aRTee

Hmm, I'll look at that next time I work on my machine...

K3B not wanting to do DAO is quite normal, even if you have cdrdao installed and permissions set right; it doesn't want to do it on my box, even though I can do a dao disc with gcdmaster. Reason: I think it tries to use cdrecord (it will for data discs) and that doesn't do dao discs with most drives. It does work for other stuff as user chris, like burning iso's right? So your stuff is properly set up I would think. At least, it may well be. Also, try if you can use gcdmaster (audio without 2sec pregaps, for live albums etc) as user chris. If that is the case, your permissions are ok.

Go for Icewm as a windowmanager, and realise that to install mdk9.1 it will be a pain, it was even a pain on a 64MB machine (pIII 766MHz of all things, it was not the cpu speed!!). Memory counts. If you can get it to install, tweak it by switching off all unnecessary services (ftp servers, mail stuff, etcetc). Don't even think about kde or gnome; icewm (very much like win95) or blackbox (which I only once had a look at) should work nicely though. Not using kde or gnome means also: not using konqueror, kmail, evolution, nautilus either... On a side note, 50USD is a bit much; I just tried to 'sell' an old P 166MMX with 48MB of ram and 3.2GB hd (plus audio, plus network) for 35USD and got my offer turned down... ;) Basically, it's likely not worth the hassle with such an old machine...

The moving of the heads is not what takes a lot of time, the beginning of the drive is where the data is read the fastest, so basically: if you have <192MB of RAM and you like to have plenty of programs open on a 'heavy' DE (kde, gnome), your swap will be used a lot. If you have >=256MB of RAM, or don't run memory hungry apps as much, you swap may never be used. (I had 512MB for a year before upgrading to 1GB, and it really didn't get used often; I only upgraded since I needed to know for sure if my mobo and the memory would hold with two 512MB sticks and the warranty was running out...) If your swap is used a lot, this slows down the system; by putting the swap file/partition at the start of the drive (which is closest to the center of the drive), it is technically on the fastest part of the HD so gives the best performance. Since the difference in speed (especially latency but also bandwidth) between the HD and main memory is a factor of 1000 or so, it is always better to avoid using the swap space. The command 'free' is your friend, it tells you if your disk swap is being used, and if so, the best way to improve is to - get more RAM, or - change your habits in your computer use, use less programs at the same time etc. Optimising with several swap partitions as you propose may or may not improve the speed of your system by 0.1%. Avoiding to use disk swap improves by a factor (maybe up to 1000x faster, since normally you'd have to wait for the disk swap to be read, now you don't) Please also realise how swap works: it will not be read and written like small (KB) chunks, when it is used it is really several MBs at once. So again, the moving of the disk heads is really not the big delay. The latency (8.5ms) and bandwidth (20-50MB/s) are. Latency of RAM is measured in us or even ns and bandwidth is up to 3.2GB/s or more...

P4 far better than XP for multitasking? ???? Sounds like FUD to me. There is no way that without benchmarking etc anyone can distinguish a P4 box from an athlon box (of similar rating/speed) by just using it, if properly / similarly set up and with the same other hardware (graphics card notably). I doubt anyone without a stopwatch (without keeping time) would actually be able to tell the difference between a P4 2.4GHz, an Athlon 2400+ and a P4 2.8 or even 3.2GHz, whereas your wallet will certainly feel the difference. Take the cheaper one, and add 1GB of mem instead of 512MB. Not having to use any swap makes much much more difference, the best way to seriously cripple a P4 is to put on only 256MB (or worse: 128MB ! ) of ram, and then open and use loads of programs at the same time.

Gowator, your points are in reality very true, though they were/are not presented as such by the movie industrie. 1) region codes: to protect cinemas around the world, originally or as an excuse. The point being that in the USA many movies came out on video when in Europe they were still making money in the cinemas. They didn't have to worry about parallel import too much due to the fact that US tapes are NTSC and in Europe PAL and SECAM are used; of course there are players that can convert NTSC but it is an inferior format anyway. (Did I write that? Yes you bet. NTSC is not a good standard compared to PAL. SECAM is technically better than PAL, which is why the French (who else ;) I'm sure Gowator understands this remark.. :) ) developed it; however, the perceived difference between PAL and SECAM is nil...) There are more than 6 region codes, there are just 6 regions, but for example dvds that are shown on airlines are code 8; and of course there is code 0 for code free dvd, a disc that plays on a player of any code. So the big excuse being: region codes are needed to keep cinemas in business is always used but completely wrong, and that's why Gowator is dead on: it stops people from buying dvds where they are cheapest. Although where they are cheapest is in Asia, 1$, and trust me, those have no region code ;) No css either btw.... :D Anyway, the reason of why this separating is simple: in the audio cd business this didn't happen, and the labels have DIVIDED UP THE WORLD! Yes that's right, they are sort of kartels, which are actually forbidden. But a true kartel sets prices. They divide into regions and in each region the movie label can set the price. For audio this went bad so they had to get politicians on their side, to get import restrictions. This has now also happened for dvd's by the way, at least here in Switzerland... In any case, for dvd they just went about it smarter. They also make sure that the language that are on the disc are not universal enough, like here in Switzerland, with 4 official languages (French, German, Italian and Rumanch), we get to see discs from Germany in the german part, french discs in the french part etc... and the french discs usually don't have german audio or subtitles. Another way to defend the territories.... The proof of all this: why are 10+ year old movies region coded????? Really, Europe has a lot of work to do against this criminal action, Brussels doesn't seem to protect the citizens/consumers very well..... 2) CSS is the real copy protection, with the DMCA as the real thing. So it is legally based, not software based. (Actually, encryption is not software or hardware based, it is algorithm based...) Why this is not real copy protection in itself is easy: make a bit by bit copy, and you will have the same movie. In other words, no matter what, as soon as something is digital a perfect lossless copy can be made. So CSS is just in place to make sure that the home user cannot legally make a backup. And the DMCA was bought for that. 3) copy protection as you call it, is known as Macrovision; actually the tv doesn't handle it at all, video recorders handle it. In the wrong way. The part where the teletext signal normally resides, betwewn 2 frames (images), the signal is alternately pumped up in luminance (white) and then down (black). The tv just tries to display this normally, but since it is outside of the image, no one notices. The vcr averages the signal with AGC (automatic gain contro) circuitry, and gets confused, so it turns to black and white etc to make the best of things. Very old vcrs with less sophisticated circuitry are actually not fooled by this, and just make a fine copy. This macrovision / copy protection signal is added by the hardware or software, and this is one of the reasons that decss cannot be allowed. Namely, with decss you can eventually pass the data without this macrovision disturbing signal. On PC's, the region code is handled in 2 levels; the first is in the dvd-drive firmware, the second is in the dvd playback software (windvd, powerdvd etc). Of course the latter is not implemented in linux dvd players, but that leaves the first. What most people don't know is that: - All dvd-rom drives that are sold nowadays (at least here in Europe) are region limited. But they don't have a region selected, for the purpose of being able to sell the same model everywhere. - on windows, the software player asks the dvd drive what code the dvd disc is, and what region code the drive itself is set to. If the drive is not set to any code, the firmware tells it to set to the region of the disc, if the (windows) software itself is also set to this region. Then the disc can be played, as all other discs of the same region. If a disc is played from another region, the drive firmware will count up by one, the max count is 5. Meaning: after 5 changes of region, the drive will not change it's region code anymore, but remain set to whatever region it was last locked into. (Btw for Linux there is a tool to set the region, it may be that if you don't use that the dvd drive refuses to play -- think it is called regionset... ) So: if you have some region 2 discs (europe) and some region 1 discs (America), and 1 region 4 disc (Australia), and you swap a bit, then the fifth time you decide to play your one and only region 4 disc: it better be good, since that will be the only one that you are going to be playing! Now, some drives were sold without this counting mechanism, they are called 'RPC-1' drives. They cannot be sold in most parts anymore, the drives that count to 5 are called 'RPC-2' drives. On the web you can find plenty of firmwares for RPC-2 drives that turn them into RPC-1 drives. See also: http://www.firmware-flash.com/ After that, you have no more problem in linux since xine and mplayer don't check the code anyway; on windows you need something like dvd-genie to fool windvd/powerdvd that the disc is from the same region as the player. Now what I don't get, so please answer if you know it, is how they manage to get around the region code in windows with locked drives. They are pointing to some software, so it seems that you can tell your firmware to ignore whatever is going on and just play... Does anyone know if that is also the case in linux?

DOlson, after they fix the icons and the exact placement and contents of the menus, people are probably going to bitch that it's free so it can't be any good...... lol dardack, yes I already posted the quote that PS is what they need, and GIMP doesn't cut it. Please read the article, you will find that it is more about what the artists are used to than what GIMP could actually do for them. I read an ad of AMD some time ago, where they were looking for designers. It was mentioned that they are using such and so software for design/development, but were willing to acquire whatever license of whatever design package that the designer would wish to get the job done. Now that is big. I know what some packages cost (I'm an electronics designer, where I work a designers seat costs 35K USD per designer...) ... some / quite a few are at or above 100K USD. So for Disney to get some licenses for good designers is not a big issue. It counts, somewhere it comes out of someones budget, but really it's not a big point. So they go the PS way, whereas with some more time for the designers to get used to and learn to work with the GIMP, I think it could be good enough for what most of them do. For a home user, it is a big point, and in most cases they can make do with GIMP.

Yeah, this sources thing is strange; I have opened a bugreport at the time, no one has done anything since, but in any case the bug seems to remain: even if you select development system, you have to explicitly install the kernel-source, or do it later with urpmi. Anyway, keep up the beta testing! (I will join you in the rc stadium, I really don't have much time..)

Photoshop can AFAIK process 48 bit images, such as those you get from scanners. Xsane by the way can also scan into 48bit images, which then cannot be opened in the GIMP. Been there, tried that (just for kicks). Never noticed any difference with 24bit, but I agree that for professionals, using many filters etc it is better (in some cases) to have the greater colour depth. To anyone considering GIMP (due to not being able to afford PS) I think this greater colour depth (or the lacking of such) is the minor of their problems. They want a perfectly new Rolls-Royce for the price of a Golf. Not that that are PS and GIMP, just to indicate the mentality of those people. And in this world, a RR for the price of a Golf is only found in emails like those where you have to front 'only' 650US$ (NO!! 699$, see SCO!! ;) ) to liberate some millions from a Swiss bank account...

On a side note, K3B is IMHO nicer. But this is personal, ... in any case, if you don't like xcdroast, just be aware there are plenty of alternatives. The main thing being: burning cd's. :)

The 16 bit colour thing is the one thing (since inclusion of CMYK colours in the latest beta's) that GIMP cannot do. Mind you, it is 16 bit per subpixel/colour, meaning 48 bit for RGB... GIMP can only do up to 24bit colour (jpg etc, gif as always is 8bit/256 colours). Apparently FilmGimp (Cinepaint) had one of the first things changed in that bit depth, because they really needed 48bit (16 bit per subpixel). For them it was very important, for most other things it is not. How many people complain that jpg does not have enough colour depth? Get real. ;)

Ok, and now it works again. What was up?? Well, I noticed the following line in my /var/log/dmesg 0: nvidia: loading NVIDIA Linux x86 nvidia.o Kernel Module 1.0-4321 Wed Mar 5 19:13:04 PST 2003 Saw in /etc/modules this: NVdriver which I just commented out. And voila, all is well, it boots into graphical login... :wink:

On a side note, from professional users: From: http://www.eweek.com/article2/0,3959,1210030,00.asp So they are using linux but want PS... sounds familiar?? ;) BTW a very good comment on Linux Today, which reminded me why I don't ever want to financially support Adobe (the second quote more than the first) : From: http://linuxtoday.com/high_performance/200...080600726NWHESW

No good reasoning here, sorry. Compare that to: If linux was good there would be no more windows. This is not the case today, so draw the conclusions for yourselves... Ok, not a fair comparison, but yours was not so fair either. Compare PS vs GIMP to BMW 850 vs VW Golf. Both good cars, but not everyone can afford or needs a BMW. Heck, I would take the Golf (spoke he as a past Golf owner... ;) ). Anyway, I agree with much of the rest you said: interoperability of Adobe stuff as a suite is nice, and yes, currently this lacks in Linux. We do have some of that in for instance OpenOffice.org though. On RTFM: show me 10 posts in the last week where a newbie was told that, and I'll show you 100 where it wasn't the case. Things change, and this is (luckily) one of them. On literature: check out the gimp website, tutorials, even free books (DOlson posted the apt-get command, now what would that be on urpmi?? ;) Hey DOlson, stop making ads for debian!! :D ). To the topic starter: as you already realised, but just to clarify: no I was not insinuating that you don't have a legal copy, just that many arguing that the GIMP is lacking and useless are such themselves.. ;)

I had problems with my hp 4200c, other models may or may not work with or without problems (sounds like a big disclaimer), ... in the end I decided I had had enough and bought an epson. And happily scanning ever since. Check my specs (see sig).

What are the 64bit benefits really? I don't have more than 1GB of ram, few people have more than 4GB which is the 32bit limit. The other advantages like the new registers will be a bigger miss, but one of the real goodies of Hammer (opteron/athlon64) is the onboard memory controller and large cache. Low latency etc. But I agree, if you can, run a real 64 bit OS on that thing, and make it fly! Thanks to AMD for bringing 64bit (and the rest) to the masses! Well, soon anyway, I hear September 22 for the desktop/workstation AND LAPTOP 64 bit hammer cpu's!

Sorry, I disagree. The full paragraph states: Which to me means that when resizing partitions for WinXP, this may lead to problems for WinXP and they can't support that. Normally, the partition program should support any such problem (partition magic or whatever). Note that if partitions don't start and end on the exact boundary, and thus are badly defined (which happens, it happened to me with partition magic), this may give problems, depending on the file system. It may well be that NTFS is such a picky filesystem. So to me this is not fud from ms against linux, it's just being clear about creating a crappy system. Honesty about being second best wrapped in hard to understand words. No linux fud. Just plain old MS software -- instable and problematic due to the user, not the system itself (yeah, right) -- and no warranty... Then again, who would ever need any info on that page? Rare are the stories of those who moved to linux and moved back to windows.

To add a bit to what mtweidmann has said: Mandrake released an Opteron distro, compiled for AMD64. I think you can download it from the usual places. This one should also directly run on Athlon xp-64. The nice thing about Opteron is that there is only linux available (ok, there are windows 64bit betas) at the moment. So no worries whether the mobo chipset is fully supported or not.... ;) Thanks to AMD for bringing out the platform and MS for being late with 64bit windows. For Athlon 64 it looks like linux will be the first to have all that is needed to use a 64bit OS; with FLOSS drivers, just recompile. For windows I don't know, but I think it was mentioned that getting all drivers is the big problem; the OS itself is said to function well. BTW you can always just run x86 versions on athlon 64/opteron, since it is fully backwards compatible. The cpu just needs another mainboard.

For me things are easy: GIMP is free, PS is not. I don't do/use pirated software. As for UI, you have to get used to things. I like the way they did things in the GIMP, right click, being able to tear of any menu and keep it. Remember, in the GIMP on linux, you just have one workspace/desktop purely for the GIMP. People used to windows hate the window cluttering, but then often they have all on one single desktop... I disagree that Linux users don't want to pay for software. But I agree with Adobe, if they port photoshop, people would just copy it and not buy. So wasted money for them. What I really can't stand is people who say that the GIMP is just not good enough due to the features it lacks. Let me clarify: some professionals really need certain features. But most people bitching are using a pirated copy and don't really need the features they are bitching about, often I think they are just pointing out the shortcomings that they heard someone else mention who really needs those features. Notably: CMYK colours (which will come in the new branch at some point), which is needed for professional press, and 16bit colour (48 in total, instead of 24). Other issues: like not liking the UI, like having problems with a moving gif etc is fair. But are you using an officially licensed PS? In that case, I see no reason why you can't afford a license of crossover office which would allow you to use PS on linux. BTW on crossover and office/ps like apps, there's not really much of a performance loss, so no point not to go that path. If you (anyone in general - not anyone here in person) don't have a license of PS, how dare you bitch about features of GIMP that don't function properly or that are lacking???? Naturally, it's fine to complain and especially indicate better solutions or indicate errors/bugs. But with bitching I mean: really putting down GIMP with no other reason then to tell how great PS is. I sometimes wish the BSA was more active in getting to home users. That would do wonders for FLOSS/linux/GNU/GIMP/KDE/GNOME etc... On the other hand, Adobe will come too late to the needs of those who won't run PS though wine (whatever version), IMHO. They won't port until there are enough (paying) users. But people don't move until the software suffices to them. So as the userbase will only be using linux when free software alternatives suite their needs, porting once there are enough users means they are already using a free alternative, and not likely pay, especially the hefty price on PS... Look at making pdf's. Is there anyone who would pay for that under linux?? Look at cd burning; Nero was supposed to get ported. Now with K3B I really don't see who would pay for Nero on linux. So FLOSS is filling the needs first, then people move over to it. The userbase of Linux is not going to buy anything they need since they already have it; they would not be using linux otherwise. To all proprietary companies: - this means: if people are still on windows due to your app not being available on linux, you can port now and sell, or you can wait for them to switch and not ever sell to them, since they will only switch once they have found a substitute. (Not counting those very few who switch and then use wine..) To get back to PS vs GIMP: I understand xaff, being used to something and being handy with it is not something that you likely throw away. One of the reasons that so many who really managed to get to the inside of MSWindows (as far as that is possible) and who actually have what it takes to really learn linux, don't want to learn/move to linux. Throw away all they learnt? Start over from scratch? With a system that is/appears to be in some ways less polished? And I agree, GIMP will not be ready for the professional market. But it doesn't have to be. It has to be ready for the enthusiast home user, who cannot afford PS. And I think we all agree that GIMP beats anything (MS Paint up to and including Paint Shop Pro) except PS.

I will have a look on how I set things up on my server tonight; upon boot, my server plays a melody through 'beep' (urpmi beep) so I know it has properly booted; no screen mouse or keyboard attached to this machine.. As last stuff before completing boot it actually starts icewm in a vnc-server as a user (not root), within this it starts the edonkey client. Don't know about that edonkey client, but all other things is what you want. Ah you're in luck, I have sent a colleague the same info, here it goes; mind you, lots could have been put in a single user script, but I built things up step by step and never bothered to change it; feel free to make one user script and only one reference to that in rc.local: In /etc/rc.local I have the following (added at the end): (my username and account is 'robert') # commands to autostart robert-account su -c /home/robert/bin/autostart.sh robert & /home/robert/bin/beepsound Then in /home/robert/bin/autostart.sh you find: #!/bin/bash cd /home/robert rm -f /tmp/.X3-lock rm -f /tmp/.X11-unix/X3 /home/robert/bin/vncdefstart cd /home/robert export DISPLAY=:3 /home/robert/bin/termstartdonkey & gkrellm & in /home/robert/bin/vncdefstart is the following: #!/bin/bash vncserver :3 -geometry 1280x1024 -depth 16 (so the display is :3, which is why after exporting it things end up in my vncsession; simple as that....) in /home/robert/bin/termstartdonkey is the following: #!/bin/bash /usr/X11R6/bin/rxvt.bin -display :3 -T -menu -fn -*-fixed-medium-r-normal-*-*-140-75-75-*-*-iso8859-1 -fg lightgray -bg black -e /home/robert/bin/donkeymultiplestart & (that's all in one line..) and in /home/robert/bin/donkeymultiplestart I put: #!/bin/bash cd /home/robert/download ./donkey_s ./donkey_s ./donkey_s ./donkey_s ./donkey_s ./donkey_s ./donkey_s Of course with a 'while' I could make it run indefinitely, but it crashes less than once a week anyway... my /home/robert/.vnc/xstartup contains the following: #!/bin/sh exec /etc/X11/Xsession -Icewm in /home/robert/bin/beepsound I have: #!/bin/bash /usr/bin/beep -f 830.61 -l 150 -n -f 698.46 -l 75 -n -f 830.61 -l 75 -n -f 932.33 -l 300 -n -f 1 -l 150 -n -f 622.25 -l 75 -n -f 698.46 -l 75 -n -f 830.61 -l 150 -n -f 932.33 -l 150 -n -f 1244.51 -l 150 -n -f 1396.91 -l 75 -n -f 1046.50 -l 75 -n -f 1244 .51 -l 150 -n -f 1396.91 -l 150 -n -f 932.33 -l 600 -n -f 830.61 -l 150 -n -f 698.46 -l 75 -n -f 830.61 -l 75 -n -f 932.33 -l 300 -n -f 1 -l 150 -n -f 622.25 -l 75 -n -f 698.46 -l 75 -n -f 830.61 -l 150 -n -f 932.33 -l 150 -n -f 1244.51 -l 150 -n -f 139 6.91 -l 75 -n -f 1046.50 -l 75 -n -f 1244.51 -l 150 -n -f 1396.91 -l 150 -n -f 1567 .98 -l 600 -n -f 1567.98 -l 150 -n -f 1760.00 -l 75 -n -f 1318.51 -l 75 -n -f 1567.98 -l 300 -n -f 1 -l 150 -n -f 1318.51 -l 75 -n -f 1174.66 -l 75 -n -f 1046.50 -l 150 -n -f 987.77 -l 150 -n -f 880.00 -l 225 -n -f 987.77 -l 75 -n -f 1046.50 -l 150 -n -f 1318.51 -l 15 0! ! -n -f 1174.66 -l 600 -n -f 1244.51 -l 150 -n -f 1396.91 -l 75 -n -f 1046.50 -l 75 -n -f 1244.51 -l 300 -n -f 1 -l 150 -n -f 1046.50 -l 75 -n -f 932 .33 -l 75 -n -f 830.61 -l 150 -n -f 783.99 -l 150 -n -f 698.46 -l 225 -n -f 783.99 -l 75 -n -f 830.61 -l 150 -n -f 932.33 -l 150 -n -f 1046.50 -l 150 -n -f 932.33 -l 150 -n -f 783.99 -l 150 -n -f 622.25 -l 150 (again, all on one line)

bvc, a new user: haven't tried, but not likely to help. Here's why: after installing the drivers (again), I can start mdkkdm as root, whereas before I can't. Since it's as root, a new user would not make a difference, mdkkdm should run as root. Also, as root before I cannot start X by doing: X but after reinstalling I can... There must just be a small setting somewhere.. James, thanks for thinking with me. /etc/modules.conf does have alias /dev/nvidia* nvidia and I did do modprobe nvidia which gives me an error as mentioned above. So I get stuck there. On the modes: I am using quite a few modelines; and no I am not using 'default' but specified resolutions. In the logfile it clearly shows that XFree never gets around to trying the modes, since it complains about no screens found. I guess this may have to do with module nvidia not being able to load... This last part is where it goes wrong; why would this module not be loaded but be 'in use' / not loadable, and how can I see which process is trying to do that? Is there any way I may try to force the loading?

I tried a bit more yesterday evening, like taking out the /dev/nvidia* part from /etc/security/console.parm in the dri section, chmod 644 /dev/nvidia* and chown root /dev/nvidia,... all to no avail. Anyone any ideas? I must just be missing something small, since reinstalling gets X up and running without any problems...

Swap should definitely be mounted (less /etc/fstab to see what I mean), and you don't always have to reboot after partitioning, I know at times I didn't have to. Basically, with a new drive it seems you don't have to reboot after partitioning, with a drive you deleted and created partitions on usually you do.

Ok, weird stuff: I have the same problem description that the topic starter has, but not the same solution... What I did to install the new 4496 drivers (the drivers I was using are the last ones that came with rpm's, forgot the number) was just start with the sh NVIDIA-Linux-x86-1.0-4496-pkg2.run (after doing ctrl-alt-F1 to go to console from mdkkdm login manager, logging on as root and telinit 3 to go to text mode only) It went fine, here's the log: -> License accepted. -> There appears to already be a driver installed on your system (version: 1.0- 4496). As part of installing this driver (version: 1.0-4496), the existing driver will be uninstalled. Are you sure you want to continue? ('no' will a bort installation) (Answer: Yes) -> A precompiled kernel interface for kernel 'Mandrake Linux 9.1 kernel 2.4.21-0.13mdk i586' has been found. executing: 'cd ./usr/src/nv; /usr/bin/ld -r -o nvidia.o precompiled-nv-linux .o nv-kernel.o'... -> Kernel module linked successfully. -> Installing classic TLS OpenGL libraries. -> Parsing log -> done. -> Validating previous installation: -> done. -> Uninstalling NVIDIA Accelerated Graphics Driver for Linux-x86 (1.0-4496): -> done. -> Uninstallation of existing driver: NVIDIA Accelerated Graphics Driver for Linux-x86 (1.0-4496) is complete. -> Installing 'NVIDIA Accelerated Graphics Driver for Linux-x86' (1.0-4496): executing: './usr/src/nv/makedevices.sh'... executing: '/sbin/ldconfig'... executing: '/sbin/depmod -aq'... -> done. -> Driver file installation is complete. -> Running post-install sanity check: -> done. -> Sanity check passed. -> Installation of the NVIDIA Accelerated Graphics Driver for Linux-x86 (version: 1.0-4496) is now complete. Please update your XF86Config file as appropriate; see the file /usr/share/doc/NVIDIA_GLX-1.0/README for details. and after deinstalling old stuff, installing the new, I could telinit 5 and it would start the login manager again. BUT: after shutting down and rebooting (the next day), I had no X. After fuzzing about a bit, I just thought, heck, I'll reinstall... And that went fine, the same as before, so I could use X again. Next day: again, no X. Reinstall, all is fine. So for your entertainment, here's the output of lsmod; the Xlogfile (which I also have) just says no screens found... modules loaded: Module Size Used by Tainted: PF lirc_i2c 4256 1 lirc_dev 9112 1 [lirc_i2c] bttv 78496 2 pwcx-gcc32 87552 0 (unused) pwc 45064 0 [pwcx-gcc32] parport_pc 25096 1 (autoclean) lp 8096 0 (autoclean) parport 34176 1 (autoclean) [parport_pc lp] snd-seq-midi 5024 0 (autoclean) (unused) snd-emu10k1-synth 5276 0 (autoclean) (unused) snd-emux-synth 31644 0 (autoclean) [snd-emu10k1-synth] snd-seq-midi-emul 6780 0 (autoclean) [snd-emux-synth] snd-seq-virmidi 4572 0 (autoclean) [snd-emux-synth] snd-seq-oss 31104 0 (unused) snd-seq-midi-event 5640 0 [snd-seq-midi snd-seq-virmidi snd-seq-oss] snd-seq 42608 2 [snd-seq-midi snd-emux-synth snd-seq-midi-emul snd-seq-virmidi snd-seq-oss snd-seq-midi-event] snd-pcm-oss 43556 0 (unused) snd-mixer-oss 14488 0 [snd-pcm-oss] snd-emu10k1 69268 0 [snd-emu10k1-synth] snd-rawmidi 17600 0 [snd-seq-midi snd-seq-virmidi snd-emu10k1] snd-pcm 77536 0 [snd-pcm-oss snd-emu10k1] snd-timer 18376 0 [snd-seq snd-pcm] snd-page-alloc 7732 0 [snd-emu10k1 snd-pcm] snd-util-mem 3008 0 [snd-emux-synth snd-emu10k1] snd-ac97-codec 40160 0 [snd-emu10k1] snd-seq-device 5832 0 [snd-seq-midi snd-emu10k1-synth snd-emux-synth snd-seq-oss snd-seq snd-emu10k1 snd-rawmidi] snd-hwdep 6368 0 [snd-emu10k1] snd 40868 0 [snd-seq-midi snd-emux-synth snd-seq-virmidi snd-seq-oss snd-seq-midi-event snd-seq snd-pcm-oss snd-mixer-oss snd-emu10k1 snd-rawmidi snd-pcm snd-timer snd-util-mem snd-ac97-codec snd-seq-device snd-hwdep] nfsd 74256 8 (autoclean) af_packet 14952 0 (autoclean) w83781d 21872 0 (unused) i2c-proc 9072 0 [w83781d] i2c-viapro 4848 0 (unused) sr_mod 16920 0 (autoclean) floppy 55132 0 8139too 17160 1 (autoclean) mii 3832 0 (autoclean) [8139too] nls_utf8 1312 3 (autoclean) nls_cp437 5148 3 (autoclean) vfat 11820 3 (autoclean) fat 37944 0 (autoclean) [vfat] supermount 15296 4 (autoclean) ide-cd 33856 0 cdrom 31648 0 [sr_mod ide-cd] tuner 11744 1 (autoclean) tvaudio 15068 0 (autoclean) (unused) msp3400 17804 1 (autoclean) i2c-algo-bit 9064 1 [bttv] i2c-core 21192 0 [lirc_i2c bttv w83781d i2c-proc i2c-viapro tuner tvaudio msp3400 i2c-algo-bit] soundcore 6276 0 [bttv snd] ide-scsi 11280 0 scsi_mod 103284 2 [sr_mod ide-scsi] NVdriver 1667200 0 usbmouse 2936 0 (unused) keybdev 2720 0 (unused) mousedev 5268 0 hid 20900 0 (unused) input 5664 0 [usbmouse keybdev mousedev hid] videodev 7872 4 [bttv pwc] printer 8448 0 (unused) ehci-hcd 18568 0 (unused) usb-uhci 24652 0 (unused) usbcore 72992 1 [pwc usbmouse hid printer ehci-hcd usb-uhci] rtc 8060 0 (autoclean) ext3 59916 8 jbd 38972 8 [ext3] I just also tried to reboot after doing a depmod -qa, didn't work... Then did a modprobe nvidia, which failed (module used.. but not loaded..) No clue on what else to try...

bvc, I think some SiS systems have only one chip in their chipset, so technically just a chip not a set... and since all interconnects are within the same die, they can have a much higher bandwidth... Apart from that, I don't really know. Don't worry about it though, the lower of the two numbers is the more important one.