Re: [GNU Crypto] JKS and reverse engineering

[Mark Wielaard]

Hi,

On Sat, 2003-06-21 at 13:03, Casey Marshall wrote:
> I've been working on an implementation of the JKS keystore algorithm,
> which is Sun's proprietary keystore format. I have done this by
> reverse engineering Sun's implementation.

Thank you! Thank you! Thank you!

> What is the position here with regard to reverse engineering? If there
> is no problem then should GNU Crypto include this class? 

We want to provide people with a free platform that they can migrate to
when they are currently locked into a proprietary implementation. If
reverse engineering is the only way to provide this then I personally
think that it should be perfectly fine to do it.

But since the FSF distributes GNU Crypto as an official GNU package it
is best just to ask their position on this (they do take some legal risk
by providing the official packages). I suggest just forwarding your
message in which you explain how you reverse engineered the
algorithm/file format (and why it is important for us to provide this to
free software developers) to Richard Stallman and ask him whether the
FSF wants to distribute this as part of GNU Crypto or that it is more
appropriate to only link to the free implementation on you homepage.

(There is even the option to take your human readable description of the
format and let someone else reimplement it from scratch if that gets
around any legal issues. If that is the case I volunteer to do that
work.)

Cheers,

Mark