adamw

We lost #1 rank in the DW six month chart around 2005 or 2006, if I recall correctly. I wouldn't say DW has ever been 'biased' against Mandriva, in the sense of determined to judge it (and us) negatively no matter what the evidence. I would say that Ladislav considered some of the moves Mandrake/Mandriva made around that same time (2004 through 2006) to be bad ones, and said so, which is of course his prerogative on his own site. It's never been the case that DW is intended to be some utterly impartial fount of inarguable wisdom. It's as much an opinion site as an information one. In the last couple of years we (myself and others at Mandriva) have had good contact with Ladislav, and he - as far as I can judge - now considers that the company and distro are both moving in more positive directions, and his coverage since then has reflected that opinion. In general what happened to Mandriva was a) Ubuntu, b) OpenSUSE, c) bad management of the user community, and d) an iffy string of releases. The first two were nothing we could do much about, but the last two were. There was a huge sea change in the Linux world around 2004 - 2005, when Ubuntu showed up, and Novell bought out SUSE and opened it up. Prior to that, the big names in the end-user distro world were Red Hat, SUSE, and Mandrake. And Gentoo for a bit, but that was more a fad than anything else. Debian was where it always is - a little out to the side. I'm ignoring it for the purpose of this discussion. What Red Hat, original-SUSE, and Mandrake all had in common is they were independent companies producing commercial Linux distributions and depending on them to survive. Red Hat and Mandriva still are, but Mandriva is the only survivor of that initial trio which still relies to any significant extent on selling products to individual end-users. Red Hat has built a massive, and very successful, business selling thousands of units of corporate products to large customers. That's what they do now, and they're great at it. They don't sell an end-user distro any more. They sponsor the development of Fedora, I believe, for three reasons: a) for the development benefits it provides, b) for mindshare in the still to some extent enthusiast-driven Linux community, and c) because, honestly, I think the leaders of Red Hat are decent people who remember where the company came from and still think it's a cool thing to make a good desktop product. Anyhow. Point is, they're not in the same market as us any more. That was the first big change. The next one is what happened to SUSE. SUSE, when it was independent, was the closest competitor to Mandrake, but Mandrake was bigger in most places (SUSE always had a strong foothold in Germany, naturally). And SUSE was actually more restrictive and closed than Mandrake ever was. SUSE development was almost entirely closed and done within the company, there was little external input. They didn't really have a development testing community. They didn't even release free ISOs; you either bought SUSE, or went through a fairly painful network installation process to install it from the public repos. Novell bought it out and completely changed it. For Novell, desktop Linux is not a money earner, it's a loss leader. They don't make money selling SUSE boxes, although they still do it just because a few die hards want them. They spend far more developing OpenSUSE than they make back selling anything to regular end users. Why? More or less the same reasons as Red Hat, actually. They use it as a development project for SLES and SLED, which they do make money selling (or, at least, they genuinely try to, it's one of their business goals). It also builds mindshare around their other enterprise products - it gives them an in with people. Anyway, why is less important than what. As I said, they totally opened up SUSE. It went from being the most closed commercial Linux to being more open than Mandrake, overnight, because Novell didn't need it to make money. They needed it to make a big splash. And finally, Ubuntu came along. Ubuntu's initial business model was (and mostly, still is) 'spend Mark Shuttleworth's money'. I'm not complaining, but making an important point. Just like OpenSUSE, Ubuntu didn't need to make money. The point of Ubuntu was more or less the opposite - it was to do as much as possible and give it all away for free. To as many people as possible. So the leading Linux desktop distros went from Mandrake, Red Hat, and independent SUSE - of which Mandrake was the most popular and the most open, having an open development process, and releasing free ISOs to the public on a short delay after the Club / PWP release - to Mandrake, Ubuntu, and OpenSUSE, of which Mandrake was suddenly the *least* open. That didn't make us look too good by comparison, but it's a big change to have to try and adjust to overnight. What we had to adjust to was a fundamental shift in the market. Prior to Ubuntu and OpenSUSE, people would pay to run Linux. You could, fundamentally, make money selling software to Linux users. We and SUSE had both done it, successfully. Once Ubuntu and OpenSUSE were around, it became very difficult. It's hard to sell someone a $99 distribution when they can get pretty much the same stuff elsewhere, for free. This was only made worse by the fact that this change unfortunately coincided with some, frankly, sloppy releases of Mandrake/Mandriva. 10.2, 2006, and - to some extent - 2007 were not very good releases, in my opinion (which was my opinion at the time as well as today, and that's on public record in various places). They all released with significant bugs. Our quality control was not great. This only made the comparison with the shiny new Ubuntu and OpenSUSE - and other distros around at the time - worse. Suddenly, instead of a choice of three commercial distros, people had a choice of one somewhat sloppy distro which we were still actively trying to get people to buy, or two that were basically totally free, and in some ways were, honestly, better. If you grab an emulator and compare, say, Mandriva 2005 with Ubuntu 5.04, or Mandriva 2006 with Ubuntu 5.10, you'll see what I mean. Early Ubuntu releases lacked a whole ton of features - all sorts of stuff Mandriva can do - but at the time, our presentation and release quality were slapdash, and there were some significant differences from the Mandriva of today. We still didn't have a public non-free repository; only PWP owners or Club members got the proprietary packages - like NVIDIA and ATI drivers, wireless firmware, Flash, Java...all those things you could get very easily for Ubuntu. Those Ubuntu releases were very minimalistic but they also had very high release quality. They shipped a great looking and great feeling product that simply didn't have obvious bugs. There were lots of things it couldn't do, but it just felt nicer than the Mandrake releases of the time. And finally, as I said, there was the user community management issue. Again, at the time, Mandrake/Mandriva still basically had a commercial company mindset. We still distinguished substantially between customers and users, and treated them differently. This was nothing unusual at the time, and we were less harsh about this than Red Hat or independent-SUSE. But as soon as Ubuntu and OpenSUSE showed up, our position started to look very bad. People realized that the distinction didn't happen with OpenSUSE or Ubuntu (or some other distros). Everyone was a user; someone who didn't pay was valued as much as someone who did (of course, no-one COULD pay for Ubuntu). So, yeah, the history is really pretty simple; there was a big confluence of factors around 2004-2005 that changed the Linux desktop sector completely, and we were in pretty much the worst position out of anyone who survived (lots of companies went under around then). We had to figure out a way to be an independent company which exists in order to produce a distribution for regular people, in a sector where it didn't make any sense to pay for a desktop Linux distribution any more. That was, and is, tough, and is basically the root of all the problems Mandriva had at the time. I don't think there was anything particular unfair in the public response to Mandriva, however tough conditions were and still are for us. The fact is, it doesn't matter to a user what the market conditions are. People were just looking at the products they had available and drawing a conclusion that Mandriva was no longer the best for them. Which was perfectly rational, for a lot of people.

Quanta is part of kdewebdev and always has been, that's not changed. That's how KDE have it organized upstream, we can't really change it.

I rather enjoyed searching for "Mandrake Linux" and then for "Ubuntu". *small sigh* ah, good times...

It's *commercial*, not *proprietary*, packages that are exclusive to PWP. The difference is important. Software that's not Open Source or Free Software but which we can legally distribute to the general public is all free on Mandriva, it's in the public /non-free repository, and the important packages of this type are included in One. The only packages that are exclusive to PWP are ones which are not only not Free Software / Open Source, but whose license does not actually allow us to distribute them to the public - we have negotiated with the publishers for the right to include it on the PWP, but we can't legally just stick it up for anyone to download. There's a more in-depth explanation at: http://wiki.mandriva.com/en/Docs/Choosing_the_right_edition

medo's advice is important - you have to install the proprietary driver from /backports to support the 9600, as it wasn't available when 2008 Spring was released, so the version of the driver in the /release repositories doesn't support the 9600.

You already can, it's been out for months. :) It's the current release. You can get it at www.mandriva.com/download . The new release, 2009, is due out soon (Oct 9th is the target date). That should work out of the box as well.

Have you tried One?

I don't think that was actually the case in 2008 (not 2008 Spring), which is what this user is on. I think it may actually just be that the proprietary driver in 2008 is too old for the card. Is there a particular reason you want to use 2008, not 2008 Spring?

There may be some other heuristic trying to detect server hardware, I wasn't aware of it, though. I don't know which bit of the installer code it's in, off hand. Sorry.

How did you 'update the drivers' exactly?

I agree with dude67 that unless you use *very* little electricity, it's almost impossible for a single computer to double your usage no matter how inefficiently it's run. A typical desktop PC will consume 200W or so of power, maybe 300. A typical laptop will consume less than 100W. Even if it never does any power saving and consumes the maximum amount *all the time*, that's a lot less than half a typical person's power budget.

RAM. 4GB or more of RAM gets you the server kernel. This was previously applied to x86-64 as well as i586 but it doesn't actually make sense, so in 2009 this will only be the case on i586.

Well...hopefully not so much now we're at RC2. :)

As the bug report says, we can't really support the use of One transformed by unetbootin. Can you try it directly from the CD with an external CD drive, or doing a network install via all.img or something? thanks. I believe the Wind is expected to mostly work with 2009...

As ffi says, you should just install the packaged versions of wine and fluxbox. When you install a packaged window manager or desktop, it will be added to the sessions list automatically (we package all the necessary files, obviously).

You didn't actually explain what the problem is.

Rhythmbox should do it automatically, if it knows the format the player requires. I know how this is handled for USB mass storage players but I'm not so sure about MTP players. So what happens at the moment - it just transfers it in the original format? Oh, wait, never mind. I found it. it's handled just like USB mass storage players. If you're interested, look at /usr/share/hal/fdi/information/10freedesktop/10-usb-music-players-libmtp.fdi and search it for the string '818'. You can see the stuff for your player there. It does look to be set to transcode to MP3 where appropriate, though. What format was the file you transferred to the device?

Rhythmbox and amaroK can both handle MTP-based players fine. Just make sure the libmtp packages are installed. For Rhythmbox that's all you need to do: run it, plug in the player, and you'll see the player in the list on the left of the screen, and you'll be able to drag and drop songs to and from it (if the songs aren't already in MP3 format you'll need the gstreamer0.10-lame package to enable Rhythmbox to transcode them on the fly). For amaroK I think you also need to go to the configuration screens and enable a plugin for mtp support.

If it offers you a proprietary driver, and you install it, and the card seems to be working fine, stick with that. If you don't get offered a proprietary driver, then as medo says, get the newer version from /backports following the instructions on that page. I am being vague because I don't quite remember whether the driver we initially shipped with 2008 Spring supported the 9800 GTX. I *think* it did but I may be wrong.

The package you're looking for in the official repos is rpm-build . Why are you looking for it, though, out of curiosity? Are you interesting in creating your own packages? Edit: I think you would benefit a lot from reading this: http://wiki.mandriva.com/en/Docs/Basic_tas...moving_software It tells you a lot about the Mandriva package repositories and package management tools. I think you will understand a lot better how to do package management tasks on Mandriva after reading it. :)

I have this page up on the Wiki for installing backport drivers: http://wiki.mandriva.com/en/Updating_propr..._from_backports as far as I'm aware, the procedure outlined in it works in 2008 Spring (and will work in 2009) in all normal situations.

I think possibly lilo / UUID has not been properly tested. It would be fruitful to file a bug on this. I also think people are overstating the readability issue. A UUID-based grub or lilo config or fstab is not at all unreadable. It's just not what you're used to. UUIDs are not that exotic, they're just a random identifier. I use the same technique for identifying them as I do when manually checking md5sums - just look at the last three characters and remember those. You're very unlikely to have two disks whose UUIDs end in the exact same three characters. Any time you actually need to *enter* the string you can just copy / paste it. There are several tools or just commands (look in /dev/disk/by-uuid ) you can use to identify drives by their UUIDs. There's nothing about UUIDs that makes them non-readable or impossible to manipulate using your brain and a text editor, just like the old hd*/sd*.

If you're waiting for it to finish, you'll be waiting a long time. :) It cycles, it just runs over and over. If it shows no errors after an all-night cycle then your RAM is fine. Just stop it and reboot.

Ah, sorry, I missed the space requirement. You could squeeze a Mandriva install into 200MB but it wouldn't be able to *do* much. :) In that case, yes, a specialist distro is the best idea, but you may well need to build openchrome yourself for decent graphics support.

task-kde4-minimal requiring openoffice.org-core is actually a bug caused by over-enthusiastic provides in openoffice (really, it depends on libraptor.so.1 and libredland.so.0 , which openoffice.org-core claims to provide but really shouldn't, as it has private copies of them that the rest of the system cannot use). I will fix this, or get our OO.o maintainer to fix it.