSD systems , in particular FreeBSD , can have > notably higer performans than Linux but this is not accross the board In > many cases , there is little or no difference in performans. In some cases , > LInux may performa better than FreeBSD . > > *Now , I wonder When or Which situations FreeBSD have more performans > than Linux ?! > > *I wondor too When or Which situations Linux have more performasn than > FreeBSD ?! > > I used Linux too much but I never use FreeBSD in production Env. because > of this I can't compare it "Suck it and see." That is, if you're interested in comparing performance for your application, run your application on both platforms and measure the performance. If you have specific questions about tuning for a particular application, then the man page tuning(7) has some information; or bring specifics to the list. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Work #90: As many pseudo-intellectual sycophants as necessary to make one inarticulate scotsman think he's a genius in command of The Profound. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: how to build Spamassassin
On Mon, 8 Dec 2003, Tony Jones wrote: > Hi. > > Over time I've got into the habit of either using packages or building > directly from the source. Last time I tried this (Postfix) and asked a Q > here, I was rapped over the knuckles :-) and told to use the Ports. > > Right now I'm trying to build spamassassin, so I decided I'd be good and do > it the Ports way. > > My current system is 4.9-PRERELEASE #5, upgraded for many years from src > using CTM. > > - Read the handbook, ran /stand/sysinstall to get the ports tree, this failed > not finding the download location on ftp.freebsd.org and telling me to > manually change it > > - So I went and got it manually. Was a little confused as many years ago > I recalled their being a ports tree per release. Now ports-stable and > ports-current both point to ports. > > - I downloaded ports,tar.gz. Unpacked it and changed into > ports/mail/p5-Mail-SpamAssassin" > > - Did a make which immediately failed with > > "Makefile", line 27: Malformed conditional (${PERL_LEVEL} < 500600) > "Makefile", line 27: Need an operator > "Makefile", line 31: if-less endif > "Makefile", line 31: Need an operator > "Makefile", line 33: Malformed conditional (${PERL_LEVEL} < 500800) > "Makefile", line 33: Need an operator > "Makefile", line 35: if-less endif > "Makefile", line 35: Need an operator > > I recalled being able to make individual ports this way in the past. Where did you unpack the ports tree to? /usr/share/mk/bsd.port.mk expects /usr/ports, or the PORTSDIR variable to be set if the tree lives elsewhere. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ printf 'cat\nhello world' | `sh -c 'read c; echo $c'` ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: x11/kde3 and libglut error
On Sat, 6 Dec 2003, Ryan Moe wrote: > [EMAIL PROTECTED] wrote: > > >Hello. I'm using FreeBSD 4.8-RELEASe p14. I'm trying to install kde3 on my > >system and get this error > > > >/usr/libexec/elf/ld: cannot find -lGL > >*** Error code 1 > I believe the problem lies with XFree86-4-libraries. I had the same > problem and portupgrade -f XFree86-4-libraries followed by portupgrade > -f libglut fixed it for me.. Are either of you using the nvidia-driver port? If so, it puts GL libs in place which break this build. Fortunately, it keeps the old libraries nearby, so rather than a full rebuild of the XFree86-4-libraries, you can pkg_deinstall nvidia-driver, do the build, then reinstall nvidia-driver. Much quicker. Incidentally pkgtools.conf can be taught to do this for you. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ "No generalised law is without exception." A self-demonstrating axiom. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Ports intelligent?
On Fri, 28 Nov 2003, Steve Bertrand wrote: > Traditionally, when installing for a new Internet server, I have > installed everything from sources in a particular method. > > This is a nightmare to upgrade, and would like to know if installing > from ports would be sensible. > > MySQL, Apache 1.x, PHP, Vpopmail, Courier-imap, Qmail etc. > > In which order should I install? I need to have mysql driven vpopmail, > and apache and php must talk to mysql also. IMAP support must be built > into PHP too. > > How do I put this mess together sensibly? Install portupgrade and use that to manage your ports. It can be a chore to get things right to begin with, but is definitely worth persevering with. Check the makefiles of the ports you're interested in for tunable build switches. Read pkgtools.conf carefully. Take copies of all your config files; port upgrades don't usually overwrite these, but if they do you'll be able to recover (and report the matter to the port maintainer). -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Recursion with grep?
On Thu, 13 Nov 2003, Francisco Reyes wrote: > On Fri, 14 Nov 2003, JacobRhoden wrote: > > > No need to hack grep plese! just use -R (it appears the man page does not > > document the -R function, but you need to use -R in grep for it to recurse. > > -R == -r > > That was mentioned in the previous emails. It only recurses directories. > It will not work with a filemask. > > You can't say > grep -r *.c > > You can only say > grep -r That's correct, but the Unix Way is* to use small tools that do one job well* and compose them for the desired effect. Someone's already given you an invocation of find and xargs that (almost) does what you want: find . -type f -name \*.c -print0 | xargs -0 grep -h searchstring If you like, wrap this up with a script. Call the script "grep" and stick it closer to the start of your path than the system one. Even the win32 shell is more capable of command composition these days. Chucking as much disparate functionality as possible into each tool is a poor tactic because you wind up with every tool being indistinguishable from its peers, modulo the totally incompatible and irregular interface :-) G'luck, jan * allegedly :-) -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ and Nostradamus never dreamed of the Church of the Accellerated Worm ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: exim not running after portupgrade?
On Thu, 13 Nov 2003, Ceri Davies wrote: > On Thu, Nov 13, 2003 at 08:50:31AM -0500, William O'Higgins wrote: > > I recently ran portupgrade to get to a state that approximates current, > > and when I rebooted exim is not running. It used to, just fine, but now > > when I boot I've got no MTA. I can tell you, fetchmail thinks this is > > quite the problem :-) > > > > What could have caused this? As far as I can tell it was being started > > *somehow* before, and now it isn't. It works fine if I just remember to > > # sudo exim -bd -q5m > > but I think it shouldn't have stopped starting just because of a > > portupgrade. I went from exim 4.12 to 4.24, but there are no changes in > > the configuration needed for that update. > > I'm pretty sure that sometime during that gap a requirement for > exim_enable="YES" in /etc/rc.conf was introduced. Yes, exactly. The exim port is now compatible with -current's rcng; if you look at /usr/local/etc/rc.d/exim.sh you'll probably see it's checking for that flag. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ printf 'cat\nhello world' | `sh -c 'read c; echo $c'` ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Newbie: The C / C++ Issue
On Tue, 11 Nov 2003, Alex Kelly wrote: > Thanks for all of the great suggestions to my previous question! > > Yet, the responses have led me to another question. If C++ is newer > and more advanced than C, will it replace C? That's unlikely. The two languages are likely to coexist for quite some time. > If so, should I learn C++ and forget C? Opinions, as you'll have figured by now, differ. C++ is a large language; it has something akin to C ("a better C") as a subset. In that regard, it can be convenient to use C++ as exactly that - a better C. However, many of the improvements to that language subset have filtered back into the C standard, so the advantages there are less clear. The real issue with learning C++ is finding a decent textbook! C++ can actually be a very handy "scripting" langauge - that is, it is well-provided with pretty high-level libraries* and just writing code that ties those together is a pleasant experience. C is a lower-level nuts and bolts language. Proficiency with C means that understanding some of the "under the hood" aspects of C++ such as the nuances of object layout can be simpler. It all comes down to what exactly you want to do. If it's just the odd command-line script you might find that the convenience of the standard library of C++ makes it worthwhile learning the bits of C++ you need. jan PS. As a first text, I would generally advise people to avoid KnR. * standard and boost.org -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ They modified their trousers secretly. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: portupgrade -arR
On Tue, 11 Nov 2003, William O'Higgins wrote: > Quite foolishly, I ran this command without thinking it through: > > portupgrade -arR > > It's been running for 15 hours or so now, and I'm wondering how much > longer it is likely to take? I realize that "that depends" Well, "that depends"; many ports can be upgraded quickly. However compilation of C++ is markedly slower than compilation of C, so whenever you see things like KDE or Qt meeding an upgrade, expect it to take a while. If you use "portversion -v | grep -v =" then you'll see the list of ports which remain that need updating. (You need to run portsdb after a cvsup for the output of this, and portupgrade's operation, to be accurate.) > Any suggestions? Thanks. If you've got packages installed that you don't want, portupgrade _can_ be safely interrupted and will pick up pretty much from where it left off when you kick it off again. It's often worthwhile checking the Makefiles for the ports you install for tunable variables. Many ports offer interactive menus to select features to build; Murphy's law would suggest that one of these might well pop up just after you leave portupgrade to do its thing and go to bed. You can normally select batch operation and choose the appropriate options by putting them into /usr/local/etc/pkgtools.conf; the exact mechanism used to indicate "non-interactive" mode isn't uniform across all ports, however. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ printf 'cat\nhello world' | `sh -c 'read c; echo $c'` ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Mesa port
On Tue, 11 Nov 2003, Lutz Kittler wrote: > Hi, > > whats going on with > > Port: Mesa-3.4.2_2 > Path: /usr/ports/graphics/Mesa3 > > It is deleted in ports database,but many packages ( eg. kde* ) > depend on this port. > So portupgrade brings up many errors. > What to do ? pkgdb -F The above fixed the issue (automatically) for me. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Goth isn't dead, it's just lying very still and sucking its cheeks in. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Backing up kernel before upgrade - 4.9 - how?
On Thu, 6 Nov 2003 [EMAIL PROTECTED] wrote: > In order to have a "safety belt" I'd like to backup my current kernel > before upgrading (i.e. "make installkernel") my newly compiled kernel. > > Which files do I need to back up? > > I thought about > > /kernel (file) > /modules (directory) > > Anything else? > > TIA for your help, "make installkernel" will rename your current kernel and modules to kernel.old and modules.old for you. I tend to explicitly grab a copy of configs (/etc) prior to using mergemaster; this is purely paranoia on my part. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ Axioms speak louder than words. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Is Java/Tomcat on FBSD practical?
On Tue, 4 Nov 2003, Heath Volmer wrote: > Hi. I'm new to FreeBSD, my previous -nix experience coming from OSX and > Debian/Suse linux. I've been generally very happy with the performance > and relative ease of setup on my 4.8 system. > > My biggest problem has been Java. I've done the diablo 1.3 package and > java seemed to work (java -v), but when I install Tomcat it won't start. > I've done the same install on mac and windows with no problem. The > stack trace is completely alien to me and I can try to get it if it > would help. > > I'm basically wondering if running Tomcat or any other java on this > machine is realistic or not. > > Thanks, Heath Check on how up-to-date your system is. There were fixes that went into the base system post 4.8-R which are probably relevant. FWIW I've got tomcat running happily under the 1.3 and the 1.4 native ports on a -stable freebsd. -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ (Things I've found in my attic, #2: A hundredweight of pornography.) ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: PostgreSQL on FreeBSD
On Tue, 28 Oct 2003, Wayne Pascoe wrote: > On Mon, Oct 27, 2003 at 04:04:28PM -0500, Lowell Gilbert wrote: > > >Why does setting them in > > > /etc/sysctl.conf or /etc/loader.conf not work ? > > > > You're doing something wrong. That's all I can say when the > > description of the failure is just "not work". > > Allow me to expand on that then... I put the options in /etc/sysctl.conf > as follows: > kern.ipc.somaxconn=512 > kern.ipc.shmmax=268435456 > kern.ipc.shmall=65536 > kern.ipc.shmmni=128 > kern.ipc.semmns=256 > > When I reboot, sysctl -a | grep kern.ipc.semmns returns > kern.ipc.semmns: 60 I believe -current now has code to pull values for these out of the kernel environment; that's missing in -stable (IIRC; not checked, but I have a vague recollection of trying to figure out how the hell the values were supposed to get into the sysctl value until I looked at the -current tree). -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ "I like oranges more than apples!?" - that's like comparing apples and oranges! ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: how do you actually download an up date for KDE
On Sat, 25 Oct 2003, hawley wrote: > How do you actually download an up date for KDE. I have been playing with > Freebsd off and on for 3 yrs now. At present I have "DSL" internet access. I > am running a 500 meg Hz intell machine. I have read untell my eyes bleed and > tryed things untell my fingers cramped: NOTHING EVER WORKS I am now to > the point where I would pay to have someone walk me through a successfull > download. Any thoughts? If you can afford the time for a compile, the simplest approach is to utilise cvsup to keep your ports tree up-to-date (you may already be doing this to update your kernel and base userland system). Details of this are in the handbook. From an up-to-date ports system, install "portupgrade". It's a handy tool which will help you to automate the maintenance of your installed packages. You can tell portupgrade to try to fetch binary packages to upgrade, or (if none are available) it can grab the sources for the packages you need and build them for you. There tends to be a teething period as you switch to portupgrade (if you've been doing things by hand up until now) but it is definitely worth persevering with. There's plenty of advice on using it contained in the archives of this mailing list (also on freebsd-stable). Note that a from-source rebuild of KDE and its dependencies, while generally painless, will take quite a bit of time on your machine - but somewhat less than three years :-) -- jan grant, ILRT, University of Bristol. http://www.ilrt.bris.ac.uk/ Tel +44(0)117 9287088 Fax +44 (0)117 9287112 http://ioctl.org/jan/ On modesty: whoever said "it's hard being perfect" obviously wasn't me. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"