Tekchip

I'm thinking of starting a buisness building computers. I definately wouldn't be the first to do so. ;-) The spin I plan to put on it is a support of linux gaming. Linux pre-installed on the systems. Possibly even providing pre-installed xqf and some cd's with good linux games like frozen-bubble etc. Once I get things rolling possibly even sell systems with UT2K3 or other linux native capable games pre-installed. Does this sound like a good idea? Would any one out there considering buying such a system either soon or in the near future? Please let me know in the poll. Thanks!

I installed Fedora with no problems. Installed the rpm version of kernel 2.6 with no problems. Now I seem to be having some issues getting the Nvidia drivers installed. According to everthing I've looked up you have to patch the nvidia driver source before running a make install. The patch and instructions are found at www.minion.de. I followed the instructions to a T with no problems till I get to the make install at the end. Then it gives me the error : You appear to be building the NVIDIA kernel module with a compiler different from the one that was used to build the running kernel. This may be perfectly fine, but there are cases where this can lead to unexpected behaviour and system crashes. If you know what you are doing and want to override this check, you can do so by setting IGNORE_CC_MISMATCH. In any other case, set the CC environment variable to the name of the compiler that was used to build the kernel. make: *** [cc-sanity-check] Error 1 I'm guessing this is due to the fact that the kernel rpm was i686 and the kernel source rpm was i386 and so there's a difference in the headers apparently. If any one might be able to make some suggestions on how to make these drivers work it would be much appreciated. Thanks in advance.

Ok I found that you can do an export IGNORE_CC_MISMATCH=1 and the driver will recompile and state that it was able to install the module. I then try X with the driver name changed to nvidia in my XF86Config. The X server still wont load. I then ran across another post that suggests that you run make in the NVIDIA-Lin.../usr/src/nv and then run make install in the root of that NVIDIA-Lin... directory in order to put everything where it should be. I attempted this revision of the install and changed my XF86config again and still no luck. Does any one have any other suggestions? I've tried both processes with both the 4496 and the 4620 drivers with no luck.

Only way to get anything faster would be to replace the entire machine. Sorry.

Spent more than a few nights trying to figure out how to make it happen. I've yet to see if the AIW tv functonality works and I havn't tested dual monitors. This is a basic guide and may be not even the right way but since I hadn't found any place else describe how to do it I figured I'd post. Hit up this site for those instructions.

See here's the thing. In Redhat or Fedora some sort of a DRM kernel module gets loaded on boot up which means that weather you run fglrxconfig or not you cant start X because the fglrx module wont load. I personally am not a kernel hacker and havn't gotten to a point where I can easily recompile my own kernel. Which means I cant hunt down or remove that DRM module. Most new users are at or near the point I'm currently at with the kernel. Not being able to run 3d applications just because you cant hunt down and remove kernel modules really isn't an option for us newbs. So I had to do a little config/bootup hacking to get the fglrx driver and the agpgart external modules to load before what ever the DRM module is. This corrected the issues I was having and got the drivers to work. I figured knowing how to by pass the issue with out doing anything to the kernel was a good thing that I should share. Believe me I spent a good solid 8 hours the other night reading about the subject and trying to find a straight forward way to make them work. I eventually ran across something on the suse message boards that said something about moving those to the head of the line in the startup. I then had to go figure out where the RH/Fedora startup file was(which wasnt tough) and add them. Turns out it worked.

I've looked through the rest of the SATA related posts but none of them seem to address the issue of running RAID0 with linux. In windows at the very start of the setup you can put down the SIL3112 driver and windows install will see the pair of drives as one large hard drive. When I attempt to install Redhat 9(shrike) it sees them as seperate drives in the partitioning screen. I have all ready downloaded the drivers for the controller from Asus(my mbd is a a7n8x deluxe). These drivers only come in rpm form for rh7.3, rh7.3-athlon and suse8.0. In my experience most 7.3 compatable programs/drivers will work just fine with newer versions of redhat. This brings me to the real question. Is there some way to install drivers for the sil3112 controller pre-setup? If so how is this to be done since no rpm dbase is available to install them into? If there's some other method please let me know! Thanks in advance!

Right now the 2.4 kernel doesnt support SATA with out a patch to the kernel. At the moment if you want a kernel with that all ready built in you have to run with Suse 8.2 or newer. If your competent at recompiling kernels then you can either get the kernel module and recompile your current kernel or you can download and compile any one of the test/beta 2.5 or 2.6 kernels. Only catch is that the current module doesnt support raid on SATA for most raid chipsets which leaves me at a loss. Any way those are your choices unfortunately. Kernel 2.6 is going to rock the house, hang in there.

You want to say yes to binutils. Basically if the nvidia drivers cant find a pre-compiled kernel module from the ftp then it has to try to compile one for it's self based on your computer. Thus the reason it needs binutils. Just run the drivers and say yes to everything. You'll also need to edit (in RH9 any way) /etc/X11/XF86Config towards the bottom is the display adapter type. You'll see a line that says 'nv' on it. Change that to 'nvidia'. That will make the nvidia kernel module load on xstartup.

Acctually I clicked the link you provided and then clicked the 'download' link on that page. I was taken here(part of the same page) where you can download it for free.[/url]

I've never met a source tarball that I liked or that was less trouble than a simple rpm. Any way the drivers are no longer real rpm's but they are Nvidia's own installer. Unfortunately the installer wont run in an Xwindows session. It'll complain. So basically switch to super user(root) and run the 'init' command to switch to a non GUI mode suck as runlevel 3. So the command you need to run should look like 'init 3'. Most distros this needs to be run from the /sbin directiory. This will switch you to a command line only mode and allow you to run the install. You can edit your XF86Config from either a text mode or a an XWindows mode. Oh yeah from the command line still as root just type 'init 5' to get back to xwindows. Hope this makes sense. Either method should get the job done for you. Good luck with it!!!

ac97 codecs are handled by the intel 817 drivers I believe. As for video you diffinately want to grab those drivers. The newer version of the drivers have to be installed from a run level of 3 so you'll need to run 'init 3' and then execute the drivers. Once you've run the drivers you need to edit your XF86Config file to reflect the install of the Nividia drivers. Typically this file is located in /etc/X11/XF86Config. Under device change the line with 'nv' to say 'nvidia'. This will get hardware support running for ya and smooth out your game play. If the board is an NForce or NForce 2 based board, which it sounds like it is, then you will probably also want to grab the Linux NForce drivers from Nvidia.

The Gateway EV700 is a monitor model #. We'd need to know the model number of your acctual computer. It would also be helpful to know what kind of hardware you have in the machine. Usually a hang at boot time is caused by some form of hardware conflict. You might also want to try a newer version of Redhat like 9 which will have a newer kernel and modules supporting a greater amount of hardware.

Well there's the Linux+, RHCT and RHCE certifications which seem to be the most often talked about and well known. From what I can tell there arnt to many jobs that call for the Linux+. It may help you out if your trying to get a job in a cross environment but it's not likely to as useful in the aquisition of a job as either of the RH's would be. At least from what I can tell. I personally am aiming high, and I know it, and going with the RHCE. Something I found out was that an RHCT can be aquired even if you dont pass the full RHCE. It's really going to come down to how hard you want to push your self and how much you want to learn to get to what job level.

tcp dump is probably not in your path. You'll need to figure out where tcpdumps executables are located and run them from there. most programs get dropped in /usr/local/someplace. That or run a search for the executable. Most executables in user mode need to be run with either sh foo or ./foo I hope this helps!

Your modem isn't a software or 'winmodem' is it? Winmodems rely on your system processor to do the connection mathmatics which means most of the connection handling gets handed on via software. These modems dont fair so well under linux. If you want to see whats supported and whats not check this out. http://www.linmodems.org/ If it's not a winmodem then you've got some other issue. You'll probably need to do some command line trouble shooting to find where the problem is. Unfortunately I dont know to much about that subject having ditched dial up for broadband several years ago when cable modem was introduced. I'm sure some one else on the forum will be able to help you out. Good luck correcting the issue!

The only way any one on the boards will be able to help you is if we know what kopete is? We'll also need to know what kind of errors it throws. The only way to tell what errors it throws is to execute it from a terminal at the command line. Other wise the errors are hidden by the desktop. If you can provide that info we'll be glad to help you out!

Threw the model number in google, came up with this link. http://www.infosys.tuwien.ac.at/Staff/tom/SiS7012/ "The SiS7012 patches have now been integrated into the standard linux kernel (version 2.4.18 ). Just enable Intel i810 audio support and everything should work perfectly fine." I dont believe the default RH9 kernel is 2.4.18 I think you need to get the 2.4.18 kernel through the redhat network updater. Might want to make sure to do that before moving ahead.

REL!c2K What type of motherboard are you running? I've only been able to find the Asus A7M266-D which only runs PC2100 ram. Are you using that board and just over clocking? I've been looking at going dual Athlon.

I dont know a whole lot about ISDN per say but I have a good handle on networking. I think this link might give you enough insight to get something working. http://sdb.suse.de/en/sdb/html/hoe_isdn_terminaladapter.html

bump

Theres a previous thread about the 3112 under linux. Here's the link http://www.linuxcompatible.org/forums/viewtopic.php?t=508

I dont know where you guys have been or if you just didnt search the forums before posting but there was a previous thread discussing the 3112 under linux. Here's the link to that thread. http://www.linuxcompatible.org/forums/viewtopic.php?t=508 Bottom line is that for 2.4.x to even see the drive you need a kernel patch that is linked to in that thread and Raid is only supported in unstable 2.5.x and the soon to be stable 2.6.x. Go read up.

Ok we're going to need more info. Exactly which players have you tried and what are you wanting to use it for? Ogle is an awesome DVD player. Mplayer is an excellent all around player of nearly any format. Have you tried both of these? What problems are you running into exactly? Dependencie issues, errors?

Something I have had to do in the past before getting a hold of my current HP monitors is to know the max refresh rate and resolutions of my monitors(usually from what I would run them at in windows) and simply choose a generic monitor configuration that matched what I wanted to run at. Basicly this may be the way you would need to know if you cant find the horizontal and verticle sync rates and your monitor isn't listed in the list of supported monitors. Hopefully that's some help to you.