I replaced the Mepis 6 partition with Mandriva One. The laptop is a dual boot with Win XP which I need to retain because of the accounting application I use for my small business. The problem was that the new naming scheme for hard drives has made a mess of what used to be a simple procedure for setting up dual boots using IDE drives. The situation was further complicated because with the release of Mepis 7, Mepis has switched from using the Ubuntu repositories to the Debian repositories. This means that to move forward you have to do a re-install rather than an upgrade. I first downloaded the iso for Mepis 7 but the install process could only see sda as a whole rather than the Win XP partition and the linux partition. Ubuntu 8.04 was a bit better inthat it saw the linux partition, but it just completely locked up at the 15% mark of the install. The Mepis and Ubuntu forums both indicate that they think the problem lies with the version of GRUB that is being used, but I'm not so sure. I think that the problem lies in Debian's preference for upgrades over re-installs. It's almost as though Debian is hard coded not to overwrite a previous Debian or Debian based install (not likely I know, but neither can it be a GRUB issue because the boot loader is not written to the hard drive until the very end of the kernel install stage). In any event, I tested my theory by choosing an RPM based distro. I discounted Fedora from the get go because I really prefer KDE. Suse was out because of Novell's deal with Microsoft. Next on the list was Mandriva and what do you know but it installed without a hitch. Mepis 7 is a great distro and it installed without a hitch on my desktop which has an all SATA set up. For the laptop, Mandriva One made it by process of elimination.

Why do we have to jump through these hoops? It seems to me that if one mainstream Linux distro can work around / eliminate the bottleneck, they all should be able to -- out of the box!!!