Re: Crippled Notes export encryption

At 20:32 01.17.1996 -0500, Perry E. Metzger wrote:
>
>Alan Pugh writes:
>> infoMCI (sm)
>> Lotus-Security - Lotus Announces Compromise for Export of Strong
>> Encryption
>
>So, Lotus thinks they can fool people by back-dooring in key escrow, eh?
>
>Time to break out the artillery.
>
>Perry

Careful... what would YOU have done, with your customers demanding stronger
crypto today and you unable to legally give it to them?

Again, folks, try to remember that this is NOT key escrow... international
Notes customers are no worse off than before, and a darn sight better off
against everyone besides Uncle Sam.

Herb

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Herb Sutter (herbs@connobj.com)

Connected Object Solutions     2228 Urwin - Suite 102     voice 416-618-0184
http://www.connobj.com/      Oakville ON Canada L6L 2T2     fax 905-847-6019

In article <3105FBFC.4DC9@netscape.com> Jeff Weinstein <jsw@netscape.com> writes:
>   The other way would be to export a binary with pluggable crypto,
> which is generally agreed to be regulated by the ITAR in the same
> way as software that actually contains crypto.

How did kerberos avoid this?  The "bones" distribution of kerberos
without crypto was not regulated by ITAR, right?

David

Ian Goldberg wrote:
> OK; so what if I have code that says:
> 
> RNG_GenerateRandomBytes(buf, size);
> Hash(outbuf, buf, size);
> /*
>  *  It would be really nice if outbuf were RSA-encrypted
>  *  with (expon,modulus) at this point and the result placed in
>  *  outbuf2, but we have to do the following instead:
>  */
>     for(i=0;i<hashsize;++i) outbuf2[i] = ~outbuf[i];
> fwrite(outbuf2, hashsize, 1, fp);
> 
> Would the above code be export-restricted because it contained wishful
> thinking about how nice it would be to use encryption?

  The problem is that the government refuses to publish the rules.
They make people ask for approval for every piece of code that is
exported.  This gives them lots of wiggle room so that they can keep
changing the rules in the face of technical, legal, or political
innovation.

	--Jeff

-- 
Jeff Weinstein - Electronic Munitions Specialist
Netscape Communication Corporation
jsw@netscape.com - http://home.netscape.com/people/jsw
Any opinions expressed above are mine.

In article <199601242330.SAA08632@toxicwaste.media.mit.edu>,
Derek Atkins  <warlord@MIT.EDU> wrote:
>> How did kerberos avoid this?  The "bones" distribution of kerberos
>> without crypto was not regulated by ITAR, right?
>
>Kerberos didn't leave the crypto plugable.  The bones distribution
>removed not only the crypto routines but also the calls to the crypto
>routines.  It would be hard to call that "pluggable".  It took a lot
>of work for someone down under to replace all those crypto calls!
>
OK; so what if I have code that says:

RNG_GenerateRandomBytes(buf, size);
Hash(outbuf, buf, size);
/*
 *  It would be really nice if outbuf were RSA-encrypted
 *  with (expon,modulus) at this point and the result placed in
 *  outbuf2, but we have to do the following instead:
 */
    for(i=0;i<hashsize;++i) outbuf2[i] = ~outbuf[i];
fwrite(outbuf2, hashsize, 1, fp);


Would the above code be export-restricted because it contained wishful
thinking about how nice it would be to use encryption?

   - Ian "Maybe I should just go back to Canada..."

> Would the above code be export-restricted because it contained wishful
> thinking about how nice it would be to use encryption?

IANAL, but my guess is that no, that code would not be exportable.  At
least not if there really is a domestic vs. export version.  Yes, it
gets really fuzzy here.  I think if you started with this code and
didn't have any hooks at all, and only had a version (domestic and
export) which contained this wishful thinking, you might get away with
it.  Then again, if that were the case it would not be export
controlled in the first place since it doesn't use encryption ;)

Yes, it is a huge can of worms.  Worse, since it is done on a
case-by-case basis, there really is no clear definition of where the
exportable vs. non-exportable line actually is.  You need to try it to
test if it will work or not.

-derek

So does that mean that it is legal to ship PGP out of the US by
shooting a diskette in a rocket???  It's launching the munition, no?
Therefore by sentence (6) it should be allowed. ;)

-derek

> How did kerberos avoid this?  The "bones" distribution of kerberos
> without crypto was not regulated by ITAR, right?

Kerberos didn't leave the crypto plugable.  The bones distribution
removed not only the crypto routines but also the calls to the crypto
routines.  It would be hard to call that "pluggable".  It took a lot
of work for someone down under to replace all those crypto calls!

-derek

> cc: Jeff Weinstein <jsw@netscape.com>, cypherpunks@toad.com
> Date: Wed, 24 Jan 1996 18:30:00 EST
> From: Derek Atkins <warlord@MIT.EDU>
> 
> > How did kerberos avoid this?  The "bones" distribution of kerberos
> > without crypto was not regulated by ITAR, right?
> 
> Kerberos didn't leave the crypto plugable.  The bones distribution
> removed not only the crypto routines but also the calls to the crypto
> routines.  It would be hard to call that "pluggable".  It took a lot
> of work for someone down under to replace all those crypto calls!

So where exactly do they draw the line?  You can still construct your
software in such a way that there is a clean boundary between the
crypto stuff and the rest.

For example, could you have an application with a  function:

	authenticate_user (int file_descriptor)

which in the exportable version sends a password, and in the domestic 
version constructs some sort of authenticator?

Could you have an xdr-like function which on in an exportable version
just does argument marshaling and in a domestic version also encrypts?

How exactly are crypto-hooks defined?  This restriction seems orders
of magnitude more bogus than even the ban on exporting actual
encryption.

David

> So where exactly do they draw the line?  You can still construct your
> software in such a way that there is a clean boundary between the
> crypto stuff and the rest.

The line is drawn, AFAIK, at the actual crypto routines.  You cannot
export the crypto routines, and the functions that call the crypto
routines.

> For example, could you have an application with a  function:
> 
> 	authenticate_user (int file_descriptor)
> 
> which in the exportable version sends a password, and in the domestic 
> version constructs some sort of authenticator?

Yes.  In fact, this is what Bones did.

> Could you have an xdr-like function which on in an exportable version
> just does argument marshaling and in a domestic version also encrypts?

Yes.  However the exported code cannot have the encryption hooks
in the code.

> How exactly are crypto-hooks defined?  This restriction seems orders
> of magnitude more bogus than even the ban on exporting actual
> encryption.

Very vaguely.  If I have a function that does something like this:

authenticate (args)
{
	...

	des_encrypt ();
	...
}

I would have to remove the des_encrypt() call from the authenticate()
routine before it can be exported...

-derek

On Wed, 24 Jan 1996, Derek Atkins wrote:

> > How exactly are crypto-hooks defined?  This restriction seems orders
> > of magnitude more bogus than even the ban on exporting actual
> > encryption.
> 
> Very vaguely.  If I have a function that does something like this:
> 
> authenticate (args)
> {
> 	...
> 
> 	des_encrypt ();
> 	...
> }
> 
> I would have to remove the des_encrypt() call from the authenticate()
> routine before it can be exported...

Would removing the call to des_encrypt() and replacing it with a
comment violate the restriction?  something like:

authenticate (args)
{
	...
	/* squeamish ossifrage */
	...
}


--
|\/|ike Gurski  mgursk1@gl.umbc.edu  FidoNet: 1:261/1062
http://www.gl.umbc.edu/~mgursk1/   finger -l for PGP public key   |Member,
1024/39B5BADD PGP Keyprint=3493 A994 B159 48B7 1757 1E4E 6256 4570|   Team
My opinions are mine alone, even if you should be sharing them.   |   OS/2

"dm" == David Mazieres <dm@amsterdam.lcs.mit.edu> writes:

dm> How did kerberos avoid this?  The "bones" distribution of kerberos
dm> without crypto was not regulated by ITAR, right?
  In the ``bones'' version not only was the encryption code eliminated
(e.g., the functionality of libdes.a), but the hooks to call such code
disappeared as well.

michael

David Mazieres wrote:
> 
> In article <3105FBFC.4DC9@netscape.com> Jeff Weinstein <jsw@netscape.com> writes:
> >   The other way would be to export a binary with pluggable crypto,
> > which is generally agreed to be regulated by the ITAR in the same
> > way as software that actually contains crypto.
> 
> How did kerberos avoid this?  The "bones" distribution of kerberos
> without crypto was not regulated by ITAR, right?

  As others have noted, they removed the calls to the crypto code.

  I don't think that the TLAs are concerned about people at
foreign universities using kerberos.  They are much more worried
about mass market products.  If we did the same thing as was done
for kerberos, then exported the code to a foreign subsidiary, I
believe that the government would try to make a case against us
that we had participated in a conspiracy to circumvent the
export restrictions.

  The government continues to use FUD to impose defacto restrictions
on what we can do.  When they decided not to prosecute PRZ they did
not clarify and said that they may decide at any time to go after
someone else.  They continue to try to wiggle out of stating a
clear, firm policy.  I think that our current efforts should be
geared towards pinning them down, then once we have specific
restrictions we can attack them.  The Phil Karn case is important
because it will help to clarify the ITAR restrictions.  Even 
Raph's RSA T-shirt CJR may help to clarify the restrictions into
something that we can really fight.

	--Jeff

-- 
Jeff Weinstein - Electronic Munitions Specialist
Netscape Communication Corporation
jsw@netscape.com - http://home.netscape.com/people/jsw
Any opinions expressed above are mine.

Herb Sutter writes:

>Careful... what would YOU have done, with your customers demanding stronger
>crypto today and you unable to legally give it to them?

Build it overseas where there are no restrictions and then import it. Just
like almost every other component in a computer.

Herb Sutter writes:
> >So, Lotus thinks they can fool people by back-dooring in key escrow, eh?
> >
> >Time to break out the artillery.
> 
> Careful... what would YOU have done, with your customers demanding stronger
> crypto today and you unable to legally give it to them?

Set up development shop overseas for the crypto plug-ins.

The solution is obvious and easy.

By the way, I really think Netscape should simply ship Jeff and other
people to the Amsterdam office or wherever else seems reasonable and
do all the crypto work from there. It will save trouble and
hassle. U.S. citizens wanting full 128 bit over the net would then get
it from Netscape's overseas download sites. No one anywhere in the
world would be forced to use crap.

Perry

On Tue, 23 Jan 1996, Perry E. Metzger wrote:

> By the way, I really think Netscape should simply ship Jeff and other
> people to the Amsterdam office or wherever else seems reasonable and
> do all the crypto work from there. It will save trouble and
> hassle. U.S. citizens wanting full 128 bit over the net would then get
> it from Netscape's overseas download sites. No one anywhere in the
> world would be forced to use crap.

<flog,flog,neigh,neigh> 
Sorry, to beat a dead horse, but isn't this like that thread re:Vince Cate's 
invitation for us to all move to Anguila? We'd still be exporting our 
thoughts in violation of the ITAR. As long as you're American (person or 
corporation) you'd still be commiting a crime by putting strong crypto on 
the net, regardless of where you are, right? (That's what I remember from 
Micheal Froomkin's arguments from the last time this revolving thread came
around).

Of course, if some OTHER company (wink,wink,nudge,nudge) that is not 
incoporated in the US were employ to export a secure version of Netscape 
from Amsterdam, that would be another story (and how would anyone know that
the code, just so happenned to have been written Jeff, et al.?)
Just as long as the portion of Netscape that nominally produces Netscape 
(the software) and distributes it does not call itself Netscape and is 
not a US corporate citizen, that is the end of the game, right?
Anyone got any spare holding companies handy? 
</flog,flog. No neighs.>

> By the way, I really think Netscape should simply ship Jeff and other
> people to the Amsterdam office or wherever else seems reasonable and
> do all the crypto work from there. It will save trouble and
> hassle. U.S. citizens wanting full 128 bit over the net would then get
> it from Netscape's overseas download sites. No one anywhere in the
> world would be forced to use crap.

I've wondered why they don't do this as well.  For people around the world
in general, it would be a very good thing.  But what kind of an effect
would it have on this country?  What if they decide it's easier to fire
Jeff and hire some Dutch guy instead?  Would the government decide that
the export ban was pointless and lift it?  Or would they stand by as big
chunks of our software industry are lost to foreign competitors? 

Remember this is the government we're talking about, the people who 
destroyed a villiage in order to save it.

It seems to me that loss of jobs is inevitable if the rules aren't
changed.  But I'm not sure it's a good thing to accelerate the process. 
I'm not sure it's not, either -- both options are unpleasant. 

There's probably a big opportunity here for some enterprising cypherpunk
who's willing to move to Amsterdam (or who lives there already).  Set up a
company that provides crypto guts and distribution services for American
software companies. 

Stand alone computers are becoming less useful and less common all the 
time.  Networking is a fact of life in the computer industry.  If you're 
doing networking you have to think about security, and if you're serious 
about security you have to use crypto.  A software industry that can't 
deploy crypto without hindrance is living on borrowed time.  

We in America are extremely vulnerable to flight.  We will lose jobs and
market share if we don't change our policies.  Because other countries
will deploy crypto, our polices will be completely ineffective in
preserving the government's ability to do surveillance.  

What's the point? 

It would be very interesting to see what would happen if Netscape
announced that it's considering moving its crypto operations overseas in a
year if the export restrictions aren't lifted.

On Tue, 23 Jan 1996 10:39:03 -0500, perry@piermont.com wrote:


>Set up development shop overseas for the crypto plug-ins.
>
>The solution is obvious and easy.
>
>By the way, I really think Netscape should simply ship Jeff and other
>people to the Amsterdam office or wherever else seems reasonable and
>do all the crypto work from there. It will save trouble and
>hassle. U.S. citizens wanting full 128 bit over the net would then get
>it from Netscape's overseas download sites. No one anywhere in the
>world would be forced to use crap.

Wrong, this would be a violation of ITAR.


Dan Weinstein
djw@vplus.com
http://www.vplus.com/~djw
PGP public key is available from my Home Page.
All opinions expressed above are mine.

"I understand by 'freedom of Spirit' something quite definite -
the unconditional will to say No, where it is dangerous to say
No.        
           Friedrich Nietzsche

Dan Weinstein writes:
 > >By the way, I really think Netscape should simply ship Jeff and other
 > >people to the Amsterdam office...
 > 
 > Wrong, this would be a violation of ITAR.

I don't understand; are you saying Jeff's brain is a munition under
the ITAR?

(Is it a citizenship thing?  If so, that's an easily solved problem:
hire Dutch (or Egyptian or Bangali or whatever) engineers.)

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
| Nobody's going to listen to you if you just | Mike McNally (m5@tivoli.com) |
| stand there and flap your arms like a fish. | Tivoli Systems, Austin TX    |
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Perry E. Metzger wrote:
> 
> Herb Sutter writes:
> > >So, Lotus thinks they can fool people by back-dooring in key escrow, eh?
> > >
> > >Time to break out the artillery.
> >
> > Careful... what would YOU have done, with your customers demanding stronger
> > crypto today and you unable to legally give it to them?
> 
> Set up development shop overseas for the crypto plug-ins.
> 
> The solution is obvious and easy.
> 
> By the way, I really think Netscape should simply ship Jeff and other
> people to the Amsterdam office or wherever else seems reasonable and
> do all the crypto work from there. It will save trouble and
> hassle. U.S. citizens wanting full 128 bit over the net would then get
> it from Netscape's overseas download sites. No one anywhere in the
> world would be forced to use crap.

  I can see two practical ways to build a netscape product outside
the US.  The first is to export the source code for the Navigator
with the crypto code removed.  All of the calls to crypto would
have to be removed as well.  I've heard some people claim that the
government could come after us on the grounds that we were taking
part in a conspiracy to export strong crypto.

  The other way would be to export a binary with pluggable crypto,
which is generally agreed to be regulated by the ITAR in the same
way as software that actually contains crypto.

  I suspect that to get around the US government in this way we
would have to develop the entire product outside of the US.  That
would be a very drastic move that is not likely to happen any
time soon.  We are going to invest some money and effort into
trying to get the current restrictions lifted first.

  Of course there are some of us who are ready and willing to go
if it comes to that...

	--Jeff

-- 
Jeff Weinstein - Electronic Munitions Specialist
Netscape Communication Corporation
jsw@netscape.com - http://home.netscape.com/people/jsw
Any opinions expressed above are mine.

Jeff Weinstein writes:

>  I can see two practical ways to build a netscape product outside
>the US.  The first is to export the source code for the Navigator
>with the crypto code removed.  All of the calls to crypto would
>have to be removed as well.  I've heard some people claim that the
>government could come after us on the grounds that we were taking
>part in a conspiracy to export strong crypto.

Didn't Netscape already promise to remove the hooks? It seems to me all of
the major software players are already in bed with the government.

Mike Tighe wrote:
> 
> Jeff Weinstein writes:
> 
> >  I can see two practical ways to build a netscape product outside
> >the US.  The first is to export the source code for the Navigator
> >with the crypto code removed.  All of the calls to crypto would
> >have to be removed as well.  I've heard some people claim that the
> >government could come after us on the grounds that we were taking
> >part in a conspiracy to export strong crypto.
> 
> Didn't Netscape already promise to remove the hooks? It seems to me all of
> the major software players are already in bed with the government.

  What do you mean by "promise to remove the hooks"?

	--Jeff

-- 
Jeff Weinstein - Electronic Munitions Specialist
Netscape Communication Corporation
jsw@netscape.com - http://home.netscape.com/people/jsw
Any opinions expressed above are mine.

Jeff Weinstein writes:

>> Didn't Netscape already promise to remove the hooks? It seems to me all of
>> the major software players are already in bed with the government.
>
>  What do you mean by "promise to remove the hooks"?

I mean they planned to have some type of crypto, but then after a visit
from the USG, they removed that plan, and even withdrew the interface so
that someone else couldn't "drop in" the crypto. The press release
describing it was within the past 12 months or so. I will find it
eventually.

Mike Tighe writes:
> Jeff Weinstein writes:

> >> Didn't Netscape already promise to remove the hooks? It seems to me all of
> >> the major software players are already in bed with the government.
> >
> >  What do you mean by "promise to remove the hooks"?
> 
> I mean they planned to have some type of crypto, but then after a visit
> from the USG, they removed that plan, and even withdrew the interface so
> that someone else couldn't "drop in" the crypto. The press release
> describing it was within the past 12 months or so. I will find it
> eventually.

I think you may be confused.  I do recall a report that some NSA folks
visited NCSA and recommended they remove crypto hooks from the NCSA httpd.
Is this maybe what you're thinking of?


-- Jeff

Jeff Barber writes:

>> Didn't Netscape already promise to remove the hooks? It seems to me all of
>> the major software players are already in bed with the government.

>I think you may be confused.  I do recall a report that some NSA folks
>visited NCSA and recommended they remove crypto hooks from the NCSA httpd.
>Is this maybe what you're thinking of?

Yes, I guess that was it. Thanks for the correction. Anyway, moving to the
actual point, it does seem most major software players are agreeing to the
USG demands.

>  I can see two practical ways to build a netscape product outside
>the US.  The first is to export the source code for the Navigator
>with the crypto code removed.  All of the calls to crypto would
>have to be removed as well.  I've heard some people claim that the
>government could come after us on the grounds that we were taking
>part in a conspiracy to export strong crypto.

If you properly apply for a license to export the source code
(explainig that the source code licensee might add the features
that he feels are appropriate), is it still a conspiracy?