Odd "which" behaviour - not finding shell script
Here's a transcript from a shell session. Can anyone explain why the "which" line doesn't find the script when both tab-completion and running it do? jaycee@bigdaddy:~$ echo $PATH ~/bin:/usr/local/bin:/usr/bin:/bin:/usr/bin/X11:/usr/games jaycee@bigdaddy:~$ ls ~/bin firewall-up.ipchains jc.disconnect jc.mail.newplan oggbitrateaverage x firewall-up.iptables jc.getwww jc.ssh-agent-start old jc.background jc.hdparm jc.template run.and.log jc.connectjc.mailogg2cd run.on.kanyon jaycee@bigdaddy:~$ ls -l ~/bin/ogg2cd -rwxr-xr-x1 jaycee jaycee114 2003-02-21 17:04 /home/jaycee/bin/ogg2cd jaycee@bigdaddy:~$ which ogg2cd jaycee@bigdaddy:~$ #GRR According to which's manpage, it examines all the directories in $PATH, and doesn't reply on a previously generated index. So - what gives? TIA, jc -- $ cat .me/contactdetails/home [listmail - contact removed] $ cat .me/contactdetails/work [listmail - contact removed] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Odd "which" behaviour - not finding shell script
On Fri, Feb 21, 2003 at 05:50:03PM +, Jonathan Matthews wrote: [snip] > According to which's manpage, it examines all the directories in > $PATH, and doesn't reply on a previously generated index. s/reply/rely/ jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Odd 'which' behaviour - not finding shell script [SOLVED]
On Fri, Feb 21, 2003 at 10:53:27AM -0800, Craig Dickson wrote:
> nate wrote:
>
> > Jonathan Matthews said:
> > > Here's a transcript from a shell session.
> >
> > > [EMAIL PROTECTED]:~$ echo $PATH
> > > ~/bin:/usr/local/bin:/usr/bin:/bin:/usr/bin/X11:/usr/games
> > > [EMAIL PROTECTED]:~$ ls ~/bin
> >
> > just a guess but I say your problem is there. The shell did not
> > expand the ~ when setting the path.
>
> Perhaps it was in quotes?
Perhaps both:
It /was/ in quotes in .bash_profile (I'm sure I'm using the stock potato
.bash_profile), and also had ~/bin rather than $HOME/bin.
I've corrected both, though I'm slightly worried that the quotes were
there for a reason - perhaps to cater for spaces in directory names?
Obviously not a great idea, but would 'PATH="$HOME/bin:$PATH"' do this?
Would it let you have, say, "$HOME/binary files" in there without
barfing?
Anyway, for all reasonable paths, it's working.
For the list archives, changing the line
PATH="~/bin:${PATH}"
to
PATH="$HOME/bin:${PATH}"
seems to work.
If you have the double quotes in there is up to you - see any replies to
this post to get the opinion of people who know more than I do.
-- jc
--
$ cat .me/contactdetails/home
[listmail]
$ cat .me/contactdetails/work
[listmail]
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Taking over your desktop, one box at a time
On Sun, Feb 23, 2003 at 02:08:57AM -0500, Patrick McFarland wrote: > This is probably offtopic for this list, but can everyone download > http://freshmeat.net/projects/debian_gel_logo/ (Debian Gel/Aqua Logo > Background) and tell me what they think of it? heh - have had it as my b/g for a week or so now. V nice. Only concern would be to perhaps have "DEBIAN" centred or L-or-R aligned, as opposed to "almost left aligned". Alternatively (and this would get my vote, cause of where my fluxbox slit is located - top left, vertical), have the text going from the top right downwards - hella cool. cheers, -- jc -- $ cat .me/contactdetails/home [listmail] $ cat .me/contactdetails/work [listmail] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Using one mailbox at ISP for many people
Hi all [BRIEF SUMMARY - How do I segregate one mail stream from one mailbox at my ISP into different users locally?] At the moment, I'm the only person using my machine. However, having installed the GNOME 2.2 backport, I've finally got a desktop that I feel happy in sitting my SO down in front of and not worrying that she'll find it so confusing that she runs screaming from the room. Yes, she's not the most technical of sorts, but she tries hard :-) I'm trying to set her up using, for the sake of simplicity and general "nicely-integrated" reasons, balsa as her MUA. She's been using mail.com for a web-based account for a while, but it sucks (slow AND featureless). I'd like to get her up and running with my machine, and using my email ISP. I receive all email to @jaycee.uklinux.net. I'd like her to have a specific one at that domain. For the benefit of any bottom-feeders harvesting the archives, let's say she's going to get [EMAIL PROTECTED] :-) The problem is that I get only one actual physical mailbox at uklinux.net, and all email to @jaycee.uklinux.net gets delivered to it. The username it's sent to /is/ preserved in the headers, however. What combination of programs do I use to get one mailbox successfully segregated when it hits my machine? I'm using fetchmail -> exim -> procmail at the moment, but I've lurked on the procmail list long enough not to ask about using it as a MDA (or is it an MTA that it shouldn't be treated as?). Any help gratefully received and appreciated. -- jc -- $ cat .me/contactdetails/home [listmail] $ cat .me/contactdetails/work [listmail] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
How to find out apt-get's reasoning
Having just installed the Gnome 2.2 backport, I'm trying to drop kde from my box totally. Never liked the "underline the desktop icons" thing anyway[1] :-) I thought I'd got it all out, but witness the following: bigdaddy:/home/jaycee# apt-get dist-upgrade -u [snip] The following NEW packages will be installed: kdelibs-data The following packages have been kept back gnu-smalltalk imagemagick libcurl2 liblcms1 liblcms1-dev libmng-dev libmng1 libpng12-0 libpng12-0-dev libpng3 libwmf0.2-7 mpeglib tetrinetx 0 packages upgraded, 1 newly installed, 0 to remove and 13 not upgraded. I'm not asking anyone to tell me why my particular mix of sources is doing this, I'm wondering more if there's a grep-available or grep-dctrl invocation that might tell me which *installed* packages *depend* on kdelibs-data. Yes, I know "kdelibs & kdelibs4", but neither of them is installed. Any takers? Cheers! jc -- [1] Joke. JOKE! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Using one mailbox at ISP for many people
On Fri, Feb 28, 2003 at 04:32:57PM +, Glyn Millington wrote: > Jonathan Matthews <[EMAIL PROTECTED]> writes: > > > Hi all > > > > [BRIEF SUMMARY - How do I segregate one mail stream from one mailbox > > at my ISP into different users locally?] > > > What combination of programs do I use to get one mailbox successfully > > segregated when it hits my machine? I'm using fetchmail -> exim -> > > procmail at the moment, but I've lurked on the procmail list long enough > > not to ask about using it as a MDA (or is it an MTA that it shouldn't > > be treated as?). > > Procmail is _exactly_ what you want for this! The package description:- [snip] Thaks Glyn - I've already got procmail doing list sorting, a bit of spam catching and so on. I'm wondering how to get a machine-wide procmail set up so that each user (well, me, really - I doubt my SO'll be at that stage for a while :-) can have their own .procmailrc applied /after/ the mail hits the box and is seperated to different users. > Have a look at man procmail > procmailrc > procmailex > procmailex gives me one match for "user", and a few for "different", none of which seem to address this problem. I know it's a very powerful program as I've gone through setting it up a few times as different hard disks have gone down (backups - I've heard of them ...), but the procmail list does seem flame people quite often for daring to ask about using procmail in this manner. "It's not an MDA", apparently. So, how do I get mail from /one/ incoming mailbox to be delivered to different local accounts and not have to make the account owners take any special precautions because of this? Thanks! -- jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
SpamAssassin weightings - am I missing something?
Quick Spamassassin question: I've got SpamAssassin 2.43 installed, and it's working well. However, I noticed the two lines quoted below in the altered body of some spam that it caught recently: SPAM: EXCUSE_16 (-0.3 points) BODY: I wonder how many emails they sent in error... SPAM: EXCUSE_14 (-0.2 points) BODY: Tells you how to stop further spam It seems strange to me that these two reasons should /decrease/ the probability of the email being spam. I know that the weightings attached to different rules are user-definable, so I'm not asking "how do I stop this behaviour" - I can easily go and redefine the weights. I'd just like to get some confirmation that these weightings are wrong. It's the stock install of SpamAssassin in testing, with no alterations made to the config at all. Should I file a bug, change my own weightings or go away in shame, having made a fool of myself publicly? Cheers! -- jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[SOLVED] Re: SpamAssassin weightings - am I missing something?
On Sun, Mar 09, 2003 at 10:19:17AM -0500, Joey Hess wrote: > Jonathan Matthews wrote: [snip] > > I'd just like to get some confirmation that these weightings are wrong. > > It's the stock install of SpamAssassin in testing, with no alterations > > made to the config at all. Should I file a bug, change my own > > weightings or go away in shame, having made a fool of myself publicly? > > The deal is that spamassassin's scores are generated using a genetic > algorithm. They "breed" scores against a corpus of known spam and > non-spam, starting with random scores and mutating them up or down, then > seeing how that does and letting the winning mutations thrive. The aim > is to get as few false positives as possible while still catching as > much spam as possible of course. So the scores are not something > hand-tweaked by a human. > > What happens sometimes is it seems that making a score negative reduces > the number of false positives, while not catching any less spam, at > least in their body of spam. And the SA guys, rightly or wrongly, trust > their GA to get it right, and leave these negtive scores in. I have > mixed feelings about this, but it seems to work. Thanks Joey - I was sure there was /something/ behind it, but I had no idea a GA was used upstream. Seems almost overkill, but if it works ... Think I'll let their best guesses take precedence over mine :-) -- jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: How to find out apt-get's reasoning
On Fri, Feb 28, 2003 at 05:34:12PM -0500, Travis Crump wrote: > Colin Watson wrote: > >On Fri, Feb 28, 2003 at 03:05:36PM -0500, Travis Crump wrote: > > > >>The more likely cause of this behavior is the "Replaces" field of > >>kdelibs-data. I think apt-get will automatically try to install on > >>dist-upgrade any package that claims to replace an installed package. > > > > > >I really hope not. That would be a misinterpretation of Replaces, which > >on its own simply means that some files from the named package(s) have > >moved to the package containing the Replaces field. > > > >Replaces in combination with Conflicts or maybe Provides, possibly ... > > > > Well, the 'Replaces' theory sounds much more reasonable to me than the > 'Depends' theory. No packages were being upgraded on OP's dist-upgrade > so the depends of every other package on his system must remain > constant. In order for a depends to cause a new package to be > installed, it follows that there was a package that somehow got > installed with an unsatisfied 'Depends'. I don't know how this could > happen without the OP knowing what was going on[ie he had been using > dpkg --force-depends]. It's still confusing me, as I'm getting kdelibs-data installed with a dist-upgrade, and removing it doesn't force any other packages out too. The three-liner that Travis posted earlier in this thread produces no output (and I'm not quite sure enough of what it does to work out if there's a simple typo/thinko in there). Another suggestion that came up earlier was to use aptitude, examine the package and then look at the "Packages which Depend on kdelibs-data" line. This showed only one package (kdelibs4) which was not installed. I've had a download.kde.org site in there prior to being able to get GNOME 2.2 under testing, which might have resulted in some weird dependancies. I don't mind /having/ them, though, It'd just be nice to /see/ them! All in all, a bit puzzling :-| Any suggestions? Any packaging command incantations that'll spit out any installed package even /mentioning/ kdelibs-data? I've having a go, but am getting nowhere, fast ... -- jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: SpamAssassin weightings - am I missing something?
On Mon, Mar 10, 2003 at 02:39:49AM -0800, Paul Johnson wrote: > On Sun, Mar 09, 2003 at 11:31:02AM +0000, Jonathan Matthews wrote: > > SPAM: EXCUSE_16 (-0.3 points) BODY: I wonder how many emails they sent in > > error... > > SPAM: EXCUSE_14 (-0.2 points) BODY: Tells you how to stop further spam > > > > It seems strange to me that these two reasons should /decrease/ the > > probability of the email being spam. > > They do. Hence minus points instead of adding them. I think we're talking at cross purposes - I meant "I find it strange that the excuses cited should be taken as reasons for the mail to be less likely to be spam", not any other reading. Were you perhaps reading it differently? -- jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: SpamAssassin weightings - am I missing something?
On Mon, Mar 10, 2003 at 03:36:09PM -0500, Alan Shutko wrote: > Jonathan Matthews <[EMAIL PROTECTED]> writes: > > > I think we're talking at cross purposes - I meant "I find it strange > > that the excuses cited should be taken as reasons for the mail to be > > less likely to be spam", not any other reading. > > Spamassassin default scores are set using a genetic algorithm[1]. > Basically, there's a large corpus of spam and non-spam, the scores > are set at some default value, then modified by the algorithm until > they accurately categorize the spam in the test corpus. [snip] Yes, Joey's already put me straight about upstream using a GA to determine the default weightings. I was replying (without enoguh context) to someone who I'd managed to mislead with my poor grammar :-) Cheers! jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Apt-get, downloading a single package
On Mon, Mar 10, 2003 at 05:31:39PM -0500, Radek Zajkowski [Deb] wrote: > What would like a command to apt-get look like if I wanted to download a > single package. # apt-get install This might pull in some extra packages. If you're hoping to avoid this, I'd suggest you don't, as apt-get only pulls in the minimum required to install the one packge you /actually/ asked for. What is it you're really trying to do? jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [newbie] Setting up network
On Tue, Mar 11, 2003 at 12:32:27AM +0200, Aryan Ameri wrote: > Hi there: > > Untill now, I only got online using a dial up connection. Now i am going to > dormitory, and in order to get online I have to connect to my dormitory's > LAN. I can setup the network using the following two commands: > > ifconfig eth0 my.ip.add.ress up > route add default gw my.gw.ip.address > > but it is so frustrating to issue these commands whenever I reboot my > computer. I know there are ways to automate this proccess and there are > progrms which setup the network on each boot. But I am not able to find out > how. Can anyone please help? Take a look at /etc/network/interfaces try adding an entry like this: (don't touch the "lo" entries!) auto eth0 iface eth0 inet static address your.ip.add.ress netmask your.ne.tm.ask #probably 255.255.255.0 or 255.255.0.0 network your.net.work.address broadcast your.broad.cast.address gateway your.default.gate.way and then you'll find that it comes up and down when you turn your computer on and off. You can even take it up and down manually while the computer's on with ifup eth0 and ifdown eth0 HTH, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Onboard rtl8139 works in 2.2 kernel but not in 2.4 kernel, please help
On Tue, Mar 11, 2003 at 11:07:10AM -0500, Fraser Campbell wrote: > Hi, [snip rtl8139 problems] No idea if this is feasible here, but my favourite way of solving 8139 problems is to put a decent nic in the box (Intel EtherExpress, Tulip, LinkSys - maybe, etc.) and ignore the PoS RTL. Seriously - drop £20-30 on a well thought-out nic and you'll not go far wrong :-) Sorry if that's not a possibility in this case, but it's the simplest and best way to deal with it! Cheers, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Onboard rtl8139 works in 2.2 kernel but not in 2.4 kernel, please help
On Wed, Mar 12, 2003 at 07:07:22PM +0100, PeterG wrote: > "Jonathan Matthews" <[EMAIL PROTECTED]> schrieb im > Newsbeitrag news:[EMAIL PROTECTED] > > On Tue, Mar 11, 2003 at 11:07:10AM -0500, Fraser Campbell wrote: > > > Hi, > > [snip rtl8139 problems] > > > > No idea if this is feasible here, but my favourite way of solving 8139 > > problems is to put a decent nic in the box (Intel EtherExpress, Tulip, > > LinkSys - maybe, etc.) and ignore the PoS RTL. > > > > Seriously - drop £20-30 on a well thought-out nic and you'll not go far > > wrong :-) > > > > Sorry if that's not a possibility in this case, but it's the simplest > > and best way to deal with it! > > > > Cheers, > > jc > > > Hi, > > this cant´t be the solution. > > I´m suffering and going crazy about this problem. > > I have a 3com 3c905b! [Please don't top-post] I wasn't addressing the problem of the nic working in 2.2 but not 2.4 (I've had this myself with a few RTL8139's), but the more general problem of using a RTL8139 at all. Friends don't let friends use Realtek nics! IIRC, the OP had a problem with which driver (rtl8139 vs. 8139too) to use - or if he didn't, he'd find that problem around the corner - which I don't think is your problem, seeing as neither driver relates to your card at all. So - what problem are /you/ having? > > peter jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Mouse/X
On Thu, Mar 13, 2003 at 10:06:16AM +, Olivier wrote: > Quoting "Paul M Foster" : > > -- was [EMAIL PROTECTED] on Mar 12, 2003 at 10:59:50 -- > > > I'll give you the solution someone gave me on my system. I'm using a > > Trident Cyberblade i/1 video card, so YMMV. > > That's quite interesting because my video card is a > Trident Cyberblade i/7 ... Same here ... > > > Go to your XF86Config-4 file (in /etc/X11) and find the section > > pertaining to your video card. Within that section, add the following: > > > > Option "SWCursor" "on" > > > > What I was told is that this causes X to position the cursor, rather > > than the video card (which apparently gets it wrong). Naturally, restart > > X afterward. > > I will investigate this way and test a little (since the problem > only appears from time to time) before reporting the results. I missed the start of this thread, but if it's "X thinks I'm clicking Zcm to the $LEFT_OR_RIGHT of where my mouse cursor actually is" (where Z=1.5 and LEFT_OR_RIGHT=left for me), then this advice is correct. I used to see the problem at least once a day (only solution was a reboot, though I learned to mentally adjust the cursor position and carry on regardless :-) but haven't seen it at all since adding the SWCursor line (as detailed above) to my X config file about 1 month ago. HTH, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [Possibly OT] can't I turn off message delivery?
On Thu, Mar 13, 2003 at 09:54:05AM -0800, linux learner wrote: > > Like it says at the bottom of every message: > > > > To UNSUBSCRIBE, email to > > [EMAIL PROTECTED] > > with a subject of "unsubscribe". > > Okay *MY BAD*, it was ambiguous perhaps. > > How do i turn off message delivery without losing my > posting privilages? Witness the occasional bit of spam getting through to the list - /everyone/ has posting rights, as it's not a subscribers-only list. You can safely unsubscribe and still post here, though I believe that you'll miss out on the most rewarding part of being here - reading interesting looking threads and helping out as your knowledge grows. HTH, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Mozilla stops accepting typed input
On Thu, Mar 13, 2003 at 12:40:51PM -0800, Martin J. Hillyer wrote: > I'm having an irritating problem with Mozilla. After what appears to > be a random length of time (often very short, eg, after one entry), it > stops accepting typing in, for example, the address box, or in a > google text input area. I have to kill that instance and start > another to get it to accept typed input. Clicking on links still > works normally. Does anyone have any ideas? > > I've thought that perhaps I'd remove a config file and let it > reconfigure, but I don't see any obvious rc files - 'locate mozillarc' > only gives /etc/mozilla/mozillarc, which has only one line. Nothing > obvious pops out from my ~/.mozilla directory - there's a binary file > in there called appreg, and two folders - /fonts, and > /; the latter only has the rather mysteriously named > folder /b3nzixf2.slt, which has lots of stuff relating to bookmarks, > cookies, etc, but nothing seems to stand out as possibly corrupted. > Nothing I've looked for in Google has turned up any clues, so I'm > turning to the list... > > I'm running testing, kernel 2.4.20, X 4.2.1 with an Athlon XP1500+, > 512 MB RAM and an nVidia GeForce2/MX-400 64MB video card. I have been > bitten by the AMD-AGP issue, but this seems to be pretty much under > control with the 2.4.20 kernel (it was a problem through 2.4.18). > > Please hold suggestions that I use a different browser; I know that > (and I've used various), but I would like to get this Mozilla problem > fixed because I like its many features (popups, cookie and login > management, eg). Well, I think you've just found another "feature" :-) Recent builds (can't remember how long ago it started) of Mozilla have included a "type-ahead" feature, roughly translating to the "find-next" feature in, say, lynx. It's loaded with the same key too - "/" - and I haven't personally found a way of telling Mozilla (Galeon, really) forget that it's in type-ahead mode. For now, I find lots of pressing of "/", , and other keys can flick it out of type-ahead. HTH, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: I need a little help
On Mon, Mar 17, 2003 at 04:00:43PM -0700, Didier Caamano wrote: > Is not that I don't want to share or soimething like that, is just I have > some scripts that need to be part of the web page code but they compromise > in some ways the security of the site and the privacy of those who are > part/members of the organization. > > As a result, I was wondering how could I hide the code, or the part of the > code that I don't want my visitors to see. I see now that there is no way, > or at least with apache. But I still need to hide at least those path for > the scripts that could compromise the site. The problem you're having is thinking that it's anything to /do/ with Apache - it's not! Even if you manage to "hide" the scripts, you're still going to come down to a single problem: if the scripts are destined to be run on the client - on the user's machine, not your server - then someone /will/ get hold of the source to them. So I'd suggest that the question then becomes "how can I write these scripts in a way that they don't compromise the security of the site/server/whatever?" The simple - but totally useless - answer is "don't trust the client." Why useless? Well, it doesn't tell you about /how/ to do it, just /what/ to do. That's all I can tell you, but I'm fairly sure it's the way you should be going. Remember - if your scripts can pass back information to your servers from the client machine, then anyone malicious can pass back carefully crafted data to take advantage of your servers. You /have/ to assume that this will be done so as to make sure that it has as little affect as possible! > By the way, thank very much to you guys for your answers. Have a nice day. > Didier. jc [CC'd you because - I don't know why - I just get the feeling that perhaps you're not subscribed :-] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Is my hard drive dying?
On Mon, Mar 24, 2003 at 10:16:28PM -, Andrew Pritchard wrote:
> I've been looking through the logcheck on one of my machines, and I've seen
> a lot of these types of messages:
>
> Mar 24 17:14:51 orion kernel: hda: dma_intr: status=0x51 { DriveReady
> SeekComplete Error }
> Mar 24 17:14:51 orion kernel: hda: dma_intr: error=0x40 {
> UncorrectableError }, LBAsect=3994439, sector=63232
> Mar 24 17:14:51 orion kernel: end_request: I/O error, dev 03:01 (hda),
> sector 63232
>
> Does this mean my drive is dying? I've not had any problems with the machine
> till now.
I'd tend to say yes - get your backups done now!
Can you tell from the logs if there was a hard start date, or did
the errors just start to trickle in? How often are they occuring?
jc
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: X whacking monitor
On Mon, Feb 03, 2003 at 12:51:53PM -0500, Mike M wrote: [snip] > This is also known as the Trident 9880 chip. It is on a Jaton 107AGP card. [snip] > I might be swapping out the Trident video cards for something else if the > problems persist. > Mike M. Mike - I think the prevailing wisdom when I enquired about this chip was "get another card"! It's a joke of a chip, with the only advantages being hardware functions that aren't (yet) available in linux. Something like video units or 3d something-or-others that are just dead silicon under linux. Yes, it works. Just not very well. If it wasn't soldered onto the d*mn motherboard, I'd have it straight out of there. If you have the option, because it's on a card, just dump it and get something decent. As far as I recall, it was being touted for mobile use because of integrated video stuff and low power - not differentiators that you're /really/ looking for in a desktop system, so $DEITY only knows why it's on my mobo! Grr. -- jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
PCI graphics cards recommendations
Hi everyone. Can anyone recommend/warn me off any PCI graphics cards that are still available at retail? I've googled a bit, but can't find any info written in the last year or so, or any relevant to X4.x FWIW I'm running X4.2, with a 15" CRT and a K6-III/550Mhz. I'd be looking to try Quake I/II/III (or as far as the K6 will let me go), and general desktoppy stuff. I'd like to have the facility to run in as high a resolution as possible at a reasonable refresh rate (where reasonable is, I spose, >65Hz). Thanks in advance, jc -- [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
How do I make quality PDFs from LaTeX?
Hi all.
I've tried googling lists.debian.org, but I don't get anything coming up
that seems to address my problem.
I'm trying to get a "nice" looking PDF from a latex document. I'm using
the normal article class, with no other packages loaded. I'm using a
couple of symbols that I wouldn't expect to find on a keyboard (namely
\times and \equiv) but apart from that I don't think I'm using anything
unusual.
I'm using pdflatex to output the PDF, and I'm aiming for print quality,
not screen quality, if they're not related.
I'm seeing a variety of problems in the .pdf output, but latex by itself
doesn't exhibit any of them when I create DVI output.
The problems I'm seeing include:
o Missing characters
- lowercase "a", fagawdsake!
o Wierd quoting
- the source document has
''a quote''
- the .dvi has
"a quote" - but with 66s and 99s (as I'd expect)
- however, the .pdf has
\a quote''
o Missing symbols
- \times
- \equiv
Generally, not a great looking document.
I'm hoping that someone can say "just install the 'foo' package, and
you're set", but I'd be willing to do a fair bit of digging around if
anyone thinks it'll help. On the other hand, I hope that maybe there's
another font package that needs installing? Prrrlease?
I'm running up-to-date sid with tetex-{base,bin,extra} installed. No
manual configuration's been done beyond apt-get's efforts.
Thanks for any help!
jc
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
OT: BT Broadband - which ADSL modem?
After a couple of years of um-ing and ah-ing, my dad's finally got round to installing broadband. Specifically, BT broadband (here in the uk). He's asked me to slip in a 486 class router/firewall inbetween his Windows machines and the ADSL modem, so I'd much rather go with a modem that has RJ45 (ethernet) connections over any USB-type port. I'm sure there are some debian users out there who can help me make this choice - Which ADSL modem o has an ethernet port o works with BT broadband o offers most bang per buck where bang/buck is measured ... however you want. The baseline is "get connected", but I'd be interested to hear any other pluses that different models provide. So - any thoughts? Cheers! jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: OT: BT Broadband - which ADSL modem?
On Thu, Sep 18, 2003 at 11:15:35PM +0100, Oliver Elphick wrote: > On Thu, 2003-09-18 at 18:21, Jonathan Matthews wrote: [snip] > > Which ADSL modem > > > > o has an ethernet port > > o works with BT broadband > > o offers most bang per buck > > I have combined the modem and firewall by installing a Bewan adsl card, > which has Linux support: > http://www.bewan.com/bewan/products/adsl/bwadslpcist.php Cheers for that - it looks rather nice. Hope you don't mind if I ask a couple of questions. Are there any issues with it being internal, as with winmodems? Does it steal much cpu (I hope to put this in a P60)? The PDF techspec says "standard ATM driver" - does this take much figuring out? Is it a kernel configuration issue? What spec PC do you have it in, and how many other NICs are there in it? Does it provide any other services? How loaded does it get? Thanks! jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: printer icon
Sebastian Kapfer had the gall to say: > On Thu, 02 Oct 2003 04:30:15 +0200, Zeeblanc wrote: > > > I have a Lexmark z 11 printer.Until I was connected to aol 9.0 optimized > > it worked fine the icon was on my task bar. > > Can anyone enlighten me why _Windows_ users keep posting their [CENSORED] > questions to this list? As I said on deb-vote recently (obviously a bit OT ... :-), I /strongly/ believe these mails to be email-gathering ploys. You'll notice that many of these 1- or 2-liners are very general, with many possible ways of diagnosing/solving the problem. I believe that this is to increase the number of replies from helpful PeeCee users who think "finally - someone with a problem that *I* can help with!!". I'd suggest tagging them as spam and letting bogofilter/spamassassin/$FILTER get rid of them. HTH, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Playing sound over the network
Hi list - I've got one server/music box attached to my hifi, networked to my main work box. My main box /has/ got a soundcard attached, but I'd really like to avoid having to compile soundcard support in to the kernel. Also, it's got on-board AC97, which I gather is a PoS. So, is there a way of getting /dev/dsp (or whatever) to be forwarded over the network, so that I can play stuff locally with apps not having to know that any jiggery-pokery is going on? It's a 10Mbps network, so bandwidth isn't /really/ a problem. The lower, the better, of course ... Any suggestions? Thanks in advance! cheers, -- jc -- It may stop, it may not. And stop calling me "dj". -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Playing sound over the network
On Mon, Oct 28, 2002 at 09:32:55PM +0100, Michael Schulze wrote: > > So, is there a way of getting /dev/dsp (or whatever) to be forwarded > > over the network, so that I can play stuff locally with apps not > > having to know that any jiggery-pokery is going on? > > i once used esound (Enlightened Sound Daemon) to forward sound over the > network. i.e. xmms comes with a output plugin for esound. Thanks Michael, but I'm really looking for something not tied into a gui, not needing any userspace programs to even know it's there. Maybe a kernel patch, or summat. Don't mind if the server box is running something hefty, but I'd like to keep things tight and lean on the client box ... Any takers? -- jc It may stop, it may not. And stop calling me "dj". -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: address book for mutt
On Fri, Nov 08, 2002 at 05:34:47PM +0530, Sandip P Deshmukh wrote: > On Fri, Nov 08, 2002 at 09:52:11PM +1100, Sam Varghese wrote: > > On Fri, Nov 08, 2002 at 03:51:28PM +0530, Sandip P Deshmukh spake thus: > > > > > i have recently migrated from windows to linux. currently i use woody 3.0. > > > in windows i used outlook and one of the important portions i will like to carry >over > > > in linux is the address book. i have converted ms outlook addressbook into csv >format. it > > > has several headings - first name, last name, work tel, home tel, etc. > > > i want to know is there anyway i can port it into linux and integrate it with >mutt? > > > if not, can someone suggest a solution? > > > > If it's Outlook 97, you can use the script available here: > > http://www.tuxedo.org/~esr/lookout/ > > sad - it is outlook 2000. but as i said, i already have the data in csv format - >fairly > universal? > > and how do i integrate it with mutt? Look at abook for a simple addressbook written for mutt. Its manpage says that it can use the --convert option to import different formats, including csv files. It's not graphical, flash, fancy, and doesn't use the mouse, but it works - sort of like mutt, in that respect :-) Also, take a look at the lbdb package - it might be of interest to you. I don't use it personally (yet), but have heard good things about it from others on this list. Hope this helps, -- jc -- It may stop, it may not. And stop calling me "dj". -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: address book for mutt
On Fri, Nov 08, 2002 at 09:08:02AM -0500, Matthew Weier O'Phinney wrote: > -- Jonathan Matthews <[EMAIL PROTECTED]> wrote > (on Friday, 08 November 2002, 01:40 PM +): > > Look at abook for a simple addressbook written for mutt. Its manpage > > says that it can use the --convert option to import different formats, > > including csv files. > > > > It's not graphical, flash, fancy, and doesn't use the mouse, but it > > works - sort of like mutt, in that respect :-) > > > > Also, take a look at the lbdb package - it might be of interest to you. > > I don't use it personally (yet), but have heard good things about it > > from others on this list. > > I have my addresses on a Palm, and use jpilot to sync to my debian box. > I use lbdb to grab that database, and thus I can use it with mutt. > > lbdb has quite a number of backends, and can be used very easily with > mutt. If you find an address book program you like, chances are you'll > be able to get at the data with lbdb. > > If you go with abook, mutt can directly query it with patches that are > included in the debian package -- in that case, you wouldn't need lbdb. I understood that mutt didn't need any patches to talk to abook - it's just a setting in the muttrc, and away you go - or is the "it" in "directly query it" your palm, and not abook? lbdb does look nice, tho, for when I actually get my ldap server up and running :-) Anyone got any nice pointers to locally serving ldap on a small scale? Went to the bookshop the other day, and there were maybe 15 published books from the last 5 years to date with "ldap" in the title! None by O'Reilly, either :-( cheers, -- jc -- It may stop, it may not. And stop calling me "dj". -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[OT] Practical differences between Netgear models
Evening all - I'm looking to take my home network up to 100Mbps, and I wondered if anyone could give me a clue as to the difference between a couple of netgear models that I've seen advertised at a fairly decent price (~35UKP+tax, I think) I'm looking at the FS105 & FS108, versus the FS608. The datasheets just don't give me much, except the number of ports and other /physical/ attributes. Are there any real, practical differences between these models that I should be aware of? The model I get will not be public facing in any way, so I believe that security, e.g. ARP spoofing, won't really be an issue. Having said that, you probably know more than I do ... in fact, I'd bet on it :-) Many thanks, -- jc PS how /do/ you spell "won't"? "wont"? Grr. -- [EMAIL PROTECTED] - use brain to mail -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [OT] Moving away from KDE to what?
On Tue, Nov 12, 2002 at 10:26:25PM +0100, Alex Polite wrote: > > > Here are the requirements: > > > > 1) Must be able to "maximize window to available space" a la > >enlightenment. > > 2) Must support multiple sequence key bindings a la emacs. > > 3) Must be fast. > > 4) Must be faster. > > Thanks everyone for all the input. > > I've installed blackbox. I doesn't meet requirements 1 and 2, contrary > to what someone in the thread said, but it does meet 3. I'll stay > with it for a couple of weeks. 1) Try "Full Maximisation" or "Maximise over slit" in the config menu. Can you describe a situation that would explain this better, if I've misunderstood? Try fluxbox, for a different take on the blackbox idea. I prefer it solely as it has a a better name. -- jc -- [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Preload Openoffice.org at X startup
Hi all - anyone got any hints on preloading Openoffice.org when X starts, sort of like the galeon -s option ... ? It's great, n'all, but a bit of a pig to start on my 550 K6-3! Failing any built-in, background server type functionality, would "cat > /dev/null &" in my .xinitrc be of any /real/ use? Thanks for any suggestions! -- jc -- [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
muttrc parse error - why??
Hi all - I'm having a problem with my .muttrc config. When starting mutt, it barfs over the line send-hook "~b [EMAIL PROTECTED]" my_hdr From: Name Surname <[EMAIL PROTECTED]> complaining "b: not supported in this mode". What's wrong with it? Can I not use ~ notation in a send hook? The online mutt manual gives the example of send-hook ~[EMAIL PROTECTED] my_hdr from: [EMAIL PROTECTED] which would suggest otherwise ... Cheers for any help! -- jc -- [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Proposal - non-free software removal
On Fri, Nov 15, 2002 at 08:50:33AM -0800, Steve Juranich wrote: > On Fri, 15 Nov 2002 10:42:14 -0200, Klaus Imgrund wrote: > > > why do people that don't want non-free .deb's just remove it from > > their sources line? > > Amen. > > Where is the original of this posting? All I can find on the > debian-user archives is the two responses. [snip] -vote, about a week ago, or so. -- jc -- [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: muttrc parse error - why??
On Fri, Nov 15, 2002 at 08:08:57PM +0100, Jens Kubieziel wrote: > On Fri, Nov 15, 2002 at 05:17:20PM +0000, Jonathan Matthews wrote: > > When starting mutt, it barfs over the line > > send-hook "~b [EMAIL PROTECTED]" my_hdr From: Name Surname <[EMAIL PROTECTED]> > ^ ^ ^ > Delete the '"' and the Space. Have done. Same error occurs. The line is now send-hook ~b [EMAIL PROTECTED] my_hdr From: name surname <[EMAIL PROTECTED]> and mutt tells me that b: not supported in this mode. Grr. > > The online mutt manual gives the example of > > send-hook ~[EMAIL PROTECTED] my_hdr from: [EMAIL PROTECTED] > > which would suggest otherwise ... > > Take a closer look at manual. ;-) Believe you me, I've spent many long hours at work looking at the manual - that's why I've posted here, since I simply can't see what's wrong with it! Has anyone any ideas how to say to mutt "If I'm writing a message where [EMAIL PROTECTED] appears /anywhere/ in the body or headers, then set the From: line to [EMAIL PROTECTED]"? -- jc PS Is it just me, or has the mutt-users mailing list dropped off the face of the earth in the last few weeks? -- [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[SOLVED] Re: muttrc parse error - why??
On Fri, Nov 15, 2002 at 05:17:20PM +, Jonathan Matthews wrote: > Hi all - > > I'm having a problem with my .muttrc config. > > When starting mutt, it barfs over the line > send-hook "~b [EMAIL PROTECTED]" my_hdr From: Name Surname <[EMAIL PROTECTED]> > > complaining > "b: not supported in this mode". > > What's wrong with it? > Can I not use ~ notation in a send hook? > > The online mutt manual gives the example of > send-hook ~[EMAIL PROTECTED] my_hdr from: [EMAIL PROTECTED] > which would suggest otherwise ... As the manual says, when evaluating one of send-hook, save-hook, fcc-hook or message-hook: "Mutt allows the use of the search pattern language for matching messages in hook commands. This works in exactly the same way as it would when limiting or searching the mailbox, except that you are restricted to those operators which match information mutt extracts from the header of the message (i.e. from, to, cc, date, subject, etc.)." I was trying to match on stuff that wasn't just in the header. Thanks to those who made suggestions. -- jc -- [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Mouse Pointer Problem
On Wed, Aug 06, 2003 at 05:56:41PM -0300, Guilherme A. Mendes wrote: > Hi everyone, > > I have a big problem with my mouse pointer: suddenly it loses the focus, > for example if I put the pointer under an icon, the pointer is not > exactly under that icon, it's some pixels left. > > I'm looking for the solution for a long time and anyone can figure out > what is happenning (I know 2 or 3 people with the same problem). > > I dont know if I can explain better, it's a strange bug! > > Does anyone know what is happening? You've got a Trident graphics card? Google for "trident xfree swcursor", possibly with my domain in there and you should come up with a solution. HTH, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Reliably transferring a large amount of data from one machine to another over DSL
On Sat, Aug 09, 2003 at 12:17:08AM +0100, Shri Shrikumar wrote: > Hi, > > I have two machines connected to the internet using DSL. What I would > like to do is backup one machine to the other. The files are compressed > and encrypted and then transferred usinc SCP. Atleast, thats what its > supposed to do. It does the first two steps fine but the transfer is > regularly unsuccessful - it aborts with "connection lost" or similart > (cant remember exact error message). The procedure is run from cron. Could it be to do with the CMs renewing their leases? Even if the IP doesn't change, perhaps the connection tracking "forgets" their connection after a new lease is acquired ... ? > Any ideas on a reliable way to transfer 1 or 2 gb of data over dsl ? How about setting up an rsync server on the machine to be backed up, and cron'ing a "get" job on the machine that hosts the backups? HTH, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
name conflict: ud / ud-ldap
I've just noticed that "man ud-ldap" gives the same manpage as "man ud", which is also the name of the uptime daemon package and binary. The ud-ldap manpage says that it represents the binary "ud", but I know from another debian box that both root and users can execute ud to get a summary of their highest uptimes. Tab-completing on "ud" without the ud package installed expands to ud-ldap, so it's not a binary name conflict, just a minor glitch in the ud-ldap manpage. I think. Is this something that is catered for in the installation of packages, or is it a conflict for which I should file a bug? If the latter, which package should it be under? Neither ldap-utils nor ud have any bugs associated with this ... Running mostly testing & a little unstable. jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: apt-get dist-upgrade bails
On Thu, Mar 27, 2003 at 08:40:39AM -0500, Patrick Wiseman wrote: > I decided to go to 'testing', so edited my sources.list, did 'apt-get > update', 'apt-get dist-upgrade' which, after installing lots of packages, > bailed with the following error: > > Errors were encountered while processing: > /var/cache/apt/archives/gdk-imlib-dev_1.9.14-6_i386.deb > E: Sub-process /usr/bin/dpkg returned an error code (1) > > So, how badly hosed is my system, and what do I need to do to complete the > upgrade? FWIW, it's advised to do apt-get update apt-get install apt dpkg tar ( ... any others anyone?) apt-get dist-upgrade ... rather than just a straight update; dist-upgrade. In this case, what does /another/ apt-get dist-upgrade do? jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Limit a process's CPU usage?
On Fri, Mar 28, 2003 at 12:26:54PM +0100, Joerg Johannes wrote: > Hi List > > I'm running [EMAIL PROTECTED] on my laptop, but the near 100% CPU usage results in > my > processor fan constatly being on. Is there a way on telling [EMAIL PROTECTED] to use > the CPU at no more than, say 30%? > I have already set it to nice 19, but this affects only the priority, so that > an other process can take over the CPU more easily, this does not reduce CPU > load. I think this might fit the bill - http://www.tls-technologies.com/CPU/cpu-main.html Any good? jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: buying a cd writer
On Sat, Mar 29, 2003 at 08:01:33PM +0100, [EMAIL PROTECTED] wrote: [snip] > Yes, agree, I have an old 12x lite on and a newer 40x. Works fine on > both windoze and debian. I'm also useing a Plextor 48x, very nice. And a > old Sony 12x and a Teac scsi 12x, also very nice. > > I'm useing scsi emulation on everything exept for the Teac ofcourse:) . > The good ting about scsi emu is that when you have a burner and a cd/dvd > you are able to copy cd's "on the fly". I've always wondered about this. Is it the case (as a local PC shop assistant tried to convince me recently) that having the reader and burner on the same IDE interface means that copying CDs is faster? As though the reader can put the data on the wire and the burner read it directly, without it having to go reader -> ide bus -> cpu (or mainboard) -> ide bus -> writer. Totally aside from this, how /do/ I copy a CD directly (in my case, from /dev/dvd to /dev/cdrom)? Can I do something like $ dd if=/dev/dvd | cdrecord - assuming that all of cdrecord's options are set correctly in /etc/default? Any caveats for audio CD versus data? cheers, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: kdm vs xdm
On Tue, Apr 01, 2003 at 11:09:08AM +0200, Mark Annandale wrote: > Hi guys > > This is probably a stupid question, but here goes. > > At the moment I boot into KDE using xdm. How do I change my setup to use kdm > instead. I changed the line in /etc/X11/default-display-manager from xdm to > kdm, however this then requires a startx to boot into kde. One way of doing this is to remove xdm. I'm sure others will point you to more satisfactory methods. # apt-get --purge install kdm xdm- Note the "-" after "xdm" - it says to remove the package even though you've specified an "install" command. The "--purge" is just there for tidyness - leave it out if you might reinstall xdm and want to keep its config files around in the interim period. Hope this helps, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: X screen shifts right !!.
On Tue, Apr 01, 2003 at 03:34:58PM +0100, Dave Restall - System Administrator wrote: > Hi, > > I'm having an irritating problem with Debian 3.0 on Intel. I'm using > olvwm on top of XFree86 Version 4.1.0.1. (it is over 12 months old). > Hardware is on board Trident Microsystems CyberBlade/i1 (rev 0). > > Machine runs fine, just a standard install, nothing special. However > from time to time (and I suspect it's xscreensaver related) I am unable > to move the mouse cursor to the far left of the screen (about 5%), I can > move it to the far right, in fact it goes off the edge of the visible > area. Top to bottom is all fine. I have tried resetting olvwm from > inside the system but the only certain way to cure the problem is to > close X _AND_ reset the box. If you don't reset, the problem is still > apparent when you reload olvwm. I have worked through the problem, > simply by moving the mouse cursor to the right of what I want to select > and clicking. If I am lucky and manage to catch the keyboard/mouse at > the right time (I think this is when xscreensaver is just about to start > a new sequence) then the cursor reverts back to its normal operation. > > The best way I can think of the problem is that the 'virtual' screen has > been offset to the right of the physical screen by about 5%. Everything > else remains the same, none of the applications move. > > Has anybody else seen this ? I've stfw and can't find anything. > Unfortunately, the problem is not easily repeatable. Used to have this myself. Solution seems (in that doing this solved it AFAICS) to be putting the SWCursor line in your X config file. Here's my compete section - Section "Device" Identifier "Trident Cyberblade 7i" Driver "trident" Option "SWCursor" "on" EndSection HTH, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: startx problem
On Sun, Apr 06, 2003 at 05:08:14PM +0300, yaron wrote: > Tran Tuan Anh wrote: > > >Hi all, > > > >I've just installed Debian, but cannot bring the window system up. > >When I run startx it gives "Fatal server error: no screens found". > > > >I am using Philips 107S monitor, > >Video adapter: GeForce2 MX, 64 MB > > > >Could anyone tell me how to set this up? > > > >Thanks a lot! > >Tuan Anh > > > > Yeah, RTFM. Useful, really useful. Tran, try this: man XF86Config-4 If it's more technical than you're used to - try again! Come and ask again if you've /really/ tried to work it out. People here aren't all as rude as the last answer you got, but they're not a hand-holding community - you get what you pay for :-) Hope this helps, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Security Questions
On Sun, Apr 06, 2003 at 09:09:47AM -0500, Thomas H. George,,, wrote: > I have read Security-Quickstart-HOWTO. > > I believe my home network has been compromised (my daughter received > returned emails she neversent) and plan to take drastic action. The > network consists of DSL modem, a wireless router and four computers. I > have no concerns about the family members and the houses in the > neighborhood are widely separated so it is very unlikely that the > wireless connection has been used by outsiders. The DSL link to the > internet is my concern. Here are my quesions: > > 1. How to erase hard drives? I plan to pull one computer off line and > reinstall Debian Woody and Windows from CD's (Regretably I still need > Windows for a few applications). Is reinstallation enough or must, and > can, the hard drives be wiped clean of any residual programs? Using Debian's installation program you can zero out all partitions easily (skipping, say, existing /home partitions ) and reinstall what you need on them. I don't believe that you need to do a low-level wipe of the disk. If you boot from CD 1, then even code/a virus in the boot sector of the disk won't be executed. > 2. What is the best Firewall? I have an old Compaq 486 machine with no > math coprocessor. I assume I can install two ethernet cards (I believe > it has two PCI slots, must look though), load Woody, set up iptables and > a sniffer and place it between the DSL modem and the wireless router. > When I am ready to put this firewall in place I have all the computers > off line. I will bring up the one that has its operating systems and > applications reinstsalled from CD's and download all the security > updates from Debian and Microsoft. The procedure can then be repeated > for the other computers. My firewall here: Dell 486sx/25Mhz, 1.2GB disk, 2 ISA NICs - one to the hub, one to my cable modem. PCI would be better, but if your f/w is only used as the router between inside and out, then you're unlikely to max out the ISA bus. I don't on my 600kbps connection. > 3. DHCP or static addresses? I have been using static addresses. I > believe I have seen in the references that it is possible to set the > wireless router to receive and transmit to these addresses only? If so, > is this the best approach? DHCP if you'd like to have the ability to plug random PCs into the network and have them find it, otherwise you'll not have a problem with static addresses. > 4. How to deal with a rogue computer? The fly in this ointment is my > grandson's laptop, a gift from his father (my daughter's ex-husband). > It came with XP Professional and I don't have the CD's to reistall it. > My grandson likes to go on the internet and also use our wireless > network to print his homework on one of the printers attached to the > fixed computers. Would it work and not compromise the system if I give > it a static address and instruct the other computer's on the network to > refuse any transmissions from this address? And could I then attach one > of the printers to the computer serving as the firewall and allow all > the computers on the network to use this printer without cmpromising the > system? Unfortunately not. I've got this problem with my SO's Win98 laptop. I just gave it an internal address that's iptable'd off by all the other machines. It isn't, though, secure. Any code that manages to execute on the XP box as Windows' equivalent of root (which is usually the result of the next exploit du jour) can just send out packets with any source address it likes. Also, it can just sit there and sniff your traffic. Without two physically distinct networks, you can't (AFAIK) really ensure that traffic isn't intercepted, and that traffic actually comes from where you think it does. That's why you should be putting everything (even locally) over ssh. That includes NFS, SMTP, whatever. A Bit Of A Bugger(tm), but neccessary. > I would greatly appreciate responses to the above questions and any > recommendations of alternate and, or additonal steps to secure the network. > > Tom George Hope this helps! If it's wrong, bring it back for a FULL refund ... jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: default run level
On Sun, Apr 06, 2003 at 07:20:12AM +0500, Ramsay D. Seielstad, KC2GMW wrote: > Timing is everything ... after several months with the basic, nothing-X > installed, I finally got around to adding sound and Xfree86. I also ended > up rebooting and rather than the usual text consoles for login got the X > login screen. > > I was going to post asking how to get rid of X unless I deliberately > started X, but rather than trying to remove the automatic start up I think > I'll use this solution for a while. # apt-get --purge remove xdm gdm kdm > Two items I'm not terribly clear about though - if I from > X into a text console, am I simply suspending X? And, can I return to X > when I'm done with my text console session or do I need to stop & restart X? Just return to it. It's still running there in the background - you're just not looking at it :-) Some graphic cards have a problem with this, evidenced by them displaying a blank screen when you change away from your X session and back again. I solved this by upgrading to X4.2. Before solving it, I just only started X when I needed a GUI, and closed it down when I'd finished. HTH, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Strange iptables behaviour (success/failure depends on presence of "harmless" logging line)
[ SUMMARY: Why does adding a "harmless" logging line to my iptables script let traffic pass (the desired behaviour) that is otherwise stopped? ] I'm having problems doing port forwarding through my old 486/25sx NAT box here. The only reason I mention its spec/age is that I'm having difficulty seeing what the issue is if it isn't a timing one. I've attached my firewall script. Any general or non-related comments are welcome, but the specific issue I have is that (after adding some forwarding lines to enable me to bypass the rather nasty "it's for performance reasons, not for security" http proxy at work so that I can connect to my home ssh server) the success or failure of the extra iptables filters seems to depend on whether I have the following line commented out or not: $IT --append drop-and-log --jump LOG --log-level info --log-prefix "Firewall: " where IT=/sbin/iptables, of course. When it's commented out, all external traffic (e.g. pinging www.google.com) doesn't return, but when it's active the traffic passes fine. As it should. The four lines (1 logging one, as above, and 3 "forward port 443 through into the lan" ones) are marked with # comments immediately before them for your ease of identification. I've included the whole file (it's only 3kb) because something this strange probably means I've fscked up somewhere /else/. Can anyone throw me a fricking clue - PLEASE? Cheers! jc PS Yes, I realise that I don't have to have the internal server running on port 443, but it's late and I've run out of coffee ... #!/bin/bash # # your system's details go here IT="/sbin/iptables" INTIP="192.168.0.1" INTIFACE="eth0" INTNET="192.168.0.0/24" EXTIP="$3" EXTIFACE="$2" DNS_SERVERS="" # what policy shall we follow when we've run out of ideas? POL="drop-and-log" # tell the kernel what we're going to be doing #echo "1" > /proc/sys/net/ipv4/ip_always_defrag echo "1" > /proc/sys/net/ipv4/ip_dynaddr echo "1" > /proc/sys/net/ipv4/conf/$EXTIFACE/rp_filter echo "1" > /proc/sys/net/ipv4/conf/$INTIFACE/rp_filter # set all chains to DROP - we'll never reach these ordinarily, # but they'll be useful when packets coming in while this # script is running, and the rules aren't all in place $IT --policy OUTPUT DROP $IT --policy INPUT DROP $IT --policy FORWARD DROP # flush all chains, all tables $IT --table filter --flush $IT --table nat--flush # delete user chains $IT --delete-chain # create new chains $IT --new-chain drop-and-log $IT --new-chain int-in $IT --new-chain int-out $IT --new-chain ext-in $IT --new-chain ext-out # start adding rules # drop-and-log chain # LOSE THE "#" AT THE START OF THE NEXT LINE, AND EVERYTHING WORKS # FINE!!! #$IT --append drop-and-log --jump LOG --log-level info --log-prefix "Firewall: " $IT --append drop-and-log --jump DROP # input chain $IT --append INPUT --jump ext-in --in-interface $EXTIFACE $IT --append INPUT --jump int-in --in-interface $INTIFACE $IT --append INPUT --jump ACCEPT --in-interface lo $IT --append INPUT --jump $POL # output chain $IT --append OUTPUT --jump ext-out --out-interface $EXTIFACE $IT --append OUTPUT --jump int-out --out-interface $INTIFACE $IT --append OUTPUT --jump ACCEPT --out-interface lo $IT --append OUTPUT --jump $POL # forward chain $IT --append FORWARD --jump ACCEPT --in-interface $EXTIFACE --out-interface $INTIFACE --match state --state ESTABLISHED,RELATED $IT --append FORWARD --jump ACCEPT --in-interface $INTIFACE --out-interface $EXTIFACE # LINE ADDED #1 $IT --append FORWARD --jump ACCEPT --in-interface $EXTIFACE --out-interface $INTIFACE --protocol TCP --destination-port 443 --match state --state NEW,ESTABLISHED,RELATED $IT --append FORWARD --jump $POL # ext-in chain #for PROTO in UDP TCP #do # for IP in $DNS_SERVERS # do # $IT --append ext-in --jump ACCEPT --protocol $PROTO --source $IP --destination $EXTIP --destination-port domain # done #done $IT --append ext-in --jump ACCEPT --match state --state ESTABLISHED,RELATED # LINE ADDED #2 # accept connections to the ssh server running internally $IT --append ext-in --jump ACCEPT --protocol TCP --destination-port 443 $IT --append ext-in --jump $POL # ext-out $IT --append ext-out --jump ACCEPT --source $EXTIP --destination ! $INTNET $IT --append ext-out --jump $POL # int-in # nasty hack to firewall off di's laptop, # except for routing to the 'net. $IT --append int-in --jump $POL --source 192.168.0.129 --destination $INTNET # standard int-in chain $IT --append int-in --jump ACCEPT --source $INTNET $IT --append int-in --jump $POL # int-out $IT --append int-out --jump ACCEPT --destination $INTNET $IT --append int-out --jump $POL # masquerade stuff going out on the external interface $IT --table nat --append POSTROUTING --jump MASQUERADE --out-interface $EXTIFACE # LINE ADDED #3 # forward traffic to bigdaddy $IT --table nat --append PREROUTING --jump DNAT --protocol TCP --destination $EXTIP --destination-port 443
Re: Problem with sqwebmail login (repost)
On Sun, Jun 15, 2003 at 02:49:58PM +0200, Sasa Babic wrote:
> /* Sorry if this is double post. I'm not sure if it got to the list (seems
> * not). */
>
> Debian stable & sqwebmail.
>
> I must be doing something wrong, because I'm not able to authenticate trough
> sqwebmail interface. Searched through Google already, but of no avail.
>
> /var/log/syslog:
> Jun 15 14:08:47 hygia webmail: authdaemon: s_connect() failed: Connection refused
[snip]
>
> Still, no success with login. What should I look for?
Total guess - /etc/hosts.{allow,deny}?
HTH,
jc
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Networked sound - any ideas?
Hi all. I've got a few machines on the internal LAN that I'd like to network the sound on. Only one of them is connected to the hi-fi, but they all have sound cards in. I've had a look at Esound and ARTS, but never had any luck with implementing a system-wide solution that my non-techie partner can "just use" by changing the hi-fi amp input selector to aux and firing up xmms. So - does anyone have any ideas and/or suggestions and/or experience with this sort of thing? Any help much appreciated, but if we're talking preferences, then I'd like it to be multi-user safe so that when two clients use it then it mixes the sound OR drops one silently on the floor - no "force the second client to hang until we can get a lock" problems. And finally, if it were possible to keep traffic low, that'd be nice. None of the clients really has the horse-power to encode high-quality oggs in realtime, but I imagine that there's a "just try to /halve/ the bandwidth - don't work /too/ hard" option out there somewhere ... Anything, anyone? Cheers! jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Networked sound - any ideas?
On Mon, Jun 16, 2003 at 12:32:25PM -0600, Jamin W. Collins wrote: > On Mon, Jun 16, 2003 at 07:58:21PM +0200, Andreas Schulz wrote: > > what's about setting up a sound server (streaming )on each machine, > > and the one with the hifi-plug joins the streams ?? > > Any suggestion on the package(s) to use for such a solution? Speaking as the OP - seconded! Are you talking icecast, or something like that? The whole idea (for me, anyway) of running a sound server on the hifi machine is that I can offload some of the processing overhead onto the server machine. The clients are 550/600/700Mhz boxes, so I'm trying to leave them as free as possible. Wouldn't running something like an icecast server on each client (so to speak) box defeat this aim? Yes - I know I didn't mention this in the OP :-P Also, I 'spose should mention that, while I /like/ the Esd idea of wrapping up /dev/dsp for all child processes of a specific early-init-time task so that all attempts to access /dev/dsp automatically use Esd without knowing it , I have heard that it can be a performance bottleneck if you have lots of processes. Has anyone any experience of this or have I been misinformed? Cheers! jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Plw
On Mon, Jun 16, 2003 at 12:50:36PM -0700, frosty frees wrote: >Hi there, u wont know me, but im wondering if you can take five minutes >out and help me out. [snip] Preaching to the converted, I guess, but I strongly suspect this to be email-address gathering spam. Reply off-list at your own risk. jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: "cut" command not working as expected
On Tue, Jun 17, 2003 at 08:54:59PM +0100, David selby wrote: > I need to get the first two file names from a directory ... If that's the spec, then I'd do ls | head -2 and that'd give you the first two files, unless you have 'ls' aliased to something else (ls -C, perhaps ..?) > My code > > directory=$(ls -r --format=single-column) > > works perfect and gives me ... > > 20030617Jun17.tar.gz 20030616Jun16.tar.gz 20030615Jun15.tar.gz > 20030614Jun14.tar.gz 20030613Jun13.tar.gz 20030612Jun12.tar.gz > 20030611Jun11.tar.gz 20030610Jun10.tar.gz 20030609Jun09.tar.gz > 20030608Jun08.tar.gz 20030607Jun07.tar.gz 20030606Jun06.tar.gz > 20030605Jun05.tar.gz 20030604Jun04.tar.gz 20030603Jun03.tar.gz > 20030602Jun02.tar.gz 20030601Jun01.tar.gz 20030531May31.tar.gz > 20030530May30.tar.gz 20030529May29.tar.gz 20030528May28.tar.gz > 20030527May27.tar.gz 20030526May26.tar.gz 20030525May25.tar.gz > 20030524May24.tar.gz 20030523May23.tar.gz 20030522May22.tar.gz > 20030521May21.tar.gz 20030520May20.tar.gz 20030519May19.tar.gz > 20030518May18.tar.gz 20030517May17.tar.gz 20030516May16.tar.gz > 20030515May15.tar.gz 20030514May14.tar.gz 20030513May13.tar.gz > 20030512May12.tar.gz 20030511May11.tar.gz 20030510May10.tar.gz > 20030509May09.tar.gz 20030508May08.tar.gz 20030507May07.tar.gz > 20030506May06.tar.gz 20030505May05.tar.gz 20030504May04.tar.gz > 20030503May03.tar.gz 20030426Apr26.tar.gz 20030419Apr19.tar.gz > 20030412Apr12.tar.gz 20030405Apr05.tar.gz 20030329Mar29.tar.gz > 20030322Mar22.tar.gz 20030315Mar15.tar.gz 20030308Mar08.tar.gz > 20030301Mar01.tar.gz 20030222Feb22.tar.gz 20030215Feb15.tar.gz > 20030208Feb08.tar.gz 20030205Feb05.tar.gz > > I want to cut the first two file names from the list ... To my way of > thinking this should be easy ... So you mean the first two files from each triple? That's not what your "ls -r --format=single-column" is going to give you, I'm afraid. > cut -d' ' -f2 $directory Notwithstanding what I've just said, I'd do the following instead of that line: echo $directory | cut -d' ' -f2 > The xterm goes nuts and ends up in hyroglyphics ! At a guess I would say > that the white space between .gz & 200... may not be space but may have > a different ascii value. I have not found a utility to display raw ASCII > for a file yet to check this out. No idea, sorry. Have you any reason to believe that the filenames have wierd characters in them? > I am a relative begginer at learning bash ... Don't worry - there's a definate "ah HA" point when it clicks :-) HTH, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Debian-user, PASSIVE MONEY GENERATOR!!!
On Sun, Jun 22, 2003 at 02:02:45AM +, Staver wrote: [snip] > JUST GO TO http://www.euroinvclub.com AND DEPOSIT NOW!!! [snip] $ wget http://www.euroinvclub.com 2>/dev/null >/dev/null $ grep "as" index.html This Account Has Been Suspended Please contact the billing/support department as soon as possible. $ Anyone else get warm fuzzies when they see this? :-) jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
HELP! Can't even ping a local host :-(
Hi - Sorry for this intrusion into your inbox for what I know will be a simple RTFM question, but the trouble is that I've RTFM lots to get this far (and things *were* working), and it's all gone pear-shaped simply (as far as *I* can see) because I moved my boxen up a flight of stairs! I can post a bunch of /etc files if needs be, but the simple problem is that I can't telnet to, ftp or even ping my local machines, on eth0. The most telling symptom (I hope) is that when host A pings host B, the hub only registers activity on host A's port, but when B pings A, both ports are active, yet both pings record 100% packet loss. Does this ring any bells with anybody? For the record, /etc/hosts.deny is disabled on both boxen, and they're both running Debian 2.2 (r0 on one, r2 on the other). Telnet, ftp (and any other service that is running on either box) ireturns a "no route to host", and unless I'm blind, /etc/network/interfaces and /etc/hosts tally on both machines. This was all working before the weekend, and although I'm sure I didn't do much to them over the duration, I was muckoing around with some scancodes to get my "Internet Ready" Compaq keyboard working. I've backed those changes out, but this hasn't improved things at all. I'd really, really (really) appreciate some help on this! As I said before, just shout out a /etc file you'd like to see, and it'll be in the post the next day. Cheers! jc
Concurrent NFS mounts
Hi - has anyone got any advice/thoughts on this situation? The setup is this. I've got two machines: one's my workstation (A), and the other's an older, stuffed-to-the-hilt-with-ram-but-not-very-fast P166 (B). A exports /home to B, and B exports /usr and /var/spool/[mail|news] to A. I can leave both on all the time, but since they're in my bedroom, I'd prefer to leave at least A off some of the time (ie night-time . . :-), and preferably leave both off. How can I solve the concurrent mounting problem that occurs if I turn them on together? I don't mind forcing a boot order - I'm thinking B first, so that A can find /usr (?), but then B doesn't find /home. I'd like to have both sets of NFS mounts in /etc/fstab so I don't have to manually run "mount A:/home /home" each time I boot B, after I'm sure that A has successfully exported /home. I suppose what I'm asking as well is: is there a way to make nfs do a mount "on-demand", i.e. when I log into B, and not at boot time? Any suggestions much appreciated! jc
autofs or automount?
Thanks to those who responded so . . succinctly to my post about needing to do on-demand NFS mounting. Just a quickie for the list, now: which should I use? What are the implications of using one over the other? Is either noticably faster, more robust, or simply better than the other? Equally, has anyone got any horror stories about either . . ? Ta! jc
Help! "nfs warning: mount version older than kernel"
Hi - I'm getting this pesky warning each time I use autofs now . . "nfs warning: mount version older than kernel" . . and it's really starting to annoy me, as I know that it's not true (well, it wasn't complaining yesterday!) Here's /proc/version: Linux version 2.2.18pre21 ([EMAIL PROTECTED]) (gcc version 2.7.2.3) #1 Sat Nov 18 18:47:15 EST 2000 And here's "mount --version" mount: mount-2.10f I did try to install some kernel-binary packages in the interim, but I cancelled this when dpkg warned me about nasty things that might happen. Is there some file I can touch and make this go away? cheers! jc
Re: Help! "nfs warning: mount version older than kernel"
From: Sean 'Shaleh' Perry <[EMAIL PROTECTED]> > > > > Is there some file I can touch and make this go away? > > > > no it is hardcoded in the code. If you simply can not survive with the > message, grab the util-linux source from woody and compile it on your box. > So why hasn't it been complaining until now? I find it hard to believe that the version of mount shipped with 2.2r2 is the "wrong" version for the stock kernel also shipped with r2 . . I'm convinced that this has a "simple" solution - be it touch, dpkg, or something seemingly unrelated! jc
Re: Help! "nfs warning: mount version older than kernel"
From: Sean 'Shaleh' Perry <[EMAIL PROTECTED]> [snip] > > the version in 2.2.r2 is the wrong version. They updated the kernel in r2 > without putting in a new mount package. Trust me, I maintained the mount > package at the time. > Ye-haw! I'll drink to that :-) Thanks for the quick response Sean! jc
vga= in lilo.conf
I'd like to have vga=9 (132Xsomething_or_other) in lilo.conf, but '9' as an option only crops up after a manual 'scan' at the "Press ENTER or type SCAN to . . " prompt. Lilo complains if I put vga=9 in, as it appears to be an invalid mode, but it works after I 'scan' for possible video modes. Any ideas on how to make it work nicely, without my having to be there at the console at boot time? Cheers! jc
Re: vga= in lilo.conf
Thanks guys. I had seen the hex mode numbers at the lilo stage, but then I remembered doc/lilo/Manual.txt: vga= alters the VGA mode set at startup. The values normal , extended , ask or a decimal number are recognized. (See also "Booting kernel images from a file".) And so I didn't think that hex would work. Moral Of The Story: RTFM, but don't T(rust)TFM :) jc
Re: Help: long file name
> When I mount Windows partition or cdrom, I receive a short file name. > How could I get the long file name supported for this? You might try mounting it as "vfat" instead of "fat" or "auto". jc
Backing up harddisk prior to failure
Hi all My server's harddisk is making weird noises and telling me at the console that various things went wrong. All in all, I reckon the drive is about to die. How would people go about transferring the whole system (dpkg details and all) over to another disk (which I'm just about to buy) and then removing the faulty one? I've got no problem with putting the two disks in concurrently, but I'm not sure how to go about moving it all, short of bzipping each partition up. Even then, would that actually guarantee a working system? As an aside, I'd also be interested in any suggestions for drive makes/models around the 10+gig mark that are competitively priced. Uk specific if poss, Oxford specific even better :-) It's to go in a P133, IDE system - nothing fancy here! cheers, jc
mutt & imap
Just been spending this morning reading the mutt manual and man pages, but can't find anything on this topic. When I use tab completion for the first time on my local vanilla UW imapd server, mutt complains, briefly, about not being able to find ~/.mh_profile. This is annoying. If I touch that file (on the server), then the folder listing that mutt comes up with has an extra folder (one of the "+" ones, that it thinks contain both messsages and folders), that contains exactly the same contents as the directory I've specified as $imap_home_namespace (e.g. ~/Mail). Annoying, this is. Is there any way to tell mutt that it shouldn't look for ~/.mh_profile and what's more, I'm not using MH at all. I'm using the latest mutt from stable, and I'd prefer not getting into compile-time options, if possible. jc
Re: Still having mail problems [New Thread] - logs, etc.
On Thu, Nov 29, 2001 at 03:22:52PM -0400, cmasters wrote:
> Greetings,
Hi
[snip]
> I'd appreciate it greatly if someone matching (or closely matching) the
> following conditions would post a sanitized versions of their relevent
> config files:
>
> 1. Mail is collected at your ISP. You currently use fetchmail to retrieve
> it and have it stored locally.
> 2. You are using exim as an MTA (I could never get sendmail to configure
> properly).
> 3. Procmail pre-sorts your email for you
> 4. You use mutt to view the pre-sorted email
I'm in exactly this situation ('cept it's split over 2 boxes, not all done on
one - shouldn't make a jot of difference)
This setup is only for one person.
You'll need a fresh/different ~/.procmailrc and ~/.fetchmailrc
for each user that wants to use this method.
Not ideal, I know, but I haven't worked out how to just run one fetchmail
instance for all users yet.
Would probably require registering the server and l/p details with some central
place so that a centralised cron job could get everyone's mail. Not what I'm
want, at the moment :-)
HTH
jc
Here're the files:
# ~/.fetchmailrc
# my options
set no bouncemail
set logfile /home/jaycee/fetchmail.log
# defaults
# NOTE: my user name there and here are the same.
# Check the fetchmail man page for syntax to map
# remote and local user names
defaults
proto pop3
user "" ssl
password ""
# uklinux.net
poll mail.uklinux.net
# end .fetchmailrc
# /etc/exim.conf
# comments and blank lines stripped.
qualify_domain = therock
local_domains = localhost:therock
local_domains_include_host = true
local_domains_include_host_literals = true
never_users = root
host_lookup = *
host_accept_relay = localhost
trusted_users = mail
smtp_verify = false
gecos_pattern = ^([^,:]*)
gecos_name = $1
smtp_accept_queue_per_connection = 1000
freeze_tell_mailmaster = true
received_header_text = "Received: \
${if def:sender_rcvhost {from ${sender_rcvhost}\n\t}\
{${if def:sender_ident {from ${sender_ident} }}\
${if def:sender_helo_name {(helo=${sender_helo_name})\n\t\
by ${primary_hostname} \
${if def:received_protocol {with ${received_protocol}}} \
(Exim ${version_number} #${compile_number} (Debian))\n\t\
id ${message_id}\
${if def:received_for {\n\tfor <$received_for>}}"
end
local_delivery:
driver = appendfile
group = mail
mode = 0660
mode_fail_narrower = false
envelope_to_add = true
file = /var/spool/mail/${local_part}
address_pipe:
driver = pipe
return_output
address_file:
driver = appendfile
address_directory:
driver = appendfile
no_from_hack
prefix = ""
suffix = ""
address_reply:
driver = autoreply
procmail_pipe:
driver = pipe
command = "/usr/bin/procmail -d ${local_part}"
return_path_add
delivery_date_add
envelope_to_add
check_string = "From "
escape_string = ">From "
user = $local_part
group = mail
remote_smtp:
driver = smtp
end
real_local:
prefix = real-
driver = localuser
transport = local_delivery
system_aliases:
driver = aliasfile
file_transport = address_file
pipe_transport = address_pipe
file = /etc/aliases
search_type = lsearch
procmail:
driver = localuser
transport = procmail_pipe
require_files = ${local_part}:+${home}:+${home}/.procmailrc:+/usr/bin/procmail
no_verify
userforward:
driver = forwardfile
file_transport = address_file
pipe_transport = address_pipe
reply_transport = address_reply
no_verify
check_ancestor
file = .forward
modemask = 002
filter
localuser:
driver = localuser
transport = local_delivery
end
smarthost:
driver = domainlist
transport = remote_smtp
route_list = "* kanyon bydns_a"
end
* * F,2h,15m; G,16h,2h,1.5; F,4d,8h
end
[EMAIL PROTECTED]${lookup{$1}lsearch{/etc/email-addresses}\
{$value}fail} bcfrF
# end exim.conf
# ~/.procmailrc
LOGFILE=/home/jaycee/procmail.log
VERBOSE=off
HOME=/home/jaycee
MAILDIR=$HOME/Mail
DEFAULT=$MAILDIR/inbox
# backup of everything
:0 c
archive/current
# debian
:0:
[EMAIL PROTECTED]
debian-news
:0:
[EMAIL PROTECTED]
debian-user
:0:
[EMAIL PROTECTED]
debian-isp
:0:
[EMAIL PROTECTED]
debian-security-announce
:0:
[EMAIL PROTECTED]
debian-security
:0:
[EMAIL PROTECTED]
debian-firewall
:0:
[EMAIL PROTECTED]
debian-java
:0:
[EMAIL PROTECTED]
debian-vote
# linux-kernel
:0
[EMAIL PROTECTED]
|/dev/null # HOW MUCH TRAFFIC? AAARGGH!
# end .procmailrc
# .forward
|/usr/bin/procmail
# end .forward
Problem with kernel modules aliasing. Maybe.
Hi all.
Perhaps not a debian question, but I have a sneaky feeling it's to do with
what's a module, what's not, and where I set up the aliases for the modules -
so I'm not *entirely* OT for the list ... :-)
After rolling my own kernel (vanilla 2.2.20), an "ifup eth0" locks the box
solid - no magic SysRq, no CtrlAltDel, nothing works 'cept a reset.
"strace ifconfig eth0 192.168.0.2" gives me a bundle of
access("/proc/sys/net/blah", R_OK) = -1
lines, ending in
access("/proc/sys/net/ash", R_OK) = -1
Everything it tries to access gives a -1.
Then the following three lines:
ioctl(4, SIOCSIFADDR, 0xba5c) = 0
ioctl(4, SIOCGIFFLAGS, 0xb9ac) = 0
ioctl(4, SIOCGIFFLAGS
and that's where it stops dead, cursor flashing the column after the second
"FFLAGS".
Things that I'd done include stripping down the stock 2.2.18pre17(?) config for
2.2.20, removing all PCI quirk, buggy RZ1000 (haven't got one, I think ..) type
entries.
Also, and this is why I think it's to do with the module
aliases/declaration/whatever, I've tried changing RealTek 8139 drivers from
rtl8139 to 8139too, simulatneously moving from a module to in-kernel driver.
So, basically what I'm saying is: does this ring any bells with anyone?
Any suggestions?
I've tried moving the 8139 driver to a module, but things are still a bit
fsck'd.
All help gratefully paid for. Next month. Really. The cheque'll be in the
post. Honest.
jc
Re: rtl8139 vs. 8139 too
On Sat, Dec 08, 2001 at 05:56:57PM +0100, Eduard Bloch wrote: > Hello, > > I would like to know which of these drivers works better for people in > kernel 2.2.x. > > Reason: one of them may be enabled as the default driver in the next > woody installation system (boot floppies). Well, on the same box, with both compiled as modules, having "alias eth0 rtl8139" in /etc/modules.conf works fine, but then changing that to 8139too crashes the box at "ifup -a" time. This is with 2.2.20 straight from uk.kernel.org. Incidentally, if anyone's got any suggestions, please shout out - it's doing my head in! So, at the moment, I'd vote for rtl8139. A lot of literature/lists mention its braindead behaviour, tho, so it may not be the best idea for debian as a whole - but just for me ... :) HTH jc
stable->testing wants to remove loads of packages ...
Hi all. Just wondering if anyone knows why the following's happening: # apt-get -f dist-upgrade The following packages will be REMOVED: console-tools gdm gnome-control-center gnome-gturing groff gs gv libguile6 locales man-db task-gnome-apps task-gnome-desktop task-gnome-games task-gnome-net task-x-window-system-core xbase-clients xf86setup xfonts-100dpi xfonts-75dpi xfonts-base xfonts-cjk xfonts-cyrillic xfonts-scalable xlib6g-dev xmanpages xpm4g The following NEW packages will be installed: cpp-2.95 defoma g++-2.95 gcc-2.95 gconf gnome-gnotski gnome-mime-data groff-base host ifupdown ipchains klogd libast1 libbonobo2 libbz2-1.0 libcap1 libcdparanoia0 libdb3 libdns5 libdps1 libefs1 libfreetype6 libgc6 libgconf11 libgdk-pixbuf-gnome2 libgdk-pixbuf2 libgimpprint1 libgnome-vfs-common libgnome-vfs0 libgpmg1 libgtk1.2-common libguile9 libimlib2 libisc4 libldap2 libltdl3 liblwres1 liboaf0 libpcap0 libpcre3 libperl5.6 libsasl7 libscrollkeeper0 libssl0.9.6 libstdc++2.10-glibc2.2 libttf2 libungif4g libwww0 libxaw6 libxaw7 net-tools netkit-inetd netkit-ping oaf perl perl-doc perl-modules perl-suid portmap python1.5-base python2.1 scrollkeeper xlibs xlibs-dev xscreensaver-gnome The following packages have been kept back base-config console-data gedit gnumeric libpaperg lilo tetex-bin xfs xserver-common xserver-svga xserver-vga16 xterm 261 packages upgraded, 65 newly installed, 26 to remove and 12 not upgraded. Need to get 135MB of archives. After unpacking 69.2MB will be used. Do you want to continue? [Y/n] n Abort. This happens after changing an up-to-date potato to point to testing, 'cept without the security sources.list entry. I'm not really sure that I want all that stuff to disappear, and it doesn't look like that cpio problem that was doing the rounds a while ago. Any help appreciated! jc
Re: smtp error during fetchmail
On Mon, Feb 04, 2002 at 06:37:14PM +, p wrote: > debs, > > when i fetchmail, i get an smtp error: > > fetchmail: smtp connect to localhost failed > fetchmail: smtp transaction error while fetching from pcisys.net > fetchmail: query status=10 (smtp) > > suggestions? > > thx. > > b. Check that /etc/hosts.allow contains a line like exim: , 127.0.0.1 or a line corresponding to whatever MTA you're using. For me, this line is: exim: 192.168.0.0/24, 127.0.0.1 There might be other reasons for your problem, but that's the one that bit me last time I moved my mail setup around. jc
Re: smtp error during fetchmail
On Tue, Feb 05, 2002 at 01:35:59PM +, p wrote:
> ...sorry (mutt, exim, & fetchmail).
ah, ic :-)
> after sleeping on it, i think i'm going to re-roll my kernel (2.4.0),
> as i recall trying to use my pcmcia modem, which would dial out but
> could never connect. (that was months ago since i last tried it.)
> ...seems like there has to be some relation to the modem problem and
> this smtp, "no-connect-to-localhost" issue.
I really don't think they're connected, but feel free to prove me
wrong :-)
> that box has the same configs as my box at the office. i must have
> screwed up the kernel.
Check you've got lo up: "ifup lo"
> i think you're right--i'll leave the "trifecta" alone. "kernel," here
> i come!
Hmmm. Let me know how you get on.
> p.s. i just realized that the box _can_ send email. i'm thinking smtp
> would be involved for that. (it's just that i can't _fetch_ my
> email. that must have been why i thought a re-install of fetch-
> mail was necessary.) (now i'm really confused.)
Nope - no reinstall req'd. The way it works is: (subject to change
once I spot any mistakes :-)
Incoming email:
Fetchmail connects to remote mail server, pulls in the mail.
Fetchmail passes the email directly to the MTA on the local machine
via the SMTP port, probably using "localhost" rather than the
hostname. (e.g. "bigdaddy" for me here, as you can see from the mail
headers)
The MTA recognises it's local mail, so it changes function to being a
MDA and (in the simplest case) sticks it in /var/mail/
Outgoing mail:
Your MUA (e.g. mutt, pine) connects to the local machine's SMTP port
and gives it the mail. The MTA listening to the SMTP port works out
that it has to send it out to your ISP's mail server (smarthost) for
delivery.
The difference is that your MUA is probably connecting to
:smtp, whereas fetchmail is connecting to
localhost:smtp. Somewhere else, say at the packet filter, you're
allowing access to :smtp, but not localhost:smtp. This
makes a difference, even tho they're essentially the same port, with
the same program listening!
Check your ipchains/iptables script, /etc/hosts.{deny,allow}, and your
xinted.conf file if you're running that (it can listen on specific
interfaces, which inetd can't (I think?)).
If none of that makes sense, blame my boss for working me too hard.
There's only so much VBA you can take in one day :-( How can they can
claim that something's Object-Oriented when you can't even extend a
class baffles me.
jc
Re: puzzling NFS problem
On Thu, Feb 07, 2002 at 02:54:03PM -0800, Henry House wrote:
> In the last few days, I've had an intermittent problem with a mail server
> that has /home mounted over NFS and delivers mail directly to user's
> directories. The following message appears over and over in the kernel log:
> Feb 7 14:45:14 romana kernel: lockd: task 168650 can't get a
request slot
[snip]
> [restarted nfs-common on romana and wotan, the nfs server]
> Feb 7 14:45:17 romana kernel: kmod: runaway modprobe loop assumed and
> stopped
> romana:~# Feb 7 14:45:17 romana last message repeated 4 times
> [all seems to be well now, no more log messages]
>
> In the place marked above, I have restarted nfs-common, which seems to cure
> the problem until it comes back. While lockd is complaining, the process
> table shows many instanced of exim in state 'D' (uninterruptible sleep,
> usually do to IO: obviously because of the NFS problem). While the trouble is
> occurring, the files in the NSF mounts still seem to be accessible to shell
> commands (unless the system has not run out of file descriptors).
>
> On Romana the NFS client drivers are modules.
Hi Henry - I had this a while ago.
It happened for me while I was fpt'ing large files from a windows box
to a linux one, and the directory I was copying /to/ was NFS mounted
from another linux box. The problem seemed to be that, due to the
inherent slowness in NFS file transfers, the available resources
of the NFS box that served the mount were requested, one after
another, faster than each finished. The write requests caught up with
the nfs server's capability to read them.
A workaround for this was to increase the number of nfsd instances on
the server, set in one of the nfs scripts in /etc/init.d. I can't
remember what the line was, and I haven't got nfs installed anymore, but
look for something like INSTANCES=8 near the top of
/etc/init.d/{nfs-server,nfs,nfs-common}. If you've never played with
it, it'll be set to the default (8). Increase this either until your
box blows, or the symptoms go away - your choice.
HTH,
jc
--
When you and friend disturb a bear, try to remember that, while running
away, it's not the /bear/ you have to outrun
Wierd 'ps ax' output
Can anyone help me work out what this entry in the process list is: PID TTY STAT TIME COMMAND [snip] 186 ?S 0:00 /usr/sbin/cron [snip] 16891 ?S 0:00 /USR/SBIN/CRON 16892 ?S 0:00 /bin/sh -c if [ -x /usr/sbin/exim -a -f /etc/exim/exim.conf ]; then /usr/sbin/exim -q ; fi 16893 ?S 0:00 /usr/sbin/exim -q 18318 ?S 0:00 /usr/sbin/exim -q 18319 ?S 0:00 /usr/sbin/exim -q 18320 ?S 0:00 /usr/bin/procmail 18321 ?S 0:00 /usr/sbin/exim -q I'm fine with #186, but why have I got #16891 there? According to pstree, it's the parent of the subsequent exim and procmail processes, but there's nothing in my /etc/crontab to suggest that anything should be going on around now. Does this smell fishy to anyone, or am I getting needlessly worried? ta, jc
Re: Wierd 'ps ax' output
On Wed, Feb 20, 2002 at 01:57:44PM -0800, Paul 'Baloo' Johnson wrote: > Needlessly worried, you probably restarted services somewhere along the > line. > > -- > Baloo I'm more worried (see previous post) about the fact that it's all in caps. Also, I know that /this/ machine was only turned on an hour ago, and nothing's been restarted. jc
Re: Wierd 'ps ax' output
On Wed, Feb 20, 2002 at 04:20:38PM -0600, Nathan E Norman wrote: > Needlessly worried. 16891 is a child of 186; you can tell because the > process name is in all caps (you could also tell if you did "ps > fax"). > > 16892 is the child of 16891; it's the script listed in > /etc/cron.d/exim, which runs every 15 minutes. Thanks! It was the caps that were bothering me. I didn't know that this was a property of child processes in 'ps' ... why is it that /all/ processes aren't in caps, being the children of init? Is it just to differentiate between parent and child where the child is a clone/fork/whatever of the parent? jc
suggestions for a console clock app . . .
Has anyone got a URL for the debs for a console clock app? I'm thinking xdaliclock sort of thing, but not under X: using the whole screen is important . . . . . so "while [ : ]; do date; done" wouldn't be helpful. . :-) cheers! jc
Re: any women here?
From: will trillich <[EMAIL PROTECTED]> > Robert wrote: > > This from debian-user: > > > http://sourceforge.net/projects/newbiedoc/ > > > > > >pick a topic that bugged you, and expose the basics for > > >tomorrow's newbie. (like me.) we need all the help we can get! > > > > Will: > > > > Like most places in GNU/Linux land, and IT in general for that matter, > > the project (newbiedoc) isn't getting "all the help" that it needs > > because: the help of women is lacking. It's sad to see 1/2 the human > > race is missing, once again. There are some women working on tech. > > projects here and there, but as we well know, it's male dominated. > > True. Sad. > > > Anyhow, that's a thought about "all the help we can get." In the case > > of NewbieDoc, there may actually be a chance that the participation of > > women can be had from the ground-up. That would be a coup. > > Unfortunately, I have no idea how one would go about reaching-out to > > the female half of the population, but it seems to me that NewbieDoc > > can do it since it's looking for newbies -- most other tech. pursuits > > are flooded with male experts. Maybe NewbieDoc can gain some women > > newbies... at least I hope so. > > That's not a bad idea. Women tend to have a more open-ended, inclusive > "take" on things than men do. > > Ladies, join us at http://newbiedoc.sourceForge.net, and help make > Debian a kinder, gentler place FWIW, the Java tutorials (trails?) at java.sun.com seem to be written by a four-women team . . . jc
Re: The Heart Is An Open Source: A Romance
From: Chris Gray <[EMAIL PROTECTED]> > On Tue, 17 Apr 2001, Jaldhar H. Vyas wrote: > > shell to /bin/false. For remote access they can use an IMAP client. > > For the IMAP server I recommend the uw-imapd-ssl package. I've > > heard the maintainer is very diligent and responsive, not to mention > > the only man on earth even more handsome than me." "But Flint," said > > (cgray4 ~)-> apt-cache show uw-imapd-ssl | grep Maintainer > Maintainer: Jaldhar H. Vyas <[EMAIL PROTECTED]> Who couldn't see *that* one coming a mile away . . . . . almost makes me want to install it just to read the man pages jc
bash woes
Ok - can someone explain the following: [EMAIL PROTECTED]:/tmp$ ls [EMAIL PROTECTED]:/tmp$ rm abc rm: cannot remove `abc': No such file or directory [EMAIL PROTECTED]:/tmp$ rm abc 2>err [EMAIL PROTECTED]:/tmp$ cat err rm: cannot remove `abc': No such file or directory [EMAIL PROTECTED]:/tmp$ rm err [EMAIL PROTECTED]:/tmp$ rm abc 2>&1 > err rm: cannot remove `abc': No such file or directory [EMAIL PROTECTED]:/tmp$ rm abc 2>&1 > /dev/null rm: cannot remove `abc': No such file or directory Why can't I silently discard the output of rm? Am I missing something subtle? Something obvious? A vital brain part? cheers, jc
Re: automounting home directories with nfs
> Hi All, > I'm trying to set up a small network and I'll be using NIS and NFS for > home directories. > > I was wondering if it's possible to only mount a user's home directory > when he tries to log in. This way I'll only have one home directory > mounted on the box at a time and I can avoid putting an unnecessary load > on the server. I recall seeing a solution for this somewhere. Went along the lines of nfs-mounting a /home/$USER directory for each user who logged on, but not mounting the whole of /home. Can't remember exactly where it was, but I think it was a Linux Gazette article. Try searching google for "nfs auto mount tutorial" and try the resulting LG pages. There were two articles, AFAIR, both by the same person. Either that, or the man pages or howtos . . . :-) They're the only places I would've seen the article . . For what it's worth, I'd just stick with automounting /home when a user logs on - it's not really costing you that much (any at all?) in terms of cpu or memory to have all of /home mounted. Then again, I could be wrong ... :) HTH, jc
NFS sharing /usr
Hi all- I'm NFS sharing /usr from one box to a number of other, with the aim of having only to do one apt update to the server machine (when I finally get my 28.8 modem working ... :-) and so as not to waste disk space on the client machines. I'd just like to know what anyone else in this situation does to keep /etc/ and other important directories up-to-date on the clients, so that /etc/alternatives works, and all the wm files under /etc (like menu hooks, etc) are consistent/present. Or am I going at this the wrong way - is NFS'ing the whole of /usr not a Good Thing? If so, what's the solution that results in a global /usr for the clients? cheers, jc
Re: NFS sharing /usr
From: nico de haer <[EMAIL PROTECTED]> > What would happen if you 'just' installed a deb on all pc's ? worst case: > they write the same file to /usr a few times. Best part: all pc's fix their > own /etc and related stuff. (do note: this is an idea, not something i've > tried) Well, ideally I'd like to have /usr mounted as ro to all the client boxes, with root_squash making sure that even as root on a client, I can't do a rm -rf /usr. Also, umounting /usr and writing to the local one on / wouldn't be great. I'm sure that there's a nicer way to do it than having to wipe the local one after each change. Otherwise, I might as well keep an independant /usr for each client, which kinda defeats the purpose of what I'm trying to do ... jc
Re: How to set time
David P James had the gall to say: > Considering that the user is (1) on ALO and (2) using a Windows > mailer (ALO it looks like) I have my doubts that the above is much > help... Agreed. Does anyone else find it easy to filter out such obvious "noise" solely on where the text versus the header appears in mutt? > On a related note, what's with all these LOA users asking about > removing ALO a.r.t files? Text above is obfuscated a touch to avoid what I'm about to describe. I reckon it's a vicious circle - they google/$BIG_ISP_SEARCH_ENGINE for the phrase "ALO a.r.t file" and since - waay back - someone asked on debian-user about it and d-u is presumably a fairly well-respected source in google's eyes, they get pointed here. Cycle repeats itself, leading to more and more OLA-users making their way to d-u to ask that one question. I keep meaning to set up a simple "Hello, user of $BIG_ISP. Here's where to go to get info on your a.r.t. files" page and post a link to it in reply to future questions on this subject. Has anyone dealt with a mom/pop/mum/dad and actually *solved* this problem? Where /should/ they go for instructions? Please avoid the unobfuscated versions of the words ALO and a.r.t. in your reply! Cheers, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: need advice on fixing my home lan
Paul E Condon had the gall to say: > I have a small LAN in my home. I need some advice on tuning it. > > I've started working on a project wherein I move large files (>3GB) > between two Debian boxes. This is a slow process. I would like to be > sure that it goes as fast as is reasonable. I think all my LAN cards > are claimed by their makers to be 10/100, but for some this might be > marketing hype. All my cables are 'CAT5'. So, some questions: > > How do I determine whether my lan is passing data at 10 or 100 MHz? I'd set up some sort of plain (i.e. not scp) data transfer test between the hosts. A temporary internal ftp/http server should do fine. Then do a "wget http://my.other.local.ip/a.large.file/"; and see what the average data rate is. ~1 meg per second == 10Mbps connection, ~10 megs per second == 100 Mbps connection. Cheers, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Woody-->Sid - any current gotchas?
Evenin' all. I'm about 2 hours away from getting my SOs new (well, third-hand) desktop machine up and running. Because she's a fairly non-technical sort, I want to give her the friendliest DE/WM I can - and I'm guessing that means something Gnome 2.4-y, which I'm only really going to get in Sid. (This isn't the main point of this email, but am I right here?) The install CDs I have are woody.r0, and I'm wondering if there any dist-upgrade gotchas in Sid at the moment? Also (it's been so long since I did it last) is it recommended to go Woody->Sarge->Sid, or is Woody->Sid preferred? Finally, is a net install a good/bad/indifferent idea? I'm fairly keen to try it for the first time, so any pointers would be appreciated! cheers, jc -- Please don't CC: me - reply on list instead. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Xfree setup: does AGP require anything special?
Doing a bit of advocay, I've managed to convince a friend to try Debian on his newly put-together PC for a bit, while he saves up the £165 WinXP tax. Unfortunately, I'm not doing a great job. I did a netinstall from debian-installer (12-01-04, FWIW, and it *rocks*!), but can't seem to get X to come up at all. I'll happily post any config files here, but (off the top of my head) the problem is that X bombs after a couple of seconds with "no screens found". He's got a NVidia FX mumble in there. If I could get the damn thing to work with even the vesa driver, I'd know where to start - but I can't. I've got mdetect, read-edid and discover installed (all before apt-get installing xserver-xfree86), but this doesn't seem to help. So - o any AGP gotchas I should look out for? I've only installed on a PCI system before .. o What's the lowest common denominator driver that I should be starting with to get over the absolutely-no-GUI-accessible hurdle? o What files can I post here to help you help me? TIA - I really need some advice! jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: simple bash loop problem ...
On Sat, Jun 28, 2003 at 03:03:27PM +0100, David selby wrote: > Hello, > > I am writing bash a bash & sed script, it has been going suprisingly > well. I need a loop to count 9 times & the variable n to the count .. [snip] for N in 1 2 3 4 5 6 7 8 9 do echo $N done I'm sure someone will point out a more elegant first line than this. HTH, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
woddy-proposed-updates vs. security
Just a quickie - Is there any difference between woody-proposed-updates and security.debian.org, for a stable machine? In other words, if I have stable and security in sources.list, am I missing out on /anything/ that's been updated by not having proposed-updates in there too? In other, other words, does proposed-updates carry non-security related updates that wouldn't be distributed on security.d.o? Any replies on-list, please ... :-) Cheers! jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: pipes, dpkg and default screen width (sort of)
On Fri, Jul 18, 2003 at 10:15:03AM -0400, charlie derr wrote: > Hi, > Often I find myself executing the following: > > $ dpkg -l '*foo*' > > to find all packages with foo in the name. > When instead I look for only installed foo packages as follows: > > $ dpkg -l '*foo*' | grep ii > > the output is truncated, and if there's a foo package with a > particularly long package name, I won't see the full name. > > Is there an envrironment variable I could set to prevent this? > Or some other workaround simpler than redirecting the first > command to a temporary file (and then using grep on that)? try COLUMNS=200 dpkg -l ... HTH, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: DVD+RW gone missing on moving to kernel 2.6.6
Graham Williams had the gall to say: > My "_NEC DVD+RW ND-2100AD, ATAPI CD/DVD-ROM drive" that is easily > identified under 2.4.25 (as /dev/hdc) goes missing under my 2.6.6 > kernel. dmesg has no hint at all of the DVD. This is my only IDE > device (using a SATA hard drive). > > Is this a matter of finding the right module? The machine is sid, > up-to-date. Just hit this problem myself. Here's what I did - let me know if it works: 1) Remove any "hdc=ide-scsi" lines from /etc/lilo.conf (grub? No idea...) and then run "lilo" as root. 2) # modprobe ide-cd 3) $ cdrecord dev=ATA -scanbus ... and use the x,y,z triple that cdrecord reports, as in $ cdrecord dev=ATA:x,y,z HTH, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Moving to a dedicated host...
[EMAIL PROTECTED] had the gall to say: > Hi all, [snip] > Any > experiences with other providers of decent non-managed Debian boxes with > a monthly datalimit of about 100GB? FWIW, I find www.bytemark.co.uk to be very good. They provide UML machines for a very reasonable amount. I can't remember what their bandwidth charges are, but I do recall a conversation with someone who indicated that b/w is at cost, not for profit. I'm just a happy customer - I've no other relationship with them! jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[OT] Taking notes
Fairly OT for d-u, but I'm wondering what people use to take notes for courses. I'm studying T171 with the Open University (it's a compulsory course on the way to their BSc in PeeCees), and a lot of the assessment is writing up what you thought about various resources (websites, reports, etc) that they throw at you. The important thing is the note-taking and subsequent writing up - /not/ the opinions you actually express in the notes. So - I'm looking for packages which let me keep a structured record of what I was looking at, where it was, when I looked, and what my thoughts were. I've found hnb, but that's about it. Any suggestions? Would a custom (v.v.v.v. simple) DTD be an idea? What emacs packages let me input the notes into a valid XML file adhering to my simple DTD in a pointy-clicky sort of way? Is this not the way forward? Cheers! jc -- This signature no verb. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Imap && imap-ssl && pop3-ssl
[Sorry for the cross-post - I think it's applicable to both -isp and -user.] I need to offer imap, imapssl and pop3ssl services. FWIW, imap would be localhost only, but -ssl services would be publically accessible. My reading thus far leads me towards Courier-imap with Exim 4 backported to stable so I can interface with ClamAV, but feel free to point out something important that I've missed. Do I need to have a different instance of the server running for each protocol? i.e. one listening on each port that the three services use as standard? Is there a server that would do the job with just one instance listening on all three ports? Would there be any advantages or disadvantages to this? I'm thinking locking/concurrency/that-sorta-thing. How do you deal with this situation? Are there any gotchas I need to know about? I'm guessing that using Maildirs will alleviate many of the problems that mboxes would create ... Any pointers/suggestions/cluebats appreciated! jc -- Have a nice cup of tea and a biscuit. Fifty-five million Britons can't be wrong. Ooo - jammy dodgers!! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: devfs.rules vs. udev.rules
csj had the gall to say:
> What's the relationship between /etc/udev/devfs.rules and
> /etc/udev/udev.rules? With the latest udev upgrade (0.024-6), I
> lost my video for linux devices. Before the upgrade they were in
> /dev/v4l/*.
>
> Sure enough I found that the rules for them had disappeared from
> /etc/udev/udev.rules. They're now in /etc/udev/devfs.rules. As
> a hack (rather than disabling udev entirely), I copied the
> relevant entries from /etc/udev/devfs.rules to
> /etc/udev/udev.rules.
I've just installed udev, and I found that moving devfs.rules to
udev.rules (after backing udev.rules up ...) solved the problems I had.
I think that devfs.rules is a compatability ruleset that creates /dev as
if devfs were running. I can't see any harm in this, so I've left my
box running with these rules active. I don't know what rules the
"native" udev.rules file propagates - my advice is not to use them until
it settles down a bit!
Cheers!
jc
--
jaycee . without . spam {at} jaycee [dot] uklinux (dot) net
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: OT: Viruses on lists
Paul Johnson had the gall to say: > "Derrick 'dman' Hudson" <[EMAIL PROTECTED]> writes: [snip] > > Almost. murphy generates a bounce and sends it to the list manager > > (mailman, majordomo, ezmlm, etc. - I don't know what one murphy is > > running). The list manager then counts that against you in its > > determination of which addresses are invalid and need to be removed > > from the list. > > It takes quite a few bounces before you get removed, though. Does anyone know a definitive figure or rate here? > > My choice is to simply drop viruses. I don't expect to have any legit > > messages falsely identified as viral, and dropping the message simply > > removes waste from the network bandwidth and disk storage of the > > world. I see no need to push the bounce back at someone else, > > particularly since the offender is rarely the one punished in that > > case. Drop /after/ accepting? Would that not mark you (in the virus' eyes, anyway) as a potential target? What with viruses having their own builtin SMTP engines these days and hence knowing for sure what response was given to the SMTP session, is that not potentially inviting future, smarter viruses (with memories for this sort of thing) to hit you first? > Which is why I reject at SMTP. Doesn't push a bounce back to forged > addresses. I should have said - I've followed Paul's instructions on ursine.ca to set this up, and am consequently rejecting at SMTP time. I'm unsure as to the difference between accepting a mail and bouncing later and rejecting at SMTP time as far as murphy is concerned. (I'm fine with the general difference for normal mail.) Can anyone venture an opinion? Do both bounces (is it correct to call a 5xx reject a "bounce"?) count similarly negatively when working out who shouldn't be on the list anymore? Should I stop asking questions (sort of like this one?) inside other questions? Answers on a postcard, please ... Cheers, Jonathan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
OT: Viruses on lists
Evenin' all. I've installed ClamAV+Exim4 to reject viruses at SMTP time. d-u's headers don't seem to mention anything about /virus/ scanning (as opposed to SpamAssassin), so I guess I'm ok asking this question here: The whole point of having virus scanning while the sender still has an open connection is a) to reduce email processing load on your system and b) to reduce bounces to forged headers - which must be sent if the email is accepted and only scanned later. I'm fine with (a) - I think that still holds - but is (b) incorrect when dealing with listmail? Since the mail has already been received and accepted by murphy, am I just pushing the sending of spoofed bounce messages one stage back up the email processing ladder? Is it an unfriendly thing to do to murphy - should I be whitelisting it instead? Any thoughts on this, or how to configure the exceptions inside Exim would be appreciated! Cheers, jc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: No access but root
Pritpal Dhaliwal had the gall to say: > no one else has a clue? > > come on guys ( and girls).. its the debian users list.. If I wanted to > start doing fresh install when things get messed up.. I could stick to > windows... > > help me out please Please remember you're talking to /volunteers/ here. Also, please learn not to put your replies at the top of email. As to your problem, I believe you've hit the fact that, if a script file does an action that is not permitted, then the individual line that causes the problem is not automatically echoed (giving you a clue about how to debug it) but instead just says "bash: : ". In other words, I believe that your problem is caused by a line /inside/ /etc/bash.bashrc, not the permissions on the file itself. Have a look through, and see if there's anything that you wouldn't expect a normal user to be able to do. As a last resort (you /should/ be able to fix this!) post the contents of the file here for people to look at. Cheers, Jonathan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: iptables; some IPs are getting through netmasks
On 23 December 2012 16:41, Mark Ford wrote:
> Here is a shortened version of the output from iptables-save (full version
> simply has more "-A pests" lines).
>
> # Generated by iptables-save v1.4.8 on Sun Dec 23 16:24:43 2012
> *filter
> :INPUT ACCEPT [252417:278747603]
> :FORWARD ACCEPT [0:0]
> :OUTPUT ACCEPT [255016:258290199]
> :pests - [0:0]
> -A INPUT -p tcp -j pests
> -A pests -s 1.85.17.0/24 -p tcp -j DROP
> -A pests -s 67.228.245.0/24 -p tcp -j DROP
> COMMIT
> # Completed on Sun Dec 23 16:24:44 2012
>
>
> Here is the complete header from the spam email...
[snipped]
I would trust what you find in /var/log/{mail,exim} more readily than
Received: lines in a spam mail, no matter how correctly you think
you're reading them. I'd check there instead.
As an aside, I wouldn't block /24s like this myself. Use something in
protocol (i.e. configured in Exim), perhaps, and be /really/ careful
about blocking entire /24s. The collateral damage could be more than
you intended.
Cheers,
Jonathan
--
Jonathan Matthews // Oxford, London, UK
http://www.jpluscplusm.com/contact.html
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive:
http://lists.debian.org/cakstx7a_nwcj81ewwfv+h8n4aunhtuebd3hft_jjnc0ho60...@mail.gmail.com
