aRTee

Ian, you're good!! One day, you're going to have to explain what exact search terms you used... Anyway, thanks a lot! Also thanks to speedball2, I would have gone that route (wine and win software) if cdemu hadn't worked so nicely. How to install cdemu on Mandriva (2009.1 64 bit in my case) Ok, so indeed cdemu works fine, all that was needed was (as root): urpmi cdemu -a and then I had to make sure the dkms module of vhba got installed - the urpmi command did install one precompiled vhba module, but not for my current kernel. It actually installed a slightly older kernel too, which I later removed. So I did (still as root): urpmi dkms-vhba which installed and compiled the vhba kernel module for my kernel. How to use cdemu Then I loaded the module: modprobe vbha which created a device /dev/vhba_ctl and started the cdemud daemon: service cdemud start No config file editing necessary. At that point, the kernel message scrolled past (tail -f /var/log/messages): May 6 23:08:02 zurich cdemud[2862]: Starting daemon in daemon mode with following parameters: - num devices: 1 - ctl device: /dev/vhba_ctl - audio driver: null - bus type: system May 6 23:08:02 zurich klogd: scsi 9:0:0:0: CD-ROM CDEmu Virt. CD/DVD-ROM 1.10 PQ: 0 ANSI: 0 May 6 23:08:02 zurich klogd: sr1: scsi3-mmc drive: 40x/40x cd/rw xa/form2 cdda tray May 6 23:08:02 zurich klogd: sr 9:0:0:0: Attached scsi generic sg2 type 5 and I checked to find: # ll /dev/sr1 brw-rw----+ 1 martina cdrom 11, 1 2010-05-02 21:19 /dev/sr1 The vbha module loading may be done automatically, didn't try to start clean yet. I also don't know yet if the cdemud will start automatically after rebooting, but I'll find out when that time has come. I would expect so. So then as regular user (since the device was owned by the regular user) I did: $ cdemu status Devices' status: DEV LOADED TYPE FILENAME 0 0 N/A N/A followed by: $ cdemu load 0 /mnt/neuchatel/giant/home/donkey/Globi_Seefahrer.nrg and checked to find: $ cdemu status Devices' status: DEV LOADED TYPE FILENAME 0 1 PARSER-NRG /mnt/neuchatel/giant/home/donkey/Globi_Seefahrer.nrg After that I started grip, and changed the cdrom device to /7dev/sr1, restarted it (it didn't want to see any contents, or perhaps I was not patient enough in my excitement), and it detected happily that there was a cd in the emulated drawer, got the track names from cddb and ripped it at high speed. After finishing, grip happily ejected the emulated disc, so I found: $ cdemu status Devices' status: DEV LOADED TYPE FILENAME 0 0 N/A N/A For gnome users there's the applet gcdemu, but I'm on kde, and perfectly happy with the CLI, so no problem for me. It would be nice for non-gnome GUI oriented users if there would be some other GUI. Or if anyone has an idea how to start a gnome applet from within kde (4), please post below.

Thanks Ian, but unfortunately that doesn't work... I forgot to mention I had also found that, but an audio cd cannot be mounted, it's not iso9660 but IIRC redbook. Also, yes, Nero can burn to an image,.. but that's an .nrg image.. which I have plenty of already. So I'd really like to find a solution which has some kind of simulated/virtual cd burner.

I'm trying to use nero .nrg audio cd images. Friends of mine have scanned their collection to .nrg files, but those are sadly not usable (also not on Win or OSX) other than to burn to cd, and they are non-compressed (or at least, they seem to be from the file sizes). So I've explained it's better to rip to flac, but it would be nice if they could convert their current rips to flac as well. I tried and managed using Nero for Linux, but I need to burn each .nrg to a cd (rw of course), then rip that one, then burn the next. So the speed and the quality both have me worried a bit. Plus, it wears out my burner. Is there any way to create a virtual cd burner under Linux? That way I can use nero to 'burn' the .nrg files to the virtual drive, then rip it as usual. Please note the following: I have not found any program that can use audio .nrg files under Linux. I have looked at nrg2iso, but that works for data discs only. There are some other similar things that I found which also didn't work, I forgot the names but can find those again, if necessary. I also looked at audiofs and cdfs, which seem to be unmaintained since a while, in any case, I couldn't find any way to employ those. [moved from Software by spinynorman]

I think I'm slowly starting to get the activies' purpose.. and why I didn't get it nor need it... I guess it's for people who are using plasmoids / widgets heavily. I only use gkrellm since I never found a good substitute, and otherwise the panel (clock / calendar). But people who need more space for widgets need these activity plains, I think if you use various programs (in windowed form), you need multiple desktops, or at least get benefit of that. If you use many many widgets, it is nice to have more space for them, and activities can do that. So: multiple desktops for multiple windowed apps, multiple activities for multiple widget collections. If anyone has another point for this, I'm all ears...

It's not very important when you sync, just do it before unmounting and wait for the disk light to stop burning. That's why I put the s double, I actually keep hitting s until the light doesn't come up anymore... You could do r s e s i s s u b or so. Whatever you do, make sure the sync is done.

I'm using KDE 3.5 on the main system (well, my wife uses it) - no way to move to KDE4 until each virtual desktop can have its own background. So I tried 2010.0 beta, and Lo and behold, individual images per virtual desktop. Lots of other interesting things, and: normally no icons. Contrary to most, I find this worthy of a "Hooray!" One thing has me puzzled though: the zoom out and multiple activities functionality. I've searched the web, read lots of things about it, such as this:http://temporaryland.wordpress.com/2009/07/23/my-kde-4-x-desktop-activities-tutorial/ and from kde: http://userbase.kde.org/Plasma... From the latter: Well, that's fine and dandy, but I did that with regular virtual desktops (workspaces, whatever you call it) back in 1992. Minus the widgets, because in those days on HP-VUE UNIX there were none. Anyway, with KDE 4.3 without using activities, just multiple desktops, I get exactly that. With KDE 3.5 I can also get exactly that. So is there anyone who can tell me what these activities are good for that I can't just do with a (extra) virtual desktop? Because I'm just not getting it...

I've had some problems on my laptop - it would not really start/boot the system. The fix? I hit alt-sysrq-e. Perhaps that needs to be preceded with a alt-sysrq-r, not sure. Something got unstuck due to that, and after that magic keystroke it booted fine. (I actually had to do that every time on 2009.1 on this laptop. Now I'm running 2010.0 beta's and this is now fixed.) I actually wanted to shut down the system since it wouldn't move anymore. (alt-sysrq-r,e,s,s,u,i,b is the thing to do to get the system to restart when only the kernel is alive, but for instance X is dead..) Obviously I don't know if this will help in your case, but it won't hurt. If the system really doesn't move, use the sequence I mentioned, instead of hard switching / power cycling or so.

Well, I disagree. Sorry to be late to the party guys - not enough time, too many things to do. But anyway... I disagree, since without bugreports, Mandriva won't get better, and with bugreports, at least it may. And I have used Mandriva without paying a cent for several years now (I did pay a few years although I got a free VIP account due to being a translator). If you don't pay the company Mandriva, don't comment on how the company treats their people. (If you do pay them, perhaps you should pay them more before you comment...?) If you use the distribution Mandriva Linux, please do contribute as much as you can. One way is to file bugs properly. Note: all bugs I filed have been handled in decent time, as far as I can remember. I've never felt like I was doing free work for a company when I was translating, I always felt like I was doing a little bit of work and getting back the whole OS in return. Small price to pay. BTW if you don't support Mandriva (linux, the distro), then how much better will Mandriva treat the remaining employees? You complain about Mandriva laying off AdamW, but none of what you do (FWICT) leads them to be in a position where they don't have to do that again the next time things get tough.

Quick question: is it now possible to have a different background image per desktop?

Late to the party as I usually am.... Adam, good luck with your next steps, and you and I know you deserved better. As for Mandriva, I'll continue to use it as long as it's usable. Creating a fork? That will be an option if things go that way - it is not that time yet, IMHO. I love Mandriva the OS as far as I love any computing platform, more than any other computing platform. I've never really loved the company though. The developers have done an amazing job on a shoestring budget, providing quite a few 'first-time' linux common user features: first freely available distro to do ntfs resizing during installation, great tools, etc. Ubuntu is only now coming closer (not there yet, though I haven't checked in a while). As I wrote, I'll continue to use it as long as it works for me.

I used to do this, and contrary to common opinion, this can work fine. I dual booted multiple linux installations, hibernating one and then going to another, etc. Just make sure you never ever write to a partition that was mounted on the hibernated system. Not actually mounting any of them helps, but if you must read data, mount them read-only. Otherwise, no problem as far as I can see. Make sure you have two hibernate swap partitions if you want to do this with two Linux installs. Hibernating Linux then booting to Windows should not have any similar conflict. The current grub has an option to immediately boot if in hibernation / suspend to disk. There is an option, forgot which file - can check if you want - to switch back to the default behaviour. If you have 2 different Linux installations, you can set up the regular grub in the mbr to boot: linux 1 windows chainloader to linux 2 partition boot sector and in the second linux installation boot sector, you can again have all kinds of entries. Now if you hibernate the second Linux, the grub in the boot sector of that root partition gets updated to autoresume, but the grub from the other Linux in the mbr doesn't, so from that one you can still boot windows. To be able to do it with only windows and one linux install, you need to find the config file. I'll post it if I find it... Edit: got it - I couldn't find it on the web, so I figured it should be something the code for the suspend stuff must include - so I opened mcc and went into the software installer, and checked which files get installed with pm-utils one of those is /usr/lib64/pm-utils/sleep.d/01bootloader where on a i586 system lib64 is likely just lib it refers to a file /usr/sbin/rebootin which is a perl script for a single one time system boot bypassing the menu. never mind that /usr/lib64/pm-utils/sleep.d/01bootloader gives error messages for Lilo, this is for grub too. all you have to do is remove 01bootloader from /usr/lib64/pm-utils/sleep.d/ and you should be set. Again: do not mount any partitions from a hibernated system in rw mode! They have not been cleanly unmounted, and the hibernated system will expect them not to have changed - Linux partitions keep track of mount-count and such!

Perhaps I'm more sensitive to tearing and other video artefacts than others, but in any case, I can hardly find any similar comments/problem reports on the usual forums nor on the web at large... Essentially, the problem is that I see some tearing even though I have overlay functional with Xv video. I figured out (from the intel manpage, man:intel) that the problem is likely that the hardware will accept any semi functional modeline as long as the lvds output is set up to use whatever config the lvds screen prescribes. But the video graphics will still use the modeline / xorg config as xorg likes; it looks like the graphics chip will tell xorg that some range of signals is fine, but then in the end pass the buffer data in the way the screen wants it. This leads to the +vsync not to have any positive effect on the tearing (it should go away), because it's not taken into account on the final signal going to the panel, but merely in between. The problem is, I tried turning off this driver feature, but then I get a 'no display found' error from X and only console. So it seems that the regular modeline that xorg.0.log tells me it uses (with -vsync no less..) is not at all what is needed. The panel is a 1280x800 one, and obviously I have no clue as to the real timings on the thing... Any hints as to what to try? Can anyone confirm what I postulate above?

The output of your free command shows no swap is used. In other words, you have enough ram, more wouldn't help. What you want is indeed to have the first line tell you that your ram is fully used, and the last line that no swap is used, in which case all is fine, ... output from my current machine: $ free total used free shared buffers cached Mem: 4052292 3853548 198744 0 255472 2575444 -/+ buffers/cache: 1022632 3029660 Swap: 10233080 148 10232932 (Oh noes, the system ate all my ram!! ) Can you check if there is some disk activity? I noticed that kerry / beagle sometimes keep the disks busy, which is a major cause of slowness. I normally run with gkrellm on the side, just to keep track of high cpu and disk loads (and temps and stuff..). BTW flash can make a Athlon 2400+ stutter, from what I gather also on windows, so go complain to Adobe about that one.. ;) Did you get the new version of FF? Should be in the backports repos somewhere (I'm running 2008.1 x64 on all machines now, and did get FF3 on all of them, just forgot from where I got it).

For http access, there is an extra package you need: tightvnc-java I messed around a bit, so I'm not sure if there's anything else to do. Otherwise, only m repeats, but to get one 'm' character, you need to press it twice... See also: this bug Did some more hacking, and found that with this way to start, all is fine: vncserver :7 -dpi 80 -geometry 800x480 -httpd /usr/share/java -httpport 5807 +kb The java vnc package must be installed and +kb at the end is there to override the wrong default of -kb that gets used otherwise.. Hope this is useful to someone...

Hi Ian, thanks for the reply... Well, I just gave it up in the end. I did try to make a truecrypt file around the size of the partition, the first try realcrypt told me 'filesystem full'... after 4 hours of chugging... then the second time it didn't work either due to some other reason, forgot why, perhaps I forgot to format before trying to mount the realcrypt drive. In any case, I decided to forego realcrypt/truecrypt on the raid drive/partition. I did create two regular partitions where everything worked fine..