How X keyboard maps work in Deb
Does Debian put the 'kbd' file in /etc/X11? Is it even needed under Debian or is some other mechanism used to do keyboard mapping? If it turns out that I need the file, what package contains it? I have been informed that I need this file to solve an application problem (described below) but before I research their prescribed fix, I want to make sure that Deb doesn't add some additional configuration indirection that will render the fix inaffectual, or, worse, mess up the Debian configuration for future upgrades. I use Blender on two Woody workstations and have noticed that the numeric keypad controls work correctly on one system and not at all on the other. The primary difference between the two systems is one was originally a Potato and the other has been a Woody all its life. The one that works is the Woody-since-birth machine. The Blender FAQ's tell me to check for a file called 'kbd' in /etc/X11 . It's not there on the broken machine. I will check for it's presence on the working machine once I get home tonight but I'd really like to get this machine working today. The Blender FAQ suggests that I re-install all the files in /etc/X11 and re-configure XFree86 but I'm not willing to perform that kind of radical surgery unless it is absolutely necessary. Thanks, -=greg
XML editor/schema developer
Hi All- I just started working with the demo version of XML Spy and it's pretty darned cool. Trouble is, it's Windows-only and I'm trying to ultimately eliminate the back-and-forthness of my desktop environment. Can anyone recommend an equivalent (more or less) package for Debian? It doesn't have to be free (but it has to be reasonable ). I mainly want the schema visualization and gener- ation capabililties. TIA, -=greg
Re: init.d: writing custom init-scripts
On Thursday, December 20, 2001 2:19 PM, [EMAIL PROTECTED] > it should be :-) Well, it is an x-less station and but to go shure i > also put the symlinks to runlevels 4,5 and 6 too. I just checked a Potato and it defaults to RL 2 . -g
Re: init.d: writing custom init-scripts
On Thursday, December 20, 2001 2:06 PM, [EMAIL PROTECTED] wrote: > Then, i placed a symlink called S99myscript pointing to the script under > /etc/rc3.d, however, it does not get stalled. The symlink and the script > [...] > When running the script manually as user root, it succeedes, but it is > not run by the init deamon on machine startup. You've probably checked, but is your default runlevel, indeed, 3? -=greg
Re: Enabling/disabling Ultra DMA for ATA drives
On Tuesday, December 18, 2001 6:48 AM, [EMAIL PROTECTED] wrote: > How it is possible to enable/disable Ultra DMA for selected ATA drives? Use hdparm. A while back someone posted a link to a decent hdparm introduction. Check subject headers for "hdparm". > Is UDMA enabled by default in Woody? No. -=greg
Re: JBoss & Linux Threads
On Sunday, December 16, 2001 11:52 PM, [EMAIL PROTECTED] > > each thread uses up one available process in the > So the "bad" thing about this is that you need one PID for each > thread? What's the advantage of lightweight threads compared to > intra-process threads. I think you mean to compare lightweight to process- based mechanisms. I would think that any disadvantage was due to the overhead of using the kernel process mechanism. Non-kernel threads can make assumptions that a kernel process cannot so additional work is required to acquire such a resource. Also, I would guess that the context switching for the process-based threads costs more as well. This is all just speculation on my part. There are many people on this list who are far more expert with this kind of thing. > Could you describe your installation a little further? How many concurrent > users are there in your installation? Did you have to tune some kernel/JVM > parameters to make it work (better)? It's strictly a development platform for proof- of-concept work. No load to speak of so the out-of-box config works fine. > What's the URL for the official JBoss manual? Haven't found it. >From jboss.org, you can purchase the manuals. Just go to the documentation area. Good luck, -=greg
Re: JBoss & Linux Threads
On Sunday, December 16, 2001 8:52 AM, [EMAIL PROTECTED] > http://www.jboss.org/online-manual/HTML/ch10s02.html > I read that Linux allegedly does not support "real threads". My questions > on this issue are: Linux is obviously multi-threaded but I think they are referring to so-called "lightweight" threads. IIRC, Linux creates intra-process threads via the same mechanism it uses to create processes--i.e. each thread uses up one available process in the table. I recall there was some big change wrt threads in libc6 but I don't remember, off-hand, what that was. > Maybe someone subscribed to this mailing list has already used JBoss on > Linux, is familiar with Linux threads and can enlighten me on this issue? I'm just getting JBoss going in Linux. It's working fine for me but I haven't checked performance vs. Windows. The OL manual's claim of doubled per- formance under Win is a bummer, if true. Remember, though, the online JBoss manual is the original one and not the official one. Peace, -=greg
Re: auto laptop interfaces [ was: apt-get update: Could not connect...]
On Thursday, December 13, 2001 1:58 PM, dman wrote: > The only problem left (for me) is : how can it automatically > determine whether or not to use DHCP (that is, is the link from home > or not). I don't think that can be automated (GPS ;-)?). What I > really need to do is configure dhcpd at home sometime. Someone here posted that he had a program that would check for the presence of another host on the LAN. I'm not sure if it was based on MAC or IP but if you can't find that, I would think some kind of arp-based script could tell you if a specific interface was nearby. Then just create a couple of virtual interfaces that are mapped via that script: one for DHCP, one for static. -g
Re: Printing HOWTOs a problem. Newbie #61
On Thursday, December 13, 2001 2:26 PM, [EMAIL PROTECTED] wrote: > I print out some of the HOWTOs for reading. > > say ... > > $ zless /usr/share/doc/HOWTO/en-txt/Net-HOWTO.txt.gz | pr -o4 | lpr > > After a few pages the printing suddenly slips down the page and is then across > the tear line of the continuous stationary that I use. This is usually because the number of lines per page your printer uses is different from your output. On a line printer, you can either use line feeds to advance your page, or send a form feed to ask the printer to do it. Some printers will automatically advance a page once a certain number of lines are printed. You probably just need to adjust the number of lines pr assumes your printer expects per page and maybe set it to use form feeds (see the -l and -f options). But, depen- ding on your hardware, you might have to change some of its settings. Some lpr's try to cook your output on the way out so you might be fighting that as well (some even use pr as a filter :) ). You'll need to experiment. Best, -=greg
Re: auto laptop interfaces [ was: apt-get update: Could not connect...]
On Thursday, December 13, 2001 12:26 PM, dman wrote: > Usually this means your network interface isn't up. (like if I turn > on the laptop, but forget to "sudo ifup eth0" first) In case you're interested, here's a strategy to configure laptop ethernics automatically--even if you have a hardware config that changes regularly (docking, PCMCIA, whatever). http://www.orthogony.com/gjw/lap/lap-ether-intro.html -=greg
Re: tin and the resolver..
On Thursday, December 13, 2001 12:23 PM, [EMAIL PROTECTED] wrote: > the problem is that despite the fact I have the > correct order in /etc/host.conf (host,bind) and all the entries in > /etc/hosts properly configured in every machine on my network tin tries to > query the dns for my local nntp server.. Check /etc/nsswitch.conf to make sure lookups are using the file method instead of the dns method (or at least trying file resolve first). -=greg
Re: kernel suddenly not linking
Ach. Looks like I'm not the only one. I gather that the linker is is now less forgiving. -g - Original Message - From: "Greg Wiley" <[EMAIL PROTECTED]> To: Sent: Thursday, December 13, 2001 9:34 AM Subject: kernel suddenly not linking > Hi all- > > I have successfully built the 2.4.16 kernel several > times but suddenly cannot. I am running an up-to- > date Woody and am wondering if some recent package > update(s) might be the problem. Does anyone have > any idea what would cause the linker output in the > attached text? > > TIA, > -=greg > > >
kernel suddenly not linking
Hi all- I have successfully built the 2.4.16 kernel several times but suddenly cannot. I am running an up-to- date Woody and am wondering if some recent package update(s) might be the problem. Does anyone have any idea what would cause the linker output in the attached text? TIA, -=greg ld -m elf_i386 -T /home/greg/build/linux-2.4.16/arch/i386/vmlinux.lds -e stext arch/i386/kernel/head.o arch/i386/kernel/init_task.o init/main.o init/version.o \ --start-group \ arch/i386/kernel/kernel.o arch/i386/mm/mm.o kernel/kernel.o mm/mm.o fs/fs.o ipc/ipc.o \ drivers/char/char.o drivers/block/block.o drivers/misc/misc.o drivers/net/net.o drivers/media/media.o drivers/char/drm/drm.o drivers/ide/idedriver.o drivers/cdrom/driver.o drivers/sound/sounddrivers.o drivers/pci/driver.o drivers/pnp/pnp.o drivers/video/video.o \ net/network.o \ /home/greg/build/linux-2.4.16/arch/i386/lib/lib.a /home/greg/build/linux-2.4.16/lib/lib.a /home/greg/build/linux-2.4.16/arch/i386/lib/lib.a \ --end-group \ -o vmlinux drivers/char/char.o(.data+0x46b4): undefined reference to `local symbols in discarded section .text.exit' drivers/net/net.o(.data+0x174): undefined reference to `local symbols in discarded section .text.exit' drivers/net/net.o(.data+0xd74): undefined reference to `local symbols in discarded section .text.exit' make: *** [vmlinux] Error 1
Re: login + passwd for 30 users
on Friday, December 07, 2001 4:57 PM, [EMAIL PROTECTED] wrote: > does any script exist, which creates 30 users automaticaly, > with the list of 30 persons containing their name, christian name, etc... > instead of using 30 times "adduser" ? It's been a number of hours so you've probably entered them already, but try using "useradd" instead. It takes its parameters from the command line so you could easily loop through a file read in sh and populate the command options. -=greg
Re: vnc and inetd - I thought I knew what I was doing . . . ;-)
On Wednesday, December 05, 2001 8:59 AM, [EMAIL PROTECTED] wrote: > login. netstat shows xdmcp only listening over udp. > > service1:/home/jvincent# netstat -l | grep xdm > udp0 0 *:xdmcp *:* Actually, that is only telling you that *something* is listening on the xdmcp port, 177. To find the process that's listening, use the -p option to netstat. Hope it gets you a step closer, -=greg
sticky directories
Aw, nuts. Brain finally kicked in. The bit I needed is the setgid bit, not the sticky. Sorry for the noise. -=greg
sticky directories
Good day- IIRC, on another Unix, I was able to use the sticky bit on a directory to cause new entries in that directory to inherit ownership and permissions. In Debian, the semantics of the directory sticky bit are to add additional re- strictions on modifying the directory's entries. Q: is there some way in Deb to cause new files to inherit ownership and permissions? In particular, I want files created by certain users in a certain directory to be owned by a group that each belongs to but that is not the primary group of any of them. TIA, -=greg
Re: exim - what is required?
On Monday, December 03, 2001 11:29 AM, dman wrote: > Say, does exim allow sourcing other filter files, like INCLUDERC in > procmail? Nope. Annoyingly, it doesn't do alot of things. Someday I'm going to get annoyed enough to learn procmail. :) -=greg
Re: finger, as it pertains to mail
On Monday, December 03, 2001 8:40 AM, [EMAIL PROTECTED] wrote: > I recently did a large mail upgrade and installed a new mail server that > incorporates mail hashing (i.e /var/spool/u/s/user). As expected, my > finger command on any user now reports that the user has no mail. Is this > an actual finger problem, or does something have to be defined in the > users home directory? Any help would be greatly appreciated. Thanks! The default finger program uses the _PATH_MAILDIR #define from , which is /var/mail, to locate a user's mailfile. There is no logic or hook to find an alternate location. There are configurable finger daemons packaged in Deb, one of them (cfingerd?) lets you run arbitrary scripts. Hope it helps, -=greg
Re: exim - what is required?
On Saturday, December 01, 2001 6:30 PM, [EMAIL PROTECTED] wrote: > What do I need for exim > to run only as an mda? It will be run by sendmail via my .forward > file as procmail is currently run. You will have to set up mailspool and queue directories, I think. Exim, like sendmail, delivers a single message by default. So you should be able to run it from procmail by simply executing exim with the destination address as an argument. > How can I tell exim to use something other than .forward for the > filter file since that is already used by sendmail? It is, I believe, the "file" configuration item in the forwardfile director section. Is this just to see how exim filtering works or are you trying to solve a specific problem that procmail can't? Good luck, -=greg
Re: Executable link rejected due to loaction or path
On Sunday, December 02, 2001 1:37 PM, [EMAIL PROTECTED] > Second how can I set up Apache to allow use > of this this script & the data files for it? Debian puts cgi programs into /usr/lib/cgi-bin. Whatever goes in there becomes part of the server namespace as /cgi-bin/pick_cgi (if pick_cgi is the name of your program) and will be executed as a CGI program rather than served as a file when it is requested. Rather than use the Debian directory, however, you might consider a local directory for your own programs. So make a place for your programs, such as /usr/local/lib/cgi-bin, and modify the Apache srm.conf file to add the new directory as a ScriptAlias. There are also options to make CGI available to user URLs (~user) if that is more appropriate to your setup. In any case, read the Apache docs. A badly designed/configured executable in in a public place can burn your house down. -=greg
wu-ftp vulnerability
According to /., wu-ftp has a vulnerability that allows root access to files. Check out: http://www.securityfocus.com/archive/1/242750 Debian 2.2 is on the list. Apologies if this has been discussed already. -=greg
advice needed on portals and weblogs
Good day! I need to set up an organizational web site that will have some characteristics of a weblog (for editorial and news) and some of a group site ( shared calendar and contacts, etc.) . I will need to develop some specialized applications over time. This is for a church so it will have some public functionality and some internal, protected functionality. My goals include extensibility, low operational maintenance, and Debian compatibility. I have looked at several content management systems for a basis and wonder what Deb users' experiences are. Some notes on the research so far: - I started installing Slash and found that it requires mods to Debian-installed components. It also appears to be a resource hog. - PHP-Nuke seems good except I am concerned about the tight coupling of logic/presentation and how it will affect extensibility and the resultant code maintenance. - Zope and Zope CMF are also interesting, particularly as they are Deb packages and are built for extensibility. My only concern is that I haven't seen any Z-CMF examples, including the dog bowl, that really connect to what I am doing. Since all the solutions require a considerable time investment for research and implementation, I am just looking for some pre-advice before I get too far down any particular road. Thanks in advance, -=greg
[SOLVED] Laptop ethernet dock and PCMCIA config
Hi all- A while ago, I asked how to configure the ethernet interfaces on a laptop such that if the system was docked in its ether-enabled station, it should config that interface and ignore any PCMCIA ethernet. But if out of its dock, it should check for the PCMCIA ethercard and use DHCP to bring it up. The ifupdown suite seemed, at first glance, to be insufficient, primarily because it uses the interfaces configuration individually for each interface and thus the interface mapping is not passed any broader context information. Yesterday, I finally attempted to tackle this thing and I have documented the solution at: http://www.orthogony.com/gjw/lap/lap-ether-intro.html It's also a tutorialish treatment of the ifupdown suite. Peace, -=greg
Re: Annoying file access behavior
On Wednesday, November 14, 2001 9:12 PM, [EMAIL PROTECTED] wrote > I have a bit of a problem. Every time I work with big files (movies, > converting mp3s to wavs for ripping), my system slows down to a crawl for > some time, while it is emptying the buffers I had a similar problem the first time I used IDE instead of SCSI. I was able to eradicate it (and how!) using hdparm--particularly the dma and irq masking options. Hope it helps. -=greg
new Xsession mechanism
Good day! I am wondering if there is a correct place to put X session shutdown code now that .../Xsession.d is used. Kind of like S vs. K in /etc/init.d . After studying all the scripts in the lifecycle of an X session, I don't see any such mechanism. That doesn't mean there isn't one, tho. If a display manager is used, no problem since it will normally have a reset script. But for startx/xinit, is there a hook for cleanup code? Specifically, I want to kill processes that have daemonized themselves. They don't always die with the X server. As analternative, I can put cleanup into ~/.xsession but that means I can't exec the window manager. TIA, -=greg
Annoying Mozilla behavior
I like Mozilla a bunch but it has an annoying behavior that deserves a wishlist bug unless it's due to my own stupidity: It doesn't remember the last page position when returning via "back". That is, it always returns to the top of the previous page whenever I hit the "back" button rather than return to the last place viewed on the page.. This might be related to a second bothersome behavior: it doesn't seem to honor anchors. Again, it always puts me at the top of the page regardless of any "#anchor" included in the URL. Can this be fixed by a configuration setting? Version is M18-3. Thank you, -=greg
Re: ldso part 2
On Saturday, October 27, 2001 4:16 AM, [EMAIL PROTECTED] > > I gave removing > > ldso a shot on a non-critical testing machine. > > [...] > > Today's upgrade put ldso right back on. > > [...] > > Now orphaner says, again, that nothing > > depends on ldso and that it may be safely > > removed. > Try 'dpkg -p ldso' and 'apt-cache show ldso'. Do either of those display > an 'Essential: yes' field? What tool are you using to upgrade? Both commands yield a "yes" in the "Essential:" field. I use apt-get to upgrade. Since the package is in the optional section, I would assume it is not essential. But that may only indicate a misunderstanding on my part of the Debian use of those terms. Thanks, -=greg
ldso part 2
As suggested by orphaner, I gave removing ldso a shot on a non-critical testing machine. No adverse affects as far as I could tell. Today's upgrade put ldso right back on. Now orphaner says, again, that nothing depends on ldso and that it may be safely removed. Is this a bug? If so, against what package should it be filed? Best, -=greg
ldso remove warning
Just slightly confused here. I tried the deborphan package on a testing machine today and was given a list of packages to remove. They included ldso. Now, I was pretty sure ldso is an important package so I checked using dselect. Hey, ldso is in opt/oldlibs. So I go ahead and try the removal but here is what I get: WARNING: The following essential packages will be removed This should NOT be done unless you know exactly what you are doing! ldso 0 packages upgraded, 0 newly installed, 1 to remove and 0 not upgraded. Need to get 0B of archives. After unpacking 684kB will be freed. You are about to do something potentially harmful To continue type in the phrase 'Yes, do as I say!' ?] Abort. The 'Abort' is because I didn't type in that phrase. Can I safely remove this package? Why is ldso in oldlibs? TIA, -=greg
Re: Messages being frozen with exim.
On Tuesday, October 23, 2001 1:50 PM, [EMAIL PROTECTED] > Lately in my log file for exim I see a bunch of messages that certain mail > messages are being frozen. > Where should I look to see the reason for this nd how do I unfreeze them? Mail that can neither be delivered nor returned is frozen by exim. You can search your logs and/or queue to determine the offending addresses. Often, it's just blind spam. Sometimes, however, mail that looks OK gets frozen. This is sometimes the result of temporary network problems (particularly if you are using a "smart host" to handle your mail) but could indicate a misconfiguration. That could take a more detailed analysis to solve. You can use eximon or or the various -Mxxx exim options to deal with problem messages. If you have a global problem, once you fix it, you can run exim -qff to thaw all frozen messages and retry delivery. Hope this helps, -=greg
Best approach for transient NICs
Good day! I know I can hack this together but I am looking for the right (meaning balance of elegant, simple, friendly to upgrades, etc.) way to do this: I have a portable that can have 0, 1, or 2 NICs installed depending on whether or not the PCMCIA card is inserted and whether or not it is docked in its port replicator. When in the replicator, the PCMCIA card, if if present should be ignored. The replicator NIC should be configured with a static IP address. When not in the replicator, if the PCMCIA card is inserted, the system should attempt to configure the card via DHCP. Any general or specific advice? Thanks, -=greg
Re: removing "beeps"
On Tuesday, October 23, 2001 9:33 AM, [EMAIL PROTECTED] > Everytime I enter something wrong into bash it sounds "beep" out of my > computer. > Can i remove that? Check out "info rluserman" . You can set the bell of readline to audible, visible, or off completely by creating a ~/.inputrc file with the command: set bell-style
Re: Why no man pages?
On Friday, October 19, 2001 1:50 PM, [EMAIL PROTECTED] wrote: > When i run the man command I get this error back: > > "bash: man: command not found" > > However, I do have directories in /usr/man/...? I had a Woody upgrade that silently removed man support from my box. apt-get install man-db did the trick. Hope it works, -=greg
Re: X/WindowManager: Mouse jumping around
On Friday, October 19, 2001 1:42 PM, [EMAIL PROTECTED] wrote: > When I startx the mouse jumps around wildly. It works for some seconds, > then it hops around, also pressing buttons. This makes working impossible, > of course :( I don't know, some days I think random input is as good as any I can come up with. > I have Debian Sid, a Logitech cordless Desktop, so a PS/2 mouse, which is > configured in X as standard "configured" mouse on dev/psaux. Sounds like a protocol problem. I don't know the Logitech cordless but start with setting protocol to "ps/2" rather than "auto". man XF86Config gives the list so you can try others. For some reason, "auto" doesn't work on any /dev/psaux mice in my experience. It has also happened to me after a bad Open/GL app exited but I think you probably aren't running anything before the problem starts. Hope it helps, -=greg
Re: exim and delayed deliveries
On Thursday, October 18, 2001 8:12 PM, [EMAIL PROTECTED] wrote: > The problem that sometimes arises is that with a mainly default > install, exim simply queues messages for delivery but doesn't actually > deliver them until a cron job runs and tells it to. > > My question is : is there a way to (asynchronously so mutt's > performance doesn't suffer) make exim send the messages immediately > rather than just queueing them? It isn't run by a cron job in the default install--rather, a -q is passed on the command line which tells exim how often to launch queue-runners. Make sure that the "queue_remote_domains" option is not set in your exim.conf (also that the -odqr option is not passed on the cmd line). queue_remote_domains is used to reduce the number of connections to remote hosts but it is not really useful for workstations. Best, -=greg
Re: XDM starting when not wanted
On Friday, October 19, 2001 8:22 AM, [EMAIL PROTECTED] > How can I stop XDM from starting on > bootup? Easiest, IMO, is to just remove xdm: apt-get remove xdm But if you still want the package, you could use update-rc.d to disable the xdm startup script. Best, -=greg
Re: a challenge
On Thursday, October 18, 2001 10:52 AM, [EMAIL PROTECTED] > If the attacker knows the algorithm (although not the prime number) this is > unfortunately trivial to crack: they just have to guess the time that is > encoded by the timestamp. :( You're right. I solved it as if the timestamp was unpredictable. Better to use a one-way and append it to the plaintext. But if an MD5 digest is used, it should probably be based on some function of the timestamp and a secret as Hahn suggested. -=greg
Re: a challenge
On Thursday, October 18, 2001 4:58 AM, [EMAIL PROTECTED] wrote: > goal: a 4-16 byte 7-bit character value that somehow encodes the time > of creation such that it can be extracted if the encoding scheme/seed > is known. the encoded value should be such that it is mostly > impossible to change it so as to yield a later time of creation to be > encoded. in general, changing the encoded value may well render the > data invalid. This is just a general idea: I'm assuming a 32-bit timestamp. You want to encode that value somewhat securely with an invertable hash into 16 * 7 = 112 bits (less if you can't use control chars). It has to be somewhat resistant to tampering (you gave no indication of the degree to which it must resist). So you can multiply your timestamp by any number < 2^(112-32) = 2^80 = 1208925819614629174706176 . and still fit in 16 7-bit chars. I would pick a prime value < 2^80 and multiply the time by it, then format it into 7-bit characters. To extract, pack it and divide. Does that work? Note that to break this, the attacker will have to search a fairly large space but not an impossible one. Some of the effectiveness of this will depend on how often you change the prime factor. Also, if it is fairly expensive for an attacker to check a single value, the search could take a while. This also doesn't take into account the value of breaking it, i.e. how big is the "prize" for finding your prime factor? Best, -=greg
New startx behavior in testing?
Since last night's Woody upgrade, startx no longer works as before. Now, it completely ignores ~/.xsession and does not even create a default session if .xsession does not exist which leads me to believe that Xsession is never sourced. This means that the X server quits as soon as it's started. I tried xinit--symlinking ~/.xinitrc to ~/.xsession--and it worked fine. I accidentally left the link in place and tried startx...voila! it worked. Removing the link breaks it again. Is this purposeful behavior? The only related bug against xbase-clients seems to be 111009 but the resolution to that seems to require man-page fix, not a change in session behavior. Best, -=greg
Re: Upgrade Catch-22
On Tuesday, October 16, 2001 2:20 PM, [EMAIL PROTECTED] wrote: > It sounds like you are going to have to use method 2. > edit /var/lib/dpkg/status to trick dpkg into thinking modutils are not installed. > Get a hold of modutils 2.4.8.1 from somewhere. > Install this manually with dpkg -i > When you use aptitude select modutils place a "hold" on it. Becareful you must place a "hold" before every get. I just finished a firewire-free kernel compile and the problem is now solved. It looks like there is another thread re: this problem and that it can also be solved by simply removing the offending ( sbp2.o ) module from the tree. doh. Thanks again, -=greg
Re: programs not appearing on desktop menu
On Tuesday, October 16, 2001 10:28 AM, [EMAIL PROTECTED] wrote: > On 16-Oct-2001 joe golden wrote: > > I am running the testing version of debian and have recently installed > > povray, zed, circlepack and cooledit among others. None of these recently > > installed programs appear on my desktop. > as root run 'update-menus'. A recent problem with the xterm menu file in /usr/lib/menu causes update-menus to fail, leaving you with whatever menu you had before that file was installed. If that is the problem, in the xterm file change: title=XTerm (Unicode)\ to: title="XTerm (Unicode)"\ Peace, -=greg
Re: Typespeed without bad words
On Tuesday, October 16, 2001 4:30 PM, [EMAIL PROTECTED] wrote: > Typespeed is a good typing tutoring program. > Any cleaned up versions of this good program available? You can edit the word files themselves in /usr/share/games/typespeed According to README.Debian for typespeed your changes will be erased whenever typespeed is upgraded so make sure to back up whatever word files you create. Also, the startup file mechanism has been disabled in the Debian version so you can't point it to arbitrary word lists. Take care, -=greg
Upgrade Catch-22
Argh. I have a Woody machine that attempted to go through an upgrade last night and is now in dpkg jail. ppp tries to upgrade but bails with: Unpacking replacement ppp ... depmod: Unexpected value (20) in '/lib/modules/2.4.9-686/kern el/drivers/ieee1394/sbp2.o' for ieee1394_device_size It is likely that the kernel structure has changed, if so then you probably need a new version of modutils to handle this kernel. Check linux/Documentation/Changes. dpkg: warning - old post-removal script returned error exit status 255 dpkg - trying script from the new package instead ... [ same depmod err as above] dpkg: error processing /var/cache/apt/archives/ppp_2.4.1.uus-1_i3 86.deb (--unpack): subprocess new post-removal script returned error exit status 255 [ same depmod err as above] dpkg: error while cleaning up: subprocess post-removal script returned error exit status 255 Errors were encountered while processing: /var/cache/apt/archives/ppp_2.4.1.uus-1_i386.deb E: Sub-process /usr/bin/dpkg returned an error code (1) But trying to remove ppp balks with: dpkg: error processing ppp (--remove): Package is in a very bad inconsistent state - you should reinstall it before attempting a removal. Errors were encountered while processing: ppp E: Sub-process /usr/bin/dpkg returned an error code (1) So I can neither go forward nor back. Upgrading modutils directly claims I have the most recent version. I am using the 2.4.9-686 kernel image from the Woody distribution. I don't have a firewire device, nor is the module listed in my module config. I don't even need ppp on this machine. I could ignore this error but it is holding up the upgrade of ~50 other packages. Any nudges in the right direction are greatly appreciated. Thanks, -=greg BTW, I broke up some of the quoted lines above so they don't exceed normal term width. Is there a convention for reporting long-line output in email?
Re: PCMCIA under potato
On Sunday, October 14, 2001 7:23 AM, [EMAIL PROTECTED] wrote: > I've installed and compiled the pcmcia source coming with Bunk's stuff > under potato and kernel 2.4.9 (of course enabling pcmcia in the > kernel). Then I've installed pcmcia-cs. > > Now when I start pcmcia service with two cards inserted (a modem and an ethernet card) the answer is as follows: > > debian:/home/dada# /etc/init.d/pcmcia start > Starting PCMCIA services: modules/lib/modules/2.4.9/pcmcia/i82365.o: init_module: No such device > Hint: insmod errors can be caused by incorrect module parameters, including invalid IO or IRQ parameters > /lib/modules/2.4.9/pcmcia/ds.o: init_module: Operation not permitted > Hint: insmod errors can be caused by incorrect module parameters, including invalid IO or IRQ parameters > cardmgr. I had the same problem upon upgrading to 2.4.9. The modules are organized differently. To get the net xfc up, I have to load modules in this order on my Dell Inspiron 5000 (obviously, your ether card may be different) : pcmcia_core yenta_socket ds cb_enabler 3c59x < I used to load 3c575_cb but now one module does it all. For some reason I still don't udnerstand, the i82365 module won't load at all on my machine. The above setup works, however. Best, -=greg
Re: I'm in a loop
On Thursday, October 11, 2001 9:56 AM, [EMAIL PROTECTED] wrote: > After making choices (series of questions), > I'm told that I've reached the moment of truth. > Excited, I follow the instruction to reboot and > find myself back at the welcome screen which > leads nowhere except through the same > questions. My guess is you need to take the floppy out of the drive before you reboot. If you chose the 'make a boot floppy' option, then put that one in. Take care, -=greg
Re: Fonts gone in Mozilla
On Wednesday, October 10, 2001 12:28 PM, [EMAIL PROTECTED] wrote: > > Instead, I get mostly dashed-rectangular boxes > Check out http://www.debian.org/News/weekly/2001/23/ on "Fonts missing after > upgrade." Thanks Ray, that was it exactly. It looks like the upgrade didn't restart the xfs daemon. A simple xfs restart was all it took. Best, -=greg
Re: Fonts gone in Mozilla
On Wednesday, October 10, 2001 8:16 AM, [EMAIL PROTECTED] wrote: > [...] > Mozilla will not display any characters in the > rendered-page area. > [...] A clue! neither xscreensaver prefs nor demo will show characters in their dialogs. Now, how do I find out what fonts they are looking for? -g
Fonts gone in Mozilla
Hi All- I just upgraded another potato to woody and ran across this weirdness. Mozilla will not display any characters in the rendered-page area. All Mozilla menus, buttons, labels, etc. are fine. But no characters from HTML pages show up. Instead, I get mostly dashed-rectangular boxes but some slashes and dots also. I have changed font sizes and tried various pages. Neither the X server nor Mozilla complain about missing fonts. No font complaints in any of the system logs, either. Any suggestions on where to start hunting down the problem? TIA, -=greg
Re: kdm, blackbox, and .xinitrc
On Monday, October 08, 2001 8:49 PM, egm2@jps.net wrote: > On Mon, 08 Oct 2001 23:27:49 -0400 (EDT), Alexander N Gould <[EMAIL PROTECTED]> wrote: > > > On Mon, 8 Oct 2001, Eric G. Miller wrote: > > > > > On Mon, 08 Oct 2001 23:03:58 -0400 (EDT), Alexander N Gould <[EMAIL PROTECTED]> wrote: > > > > > > > My .xinitrc file does not seem to run when I start a blackbox session form > > > > > > Rename it to ~/.xsession. > > > > > > > Thanks; I tried that, but still nothing happens. Any other ideas, please? > > Maybe you could clarify. Does blackbox get started? Does some other > window manager get started? Does nothing happen? I'm not too familiar > with kdm these days, does it give you an option about what type of > session to start? If so, did you choose "Xsession" or "Debian"? With KDM, the ~/.xsession is run only when you choose a "default" session type from the login screen. ~/.xinitrc is ignored as far as I can tell. In the past, I have modified the KDM session scripts so they source an init script from the user's home regardless of the session type. Regards, -=greg
Re: Sources for tulip
On: Monday, October 08, 2001 1:55 PM, [EMAIL PROTECTED] wrote: > But there is nothing in the /usr/src directory since I installed from the > net, and I guess I didn't install any sources... > Coan someone tell me what I need to apt-get apt-get install kernel-source- Where is the, well, version. > and then how to recompile the > module? I know that recompiling the kernel will recompile the module, but > I don't even know yet how to do that... Kernel-HOWTO.txt.gz on your system (if you installed the docs) gives you a great step-by step. Best, -=greg
Re: X fonts problem
On Wednesday, October 03, 2001 11:32 AM, [EMAIL PROTECTED] wrote: > certain programs, like "plan", "dia", > display text as dotted rectangles, and not even one per letter. In some programs, that means appropriately-sized screen fonts are not available. -=greg
nslookup disappearing?
This message appears when I run Woody's nslookup: Note: nslookup is deprecated and may be removed from future releases. consider using the 'dig' or 'host' programs instead. Run nslookup with the '-si[lent]' option to prevent this message from appearing. Has nslookup really fallen out of favor with users? What's wrong with it? I've always considered dig a low-level tool mostly for problem-solving. Host is not interactive AFAIK. If nslookup goes, how long before we lose ping and traceroute? :) Best, -=greg
ide tuning ( results regarding: hdparm or ide howto)
Thanks to those who pointed to resources. They gave me a great start and I have solved some of the problems I had resigned myself to live with since switching to IDE. I am now investigating the kernel's support for my particular chipset and think I can get an additional siginificant bump in sustained data rates. For anyone with newer IDE drives that hasn't read these texts, do it. The default drive config is really conservative. You will be amazed at the results. Best, -=greg
hdparm or ide howto
Good day all. Is anyone aware of a document or set of ducuments that describe Linux (E)IDE support? It seems to me that some sort of primer on tuning IDE and demysti- fying the newer technologies such as ATA and UDMA would be a significant HOWTO asset. This is prompted by the recent discussion on this list which poked at a buried suspicion I have had that my (pretty new) IDE HDDs are not performing at their best. I get unexplained system slowdowns that I never had on much older/slower systems that used SCSI. TIA, -=greg
Re: auto power off capabilites under linux
On Thursday, September 27, 2001 5:34 PM, [EMAIL PROTECTED] wrote: > Is it possible to get some sort of driver to allow > debian to auto power-off the machine like windows does? What's cool is that the driver is in the kernel already! You just have to turn it on. Append the string, "apm=on", to your kernel parameters at startup. Best, -=greg
Re: Handling with BASH variables
On Tuesday, September 25, 2001 1:17 PM, [EMAIL PROTECTED] wrote
> BTW: what is the difference between $KDEDIR and
> ${KDEDIR}? When is wich used?
They refer to the same data. The braces form is to
seprate the variable from the context. For example:
rename $fname $fnamebackup # is ambiguous,
rename $fname ${fname}backup # is not.
-g
Re: restarting a daemon
On Tuesday, September 25, 2001 11:48 AM, [EMAIL PROTECTED] wrote: > I am unable to determine how to HUP a daemon Your doc is telling you that you can send the process a signal--in this case a hangup or HUP--to ask it to reload its configuration. The signal-sending program in unix is 'kill'. So the command: kill -HUP will send the proper signal to the process with the given number. There are several ways to find the process number. A simple one that usually works is to search the output of the ps command for the process name. In your case: ps -Al | egrep 'exim' will return ps report lines that contain the string 'exim'. Substitute the number in the 4th column--the PID field--for in the kill command above (sometimes the "egrep 'exim'" command itself will be returned in the report. Don't use that line.). If your daemon can be down for a few moments, you can simply execute the command: /etc/init.d/exim restart Or, if that doesn't work (sometimes doesn't, don't know why), issue two commands: /etc/init.d/exim stop /etc/init.d/exim start Hope this helps. Best, -=greg
Re: Handling with BASH variables
On Tuesday, September 25, 2001 11:24 AM, [EMAIL PROTECTED]
>[...] there is a file "conf_kdelibs"
> the file contais that line:
> ./confugure --prefix=/opt/kde --enable-final ...
> Now i want to set for the prefix /opt/kde the $KDEDIR variable.
> This "conf_kdelibs" file is called by my compile script with the command "
> 'cat conf_kdelibs' ". Now bash is not taking the $KDEDIR varibale contents
it
> takes the string "$KDEDIR". How can i bring BASH to take KDEDIR as a
variable?
I think what you are saying is you want to put
a line: ./configure --prefix=${KDEDIR} --enable-final ...
in your conf_kdelibs file and then have the main script
execute that (and other ) lines in conf_kdelines, substituting
the current value of $KDEDIR.
If that is the case, don't use "cat" for that purpose. Instead,
source it within the context of your compile script. For
example:
...
KDEDIR='/opt/kde'
...
. conf_kdelibs # <--note the period--that's
# the source operator
...
Hope this helps. If you are trying to do something
completely different, ignore this. :)
Take care,
-=greg
Re: Mail Server
On Tuesday, September 25, 2001 5:33 AM, Wyatt Rowe - O.S.N. wrote: > [...] please advise on a good package / solution to monitor > and track a Mail / Internet server that runs on Debian 2.1. What do you mean by "monitor and track?" Do you want an application or system monitor to assist with technical operations? Do you want to read users' mail or apply some heuristic to messages? Something else? Please clarify. Best, -=greg
Re: exim and aliases
On Friday, September 21, 2001 10:17 AM, [EMAIL PROTECTED] wrote: >[...] I offered to give everyone on the box aliases through > /etc/aliases but I get an error in testing. > > In aliases - > > localemail: [EMAIL PROTECTED] > > When I send an email to [EMAIL PROTECTED] exim freezes the > message with the following error: > > Message 15kDqZ-0002Hs-00 has been frozen. The sender is <[EMAIL PROTECTED]>. > > The following address(es) have yet to be delivered: > [EMAIL PROTECTED]: lowest numbered MX record points to local host > > I think that bind is configured correctly, but I'm not sure that the > 'loverangers.org' domain has filtered out to the net at large. Could > that be affecting it. I help administer a freebsd box with sendmail, > and this exact kind of aliasing works great. I just want to have handy > aliases for my friends and not really tweak the box. What can I change > to get this to work? You used obfuscated addresses in the problem desc but gave specific error messages: 1) do you mean that loveranger.org mail is handled by your host but addresses are aliased to addresses in other domains? i.e. aliases file contains: chaka: 2) if so, have you told exim that your host is, indeed, re- sponsible for loverangers.org mail? 3) if not, then your host should not be listed in DNS as the lowest-numbered mail exchanger 4) just what, exactly, is a loveranger? :) Best, -=greg
Re: loading mods from boot [was kernel compile step 1]
On Friday, September 21, 2001 7:25 AM, [EMAIL PROTECTED] wrote: > ok. Just to make things easy I only upgraded from 2.2.29pre-17-compact > to 2.2.19. Everything was painless except that on boot up the drivers > for my 3com nic cards don't automatically load as they did with > ...pre17-compact. I sort of assume they load from a file in > /etc/init.d/ Searching google I get the idea that kerneld is the script > that loads them but looking at the script I don't really see > how it would load them from the compact image and not the 2.2.19 image. You might try running modconf to setup the module configuration. You should be able to find your NIC there. -=greg
Re: disk partitioning problems
On Friday, September 21, 2001 6:56 AM, [EMAIL PROTECTED] wrote: > [...] I have tried to partition my new 40GB drive using fdisk in DOS, > [...] > With fdisk I can create a primary DOS partition, and a number of further > extended and logical partitions, that, when formatted in DOS produce disk > drives labelled D:, E:, F:, G: etc. If you use fdisk, only create one primary partition. Do not create any additional partitions from fdisk. It only understands things Microsofty. Leave the rest of the space free. > Using cfdisk, only 2GB of the new disk seems to be available, and I can > create a primary DOS partition quite easily, or a Linux partition, with > extended and logical partitions; however, I don't seem to be able to gain > access to the rest of the diskspace. If your hardware is fairly recent, you should be able to enable large disk support from your bios setup. You should enable logical block support at least, but check your mainboard manual or mfgr website. You may want to post the details of your mainboard and HDD if you can't get it working. Logical partition is a MS-only term AFAIK. You will be creating primary and extended partitions for Linux. The extended partition concept is just a way to work around the limitations of the original PC partition table format. Primary vs. extended doesn't make any real difference to Linux. Just know that if your total number of partitions is greater than 4, you cannot have more than 3 primary partitions (the 4th is used to create the extended partitions). Also, depending on your HW, you might want to create a very small partition for /boot as your first primary partition--just in case your system cannot boot a second partition in the presence of a 10Gb first partition. Take care, -=greg
Re: .config in kernel source?
On Thursday, September 20, 2001 3:10 PM, [EMAIL PROTECTED] wrote: >also sprach John Patton (on Thu, 20 Sep 2001 02:14:26PM -0500): >> The config file does not come with the package (although perhaps it >> should). It is generated by make config... but of course you will >> have to go through everything the first time through and configure >> things the way you like. >so self-made kernel-source packages do not contain the kernel-source >of your kernel, but the generic kernel source? I always wondered that. It may be buried somewhere in the source dirs but I couldn't find it. It was installed in /boot with the -image package. -g
Re: disk partitioning problems
Sent: Thursday, September 20, 2001 2:48 PM, Duncan Smith wrote: > As a new Linux enthusiast, I am trying to configure > and install Debian Linux on my PC, and I am having > problems partitioning my new hard drive for this. I > would be reallly grateful for some advice. The new > disk is approx 40 GB, and I have tried lots of things. > What I don't understand from all the literature I have > been reading is how to partition the drive. I thought > I should make a DOS primary partition of say 40% > of the total disk space for use with Windows 98. > Should I then create another DOS partition in which > to install Linux, or have the Linux installation process > create a new partition in the remaining space? Right > now I get an error message saying that the primary > partition table is corrupted, when I try to partition >the disk in the Linux installation process. I have no > data on the new drive, and am a little stuck! Hi Duncan- What has worked for me is to use the Debian installer to partition the drive. If you want your MS partition to reside as the first one, create /dev/hda1 to the size you want using cfdisk from the installer. Then create all your Linux partitions according to your needs (see the various HOWTO's to help with the partitioning scheme). Finally, delete the partition you reserved for MS or change it to the correct type for the OS ( I usually delete it so the MS installer can pick its own type). Another tip: if you don't mind jumping out of the Debian installer, quit it after writing the partition table, then install your MS OS. That way, you won't have to boot from a floppy in order to get Debian to start after installing Win98 (you will have to modify the lilo.conf to get MS to start :) ). If the error message is happening on boot, don't worry, it will be taken care of once you write your first partition table. I hope this helps and that some more experienced folks here can give you even better advice. -=greg
Re: nimda probes
On: Thursday, September 20, 2001 2:09 PM, [EMAIL PROTECTED] > > > the worm wouldn't even know the difference, to it it looks like it would > > > hit microsofts site from your url if it tries those extentions. > > Not correct, it gets a Redirect as the response, and it's its > > responsibility to follow it, unless it's using a toolkit that does so > > automatically. > > > > Code Red, for instance, wouldn't follow redirects. > try calling default.ida from my server -- Here is the request: GET /default.ida HTTP/1.0 Here is what your server returns: HTTP/1.1 302 Found Date: Thu, 20 Sep 2001 22:18:42 GMT Server: Apache/1.3.9 (Unix) Debian/GNU Location: http://www.gnubies.com/mess.html Connection: close Content-Type: text/html; charset=iso-8859-1 302 Found Found The document has moved http://www.gnubies.com/mess.html";>here. The Location: header signals the user agent that the resource is at a different location (redirect). The user agent is usually a browser that knows how to do this (the HTML code is there in case it does not). Code Red ignores Location:. Don't know if nimba does or not. -g
Re: .config in kernel source?
On Thursday, September 20, 2001 10:42 AM, [EMAIL PROTECTED] wrote: > I have downloaded the kernel-source-2.4.9 and I couldn't find the debain > .config file for this version of the kernel. Is debian giving the configure > options used in the kernel-compilation ? If the kernel-source is a Debian package with a corresponding kernel-image, the kernel-image package installs a config- file in /boot . -=greg
Re: nimda probes
On Wednesday, September 19, 2001 11:55 PM, [EMAIL PROTECTED] > Nicholas Petreley had this suggestion for redirecting > nimda probes using Apache: > RedirectMatch ^.*\.(exe|dll).* http://support.microsoft.com Heh. I wonder if nimda actually responds to redirects. -=greg
Re: watch out for email "PRINT Summary Report"
On Wednesday, September 19, 2001 11:58 AM, [EMAIL PROTECTED] wrote: > nasty auto-running email -- look out, those of you who read using > microsoft email readers: i strongly recommend that if you get a > message entitled "PRINT Summary Report" from SOMEONE YOU DO NOT > KNOW, delete it instead of viewing it. Don't read it from anyone you DO know, either. :) This mail is also offered to you, via HTTP, from nimba-infected IIS boxes. -=greg
Re: Dual Serial port
On Tuesday, September 18, 2001 2:33 PM, [EMAIL PROTECTED] wrote: > On Tue, Sep 18, 2001 at 11:25:45AM -0700, Lazar Fleysher wrote: > > As far as I understand a regular 25 pin serial port has 2 serial ports > > built-in on one chip. > Uhm, i dont know what kind of computer you're using, but last i checked > DB25 is exactly like DB9 except it re-arranges the pins and carries more > redundant info. Right. The 25-pin EIA-232-D includes lines not normally used by PCs--mainly dealing with synchronous communication. All the signals needed for async can fit into 9 pins ( or 7 pins or 6 pins or 4 pins... :) ). -g
Re: Arentcha glad?
On Tuesday, September 18, 2001 12:44 PM, Greg Wiley wrote: > This new W32.nimda thing hits my box with 9 seperate > URLs for each attempt. Whoops, no, it's 16 per attempt. Will MS ever have to answer for this waste? -g
Arentcha glad?
Aren't you glad you use Debian? This new W32.nimda thing hits my box with 9 seperate URLs for each attempt. -=greg
Re: SUID
On Tuesday, September 18, 2001 10:27 AM, [EMAIL PROTECTED] wrote: > I can't figure out what change has to be made...I tried > RTFM, but didn't see anything that seemed relevant Yeah, I'm not sure why,but neither 'man chmod' nor 'info chmod' answer that question. For suid, you set the (user) sticky bit. In Linux: chmod u+s filename -=greg
Re: decompress tgz
On Thursday, September 13, 2001 10:34 AM, [EMAIL PROTECTED] wrote: > What is tgz and how do you decompress it ? short for tar.gz. just gunzip it and it will become tar -=greg
Re: little script for log watching
On Wednesday, September 12, 2001 4:25 PM, [EMAIL PROTECTED] wrote > Mike Egglestone <[EMAIL PROTECTED]> writes: > > I would like to write a script that would > > output to the screen the latest contents of a log, > > and continue to do so until I abort the script. > No need to write it yourself -- > $ apt-cache show console-log Will 'tail -f' do the job? -=greg
Re: APCUPSD doesn't shutdown machine
I don't use apcupsd but in order to get the machine to respond to poweroff, I must append "apm=on" to the kernel params on startup. The kernel turns off power management by default even though it is compiled in. -=greg - Original Message - From: "Dean A. Roman" <[EMAIL PROTECTED]> To: "Debian User List" ; <[EMAIL PROTECTED]> Sent: Thursday, September 13, 2001 8:36 AM Subject: APCUPSD doesn't shutdown machine > Hello all, > >I am having a problem with apcupsd. It won't shut the machine down. > > I can see in the syslog that it is writing an entry that says shutting down > machine, but it never happens. > The config is default from setup(standalone) except I changed MINUTES=10, > BATTERYLEVEL=70 so I could test the UPS without having to wait too long before > it shutdown the machine. > > apcupsd version=3.8.1.5-1 > Debian version=woody > kernel=2.2.19 > > Any ideas on what I'm missing here? > Has anyone else seen this problem? > > > Thanks, >---Dean Roman. > > > > > == > SYSLOG BEFORE, DURING, AFTER POWER LOSS > == > Sep 12 20:23:01 srfs1 /USR/SBIN/CRON[16425]: (mail) CMD ( if [ -x > /usr/sbin/exim -a -f /etc/exim/exim.conf ]; then /usr/sbin/exim - > q ; fi) > Sep 12 20:25:01 srfs1 /USR/SBIN/CRON[16473]: (root) CMD (/usr/sbin/ddns.cron.pl) > > Sep 12 20:25:01 srfs1 /USR/SBIN/CRON[16474]: (root) CMD (/usr/bin/ARKPER > 1>/dev/null 2>/dev/null # Auto added (NL_MARKER)) > Sep 12 20:26:48 srfs1 apcupsd[16417]: Power failure. > Sep 12 20:26:51 srfs1 apcupsd[16417]: Running on UPS batteries. > Sep 12 20:30:02 srfs1 /USR/SBIN/CRON[16550]: (root) CMD (/usr/sbin/ddns.cron.pl) > > Sep 12 20:30:02 srfs1 /USR/SBIN/CRON[16551]: (root) CMD (/usr/bin/ARKPER > 1>/dev/null 2>/dev/null # Auto added (NL_MARKER)) > Sep 12 20:30:42 srfs1 gnome-name-server[16563]: server_is_alive: > cnx[IDL:GNOME/Terminal/TerminalFactory:1.0] = 0x8055c88 > Sep 12 20:30:42 srfs1 gnome-name-server[16564]: server_is_alive: > cnx[IDL:GNOME/Terminal/TerminalFactory:1.0] = 0x8055c88 > Sep 12 20:33:33 srfs1 apcupsd[16417]: Battery charge below low limit. > Sep 12 20:33:33 srfs1 apcupsd[16417]: Initiating system shutdown! > Sep 12 20:33:33 srfs1 apcupsd[16417]: User logins prohibited > Sep 12 20:35:01 srfs1 /USR/SBIN/CRON[16668]: (root) CMD (/usr/sbin/ddns.cron.pl) > > Sep 12 20:35:01 srfs1 /USR/SBIN/CRON[16669]: (root) CMD (/usr/bin/ARKPER > 1>/dev/null 2>/dev/null # Auto added (NL_MARKER)) > Sep 12 20:35:11 srfs1 gnome-name-server[16675]: server_is_alive: > cnx[IDL:GNOME/Terminal/TerminalFactory:1.0] = 0x8055c88 > Sep 12 20:35:11 srfs1 gnome-name-server[16676]: server_is_alive: > cnx[IDL:GNOME/Terminal/TerminalFactory:1.0] = 0x8055c88 > > > == > # apcaccess status ## BEFORE > == > DATE : Wed Sep 12 20:35:56 PDT 2001 > HOSTNAME : srfs1 > RELEASE : 3.8.1-5 > CFGNAME : Default > CABLE: Custom Cable Smart > MODEL: Smart-UPS 700 RM > UPSMODE : Stand Alone > STARTTIME: Wed Sep 12 20:21:37 PDT 2001 > UPSNAME : SU700LAB > STATUS : ONBATT > LINEV: 000.0 Volts > LOADPCT : 48.1 Percent Load Capacity > BCHARGE : 063.0 Percent > TIMELEFT : 14.0 Minutes > MBATTCHG : 70 Percent > MINTIMEL : 10 Minutes > MAXTIME : 0 Seconds > MAXLINEV : 000.0 Volts > MINLINEV : 000.0 Volts > OUTPUTV : 119.0 Volts > SENSE: High > DWAKE: 000 Seconds > DSHUTD : 060 Seconds > DLOWBATT : 02 Minutes > LOTRANS : 106.0 Volts > HITRANS : 127.0 Volts > RETPCT : 000.0 Percent > ITEMP: 24.3 C Internal > ALARMDEL : 30 seconds > BATTV: 23.6 Volts > LINEFREQ : 60.0 Hz > LASTXFER : Line voltage notch or spike > NUMXFERS : 1 > XONBATT : Wed Sep 12 20:26:48 PDT 2001 > TONBATT : 551 seconds > CUMONBATT: 551 seconds > XOFFBATT : N/A > LASTSTEST: N/A > SELFTEST : NO > STESTI : 336 > STATFLAG : 0x10 Status Flag > DIPSW: N/A > REG1 : 0x00 Register 1 > REG2 : 0x00 Register 2 > REG3 : 0x00 Register 3 > MANDATE : 12/18/00 > SERIALNO : XS0131001479 > BATTDATE : 12/18/00 > NOMOUTV : 120 > NOMBATTV : 24.0 > HUMIDITY : N/A > AMBTEMP : N/A > EXTBATTS : 0 > BADBATTS : N/A > FIRMWARE : 152.3.D > APCMODEL : GWD > END APC : Wed Sep 12 20:35:59 PDT 2001 > > == > > == > # apcaccess status ##DURING (AFTER MINUTES EXPIRED) > == > DATE : Wed Sep 12 20:41:27 PDT 2001 > HOSTNAME : srfs1 > RELEASE : 3.8.1-5 > CFGNAME : Default > CABLE: Custom Cable Smart > MODEL: Smart-UPS 700 RM > UPSMODE : Stand Alone > STARTTIME: Wed Sep 12 20:21:37 PDT 2001 > UPSNAME : SU700LAB > STATUS : ONBATT > LINEV: 000.0 Volts > LOADPCT : 47.4 Percent Load Capacity > BCHARGE : 043.0 Percent > TIMELEFT : 9.0 Minutes > MBATTCHG : 70 Percent > MINTIMEL : 10 Minutes > MAXTIME : 0 Seconds > MAXLINEV : 000.0 Volts > MINLINE
Upgrade report
Hi all- Just some notes on a recent upgrade from stable to testing on a box with the Ivan Moore II KDE packages. Since the packages are not a part of the official Potato version, I was unsure of how the upgrade would proceed. --- The machine is a Dell Inspiron 5000 notebook with the 1400x1040 display. As expected, apt-get dist-upgrade has to be run several times. All the KDE packages are held back the first time. After a clean run through dist-upgrade, I tried to launch kdm. This failed because kdm was gone. Okay, install kdm. This also installs kdebase although I thought it was already there. But still no luck starting the kdm . Turns out XFree86 is still in version 3 and my weird configuration ( for the Dell display ) has been replaced with a default one. For some reason, offhandedly, I run dist-upgrade again. XFree86 version 4 starts coming down. I choose the new XFree86 server but decline to setup since my resolution is not supported by the package install scripts. I run XFree86 -configure and get a new configuration file. All video and LCD setup is magically generated for my weird hardware (that is really cool). But upon testing the server, mouse is frozen. After some playing, I discover that I have to remove gpm (bummer, it used to work with X no problem, I just used /dev/gpmdata for my mouse device) AND I have to specify the PS/2 protocol (-configure set it up for a serial device). I resolve to figure out the gpm problem someday. OK, X server works. I replace X config file, make sure Xserver and /etc/X11/X are pointing to the right place (Q: are both necessary? Why do we list the server in Xserver AND link X to it?) and start kdm. Bingo. Observation: KDE is still version 2.1 (I thought it would be 2.2), but everything seems much faster. Observation: The vertical bars on the panel that seperate the applets are now vertical rows of dots. I don't think that's right. The vertical bars are still there on application tool panels. -
How to handle "unofficial package" upgrade
Good day all- On a Debian Potato, I am using KDE 2.1 packages that are, obviously, not part of Potato but are from a fairly common source. Since KDE 2.2 is slated for inclusion in the upcoming Debian release, what is the best way to prepare for the upgrade? I cannot assume that the new packages will be aware of the old and will upgrade them automatically ( will they?). So, am I best off finding every trace of the non-Debian KDE and eradicating or will things just sort of work out if I leave it all alone? Any suggestions? Thanks, -=greg
Re: dual boot problem
On Friday, August 31, 2001 9:00 AM, [EMAIL PROTECTED] wrote > > > I don't know if the windows 2000 boot loader is different than windows 9x, > > > but when I used to use windows 98, I had in /etc/lilo.conf: > > > other=/dev/hda1 > > >label=win > > > > Works just fine for w2k. > > > I am not sure though cos yday I removed Win2k , put back Win ME and only > then did the dual boot work. Owise i used to get that > "inlcuding Win9" > "Fatal: no image installed" error. Havent still figured out the problem. > Giri I saw in another message that the choice of NTFS or FAT-32 might make a difference. Mine is NTFS and it works with the simple config above but I can't speak to the FAT-32 case. -g
Re: dual boot problem
On Thursday, August 30, 2001 4:40 PM, [EMAIL PROTECTED] wrote: > > On Thu, Aug 30, 2001 at 05:56:39PM +, Giri X wrote: > > | I recently installed Debian on a ayatem that had Win2k on it b4. I made 2 > > | [...] > > | thing). There was no Windows image. So now i can boot only into Debian but > > | now windows. > > | [...] > > | what is the problem and the solution ?? > I don't know if the windows 2000 boot loader is different than windows 9x, > but when I used to use windows 98, I had in /etc/lilo.conf: > other=/dev/hda1 >label=win > > (remember to rerun lilo after editing /etc/lilo.conf) Works just fine for w2k. -g
Re: Choosing a Debian Variant
On Friday, August 24, 2001 1:39 PM, [EMAIL PROTECTED] wrote > I can more easily revert back to my old kernel without a .deb, because > the old kernel is still there, ready to be used. Whenever I build a > kernel, I keep the old one around for a while > > Lastly, and most importantly, the resulting .deb did not modify > lilo.conf and re-run lilo at install time, which is, for me, the single > most dangerous and easy-to- forget thing about installing a new kernel. > Without automating that, kernel-package was, in my view, useless. Now, > granted, not everyone uses lilo. For the kernel packages I have installed via apt, the installation updates the symlink from /vmlinuz to reference the new kernel image in /boot. It modifies the /vmlinuz.old symlink to reference the former kernel. The LILO config doesn't require update since the labels for /vmlinuz and /vmlinuz.old are Linux and LinuxOLD respectively so the boot-with-former-kernel feature comes for free. Regards, -=greg
Where are config files for pre-compiles kernels
Hi all- I have the feeling I am missing something obvious: If I install one of the pre-compiled kernel packages from stable, is a copy of the configuration file used for its build stored somewhere on my system? Thanks, -=greg
Re: XDM
On Tuesday, August 21, 2001 10:37 AM, [EMAIL PROTECTED] wrote: > I'd like to revert back to logging into a command > prompt and starting X from there. apt-get remove xdm Best, -=greg
Re: how do i extract a bullet from my foot (tar woes)
On Monday, August 20, 2001 2:26 PM, [EMAIL PROTECTED] wrote:
> silly me filled up my current directory with a file called
"--remove-files".
> my question is: how the heck to i get rid of this beast
>
> i've tried
Here is a C program that will do it:
#include
int main() {
char *fname = "--remove-files";
unlink( fname );
return 0;
}
Hope it helps. -=greg
various for pre-compiled kernels
1) I used dselect to request a new kernel version from stable. It has some obvious configuration differences from the former version. Is the kernel configuration file used for the build also stored somewhere on my machine? I want to compile a slightly modified version but maintain Potato compatibility. 2) Are the modules listed in /etc/modules distinguished somehow between different kernel images? It doesn't look like it but I wonder if there is some mechanism for specifying modules to load based on the kernel image booted. Thanks, -=greg
Re: XF86Config for potato/dell inspiron
>From the "Answer your own quesition department": I got this kind of working. Here are the basics for those in the same boat. XServer: XF86_Mach64 XF86Config: [...] Section "Monitor" Identifier"LCD" VendorName"Samsung" ModelName "LTN150P1-L01" HorizSync 30-100 VertRefresh 50-100 ModeLine "640x480" 36.00 640 696 752 832 480 481 484 509 -hsync -vsync Modeline "1400x1050" 155.00 1400 1464 1784 1912 1050 1052 1064 1090 -hsync -vsync EndSection Section "Device" Identifier"Mobility" VendorName"ATI" BoardName "Rage Mobility 8MB" EndSection Section "Screen" Driver"Accel" Device"Mobility" Monitor "LCD" SubSection "Display" Depth24 Modes"1400x1050" "640x480" Viewport 0 0 EndSubSection EndSection I have a little flickering row of dots in the lower left-hand corner of the display but have been told that it will go way once I enable frame- buffer support. To anyone with LCD expertise: Does this config look reasonable? I don't know if LCD's are susceptible to the same overdriving problems as CRT's. Thanks, -=greg - Original Message - From: "Greg Wiley" <[EMAIL PROTECTED]> To: Sent: Sunday, August 19, 2001 12:09 AM Subject: XF86Config for potato/dell inspiron > Does anyone have a working X config file > for the Dell Inspiron 5000 with the > 1400x1050 panel? I can't get this bugger > to work in anything other than 640x480. > > The Rage Mobility is correctly identified > by the SVGA server and the panel and its > resolution is also detected. However, > I cannot seem to create a valid 1400x1050 > mode (XF86Setup doesn't even offer> the resolution as an option). > > I found a document describing how to > install Potato on the 5000 but the author > didn't use Debian for X--opting, instead, > to immediately upgrade to XF86 v4. > > I would rather stick to a pure Potato if > possible. > > TIA, > > -=greg
Re: X Windows Problems
On Sunday, August 19, 2001 2:55 AM, [EMAIL PROTECTED] wrote: > My ATI Radeon 32DDR video card is not on the > list, and I'm not sure what kind of default to use for this. While I have not tried X with the Radeon, I have found that the current svga server works like magic with ATI hardware. Get the svga server with: apt-get install xserver-svga and get the standard vga server (for the setup program) with: apt-get install xserver-vga16 Then type 'XF86Setup', if it asks you about using the current config, answer "no". Under the cards section, don't use the card list but go to the "Detailed" setup and make sure that the "SVGA" server button is depressed. Under the Monitor section, enter the horizontal and vertical ranges (low-high) as found in your monitor spec (you DO keep those, right? :) ). If you don't have the spec, choose one of the canned configs (just make sure your monitor can do what you indicate it can do). Under the Modeselection section, choose the color depth and resolutions you want. Press "Done" and see if it works. If all is well, save the config. You can re-run the program to setup the mouse and keyboard stuff if you want. > Also, I believe I am configuring xdm (or something like it) to manage my > windows. I'm not sure if it has something to do with this or not, but > everytime I boot up (even with an emergency backup floppy), the X-Windows > tries to start up and the screen goes black again. Until you get X running, remove the display manger with: apt-get remove xdm and just use 'startx' for testing. I hope this helps. Take care, -=greg
XF86Config for potato/dell inspiron
Does anyone have a working X config file for the Dell Inspiron 5000 with the 1400x1050 panel? I can't get this bugger to work in anything other than 640x480. The Rage Mobility is correctly identified by the SVGA server and the panel and its resolution is also detected. However, I cannot seem to create a valid 1400x1050 mode (XF86Setup doesn't even offer the resolution as an option). I found a document describing how to install Potato on the 5000 but the author didn't use Debian for X--opting, instead, to immediately upgrade to XF86 v4. I would rather stick to a pure Potato if possible. TIA, -=greg
Re: how to use BTS for install disks
On Saturday, August 18, 2001 12:28 AM, [EMAIL PROTECTED] wrote: >> I would like to report a bug on the install disks >> for the "testing" [...] > BTW, which bug did you discover? Any problems with > PCMCIA hardware installation? Yep. The running log on one of the consoles claims that some files or a directory are missing. I'll run it again to get the exact text before I report. But it sounds like the problem has already been reported, no? -=greg
how to use BTS for install disks
I would like to report a bug on the install disks for the "testing" distribution. My inclination is to file against "installation" but I am not sure. How does one report a possible defect in the installation disks? Thanks, -=greg
Re: package to get exchange mail?
On Monday, August 06, 2001 9:18 AM, [EMAIL PROTECTED] wrote: > well, Exchange supports POP3 and IMAP -- if those are enabled and > configured on your server, you could use something like fetchmail for > that. Right, unfortunately, those services are disabled. I was hoping that whatever protocol links Outlook to Exhange could be used by a fetchmail-like process. Thanks, -=greg
package to get exchange mail?
Good day! Does anyone know of a package that will let me fetch email from MS Exchange? I have just been notified that I will have to use a central Exchange server for a certain class of email. I would like to have a Debian mail server grab that mail and deliver it to my normal mailbox. Any idee-ers? TIA, -=greg
Re: OT:Perl %= operator
On Thursday, August 02, 2001 5:28 PM, [EMAIL PROTECTED] pondered: > [...] I'm just not sure what 2 % 7 equals. for m,n integers with m>=0, n > 0 , m < n: ( m modulo n ) = m so, yes, 2 % 7 is 2 . Best, -=greg
Re: How to give non-root user the right to start X
On Wednesday, July 18, 2001 1:07 PM, [EMAIL PROTECTED] wrote: > I have Debian 1.2. root can start X and other users can't. > How to give non-root user the right to start X? I don't think anything prevents a normal user from running 'startx'. Just to test, I just now shut down kdm on a Debian box, logged in to a nothing-special user account, and typed 'startx'. Other than the fact that widow-maker showed up, it started just fine. How do your normal users start X and how does the system respond? -=greg
Re: Screwed up cursor under X
On Wednesday, July 18, 2001 8:21 AM, [EMAIL PROTECTED] wrote: > I'm running an ATi AGP Rage (pro?) card with a proview monitor, and X 3.3.6. > Under X my mouse point, rather than being a normally arrow, is now a white > long and thin rectangular box with a couple of vertical transparent lines > passing through it, and it happens under all window managers. It moves around > fine, but causes me difficulties when trying to do precision work! Can you boot any other OSs to determine if it only happens under Linux? I have run across this problem in the past (not with a Rage) and it turned out to be broken hardware (or firmware, I guess, anyway, a replacement card fixed it). -g
Re: .bashrc
On Tuesday, July 17, 2001 1:04 PM, [EMAIL PROTECTED] wrote Subject: Re: .bashrc > Hi, > > On Mon, Jul 16, 2001 at 08:00:16AM -0700 or thereabouts, Greg Wiley wrote: > > The problem with ~/.xsession for kdm is that > > you lose the session selection capability of > > the login app. I ended up creating a new > > mechanism that sources a local user init file > > before running the main Xsession. > > For me this is wrong. I have an ~/.xsession file and am using kdm. As > long as I use the "default" session my .xsession is executed, otherwise > the window-manager i specified. I remember the kdm manpage explaining > this quite nicely. You are right, Karsten, but I am referring to getting the best of both. I wanted a local initialization mechanism, like ~/.xsession, independent of the choice of session type. If I use ~/.xsession, those settings don't take effect in "non-default" sessions. -=greg
Re: .bashrc
On Friday, July 13, 2001 1:54 PM, [EMAIL PROTECTED] wrote: > If you use xdm, the trick is [...] > Since programs in your x session are usually descendants of ~/.xsession, > they will automatically be run with the right environment settings. Right, just make sure your bash-specific com- mands (like alias defs) are in .bashrc so they will be run at term start. > Since I am not a gdm/kdm user, I cannot tell how it works there, but the > idea shoudl be the same, I guess. The problem with ~/.xsession for kdm is that you lose the session selection capability of the login app. I ended up creating a new mechanism that sources a local user init file before running the main Xsession. There might be a better way that doesn't require modifying /etc/X11/kdm/Xsession and I'd be happy to hear it. -=greg
Re: .bashrc
On Thursday, July 12, 2001 8:51 AM, [EMAIL PROTECTED] spoke: > On Thu, Jul 12, 2001 at 10:20:52AM -0500, Larry W. Irwin Sr. wrote: > > [...] having a problem with .bashrc. It works fine as > > root but does not get executed when I log in as a > > user. > maybe you have a .bash_profile file that is taken in place of .bashrc. Also, in case you are in X and launching a term, .bashrc is automatically loaded (instead of .bash_profile, .profile, .login) when bash is not a login shell. -=greg
[OT] Why attached text messages?
Why do certain peoples' posts to this list show up as attached text files in my mail client (OE)? I get a blank message with two attachments: .txt and .dat. The text file contains the actual message. -=greg
Re: Exim & include files
On Tuesday, July 03, 2001 1:14 AM, [EMAIL PROTECTED] wroted: > I'm trying to get exim to [...] > do something along the lines of : > > if $h_from: contains "" > then > freeze text "Hello big boy" > endif A couple of ideers: If the file doesn't change much (and isn't large), maybe write an m4-or-something script that builds a filter file with the target values hardcoded--possibly a series of ifs or a single long delimited target string for one contains test. The pipe command might be a useful mechanism. Send all candidate messages (hopefully, there is a way to roughly qualify messages for this process) to a pipe that checks against your file. The pipe could add a header to the message that indicates the result of the test, then resubmit the message to exim. Neither of these sounds all that great so I hope you receive a more intelligent response from someone else. -=greg
