Re: Long Name Requirements

Stephen Farrell <stephen.farrell@cs.tcd.ie> Tue, 24 February 2009 10:15 UTC

Return-Path: <owner-ietf-pkix@mail.imc.org>
X-Original-To: ietfarch-pkix-archive@core3.amsl.com
Delivered-To: ietfarch-pkix-archive@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 8A14C28C13A for <ietfarch-pkix-archive@core3.amsl.com>; Tue, 24 Feb 2009 02:15:52 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.293
X-Spam-Level:
X-Spam-Status: No, score=-0.293 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, FH_HOST_EQ_D_D_D_D=0.765, FH_HOST_EQ_D_D_D_DB=0.888, HELO_MISMATCH_COM=0.553, RDNS_DYNAMIC=0.1]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id X59epxHYlOTP for <ietfarch-pkix-archive@core3.amsl.com>; Tue, 24 Feb 2009 02:15:51 -0800 (PST)
Received: from balder-227.proper.com (properopus-pt.tunnel.tserv3.fmt2.ipv6.he.net [IPv6:2001:470:1f04:392::2]) by core3.amsl.com (Postfix) with ESMTP id 8137028C0FB for <pkix-archive@ietf.org>; Tue, 24 Feb 2009 02:15:49 -0800 (PST)
Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.14.2/8.14.2) with ESMTP id n1O9q7s0051173 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Tue, 24 Feb 2009 02:52:07 -0700 (MST) (envelope-from owner-ietf-pkix@mail.imc.org)
Received: (from majordom@localhost) by balder-227.proper.com (8.14.2/8.13.5/Submit) id n1O9q7bx051172; Tue, 24 Feb 2009 02:52:07 -0700 (MST) (envelope-from owner-ietf-pkix@mail.imc.org)
X-Authentication-Warning: balder-227.proper.com: majordom set sender to owner-ietf-pkix@mail.imc.org using -f
Received: from mail.newbay.com (87-198-172-198.ptr.magnet.ie [87.198.172.198]) by balder-227.proper.com (8.14.2/8.14.2) with ESMTP id n1O9psRB051156 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO) for <ietf-pkix@imc.org>; Tue, 24 Feb 2009 02:52:06 -0700 (MST) (envelope-from stephen.farrell@cs.tcd.ie)
Received: from localhost (localhost.localdomain [127.0.0.1]) by mail.newbay.com (Postfix) with ESMTP id A045F10041539; Tue, 24 Feb 2009 09:51:53 +0000 (GMT)
X-Virus-Scanned: amavisd-new at newbay.com
Received: from mail.newbay.com ([127.0.0.1]) by localhost (mail.newbay.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id aKdctYSE2-K1; Tue, 24 Feb 2009 09:51:50 +0000 (GMT)
Received: from [192.168.3.55] (unknown [192.168.3.55]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by mail.newbay.com (Postfix) with ESMTP id A70CD1003EB59; Tue, 24 Feb 2009 09:51:50 +0000 (GMT)
Message-ID: <49A3C33A.7010502@cs.tcd.ie>
Date: Tue, 24 Feb 2009 09:51:54 +0000
From: Stephen Farrell <stephen.farrell@cs.tcd.ie>
User-Agent: Thunderbird 2.0.0.16 (X11/20080707)
MIME-Version: 1.0
To: ersin.gulacti@kamusm.gov.tr
CC: ietf-pkix@imc.org
Subject: Re: Long Name Requirements
References: <200902240835.n1O8ZjYW047493@balder-227.proper.com>
In-Reply-To: <200902240835.n1O8ZjYW047493@balder-227.proper.com>
X-Enigmail-Version: 0.95.7
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 8bit
Sender: owner-ietf-pkix@mail.imc.org
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-pkix/mail-archive/>
List-ID: <ietf-pkix.imc.org>
List-Unsubscribe: <mailto:ietf-pkix-request@imc.org?body=unsubscribe>

So I reckon if this is a real use-case then we should think
about a change. And while it sounds like it is a real use
case it'd be good to see some example names that are longer
than those allowed by 5280. (And any issued certs that are
in the wild containing those.)

Having said that, I do wonder whether the organisations
actually also have unique shorter names - hard to imagine
people always using such long names - in which case why
not modify the policy to work within the 5280 limits? A
policy change like that is much cheaper than modifying
and deploying new s/w everywhere. If the authors of the
policy are offered the choice between their current
scheme (full long names in subject field) vs. the cost
of doing so (for everyone else), and the lack of benefit
during Internet-scale deployment, and if they are reasonable,
then they may well change the policy. If OTOH, the policy
only covers certs for some particular application then
I don't care too much anyway that they are x.509 but
not 5280 compliant.

E.g. in our case, we like to be called Trinity College
Dublin even though our name in contracts is much longer (*)
and would in fact break the 5280 limit. (A certain
irony there since my affiliation is mentioned in 5280;-)
But I don't see a need for that name to be in any X.509
certs so I don't mind.

Stephen.

(*) In contracts, but essentially nowhere else, TCD is called
"The Provost Fellows and Scholars of the College of the Holy
and Undivided Trinity of Queen Elisabeth Near Dublin" which
is a 5280-unfriendly 113 characters. I guess the short-sighted
folks 400 years ago just didn't plan ahead for x.509:-)



Ersin Gülaçtı wrote:
> 
> First, I like to clarify some points. What I am trying is to put a legal
> person's name (i.e Company Title) in the Subject field of the certificate
> (CN=ACME Company ...). The company will be the owner of the certificate.
> Since company names are longer than 64 characters in nature, I have to
> devise a solution for creating standards compatible certificates. 
> 
> I have reviewed the discussions between the PKIX and ITU on upper bounds. I
> wish and hope that RFC 5280 bis will follow X.509's lead on handling longer
> names. In practice our CA software and verification API had strict
> enforcement of the 64 characters based on RFC 3280, 4 years ago. At that
> time we encountered some root certificates that had longer CN and OU
> components than 64 characters. To make our API compatible with those
> certificates we relaxed our upper bounds limit to 128 characters for the
> decoding of certificates. Our CA software still creates certificates
> according to the 64 characters upper bound limitation. I think it would be
> better to modify our CA software and create certificates with no upper bound
> limits on name structures. 
> 
> -----Original Message-----
> From: Kemp, David P. [mailto:DPKemp@missi.ncsc.mil] 
> Sent: Monday, February 23, 2009 8:29 PM
> To: egulacti@uekae.tubitak.gov.tr; ietf-pkix@imc.org
> Subject: RE: Long Name Requirements
> 
> Ersin,
> 
> While X.509 permits unlimited-length strings in the Subject field, it may be
> worth considering usability before making a decision to exercise that
> freedom.
> 
>>From a practical perspective, applications may handle long names with
> varying degrees of grace - some may have hard-coded upper bounds based on
> 5280 and will refuse to accept longer strings; others may have user
> interface layouts based on length assumptions and may truncate or wrap for
> display purposes without rejecting the certificate.  If you buy or develop
> applications designed to handle long strings, then no problem as long as
> every application that touches your certificates can handle them properly.
> 
>>From a theoretical perspective, title (regardless of length) probably
> doesn't belong in the Subject field.  The Subject name is an identifier that
> tells "who" the certificate applies to.  Other information about the user,
> such as height, weight, date of birth, fax number, or title/role/position
> within the organization can be included in certificates but are not
> semantically identifiers.  Contact information can go either way - at the
> office a fax number is probably shared but you have a private voice line
> that uniquely refers to you, while at home the voice line may be shared with
> the wife and kids but the only fax machine is in your home office.  If your
> purpose is to issue a "role certificate" to a position without regard to who
> occupies the position, then title is an identifier and does belong in the
> Subject name.  But in most cases, title is an attribute of the user, not
> part of the user's name.  The X.520 title attribute is most appropriately
> placed in the Subject Directory Attributes extension, with the same
> implementation considerations as above.  
> 
> As Eric says, defining a new attribute type is not necessary because it
> would be at least as difficult to get applications to support a new type as
> it would be to support the new syntax of the existing attribute.
> Hopefully RFC 5280 bis will follow X.509's lead, making upper bounds minima
> that implementations MUST support, not maxima that implementations MUST NOT
> exceed.
> 
> I'm sure someone will point out that there is precedent for putting almost
> anything you want in the Subject name, including certificate policy
> statements and mpeg videos of cats.  But don't be persuaded that everything
> that is syntactically valid is necessarily wise.
> 
> 
> 
> -----Original Message-----
> From: owner-ietf-pkix@mail.imc.org [mailto:owner-ietf-pkix@mail.imc.org]
> On Behalf Of Erik Andersen
> Sent: Monday, February 23, 2009 9:57 AM
> To: egulacti@uekae.tubitak.gov.tr; ietf-pkix@imc.org
> Subject: RE: Long Name Requirements
> 
> 
> Hi Ersin,
> 
> In the latest edition of X.509, we removed the length restrictions on
> attribute types. The PKIX group did not follow suit, but retained the length
> restriction. A field of 500 characters will not comply with RFC 5280, but it
> will comply with the X.509 itself, and that is what is important.
> 
> You can find links to the lasted documents on
> http://www.x500standard.com/index.php?n=Extension.Ed6.
> 
> You do not need to define own attribute types.
> 
> Erik Andersen
> Andersen's L-Service
> Elsevej 48, DK-3500 Vaerloese
> Denmark
> Mobile: +45 2097 1490
> email: era@x500.eu
> www.x500.eu
> www.x500standard.com
>  
> 
> -----Original Message-----
> From: owner-ietf-pkix@mail.imc.org [mailto:owner-ietf-pkix@mail.imc.org]
> On
> Behalf Of egulacti@uekae.tubitak.gov.tr
> Sent: 23. februar 2009 14:42
> To: ietf-pkix@imc.org
> Subject: Long Name Requirements
> 
> 
> Hi,
> 
> I have read through the PKIX mailing list archives to find a method for
> using name components longer than 64 characters in the Subject field of an
> X.509 v3 certificate. Unfortunately I could not find a solution which
> satisfies RFC 5280 (3280) by using standard CN, Title, O or OU components.
> 
> Now I plan to use a custom attribute-value pair in the Subject field. Is
> there any widely used OID for really long names in the Subject field? I need
> to put full company titles, up to 500 characters long, in the Subject field.
> 
> Regards,
> 
> Ersin
> 
> 
> 
> 
>