Hadders wrote: > If found this driver, http://www.fs-driver.org/ . Sound goods, but only > has ext2 support. Jonathan Berry replied: > ext3 is completely backwards compatible with ext2. Thus, you can > mount an ext3 partition as an ext2 one, you just won't get the added > benefits of ext3 (like journaling). I have used a Windows ext2 driver > in the past and it worked fairly well. I stopped using it, though, > because one day I had to pull the power when in Windows and the next > time I booted Linux, the ext3 filesystems had error and had to be > fixed (which wasn't as easy without the journal). That scared me :-). > > For what you want to do, NTFS may be a good solution. As others have > noted, the new ntfs-3g driver for Linux seems to work fairly well. Hadders responded: > Hmmm, so what you're saying is the driver works fine unless you drop > the power to the box. That's not a problem for me, I have a UPS and I > can't remember the last time I had to hold in the ATX switch for 10 > seconds to force a cut-out. Also, I'll be running the partition on a > RAID 1 mirror, that may fix an odd half-write inconsistency from the > other disk, .. maybe, depends which it believes is the correct disk. We are talking Windows here. In my experience, Windows 2000 and XP can be pretty stable, but it does depend on the quality of the drivers -- there are a lot of very badly written drivers our there. It's arguable that Linux's biggest stability advantage is that most drivers have to go through the Quality Assurance process of the Linux Kernel Mailing List (and most user-space drivers don't get to affect the stability of the system). [1] So it will only take one dodgy driver (especially if you've got a dual-core or hyperthreading system [2]) to challenge the stability of your system. You may find that a large FAT32 partition is best for your needs. (Windows won't create them above 32 GB, but it will read larger partitions.) Hope this helps, James. [1] Closed source and out-of-tree drivers don't get this advantage, of course. [2] A *lot* of potential bugs in kernel-mode code will only come to light occasionally and on SMP systems. -- E-mail: james@ | Words fail me. Thank goodness I can make gestures. aprilcottage.co.uk | -- Mark Hughes
