mariella wrote on Dec 16^th, 2016 at 3:31pm:


Frankly, I've never been impressed with "top-rated" or "award-winning" appellations. In most cases, reviews only cover simple, basic tasks that most programs can do. Testing is not comprehensive enough to get into the kinds of tasks that set programs apart.

But that aside, my post was directed at disk and partition imaging programs that I have had occasion to test over the last decade or so. (Programs I tested more than a decade ago are no longer compatible with today's hard disks, principally due to the 128 GB barrier.)

Some of these programs I may have tested several years ago. While a newer version might possibly remedy some of a product's earlier limitations, there would need to be a really good reason for me to go back and re-review a product I tested earlier unless I thought it could be better than my preferred alternatives.

In the case of dd, the version I tested did not modify the boot.ini or BCD when necessary to make the restored partition bootable. It also seemed incapable of compressing the image file (at least I wasn't able to find such an option in the version I tested). When there are so many other free programs that can do those things, I would never suggest dd to a user asking me for a recommendation.

In addition, I have seen too many failures caused by linux-based partition managing tools, so I make it a rule to never use linux-based tools to modify NTFS or Windows-based partitions unless there is no other choice.

I suspect the underlying problem is linux doesn't expect the same kind of boot or partition rules Windows expects, so linux thinks it's doing fine while it's actually driving Windows batty. For instance, Clonezilla can at times restore a partition to start at any sector on the disk, and while that kind of layout may not be an issue with linux it makes for a very un-Windows like partition layout.

In the past I have documented cases where linux tools have mangled partition tables because linux only cares about the LBA numbers, and so doesn't care if it trashes the legacy CHS numbers in the partition table. That can have catastrophic consequences for older Windows and DOS partitions--including the Utility partition Dell installed at the front of OEM hard disks for more than decade.

My post was not meant to extend to floppies, USB flash drives, or virtual disks. But since you asked, I do not use any of the aforementioned programs for backing up those devices.

For making/restoring an image of a floppy disk, I use TRACKCPY.EXE, a program I wrote myself 20 years ago. However, if someone asked me for a recommendation I would suggest the widely available RAWREAD.EXE and RAWRITE.EXE, which do essentially the same thing. I have a library of images of all my old floppy disks, and can still mount and run them in virtual machines whenever I want.

For USB flash drives, I do not recommend any kind of imaging. First of all, if the device is only storing data then imaging is the wrong tool, anyway. Copying is a more straightforward backup solution. Imaging is only necessary for bootable devices.

The problem with imaging is flash devices are very particular about the "drive" parameters they simulate to the boot process, and those parameters are not standardized like they are with hard disks. There's a high probability of failure if an image is taken from one flash drive and restored to a different flash drive.

If you're restoring back to the very same device, then there's a good chance any of the usual partition imaging tools will work--even Ghost 2003 can handle this if you do a partition restore instead of a disk restore. If I'm trying to restore to a different flash drive, though, I don't bother. It's better to create the bootable flash drive anew using the Rufus tools.

As for virtual disks, any disk or partition imaging program is the wrong tool. A virtual disk is merely a file on the host system. Backing up a virtual disk is as simple as copying that file. Using an imaging program from inside a VM strikes me as a Rube-Goldberg solution to a simple task. (Remember that old joke about how NASA spent 20 million dollars developing a pen that would write in zero gravity, while the Russians just used a pencil?)