Re: ?? Lilo, Win NT, & MBR problems: a catch 22
Pending further investigation, we now allege that [EMAIL PROTECTED] wrote: > At home I have a laptop where win98 and linux co-exist, with lilo offering > to boot either one at startup. But at office, after linux installs lilo to > MBR, > NT won't boot. So, I went to boot floppy, ran fdisk /mbr; but now > machine won't boot linux. Boot floppy won't work for linux either. > > Reinstalled NT from scratch,and it is now the default boot. Since > can't boot linux from floppy, I am now reinstalling Potato. But, I > get stuck at ? where to put lilo. The /boot thing doesn't work for > me. I can boot with shift key down, but then computer hangs and > won't accept a number like 2. > > Shouldn't complain, I guess, but I lost a lot of time today because > of this problem. Newbie always appreciates help. TIA,,,dave NT can boot from lilo, just set it up exactly as you would win98. You can also boot linux from ntloader; there's a howto that explains how to do this.
Re: vpn -> how?
Pending further investigation, we now allege that will trillich wrote: > anybody care to suggest a few simple steps to get VPN > up and running? i've looked at the nano-HOWTO and don't > have any luck. other methods are delightfully accepted! > > rather than tell you how many methods and in what glorious > ways i've failed, i thought i'd start with tabula rasa and > let you experienced gurus start me off... > > [care to write a newbieDoc about it, while you're at it?] > I'm not sure if this is the same as the nan-HOWTO to which you refer, but there is a VPN howto (http://www.linuxdoc.org/HOWTO/VPN-HOWTO.html) which outlines a technique that involves running ppp over ssh. I do almost exactly the same thing to connect my laptop at work to my server at home, and it works pretty well. If you're looking for something to connect two offices over the Internet, it might not be the right choice. In that case you can check the links in that howto for info on CIPE and IPSec.
Re: lilo
Pending further investigation, we now allege that Stefaans Mostert wrote: > Hi all > > Well I was wondering can one transfer lilo onto a stiffy and boot from > that? That sounds a bit kinky. You might want to ask on debian-adult. > > Scenario: > You reinstall windows (wich is on the same drive as linux) and it > overwrites your mbr.After wich you suddenly and to your horror discover > that the dog lying in the corner looks so content beacause he has just > devouered you bootdisk!! > > What now? > I realize you will need lilo.conf in there somewhere to specify your > root partition. > lilo.conf is stored in /etc. Any bootable disk that can mount that partition can get to that file. If you have the Debian CDs, you can boot from the first one. When the setup starts, switch to the second terminal and you have access to a console. From there you can mount paritions on your hard drive, run lilo (using chroot or the -r switch), etc. Linuxcare also makes a bootable CD that is even better; you can download an image from http://open-projects.linuxcare.com/BBC/ (~50 MB I think) and burn a couple copies.
Re: question regarding samba usage
Pending further investigation, we now allege that Ethan Benson wrote: > On Wed, Dec 20, 2000 at 12:01:51PM -0500, Andy Bastien wrote: > > [snip] > > NTLMv2 improves on NTLMv1 by going from an 56 bit MD4 hash to a 128 > > bit MD5 hash. NTLMv2 is very difficult to attack with a brute-force > > your still missing the point, NT still just sends this md5 hash (which > is think is still unsalted) over the network where its used _as_ the > password. that is the hash is simply compared with the hash stored in > the password database. so if you sniff the hash you can use it to > authenticate yourself against an NT server, even if you *don't* know > the actual password. (this only requires a slightly modified > smbclient utility) That only works if the password is less than eight characters. This because there are two session keys returned by the server. One is the nthash, and the other is half of the lmhash. If your password is no more than seven characters, the whole thing fits into that half of the lmhash and you can use it to authenticate in the future. The nthash does you very little good. Of course, you can disable lmhashes on the server and fix this problem. That's what I was referring to when I mentioned that registry key. Let's not forget that NFS can be subject to address spoofing. You can sniff telnet, POP3, HTTP, and FTP sessions to get plaintext usernames and passwords. All of these exploits are at least as bad as the lmhash thing with SMB. There are other ways to get around the security that SMB provides. You can hijack sessions, or pretend to be a DC, for example. I would never say that it's perfect, but just because it is possible to crack it doesn't mean you should throw away the security that it does provide.
Re: question regarding samba usage
Pending further investigation, we now allege that Ethan Benson wrote: > > the smbpasswd crap comes from MS changing win95b, win98, NT4sp4, W2K > etc to send a unsalted password hash instead of the password to the > server, where the hash is compared with the hash stored in the local > password file, if the two hashes match the authentication suceeds. > its hardly better then sending clear text passwords over the network > since you can simply use the hash *as* the password. (to make it > worse the hash is rather weak anyway, especially since its unsalted > which makes it quite easy to brute force) in short your not really > losing any security by disabling MS's so called `encryption'. MS > didn't make this change for security purposes, they made it to break > samba. > This is not true. The original LANMAN authentication that was used by IBM and Microsoft in Lan Manager, and kept in Windows 95 by Microsoft was very weak. Microsoft added NTLMv1 in the original NT4. NTLM is much more secure than LM, because it creates a 56 bit MD4 hash based on the entire password and removes the whole 7 character issue in the original LM. The default on NT4 was to allow servers to accept LM requests. There's a registry setting that disables this, but since Windows 95 did not at that time support NTLM, hardly anyone used it. Later on, Microsoft added NTLMv2 in SP4, and they also included NTLM support for Windows 98 and Windows 95, so they made the default config to not accept LANMAN authentication. This had nothing to do with samba, and a whole lot to do with all of the bad press generated by l0phtcrack. NTLMv2 improves on NTLMv1 by going from an 56 bit MD4 hash to a 128 bit MD5 hash. NTLMv2 is very difficult to attack with a brute-force method. This is a good thing, and it bothers me when Linux advocates criticize Microsoft for increasing the security of their products (especially when there are so many perfectly valid reasons to criticize Microsoft). If you disable encryption, you lose a great deal of security. I'd like to see evidence of you cracking an NTLMv2 password sent over the wire that consists of at least 8 characters with a mix of upper case, lower case, digits, and punctuation. You can get L0phtcrack at http://www.l0pht.com/l0phtcrack/.
Re: Devices for SCSI Tape Drives
Pending further investigation, we now allege that Victor R. Cain wrote: > Could some one tell me where to find descriptions of what > the different device designations mean, especially the > /dev/nst* and /dev/st* devices? > /usr/src/linux/Documentation/devices.txt The /dev/nst* devices are the non-rewinding versions for the /dev/st* devices.
Re: kernel panic
Pending further investigation, we now allege that q wrote: > debs, > > in one of my dual-boot deskboxes, i have a kernel panic: > > unable to load NLS charset cp 437 > kernel panic: VFS: unable to mount root fs on 03:03 > > my potato on custom kernel 2.2.17 was functioning well until yesterday. > > i don't mind a re-install, but i'd really like to know what's going on, as > this happens like every couple of months to one of my machines. > > (i know i have a line-wrap problem w/ vi. when i get some time to learn how > to fix it, i'll be very happy. until then, my apologies.) > > anyway, when i use the rescue disk and mount a partition, i only get the > choices of /dev/hda1 (wincrap) /dev/hda3 and /dev/hda4. i'm sure i had > /dev/hda2 (linux) and /dev/hda5 (swap). and when i try to mount > /dev/hda3-hda4, i get, "invalid argument." > > i think i'm screwed; but if anyone has suggestions, please cc me. > > ia, t. > > bentley taylor. > Is /dev/hda3 is a FAT volume?
Re: partitionmagic
Pending further investigation, we now allege that Marcelo Chiapparini wrote: > Hi! > I am curious about how does Partition Magic work under Debian ext2 > filesystem. > It works very well, in my experience. PM doesn't work _in_ linux, but there is a "linux install" option that creates a bootable DOS floppy. Sort of cheesy, but I suppose it was the only way for them to support linux-only users without spending a large amount of cash on porting the entire program. parted is a [GPLed|Open Source] program that accomplishes the same sort of tasks. Of course, you should backup all of your data before repartioning, etc., etc.
Re: Network Monitoring recommendation
Pending further investigation, we now allege that sena wrote: > On 14/12/2000 at 17:26 +, Craig Coles wrote: > > I am looking for a tool to monitor the through-put of my connection to the > > internet. I want to know if my connection is maxing out, or my users just > > have slow connections to the internet. Is there a debian package to do > > this, or a recommendation otherwise? Can these tools determine what kind > > of > > traffic it is (http,ftp,etc.)? > > > > I'm using apache on woody, I know apache does some statistics with > > server-status but seems to only show averages over uptime. > > > And do you know of any program that also (or just) measures the quantity of > data (er.. you know what I mean :)) that passes through an interface? > > I think ipchains does that, but I think it just counts packets... Is there > any way of counting GB/MB/whatever? > iptraf, also check out the netdiag package.
Re: potato and netsaint
Pending further investigation, we now allege that Andy Bastien wrote: > > I'd like to install netsaint on a potato box. Unfortunately, the most > recent version requires glibc 2.2. Is is reasonably possible to > upgrade potato to glibc 2.2? This is a production box, (mainly DNS, > but it does one or two other things as well), so I can't screw it up > ;-). Should I just try to live with an older version of netsaint? Do > I have any other options? > Never mind. I had failed to remove the partially installed package from woody, so I was executing the wrong one. I've already banged my head against the wall a couple times. Still, I supose the question still stands, since this problem has happened before, as to whether you can or should upgrade a potato box to glibc 2.2.
potato and netsaint
I'd like to install netsaint on a potato box. Unfortunately, the most recent version requires glibc 2.2. Is is reasonably possible to upgrade potato to glibc 2.2? This is a production box, (mainly DNS, but it does one or two other things as well), so I can't screw it up ;-). Should I just try to live with an older version of netsaint? Do I have any other options?
Re: vnc problem: unable to connect to vnc server
Pending further investigation, we now allege that Giulio Morgan wrote: > > I am trying, unsuccessfully, to make a VNC connection from Windows98 (office > of > my client) to my woody machine. Perhaps I am misunderstanding something about > the usage of VNC, and if anyone could help, I would really appreciate it. > > I start the vncserver by > $ vncserver :4, > then check for my password in ~/.vnc/passwd. I discover it to be > \111A\222\333P\444L\555, > > I then launch vncviewer from the windows machine, and am prompted for the > address I wish to establish a connection to, I type in "my dynamic ip > address:4", and I am then prompted for "session password", and type > \111A\222\333P\444L\555, > and I get a notification: VNC authentication failed! The password that is stored in ~/.vnc/passwd is encrypted. Use vncpasswd to set or change it.
Re: SCSI
Pending further investigation, we now allege that john gennard wrote: > I've not had anything to do with the Scsi interface. A little while > ago in anticipation of trying it, I picked up a PCI board made by > Iwill and called a 2930U+ plus a secondhand 2.1Gig IBM OEM > harddrive (also a S/H CDRom Drive which I shall leave for the time > being). > > There's plenty of literature with the Scsi Host Adaptor - little of > which I understand at the moment, but I live in hope. Will a Potato > installation recognize this board - I haven't yet recompiled the > 2.2.17 kernel for Scsi so don't know what will be involved. > > The harddrive although old seems in good condition. Again, a > problem as I can find nothing on IBM's website which helps with > the jumper settings. There are 12 of these:- > 1. Spare(50pin) > ID bit3 > (For Wide) > 2. ID bit 2 > 3. ID bit 1 > 4. ID bit 0 > 5. Disable Auto Spin > 6. SCSI term on > 7. Disable unit attn > 8. Spare > 9. Auto start delay > 10. Delay start 6/12 > 11. Disable parity > 12. LED out > How many jumpers do I need to set? And do I use 4 or 3 if > the drive initially will be the only device attached? If it's the only drive or the last drive on the cable, you'll need to set termination on. You probably don't want to set any of the ID bits so that the ID of the drive will be 0. (The ID bit jumpers set the device number to 2^x. Bit 1 makes the drive ID 2, bit 2 makes it 4, bits 0 and 2 makes it 5, etc.) The rest of the jumpers you probably want to leave off unless you have problems with the drive. > If I can get to understand Scsi and find it a viable alternative to > IDE, it was my intention to attach a CD writer and some new > harddrives externally (the board provides for this). Here again > I'm lost - some people I've spoken to insist only one device can > be attached externally whilst the literature clearly says up to 7. > Furthermore, here in the UK no-one appears to stock any 'boxes' > or whatever for more than one device. A stock explanation seems > to be "haven't seen any for years - there's no call for them" > (One 'charming' individual said he knows nothing technical > about computer parts "I only sell them". You can attach up to 7 devices to a SCSI-2 chain, and 15 to Ultra SCSI or better. Whether they are internal or external is irrelevant, although you do need to stay within the cable length limitations. These limitations vary, but, as a rule of thumb, 2m total cable length is usually fine. Seagate's website has some good information about SCSI specs if you want more detail. There are companies that sell external SCSI chassis over the web. I don't know any off-hand, but if search around you'll find them. There probably are some within the UK.
Re: 2.4.0-test11, PCMCIA ethernet card through yenta-socket
Pending further investigation, we now allege that Marvin Stodolsky wrote: > My PCMCIA "3Com Corporation", "3CCFE575BT", "LAN Cardbus Card", "001" > manfid: 0x0101, 0x5157 > function: 6 (network) > PCI id: 0x10b7, 0x5157 > is working fine under 2.2.17 and under the 2.4.0-test11 with the "old > way" PCMCIA support of compiling all the pcmcia modules, > # lsmod > Module Size Used by > 3c575_cb 20064 2 > cb_enabler 2736 2 [3c575_cb] > ds 6672 2 [cb_enabler] > i82365 23232 2 > pcmcia_core43392 0 [cb_enabler ds i82365] > > Thus far I've been unable to get eth0 functional with PCMCIA support > specified "the new way" through the kernel .config file, after which > support would presuably come from: > 3c59x > cb_enabler.o > pcmcia_core.o > yenta_socket.o It doesn't work. The 3c575 chipset is different than the 3c59x series. You might have luck with using the kernel-level support with the 3x575_cb.o module from pcmcia-cs, but I wouldn't lay money on it.
Re: NT4 in VMWare VM can't see Samba over network - Yes, PlainTextPassword set :)
Pending further investigation, we now allege that Robin Collins wrote: > Not really a Debian issue but maybe someone can help me out here. > > There are two computers here, for the sake of discussion called A and B. B > runs Samba 2.0.7a on Potato. A is multi-boot. > > When booted to native NT4 SP6a with PlainTextPassword set A can see B and > access B's shares perfectly. > > When booted to Linux, SMBCLIENT can see B's shares, use them and can also > NFS mount them. > > BUT, when I run NT4 SP6a in a VMWare VM the VM can see the host, can see a > Win98 system also on the network but can't see B nor map a share using NET > VIEW: attempts fail with an error 53, name not found. Network Neighbourhood > is also missing B, but the stange thing is that after a while (where "while" > seems indeterminate!) B appears in Network Neighbourhood but it's shares > cannot be seen. > > There are no errors logged by smbd or nmbd on B but the problem "feels" like > some sort of name resolution problem; though clearly not DNS since the VM > can ping B by name. PlainTextPassword is set since the VM can map drives > from the host's Samba. SMBCLIENT on B can get to the VM's shares so B in > able to "see" A's VM by name. > > I've installed many NT4/Samba systems in my time but I've never seen this, > anyone any ideas? > Is the VM a host-only or bridged network? Do you have WINS server support on B? If so, is the VM using it? My first thought on hearing this is that your VM is a B-node running on VMWare's virtual network, so it can't resolve NetBIOS names for machines on the real network. If this is what's happening, you'll either need to set up a WINS server somewhere (B would be ideal), or you'll have to create a lmhosts on the VM. Also, if the VM has a host-only network, you'll want to make sure its workgroup is different than B's. btw, do you have any shares on A, and if so can the VM see them?
Re: tnt2 vanta and dell 21"
There are those who would have you believe that Anders Lennartsson wrote: > It seems Andy responded to a question by Anders Lennartsson: > > >There are those who would have you believe that Anders Lennartsson wrote: > >> Hi > >> > >> I'm trying to configure X on a PII 300 Mhz dual proc with a Creative > >> TNT2 Vanta PCI card and running Woody. > snip > >> The problem is for instance that if I click on the desktop the menu > >> (fvwm menu) comes up with an offset to the right > >> and is unstable and has stripes. The system however thinks it is located > >> close to the mouse pointer so there is some > >> guessing to choose the correct line etc. Bigger windows case even more > >> trouble. > snip > > Further, I have since discovered that the file /var/log/xdm.log contains > the following > lines > > snip > > (II) NV(0): Supported VESA Video Modes: > (II) NV(0): [EMAIL PROTECTED] > (II) NV(0): [EMAIL PROTECTED] > (II) NV(0): [EMAIL PROTECTED] > (II) NV(0): [EMAIL PROTECTED] > (II) NV(0): [EMAIL PROTECTED] > (II) NV(0): [EMAIL PROTECTED] > (II) NV(0): [EMAIL PROTECTED] > (II) NV(0): [EMAIL PROTECTED] > (II) NV(0): Manufacturer's mask: 0 > (II) NV(0): Supported Future Video Modes: > (II) NV(0): #0: hsize: 640 vsize 480 refresh: 85 vid: 22833 > (II) NV(0): #1: hsize: 800 vsize 600 refresh: 85 vid: 22853 > (II) NV(0): #2: hsize: 1024 vsize 768 refresh: 85 vid: 22881 > (II) NV(0): #3: hsize: 1280 vsize 1024 refresh: 85 vid: 39297 > (II) NV(0): #4: hsize: 1600 vsize 1200 refresh: 75 vid: 20393 > (II) NV(0): #5: hsize: 1600 vsize 1200 refresh: 85 vid: 22953 > > snip > > Where the last section indicates that the 1600x1200 mode is not yet > supported. > I guess this closes my question. > I think you're right...I just noticed this in the readme while looking for something else: 6.5.6 Corruption at very high resolutions Late in the process of testing our driver, we found that it is possible to see some image corruption if your physical screen size is set to 1600x1200. We will fix this in the very near term. There also seem to be many people with the Vanta boards complaining about NVidia's drivers. They're a little different from the other TNT2-based cards, so there might be something in those differences that the driver can't handle.
Re: [debian-user] GeForce2 w/ XFree86 4.0.1 [ a little new ]
There are those who would have you believe that Rüdiger Kuhlmann wrote: > > Hi! > > >--[Andy Bastien]--<[EMAIL PROTECTED]> > > >--[Rüdiger Kuhlmann] > > > I have a GeForce2 MX, but I still don't get 3D running. At first, kdm > > > crashes unless there is also an libGL.so.1 pointing to the _old_ libGL.so. > > You might have a conflict among your libraries somewhere. Go through > > all of the instructions that came with the drivers again and verify > > that you've done everything, especially the steps that explain which > > libraries you need to remove. Compare dates and file sizes with the > > drivers you compiled. > > I fiddled around with it by hand after it didn't work. This way, 2D, but not > 3D, works. If I let libGL.so.1 point to libGL.so.1.0.5.nvidia as well, kdm > crashes on startup. Can someone check filesizes with this list? That's interesting...can you try it with Gnome or only with a window manager? (I'd think twm would be the best bet to get rid of possible interactions with the window manager.) > --snip-- > [EMAIL PROTECTED]:/usr/lib> ll *GL* *glx* > ls: *glx*: No such file or directory > -rw-r--r--1 root root 519926 Nov 2 18:44 libGL.a.debian > lrwxrwxrwx1 root root 21 Nov 6 19:30 libGL.so -> > libGL.so.1.0.5.nvidia > -rwxr-xr-x2 root root 19 Nov 2 18:44 libGL.so.1 -> > libGL.so.1.2.debian > -rwxr-xr-x1 root root 184112 Oct 23 22:30 libGL.so.1.0.5.nvidia > -rwxr-xr-x2 root root 349804 Nov 2 18:44 libGL.so.1.2.debian > lrwxrwxrwx1 root root 13 Nov 5 13:44 libGLU.so -> > libGLU.so.1.3 > lrwxrwxrwx1 root root 13 Nov 2 08:39 libGLU.so.1 -> > libGLU.so.1.3 > -rwxr-xr-x1 root root78412 Oct 18 15:59 libGLU.so.1.1.030201 > -rwxr-xr-x1 root root 487348 Nov 2 18:44 libGLU.so.1.3 > lrwxrwxrwx1 root root 25 Nov 6 19:22 libGLcore.so -> > libGLcore.so.1.0.5.nvidia > lrwxrwxrwx1 root root 25 Nov 6 19:22 libGLcore.so.1 -> > libGLcore.so.1.0.5.nvidia > lrwxrwxrwx1 root root 25 Nov 6 19:22 libGLcore.so.1.0.5 -> > libGLcore.so.1.0.5.nvidia > -rwxr-xr-x1 root root 1499464 Oct 23 22:30 > libGLcore.so.1.0.5.nvidia > [EMAIL PROTECTED]:/usr/lib> cd /usr/X11R6/lib/modules/extensions/ > [EMAIL PROTECTED]:/usr/X11R6/lib/modules/extensions> ll *GL* *glx* > -rw-r--r--1 root root 1821046 Nov 2 18:44 libGLcore.a.debian > -rw-r--r--1 root root 259916 Nov 2 18:44 libglx.a.debian > lrwxrwxrwx1 root root 22 Nov 6 19:32 libglx.so -> > libglx.so.1.0.5.nvidia > lrwxrwxrwx1 root root 22 Nov 6 19:32 libglx.so.1.0.5 -> > libglx.so.1.0.5.nvidia > -rwxr-xr-x1 root root 450924 Oct 23 22:32 libglx.so.1.0.5.nvidia > [EMAIL PROTECTED]:/usr/X11R6/lib/modules/extensions> > --snip-- Here's listings from my desktop machine: elder:/usr/lib$ls -l *GL* *glx* ls: *glx*: No such file or directory -rw-r--r--1 root root 1514844 Aug 2 00:15 libGL.a lrwxrwxrwx1 root root 18 Oct 29 08:56 libGLcore.so.1 -> libGLcore.so.1.0.5 -rwxr-xr-x1 root root 1499464 Oct 29 10:53 libGLcore.so.1.0.5 lrwxrwxrwx1 root root 10 Nov 5 19:13 libGL.so -> libGL.so.1 lrwxrwxrwx1 root root 14 Oct 29 10:56 libGL.so.1 -> libGL.so.1.0.5 -rwxr-xr-x1 root root 184112 Oct 29 10:53 libGL.so.1.0.5 -rw-r--r--1 root root88786 Aug 2 00:15 libGLU.a lrwxrwxrwx1 root root 11 Nov 5 19:13 libGLU.so -> libGLU.so.1 lrwxrwxrwx1 root root 20 Oct 29 00:12 libGLU.so.1 -> libGLU.so.1.1.030201 -rw-r--r--1 root root78412 Aug 2 00:15 libGLU.so.1.1.030201 lrwxrwxrwx1 root root 10 Oct 29 00:12 libMesaGL.so.3 -> libGL.so.1 lrwxrwxrwx1 root root 11 Oct 29 00:12 libMesaGLU.so.3 -> libGLU.so.1 elder:/usr/X11R6/lib/modules/extensions$ls -l *GL* *glx* -rw-r--r--1 root root 1821046 Nov 2 12:44 libGLcore.a -rw-r--r--1 root root 259916 Nov 2 12:44 libglx.a lrwxrwxrwx1 root root 15 Oct 29 10:53 libglx.so -> libglx.so.1.0.5 -rwxr-xr-x1 root root 450924 Oct 29 10:53 libglx.so.1.0.5 That libGL.a is from mesag-dev. I also don't have a libGLU.so.1.3. I have mesag-dev 3.2.1-1 and mesag3 3.2.1-1. You don't have mesag3-glide, do you? When I installed qt2.2, dselect really wanted to install the glide libraries and I had to change several of it's default selections to prevent that. > > > Then, when starting X, the screen is
Re: [debian-user] GeForce2 w/ XFree86 4.0.1 [ a little new ]
There are those who would have you believe that Rüdiger Kuhlmann wrote: > > Hi! > > >--[Jonathan Wheelhouse]--<[EMAIL PROTECTED]> > > On Sat, Nov 04, 2000 at 11:01:46PM -0800, Cisco.Addict wrote: > > > Sorry, is there any way i can get my GeForce2 GTS 64mb working in XFree86 > > > 4.0.1 > > > (Im hopeing to be able to play Quake3) > > I've got a GeForce2 GTS 32mb; I followed tomshardware instructions (+ > > others) to install and now I can play the quake 3 arena demo (looks > > great but is really gory). > > I have a GeForce2 MX, but I still don't get 3D running. At first, kdm > crashes unless there is also an libGL.so.1 pointing to the _old_ libGL.so. You might have a conflict among your libraries somewhere. Go through all of the instructions that came with the drivers again and verify that you've done everything, especially the steps that explain which libraries you need to remove. Compare dates and file sizes with the drivers you compiled. > Then, when starting X, the screen is black for ~8 secs with some blue > flickers, before it switches to the right display mode (you can see an "old" > image there for a short time). That's strange, but doesn't really hurt. That's (more or less) normal. > But when I play anything with eg realplayer, it isn't accelerated - the CPU > usage goes up endlessly, and it's still slow (I'M just playing the > realplayer credit intro, so nothing too complicated). You could also try some OpenGL screensavers, or go to http://www.linuxgames.org, > I don't get any error messages in XF*log (except the one you noted, but DRI > is a driver architecture not supporting nvidia, so I guess it will stay > unused if you load it). Yes I do have the NVdriver compiled, although I got > some warnings (I don't remember too precisely right now). I don't think you're supposed to get any warnings. You could try compiling the drivers again and posting the warnings. > > But doesn't seem to hurt anything. quake3 is really quick; looks good > > so I assume 3d performance is OK. > > Okay, are there any .debs for Quake3, or where do I get it? http://www.lokigames.com. I'd recommend buying from a store, even if it costs a few dollars more. That way, you not only support the development of games on Linux, you also help justify shelf space for Linux software.
Re: tnt2 vanta and dell 21"
There are those who would have you believe that Anders Lennartsson wrote: > Hi > > I'm trying to configure X on a PII 300 Mhz dual proc with a Creative > TNT2 Vanta PCI card and running Woody. > > X works in 1280x1024, but both xfree 3.3.6 xserver-svga and xfree 4.0.1 > xserver-xfree86 > have problem with stable picture in 1600x1200. The card should be able > to do this with 85Hz vertical > according to the documentation. In fact it is only 220Mhz pixelrate of > the possible pixelrate of 350MHz. > Even my old Millenium II can take the Dell 21" D1626HT to 1600x1200 but > with less colors. > The problem is for instance that if I click on the desktop the menu > (fvwm menu) comes up with an offset to the right > and is unstable and has stripes. The system however thinks it is located > close to the mouse pointer so there is some > guessing to choose the correct line etc. Bigger windows case even more > trouble. > > Anyone got some input or experience of such a combination? What is the > quality of the xfree86 drivers for this card? > Should one try to compile the drivers from Nvidia's ftp site? > Does the picture wrap around one side of the display, or is it correctly centered? Is it 'wiggly'? Does it have banding (horizontal, vertical or diagonal stripes of varying colors)? Is there a bright stripe on either side of the display? Is it rectangular with perpendicular sides, or does it bow in or out on the sides or appear as a trapezoid or other non-rectangular quadrilateral (i.e. a 'leaning' rectangle)?
Re: soundcore module
There are those who would have you believe that robert_wilhelm_land wrote: > What is the soundlow & the soundcore used for? > > I'm just trying to get my Sonic Impact to work (which should be a OSS > card) and cannot find any of the below modules in 'make xconfig'. The > present deb 2.2 has both modules loaded but I have no idea what they > are used for. The installation routine (by booting from CD#1) had as > far as I can remember a various list of sound options which I'm not > able to find in the xconfig app. > > > Would someone help to clarify? > > Partly out of 'lsmod' (current installation): > > sound 57592 0 (unused) > soundlow 416 0 [sound] > soundcore 2628 3 [sound] > Pick "Sound card support" to get soundcore.o and "OSS sound modules" for sound.o (both as modules). I'm not sure about soundlow.o, but I believe you get that if you go into the "Additional low level sound drivers". You can probably find more details under Documentation/sound.
Re: Sound Card: Turtle Beach Montego A3D 64 Voice PCI
There are those who would have you believe that [EMAIL PROTECTED] wrote: > > > Dear community, > > I' ve a Sound Card: Turtle Beach Montego A3D 64 Voice PCI (with a DELL > computer). The Turtle Beach device drivers proposed by Debian 2.2 are for > other > types of cards and don't work. Chipset is Aureal AU8820 and AC'97 Codec > according the documentation. > Perhaps someone can shed some light on what device driver to actually use. > Montegos are based on the Aureal Vortex chipset. There used to be a driver page at http://linux.aureal.com, but it seems to be down. Try http://www.vortexofsound.com. You could have a problem with it being a Dell card, as Dell (other companies do too, I'm not picking on Dell here) sometimes uses not-quite-standard versions of retail components.
Re: sticky bit, powersaving & hdd spindown
There are those who would have you believe that Krzys Majewski wrote: > Damien <[EMAIL PROTECTED]> writes: > > > i've been working on a stand alone music player for the last couple of > > months. > > it's currently a bit loud (being based around an old p166 with a very noisy > > hdd & powersupply fan. > > I can't help wondering if you could do away with the hdd altogether, > e.g. booting the thing once and for all from floppies. The P/S fan > solution I've posted before -- if it's an old and wimpy machine and > you're daring you could maybe try even more extreme things like nuking > the fan altogether. (Check your household fire insurance?) Along the > same lines, you could try disabling the CPU fan as well if you have > one. I ran my PIII-500E for a few hours with the CPU fan unplugged and > it didn't even blink. It seems to me that there is a bit of room for > trial and error here, especially if you can afford it: if the CPU > overheats, the machine will crash. Whether this damages the CPU > permanently I don't know, any electronics types here that know > anything about this? > This is a good idea about just tossing the hard drive. You could also write a bootable CDR and just use that. I would be very careful about running a CPU without a fan. IIRC, you said you were using a Pentium 166. With a good heatsink and enough airflow past the heatsink you could do without a fan. Newer CPUs cannot run without a fan; AMD specifically recommends that you never power on an Athlon without a heatsink and fan, not even for a couple of seconds just to make sure it works.
Re: New X Server crashes
There are those who would have you believe that Michael Abraham Shulman wrote: > I wrote: > > MAS> Okay, it appears to be working now. > > Well, more or less. Everything runs fine with `startx', but I'm > having some peculiar difficulties with xdm. Specifically, > > 1. xdm now takes over a minute from when the daemon is started, either >by init or manually, to when the login widget appears. Is this >normal now, or is there anything I can do to speed it up? I don't know. Maybe it's trying to resolve your hostname? > > 2. Does anyone have any clue what could cause an xdm-started X server >to use a different video mode than an xinit-started X server? >Because that's what it's doing. I copied a 1152x864 mode line from >my old XF86Config file, and it works fine with startx, but when run >from xdm, the X server reverts to 1024x768. startx looks in the appropriate Screen section of your XF86Config for the DefaultDepth, then uses the first entry of the Mode parameter for the Display subsection that matches that color depth (I believe if DefaultDepth is not specified, it just uses the first Display). xdm uses the entries in /etc/X11/xdm/Xservers; if it has a different setting for the color depth it could affect which resolution you end up with.
Re: aside: perl's unpack()
There are those who would have you believe that Damien wrote:
> > > I can tell you an easier way to decipher them, but I too would like to
> > > know how they are contructed, i.e what's the algorithm.
> >
> > a.b.c.d
> >
> > n = d + c * 256 + b * 256 ^ 2 + a * 256 ^ 3
>
> after reading the previous mail, this is what i was thinking. i went to test
> it in perl afterwards, but got the wrong result. could anyone tell me what's
> wrong with my interpretation?
>
> perl -le '$, = "."; print unpack("C4", "2704935062");'
>
IP networks use big-endian notation and PCs use little-endian; you
need to convert the value before splitting it back into octets:
($a, $b, $c, $d) = split(/\./, "50.55.48.52");
$n = $d + ($c << 8) + ($b << 16) + ($a << 24);
print "$n = ", join(".", unpack("C4", pack("N", $n))), "\n";
I'll leave it as an exercise to make this into a one-liner ;-).
Re: Upgrading video card: Best of these three?
There are those who would have you believe that wulfie wrote: > I'll second the anti-Nvidia driver lobby. I too have had severe lockups > either starting XF86 4.0 or ctl-alt-Fx'ing to a text console. My specific > hardware is a 16Mb Sparkle Nvidia Vanta AGP 2x card, this on my main SuSE > (sorry!) box. > Are you sure it's the drivers? Have you verified that your motherboard fully follows the AGP spec? When the TNT2 cards first came out, there were many people who complained about instability with them. Many, if not most, were using them on motherboards that did not have a correct AGP implementation. Some motherboards have AGP settings that can be used to improve performance or stablity. You should check your motherboard manufacturer's (or computer vendor's) website and/or do a search on deja to find out if other people with your hardware have had problems and if they've found solutions.
Re: Upgrading video card: Best of these three?
There are those who would have you believe that Kristian Rink wrote: > > > Any other sugestions appreciated. I left off the NVIDIA 5or close to it) > > crds because the GeForce2 seem to prefer XF 4.0x. That is true. > > Basically, nvidia`s chipsets (at least for TNT2 I tried) also are >supported by XF 3.3.6 and utah-glx... Besides this, after trying out >for several times, I don't at all recommend running nvidia-based cards >in XF 4.x for the following reasons: > > (a) NVIDIAs drivers for those cards (even while being fast) are >obviously still *very* unreliable, which (at least with my TNT2 board) >caused several severe crashes of the x-server (especially while >running "gears" or other of those problems with -root option set, and >while trying to switch to the text console and back to X again). I've used NVidia's drives with both TNT2 and GeForce 2 cards and haven't found this at all. In fact, on my desktop system at home I have more problems with the Win2000 drivers than the Linux ones. Your problems could be a hardware issue; there are many older (i.e. more than 18 months or so) motherboards that don't supply enough power to the AGP port and this can cause problems with NVidia's cards. However, it is true that they are beta drivers and if you require 100% stability you shouldn't use them (of course, you shouldn't be playing Quake either ;-). > (b) besides that, those drivers are binary-only and they seem to >have some problems with certain 2.4.0-testx kernels (don`t know if >they fixed by now). Installation seems to be difficult sometimes due >to unresolved conflicts with installed MESA-version while using those >drivers on systems which are not RedHat-based (because nvidia only are >offering .rpm - packages and `generic` archives as .tar.gz ...). The binary module is a problem and does force users to rely on NVidia keeping up to date on kernel developments. Of course both the drivers and the 2.4 kernel are in beta, and combining beta products often seems to multiply problems. Once again, if stability is important, then don't use 2.4. I haven't found installation difficult. The readmes contain very specific and complete instructions for installing the drivers. The only problem I've run in to was a system that was missing a directory (I don't remember which) and the install scripts just assumed that it existed. The error messages made it quite obvious what went wrong and it was easy enough to fix. > (c) nvidia still is not very cooperative in providing the developers >of `open` driver systems with information, that's why utah-glx is >still very much slower than their native drivers because it`s not >possible for those to use DMA on that card. This is why I am tending >to ask people to please DO NOT SUPPORT companies like nvidia and their >attempt of bringing proprietary drivers to Linux... Thank You! This is a good point. However, since they make good hardware I'm willing to give them the benefit of the doubt for now that their current open-source efforts are in transition and that they are working in good faith to resolve any licensing issues or other problems that are preventing them from releasing fully open drivers. A year from now I'll probably feel differently. The truth is that right now you don't have many options. ATI cards don't even compare in terms of 3D performance (with the possible exception of the Radeon, but IIRC it's not supported by the Linux drivers yet), and Matrox' and 3dfx' cards will end up costing much more for a similar level (and ATI is on my sh!t list until I can get hardware OpenGL support on my laptop; llxdoom is OK, but I want Quake 3, dammit ;-). Without good drivers, there will be very few games. Without games, the number of Linux users will be limited, and there will never be a reason for video hardware manufacturers to go out of their way to support open source or free software.
Re: sawfish heads-up
There are those who would have you believe that Colin Watson wrote: > Andy Bastien <[EMAIL PROTECTED]> wrote: > >There are those who would have you believe that Colin Watson wrote: > >> If you upgrade sawfish to the one in unstable at the moment, don't > >> follow the instruction that says: > >> > >> execute 'kill -9 `pidof sawfish`;sawfish&' in a terminal > >> > >> I did so without thinking, and of course lost all my state. I don't see > >> any reason not to use the old 'sawfish-client -f restart', apart from > >> the cosmetic problem of the spurious error message, but no doubt I'll > >> find out as a result of the bug report I've just filed. 'kill -1' might > >> work, but I've no intention of blowing up my X session again just to > >> find out ... > > > >What do you mean when you say you lost all of your state? FWIW, I did > >this on my laptop and didn't have any problems. > > Maybe you're running something like gnome-session as your session > manager (the last thing in your .xsession). If you're like me and don't > use full-blown GNOME, the last thing your .xsession does is often to run > your window manager, and thus being advised to kill it is bad news. > > Looking before I leaped would have helped, I admit :) > You're right, that's exactly what I do. I just wanted to be sure I understood what you were saying before I do this on any other machines ;-). How about this instead: pidof gnome-session && kill -9 `pidof sawfish`;sawfish&
Re: sawfish heads-up
There are those who would have you believe that Colin Watson wrote: > If you upgrade sawfish to the one in unstable at the moment, don't > follow the instruction that says: > > execute 'kill -9 `pidof sawfish`;sawfish&' in a terminal > > I did so without thinking, and of course lost all my state. I don't see > any reason not to use the old 'sawfish-client -f restart', apart from > the cosmetic problem of the spurious error message, but no doubt I'll > find out as a result of the bug report I've just filed. 'kill -1' might > work, but I've no intention of blowing up my X session again just to > find out ... > What do you mean when you say you lost all of your state? FWIW, I did this on my laptop and didn't have any problems.
Re: En snabb fr?ga...
There are those who would have you believe that V.Micic wrote: > Var kan jag ladda hem den senaste versionen av Debian i ISO format,jag har > letat men inte hittat... > Tack i Förhand I am the only one who tried to rot13 this?
Re: GeForce2 GTS & X 4.0.1c
There are those who would have you believe that Jonathan Wheelhouse wrote: > Recently I installed Branden's debs for X 4.0.1c (deb > http://samosa.debian.org/~branden/ woody/$(ARCH)/) because, frankly, > vga16 @ 640 x 480 sucks with a GeForce2 GTS card. > > So, now, woohoo, the desktop looks great and I don't spend much time > in the console. > > But my question is - I thought that I had to grab the Linux drivers > from the NVIDIA site to make things work; it seems I don't. > > /var/log/gdm says the NV driver supports the GeForce2 GTS. > > So, is it that for 3d I need the NVIDIA drivers? > Yes, for accelerated GL support. > And what are some good apps for checking frame rates and also showing > off the power of the card? > Quake 3, Unreal Tournament
Unidentified subject!
To : debian-user@lists.debian.org Bcc: Subject: Re: 2.2: MBR install whacks my system boot Reply-To: In-Reply-To: <[EMAIL PROTECTED]>; from [EMAIL PROTECTED] on Thu, Oct 26, 2000 at 05:47:56PM -0400 X-Warning: Vs lbh pna ernq guvf, lbh'er n trrx. X-Wisdom: Nobody gets out of the Bermuda Triangle. Not even for lunch. There are those who would have you believe that Paul D. Smith wrote: > %% Andy Bastien <[EMAIL PROTECTED]> writes: > > ab> In any case, I'd set sda1 active and then set the value for root in > ab> lilo.conf to /dev/sda and then go from there. > > I set sda1 active and changed the boot to /dev/sda in /etc/lilo.conf, > and re-ran lilo, and now it works; thanks. > > Seems like the Debian install picked the wrong thing to do by default in > this case? > > S... anyone know why my apt-get install won't work (see my other > message)? :) > I glad to hear it works. I'm also glad you caught my error ;-).
Re: I/O to ZIP drive temporarily locks up machine
There are those who would have you believe that Krzys Majewski wrote: > Krzys Majewski <[EMAIL PROTECTED]> writes: > > > Hm, I compiled parallel port support as a module. I already had > > > > options parport_pc io=0x3bc irq=7 > > > > in my /etc/modutils/options, so I appended "dma=3" to give > > > > options parport_pc io=0x3bc irq=7 dma=3 > > > > Now I can modprobe parport_pc, but when I try to modprobe imm I get: > > > > 17:52:05<~>$ sudo modprobe imm > > /lib/modules/2.2.17/scsi/imm.o: init_module: Device or resource busy > > Hint: insmod errors can be caused by incorrect module parameters, including > > invalid IO or IRQ parameters > > /lib/modules/2.2.17/scsi/imm.o: insmod /lib/modules/2.2.17/scsi/imm.o failed > > /lib/modules/2.2.17/scsi/imm.o: insmod imm failed > > 17:52:09<~>$ > > > > I don't see any imm options in my /etc/modutils. How do I find out > > what parameters the imm module takes? Does it need any parameters? > > What's going on? -chris > > > > OK, looks like I got things to work by taking out the io=0x3bc > parameter in the parport_pc options. But, I'm doing a backup now and > it's still doing weird things to my machine. For example, I'm playing > a streaming mp3 and it's very choppy. I've tried both the > "ECP" and "ECP+EPP" options in the BIOS. > Your earlier post (where you included the output from /proc/parport/0/hardware) indicated that the port was using 0x378 for the io port, so that might be why removing that parameter fixed the problem with the module not loading. The values have to agree with what your hardware is expecting, so you might have to set the parallel port's DMA to be 3 in your CMOS settings. Have you tried using the ppa module instead of imm? I'm not sure what the difference is, except that ppa is for "older" zip drives, and imm for the "newer" ones. You might get better results. I found this in the README.ppa from the kernel source, maybe it'll help: Terse where to get ZIP Drive help info General Iomega ZIP drive page for Linux: http://www.torque.net/~campbell/ Driver achive for old drivers: http://www.torque.net/~campbell/ppa/ Linux Parport page (parallel port) http://www.torque.net/parport/ Email list for Linux Parport [EMAIL PROTECTED] Email for problems with ZIP or ZIP Plus drivers [EMAIL PROTECTED]
Re: I/O to ZIP drive temporarily locks up machine
There are those who would have you believe that Krzys Majewski wrote: > On 20 Oct 2000, Krzys Majewski wrote: > > I don't know how to say this properly.. when backing up files from my > > IDE hard drive to my parallel port ZIP100 drive, the machine > > occasionally blocks: it's like everything freezes for about half a > > minute before I can do anything again. This only happens when the ZIP > > drive is running (not sure if it's R or W or both). This is with > > kernel 2.2.17, using the imm driver as a module. I've set all the > > default "safe" options for the driver, like "slow and reliable" or > > whatnot. Is this a kernel bug? IMM driver bug? User bug? Feature? > > Help? chris > > > > I got a great response to this from someone who suggested enabling ECP > in the BIOS and assigning a DMA. I did the first (BIOS now says: > "ECP+EPP > ECP Mode use DMA 3 > EPP Mode Select EPP1.7") > but don't know how to do the second. Reading > /usr/src/linux/Documentation/parport.txt > didn't help. > > 11:24:02<~>$ cat /proc/parport/0/hardware > base: 0x378 > irq:none > dma:none > modes: SPP,ECP,ECPEPP,ECPPS2 > > Notice the lack of DMA. How to fix this? I'm getting lots of > errors on the ZIP drive now, e.g. > > localhost kernel: SCSI disk error : host 1 channel 0 id 6 lun 0 return code = > 2701 > localhost kernel: scsidisk I/O error: dev 08:04, sector 32345 > localhost kernel: file_cluster badly computed!!! 2248 <> 3784 > > and more. I can occasionally read/write the ZIP drive but basically it > is unuseable in this state. > In your modules.conf, put a line something like: options parport_pc irq=7 io=0x378 dma=3 but put the values that you are actually using (the io and irq are probably the same, the dma is probably set in your CMOS).
Re: 2.2: MBR install whacks my system boot
There are those who would have you believe that Paul D. Smith wrote: > I've installed Debian on a lot of systems, and this is the first time > it's ever done this. > > I'm installing 2.2 from a DOS partition on a 200MHz ppro, with 64M ram. > It has a 3c595 Vortex 10/100 autoselect enet card, and an Adaptec > AHA-294x Ultra SCSI card. > > The SCSI card has one disk (/dev/sda), which is a Seagate ST32430N 2G, > and one CDROM. There is also a floppy. > > That's it: no IDE disks, etc. > > The partition table for this system looks like this (from fdisk): > > /dev/sda1 126208813+ 6 FAT16 > /dev/sda2 *27 261 1887637+ 5 Extended > /dev/sda52746160618+ 83 Linux > /dev/sda647 142771088+ 83 Linux > /dev/sda7 143 155104391 82 Linux swap > /dev/sda8 156 261851413+ 83 Linux > > Don't ask me why there's only one primary partition and everything else > is in the extended partition; it was configured like this before I got > it. It had RedHat 6.1 on there and booted fine. The previous install > booted into Linux (/dev/sda5 is the root partition), with an optional > boot into DOS (/dev/sda1). > > When I installed Debian I had it overwrite the data in all the Linux > partitions. > > When it asked me about installing the MBR, I took all the defaults. > After it installed (or tried to install) an MBR for /dev/sda, it gave me > a screen about how the boot partition was on an extended partition, so > it needed to install into /dev/sda2, and I again took the default (OK). > > Now when I boot the system it prints "MBR", then does a hard hang. I > can't even C-A-D or use the reset button, I have to power off/on. > > If I hold the SHIFT key down, it prints "MBR 2AF", then does the same > hard hang. > > I can boot off of my boot floppy and it comes up OK, though. > > What should I do? I'd hate to have to reinstall, and it used to boot OK > so it must be possible with the above partition setup. > > /etc/lilo.conf says: > > lba32 > boot=/dev/sda2 > root=/dev/sda5 > install=/boot/boot.b > map=/boot/map > delay=20 > vga=normal > default=Linux > > image=/vmlinuz > label=Linux > read-only > > image=/vmlinuz.old > label=LinuxOLD > read-only > optional > > Help! > I'd bet that there is only one primary partition because that is all that DOS could use, so one primary and and an extended to fill the rest of the disk used to be a standard configuration. In any case, I'd set sda1 active and then set the value for root in lilo.conf to /dev/sda and then go from there. OTOH, I don't see any reason why you couldn't just repartition the drive how you'd like it to be, since there is apparently nothing that you need to save in the extended partition.
Re: domain change
There are those who would have you believe that Debian Ghost wrote: > Hey Guys, > I posted this question earlier in another way, but no one replied to it. > Maybe if I ask in another way: > > I changed the domain name of my machine (via DNS) and was wondering > if there was a global file that I need to change to tell my machine it is > > host.new.domain rather than host.old.domain. > > I changed the exim.conf file (sendmail) but when I telnet into my machine, > somehow exim still thinks it is the old name even though I changed the > conf file. Is there any global file that I need to edit to get rid of the > old.domain and tell the machine it is new.domain? > Maybe /etc/hosts or /etc/hostname ? btw, you aren't related to Space Ghost, are you?
Re: Linux Tough guy
There are those who would have you believe that Cavaiani, Don wrote: > You seem to be much smarter by far than 98% of the people on this list. I > have been trying for a couple of days - via the Debian list yesterday - to > find out how to make my linux part. bigger w/o losing everything. I'm sure > the many e-mails must 'bout drive you batty some days! > > I tried to find a FIPS expert via the web, but he's probably no longer > there, so I forwarded it directly to you for one last try. > I don't think I'm the one you meant, but I know Partition Magic 5 can resize ext2 partitions. I don't remember whether it handles swap partitions, but you can just delete and recreate those.
Re: fullscreen video playback
There are those who would have you believe that Henrique M Holschuh wrote: > On Mon, 23 Oct 2000, matt garman wrote: > > On Mon, Oct 23, 2000 at 10:42:03PM -0200, Henrique M Holschuh wrote: > > > On Mon, 23 Oct 2000, [EMAIL PROTECTED] wrote: > > > > without proper video acceleration in X this is not > > > > possible/reccomended. try using the double option in mpegtv and you'll > > > > see > > > > > > Well, if you define a videomode closer to the resolution of the video > > > you're > > > playing, SDL fullscreen mode should switch the Xserver to that mode, I > > > think. > > > This is obviously not even close to a good hardware-based > > > scale-and-dirther > > > solution, but at least the movies will not be played in a small rectangle > > > with huge black borders anymore :-) > > > > On a local newsgroup, someone said I need the "dbe" module loaded in order > > to let SDL to video mode switching. What package provides the dbe module? > > > > I'm running Debian 2.2, with the xfree 3.3.6 package. I can't find any > > relevant dbe files on my system. In fact, I don't even have the directory > > /usr/X11R6/lib/modules on my computer. > > Well, SC3000 did that for me, under xf 3.3.6 and SDL in its default (debian) > configuration. The only thing I did was to provide the proper videomodes (as > in I can select them using Ctrl+Alt++ and Ctrl+Alt+-). I don't even know > what a dbe module is :-) > This is key: you need to have the right modelines in your XF86Config, and MPEGs tend to use odd modes such as 352x240. I don't know where you'd find modelines for these resolutions, but I'd check the mpegtv docs and website first. The program might be smart enough to pick the closest available resolution, so if the closest mode you have available is 640x400 or 512x384 it might use one of those instead.
Re: I/O to ZIP drive temporarily locks up machine
There are those who would have you believe that Damon Muller wrote: > Hi Chris, > > Quoth Krzys Majewski, > > I don't know how to say this properly.. when backing up files from my > > IDE hard drive to my parallel port ZIP100 drive, the machine > > occasionally blocks: it's like everything freezes for about half a > > minute before I can do anything again. This only happens when the ZIP > > drive is running (not sure if it's R or W or both). This is with > > kernel 2.2.17, using the imm driver as a module. I've set all the > > default "safe" options for the driver, like "slow and reliable" or > > whatnot. Is this a kernel bug? IMM driver bug? User bug? Feature? > > Help? chris > > After having recently resurected my parallel zip drive and connecting it > once again to my debian machine (decided that it might be a good idea to > back up my thesis...), I've also experienced similar problems. > > Using the zip drive under both windoze and linux I've seen it eat up so > much cpu time that the system goes very choppy and is unusable while the > transfer is going on. I don't know if this is a bug in anything in > particular, or if writing that much data through the parallel port is > just a really tough job. > > Unless anyone else can offer a more concrete suggestion, I'd recomend > that when copying large files to your zip, it's probably a good > opportunity to go and make yourself a coffee! > It's possible that Linux is using PIO mode for data transfers over your parallel port instead of DMA. Check /proc/parport/0/hardware (assuming you're using the first parallel port). If there's no DMA, make sure the port is in ECP mode with a DMA assigned to it in your CMOS settings, and read the parport docs in the kernel source.
Re: XFree 4.0.1 and Quake
Why did you need to edit out moduleversions.h? Did you get both the kernel and the GLX archives? There are those who would have you believe that Brandt Dusthimer wrote: > Yup. Did all of that. Do you think that editing the Makefile code so that > it would not load moduleversions.h had anything to do with this? The error > reported from Quake is this: > > XError of failed Request: GLXBadRenderRequest > Major opcode of failed request: 142 (GLX) > Minor opcode of failed request: 1 (X_GLXRender) > Serial number of failed request: 4510 > Current serial number in output stream: 4514 > > Brandt Dusthimer > > > > > There are those who would have you believe that Brandt Dusthimer wrote: > > > I'm having problems with XFree 4.0.1, NVidia drivers (I have a > > >TNT2 Elsa Erazor III, not available in the XF86Config), and Quake III. > > >Quake 3 runs fine until I try to start a game. It load game media and > > >all that other stuff, but as soon as I go into the real game it dumps > > >me with a GLX error. I compiled the garbage myself and had to do a > > >little editing (comment out the modversions.h because it isn't in > > >kernel 2.2.17), and after I did that it ran everything fine. In fact, > > >everything seems to run fine except for quake III ( and a little bit > > >of X with the window managers and the screensavers not working.) > > >Anyone know anything about this or had something like it happen to > > >them? > > > > > > > Are you using the NVidia drivers included with X, or the ones from > > Nvidia's website? In the latter case, did you follow the instructions > > included with the driver about GL support (I don't remember the > > details, but IIRC it concerns making sure that you're loading the > > library from /usr/lib/X11R6/lib/modules instead of /usr/lib/X11R6/lib)? > >
Re: 2.2 - install - cdrom_module
There are those who would have you believe that robert_wilhelm_land wrote: > I'm just in the procedure of installing potato on a virgin system. > Everything seems to be ok, I'm just unsure when it comes to choose the > appropiate modules. > > This box has a ATAPI CDROM which is currently used by the debian > boot/install procedure. > > What are all those cdrom devices in the modules list for? Are these > devices without a ATAPI interface? > > I assume to choose non but unsure having a ready system without cdrom > support. > > Any comments would be helpful. > Those are for old proprietary CD-ROM interfaces. Hardly anyone uses them anymore, but they were common when CD-ROMs first came out. As long as you have a standard ATAPI or SCSI CD-ROM you don't need to worry about them.
Re: PPP Problem (Timeout LCP ConfReq)
There are those who would have you believe that Michael Croon wrote: > Hi all, > > I'm stuck with a problem related to ppp that is similar to the one already > brought up here by Ashby Gochenour's (see his post > http://lists.debian.org/debian-user-0007/msg00157.html in July). > Unfortunately, no solution working for me was given in the thread. > > Short description of the problem: > - > > * Config: > > - Debian GNU/Linux 2.2r0 _Potato_ Official i386 > - Linux Kernel 2.4.0-test9 > - PPP compiled as modules (ppp_generic.o, ppp_async.o) > that kerneld inserts successfuly > - ppp_2.3.11-1.4.deb > - pppconfig_2.0.5.deb > > * Problem: > > I configured the scripts with pppconfig. I use PAP and chat. pppd calls > the ISP, establishes a serial connection and dies with a LCP timeout after > having failed to send out ConfReqs. It seems to be a specific problem to > the above mentioned configuration as my SuSE 6.1 box (Linux 2.2.10, > pppd 2.3.5) works fine with the same ISP (i.e. it sends out the LCP > ConfReq and receives one shortly afterwards). I tried to use the SuSE > scripts on Debian, but pppd died with 'alarm'. > You could try using the SuSE scripts on the Debian system, but comment out all of the lines. Then uncomment them individually to find out which one it doesn't like. You could also put the "kdebug 1" option in your configuration on the Debian system so that you can see the LCP handshaking on that one.
Re: XFree 4.0.1 and Quake
There are those who would have you believe that Brandt Dusthimer wrote: > I'm having problems with XFree 4.0.1, NVidia drivers (I have a >TNT2 Elsa Erazor III, not available in the XF86Config), and Quake III. >Quake 3 runs fine until I try to start a game. It load game media and >all that other stuff, but as soon as I go into the real game it dumps >me with a GLX error. I compiled the garbage myself and had to do a >little editing (comment out the modversions.h because it isn't in >kernel 2.2.17), and after I did that it ran everything fine. In fact, >everything seems to run fine except for quake III ( and a little bit >of X with the window managers and the screensavers not working.) >Anyone know anything about this or had something like it happen to >them? > Are you using the NVidia drivers included with X, or the ones from Nvidia's website? In the latter case, did you follow the instructions included with the driver about GL support (I don't remember the details, but IIRC it concerns making sure that you're loading the library from /usr/lib/X11R6/lib/modules instead of /usr/lib/X11R6/lib)?
Re: Wanted: Advice on Video Cards
There are those who would have you believe that Chris Gray wrote: > On my home box I run Windows 98 for games and Debian(Woody) for serious > (non-graphics-intensive) work. I want to upgrade my video card. I'm more > interested in something that will be straightforward to install on both > platforms and will give respectable performance for a while to come. I want > to avoid the situation I'm in now. I have a cheap S3 Savage4 based AGP card > for which there is no X server in the Debian packages. I can get X servers > elsewhere but on exit they crash my system. > > At the moment I'm leaning toward an ATI Fury. I'd be grateful for any > advice, opinions, or caveats. > The most important single factor is going to be whether you're interested in a 3D card or not. If you are, the NVidia GeForce 2 cards are probably the best out there (although Voodoo fans will probably argue this point). The GeForce 2MX is the bargain version, but as far as price/performance goes you can't beat it at ~ $100. The GeForce 2GTS is the more expensive one, going for around $200. Be aware, though, that these require an AGP 2.0 motherboard. If you don't card about 3D, then the Millenium that was mentioned isn't a bad way to go if you can get your hands on one cheap, and the Millenium II comes in an AGP version. You can also get an NVidia TNT2-based card for under $50, and that has good 2D performance (AGP 4X, 16 MB, I think the RAMDACs are 300 Mhz). The TNT2 cards still have decent 3D performace, although you might find yourself limited to 800x600 or less in newer games and/or if you have a low-end CPU. The Matrox Millenium G400 are good if you primarily want a good 2D card but also want accelerated 3D. ATI's advantage seems to lie mostly in their hardware DVD/MPEG2 support. I really don't know if this is supported in their Linux drivers. The NVidia cards have a pretty good accelerated driver for XFree 4.0.1 that you can download from NVidia's website. For 3D on XFree 3.3.6, you might be better off with a Voodoo card, although IMHO you'll end up spending more for no overall gain in peformance. Enough rambling...
Re: What windows ssh client you use?
If you go to the TeraTerm website at http://hp.vector.co.jp/authors/VA002416/teraterm.html you'll find a tgz'ed terminfo file for TeraTerm. This includs a terminfo file and a keyboard.cnf and instructions for setting things up, but I've found that if you just set your terminal type to "Linux" in Teraterm and use the keyboard.cnf file, everything works. There are those who would have you believe that Diarmuid Drew wrote: > I'm using TeraTerm with the TTSSH extension, but I can not get my keyboard > map correct. Cant seem to map F5 or the page up. I know its way off topic, > but has anybody got a keyboard .cnf file they would like to share ?? > > Derm. > ----- Original Message - > From: "Andy Bastien" <[EMAIL PROTECTED]> > To: > Sent: Sunday, October 15, 2000 3:52 PM > Subject: Re: What windows ssh client you use? > > > > There are those who would have you believe that kmself@ix.netcom.com > wrote: > > > On Sun, Oct 15, 2000 at 03:21:37PM +0400, Rino Mardo ([EMAIL PROTECTED]) > wrote: > > > > I'm using PuTTY from windows to ssh to my debian box but I find that > it > > > > disconnects my session after some idle time. > > > > > > > > Is it a ssh setting (coz I can't find it) or it's my ssh client? > > > > > > Could be either. If it's a high-latency connection, you may be getting > > > a timeout, though I'm not sure where that's configured. > > > > > > There are several java ssh clients which run inside Netscape (or IE) > > > which you might want to play with as well. Try searching Google. > > > > > > > > > Another good (and free) client is the TTSSH extension to TeraTerm. > > > > > > > > -- > > Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] < > /dev/null > > > > > > > -- > Unsubscribe? mail -s unsubscribe [EMAIL PROTECTED] < /dev/null
Re: x font server
There are those who would have you believe that Joseph Holland King wrote: > i just installed debian yesterday. today i downloaded and compiled the x > server and tried to get x running. in oreder to do so i rebooted my system > to make it recognize the mouse. when it tried to reboot it stopped with > the error message that the x font server did not recognize the argument > passed in. now it won't boot and since the master boot record won't allow > me to boot directly into single user mode (at least not in any way that i > am aware of). also i was not able to make a boot disk during installation > due to some other problems. is there anything that i can do in order to > get my system to boot correctly and to get x to work? thank you. > At the lilo boot prompt you can enter "linux single" to boot into single-user mode (as long as your linux image is called linux in lilo.conf). You could boot from the installation CD or floppies, mount your root partition, and either edit the inittab to not start X at boottime or remove the link that starts xfs.
Re: Mouse freezing in X
There are those who would have you believe that Bob Nielsen wrote: > I have noticed in the past few days that if I switch from the X display > to a VC and back again that as soon as I move my mouse the cursor jumps > to the upper right corner of the display and stays there. This is on a > system running woody. Has anyone else noticed this? > This can happen if you're using different protocols in gpm and X; for example, configuring the mouse as ps/2 in gpm and imps in X. It also seems that Microsoft mice will work with either ps/2 or imps protocols, but once you use the imps protocol you can't go back to ps/2.
Re: What windows ssh client you use?
There are those who would have you believe that kmself@ix.netcom.com wrote: > On Sun, Oct 15, 2000 at 03:21:37PM +0400, Rino Mardo ([EMAIL PROTECTED]) > wrote: > > I'm using PuTTY from windows to ssh to my debian box but I find that it > > disconnects my session after some idle time. > > > > Is it a ssh setting (coz I can't find it) or it's my ssh client? > > Could be either. If it's a high-latency connection, you may be getting > a timeout, though I'm not sure where that's configured. > > There are several java ssh clients which run inside Netscape (or IE) > which you might want to play with as well. Try searching Google. > Another good (and free) client is the TTSSH extension to TeraTerm.
kde2 problem
I've installed KDE2 from Debian's servers, but I haven't had much luck getting it running. What happens, basically, is kicker dies with a signal 11 almost as soon as it starts up (the status indicator is on "Restoring session"). What I do get is a placid blue screen with a couple of icons on it. It looks to me like I'm missing a library, but I have no unresolved dependencies. I've included what seems to be the relevant X output. Does anybody have any ideas? libpng warning: Incomplete compressed datastream in iCCP chunk libpng warning: Incomplete compressed datastream in iCCP chunk libpng warning: Incomplete compressed datastream in iCCP chunk libpng warning: Incomplete compressed datastream in iCCP chunk Could not dlopen library kdesktop.la: undefined symbol: prettyURL__C4KURL Could not load library! Trying exec KLauncher doing clientStarted(`kio_file') Could not dlopen library artsd.la: file not found Could not load library! Trying exec Launched ok, pid = 5049 KLauncher doing clientStarted(`kmixctrl') KLauncher doing clientStarted(`kxmlrpcd') xrdb: "Netscape*background" on line 507 overrides entry on line 333 xrdb: "Netscape*foreground" on line 508 overrides entry on line 334 xrdb: "Emacs*Background" on line 516 overrides entry on line 255 Could not dlopen library kicker.la: undefined symbol: prettyURL__C4KURL Could not load library! Trying exec KCrash: crashing crashRecursionCounter = 2 KCrash: Application Name = kicker path = DCOP aborting call from 'anonymous-5061' to 'kicker' Could not dlopen library kwrited.la: undefined symbol: prettyURL__C4KURL Could not load library! Trying exec KSMServer: SetAProc_loc: conn 0, prot=local, file=/tmp/.ICE-unix/5072 QSocketNotifier: Multiple socket notifiers for same socket 6 and type read KLauncher doing clientStarted(`kwin')
Re: X-window
Please type with two hands. Please used a fixed font when composing your messages. Please sit in an upright position before replying. There are those who would have you believe that kmself@ix.netcom.com wrote: > Please do not CC: me. > Please use postfix followup. This means your response goes **AFTER** > the body text. 2nd request. > Please set your mail client linewrap to 72 characters. > > On Wed, Oct 11, 2000 at 03:12:24PM -0400, Antonio Rodriguez ([EMAIL > PROTECTED]) wrote: > > kmself@ix.netcom.com wrote: > > > > > Please us postfix followup. > > > > > > On Wed, Oct 11, 2000 at 02:41:10PM -0400, A R > > > ([EMAIL PROTECTED]) wrote: > > > > kmself@ix.netcom.com wrote: > > > > > > > > > On Mon, Oct 09, 2000 at 12:45:49PM -0500, Cavaiani, Don > > > > > ([EMAIL PROTECTED]) wrote: > > > > > > Can anyone tell me why my X-window appears in a "letterbox" > > > > > > format - i.e., it only uses about 2/3 of the 15" monitor? I > > > > > > can't figure out how to get it to the full screen format? > > > > > > > > > > Do: > > > > > > > > > > $ startx 1> startx.log 2>&1 > > > > > > > > > > Post the output file (startx.log) along with your XF86Config file. > > > > > > > I have the same problem, is is only a small square in the center if > > > > the screen. My XF86Config file is here > > > > > > Repost with XF86Config **AND** the output from your X server, as > > > specified above. > > > > > > I suspect bad modelines. > > > I have xdm taking over, so I need to kill it before I do startx > > whatever, but I want to kill it in a way that does not remove it. How > > do I prevent it from taking over just once, that is, so that I can do > > the indicated procedure? > > One-time fix: > > $ /etc/init.d/xdm stop > > To prevent xdm startup: > > $ update-rc.d -f xdm remove > > To prevent running xdm on console, edit /etc/X11/xdm/Xservers and > comment out local displays. >
Re: Fetchmail and NTLM configuration
There are those who would have you believe that kmself@ix.netcom.com wrote: > On Mon, Oct 09, 2000 at 02:06:12PM -0700, Aaron Brashears ([EMAIL PROTECTED]) > wrote: > > Andy Bastien wrote: > > > You might not even need to use NTLM to access the server, as Exchange > > > servers accept plaintext authentication by default. If your NT account > > > is the primary account for the mailbox, try just using your username > > > and password or domain/username and password. IMAP is more > > > complicated than POP3, but you can telnet in on port 143 to test your > > > username and password: > > > > > > frodo:~$telnet mail.foobar.com 143 > > > Trying 10.10.10.10... > > > Connected to 10.10.10.10. > > > Escape character is '^]'. > > > * OK Microsoft Exchange IMAP4rev1 server version 5.5.2650.23 (MAILSERVER) > > > ready > > > a001 login username password > > > a001 OK LOGIN completed. > > > a002 logout > > > * BYE Microsoft Exchange IMAP4rev1 server version 5.5.2650.23 signing off > > > a002 OK LOGOUT completed. > > > Connection closed by foreign host. > > > frodo:~$ > > > > > > Telnetting into the port worked, but that doesn't match the log given my > > fetchmail. Fetchmail has a lot of strange stuff going across the wire > > according to the -v log. How do I configure fetchmail to use imap > > plaintext? > > Set proto to IMAP, possibly IMAP-LOGIN? > > I don't have experience with this, but can read a man page . > The log that you posted shows IMAP the way it's supposed to be used. The stuff the I posted was just enough to show how to check if your server will accept plaintext authentication. Any mail client that does what I did is probably broken, but I'm a human so I have an excuse ;-). With the NTLM, It looks like you're trying to use "domain/username/smtp" to access this mailbox, when what I think you want is "domain/username/mailbox". Remember, Exchange is not a native SMTP server, and your SMTP address is not directly related to the mailbox name. It's a pretty standard procedure to make the mailbox name to be the same as your username, so you could try that (i.e. corp/abrashears/abrashears). The instructions for regular IMAP are in the man page. Fetchmail might be out-thinking itself and automatically using NTLM authentication because it sees that the server supports it. If this is the case, you should be able to force fetchmail to use password authentication. I don't know the option off-hand.
Re: Fetchmail and NTLM configuration
There are those who would have you believe that Aaron Brashears wrote: > Hi there. I am running woody, and have the sad requirement that I must > use the company's exchange server. I've been using netscape for a while, > and it works fine using the format 'domain/ntusername/mailboxname' for > access the imap gateway on the server. I've also tested mutt, and it's > imap support seems to work fine when accessing the server, but it takes > a lot of typing and the documentation for mutt suggests using fetchmail. > > So, after a quick apt-get install fetchmail fetchmailconf I have > fetchmail running, but I can't get the configuration right. When running > fetchmailconf I set up the server, and the user as > 'domain/ntusername/mailboxname', give it my password, and then set local > delivery to be my localhost account name. When I test fetchmail, the log > indicates that fetchmail recognizes NTLM, spews a lot of characters, and > then complains that I can't login. > > Is there trick I'm missing? I already tried ESR's reccomendation for > '/domain/ntusername/mailboxname' which also failed, and isn't the format > used in netscape or mutt. > > Thanks. You might not even need to use NTLM to access the server, as Exchange servers accept plaintext authentication by default. If your NT account is the primary account for the mailbox, try just using your username and password or domain/username and password. IMAP is more complicated than POP3, but you can telnet in on port 143 to test your username and password: frodo:~$telnet mail.foobar.com 143 Trying 10.10.10.10... Connected to 10.10.10.10. Escape character is '^]'. * OK Microsoft Exchange IMAP4rev1 server version 5.5.2650.23 (MAILSERVER) ready a001 login username password a001 OK LOGIN completed. a002 logout * BYE Microsoft Exchange IMAP4rev1 server version 5.5.2650.23 signing off a002 OK LOGOUT completed. Connection closed by foreign host. frodo:~$
Mutt time/date display
I use Mutt for my mailer, and I currently have it set up to display the time and date of incoming messages like "09:23 Sep 12". This is ok, but it's not the information than I like to have for every message. What I would like to do is display only the time for messages I've received today, only the day of the week for messages I've received this calendar week, the month abbreviation and day of month for messages I've received this year, and the month and year for remaining messages. This is similar to how ls -l displays file date/times. >From what I've been able to determine, the "index_format" is static and not capable of performing this sort of logic. Is there some sort of index_display_hook or some other way of doing this?
Re: ODP: Debian chicken (was: Article: Debian's Daunting Installation)
There are those who would have you believe that Keith G. Murphy wrote: > > George Bonser wrote: > > > > > > > > I for one would buy a t-shirt with the chicken printed on it. Any > > > takers? > > > > > > > hmmm ... now that sounds really cool. Make the Debian mascot a > > chicken! I kinda like it. How about a logo with a penguin and a chicken, > > wing/flipper over each others shoulder! So cool! > > > Well, the old logo did look like a chick, after all. > > On the other hand we could make it a cock, er, I mean rooster. > I suppose that would go with woody.
Re: dac960 problems
I've run in to the same (or similar) problem. In my case, another driver in the kernel prevented the DAC960 driver from loading (I don't remember the exact error messages). What I ended up doing was loading the module from another floppy to complete the installation, then building a kernel on another system with the DAC960 driver compiled-in and using that instead of the stock kernel to get the system running. There's probably an easier way to fix it, but that worked for me. There are those who would have you believe that armin glaab wrote: > Date: Wed, 04 Oct 2000 22:32:35 +0200 > From: armin glaab <[EMAIL PROTECTED]> > X-Mailer: Mozilla 4.7 [de] (Win98; U) > To: "debian-user@lists.debian.org" > Subject: dac960 problems > > I try to install debian (potato), my problems if I boot with the first > CD the kernel doesn´t detect the mylex raidcontroller. I made me two > bootdisk 1. install/rescue disk & 2. root disk, with the compact kernel, > > but the rescue disk doesn´t boot, what did i wrong?? > > thanks for help > > armin glaab >
Re: of Xservers and XSetup, of sabotage and kinks...
There are those who would have you believe that S.J. Black wrote: > Date: Fri, 15 Sep 2000 23:23:36 -0700 > From: "S.J. Black" <[EMAIL PROTECTED]> > X-Mailer: Mozilla 4.75 [en] (X11; U; Linux 2.2.16 i586) > To: debian-user@lists.debian.org > Subject: of Xservers and XSetup, of sabotage and kinks... > > Hi All - > > Just downloaded 2.2 - verrry sweet ftp install! I've run into a small > problem with getting/finding/installing the XServer, though, and was > wondering if someone could help with this. > > I did grab the XFree 4.0 stuff, as XF86Setup was quite insistent that I > did not have the "required server" for my Diamond Viper 770. Looked at > the documentation for NVidia, and came out of the 2 hour doc-search no > wiser than before. Again, was surprised to find setup telling me the > required server didn't exist. > > Any suggestions? > The Diamond Viper 770 is based on the NVidia TNT2 chip. It's supported by the SVGA server in X 3.3.6. For 4.0.1, you can use the "nv" driver, or you can go to NVidia's website (http://www.nvidia.com) to download their accelerated server.
Re: Off Topic: SCSI Recommendations?
There are those who would have you believe that Leen Besselink wrote: > Date: Fri, 15 Sep 2000 23:48:39 +0200 (CEST) > From: Leen Besselink <[EMAIL PROTECTED]> > To: debian-user@lists.debian.org > Subject: Re: Off Topic: SCSI Recommendations? > > > > I haven't run into any problems with my onboard U2W 7xxx chip, so... > > > > So have I, or haven't if you like. :) > The drivers developers have complained in the past that Adaptec is less than helpful. Adaptec certainly makes quality hardware but they haven't been the best friends for Linux. I've had good experiences with Symbios-based boards, and I've never had a problem (except for the Diamond Fireport, but that board's in the landfill now). I've got a Tekram in my home PC and my work PC, and both work great. I have a different brand in my home server (I'm spacing the name) and it's also been flawless.
Re: Debian Gnome vs. Helix Gnome
There are those who would have you believe that Michael Soulier wrote: > Date: Fri, 15 Sep 2000 17:48:01 -0400 (EDT) > From: Michael Soulier <[EMAIL PROTECTED]> > To: Debian Users List > Subject: RE: Debian Gnome vs. Helix Gnome > > On Fri, 15 Sep 2000, Sean 'Shaleh' Perry wrote: > > > gnome in debian came from gnome.org, not helixcode. > > Yes, my terminology was incorrect, but the question stands. If > gnome is running the gnome project, but helix is extending it, what's the > current status between them? Does helix have commercial interests that I > should avoid? I find the information on their homepage a little scarce. > > Mike > They hope to eventually make money off of services: http://www.boston.com/dailyglobe2/236/business/A_user_friendly_face_lift_for_Linux+.shtml Plus, you have to trust them in that the sources that they put up are the ones that they use to build the binaries. I do, but there are some people who insist on compiling everything themselves.
