Turb0flat4

But can one do that with an installed mplayer binary (installed with urpmi) ? I always thought I need to compile from source for new codecs - not true ? I was using xine as an alternative - seeing if I could get the xine win codecs working with xine to view the file. Unfortunately I couldn't. EDIT : I think it's the files themselves that are corrupted. I can view other .wmv files off the web. BTW, Ian is right : now I can view wmv files with mplayer - apparently the codecs are accessible by mplayer too ! That's cool.

I just tried urpmi xine-win32 but the new codecs don't work. When I play the files with xine, the video is all scrambled. I know mplayer win32 codecs work, but under this distro, I am unable to get them working. :(

I need to play a couple files with extension wmv (Win Media 9 encoding). The mplayer I got with urpmi mplayer cannot play these. How do I install the needed Win32 codecs into the rpm based install ? I tried compiling from source but the gcc version I have (4.0.1-5mdk) is not supported by MPlayer developers who claim to only support gcc 2.95 and gcc 3.x I tried compiling with gcc check disabled in the configure script, but there were errors on make. So I can't install this program this way. These were the errors : In file included from libpostproc/postprocess.c:655: libpostproc/postprocess_template.c: In function 'do_a_deblock_MMX2': libpostproc/postprocess_template.c:2901: error: memory input 4 is not directly addressable libpostproc/postprocess_template.c: In function 'postProcess_MMX2': libpostproc/postprocess_template.c:3510: warning: pointer targets in passing argument 6 of 'blockCopy_MMX2' differ in signedness libpostproc/postprocess_template.c:3656: warning: pointer targets in passing argument 6 of 'blockCopy_MMX2' differ in signedness libpostproc/postprocess_template.c:3757: warning: pointer targets in passing argument 4 of 'tempNoiseReducer_MMX2' differ in signedness libpostproc/postprocess_template.c:3781: warning: pointer targets in passing argument 4 of 'tempNoiseReducer_MMX2' differ in signedness make[1]: *** [libpostproc/postprocess.o] Error 1 make[1]: Leaving directory `/home/deepak/apps/tmp/MPlayer-1.0pre7try2/libavcodec' make: *** [libavcodec/libavcodec.a] Error 2 OK, so is there any way I can view Windows Media video with this thing ? Thanks, appreciate the help.

I have the winning argument here. Hanes, would you please post your IP ? Yeah, didn't think so.

Troll, troll, troll. Being passwordless for root is a *terrible* idea. One of the reasons I tired of Ubuntu was that nearly everything could be done with sudo and a non-privileged user password. That's bad enough. This is infinitely worse. BTW, don't use WinXP to bolster your feeble arguments. True, by default it's setup that way, but it's the easiest thing to set a password. Even better would be to make an additional limited user account to do day to day stuff. In any case, Windoze XP should not be used as a benchmark for anything. The last OS where Redmond got nearly everything right in an OS was Windows 2000 Pro. It hadn't happened before, and it hasn't happened since. WIn2k had *excellent* security policies *by default* - you HAD to login, and they even encouraged you to use a reserved key combo (Ctrl-Alt-Del) to enter the login screen. You could even change the name of the Admin account from the default to something non-intuitive for added security, something that (as far as I know), still cannot be done in Linux. And you couldn't switch users "on the fly" like you can with WinXP. That was a brilliant security policy, it's how it should be. But lusers like you seem to want insecure machines as long as they're "easy" to use, so I guess that's why WinXP is done the way it is. That's not a good thing though !

I let it do it automatically. It did it correctly. I didn't want to do it manually because I couldn't remember the location of the new config file - it's no longer at /etc/X11/XF86Config, I think. EDIT : It's at xorg.conf in the same directory. Things have changed since I used Mandrake two versions ago.

Yup, you're right. No need to make executable if using "sh". Learnt something new. :D

Hold on, was your NVIDIA installer script made executable ? It isn't by default. Go to the directory you d/led the file and type "chmod a+x NVI*run". That should make it executable. Then make sure you have kernel sources for your installation. For this you need to run as root, "urpmi kernel-source", then press enter to resolve any dependencies. Note that for some reason the latest version of the NVIDIA drivers (1.0-8756) does not work with kernel 2.6.12-12mdk, you will not be able to startx. If you are using this kernel, get the previous edition. On the whole, it is better to upgrade your kernel to 2.6.12-18mdk as I did today, you can do it with urpmi also, post if you need help. Assuming you're using the right version of drivers for your kernel, now to go nongraphical mode "init 3" , login as root, and "cd" to the directory where you had the NVIDIA file. Then just type "sh NVI*run" or I think even "./NVI*run" should work as well. This is provided you only have one version of the driver in that directory, make sure of this to avoid problems with the wildcard. Alternatively, you can type the whole name of the file out, using tab completion as needed. Then just click accept and allow it to recompile a driver for the kernel (this is why it needs kernel-source). Allow it to modify your X config files for you. After it's done, type startx and pray. ;)

Upgraded to 2.6.12-18mdk and the latest version (1.0-8756) installed without a hitch.

I have another problem printing via SMB printing to a Windows 98 PC on my LAN which has a HP Deskjet 845c printer connected via USB. I've installed Samba, etc. and tried configuring the printer every which way. I used the web config for CUPS (http://localhost:631), as well as the Mandrake Control Center and finally the "kups" printer wizard (which wasn't even properly documented). At first the HP printer remained completely unresponsive, but after the final try with Kups, I managed to get it going. However, the test page printout wasn't successful - it prints out formatting type text on the first and last pages of a sheaf of blank papers. This sounds more like a driver problem. I've tried the recommended "foomatic" driver as well as the CUPS/Ghostscript driver to no avail. I even installed the hplip rpm (urpmi hplip) but didn't know what to do with the installed files, since no new driver files became available on kups etc. I'd really like to get this SAMBA printer working with my linux box. Please help ? Thanks.

What kernel are you using ? I am using Mandriva Free 2006 and kernel 2.6.12-12mdksmp. I could not get the latest version of the driver working on my system - X wouldn't start. I used the previous version and it's OK. Could be a buggy driver or some sort of kernel incompatibility. Checking the config files etc. didn't help in my case, it was a more fundamental problem. There's no other reason for the old driver working while the new one didn't.

Probably true that it wouldn't work perfectly but it does work - the output is tolerable. But this is not a print driver problem - this is a CUPS/spooling problem. When I cancel a print job thru "cancel" or "lprm" something should be telling CUPS to stop sending the job out to the printer. This is not happening.

Apparently, others have been experiencing very similar things. These guys also use kprinter and CUPS. http://lists.kde.org/?l=kde-print&m=110684553922160&w=2 Almost identical symptoms. The issue is still unresolved, with people saying one needs root privileges to kill a print job. This is very odd. EDIT : More corroboration : http://www.leuf.net/ww/wikidn?FlushPrintQueueData "parallel" was one of the dependent processes so the same "pkill parallel" as root would have worked here too.

Whoa, this is bad. I just tried printing 3 pages of a pdf. Midway thru the first page, I cancelled the job from the gui in Document Viewer (which had the pdf opened), and then in an opened konsole, I did "lpstat" and cancelled the job. Guess what ? The computer kept printing all the way to completion to 3 pages ! No stopping it, I would've had to power off to cancel the job immediately. I did "ps aux | grep parallel" and this is what I found : "root 20385 69.8 0.0 3076 852 ? R 16:41 0:18 parallel:/dev/lp0 9 user 1 /var/spool/cups/tmp/9-user-cupsjob20374" If I had wanted to kill the job immediately, I would've had to become root and pkill parallel again, after which I would've had to power cycle the printer. I can't figure out why a cancelled job should keep printing. I'm aware the file has already been spooled, but surely there should be a process in place to stop that from going out to the port ? EDIT : It's definitely something to do with the spooling. Once the job has been offloaded from lpstat to /var/spool/cups/ it cannot be stopped with cancel. Only pkilling parallel has any effect. :(

"cancel" kills the jobs (they disappear from lpstat) but the printer keeps printing for a while. To save paper and toner, I am forced to use the soft switch to discontinue printing immediately. This was never a problem on Windows - I would power off the printer, cancel the print job and when I powered the printer back on, it would be totally quiescent and ready to accept new jobs. It's not the printer buffer, of this I am fairly sure. If you reread what I wrote, I've tried powercycling (including disconnection from supply) many times to no avail. It's a Linux problem (my guess being specifically a CUPS problem) - there were some running processes that were insidiously sending data to the printer without appearing on lpstat. I got them out with ps aux, as I detailed in the post above. But that doesn't change the fact that this shouldn't have to be done, there should be a cleaner way to handle a terminated print job. I'm using KDE. But I prefer the command line in a pinch.

No go, I'm afraid. :( I actually know how to use the localhost:631 interface to CUPS, but the Mandriva Control Center Wizard has the same functionality. I've tried removing and adding the printer to no avail. It can't be a print hardware buffer problem - because powering it off and disconnecting it from the mains for a couple of minutes has no effect. So I'm guessing it must be something on the PC sending it spurious data. Would lpstat miss anything ? Would I be able to tell if something is sending rubbish data to the printer port ? ------------ OK, I've managed to solve the problem for now. But this is an ugly workaround and I can't figure out why I would need to do it. I went to console, became root and did "ps aux | grep lp" (grepping for "parallel" would also work). I saw a TON of processes listed starting with parallel:/ blah blah.. printing to /dev/lp0. None of these were showing up on lpstat, even as root. I just "pkill parallel" , power cycled the printer one more time and everything's OK. But why don't these rogue jobs show up under lpstat ? And don't tell me I need to pkill processes every time I stop printing midway thru a job ?!

Hi. I'm on Mandriva 2006 and have managed to get most things working right. But I have a problem with printing. I can send jobs OK, but when I want to cancel the job immediately I have problems. Printer is Canon Bubblejet BJC255SP on parallel port (/dev/lp0) using distro included Ghostscript drivers. This is what I do to start off jobs : either thru CL : lpr (job), or print thru a GUI from an app. When I want to stop the job : turn off the printer with the soft switch : the printer will stop and eject out the page. Then I go to CL and type lpstat to list jobs and cancel the job. When I turn on the printer after doing that, it just starts printing nonsense characters in a compressed font - a few chars a page, then form feed. Nothing helps - on the software side I've tried restarting CUPS, using the Mandriva Configuration wizard to disable and re-enable the printer. On the hardware side, I've tried power cycling the printer, disconnecting the data cable, etc. Nothing helps, the moment I put a new piece of paper in, the printer is happily churning out pages of garbage. Webmin and lpstat show no print jobs at this stage. What to do please ? I don't want to reboot, nor should I have to. Thanks and appreciate the help. :)