IPv4 Link-Local (RFC 3927) patents

Hi all,

As promised, I'll follow up on the patent stuff.

As far as I can see, ZCIP implements IPv4 link-local addressing. This is 
standardized in RFC 3927, or will so shortly (it is in "Author 48-hour 
hold" since August 2004...). ZCIP does only partly do so, since some 
things have to be fixed in the kernel (for example, ARP replies must be 
broadcasted, but this is a kernel issue, not a ZCIP issue)

With the IETF, two distinct patent claims have been submitted:
A claim by Microsoft Corporation:
http://www.ietf.org/ietf/IPR/MICROSOFT-499.txt
A claim by Apple Computer:
http://www.ietf.org/ietf/IPR/apple-ipr-draft-ietf-zeroconf-ipv4-linklocal.txt

A bit of history here. Around 1985, Apple used AppleTalk for local 
communications. In this, they had a rather noval way to randomly select 
addresses and prevent collisions. This was patents in the USA by patents 
  4,661,902 and 4,689,786. In 1998 Microsoft released Windows 1998, 
which included link-local addresses (which was then called "IP Auto 
Configuration" or "IPAP", it was renamed to "Automatic Private IP 
Addressing (APIPA)" with Windows 2000 -- see MS Knowledge Base Q220874).
Microsoft filed patent in the USA as well in April 1998, just before the 
Win98 release in June, with patent 6,101,499. Since Apple and Microsoft 
got in a patent cross licensing in August 1997, Apple could implement it 
is well (That deal expired in 2002, but Microsoft still owns about 18.2 
mln Apple shares, 4.5% of total). Interestingly, Apple forgot that they 
had the prior patents regarding random addresses selection and collision 
prevention. The original authors where employed by Microsoft in 1998, 
and Microsoft had to point Apple to it. The technology was then 
described in miscellaneous Internet-Drafts, starting with 
draft-ietf-dhc-ipv4-autoconfig-00.txt by Ryan Troll in September 1998, 
draft-cheshire-ipv4-linklocal-00.txt by Stuart Cheshire of Apple in 
March 2000, up to draft-ietf-zeroconf-ipv4-linklocal.txt by Stuart 
Cheshire (Apple), Bernard Aboba (Microsoft) and Erik Guttman (Sun). It 
will hopefully be published as RFC 3927 soon (it was already published 
at http://www.ietf.org/rfc/rfc3927.txt last month, but the RFC editor 
withdraw it after some discussion on the zeroconf IETF list).

So, in summary:
- The IETF (and thus I) am only aware of IP claims by Apple and
   Microsoft.
- All patents are only valid in the USA, not in Europe (where I live)
   or Australia (where Brad Hards lives).
- No-one ever asked ZCIP to be removed (Stuart Cheshire was also not
   aware of any such actions by Apple).

So, do these three patents prevent us from publishing ZCIP? I think not.
First of all, I doubt the validity of some of these patents. But the 
answer may be more simple. I'm not living in the USA, so I can use and 
publish it freely. However, I love to have it available to anyone, so I 
did some more research.

- Microsoft clearly states to the IETF that "[If] these Internet Drafts 
become[] an IETF Standard [..], Microsoft will grant a royalty-free 
license [...], provided a reciprocal license is granted to Microsoft"
Source: http://www.ietf.org/ietf/IPR/MICROSOFT-499.txt
Well, ZCIP uses a very nice X11/MIT/3-item BSD license, so that is set.
- Apple's patents did expired in March 2005. I contacted Apple for 
confirmation, and got that. You will find the original mail of one of 
Apple's lawyers below. Though he did confirm that the patents have 
expired, he is not prepared to give us a "carte blanche" (they reserve 
the right to still sue us if they later find out they forgot about 
another patent ;-) ). That is fine. Remember that this is a lawyer, and 
he's only paid to protect the best interest of his company; he's not 
paid to protect our rights. In fact, for a lawyer, this is a lengthy and 
very friendly and kind letter. I can assure you that anything else would 
have first gone through at least the vice-president or higher (if that's 
not already the case for this mail).


In short --- I do not see any problem whatsoever, and in fact we have 
_written_ statements that we are free to implement this technology.


To prevent this from happening again (personally, I feel that 
withdrawing is bad for the open source community), I decided to
make a statement for any further intelectual property claims. I put a 
draft at http://zeroconf.sourceforge.net/?selected=patents
Please review it, and let me know what you think.


Regards and let's get back to coding,
Freek

-------- Original Message --------
Subject: Fwd: IPv4 Link-Local (RFC 3927) patents?
Date: Mon, 11 Apr 2005 15:03:13 -0700
From: Chip Lutton <[email protected]>
To: Freek Dijkstra <[email protected]>

I'm not sure I can answer all your questions.  I do believe the patents
you mentioned have expired, but I do not believe that they are
necessarily all the intellectual property that Apple has that may be
relevant to what you want to do.  I am not in a position to make this
determination for you.  However, I can say that Apple reserves any
rights it may have in intellectual property unless explicitly granted
otherwise, which we have not done in this case.  Thus, I can offer no
assurance that you are free to practice and distribute the program
without concern for Apple patents.

Chip Lutton
Apple


> -------- Original Message --------
> Subject: IPv4 Link-Local (RFC 3927) patents?
> Date: Mon, 21 Mar 2005 20:32:08 +0100
> From: Freek Dijkstra <[email protected]>
> To: [email protected]
>
> To the Director of Patents, Apple Computer,
>
> Dear Sir, Madam,
>
> Background:
> I am a scientist at the University of Amsterdam and like to use
> so-called IPv4 Link-Local Addresses on a Linux cluster. There is a
> software package which implements this, called zcip [1], but about a
> year ago, it was withdrawn by the author because he was not sure about
> IP claims by Apple [2] regarding this technology. I like to use it, and
> in the interest of other people, I hope you can give me some
> clarification on short notice (since I have the impression that the
> claim is now expired and/or freely available). Sorry if my questions  
> are
> not 100% clear -- I'm a computer scientist, clearly not a lawyer.
>
> I understand you can not evaluate each and every software package, so
> let me only ask about the technology, not the software package. The
> technology is described in RFC 3927 [3], and Apple's Intelectual
> Property claims have been submitted to the IETF [4] in April 2004 [2],
> based on two patents [5,6] hold by Apple (U.S. Patent Numbers 4,661,902
> and 4,689,786).
>
> Please, note two significant events this month:
> 1. In March 2005, this technology was published as an official IETF
>     standard.
> 2. March 21, 2005, both patents mentioned in your claim seem
>     to have expired, since the have been filed in March 1985,
>     exactly 20 years ago.
>
> My questions are:
>
> - Does Apple still claim Intelectual Property regarding IPv4
>    Link-Local Addresses, as described in RFC 3927 [3]?
> - Are the patents mentioned in the claim by Apple to the IETF [2]
>    still valid or have they expired?
> - Can I freely implement the technology as described in RFC 3927?
>    With freely meaning "gratis" (without a fee), and can I distribute
>    my implementation to others?
>
>
> By the way, If you want to know what this technology is:
> IPv4 Link-Local addresses are those "self-assigned" IP addresses (the
> 196.254.xx.xx ones) address you see in  the Network configuration of
> your Mac, when you directly connect your Mac to another machine,  
> instead
> of to your company network.
> Your technology "Rendezvous" (to be renamed "Bonjour", as I was told by
> our Apple Europe Sales Manager Last week), as implemented by Apple,
> partly relies on this technology. Based on mailing list activity, I
> belief Marc Krochmal and/or Stuart Cheshire may tell you more then I  
> can
> about it.
>
> With highest regards,
> Freek Dijkstra
>
> References:
> [1] Zcip (Zeroconf IP) software package --
> http://sourceforge.net/projects/zeroconf/
> [2] Apple IP claim -- http://www.ietf.org/ietf/IPR/
> apple-ipr-draft-ietf-zeroconf-ipv4-linklocal.txt
> [3] "Dynamic Configuration of IPv4 Link-Local Addresses", S. Cheshire,
> B. Aboba, E. Guttman, RFC 3927 (IETF Standards Track) --
> http://www.ietf.org/rfc/rfc3927.txt
> [4] IETF, The Internet Engineering Task Force, a standardization body  
> -- http://www.ietf.org/
> [5] U.S. Patent Number 4,661,902, "Local Area Network With Carrier  
> Sense Collision Avoidance", filed March 21, 1985 --
> http://portal.uspto.gov/external/portal/!ut/p/_s.7_0_A/7_0_CH/.cmd/ad/ 
> .ar/sa.getBib/.c/6_0_69/.ce/7_0_1ET/.p/5_0_18L/.d/5
> [6] U.S. Patent Number 4,689,786, "Local Area Network With Self  
> Assigned Address Method", filed March 21. 1985 --
> http://portal.uspto.gov/external/portal/!ut/p/_s.7_0_A/7_0_CH/.cmd/ad/ 
> .ar/sa.getBib/.c/6_0_69/.ce/7_0_1ET/.p/5_0_18L/.d/1
>
>
> ------ End of Forwarded Message





-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.
Get your fingers limbered up and give it your best shot. 4 great events, 4
opportunities to win big! Highest score wins.NEC IT Guy Games. Play to
win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20