Re: Suggestion about error control

[Antonio Diaz Diaz]

Scott Dwyer wrote:
> The tools mentioned (hddsuperclone and DMDE) are not open source and
> have paid versions for a reason. The authors have spent much time and
> effort working on them to make them special.

Do you mean that I have not spent much time and effort working on ddrescue 
to make it special?

> Realizing the different error conditions of a device cannot be done with
> normal commands.

IMO, non-standard commands should be used only when standard commands don't 
suffice, because non-standard commands have a much higher probability of 
causing a catastrophic data loss because of an incompatibility.

Ddrescue is a low-risk project. Maybe it won't maximize the probability of 
recovering the data in difficult cases, but it won't risk destroying your 
data by trying to be more clever than the kernel.

It is good that there exist low- and high-risk projects. This allows you to 
try the low-risk ddrescue first, and try the high-risk non-standard software 
if ddrescue can't recover your data.

> Comparing those tools to ddrescue is like comparing apples to oranges.

Certainly. Ddrescue exposes its code so that everybody can see (and improve) 
it. OTOH, we only have your word that the secret, non-standard commands you 
use in your program won't eat the user's data. :-)

IMO every piece of software should either publish the full source code (so 
that users can decide if they trust it) or offer an unlimited warranty in 
case of misbehavior of the code.

By the way, would you mind toning down the ads in your emails? This mailing 
list is for improving GNU ddrescue, not for advertising other software. Thanks.

Antonio.