Secret Sharing (again)

[Phil Sutter]

Hey, all!

Long time no read, so there's news:

First of all, I changed my email address. The prior
<sutter at informatik.hs-furtwangen.de> will run out within the next few
months as I've finished studying there.

For further development of the Secret Sharing extension I set up a Git
repository. The main reason why I did not use Subversion is the ability
to check in code. The old approach was to have a checkout of trunk,
heavily modified and the output of 'svn diff' regularly being checked in
to my own Subversion repository. Not only that this was rather a risk of
loosing code than a way to asure the opposite, an at least half way
transparent development process is not possible on this base. When using
Git this all becomes easy: I mirror current trunk into my Git
repository, do my development and commit changes as small as I want.
With a history of patches and commit messages, as it should be. Finally,
as long as I use 'git rebase' to merge updates from SVN, the point in
history where official development ends and mine starts is always clear,
so creating a patch containing only my changes is piece of cake.

I set up Gitweb on my server, so the repository can be accessed
comfortably. This is the URL:

http://nwl.cc/cgi-bin/git/gitweb.cgi?p=ssd.git;a=summary

Since my last email sent here I've continued developing, basically
implementing the first TODO on my list (i.e. changing the Galois Field
being used). The main reason for this is the direct implication of the
field size on the maximum number of shares which can be created. As I
used GF(2^8), this was 255 shares at max. The used field size is now
selectable when initialising the sharing of a key, the choices are
either 2^8, 2^16, 2^24 or 2^32, while the last one needs quite a lot of
CPU power (my X40 wasn't enough).

I would really appreciate to get some feedback from you on all this. So
here are some questions (of decending importance):

* is there still interest in the proposed extension?
* How good are the chances to eventually getting this mainstream?
* Do you ACK the general concept as is? (See my earlier mails for
  reference, I did not change the concept at all.)
* What else should I have asked here instead of this dummy question? ;)

Greetings, Phil