Christophe,

Fair enough, and point taken.  Thanks for the clarification.

I am puzzled by the need to modify buffers on the fly inside the FSM.
This is straight out of some unpublished information, so it struck a raw
nerve.  Just be aware this tends to follow the design and could be used
for such.

Cheers,

Andre Hedrick
LAD Storage Consulting Group



On Fri, 2 Jan 2004, Christophe Saout wrote:

Am Fr, den 02.01.2004 schrieb Andre Hedrick um 05:43:

quoted

I am sorry but adding in a splitter to CPRM is not acceptable.
Digital Rights Management in the kernel is not acceptable to me, period.

Maybe I have misread your intent and the contents on your website.

Device-Mappers are one thing, intercepting buffers in the taskfile FSM
transport is another.  This stinks of CPRM at this level, regardless of
your intent.  Do correct me if I am wrong.

I can assure you I was never having DRM or anything like this in mind
nor making fundamental changes to the IDE layer. It was just that
++bi_idx that bugged me. Must be a misunderstanding, sorry. :)

The only thing I'm having on my website is a device-mapper target that
does basically the same as cryptoloop tries to. It's just about
encrypting sensitive data on top of any other device, nothing else.