Re: [expert] GnuPG - pass phrase solved
LOL, if only more people would do that! Of course some of us could do well by reading our own posts before we send them, or reading other peoples posts before we reply... Oops that's just me :-) On Friday, July 13, 2001, at 07:05 PM, Phil wrote: Hello All, I solved the problem and I did read the manual. It wasn't until the next day that I realised what the answer was. -- Regards, Phil [EMAIL PROTECTED]
Re: [expert] Multiple network cards in a Mandrake firewall/switch combination
Hi... This is probably all my fault for confusing the issue, so I guess that I will have to try and help clean it up :-) If you are using something resembling my silly little setup, BOTH nics need to be set up with 255.255.255.128 netmasks. If you are using something more like what the wise civileme wrote, is to use a completely different network for the 10baseT hosts. Something in the 172.16.*.* range (as he suggested) or the 10.*.*.* would be appropriate (I believe these are the IP ranges, along with 192.168.*.*, researved for private networks) Hope this helps, sorry for making muddying the waters. On Thursday, July 12, 2001, at 12:46 PM, Darcy Brodie wrote: I have added the additional nic into the firewall / masqurading machine, and configured it to 192.168.1.128/255.255.255.128 I can ping the address from the 100mhz network(both from the server, and from remote workstations). However, I connected a workstation to the new nic card (eth2) through a 10mhz hub, and I can not ping either the eth2 card from the remote 10mhz workstation, or the remote workstation from the server. I have verified that the hub and the cables are working. I have even used a cross over cable from the workstation to the server, but I still can not ping the eth2 card Darcy Nathan Callahan wrote: There is another option. You could set the machine up as an ethernet bridge as I am doing here so that I can use my powerbook on our local coax network, and get to the masquerading host easily and so that the other people on the network don't need to change their settings to see my machine. It means that hosts on two subnets can see each other as though they were on the same subnet, basically like a switch (only cheaper). Although this is much easier to do on a 2.4 kernel, it can be done under 2.2, I just can't remember how at the moment, but I remember that it does require a special utility (and there is a howto) If you _are_ running 2.4... here's how to do it. configure one card to have an address in the range 192.168.2.1 through .127, and the other in 192.168.128 through .254 and give both a 255.255.255.128 netmask. All machines on the 1-127 side need to have ip addresses in this range, all machines on the other side, ip addresses in 128-254 then issue the commands... (assuming that the cable modem is on eth0, the local cards being eth1 and 2) echo 1 /proc/sys/net/ipv4/conf/eth1/proxy_arp echo 1 /proc/sys/net/ipv4/conf/eth2/proxy_arp and turn on forwarding between the interfaces... echo 1 /proc/sys/net/ipv4/ip_forward The proxy arp bit basically make the machine transparent as far as the local network is concerned all machines can carry on having 255.255.255.0 netmasks. The other thing is that if you do have a firewall set up on that box, and as civileme has suggested, the forward policy is DENY, you will probably need something along the lines of. ipchains -A forward -s 192.168.1.0/24 -d 192.168.1.0/24 -j ACCEPT I make no guarantees as to the completeness or robustness of this solution, it works for me, YMMV. Hey, even if this doesn't help you a bit, I think it's pretty cool and felt like showing it off anyway :-) Plus, it may help someone else. Regards, Nathan Callahan On Tuesday, July 10, 2001, at 04:18 PM, Darcy Brodie wrote: and in his usual, amazingly helpful style... civileme wrote: On Tuesday 10 July 2001 04:47, Darcy Brodie wrote: Hello I hope that this can be done. I currently have a LM7.2 box as a firewall for our internet access. Cable modem from ISP is going to eth0. eth1 (100baseT) is going to the internal network. What I need to do, is add a 3rd network card to allow me to also have a 10baseT network within the local netwok. Can this be done with Linux? Have not been able to find any information in the how-to's on this configuration. I also, if need be, have access to a second Linux file server, that I could add additional network cards into (it currently only has 1 card in it) I am currently using class C IP 's in the 192.168.1.X range, but this is flexible if required. Thanks Darcy Just add the card and setup adaptor. If you are making this a different network and want the two to talk, you will need to setup a route and make sure your internet masquerading rules apply only to forwards pointed at the internet interface. Since the first instruction in many masquerading setups is ipchains -P forward DENY you will need to write a series of rules in terms of -i ethx -o ethy to cover all possible combos. Of course if you set up netmasks so they are effectively on the same network, then the route does not need to be added, but you still need the rules for forwarding. Another approach, using your other box, is to make it a masquerading gayeway from the 10baseT net to the 192.168 net, and use some other schem for the others like 172.16.x.y
Re: [expert] Multiple network cards in a Mandrake firewall/switch combination
OMG I did F...F...F..Foul! up that first part... Gareth is completely right. Sorry again. On Thursday, July 12, 2001, at 09:33 PM, Gareth Allen wrote: Just for info, when using a 255.255.255.128 netmask, the valid address ranges are 192.168.1.0 Network Address 192.168.1.1-126 Host Address 192.168.1.127 Broadcast Address 192.168.1.128 Network Address 192.168.1.129-254 Host Address 192.168.1.255 Broadcast Address Basically in setting 192.126.1.128, you are using an invalid address. Regards Gareth -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of Nathan Callahan Sent: 12 July 2001 12:12 To: Darcy Brodie Cc: [EMAIL PROTECTED] Subject: Re: [expert] Multiple network cards in a Mandrake firewall/switch combination Hi... This is probably all my fault for confusing the issue, so I guess that I will have to try and help clean it up :-) If you are using something resembling my silly little setup, BOTH nics need to be set up with 255.255.255.128 netmasks. If you are using something more like what the wise civileme wrote, is to use a completely different network for the 10baseT hosts. Something in the 172.16.*.* range (as he suggested) or the 10.*.*.* would be appropriate (I believe these are the IP ranges, along with 192.168.*.*, researved for private networks) Hope this helps, sorry for making muddying the waters. On Thursday, July 12, 2001, at 12:46 PM, Darcy Brodie wrote: I have added the additional nic into the firewall / masqurading machine, and configured it to 192.168.1.128/255.255.255.128 I can ping the address from the 100mhz network(both from the server, and from remote workstations). However, I connected a workstation to the new nic card (eth2) through a 10mhz hub, and I can not ping either the eth2 card from the remote 10mhz workstation, or the remote workstation from the server. I have verified that the hub and the cables are working. I have even used a cross over cable from the workstation to the server, but I still can not ping the eth2 card Darcy Nathan Callahan wrote: There is another option. You could set the machine up as an ethernet bridge as I am doing here so that I can use my powerbook on our local coax network, and get to the masquerading host easily and so that the other people on the network don't need to change their settings to see my machine. It means that hosts on two subnets can see each other as though they were on the same subnet, basically like a switch (only cheaper). Although this is much easier to do on a 2.4 kernel, it can be done under 2.2, I just can't remember how at the moment, but I remember that it does require a special utility (and there is a howto) If you _are_ running 2.4... here's how to do it. configure one card to have an address in the range 192.168.2.1 through .127, and the other in 192.168.128 through .254 and give both a 255.255.255.128 netmask. All machines on the 1-127 side need to have ip addresses in this range, all machines on the other side, ip addresses in 128-254 then issue the commands... (assuming that the cable modem is on eth0, the local cards being eth1 and 2) echo 1 /proc/sys/net/ipv4/conf/eth1/proxy_arp echo 1 /proc/sys/net/ipv4/conf/eth2/proxy_arp and turn on forwarding between the interfaces... echo 1 /proc/sys/net/ipv4/ip_forward The proxy arp bit basically make the machine transparent as far as the local network is concerned all machines can carry on having 255.255.255.0 netmasks. The other thing is that if you do have a firewall set up on that box, and as civileme has suggested, the forward policy is DENY, you will probably need something along the lines of. ipchains -A forward -s 192.168.1.0/24 -d 192.168.1.0/24 -j ACCEPT I make no guarantees as to the completeness or robustness of this solution, it works for me, YMMV. Hey, even if this doesn't help you a bit, I think it's pretty cool and felt like showing it off anyway :-) Plus, it may help someone else. Regards, Nathan Callahan On Tuesday, July 10, 2001, at 04:18 PM, Darcy Brodie wrote: and in his usual, amazingly helpful style... civileme wrote: On Tuesday 10 July 2001 04:47, Darcy Brodie wrote: Hello I hope that this can be done. I currently have a LM7.2 box as a firewall for our internet access. Cable modem from ISP is going to eth0. eth1 (100baseT) is going to the internal network. What I need to do, is add a 3rd network card to allow me to also have a 10baseT network within the local netwok. Can this be done with Linux? Have not been able to find any information in the how-to's on this configuration. I also, if need be, have access to a second Linux file server, that I could add additional network cards into (it currently only has 1 card in it) I am currently using class C IP 's in the 192.168.1.X
[expert] Software Manager Problem
Hi All, I just got LM 8.0 Powerpack. I've been trying to add the CDs as extra sources in Software Manager, but it always seems to fail to read the second disk that I put in, and the drive becomes unusable until I reboot (i/o error is the message that I seem to remember) . I'm running a Pioneer 16x DVD ROM drive, of an ASUS Mobo (Can't remember the model at the moment) with a Slot A 700Mhz Athlon. Any help greatfully accepted.
Re: [expert] Multiple network cards in a Mandrake firewall/switch combination
There is another option. You could set the machine up as an ethernet bridge as I am doing here so that I can use my powerbook on our local coax network, and get to the masquerading host easily and so that the other people on the network don't need to change their settings to see my machine. It means that hosts on two subnets can see each other as though they were on the same subnet, basically like a switch (only cheaper). Although this is much easier to do on a 2.4 kernel, it can be done under 2.2, I just can't remember how at the moment, but I remember that it does require a special utility (and there is a howto) If you _are_ running 2.4... here's how to do it. configure one card to have an address in the range 192.168.2.1 through .127, and the other in 192.168.128 through .254 and give both a 255.255.255.128 netmask. All machines on the 1-127 side need to have ip addresses in this range, all machines on the other side, ip addresses in 128-254 then issue the commands... (assuming that the cable modem is on eth0, the local cards being eth1 and 2) echo 1 /proc/sys/net/ipv4/conf/eth1/proxy_arp echo 1 /proc/sys/net/ipv4/conf/eth2/proxy_arp and turn on forwarding between the interfaces... echo 1 /proc/sys/net/ipv4/ip_forward The proxy arp bit basically make the machine transparent as far as the local network is concerned all machines can carry on having 255.255.255.0 netmasks. The other thing is that if you do have a firewall set up on that box, and as civileme has suggested, the forward policy is DENY, you will probably need something along the lines of. ipchains -A forward -s 192.168.1.0/24 -d 192.168.1.0/24 -j ACCEPT I make no guarantees as to the completeness or robustness of this solution, it works for me, YMMV. Hey, even if this doesn't help you a bit, I think it's pretty cool and felt like showing it off anyway :-) Plus, it may help someone else. Regards, Nathan Callahan On Tuesday, July 10, 2001, at 04:18 PM, Darcy Brodie wrote: and in his usual, amazingly helpful style... civileme wrote: On Tuesday 10 July 2001 04:47, Darcy Brodie wrote: Hello I hope that this can be done. I currently have a LM7.2 box as a firewall for our internet access. Cable modem from ISP is going to eth0. eth1 (100baseT) is going to the internal network. What I need to do, is add a 3rd network card to allow me to also have a 10baseT network within the local netwok. Can this be done with Linux? Have not been able to find any information in the how-to's on this configuration. I also, if need be, have access to a second Linux file server, that I could add additional network cards into (it currently only has 1 card in it) I am currently using class C IP 's in the 192.168.1.X range, but this is flexible if required. Thanks Darcy Just add the card and setup adaptor. If you are making this a different network and want the two to talk, you will need to setup a route and make sure your internet masquerading rules apply only to forwards pointed at the internet interface. Since the first instruction in many masquerading setups is ipchains -P forward DENY you will need to write a series of rules in terms of -i ethx -o ethy to cover all possible combos. Of course if you set up netmasks so they are effectively on the same network, then the route does not need to be added, but you still need the rules for forwarding. Another approach, using your other box, is to make it a masquerading gayeway from the 10baseT net to the 192.168 net, and use some other schem for the others like 172.16.x.y This permits both local net and internet access and keeps the networks separated without a lot of rules complexity. internet _| | Gateway | |Current | | Local| |_| | _|___ || |_ __| | | | | | | Other box | (current local net) | Interface to | | other | |__| | __| | || | (new local net) In the ASCIIgram above, the boxes shown both use masquerading and the one handling the 10MHz net is 100MHz on the main net, something like a data compression switch. It can also be peered with the other local net computers. Finally, how about just using one port off
[expert] SBLive! (lots of them)
I seem to have some weird stuff going on with the sound on my machine at the moment. I have 1 Soundblaster Live! installed in my machine. In an effort to get midi going, I have tried running alsaconf from the command line as suggested earlier on this list, I have also tried reconfiguring the card from Harddrake. The upshot of all this is that I now have three sound mixers. Two are labelled Creative SBLive and say - Invalid mixer Creative SBLive - when I bring them up in the GUI sound mixer. The third is labelled as TriTech (23) and sort of works, but the mixer is all confused and I have to control the volume to my front and rear speakers independently, I used to be able to control them both from the master volume when I had one Creative SBLive Mixer and no more. Can anyone help me get rid of these strange settings? BTW I just bought Mandrake 8 Powerpack today... Hey, maybe I could use some of my techsupport :-) Thanks in advance.
Re: [expert] script help ...
HI, If you want the same answer to all questions for a particular command, there is an amazing little utility by the name of yes By default it echos an endless stream of y characters. If you pass it anything as an argument, then that is what is echoed endlessly instead. You can almost certainly find it in /usr/bin/yes. Just pipe the output to the command that you want the responses for. As has already been pointed out, if the answer varies for any given command, you should probably use expect. On Thursday, July 5, 2001, at 05:35 PM, faisal gillani wrote: hello there can any one help me in making a small script all i want is that my system runs itself in single user mode , then run fsck on / , then return to normal mode... i know the commands 1, init 1 2, fsck / 3, init 6 i am stuck in how can i give the answers to the command automatically ... is it possible ? thanks Faisal __ Do You Yahoo!? Get personalized email addresses from Yahoo! Mail http://personal.mail.yahoo.com/
Re: [expert] dual Athlon CPU not recognised under Mandrake 7.2?
On Wednesday, July 4, 2001, at 07:10 PM, Peter Varnai wrote: Dear Experts, I installed Mandrake 7.2 on a test machine: Tyan thunder k7 board + 2 Athlon 1200 MHz Mandrake 7.2 a) Bastard :-) If you can't get it to work, send it here and I will. /proc/cpuinfo: only CPU 0 recognised ... JHow can I get the kernel notice the second CPU? Any help would be appreciated! b) Checking the bleedingly obvious, I'm assuming that you do have an SMP (Symmetric Multi Processing) enabled kernel installed. c) If so, try the very latest kernel sources that you can lay your hands upon. This is a very new board, and I reckon that linux support is only just appearing.
Re: [expert] Can't get xawtv working with bttv
Hi, Have to go shortly so I can't take the time to look up things. I recall that in order to get the bttv chip working, you need to pass the module some parameters. The readme in the documentation subdir of the kernel sources is kind of handy. On Monday, July 2, 2001, at 07:52 AM, JR Lefty wrote: HELP!, ok, heres the details: lsmod results. msp340013904 0 (unused) tuner 4624 0 (unused) bttv 57648 0 (unused) i2c-algo-bit7200 0 [bttv] i2c-core 12720 0 [msp3400 tuner bttv i2c-algo-bit] videodev4704 0 [bttv] /var/log/messages info. - Jul 1 16:18:11 jrh6pc kernel: bttv: driver version 0.7.65 loaded Jul 1 16:18:11 jrh6pc kernel: bttv: using 2 buffers with 2080k (4160k total) for capture Jul 1 16:18:11 jrh6pc kernel: bttv: Bt8xx card found (0). Jul 1 16:18:11 jrh6pc kernel: bttv0: Bt848 (rev 17) at 00:0c.0, irq: 10, latency: 64, memory: 0xebbfe000 Jul 1 16:18:11 jrh6pc kernel: bttv0: model: BT848(Hauppauge old) [insmod option] Jul 1 16:18:11 jrh6pc kernel: bttv0: IRQ 10 busy, change your PnP config in BIOS Jul 1 16:18:16 jrh6pc kernel: i2c-core.o: driver i2c TV tuner driver registered. Jul 1 16:18:26 jrh6pc kernel: i2c-core.o: driver i2c msp3400 driver registered. For info below: My card is a hauppauge 848 card with the phillips NTSC tuner modules.conf contents - alias char-major-81 videodev alias char-major-81-0 bttv options bttv card=2 tuner=2 options tuner debug=1 type=2 v4l-info output - When running v4l-conf, the following messages come through. v4l-conf v4l-conf: using X11 display :0.0 dga: version 2.0 mode: 1280x1024, depth=16, bpp=16, bpl=2560, base=0xe0041000 can't open /dev/video: No such device ls -l /dev/video* ls -l /dev/video* lrwxrwxrwx1 root root6 Apr 21 18:04 /dev/video - video0 crw-rw-rw-1 jrh sys 81, 0 Apr 14 07:06 /dev/video0 crw---1 jrh sys 81, 1 Apr 14 07:06 /dev/video1 crw---1 jrh sys 81, 2 Apr 14 07:06 /dev/video2 crw---1 jrh sys 81, 3 Apr 14 07:06 /dev/video3 Help! All the modules appear to be installed yet.. the device is not reacting.. ___ Send a cool gift with your E-Card http://www.bluemountain.com/giftcenter/
Re: [expert] Network sniffing, how?
On Tuesday, June 26, 2001, at 11:57 PM, Craig Sprout wrote: Laurent Duperval wrote: Hi, How do I sniff the packets coming thru my network connection? At home, I have a cable modem and last night I noticed some activity on it, though I know I'm not doing anything using the network. How can I sniff what's going on there? Ethereal is a packet analyzer that is, IMO, one of the most useful network tools you can have in your arsenal. Or, for something a little more raw you could try tcpdump.
[expert] MandrakeFreq + Mandrake Update
Probably not the forum for this. But does anyone know if the MandrakeFreq files are available for download individually in a form that you can point Mandrake Update at? I have had nothing but trouble with cooker RPMs and this seems like a nice safe way keep somewhat up to date, but 2 whole CDs a month is going to blow my download limit. Perhaps I should just bite the bullet and subscribe.
Re: [expert] icons disappered
You could try the update-menus script. It fixed a similar problem for me, once after installing a new version of KDE, and again after installing some software. This was back when I was still running 7.2 though. On Saturday, June 23, 2001, at 02:32 PM, Alden Torres wrote: almost all my icons in the panel and the main menu disappearead. Im use Mandrake 8.0.
Re: [expert] GCC 3.0 - Question on GPL
Hi! I thought I should chime in on this one, even though JoeLX seems to be happy. As I understand it, ALL of GCC3.0 is published under the GPL, which is viral. The GNU Lesser General Public License is one usually used for libraries, which is far less viral. However, GCC3.0 has a special exception for the GCC library. I haven't looked at what this is precisely, but I know that its intention is to allow the production of proprietary software with the compiler system. I cannot tell you what the distribution requirements for that library are, as I have not yet read the exception in the license. BTW M$, AFIK, does not use GPLed code in any of its operations. It uses BSD licensed code, which, by not having the contagious elements of the GPL, lets this sort of thing go on. I hope that this has cleared a couple of things up. But the short answer is AFIK you can distribute code made with GCC3.0 under any license that you feel like; although, if the libraries are under the LGPL as has been suggested (I can't find any reference to this, just an exception) then there are a few minor restrictions that you must be aware of. Regards, Nathan Callahan On Thursday, June 21, 2001, at 02:45 PM, JoeLX wrote: Thanks Tom Thanks Craig The reason for my question was because I saw some software, having developed using GCC, were distributed as proprietary and some are labeled as commercial. I guess these developers should give many thanks and appreciation to the developers of GCC and should contribute to Free Software cause... unlike MS Hotmail, Interix etc... use GPL but never admit it nor appreciate it, rather it chose to bash and bash and bash... Stupid people never learn. OK, thank again sirs! Have a good day! Best Regards, Joe RLU# 186063 Reading is the essence of knowledge - Cut Paste: Tom Badran [[EMAIL PROTECTED]]: Anything you create under gcc you are free to distribute under any terms you wish. The libraries are released under the lgpl which means you can dynamically (and possibly statically) link to them with no requirements on your part. Basically, unless you use anyhting more than libc and glibc you dont have to release your software under the gpl -- Imperial College, Department Of Computing email: [EMAIL PROTECTED] tel: 020 785 22277 -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] Sent: Wednesday, June 20, 2001 07:42 PM To: [EMAIL PROTECTED] Subject: Re: [expert] GCC 3.0 - Question on GPL Shahrimi Johann wrote: Hi all, just a question. GCC is a GPL, open source software and it comes with libraries that will be used in the programs that are created using them. I'm no expert on the GPL, but, as I understand it, it only becomes viral when you use source from another GPL'd program. It doesn't much matter if you compile it with gcc, cc or any other compiler. Libraries are a little different, and fall under the LGPL, and I'm not very familiar with that at all. Just my .02. -- Craig Sprout Network Administrator Crown Parts and Machine, Inc. http://www.crownpartsandmachine.com/
Re: [expert] How do I set up a joystick?
If this is a USB sidewider that you are talking about, try modprobe joydev. Also: rm /dev/js0 ln -s /dev/input/js0 /dev/js0 might make life a bit easier. Hope this helps. By the way, can anyone tell me how to calibrate said joystck? On Sunday, June 17, 2001, at 02:58 PM, Ken Arromdee wrote: I am running Mandrake 8, with fvwm2 as a window manager and no Gnome or KDE. I'm trying to compile/run xmame and need joystick support, but can't figure out how to get it. harddrake shows a line for joysticks, but if I click on it nothing happens. probing options lists no configuration tool for joysticks. rpmdrake lists no packages that contain the string joy. xf86cfg doesn't seem to anything useful that I can figure out (like put a joystick section in my XF86Config-4). I can try modprobe analog or turn on my Sidewinder and try modprobe sidewinder. This doesn't give an error message, but doesn't do anything useful that I can see, either. What can I do to get a working joystick?
Re: [expert] spelling
Hi, I can't quite remember what the rpms are called, but the LM cds have ispell dictionaries for UK english, as well as a there being a UK english locale included in the en qt locale. Once these are installed, you should be able to go to Personalis(z)ation-Country Language in the KDE Control Centre(er) and select UK English as your Language. Hope that this is the right path, not the garden variety. Regards, Nathan On Thursday, June 14, 2001, at 05:16 AM, Paul Stear wrote: Hi all, My spelling could be better and I use spelling checkers alot, however, I can only get the ammerican way of spelling words using any application in KDE. I know the differences are only minor like color and colour, neighbor and neighbour etc. Can somone point me in the right direction please -- regards Paul This message has been sent using Mandrake Linux and kmail
Re: [expert] cvs and X forwarding
Hi, My guess is that you are not starting ssh with the -X flag. This X11 forwarding. Possible other problems, the DISPLAY environment variable is not correctly set (although it should be in an xterm). If this is the case, assuming that you only have one X11 server running on one monitor, try export DISPLAY=:0. Elsewise, I think I need more information to come up with more useless ideas :-) On Wednesday, June 13, 2001, at 06:30 PM, Eivind Myrold Eriksen wrote: Hi! I have several machines running Mandrake 8.0. When I use ssh (as normal user) between two of these machines and want to start an X application from the remote machine, I get: Error: Couldn't find per display information. If I su before I initiate the ssh connection, there is no problem at all. Is there a remedy for this? Thanks!
Re: [expert] Poor OpenGL performance in LM 8.0
Don't know what causing tuxracer to not report FPS correctly, but I think know what the problem with the Radeon is. The XF86 version that comes with LM8 does not support DRI (Direct Rendering Infrastructure) on the radeon. You will have to obtain a new version of XF86 either from the DRI project (dri.sourceforge.net) or I believe that XF86-4.1.0 includes radeon DRI support. One gotcha is that you need agp support enabled in the kernel (either through compile in, or module load) before the radeon driver is loaded by XF86, or it won't work. Otherwise the installation instructions on the DRI homepage worked fine for me. Hope this helps. On Wednesday, June 13, 2001, at 08:00 AM, Ty Auvil wrote: When I run any OpenGL program (Tuxracer etc...) the framerate is extremely poor. However, the Mesa demos look allright. What confuses me is that I set the FPS display on in Tuxracer and it showed between 50 and 90 FPS. What I see is only about 1 or 2 FPS. When I run Tuxkart, the load up info tells me that Utah GLX is running in indirect mode. Is this the problem? My Hardware is: Radeon 32MB DDR Duron 750 w/ Via Chipset I'm running XFree86 4.0.3 with 'Experimental 3D support.' Is there a simple fix for this? If not, can anyone point out a FAQ. Thanks, _ Do You Yahoo!? Get your free @yahoo.com address at http://mail.yahoo.com
Re: [expert] Terminal screen unusable
Don't really know what could cause it, I do know a workaround for a similar problem that I hat under MD7.2 If you log in blind to the virtual terminal and reset is (by typing reset) it may fix the problem. Regards, Nathan Callahan On Thursday, June 7, 2001, at 03:02 PM, M L Cates wrote: I am running LM8.0, XFree3.3.6, with an Alliance AT3D video card. My problem is this: After starting X (KDE) if I do a CTRL-ALT-F1 or F2,etc. I cannot read the screen. Also even after I shutdown X the problem remains. The screen is out of sync basically with about 8 - 10 vertical frames displayed each about 1 inch high stair stepped down my screen. I previously was running RH7.0 with no problem and have run Mandrake 7.1 on this machine with no problem. Does anybody out there have any idea what could cause this? Thanks Mike Cates
Re: [expert] samba/cups
This is an interesting little problem, one that I have come across here at home. Firstly, if there are cups drivers for the printer model, it really shouldn't be a problem, just share the printer, find out its name, and point cups at it. This assumes that samba is set up alright, which it seems to be out of the box on LM8.0. Now the fun part comes if you do not have such a driver. You need to get Aladin Ghostscript for Windows and a little utility by the name of Redmon. I can't remember exactly how this is set up, but basically you share a generic postscript printer (Apple Color Laserwriter works well) and redirect the input via Ghostscript to the windows drivers using Redmon. If this last thing is what you need, I can find out more details for you. Regards, Nathan Callahan On Tuesday, June 5, 2001, at 12:59 AM, mike wrote: Where is some new documentation of how to get kups and samba to play nicely printing to a printer connected to a Windows 98 machine. Thanks. Michael H. Collins
Re: [expert] routing problem
You have it set so that 131.103.1.10 and 10.10.90.99 are gateways. This probably isn't what you want, as it means that these hosts are assumed to be responsible for all traffic bound for their respective networks. If you remove the gw x.x.x.x parts from the respective routing tables, it will probably work. The other thing is that you may need to turn on proxy arp if you want the computer to act as a bridge between these networks. This can be done with echo 1 /proc/sys/net/ipv4/conf/eth0/proxy_arp echo 1 /proc/sys/net/ipv4/conf/eth1/proxy_arp I think that this is only nessessary if you need the box to be transparent (like a switch) and probably only if the machines on either side don't know that they are on different networks. Regards, Nathan Callahan On Wednesday, June 6, 2001, at 09:33 AM, Doug Gough wrote: I'm not able to get my LM8.0 box to work as a router between to LANs. When it boots, I get a message saying IP forwarding is on. My routing table is very simple, using static routing as follows 131.103.1.0 131.103.1.10255.255.255.0 UG 0 0 0 eth1 10.10.0.0 10.10.90.99 255.255.0.0 UG 0 0 0 eth0 127.0.0.0 * 255.0.0.0 U0 0 0 lo It couldn't get much simpler. I have checked and rechecked the IP addresses and netmasks, and found everything to be correct. From the 131.103.1.0 network, I can ping 131.103.1.10 and 10.10.90.99, but I can't reach anything else on the 10.10.0.0 network. I've read as much as I can find on the subject, undoubtebly missing the most simple and obvious :-) Any hints and help would be appreciated. Thank You Doug Gough Computer Services Pacific Academy
[expert] Routing / NAT problem
I've got a problem which must be solved by tomorrow. I need to be able to take all packets bound for a particular local subnet (eg 192.168.100.0/24) and instead send them off (probably using GRE encapsulation) to an internet address (eg 123.456.78.90) instead. I cannot set up a VPN at the moment, it will be done in the near future. If anyone has a good clue on this one, please tell me. Nathan Callahan
Re: [expert] Routing / NAT problem
Thanks for the input. On Monday, June 4, 2001, at 10:26 PM, Randy Kramer wrote: This is probably a bad clue, but I thought I'd throw it out and see if it might be workable: How about adding a line to your routing table to set up the internet address (123.456.78.90) as a gateway to subnet 192.168.100.0/24? Tried that, didn't work. Unfortunately the pack is not translated for the new network and gets thrown out onto the net as a packet bound for 192.168.100.?... Not good. I have actually got the answer now, I think. What I need to do is masquerade the packet, then port forward it to the port that it came in on, on the target host. ipchains can't do this, but someone put me onto ipmasqadm, which looks like it can. iptables can do it too, but the gateway in question is running a 2.2 kernel. Thanks people. If anyone notices a glaring flaw in my logic, feel free to put it out. I can't tell you more about how to do it -- is there a command like addroute or routeadd, or can you do this in netconf? And, I don't know if it will work, And, if it does work to get the packets there, I'm not sure that the internet machines will do something useful with them or just attempt to send them back to you (or /dev/null). Sorry, I know I'm not being real helpful, more curious than anything, Randy Kramer Nathan Callahan wrote: I've got a problem which must be solved by tomorrow. I need to be able to take all packets bound for a particular local subnet (eg 192.168.100.0/24) and instead send them off (probably using GRE encapsulation) to an internet address (eg 123.456.78.90) instead. I cannot set up a VPN at the moment, it will be done in the near future. If anyone has a good clue on this one, please tell me. Nathan Callahan
Re: [expert] IP Masquerading Problems
You might want to try: echo 1 /proc/sys/net/ipv4/ip_forward Also I noticed that the routing for eth0 is in the routing table twice, this seems a bit weird. If the above command doesn't fix it, try taking down the card (ifdown eth0) and bring it back up (ifup eth0) to see if this clears up those routes. Or you could use the route command, but I know that I try to avoid it :-) Nathan Callahan On Friday, June 1, 2001, at 01:47 PM, Abiel Reinhart wrote: After reformatting my system and upgrading to Linux Mandrake 8.0 from 7.2, I am unable to get IP masquerading to function. I was able to get it working with 7.2 and with Redhat 7.0 and earlier, with the same hardware configuration and client configuration I am using now. I am using kernel 2.2.19 (my modem driver does not function with the 2.4.x series), with all masquerading related options enabled. I am using a ppp modem connection, with a dynamically assigned IP. Linux router: 192.168.0.1 Windows 2000 client: 192.168.0.2 (worked with Mandrake 7.2, so already configured.) netstat -rn: Kernel IP routing table Destination Gateway Genmask Flags MSS Window irtt Iface 12.7.120.2510.0.0.0 255.255.255.255 UH0 0 0 ppp0 192.168.0.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0 192.168.0.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0 127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0 lo 0.0.0.0 12.7.120.2510.0.0.0 UG0 0 0 ppp0 ipchains -nL: Chain input (policy ACCEPT): Chain forward (policy DENY): target prot opt sourcedestination ports MASQ all -- 192.168.0.0/24 0.0.0.0/0 n/a Chain output (policy ACCEPT): ifconfig: eth0 Link encap:Ethernet HWaddr 00:20:78:10:1D:D6 inet addr:192.168.0.1 Bcast:192.168.0.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:9 errors:0 dropped:0 overruns:0 frame:0 TX packets:4 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:1010 (1010.0 b) TX bytes:264 (264.0 b) Interrupt:5 Base address:0xe000 loLink encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 UP LOOPBACK RUNNING MTU:3924 Metric:1 RX packets:44 errors:0 dropped:0 overruns:0 frame:0 TX packets:44 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:3248 (3.1 Kb) TX bytes:3248 (3.1 Kb) ppp0 Link encap:Point-to-Point Protocol inet addr:12.7.121.89 P-t-P:12.7.120.251 Mask:255.255.255.255 UP POINTOPOINT RUNNING NOARP MULTICAST MTU:1518 Metric:1 RX packets:213 errors:0 dropped:0 overruns:0 frame:0 TX packets:214 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:10 RX bytes:115513 (112.8 Kb) TX bytes:24652 (24.0 Kb) Local network connectivity is operating correctly (I can ping both ways). Tcpdump on the router shows incoming activity on eth0 when I try to access the Internet from 192.168.0.2, but no outgoing packets on device ppp0. I am unable to ping my ppp gateway (12.7.120.251). Any help is greatly appreciated. Thank you. Abiel Reinhart [EMAIL PROTECTED]