[Bug-ddrescue] configure DDRescue to retry disks that unexpectedly go offline

[Robert Imhoff]

Hello !

First of all many thanks for an incredibly useful piece of software!

A small suggestion for a feature improvement is the following:
I've just been working on recovering a defective WD10TMVW drive that has 
the USB interface integrated on the drive's PCB. Due to the hardware 
failure, the drive goes completely offline after trying defective 
sectors, causing DDRescue to abort with an error message that the input 
file cannot be found (it is being accessed directly, as /dev/sdh, not as 
a partition). After a short pause, the drive comes back online, but 
DDRescue has already aborted and needs to be manually relaunched.

So I thought it would be a useful feature that in such cases, instead of 
aborting, DDRescue could be configured to automatically retry input 
files after a certain interval.

If I may also ask a question here, it concerns the importance of the 
physical sector size of the input and output drives being identical.
In particular, the defective drive has physical sectors of 4096 bytes, 
and logical sectors of 512. The only destination drive I had available 
has both physical and logical sectors of 512 bytes.
The files can still be recovered on the clone, but the main and  backup 
MFT are lost (it's an NTFS partition), and the recovered files’ names 
are not retrieved (for example with R-Studio or ZAR).

My question is whether this could be due to the mismatch in physical 
sector size, or whether it's just because the directory data is in the 
defective region of the drive? (it lost the first 50 GB out of a total 
1TB drive). Doesn't the file access happen in the logical sector size 
anyway, no matter what the physical sector sizes are?

Many thanks!
Robert