[lkml]   [1997]   [Apr]   [20]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
    SubjectRe: [Crypto] Re: vs.

    > In article <199704190944.FAA04829@jenolan.caipgeneral> you write:
    > > Won't work, there are laws that say it is also illegal to export
    > > anything which "allows non-exportable crypto to be plugged in
    > > easily".
    > Then does this apply to the loopback device? It has hooks for
    > encryption where you can semi-easily (some source tweaking necessary)
    > plug in a DES module.
    > I'd hate to see the official kernel distribution "Bones"ed.
    > > There were changes many months ago to Glibc specifically because of
    > > this problem.
    > As I assume the FSF lawyers know this stuff better than I: what needed
    > to be taken out?

    They use the add-on mechanism of GNU libc. This is a totally generic
    mechanism that allows to include extra code into the library without
    hacking the library itself. All the encryption code is now in a
    des-crypt add-on. One of the side effect is that GNU libc now has
    a libcrypt - I stopped counting the number of Makefiles I had to fix ...

    I hate the idea of a diced libc/kernel etc. just like you but the FSF
    is paranoid about legal things and they're probably right because
    being sued about such an issue might ruin them. As I understand law
    it would have been fully sufficient to modify the DES code of crypt()
    such that it can no longer be used for encryption and decryption to
    satisfy ITAR.

    I'd appreciate if someone'd post references to the text of EAR.


     \ /
      Last update: 2005-03-22 13:39    [W:0.019 / U:4.532 seconds]
    ©2003-2017 Jasper Spaans. hosted at Digital OceanAdvertise on this site