Re: X quality really bad on 4mb video card on debian
On Sat, 12 Jul 2003, Paul wrote: > and 196 mb of ram. The speed of X is fine but the quality is horrible. > Colours get stuffed up often, pixel get messed up when I move windows. I > am using the amp driver for a allience promotion AT25 with xfree86 4.* . > >From my research on the net I found out that this is the right driver > for the AT25. Hi Paul, I have a Alliance Promotion AT3D 4Meg card on Slackware which uses the same "apm" driver (I presume "amp" was a typo),the trick is to select "no_accel" since it is a hardware accelerated card which does not require software acceleration iirc. An older monitor should work with DefaultDepth on 16. below are relevant sections of my XF68Config file. hth, Keith. -cut--- Section "Device" Identifier "apcard" Driver "apm" #VideoRam4096 Option "no_accel" #Option "swcursor" Section "Screen" Identifier "Screen 1" Device "apcard" Monitor "dell770" DefaultDepth 24 Subsection "Display" Depth 8 Modes "640x480" "800x600" #"1024x768" "1280x1024" ViewPort0 0 EndSubsection Subsection "Display" Depth 16 Modes "640x480" "800x600" #"1024x768" "1280x1024" ViewPort0 0 EndSubsection Subsection "Display" Depth 24 Modes "640x480" "800x600" #"1024x768" "1280x1024" ViewPort0 0 EndSubsection EndSection ---uncut--
Re: KPPP
On Sat, 9 Aug 2003, Fred Wyatt wrote: > 1) I can use Kmail if and only if I start Kppp. Any problems with starting Kmail on it's own? > 2) I can start Kppp if and only if I type the root password. Redhat used to set kppp as root user only using pam authentication, then about RH-6.2 they also added the "consolehelper" script. A workaround is to bypass the use of "consolehelper" for /bin/kppp. As #root: 'which kppp'(should show kppp in /bin) 'ls -la /bin/kppp' (shows kppp is softlink to 'consolehelper') 'rm /bin/kppp' 'ls -la /sbin/kppp' 'chmod u+s /sbin/kppp' 'ln -s /sbin/kppp /bin/kppp' If you get "Device Busy" messages when starting kppp that means your "/dev/modem" softlink maybe pointing to the wrong modem device. In Kppp setup, select the correct device (/dev/ttyS1,S2,/dev/LTO etc) or fix the /dev/modem softlink. > 3) KDE Help Center (kppp Handbook) says it is a "bad thing" to dial in as root. The Kppp book can be confusing for Redhat user's because the recommendation of setting up a "modem" group does not work around the Redhat "consolehelper" script iirc. By dialing in as root I think the help document is referring to being logged in as root user and dialing up,browsing etc, not to being an ordinary user who just uses the root "password" to start kppp via consolehelper. cheers, Keith.
Re: gcc issue
On Sat, 9 Aug 2003, Nick Rout wrote: > http://linmodems.technion.ac.il/resources.html#GCC3 > > On Sat, 09 Aug 2003 > 19:55:11 +1200 Rik Tindall <[EMAIL PROTECTED]> wrote: > > On driver compilation, it fails, saying that my gcc3 working on the > > available gcc2 files downloaded is "known not to work". I'm running > > RH9 on an Athlon XP1800. Are you using a old ltmodem.tar.gz file (8.22a5), download the latest one from http://www.linmodems.org (at least 8.26a9 a few months ago) The install script will set up the /dev/modem link to /dev/LT0 iirc not /dev/ttyS15, but I think the script does also offer making a softlink from /dev/tty14---/dev/LT0 Make sure the booted kernel (2.4.20-8 whatever) is the same as the kernel source /usr/src so that you can build the modules in /lib/modules/2.4.20-8/ltmodem cya, Keith.
Re: Your Solution
On Mon, 11 Aug 2003, Fred Wyatt wrote: >lrwxrwxrwx 1 root root /usr/bin/kppp --> /usr/sbin/kppp > > As a point of interest, does the leading letter "ell" in the permissions > string indicate a linked file? yes, section 15.1 of the Rute manual by Paul Sheer has info. http://rute.sourceforge.net
Re: Firewall woes with KDE 3.1.0...
On Sun, Sep 14, 2003 at 10:38:20AM +1200, Chris Wilkinson wrote: > Hi there, > > I installed an RPM of Firestarter last nite, since I'd heard it was > better than Guarddog (my previous iptables config app) for managing > rules etc...for a firewall... > > KDE now will not boot if iptables is running, and it hangs at the point > where it trys to intialise system services. looks like the iptables script (/etc/sysconfig/firewall?) is blocking the X server on port 6000. Check the Firestarter Readme in /usr/doc or the manual to see if there is any mention of tcp 6000. A medium or low security firewall may also block the default system ports below 1024 but leave 6000 open. > I logged in as root in a console, removed firestarter, then rebooted > with no joy. I next removed iptables, which allows KDE to start, but > trying to download guarddog (which works fine, and is MUCH easier to use > than firestarter IMHO) fails, because dependencies require iptables > to install first (and then the internet download of guarddog is refused > by iptables)...Grr Are you doing a rpm download/install here? Would a download using 'lynx' or 'wget' then starting the firewall and installing guarddog with 'rpm -Uvh' work. cheers, Keith.
Re: DVDs under Linux
On Wed, Nov 12, 2003 at 02:42:45PM +1300, CF wrote: > Yes - I can read the udf filesystem, and the disks work in a regular dvd > player. gui programmes such as xine,mplayer usually use the soft link /dev/dvd to read the disk so check the link exists. 'ln -s /dev/hdd /dev/dvd' also have you installed the libdvdnav files from http:www.xine.org it is a plugin which enables ogle,xine,mplayer to use dvd menu and navigation features using the libdvdread library. hth, keith.
Re: modem help
On Thu, Dec 04, 2003 at 01:24:01PM +1300, Volker Kuhlmann wrote: > > Hi, im a newbie at Linux and have just installed mandrake 9.1, it > > seems that the o.s.detects the modem but does not have drivers for any > > modem with a conexant chipset. Can anyone out there help with this > > problem, the modem is a 56k Puretek 3515 with conexant chipset. > > Bad luck. There are only commercial drivers for connexant chip modems > available, US$15. You can try at reduced speed (33k) before you buy. Connexant hcf/hsf modem drivers are available from- http://www.linuxant.com/drivers The free drivers there run at only 14k and the 56k drivers cost US14.95. This is the site that www.linmodems.org points to for generic conexant drivers. Dick Smiths sell a 56k pci ham modem for $34 which is sometimes reduced to sale price. Its very fast and comes with linux drivers on the cdrom. Product XH1154. > For $30-35, you can get a new modem card with lucent chip which has > costless, but no opensource drivers. Installing the drivers will be a > little hassle as you'll need to recompile/link them for your kernel. I agree, buying another card is worth it since the cost can be recovered by selling the conexant chipset modem. It's not worth supporting a product that linux users have to pay twice for when other 56k modems are sold cheaper with drivers included. > There are Intel 536EP chip modems too for the same price, driver > installation and configuration of Linux etc is out of the box and fully > automatic on SuSE 8.2, but I could not yet get the modem itself to talk > to any ISP's dialup, probably because I haven't yet found the correct > init strings for that modem. You will have this kind of problem > potentially with any modem. The 536EP driver is also costless, but not > opensource, if it's not out of the box for your, crank your compiler. have you tried X3 in the init string, this forces dialing without waiting for a dialtone first. hth, keith.
Re: Help! undelete for ext2/3??
On Mon, Jan 12, 2004 at 04:22:25PM +1300, Nick Rout wrote: > dogdamn who would use ext3 huh? > in ext2 you delete a file and the inode info is kept intact. you can revert the ext3 back to ext2, see line 240 of man tune2fs. - umount the /dev/hdb1 partition. - 'tune2fs -O ^has_journal /dev/hdb1' (the caret negates the feature) - e2fsck -y /dev/hdb1 - remount hdb1 and change to the root level of that partition. - 'remove -f .journal' (delete the jornal file in hdb1) use mc's command-undelete ext2 tool or the 'debugfs /dev/hdb1' and it's #help and #lsdel commands to do the same thing manually. also "undeletion howto" hth, keith.
Re: Next Workshop.
On Sun, Feb 01, 2004 at 04:59:49PM +1300, Benjamin Devine wrote: > I have tried installing an old version > of Slackware with no avail. I think that is because I have no idea on > kernal modules but the Original isa scsi controllers(2) have the adaptec > AIC6360Q Chipset. Hi Ben, check out the /etc/rc.d/rc.modules file on the slackware you have installed. This script should have a scsi controller section which you uncomment lines to load the various modules you require. There might be entries for aic6XXX adaptec cards. Also during the install did you select adaptec.s as the kernel image file. Using the default bare.i will not include scsi drivers. Apparently Slackware-9.1 supports aic7xxx cards only, there is no mention of aic6xxx in /usr/src/linux/drivers/scsi/README. The newsgroup alt.os.linux.slackware is helpful although it is usually faster just to use http://www.google.co.nz > I'm going to download Slackware 9.1 on monday. I would like to try slackware > in a fileserver enviroment unless anyone else has any better suggestions. Slack-9.1 Disk 1 contains everthing you need for a fileserver or X with the smaller window managers such as windowmaker. KDE,Gnome and extras are on Disk 2. hth, keith.
man colour in console.
picked this up from a newsgroup to enable man pages to display bold text in console. >You may want to try "most" as your pager. "man -P /usr/bin/most man" to see; >and set your color preferences in $HOME/.mostrc an alias such as the following can be put in /etc/profile or .bashrc. alias manc='man -P /usr/bin/most $1' --- keith.
Re: was: An Internet Dialer now kpp & network
On Sun, Jun 20, 2004 at 01:16:36PM +1200, Kerry Mayes wrote: > Tried some of the others like wvdial, but it needs to have > root access to work. Ordinary users can use wvdial if pppd is suid root. #which pppd #chmod u+s /usr/sbin/pppd(if that is your path) #ls -l /usr/sbin/pppd (check for rwsr-xr-x permissions) > However, it seems to be interfering with the network. Internet works if > I have the network connection down, but not if it is up. For packets to go out the modem put "defaultroute" in the ppp options file. (probably /etc/ppp/options) Kppp can automatically put the "defaultroute" option in /etc/ppp/options every time you use kppp. In kppp setup select your account then "edit account => gateway => => assign default route to this gateway" -- keith.
Re: Kde troubles
On Sat, Jul 10, at 11:13:32PM +1200, Ross Drummond wrote: > .. Home partition only 4% full. > I have been floundering around with xinit and startkde. Nothing seems to > work. have you installed kdelibs. --- keith.
Re: Kde troubles
On Sun, Jul 11, at 11:23:20AM +1200, Ross Drummond wrote: > Used to work. I installed some gimp stuff which had heaps of > dependencies and 'poof' kde doesnt work anymore. Ross, check if the gimp deps have been replacing the original kdelibs. knotify.so and kded.so which crash both call libkdeui.so.4 which is in it. use 'ldd /usr/bin/' to track down the problem. an eg 'ldd /opt/kde/bin/startkde' if involved then could try backup of ~/kde and uninstall/reinstall kdelibs again. gimp might still work with the original kdelibs. or if the new kdelibs are required for gimp, then a matching version of kdebase to kdelibs may be needed for startkde. looking at whether the (new?) gimp package is actually needed may also be worth consideration if your distro is relatively recent and the gimp package is not required for your printer/xsane. (--no-deps manual install works in some cases also.) -- keith.
Re: OpenCD
On Thu, Jul 22, at 04:44:08PM +1200, Carl Klitscher wrote: > > Oooh you're giving me far too much credit... it is actually developed and > delivered from http://www.theopencd.org and is about 290Mbs worth. The "gnuwin2" cdrom contains the same programmes and more including apache, mysql,php,c++ compilers,perl,vim,xemacs,nmap,xchat and litestep,a afterstep clone for the win32 desktop. programme list- http://gnuwin.epfl.ch/apps/en/bestlist.html The iso can also be downloaded from this site. Note that mozilla-1.0-win32.exe on this disk may require the security update that was announced last month on http://www.mozilla.org If anyone would like to pick up a copy ($2) email me offlist. --- keith.
Re: Screens, colours, and text sizes.
On Wed, Aug 11, at 12:08:25PM +1200, Andy George wrote: > Further to the Ansi colours/size discussion, the solution you present below, > will change the console, in a similar vein to Slackware, yes? slackware has the "setconsolefont" programme (a part of 'pkgtool') to change the screen font. edit /etc/DIR_COLORS file for setting of the 'ls' command color output. an alias for ls in your ~/.bashrc file as follows will ensure that colour is for terminals only. see 'man ls' line 380 for info. --snip-- alias ls='ls --color=auto' alias ll='ls -la --color=auto' to colourize the strings output by your init scripts (/etc/rc.d) see 'man console_codes' line 200. example for cyan string output. --snip-- echo -en "\\033[0;36m" " stringtexthere "\\033[m" the pager "most" is ideal for colorizing man pages. you can set the "most" pager as your environment variable in /etc/profile, (other distros /etc/bashrc) export PAGER="/usr/bin/most" 'grep' can also be colorized with an alias. see man grep line 60. alias grep='grep -i --color=auto' > I'm looking at the console output of my IPTables Firewall, and honestly, > if I can pretty that up some, it'd be almost readable. if you use 'elvis' a ~/elvisrc config file as follows might help. --snip-- color bold yellow color italic brown color error red color fixed green color link green color cursor cyan color comment green color string cyan color char red color keyword yellow set ruler set showmode Most newbie linuxers would prefer coloured mc (midnight commander) to vi. I have a page for mc colours here. http://homepages.e3.net.nz/~keithmg note that cat can be used to append (>>) your preferred mc colours file to ~/.mc/mc.ini. 'lynx' colours are set by the /usr/lib/lynx.cfg (line 2750) and lynx.lss file. lynx.lss tends to have priority over lynx.cfg. in general (imho) cyan and green on black bg are the easiest colours to use as they are clear colours that do not strain the eye. hth, keith.
Re: GNUWin and the OpenCD
On Thu, Aug 19, 2004 at 04:20:18PM +1200, Wesley Parish wrote: > Any chance of someone with a decently speedy connection downloading > and burning them for me? > http://theopencd.sunsite.dk/ > http://gnuwin.epfl.ch/en/index.html Hi Wes, email me if you want me to burn a copy. I have both although gnuwin is the better of the two imo. cheers, keith.
Re: Mandrake 10 2.6.3-7 modem install
On Mon, Aug 09, 2004 at 12:40:10AM +1200, Robert Nairn wrote: > I am new to Linux and have got a copy of Mandrake10 from Dick Smith > I have tried three different modems and god only knows how many different > drivers in my computer and cannot get any to work properly. > The original modem in this computer...was the only driver that compiled. Hi Rob, if you bought a internal 56k pci ham modem from Dick Smiths it may not be your fault that you could not compile it. This modem (product XH1154) has the linux logo on the box and includes linux drivers on cdrom. Drivers supplied as "Intel-536ep-4.51" for the XH1154 have a bug which prevents the 536ep modules being compiled. I bought this modem last week and the build number was 4.51, but if dse are selling newer XH1154's they will probably include updated drivers without this bug. The easiest fix is to download the latest drivers "Intel-536ep-4.68" from the dse website http://www.dse.co.nz The buggy file in "Intel-536ep-4.51" is "clmdrvr.c" and the error message is: Error message: lines 481,482 of clmdrvr.c file. clmdrvr.c:481:22: missing terminating " character clmdrvr.c:482:37: missing terminating " character Details and Patch file follow. Product: XH1154_PCI 56k modem with Intel ham 536ep chipset. Driver Packages: distro and generic drivers of build release 4.51 Buggy Source File: /Intel-536ep-451/serialdrv/clmdrvr.c Symptom: Compile using the 'make 536ep' command fails with the following error message. Error message:lines 481,482 of clmdrvr.c file. clmdrvr.c:481:22: missing terminating " character clmdrvr.c:482:37: missing terminating " character patchfile: --- cut 481,482c481,482 < printk(KERN_ERR"ham:rs_open: DSP did not reset. try again or restart computer\n < count now %d\n",MOD_IN_USE); --- > printk(KERN_ERR"ham:rs_open: DSP did not reset. try again or restart > computer\n" > "count now %d\n",MOD_IN_USE); -- cut --- Fix: copy the patch to the /Intel-536ep-451/serialdrv/ directory. (parent directory name will be different as per distro) change to the serialdrv directory containing the patch and use the command: 'patch clmdrvr.c patchfile' change to the parent directory and compile as per instructions, make clean, make 536ep, make install. hth, keith.
Re: setting up dialup
On Mon, Oct 11, Volker Kuhlmann wrote: > If it's Intel 536, the driver is missing from > SuSE 9.1 because it keeps locking up on the 2.6 kernel. new 536ep drivers for Suse-9.1 2.6.X kernel are available since 20.7.04 from: http://developer.intel.com/design/modems/support/drivers.htm select the modem type (536ep) and suse version (Suse) to get to the download page. cheers, keith.
Re: Quadriple booting
On Fri, Nov 05, Julian Visch wrote: > I have been searching google to find if there is a way of triple booting > 2 Linux hard drives and 2 windows hard drives. > other = /dev/hdg1 > map-drive=0x80 > to=0x83 > map-drive=0x83 > to=0x80 > label = windows2 > table = /dev/hdg Possibly secondary slave hdd1 (0x83) rather than hdg1 which is 0x86. --- keith.
Re: Quadriple booting
On Sat, Nov 06, 2004 at 06:17:54AM -0800, Julian Visch wrote: > > > other = /dev/hdg1 > > > map-drive=0x80 > > > to=0x83 > > > map-drive=0x83 > > > to=0x80 > > > label = windows2 > > > table = /dev/hdg > > How does the numbering system work? > > 0=a, 1=b, 2=c,3=d,4=e,5=e,6=f? Does it ignore cd drives? cdroms are connected identical to ide disk drives and most atx motherboards use only hda->hdd. Try running your lilo config prog again to find out how the hdg got into lilo.conf. also 'man lilo.conf' mentions that since lilo version 22.5 the per image option map-drive= has been replaced by boot-as=. -- keith.
Re: Anyone Compaq conversant???
On Sat, Nov 06, Steve Holdoway wrote: > 128 processes: 127 sleeping, 1 running, 0 zombie, 0 stopped > CPU states: cpuusernice systemirq softirq iowaitidle >total 41.8%0.0%3.6% 0.0% 0.0% 114.2% 40.0% >cpu00 20.1%0.0%1.5% 0.0% 0.0% 53.3% 24.9% >cpu01 21.7%0.0%2.1% 0.0% 0.0% 60.9% 15.1% > Mem: 1668088k av, 1650880k used, 17208k free, 0k shrd, 43216k buff > 931796k actv, 496804k in_d, 24260k in_c > Swap: 2097096k av, 0k used, 2097096k free 1368716k > cached > > PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME CPU COMMAND > 1870 oracle16 0 50588 48M 47360 S30.8 2.9 0:09 0 oracle > 1864 oracle15 0 55052 52M 51820 S 5.7 3.2 0:15 1 oracle > As you can see, a huge amount of time seems to be spent in iowait. Steve, 114% for the iowait is cumulative so perhaps it is not a totally accurate figure compared to the two individual cpu stats. Also the procps-3.2.1 TODO doc mentions: Supply better data for top's CPU state display. Currently top has to subtract old numbers from new numbers and divide that result by the number of processors. The kernel won't even supply the number of processors in a portable way. btw how do you display iowait? No mention of it in man top. hth, keith.
top + pinfo colours
I was surprised to find that 'top' can now display screen output in colour. Seems to have had added features since 2.x.x I don't recall the top man mentioning a couple of years ago including display of multiple windows. key mappings are- z - toggle colour Z - config colour screen W - write ~/.toprc config file h - show help "pinfo" is a lynx like browser that displays info and man files in colour. General movement is a lot faster due to arrow keys and options can be set to set your favourite prog to run when http,ftp,mail links are selected. The config file is ~/.pinforc and pinfo-0.6.8.tar.gz (312kb) is built running configure,make,make install. download from- http://dione.ids.pl/~pborys/software/pinfo/ sourceforge- http://sourceforge.net/projects/pinfo (no links at sourceforge to the download page) cheers, keith.
Re: File associations for Firefox in KDE
On Thu, Jan 13, at 03:37:28PM, Tom Munro Glass wrote: > . But if I use Firefox as the browser, it > usually pops up a dialog asking if I want to save the file or open it with > some application that I have to browse for. > How do I make Firefox use/inherit the standard file associations? Hi Tom, Firefox may have an entry field in the dialog box for "open file with (application)" and also a button to "always show this dialog". (Uncheck the button) The Yolinux site has a helper apps configuration guide including a wrapper for automating the opening of octet-stream mail attachments. http://www.yolinux.com/TUTORIALS/LinuxTutorialMozillaConfiguration.html#PLUGINS The following mozilla configuration for pdf files should apply to Firefox. go to Edit --> Preferences --> Navigator --> Helper Applications and add the "new type" application. pdf files require mime type application/pdf, extension pdf and open with /usr/X11R6/bin/xpdf. The dialog box that pops up when hitting a file link will mention the mime type of the file you are wanting to read so use that definition for the MIME type entry in Helper Applications. The helper config file is ~/.mozilla/default/ad49qtsw.slt/mimeTypes.rdf One of its entries is- NC:alwaysAsk - "true" if the user should always be prompted before handling data of this type, false otherwise. /usr/lib/mozilla/defaults/profile/mimeTypes.rdf contains comments explaining the syntax of this file. -keith.
Re: device sr0
On Thu, Jan 20, at 07:55:40PM, Gary Durey wrote: > In the /dev directory are files sr0, sr1 etc to sr15. > How can I increase the number of these to sr42 Gary, the /dev/MAKEDEV script only creates sr0-sr15 so you have to make /dev/sr16 - sr42 manually using the 'mknod' command. see man mknod for details. as root in /dev- 'mknod -m 660 sr16 b 11 16' 'mknod -m 660 sr17 b 11 17' .. .. 'mknod -m 660 sr43 b 11 43' hth, keith.
Re: make install help - kxstitch
On Sun, Jan 23, at 10:19:51PM, Barry wrote: > I recently downloaded source code for kxstitch and worked my way through > configure, make and make install. > 1st Problem is that the executable is in /usr/local/kde/bin instead of > /usr/bin. You can uninstall kxstitch with- 'make uninstall' and remove remaining source dir binaries- 'make clean' check configure options- './configure --help' probably './configure --prefix=/usr' (or for some distros kde path is) './configure --prefix=/opt/kde' > It starts but crashes on pressing button3 check './configure --help' for options to disable any support libs that you think may not be installed. > 2nd problem is that the help button can not find the documentation > which is in /usr/local/kde/doc/HTML/en/kxstitch Not sure about Mandrake, I know a few distros install kde in /opt/kde therefore you would use './configure --prefix=/opt/kde'. -- keith.
Re: make install help - kxstitch
On Sun, Jan 23, 2005 at 10:53:31PM, Nick Rout wrote: > the cure for the program not being where you want it is the mv command. I have found that some programmes will not run if moved from their installed location, the use of a soft link was a quick fix. 'ln -s /usr/kde/bin/kxstitch /usr/bin/kxstitch' hth, keith.
Re: slingshot dialup not authenticating
On Wed, Jan 26, at 08:12:14PM, [EMAIL PROTECTED] wrote: > Modem dials, does handshake, sends password/username but slingshot doesn't > reply. I'm using PAP authentication which I'm pretty sure is correct. > pap-secrets exists with correct information in it. Hi, this works for me using pci lucent winmodem. /etc/ppp/options lock defaultroute noipdefault modem /dev/modem 115200 crtscts #noauth passive asyncmap 0 name "" #identical to pap-secrets . /etc/ppp/pppscript TIMEOUT 60 ABORT ERROR ABORT BUSY ABORT "NO CARRIER" ABORT "NO DIALTONE" "" "AT&FH0" OK "atdt087XX" TIMEOUT 75 CONNECT > I've googled for the problem and answers point to modem init strings. Often it is easier to leave init strings to the default AT&FH0 setting and check out permission/authentication issues with pppd and pap first as well as the options file. also have a look for backup files of the previous paradise configuration in /etc/ppp, can't remember if pppconfig creates them or not. cheers, keith.
Re: More progress
"/usr/bin/smbmnt" is the /usr/bin/smbmount helper application. See #man smbmnt #which smbmnt #ls -la /usr/bin/smbmnt smbmnt is probably set as: rwxr-xr-x root root Use: #chmod ug+s /usr/bin/smbmnt(to set the suid bits rwsr-s--r-x) In /etc/fstab add the "suid" option //Sydog/C/winshare/Sydog_Csmbfs noauto,_netdev,user,suid 0 0 Since you are trying to mount SMB shares using the mouse on the desktop icons they might be using the /bin/mount command. In that case set /bin/mount as "r-sr-sr-x root root" (with chmod,chown) The "user,suid" options also apply to the /dev/fd0 and /dev/cdrom entries in /etc/fstab to allow ordinary users to mount/umount those devices if /bin/mount is set r-sr-sr-x root root. hth, keith Fisher Family wrote: > > Tried this but no go. I first tried exactly as you suggested > then tried username=myusername,password=mypassword, but neither worked. > > I got Reported error: > > smbmnt must be installed suid root for direct user mounts > (500,500) smbmnt failed:1 >
Re: New Modem Question(2)!@#$%
Check the kppp settings for flow control. (Setup-->Device-->Flow Control) "CRTSCTS" usually works but try the others, xon-xoff,none. Add the "debug" option to /etc/ppp/options and check /var/log/ppp.log for clues as to what happens after kppp starts connection. It should send an ATZ command to the modem and recieve OK before going on to the authentication handshake with the isp. hth, Keith. Bowen Denning wrote: > but now when I try to connect it gets as far as saying "initializing modem" >but never seams to get any further.
Re: New Modem Question(2)!@#$%
I made a mistake there. It should be the Setup-->Device-->Line Termination CR-LF-CR/LF settings that you adjust. The kppp handbook FAQ (chapter 7) has info on troubleshooting. __ Keith > > Check the kppp settings for flow control. (Setup-->Device-->Flow Control) > "CRTSCTS" usually works but try the others, xon-xoff,none.
Re: Modem Problems Again :(
Hi, Does the OH light flash on the dynalink modem.That means that kppp has initialised the modem and has entered the dialing phase. You'll just have to keep working on the Device Setup untill it is correct for the Dynalink. Refer to #man pppd for info on the details of ppp and pap configuration. /etc/ppp/options requires the "user" option to point to /etc/ppp/pap-secrets. ie 'user name' will be your login name to xtra.(Edit the file manually) "rp3" is the gui frount for the console dialup programme called "wvdial" See #man wvdial, #man wvdialconf. Just run "wvdialconf" (in console) and then edit the file /etc/wvdial.conf to enter your username,password and telephone number into the [Dialer] section. After that it should run when you use the console command #"wvdial xtra" (if you gave it the dialer name "xtra") and since rp3 is the gui frount, it should work also. regards Keith. Bowen Denning wrote: > > Well im sure your probably going to be sick to death of these questions i am > asking but my modem is no better off. > > It seams that i can dial out fine using minicom but not kppp. So why aren't > i using minicom? well i would love to but it seams that all other > applications are oblivious to the fact that i am connected to the internet. > so i can connect but thats it. > > > Check the kppp settings for flow control. (Setup-->Device-->Flow Control) > > "CRTSCTS" usually works but try the others, xon-xoff,none. > > I have tweaked around with all the settings but none seam to get me any > further past the "initializing modem..." message. another thing i did try > was to change some of the modem command settings (as pointed out in the kppp > manual FAQ) to the same as minicoms settings (no idea what i was doing) but > this also seams to get me no further. > > So i am still in the situation that (1)redhats dialer won't even detect my > modem (2)kppp will queried the modem but get as far a "initializing > modem..." and go no further (3)and minicom will dial and connect but no > other programs will be able to use the connection. (this is probably coz i > dont know how to use minicom) > > well once again any suggestions would be good. > -- > == Bowen Denning (linux neophyte)
Re: Samba help
Rob Fisher wrote: > I am having no success with Samba. Well actually I have had progress but > not enough to use it yet. > My Win2000 and WinXP PC's can see the Samba Server but are denied access >or not accessable. WinNT bringing up "denied access" and "not accessable" boxes may be a NetBIOS "name resolution" issue not a encryped passwd problem. This involves setting the Linux box as the WINS server. As Chris mentioned if you have set encrypted password in /etc/samba/smb.conf and have the same user/passwd on both machines it should work ok. This "access" error occurs when you try to get to the Linux samba box by clicking on the Network Neighbourhood folders in NT Explorer. If you can reach the Linux server using its ip address then you know it is a NetBIOS issue. In the NT start menu -"run command" box, as ordinary user enter the: \\linux-host-ip-address\sharename ie probably \\192.168.1.1\homes (if that is grunters address) Section 7 of the Samba Howto does mention this but does not provide a solution to fix the "name resolution" problem of accessing the the "Local Master Browser" machine which is your Linux box. The Howto just says forget using the host name, just type in the ip address. I found that going to the NT "Control Panel" "Network"-"Identity" and selecting "change" to get the protocols-tcp/ip "properties" window I was able to set the Linux box as the Primary WINS Server by entering the Linux ip address. After that being WinNT, you have to reboot. Additionally you will have to set /etc/samba/smb.conf entries for being the WINS machine. Probably: WINS support = yes domain master = yes preferred master = yes Then run the console command "samba restart" (as root) to initialise the new smb.conf settings. >From NT you can then use the linux hostname in the "run command" box ie \\grunter\homes Also in NT Explorer use "tools"-"find computer". Grunter should appear in the search result and if you double click on that the shares should open in a window like Network Neighbourhood. So far I havn't figured out if network neighbourhood does work in the left window pane of the NT Explorer window. The Samba "DIAGNOSIS.txt" file has step by step info on troubleshooting. /usr/doc/samba/docs/textdocs/DIAGNOSIS.txt In Linux to test the user account try: # smbclient //localhost/homes (then login user-password) or # smbclient //localhost/homes -U username%passwd ? brings up the help menu after login. For the NT dos command window, to see shares. # net view hth, keith.
Re: Samba help
Thanks for the info. I'm new to NT and didn't know how it handles dns. Since its a home network all the Linux box has is /etc/hosts for name resolution. regards, Keith. Will Pilvio wrote: > > If \\linux-host-ip-address\sharename works, just edit the Win-machines > HOSTS file to handle the DNS resolution without WINS. > Should work, shouldn't it?
Re: Displaying part of an X desktop
There is a .deb package called "Xtv" which grabs a screenshot of a remote display and updates it periodically. Maintained by Jay Kominek and might be on http://www.rpm.find.net Not exactly what you want but similar. regards, Keith. Vik Olliver wrote: > > Does anyone know of a utility that can display a selected area of an X > desktop on a remote machine in real time?
Re: kde failure
Michael, startx reads the /home/user~/.xinitrc file for the default windowmanager. It will use kde if you put a kde entry in. Comment out the gnome-session line. This is a basic example /home/user/.xinitrc ---cut-- #!/bin/sh #exec /usr/bin/gnome-session #exec /usr/bin/wmaker exec /usr/bin/startkde ---cut-- If you want a complete xinitrc file with resource settings check your /var/X11R6/bin/xinit directory. I don't think that a reinstall of RH will make any difference. Afaik other distros have the same startx problem. (eg slackware 7.1) Probably a user permission problem with Xserver. Gnome should run ok from kdm or gdm in runlevel 5. cheers, Keith. Michael wrote: > > I run Redhat 7.2. Whenever I login (run level 3) and startx, gnome appears > to load up but I end up with a blank blue screen without any icons etc - > absolutely nothing. I can crash back into level 3 without a problem using > the keyboard.
Re: corrupt partion table :(
Hi, you can run 'cfdisk' and 'gpart' from the RH7.2 partition. 'gpart' tries to read the sectors and locate the start/finish points. Both tools should be on in RH7.2 disks or from http://www.rpmfind.net Toms rescue disk also has cfdisk. http://www.toms.net/rb There is a mini howto call "Partition Rescue mini HOWTO" by Jean-Daniel Dodin which deals with fixing corrupt pt's. http://www.linuxdocs.org Don't alter the Windows or Redhat partition sizes. Just try to get the Mandrake start sector first. Sometimes it will boot with the finish sector written short of the actual finish mark but the start sector ok. If you are wanting to repair the Mandrake partition size proper it may take more rewrite attempts to get the correct start-finish points. cheers, Keith. Chris Bayley wrote: > > DiskDraK (Mandrake 8.1) hung during an install of M8.1 and has left me with > a corrupt partition table, > PartionMagic6 Spews.what else ??
mc colour
Nick Rout wrote: > I would do mc, but i'm sure there are people who know more about it > than me. :-) Not really. But I really enjoy using it. One thing I have noticed is that just about every distro I have used alway sets mc as default colour blue and white. I just wonder how many users do configure mc with their own colour style. The ~/.mc/ini file for each user can be edited with the addition of a [Colours] section to the bottom of the file. Info on mc colour is in #man mc (line 2400) An example is the following with the format being- =,: --cut-- [Colors] color_terminals=linux,xterm,gnome-terminal base_color=normal=cyan,black:directory=yellow,black:selected=blue,cyan: marked=brightcyan,brown:menu=blue,cyan:link=green,black: executable=brightmagenta,black:core=red,black: editnormal=cyan,black:editbold=yellow,black;editmarked=magenta,black; --cut-- It has to be put in as three lines only. [Colors] color_terminals...this defines the terminals base_color=normalthe colour options line is continuous without gaps. I had to use vi to edit ~/.mc/ini. For some reason using mc or kedit do not save permanently and the text is overwritten by a default ini file Also for Redhatters you can change the screen output colour of the boot up init scripts to cyan by editing the "normal" line in the "/etc/sysconfig/init" file. SETCOLOR_NORMAL="echo -en \\033[0;39m" --- change the number 39 (white) to 36 (cyan) --- SETCOLOR_NORMAL="echo -en \\033[0;36m" #"manconsole_codes" has info on the ESC colour codes.(line 200) cheers, Keith.
Re: Kde panel lost, how does one get back?
Hi, kpanel has had a name change to "kicker" since kde2 so that would be what is put into the commandline box. Its config file is ~/.kde/share/config/kickerrc My own experience with kde3 is that after a new install it is usually sluggish untill it has settled down after a few logins. It is usually the cause of a few crashes untill then. Also I have found kde3's menu editor buggy after choosing the option to "hide" a submenu in the kicker menu. The problem is that the submenu's cannot be bought back by selecting "unhide". The "hidden" submenu's are put in the ~/.kde/share/applnk directory and can be returned by deleting it's entry in that directory, followed by a mouse right click on the kde desktop to choose the "Desktop Refresh" option. cheers, Keith. On Tue, 13 Aug 2002, Yuri de Groot wrote: > Press [Alt] + [F2] to bring up a single commandline box. > Type "kpanel" without the quotes and press [Enter] > > It happens to me sometimes - a glitch in kde perhaps.
Re: Acronyms
On Mon, 26 Aug 2002, Michael Beattie wrote: > ## Colorize smileys: :-) ;-) :-/ :-( > color body yellow black "[ \t]>?[;:][-=]?[PpD)/(|}{><]" Hi Mike, I had a few problems getting the -=[]<> characters from your above expression to work in my ~/.muttrc file. The only way I could get yellow smileys was by splitting the expression up into three lines. Also added forward slashes to the first line so characters =<> were taken literally. ___ ### Colorize smileys: :-) ;-) :-/ :-( ### color body yellow black "[ \t]>?[;:][-=]?[PpD)/(|}{><]" ___ color body yellow black "[t\>\<\=;:?PpD)(|}{><]" color body yellow black "[][]" color body yellow black "-" I'm using mutt 1.4i. Does your expression work ok on your mutt mua. thanks, Keith.
Re: Acronyms
Thanks Mike, I don't always make use of reg expression's and was uncertain about the muttrc syntax. sincerely, Keith. On Mon, 26 Aug 2002, Michael Beattie wrote: > I'm not sure you understand the regular expression. > > > ### color body yellow black "[ \t]>?[;:][-=]?[PpD)/(|}{><]" > above is split into 5 parts:
Re: Ye Olde Meeting
Hi, Carl certainly can speed things along in the cli. Interesting evening. Somebody asked about how the manual pages appeared in colour. Carl said this was a Redhat feature. (Most new distro's use XFree86 4.2.0) Top was also in colour but information on this is sparse and I havn't found out how top can do this. The terminal window was an Xterm which has its system config file in /usr/X11R6/lib/X11/app-defaults (in Redhat) although other distro's such as Slackware use /etc/X11/app-defaults. The Xterm config file is /etc/X11/app-defaults/XTerm. There is also another file called /etc/X11/app-defaults/XTerm-color but I found that it didn't work, probably because of the #include XTerm header. It is virtually a copy of the colour section of Xterm anyway so I just stuck to editing the Xterm file. All lines starting with a bang are comments. The colour section down the bottom is called "Enable Color by default" and contains several attribute entries that need to uncommented for bold "BDMode" and underlined "ULMode" modes to become switched on. (During the presentation it was the bold and underlined man text that was appearing in colour. Redhat has them on by default.) *VT100*colorBDMode: on (these switch BD and UL on) *VT100*colorULMode: on *VT100*colorUL: cyan (these define UL, BD colours) *VT100*colorBD: yellow The easiest method is to make a user ~/.Xdefaults file in your home directory and leave the /etc/X11/app-defaults/XTerm as is. (unless you prefer editing the sys file) Colours in .Xdefaults are (literally) pretty neat. You can fine tune the colour tone for you liking by adding 1,2,3 or 4. If you run console progs like mc in xterm you may have problems with .Xdefaults having priority over mc custom colours. The *VT100*color0 - *VT100*color15 entries to have to be changed to fix it. *VT100*color0 is the background colour, but I added another two entries called *VT100*Background: and *VT100*Background: because they work better with xterm. The .Xdefaults file below is slightly changed from the sys XTerm with underline and bold colours being cyan and yellow. If you already have a ~/.Xdefaults file you will probably just need append your colour preferences. color3 was changed from "yellow" to "brown4" because I liked brown4 in mc. *VT100*Background: black *VT100*Foreground: green were added because they cause xterm to run in black and green right from the start, not just for applications. (this is identical to using the command 'xterm -bg black -fg green') If anyone can find out how Redhat gets "top" to appear in colour I'd be glad to know. cheers, Keith. ~/.Xdefaults -- cut --- *VT100*Background: black *VT100*Foreground: green ! Enable Colour by default. *VT100*colorMode: on *VT100*boldColors: on *VT100*dynamicColors: on ! Uncomment this to use color for underline attribute *VT100*colorULMode: on !*VT100*underLine: off ! Uncomment this to use color for the bold attribute *VT100*colorBDMode: on ! Uncomment this to use the bold/underline colors in preference to other colors *VT100*colorAttrMode: on ! colours defined here *VT100*color0: black *VT100*color1: red3 *VT100*color2: green3 *VT100*color3: brown4 *VT100*color4: blue3 *VT100*color5: magenta3 *VT100*color6: cyan3 *VT100*color7: gray90 *VT100*color8: gray30 *VT100*color9: red *VT100*color10: green *VT100*color11: yellow *VT100*color12: blue *VT100*color13: magenta *VT100*color14: cyan *VT100*color15: white *VT100*colorUL: cyan *VT100*colorBD: yellow -- cut ---
Re: Ye Olde Meeting
sorry about the typo in paragraph 7, should be - *VT100*Background: and *VT100*Foreground: > Keith McGavin wrote: > *VT100*color0 is the background colour, but I added another two entries > called *VT100*Background: and *VT100*Background: because they work better ^ > with xterm.
Re: Ye Olde Meeting
Hi Stephen, lucky for you that debians bold font in "top" comes out in colour. In Slackware it doesn't and I think the binary is responsible. I have found also that 'ls' dir_colours is achievable in xterm and rxvt. This is usually not configured by default although most distros do have dir_colors in console. It seems that ~/.bashrc file in the home directory is read by xterm so it is just a matter of putting a couple of aliases in. #touch .bashrc (create the file if you don't have it already) cut - alias ls='ls --color' alias ll='ls -la --color' - cut These aliases can be put into /etc/profile or /etc/bashrc but xterm wil not display 'ls' dir_color unless ~/bashrc is pointing to those files. --cut -- if [ -f /etc/profile ] ; then . /etc/profile fi --cut -- cheers, Keith. On Sun, 1 Sep 2002, Stephen Nicholas wrote: > Top "just works" in colour now to, because the headings are bold, they > get the bold colouring from the xterm.
Re: Networking setup problem.
Hi Bill, try adding a localdomain name the hostname in /etc/hosts. A name like 'localnet.net' will do. The "redhat7.0" hostname is not required for the localhost which is for internel pinging to host 'localhost' only. eg /etc/hosts 127.0.0.1 localhost.localdomain localhost 192.168.1.1 pentium.localnet.net pentium 192.168.1.2 duron.localnet.net duron Have a look at man #ifconfig and #route. Type #route in console to check it is up and the commands are like this. #ifconfig eth0 192.168.1.1 netmask 255.255.255.0 up #route add -host 192.168.1.2 eth0 cheers, Keith. >/etc/hosts >192.168.1.2 duron duron >127.0.0.1redhat7.0 redhat7 localhost.localdomain > localhost >192.168.1.1pentiumpentium On Tue, 3 Sep 2002, Bill Evans wrote: > I am aiming to set up a SAMBA network between two computers. > They are a Duron 800 with 128 MB RAM running Windows98 (1st edition) > and a Pentium 133 with 32MB RAM running Redhat 7.0. > > As a first step I am trying to establish an ethernet connection between the > two. > Previously when the Pentium was running Windows98 I had successfully > networked the two together. > The NICs and cable have not been altered since then (apart from checking the > cable connection just now.) > > I cannot ping between the two computers. > Both computers can ping their own local host. > The Pentium, running Linux, can ping itself at the address 192.168.1.1. > (using either the address or nickname) > Perhaps significantly, the Duron, running Windows, cannot ping itself > (192.168.1.2) > > On the Duron: > Network configuration shows: > Primary Network Logon is set to Client for Microsoft Networks > File and printer sharing for MS Networks is installed > The NIC is listed as PLANET ENW-9503/04 PCI Fast Ethernet NIC > Both NetBEUI and TCP/IP are installed for this card. > TCP/IP properties: > IP Address 192.168.1.2 > Mask 255.255.255.0 > > I have edited the c:\windows\hosts file to contain the following: > 127.0.0.1 localhost > 192.168.1.1 pentium > 192.168.1.2 duron > > > On the Pentium: > I have looked around for files I thought might be relevant. > Here they are: Don't ask where I configured each of these. I have been > playing around in both the CLI and in Gnome's > Network Configuration for the last couple of days :) > > the NIC, the subject of an earlier thread, is now recognised ok >/etc/modules.conf > alias eth0 ne2-pci > alias parport_lowlevel parport_pc > >/etc/HOSTNAME > pentium > /etc/hosts.conf > order hosts,bind > > >(Initially either the localhost or pentium, I forget which, was called > wb.invalid by default, but I changed it. Given the following DOMAINNAME, > this might be significant.) > > /etc/sysconfig/network > NETWORKING=yes >HOSTNAME=pentium >DOMAINNAME=invalid >GATEWAY=192.168.1.1 >GATEWAYDEV=eth0 > >/etc/sysconfig/network-scripts/ifcfg-eth0:0 >DEVICE=eth0:0 >BOOTPROTO=none > >/etc/resolv.conf > nameserver 192.168.1.1 > >/etc/networks > this file is empty > > Network Configuration. >Gnome's Network Configuration shows settings as follows: > Hostname, nameserver, hosts, default gateway, and default gateway > device all as above. > >Interfaces: > Interface lo > IP127.0.0.1 > Proto none > atboot yes > active active > > Interface eth0 > IP192.168.1.1 > Proto static > atboot yes > active active > > Interface eth0:0 > IP > Proto none > atboot > active active > >Interface eth0:0 is a mystery to me . Does this get set automatically? > >Any help would be appreciated. >I would guess the problem lies in the Windows machine, given that it > cannot ping itself. > > Bill > > > > > >
Re: Ye Olde Meeting
Hi Tim, what could 'ls --color=auto' be useful for other than being neater for not having escape codes? The =auto output in my kdedit and gedit editors is plain black print. 'ls --color' defaults as 'ls --color=always' so in xterm generally the two options will work the same. yes, ll='ls -la' is shorter, thanks. Slackware's default install uses /etc/profile so I was putting things in that perspective. cheers, Keith. On Tue, 3 Sep 2002, Timothy Musson wrote: > I reckon: > > alias ls='ls --color=auto' > alias ll='ls -la' > > Have a look at tmp.txt: > > /usr/bin/ls --color > tmp.txt > /usr/bin/ls --color=auto > tmp.txt > > _ > > These aliases can be put into /etc/profile or /etc/bashrc but xterm > > wil not display 'ls' dir_color unless ~/bashrc is pointing to those > > files. > > > if [ -f /etc/profile ] ; then > > . /etc/profile > > fi > _ > > Usually, ~/.bash_profile (or ~/.profile) sources ~/.bashrc - e.g.: > > # ~/.bash_profile > > if [ -f "~/.bashrc" ] > then > . ~/.bashrc > fi > > > Tim > -- > Timothy Musson - [EMAIL PROTECTED] > http://homepages.ihug.co.nz/~trmusson/ > Who messed with my anti-paranoia shot? >
Re: PCI parallel & serial -- problem??
Hi Andrew, the parallel card might be like some ethernet cards that just need to have the irq option and not the io=0xnnn which is probed for. parport.txt in my 2.4.18 kernel doc says that the Base I/O addresses (for supported cards) need not be specified. insmod parport_pc irq=7 hth, Keith. On Tue, 3 Sep 2002, Andrew J Sands wrote: > However, even if I add it to my modules.conf on my custom 2.4.18 kernel when > I attempt to access the port it reports "Device is busy" yada, yada.., > If I don't add it to the modules.conf as suggested by parport.txt because > PCI devices are detected automagically then its not found and when accessing > the port the response is "No port at address."
Re: Ye Olde Meeting
Hi Tim, I see, the escape garbage from 'ls --color' in bash scripts will mess the system up very quickly. 'ls color=auto' is the viable option. There is also 'ls color=tty' although it is not in man 'ls' Sorry for not thinking before posting. 'ls' without options in bash scripts ignores the aliases and outputs normally. thanks, Keith. On Tue, 3 Sep 2002, Timothy Musson wrote: > On Mon, Sep 02, 2002 at 09:42:04PM +1200, Keith McGavin wrote: > > ~/.bashrc > > > alias ls='ls --color' > > alias ll='ls -la --color' > > I reckon: > > alias ls='ls --color=auto' > alias ll='ls -la' > > Have a look at tmp.txt: > > /usr/bin/ls --color > tmp.txt > /usr/bin/ls --color=auto > tmp.txt > > > > These aliases can be put into /etc/profile or /etc/bashrc but xterm > > wil not display 'ls' dir_color unless ~/bashrc is pointing to those > > files. > > > if [ -f /etc/profile ] ; then > > . /etc/profile > > fi > > Usually, ~/.bash_profile (or ~/.profile) sources ~/.bashrc - e.g.: > > # ~/.bash_profile > > if [ -f "~/.bashrc" ] > then > . ~/.bashrc > fi > > > Tim > -- > Timothy Musson - [EMAIL PROTECTED] > http://homepages.ihug.co.nz/~trmusson/ > Who messed with my anti-paranoia shot? >
Re: KMail & pst files & converting to LOTD.
Hi Jeremy, the enlightenment OSX type theme you are thinking of is probably the AQUA "Say hello to imac" theme that was mentioned in the "multihead X" thread that was posted last month. One prog that is also useful is the 'Esetroot' command which is in the "eterm" (rpm,deb.tgz) package. This can be used to set a desktop .jpg background in the blackbox/fluxbox/waimea window managers. Just add the command to the ~/.xinitrc file or use a bash script. #!/bin/bash /usr/bin/Esetroot -s /path/to/your/file.jpg exit appreciate it if you can post me the kde XOS icon pack links. cheers, Keith. On Wed, 18 Sep 2002, Jeremy Bertenshaw wrote: > Someone on here mentioned that they were running > enlightenment with an OSX like theme on it, I really like > OSX and have added some icon packs and themes and so forth > to kde so it looks and feels a lot like OSX, can post links > if anyone is interested :-).
Re: KMail & pst files & converting to LOTD.
On Thu, 19 Sep 2002, David A. Mann wrote: > A theme is more than just a background, you know :) Yes and some themes are very striking.I suppose many linux users have spent hours trying to figure out what their favourite theme or window manager is. A regular topic on irc.openprojects.net#linuxhelp after "I have a dependency problem" and "how do I set up a router/gateway" is "Wow look at this cool theme at cheers, Keith.
Re: Re: KMail & pst files & converting to LOTD.
That is the "External Taskbar" of Kicker. - right click the kicker - Add-->Extension-->External Taskbar. - right click the kicker again and select "preferences" - Select the Extensions tab in the settings window. - For the "External Taskbar" option set its panel location to bottom and enable hide buttons to be medium size. One noticeable feature of using the external taskbar is that it has the iconify window outline type behaviour similar to kde1.1. I have the external taskbar set to hide to conserve desktop space. The "child panel" extension is ideal at the top of the destop with a 3 second hide delay. You can click and drag programmes/submenus from the K menu to the child panel. cheers, Keith. On Fri, 20 Sep 2002, Jeremy Bertenshaw wrote: > http://www.kde-look.org/content/preview.php?file=153-1.jpg > > Does anyone know how to put the task bar above the panel > like in that picture? I've tried a few things and it ends > up being all over the show..
Re: Modem troubles and ISPs
Hi, your isp is probably using pap so you have to work that way with both progs. With minicom I do the manual login with username and passwd then I press alt+q to leave 'without reset'. Then type in 'pppd' in another console to start the protocol link. You can do this as root or ordinary user if pppd is 755 and /dev/ttySx 666. 'killall pppd' to quit. For /etc/ppp/pap-secrets have the line username * passwd Both progs require these entries in /etc/ppp/options.This may be why wvdial was failing. name#points to pap-secrets defaultroute /dev/ttySx#device hth, Keith. On Thu, 26 Sep 2002, Carl Cerecke wrote: > Nick Rout wrote: > >>In any case, the winmodem is not the problem, because the ISA modem was > >>the same result. Connects, then line dropped after at the login: > >>prompt. > >> > > > > At the risk of sounding repetitive, have you tried pap/chap? > > > > also you might like to try wvdial, it is brilliant and may sort you out. > > It is available in packages for most distros. http://open.nit.ca/wvdial/ > > wvdial has the same problem.
Re: Destop variety
Hi Vik, I agree with Chris S. that Enlightenment menus are difficult to read if the themes are like 'bluesteel'. Not a bad manager if the visuals are ok. To get things just right in fvwm2 can take several days editing fvwmrc which most ppl are just not prepared to do. Can be very good if you persist. Afterstep is one I still havn't got used to. sWM (small Window Manager with 4 workspaces) is probably the smallest (32k binary), and looks modern due to its titlebar buttons. Won't ever turn you off visually as there is just xterm. I prefer kde to gnome/sawfish because the desktop/window menus are more ordered. My favourite window manager is wmaker because of its speed, reliablity, appearance and ease to configure and work with. One hassle is wmakers default Ctrl-Alt+arrow keys to swap workspaces because I have to take my hand off the mouse to press the arrow. Changing the shortcuts to Ctrl-Tab and Ctrl-q is a easy workaround and much faster. cheers, Keith. On Sat, 21 Sep 2002, Christopher Sawtell wrote: > I have fiddled with Enlightenment and find it difficult to read. > FVWM is just too difficult to re-configure. Wonderful to stop the staff > wasting time meddling, but too difficult to change anything when you actually > need to.
Re: 2/10 meeting - t-shooting workshop?
Hi Matt, sounds like you did not install lilo/grub. Have you got a boot disk? re monitors in cars to the meeting. The best way to put a monitor in the car is upside down in the back seat with the glass facing into the seat. Usually the curvature of the seat will hold it in place but stuffing a towel underneath/on the side can give extra protection and stability. Be careful if you put it in the frount passenger seat as the monitor can rip the stiching in the seat. cheers, Keith. On Tue, 24 Sep 2002, Matt Agnew wrote: > There was talk at the last meeting of a troubleshooting session for the > next meeting - is this still on, and is there a signup sheet? I've > installed Mandrake alongside Win2k and Win98 and can't get into the > Linux partition. > > Alternatively, is someone willing to come to my house and help me in > exchange for free beer? This would be ideal as my monitor won't fit in > my car (Hillman Imp).
Re: Debian on Steroids: Libranet 2.7:
Hi Will, yes it can be worth paying for the commercial version if you want the lastest version and support from the vendor as well. Most ppl who have done this are the non-geek users who believe in buying a product. The $15 non-commercial cd that you can get from Phil at http://www.copyleft.co.nz is suitable for linuxers not concerned about support and willing to use apt-get update to upgrade. The #libranet email list which anyone can join is very newbie friendly with its subscribers having some inclination towards using multimedia like mp3,tv,mpeg and gimp/blender. I have used the $15 Libranet 1.9.1 cdrom. As a distro I would rate it 7/10. It's faults? are that it is a 150meg base.tar.gz that unzips to something like 350meg so you can't use a custom install to set up a mail/print/web server on a 100 meg drive. And if something goes wrong it can be time consuming to peruse the /etc directory to figure things out due to the numerous debian type bash scripts. This would not happen any more than redhat/mandy thou. It's console/X frountend tool for setup called "adminmenu" is excellent and includes a package tool called "capt" which is easier than "aptitude" or "dselect" (Debian users can 'apt-get install capt') Certainly this is the way to get home desktop users into a deb distro and it is far better than redhat which is concentrating on the server enterprise market now and doesn't really put tlc into the home user/desktop compared to the other distros. I don't know about Libra's 2.7 speed increase but it is not hard to be faster than RH for using KDE 3. cyou, Keith. On Thu, 10 Oct 2002, Will Pilvio wrote: > Any one with experience on this one? Is it worth the $$ vs Debian? > > From the Linux Journal Newsletter Oct. 9: > > Debian on Steroids: Libranet 2.7: > http://www.linuxjournal.com/article.php?sid=6358 -- The Libranet > distribution is promoted as offering all the security and stability of > Debian but with an easier installation process. Leon Goldstein > recently took version 2.7 out for a spin, and he was rather impressed > with the performance and speed increase. > > Cheers Will >
Re: Annoying Konqueror problems.
Hi John, the konqueror "Handbook" is in the menu under Help. re "Printer not receiving jobs" message, after you select Location-->Print are "properties","printer name" and "print system" in the print window appropriate for your setup? Selections like lp0,greyscale,generic-unix-lpd are available. Select the same as per netscape and it should be ok. cheers, keith. On Fri, 11 Oct 2002, John Williams wrote: > On Thursday 10 October 2002 09:19 pm, you wrote: > Lee, thanks for that. Who would have separated the configuration of the > window from the 'settings'? Only a geek as it is logical I suppose. > I must have deleted index.html somehow as it really is not on my system. I > must look for a Konqueror manual so I don't have to ask any more of these > questions.
Re: SMC 9000 Network card
Hi Rob, the module may be the generic smc9194 by Eric Stahlman according to the info in /usr/src/linux/drivers/net/smc9194.c By "auto detected" do you mean by the "kernel" screen output on boot or by "kudzu"? Sometimes kudzu will work if you remove the card, remove settings if asked by kudzu on next boot,then replace the card again. Not sure what the console tool is other than 'setup' although in earlier RH versions 'webmin'and 'linuxconf' could configure the network devices. You may be able to configure manually using mc, kate or pico. Probably- #/etc/rc.d/rc.local /sbin/modprobe smc9194 #/etc/modules.conf alias eth0 smc9194 options smc9194 io=10 io=0x300 ifport=0 #usually io=xx is all you need. hth, keith. On Sat, 12 Oct 2002, Robert Fisher wrote: > I have just installed RedHat8 on a Dell laptop with a docking station > with an SMC 9000 network card. > > The NIC was not auto detected. Is there a manual configuration tool for > it? (Like there is sndconfig for manual configuration of sound cards)
Re: SMC 9000 Network card
Hi Gareth, not as far as I know but I havn't set up modules.conf for quite a while. not a typo but probably better known as io=XX (Y=X) cyou, keith. On Sun, 13 Oct 2002, Gareth Williams wrote: > > would it not be better to just stick smc9194 in /etc/modules? > > > > > #/etc/modules.conf > > alias eth0 smc9194 > > options smc9194 io=10 io=0x300 ifport=0 #usually io=xx is all you need. > > (was that a typo? perhaps you meant irq=10?)
Re: SMC 9000 Network card
Great. My apologies to Gareth as I have just realised that I did make a typo by not specifying irq=10. and X syntax should have been irq=XX io=0xXXX cheers, keith. On Sun, 13 Oct 2002, Robert Fisher wrote: > Hey Keith, your advice worked a treat (only slight change I had irq=10 - > not io)
Re: install vs upgrade: rh 7.3 -> 8.0
Hi, the Redhat upgrade is ok and will leave /home and /etc alone but it does pay to backup the files in /etc as any particular setting unique to a 7.3 prog version can be altered. Easy to back up /etc as it is only 6 meg. Can't help you with grub but in lilo it is possible to boot old and new kernel images. man lilo and lilo,kernel howtos have info. Running several distro's and having each use the same /home partition is easy. Set each /etc/passwd and /etc/group to be identical in each install so that each can access the user on the /home partition no prob. example- passwd- don:x:500:500::/home/don:/bin/bash group- don:x:500: chown- 'chown -R don:don /home/don' I strongly advise that you do not store any personal or valuable stuff on an extended partition. When a hardrive has superblock problems it is a lot harder to repair extended partitions. Put swap partition last for flexibility, putting it in the middle just makes a mess of things. Assign partitions in units of hundreds so they are easy to remember. If you have an old hardrive on /dev/hdb1 you can use that for backup of /home. Example: /dev/hda1 OS/2 Warp4 800meg /dev/hda2 / 3 gig /dev/hda3 /home 2 gig /dev/hda5 slackware partition /dev/hda6 debian partition /dev/hda7 gentoo partion /dev/hda8 swap hth, keith. On Sat, 30 Nov 2002, Marcus Liddle wrote: > q) > whats the difference between an install > and an upgrade?? > > currently running rh 7.3; have rh 8.0 install set > > upgrade seems to blow away any prevous kernel versions > in the grub menu on startup > > a) I waz woundering if by doing an install, I would have > kept the prevous boot options? > > I have apps such as VMWare which prevously have been > very very picky about what kernel they play in the > sandpit with [:(] ...thus have been caught out in prevous > upgrades by being forced to upgrade all > > b) I see there's 7.3 compatablity libraries; is it > possible to run both 7.3 & 8.0 with ease? > > c) also, as a on-the-side question - how hard is it to > run a few distro's together? ie say... Mandrake / RedHat > & Debian (maybe). > > I'd imagine careful planning of partitions would be required, > but all-in-all very possible ...no? > (have large HD capacity & want to play) > > cheers > Marcus > -- > Tutor Assistant [mailto:[EMAIL PROTECTED]] > Dept. of Computer ScienceUniversity of Canterbury > Phone: +64 3 366-7001 ext: 7871 Office: 323 > [ http://www.cosc.canterbury.ac.nz/~marcus/index.html ] > >
Re: Lightwieght Linux
Hi Lance, interesting.. uClibc and busybox is about as small as it can get. I noticed that Todd Sundsted referred to the use of compile/installing lilo. Cluggers may be interested to know that lilo now has a bitmap image (16 colours) that can be viewed to make the selection on bootup. Users of the peanut distro will be familiar with this as peanut has used this in peanut-9,2. but the lilo version also provides some nice looking pics of the tux penguin. In the article it is stated that lilo is built using the command 'make'. There is also a quickinstall' script provided by the present developer John Coffman called 'QuickInst' which makes the install easy and safe. The steps to install lilo are: -download the lilo-22.3.4.tar.gz from lilo homepage http://www.brun.dyndns.org/pub/linux/lilo/ - make a floppy bootdisk first, 'mkbootdisk' in redhat I think and also in the utilities manu. - backup /etc/lilo.conf and the /boot directory. - uninstall the redhat/debian distro lilo proggie. rpm -e lilo or dpkg -r lilo, the kernel image stays in /boot but the lilo files are removed. - unzip and cd into lilo-22.3.4 , there is a README and README.bitmaps file. - ./QuickInst (runs the install script which will build and asks how many os's you want written in /etc/lilo.conf and where the lilo boot sector should be installed) If you prefer lilo written to the MBR you type /dev/hda Thats about it but the /etc/lilo.conf file usually requires some tweaking to get the tux pic and menu selections in the x,y positions of your choice. The first time I installed lilo from source it took 20 minutes to figure out how to configure the x,y tux positions in /etc/lilo.conf, but that aside it is only a couple of minutes to install/setup lilo-22.3.4 once you know how and you just use the same header section of /etc/lilo.conf on any system to bring up the appearance. Any cluggers/dunluggers lotl wanting a copy of my /etc/lilo.comf is welcome to email me. cheers, keith. On Wed, 27 Nov 2002, Lance BLACKLER wrote: > I recall that there was quite a bit of discussion sometime back about > distro's suitable for older hardware. I have just been reading an article > on IBM's site that might be of interest. > > http://www-106.ibm.com/developerworks/linux/library/l-lwl1/ > > ciao > > Lance B
re: upgrade 7.3 -> 8.0
Hi Marcus, hth means hope this helps. Obviously I did not really explain what exactly happens to the /etc directory during an upgrade. You are right that it is changed to reflect the settings of latest versions of "Myapps". Programmes such as gimp,mutt,ssh,tin store their system setting in /etc. Network and ppp settings are preserved so you don't have to setup the hostname and addresses again. Since you have plenty of disk space it is probably easier to just "install" the 7.3 and 8.0 RH distros on different partitions.I think this is what you mean when you mention installing "beside" the already installed RH partition. When making an install you have to define the partition that / will be installed to. This is known as the "target" partition. To obtain an RH8.0 install "on top" of an already installed partition (eg RH7.3 on /dev/hda2) you would have to use Disk Druid to specify /dev/hda2 as the / target for the install and select not to "format" that partition. RH8 would be written over RH7.3 and /etc settings lost. iirc Since the RH-8.0 upgrade has removed the RH 7.3 kernel it looks like you will have to copy the RH-7.3 kernel into the /boot directory (vmlinuz-2.4.16 ?) and add a stanza (section) to grub so it boots that kernel. I have no idea how the previous "move" that you did kept an old kernel other than choosing not do format the partition when asked by Disk Druid. multi distro system --- Sorry I have no experience with grub or installing the boot loader in /dev/hda1. I have always installed to /dev/hda (mbr) Your setup with /dev/hda1 partition as /boot and /dev/hda1,hda2... as / targets for each distro should work thou provided you do not format /dev/hda1 or set up grub/lilo with each install of the other distros. I have no experience with this so I recommend that you do a search on http://www.google.com for more info. Admin - "Myapp" versions such as mozilla should be the same or similar across all the distros so you can use mc to copy config files for proggies across each partition. Bookmarks, colour and window manager setups usually work ok but don't bother tring to copy anything from or to debian as debian is full of so many hook and call scripts that makes it incomparable to the other distros. It is posslibe to copy files basic /etc files such as dircolors from RH to other distros but in general the /etc directories of both RH and debian are exclusive so generally it is not possible. I admitt that Slackware and SuSE /etc directories are exclusive too however they tend to be easier to work with with very well commented scripts and fewer calls. Installing -- Yes you can define another partition to install to. Use Disk Druid to set the other partitions as the target for the install of /. During the final stage of the install (after hostname setup and before X configuration stage iirc) RH should ask you if you want to configure grub or lilo. Either of them will also ask if you want to add already installed os'es such as OS/2 or RH7.3 to the grub/lilo stanzas. If you want to use grub/lilo of a previously installed distro just select "no -do not set up grub" and after the install has completed boot back into the first distro (RH 8.0?) and run the grub configuration from there. Don't know the name of the programme but it should be in a "Utilities" or 'setup' menu somewhere. cheers, keith. On Sat, 30 Nov 2002, Marcus Liddle wrote: > (what does hth stand for?) > okay ...a little confuzzzed > /etc is left alone ...but still changes particular settings?
re: upgrade 7.3 -> 8.0
correction to typo, Your setup with /dev/hda1 partition as /boot and /dev/hda2,hda3... as / targets for each distro should work .. On Sun, 1 Dec 2002, Keith McGavin wrote: > multi distro system > --- > Your setup with /dev/hda1 partition as /boot and /dev/hda1,hda2... > as / targets for each distro should work thou provided you do
Re: help with extracting tar.gz
Hi Kevin, no idea how to extract from a remote desktop. Are you using ftp? To extract a tar just use command: # tar -xvh tarfile see man tar. -xvh is extract,verbose,hash The easy way to work with tar and tar.gz files is to use 'mc'.Just hit enter on any tar, tar.gz, rpm or tgz and mc will open it up for you. Use the F5 key to copy the tarred directory to /usr/local/src. To make a proggie check the README.txt file in the untarred directory, the default prefix is usually set /usr/local which is what most people use, ./configure --help (check options) ./configure make make install cheers, keith On Sat, 7 Dec 2002, Kevin Wilkinson wrote: > Iam working from terminal > > I am trying to extract the remote desktop tar file but can not remeber > the options extract, > can some help please > the other thing is to install the remote desktop program you have to run > the make file how do I this > > Regards, > > Kevin > [EMAIL PROTECTED] > > > >
Re: PCMCIA settings
Hi Isaac, the pcmcia boot scripts of Redhat's /etc/init.d/rc directory differs from Slackwares /etc/rc.d scripts so it is not possible to copy pcmcia boot files over as is. In Slackware the pcmcia card description files are in /etc/pcmcia, if Redhat has these files you could copy them over but I can't guarantee that they are identical. Probably the only file they both have in common is /etc/pcmcia.conf. Configuration of pcmcia on Slackware is quite easy so you may be able to set things up in Slackware. If you have trouble just post back to the list with more details about the pcmcia card. In Slackware the pcmcia "card description" files are in the /etc/pcmcia directory. You can use 'mc' in console to do most of the setup as it is fast, easy to use and ideal for Slackware. If you do not have the pcmcia packages installed yet mount the Slackware 8.0 cdrom and cd to ./a1 directory, then type 'installpkg pcmcia.tgz'. Chances are that pcmcia.tgz is already installed. By default the pcmcia "cardmgr" is set to poll your pcmcia card for its irq and io port address and read the driver settings from /etc/pcmcia/config. 'dmesg' will output pcmcia messages on the end of the file. Also use the following commands to check that 'cardctl' can see your pcmcia card. 'cardctl ident' 'cardctl config' 'cardctl status' Using 'mc' go to the /etc/pcmcia/network.opts file and press the F4 key to edit it. Edit in your network settings in the appropriate lines and press F2 to save. In my case all I had to do was, IPADDR="192.168.1.2" NETMASK="255.255.255.0" NETWORK="192.168.1.0" BROADCAST="192.168.1.255" GATEWAY="192.168.1.1" It may be easier for you to reboot the machine. When Slackware boots it's init scripts it first reads /etc/rc.d/rc.S which runs /etc/rc.d/rc.pcmcia which contains entries for the environment variables. If rc.S is not running rc.pcmcia use 'mc' to check that the pcmcia lines in rc.S are uncommented. The quick way to find these lines with 'mc' is to use the F7 search key. Just type in pcmcia. If /etc/rc.d/rc.pcmcia is being run check the environment variable entries PCIC, PCIC_OPTS. Usually they are set to poll the ethernet card. Note that rc.pcmcia reads the /etc/pcmcia/pcmcia database file. You can 'cd /etc/rc.d' and use './rc.pcmcia staart' , './rc.pcmcia stop' to start/stop it manually to reload modules. I don't think that /etc/pcmcia.conf is required to be used as by default /etc/rc.d/pcmcia should set up most of the required environment variables. All I have in my pcmcia.conf file is- PCIC=i82365#an intel chip PCMCIA="yes" Re Slacware as a distro I find it very concise to work with. I used to have a lot of trouble with Redhat /etc directory and rpm packaging due to Redhat not being the same as what the 'man' or 'info' files said. In Slackware all you have to do is use 'mc' F4 and F7 keys to edit/search what you want to fix in /etc and go to /var/log/packages to read info about what is installed. One website I can recommend that has Slackware news and updates is http://www.userlocal.com Your 133Mhz should handle kde 1.1 ok. I have a 486/20mb ram laptop which runs blackbox/fvwm2 no problems. Netscape 4.73 should be ok on 133Mhz. My ~/.fvwm2rc configuration file is set up with a matrix theme and is easier than the default .fvwm2rc file to work with. Feel free to email me if you would like a copy. cheers, Keith. On Mon, 27 Jan 2003, Isaac Devine wrote: > > I am wanting to replace redhat 7.3 on my laptop with Slackware 8 and I > have prickly PCMCIA settings. Does anyone know where the pcmcia settings > are kept so I can copy them over when slackware is installed. (i.e. put > on floppy then put back on laptop). Also opinons on Slackware welcome. > (mainly using for small size). No gentoo sorry , It's a P133 with 32 Mb > Ram and 2 GB and 56k connection, aka not suitable. > > Isaac > -- > "Don't like someone? try dd if=/dev/random of=/dev/hda" > >
Re: Sendmail
On Fri, 31 Jan 2003, Gareth Williams wrote: > I am setting up mutt on my laptop (low specs, so I want to use as many text > based apps as possible). I can recommend 'slrn' for reading newsgroups. It is a threaded newsreader available from http://www.slrn.org Only a 1 meg download and probably available with 'apt-get' Pine is very easy to read newsgroups with but unfortunately is not threaded or available with debian by default. http://www.washington.edu/pine cheers, Keith.
Re: installfest etc
On Thu, Jason wrote: > Or it could be firewall related. I set up Rays kppp dialup and iptables firewall without checking the /etc/sysconfig/firewall script. Sorry about that. The firewall can be disabled from within KDE under "Settings-Security" in the menu if I recall correctly. It will ask you to enter the root password before you make the changes. Can you try this yourself Ray? Bjorn Nilsen wrote: > Hi Ray, > try to ping an IP address eg: "ping 202.50.167.4". > If this works you probably need to set your DNS > servers manually, this is done in the /etc/resolv.conf file. I think Ray is using kppp which temporarily inserts its own dns settings into /etc/resolv.conf and comments out any dns entries that are already in that file. When kppp is stopped it removes its entries and returns the file to normal. If kppp is being used to dial up paradise the dns ip numbers (for paradise) should be checked in kppp. As far as I know they are correct as Ray had a print out of the paradise numbers that was referred to. > > -Original Message- > > From: Raycl [mailto:[EMAIL PROTECTED] > > This brings me to my next question ,Someone at the install said > > you could load my e-mail manually,If you could that would be great. That would be "fetchmail", a mail retrieval agent (mta) that can fetch mail from pop3.paradise.net.nz. See 'man fetchmail'. It is probably not set up yet. 'mozilla' downloads mail from pop3.paradise.net.nz with its own internal mta so you are more or less loading your e-mail manually now. The advantage of 'fetchmail' is that it can run either manually or automatically in the background to poll pop3.paradise.net.nz every few miniutes. > > I tried To connect to my isp and was successful on the second attempt > > (for some reason the modem isn't activated on the first try),but when > > I try to connect to any of the sites I get a error message saying It > > is an unknown host. Sorry about the mistake with the firewall, try disabling it from within KDE. The firewall configuration/set-up is in the KDE-Settings menu under "Security" I think. sincerely, Keith.
Re: X configuration
On Tue, 1 Apr 2003, Daniel Fone wrote: > This is regarding the Mandrake install I had trouble with on Monday. I have > told him that these sort of queries should be directed to the CLUG. The configuration problem was with a Riva 128 videocard and HP Pavillion M70 monitor that was bringing up X only half successfully. The window managers were not starting and the windows that did appear on occaasions had a zagged apearance, even with 640x480,16 colours etc. I suspect a riva option in /etc/X11/XF86Config as a monitor problem is usually more obvious. Monitor was set to H 30-70, V 50-120. 640x480, 16 colours. This should get sorted out eventually as the XF86Config was set up fine at the installfest with Mandrake 9.0. > Date: Tue, 1 Apr 2003 17:54:21 +1200 > From: "Raycl" <[EMAIL PROTECTED]> > No problems about last night I appreciate you helping. > I decided to try a couple of things myself, All part of the learning curve. :-) > Did a reinstall, the screen was still stable but the desktop wouldn't > or couldn't load. > I then decided to (2nd) reinstall, Screen was stable but still unable to > load the desktop, I suspect a riva 128 option. > when I did the reinstall it picked hd5 as the partition to be used, not > sure if this was the correct one. I would advise against storing valuable data on a extented partition as they are very difficult to recover, althogh having the / on extended is alright. Apparently Dan did the 9.1 install on primary /dev/hda2 and it looks like Mandrake has done the reinstalls on /dev/hda5. cheers, Keith.
Re: X configuration
> From: "Raycl" <[EMAIL PROTECTED]> > > Did a reinstall, the screen was still stable but the desktop wouldn't > or couldn't load. X configuration has been fixed by Ray using a ugrade from 9.0 to 9.1. 9.0 did set up X but 9.1 had problems. Using the upgrade preserved the XF86Config file to make X usuable in 9.1.
Re: Mounting Zip drive...
On Thu, 3 Apr 2003, Zane Gilmore wrote: > I'm fairly sure that Zip drives use some sort of > SCSI interface. Externel parallell port iomega zip drives require the ppa module to be loaded before the lp (printer) module. The ppa modules will detect the drive and use scsi emulation to communicate with the device on /dev/sda Mini-Howto Zip-Drive has info on this. Since Craig is using an internel zip I don't think that scsi emulation is used.(/dev/hdb) New Zip disks 'vfat' are recognised as partitions /dev/hdb4. After formatting as 'ext2' they are named /dev/hdb1 so the /etc/fstab entry has to be changed to mount that device. cheers, Keith.
Re: Julia's Laptop/winmodem - update (Was: Good News for winmodems: mandrake 9.1)
On Thu, Yuri de Groot wrote: > Then when it got to "Logging onto network", it spat the dummy, some kind of > error with ppp. Check /etc/ppp/options. As I understand it Mandy doesn't add these options. user yuri(or whatever 'username' is for pap) defaultroute /dev/modem cheers, Keith.
Re: Mice and RSI
On Wed, Feb 16, at 02:19:34PM, Vik Olliver wrote: > Yeah, I hardly ever use mice. Cordless thumb-based optical trackballs > just beat them hands-down. Agreed.I recently bought a 4d trackball from dick smiths XH1866, a lot better than swapping the mouse over to the left hand. With three buttons and two wheels it was just a matter adding to xorg.conf Identifier "Mouse1" Driver "mouse" Option "Protocol""IntelliMouse" Option "Device" "/dev/mouse" Option "Buttons" "7" Option "ZAxismapping" "6 7 4 5" I know this is a shameless plug but from what I have seen trackballs are a lot cheaper now compared to six years ago at $95. Reversing the buttons for left handers requires xmodmap -e "pointer = 3 2 1" in the .xinitrc file. see man xmodmap. -- keith.
Re: Mice and RSI
On Wed, Feb 16, at 09:47:51PM, Vik Olliver wrote: > I looked at that mouse, but it seemed I might accidentally bash the > buttons on the "wings". Do you find this is a problem? No, the trackball is well away to the right and I use the keyboard most of the time to maximize/iconify windows. A falling coffee cup would probably break a wing but it is less vulnerable from the side because the trackball slides when hit. -- keith.
Re: DSE XC3532 Laptop and Linux Compatability
On Sat, Feb 19, at 12:05:46AM +1300, Brendan Greer wrote: > Does any one know where I can go to find out about linux on > laptops?(Specifically compatability) http://www.tuxmobile.org -- keith.
Re: Dual Network/Internet Connection - Advanced Routing
On Thu, Mar 31, Kim Robertson wrote: > I want the full firewall on both connections. If you can work out a way > to use IPCop that would be great, but I think to make it easier I will > stick with slackware. Hi Kim, configure interfaces in /etc/rc.d/rc.inet1.conf firehol has been recommended on this list as a firewall generator. http://sourceforge.net/projects/firehol There are also a few links at http://www.userlocal.com to ready some ready made scripts. my slackware firewall script is here,it blocks by ethernet device and subnet addresses if required. http://homepages.e3.net.nz/~keithmg/rc.firewall.txt hth, keith.
Re: Mepis freezing when trying to dial-uo - Linz
On Sun, Apr 03, Dave G wrote: > need a bit more info > what program are using to dial-up (I assume kppp?) > what modem install/external ?? did the modem install put alias lines in /etc/modules.conf and was 'depmod -a' run after updating this file. also what does /var/log/messages say about the loading of modem modules and the dialup? -- keith.
Re: Mepis freezing when trying to dial-uo - Linz
On Sun, Apr 03, Lindsay wrote: > Correct kppp & internal modem (56k) 'lspci -v' should give info on the 56k modem. -- keith.
Re: Mepis freezing when trying to dial-uo - Linz
On Sun, Apr 03, Lindsay wrote: > Perhaps I haven't set up something? Did the setup of kppp occur the same way as to other dialer programmes. I believe Memphis has other dialers such as p-on, gppp and wvdial. Are you able to use those. hth, keith.
Re: Mepis freezing when trying to dial-uo - Linz
On Sun, Apr 03, Lindsay wrote: > Found something I think. > Lucent Microelectronics 56k WinModem (rev 01) Same as my 56k. Usually a hurdle for newbies to install as you have to compile the Lucent modules but the modem itself works ok. Did Memphis automatically install its own pre-built Lucent modules for the Lucent modem or did you compile the modules? -- keith.
Re: Re: Need advice on baking up files
On Tue, Apr 05, IT Support NZ wrote: > I found an excellent wee backup prog for linux called sitback. quite configurable with some handy features such as report printing over a samba network and ncurses interface. Its a good idea to give the url when recommending a prog. Sitback http://www.mrbean.dk cheers, keith.
Re: Debian Newbie - network Q 'hard'
On Wed, Apr 06, Steve Holdoway wrote: > As for firewalls, just switch them off until you know things work! the gateway (254) firewall may be dropping pings on the internal interface to the lappie (1). Something like- drop pings $IPTABLES -A INPUT -i $INTERNAL -p icmp --icmp-type echo-request -j DROP drop pongs $IPTABLES -A INPUT -i $INTERNAL -p icmp --icmp-type echo-reply -j DROP take a look at the iptables script and change icmp echo's to ACCEPT for the internal network. hth, keith.
Re: gentoo -dialup probs
On Sun, Apr 10, Nick Rout wrote: > did we compile ppp into your kernel? (as a module is best IIRC) Howard, check these options in /etc/ppp/options. modem /dev/modem# or the modem device. also that the device stated in kppp's configuration is correct. -- keith.
Re: Booting to Console
On Mon, Jun 13, Ralph Stoker wrote: > When I rebooted it booted to console (runlevel 5 showing). Don't seem to be > able to get it to boot to KDE desktop.(can't find kde.init error)... Ralph, check output of the 'mount' command to verify that the partition that kde was installed to is mounted and fix /etc/fstab accordingly. check that the X session is running on Ctrl-Alt+F7 check runlevel entries for each tty in /etc/inittab if stuck in console you can start kde with- 'xinit /opt/kde/bin/startkde &' -- keith.
Re: file system
On Sun, Aug 14, 2005 at 05:53:58PM +1200, John Mallett wrote: > With out mounting a file system. How would I go about finding out what > filesystems I have on my system. ie Vfat or ext2 > is there a simple command 'cfdisk' displays filesystem partitions and types. 'fdisk -l' is used in shell scripts. If you intend to partition your hardrive I recommend 'cfdisk' Newbies using 'fdisk' have a habit of writing overlapping partitions that result in filesystems being lost. from the 'fdisk' manual. BUGS There are several *fdisk programs around. Each has its problems and strengths. Try them in the order cfdisk, fdisk, sfdisk. (Indeed, cfdisk is a beautiful program that has strict requirements on the partition tables it accepts, and produces high quality partition tables. Use it if you can. fdisk is a buggy program that does fuzzy things - usually it happens to produce reasonable results. Its single advantage is that it has some support for BSD disk labels and other non-DOS partition tables. Avoid it if you can. --- keithmg.
Re: August Meeting.
On Tue, Aug 16, 2005 at 01:41:47AM +1200, Ross Drummond wrote: > Try; > http://www.wagemakers.be/uploads/37/7/lookat_bekijk-1.4.0.tar.gz > > A pager with a Midnight Commander like look and interface. Ross, you can use a shell script to have midnight commander do this since mc can format man documents. #!/bin/sh MAN=`man -w $1` mc -cv $MAN Have you heard of 'pinfo', it is a lynx like info and man browser that is capable of following links in the manuals. 300kb download and compiles similar to lookat. Most distros have it as a binary. http://pinfo.sourceforge.net -- keithmg.
Re: embarrassed by ssh hacking.
On Fri, Sep 16, Nick Rout wrote: > I thought I had better start taking a look around and tried to emerge > chkrootkit, but this bombed telling me it failed to untar the source > code :( Nick, trojan binaries are often commands such as ls,find,ps,netstat,less. Chkrootkit uses these commands so for a proper scan use the -p path option 'chkrootkit -p /path/to/trusted/binaries' by copying trusted binaries from the linux install cds. 'chkrootkit' is a shell script which works the "first" time provided you use trusted binaries. > this weekend I have the choice of doing further tests, or doing a complete > re-install (/home is on a separate partition). What do people recommend? The 'find' command might reveal something but a reinstall is the favoured option. find / -mtime -3 > foobar.txt find / -user root -perm -4000 -print > foobar.txt Recommended sites http://www.linuxsecurity.com http://www.cert.org newsgroup comp.os.linux.security Samhain has superceded tripwire. http://la-samhna.de/samhain/index.html Samhain guide http://www.newsforge.com/article.pl?sid=03/07/29/1727249 > I guess the real concern is how they managed to log in in the first > place. Yes, I should not have had the (default) option of allowing root > login via ssh. I don't agree with openssh being installed insecurely with default root login. Dictionary attacks and probes can be avoided by using alphanumeric user names and moving ssh from port 22. hth, keith.
Re: embarrassed by ssh hacking.
On Mon, Sep 19, 2005 at 05:44:48PM +1200, Hadley RIch wrote: > It has it's part of an overall security system. Nothing wrong with avoiding > all the general human/bot scans. Apparently moving port numbers away from their defaults can become awkward when administering large networks. Openssh provides both protocol 1 and 2 for host authentication before login. Protocol 1 is insecure as there is no exchange of public/private keys and ip address spoofing is possible. The protocols are configured this way in sshd_config. Protocol 2,1 This means allow "both" protocols 1 & 2 not a preferential selection of use 2 "before" 1. It is the client request that determines what protocol to use and the only way to force protocol 2 is to configure it as the only protocol provided. See 'man sshd_config' I would assume that most of the exploits are aimed at protocol 1 and since trusted clients are able to use protocol 2 it seems pointless to have sshd enable protocol 1. -- keith.
Re: Warning - xfe-0.84-1 file manager
On Tue, Sep 20, Barry wrote: > It only provides for a tree & 1 directory listing or 2 directory > listings as opposed to a tree & 2 directory listings. Do you mean the Tree and Panels options from the "View" menu? They provide a "tree" and "2" directory (panel) listings. > It comes with its own text viewer and editor which replace those > provided with XWC and it cripples other functionality of XWC. Try "Edit -- Preferences -- Programs" to select viewer & editor of your choice. -- keith.
Re: Second copy of kdm
On Sun, Sep 25, Vik Olliver wrote: > Can anyone give me a clue as to how to start a second kdm using > Ctrl-Alt-F8 ? it can also be done with the kdm Xservers file usually /opt/kde/share/config/kdm/Xservers. :0 [EMAIL PROTECTED] /usr/X11R6/bin/X :0 vt7 :1 [EMAIL PROTECTED] /usr/X11R6/bin/X :1 vt8 The second line sets the Xserver to vt8 but it can be any number that is not used by a console (getty) session. If using Portsentry free up its binding to port tcp 6001 in portsentry.conf as the second Xserver requires this port. In the same directory is "kdmrc" which has options to prevent users shutting down the machine using kdm. ## "Core Config for Local Displays" section. AllowShutdown=Root kdmrc is edited by the kde control centre "Logon Manager" tool. -- keith.
Re: skype
On Sun, Oct 16, 2005 at 10:17:25AM, orac wrote: > Does anyone have Skype working on slackware 10 or above Gary, I havn't used skype but it looks like all the libs are available in the l directory on the 10.1 install disk. Skype.com recommends unblocking outgoing tcp connections below port 443 to allow it to work reliably. wget http://download.skype.com/linux/skype-1.2.0.17.tar.bz2 unpack to /usr/share/skype and set the path variable in /etc/profile. export SKYPE_RESOURCES_PATH="/usr/share/skype" uses alsa sound system and their are "faq" pages at skype.com. let us know any details of what you have trouble with, eg lsmod, ldd skype, alsamixer, amixer. cheers, keith.
Re: Joe editor
On Mon, Nov 14, at 11:34:24AM, Anthony Brown wrote: > I've chosen 'joe' as an editor for some C files I am writing but can't find > how to Make Selection, then Cut/Copy and Paste. It seems this should be very > easy, but I have not been a Linux user in the past so I'm unsure of a few > things. Ant, Ctrl-K H brings up the help menu in joe. Have you tried "jed" which has menus and the ability to split windows and accessible help documentation. Developed by John Davis who also made the "slrn" newsreader. -- keith.
Re: MS to open their open XML format
On Tue, Nov 29, at 11:23:25PM, Volker Kuhlmann wrote: > No, MS announced they were going to register their Office 12 format with > the European ECMA organisation > (http://www.heise.de/newsticker/meldung/66699). This is not ISO! According to David Wheeler (author of the Linux Secure Programming-Howto) MS Open XML format is not standardised to the extent of OpenDocument and is incompatible with GPL applications. http://www.dwheeler.com/essays/why-opendocument-won.html Registration with the European ECMA doesn't mean that MS has declared some sort of alliance with free software, GPL or OpenDocument. -- keith.
Re: Mysql ODBC drivers
On Thu, Dec 08, Nick Rout wrote: > surely the first place to go is your distro's packaging system? Thanks for mentioning it. Certainly easier as distros often supply the MyODBC drivers on the second disk. The "OpenOffice1.1-unixODBC-Mysql-Howto" explains the install and setup procedure for Redhat. Two packages are required, "UnixODBC" and "MyODBC". http://www.unixodbc.org/doc/OOoMySQL.pdf The use of these two packages with OpenOffice is that UnixODBC is the "driver manager" that interfaces between OpenOffice and the Mysql drivers called MyODBC. OpenOffice --> unixODBC --> MyODBC drivers --> mysql database. | --> Posgresql drivers --> posgresql database. The driver manager can interface with more than one database if additional drivers (like posgresql db) are added. After installing the packages run the command 'ODBCConfig' in an xterm to add the path to the selected drivers. For mysql /usr/lib/libmyodbc3.so Check /etc/odbcinst.ini for server,host and password configuration and /etc/odbc.ini for the driver path. Then start OO's 'sbase" wizard to log on to the database. Info that explains further about unixodbc and myodbc configuration is at. http://www.unixodbc.org http://dev.mysql.com/doc/connector/odbc/en/faq_toc.html -- keith.
Re: server live CD
On Sun, May 07, 2006 at 10:48:55PM +1200, Neil Stockbridge wrote: > > does anyone know of a "server live CD"? i want to download a live CD > ISO with a system that when booted, configures all network interfaces > via DHCP and then starts sshd. if it has parted too that would be > great. Neil, Most live cd's store configuration a floppy disk or pen drive so enabling dhcp, the interfaces and route should be easy enough. Having the required modules loaded for the ethernet controller may be by default or need a 'modprobe' command in /etc/rc.d/rc.local. http://www.devil-linux.org/ is a gateway that provides http, ftp and ssh servers. A sample etc.bzip file is provided for the floppy disk on which you can configure the passwd, hosts, rc.local and /etc/sysconfig/network files. http://www.devil-linux.org/documentation/1.2.x/ch01s02.html download url- http://optusnet.dl.sourceforge.net/sourceforge/devil-linux/devil-linux-1.2.9-i486.tar.bz2 I havn't tried this devil-linux distro. There is a mini floppy distro called MuLinux which can also do the job if you edit the boot and server floppies, but it is no longer under active development and uses a 2.2.X kernel. hth, keith.
Re: iptables
>> Steve Holdoway wrote: >> >> iptables -F INPUT >> iptables -A INPUT -s -j DROP ( x 10 ) >> >> Using this method, it also stops access to localhost. >> What am I missing??? >> >> > On Fri, May 12, at 12:32:57PM, Craig FALCONER wrote: > > iptables -A INPUT -s 127.0.0.1 -j ALLOW > iptables -A INPUT -s -j ALLOW 127.0.0.1 and are reserved addresses that can be spoofed in packets coming from external hosts, so you should add rules that apply to the input -i and output -o interfaces not just a -s . # loopback rules iptables -A INPUT -i lo -j ACCEPT iptables -A OUTPUT -o lo -j ACCEPT #drop addresses that may be spoofed from external hosts. iptables -A INPUT -i ppp0 -s 10.0.0.0/8 -j DUMP iptables -A INPUT -i ppp0 -s 192.168.0.0/16 -j DUMP -- keith
Re: iptables
On Fri, May 12, 2006 at 07:29:53PM, Keith McGavin wrote: > #drop addresses that may be spoofed from external hosts. > iptables -A INPUT -i ppp0 -s 10.0.0.0/8 -j DUMP > iptables -A INPUT -i ppp0 -s 192.168.0.0/16 -j DUMP DUMP is a user-defined target and the default DROP target was intended. iptables -A INPUT -i ppp0 -s 10.0.0.0/8 -j DROP iptables -A INPUT -i ppp0 -s 192.168.0.0/16 -j DROP apologies for the error. --- keith.
Re: Firewall suggestions?
On Wed, May 17, at 07:36:20PM, Volker Kuhlmann wrote: > One could periodically check up on the IPs automatically... so far > I've seen nothing that can do that though. a cron job can run the 'host' and 'iptables' commands in a script such as rc.local. This script is an example only. Iptables will have to flushed/ restarted to accept old dropped addresses if the smtp relay keeps changing all the time. rc.local-smtp --- #!/bin/sh # Restart iptables if the smtp relay address keeps changing. # Every distro has a diiferent way of doing this. Can be run # as a cron job or done manually instead. # /etc/rc.d/rc.firewall restart # variable for some smtp relay servers address. Edit the hostname. SMTPADDRESS=`host smtp.some-isp.co.nz | awk -F " " '{print $4}'` # allow smtp access to that server only. iptables -A OUTPUT -p tcp -d $SMTPADDRESS --dport 25 -o ppp0 -j ACCEPT iptables -A OUTPUT -p tcp -d ! $SMTPADDRESS --dport 25 -o ppp0 -j DROP # initialise a temp file with a localhost address # first time the script is run. if [ ! -f "/tmp/smtpaddress.txt" ]; then touch /tmp/smtpaddress.txt echo "127.0.0.9" > /tmp/smtpaddress.txt echo "smtpaddress.txt created" break; # update the temp file and drop the previous smtp relay address. else if [ -f "/tmp/smtpaddress.txt" ] && [ `cat /tmp/smtpaddress.txt` != $SMTPADDRESS ]; then PREVSMTP=`cat /tmp/smtpaddress.txt` iptables -A OUTPUT -p tcp -d $PREVSMTP --dport 25 -o ppp0 -j DROP echo "$PREVSMTP denied" echo $SMTPADDRESS > /tmp/smtpaddress.txt echo "smtpaddress.txt updated" fi fi --
Re: Firewall suggestions?
On Sat, May 20, at 06:08:05AM, Keith McGavin wrote: > This script is an example only. Iptables will have to flushed/ > restarted to accept old dropped addresses if the smtp relay > keeps changing all the time. I didn't get this part right, it is not required to flush/restart iptables since all it achieves is keep the rules tidy by not repeating the DROP rule on the same ip address and port twice when $PREVSMTP keeps taking on new values. > iptables -A OUTPUT -p tcp -d ! $SMTPADDRESS --dport 25 -o ppp0 -j DROP > iptables -A OUTPUT -p tcp -d $PREVSMTP --dport 25 -o ppp0 -j DROP --- keith.
Re: Firewall suggestions?
On Sat, May 20, 2006 at 09:40:15AM +1200, Volker Kuhlmann wrote: > This isn't good enough, you are relying on the assumption that > that when the host command runs, the rest of the iptables rules > are already set up, especially the DNS related ones. I understand your point about the difficulties of adding rules for a DNS address when using bulk commit but I was writing about a changing smtp-relay addresses. > You will have to keep the IP numbers you wish to look up for your > iptables script in a file, which is updated independently of your > iptables service, at a time when the iptables service is guaranteed to > be already up. This may cost you 2 iterations - start iptables, look up > your variable IP numbers, restart iptables. thanks for the advice. --- keith.
Re: parsing
On Fri, May 26, 2006 at 11:29:54AM, Nick Rout wrote: > what is the best way to parse out the first load average figure, ie in > this case 18.73 'awk' accepts an "array" of field separators -F'[fsfsfs]' between square brackets so the required field $12 can be printed out. Not mentioned in the manual. echo $THAT_LINE | awk -F'[,:]' '{print $12}'
Re: parsing
On Sat, May 27, Hadley Rich wrote: > I must play with awk some more. explaination of some syntax in the awk/gawk manual is not always obvious but the awk/gawk gurus on news://comp.lang.awk will clarify how it is used if google fails. An active newgroup. --- keith