Network Working Group Steve Crocker INTERNET DRAFT Ned Freed draft-ietf-pem-mime-06.txt Jim Galvin Sandy Murphy July 1994 PEM Security Services and MIME 1. Status of this Memo This document is an Internet Draft. Internet Drafts are working documents of the Internet Engineering Task Force (IETF), its Areas, and its Working Groups. Note that other groups may also distribute working documents as Internet Drafts. Internet Drafts are valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet Drafts as reference material or to cite them other than as ``work in progress''. To learn the current status of any Internet Draft, please check the 1id-abstracts.txt listing contained in one of the Internet Drafts Shadow Directories on ds.internic.net (US East Coast), venera.isi.edu (US West Coast), munnari.oz.au (Pacific Rim), or nic.nordu.net (Europe). 2. Abstract This document specifies how the services of MIME and PEM can be used in a complementary fashion. MIME, an acronym for "Multipurpose Internet Mail Extensions", defines the format of the contents of Internet mail messages and provides for multi-part textual and non-textual message bodies. PEM, an acronym for "Privacy Enhanced Mail", provides message authentication/integrity and message encryption services for Internet mail messages. An Internet electronic mail message consists of two parts: the headers and the body. The headers form a collection of field/value pairs structured according to RFC822 [1], whilst the body, if structured, is defined according to MIME [2]. MIME does not provide for the application of security services. Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 1]
INTERNET DRAFT PEM and MIME July 1994 PEM [3-6] specifies how to apply encryption and authentication/integrity services to the contents of a textual electronic mail message but does not provide message structuring or type labelling facilities. This document specifies how to use PEM with the multipart/signed and multipart/encrypted MIME content types to provide authentication/integrity and encryption services. We refer to the authentication/integrity service as a digital signature service. This document specifies a number of changes to the message encryption and signature procedures of PEM and broadens the name forms that may be used to identify public keys. Many of the changes represent a departure in mechanism, not in effect. 3. Introduction This document updates the message encryption and signature procedures defined by [3] and obsoletes the key certification and related services defined by [6]. The changes to [3] include the separation of the encryption and signature services, the removal of the limitation to enhance only text-based messages, the removal of the transfer encoding operation, the deprecation of the Content-Domain: and Proc-Type: headers, and the separation of certificate and certificate revocation list transmission from the security enhancements. These changes represent a departure in mechanism, not in effect, and are detailed in Section 10. In addition, this document specifies three technical changes to PEM: symmetric key management in [3] is deprecated, the canonicalization operation in [3] is generalized, and the allowable name forms for the identification of public keys is broadened to include arbitrary strings and email addresses, and users may distribute their public keys directly in lieu of certificates. The key certification and related services document [6] is obsoleted by the specification of two new MIME content types: application/key-request and application/key-data. These new content types are used to transmit requests for key operations (storage, retrieval, certification, revocation list retrieval, etc.) and the responses to those requests. These two content types are independent body parts and are not required to be encapsulated in any other body part. These changes represent a departure in mechanism, not in effect, and are detailed in Section 10. In order to make use of the PEM services, a user is required to have at least one public/private key pair. Prior to this specification, the public key was required to be embodied in a certificate, an object that Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 2]
INTERNET DRAFT PEM and MIME July 1994 binds a public key with a distinguished name, a name form that identified the owner of the public key. The embodiment was issued by a certification authority, a role that was expected to be trustworthy insofar as it verified the identity of the owner prior to issuing the certificate. However, the deployment of certificates and the creation of the hierarchy of certification authorities has been problematic. Instead, this specification bases the PEM services on a public/private key pair. Each key pair is required to belong to a user (where user is not limited to being a human, e.g., a process or a role) which has a name. There are 3 name forms specified by this document. For backward compatibility (and forward compatibility if the X.500 Directory becomes a ubiquitous service) one of the name forms is a distinguished name. In addition, email addresses and arbitrary strings are allowed. Since a user may have more than one key pair, a name form is insufficient for uniquely identifying a key pair. The owner of a key pair must assign a key identifier to each key pair. The combination of a name form and a key identifier uniquely identifies a key pair and each key pair is uniquely identified by a name form and key identifier combination. Throughout this document, this combination is called an identifier. There are 6 identifiers specified by this document. With a key pair for one's self and software that is both MIME and PEM aware, an originating user may digitally sign arbitrary data and send it to one or more recipients. With the public keys of the recipients, a user may encrypt the data so that only the intended recipients can decrypt and read the it. This specification separates these two services so that an originator may apply either or both, in either order. The name forms and identifiers are described in detail in the next section. Succeeding sections specify how PEM and MIME are used together and other ancillary details. 4. Name Forms and Identifiers Currently, [3] requires the use of certificates to identify the public key (and corresponding private key) used to create a PEM message. Within certificates, [4] requires the use of distinguished names as specified by the X.500 Series of Recommendations. However, the Internet community has a great deal more experience with the use of electronic mail addresses as a name form and there is a desire to be able to use arbitrary strings to identify the owners of public keys. Hence, there is a need to support name forms which do not conform to the expected Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 3]
INTERNET DRAFT PEM and MIME July 1994 usage of distinguished names. When processing PEM messages it is necessary to be able to uniquely identify the key pair used to create the message. A certificate is uniquely identified by the combination of its issuer's distinguished name and its serial number. Thus, the issuer name and serial number uniquely identifies a key pair. Since a user may have more than one key pair, a name form is insufficient for this purpose. An identifier is required that consists of both a name form and key identifier, a value assigned to a key pair by its owner. In addition, users may distribute their public keys via mechanisms outside the scope of the PEM specification, for example, in a file via FTP. As a result, it is desirable to be able to explicitly specify the public key used rather than an identifier of the public key. A significant benefit of this mechanism is the ability to support encrypted, anonymously signed mail. The objective of the various Originator and Recipient fields specified in [3] is to identify which public key has been used or is required. This document simplifies the set of fields by specifying exactly two: Originator-ID: for originators and Recipient-ID: for recipients. This specification defines six (6) identifiers with which the public key used may be indicated in each of these fields. In the next section the 3 name forms are described in detail. Following that is the specification of the 6 identifiers. 4.1. Name Forms There are 3 name forms specified by this document: email address, distinguished names, and arbitrary strings. 4.1.1. Email Addresses The email address (grammar token <emailstr>) used must be a valid RFC822 address, which is defined in terms of the two grammar tokens <addr-spec> and <route-addr>. The grammar for these two tokens is included in the Appendix as a convenience; the definitive source for these tokens is necessarily RFC822 [1]. <emailstr> ::= <addr-spec> / <route-addr> ; an electronic mail address as defined by ; these two tokens from RFC822 Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 4]
INTERNET DRAFT PEM and MIME July 1994 For example, the string "galvin@tis.com" is an email address. 4.1.2. Arbitrary Strings The arbitrary string (grammar token <string>) must chosen from the us- ascii character set and must have a length of at least 1. It is possible to encode the actual string in such a way that only characters from the us-ascii character set are generated, but there is no mechanism for conveying to a recipient the encoding that was used. <string> ::= ; a non-null sequence of us-ascii characters For example, the string Jim "the SAAG mailing list maintainer" Galvin is an arbitrary string. 4.1.3. Distinguished Names The distinguished name (grammar token <dname>) must be constructed according to the guidelines of the X.500 Directory. For the purposes of conveying a distinguished name from an originator to a recipient, it must be ASN.1 encoded and then printably encoded according to the base64 encoding defined by MIME. <dnamestr> ::= <encbin> ; a printably encoded, ASN.1 encoded ; distinguished name ** EXAMPLE DISTINGUISHED NAME ** 4.2. Identifiers There are 6 identifiers specified by this document: email address, arbitrary string, distinguished name, PGP key identifier, the public key itself, and the issuer name and serial number pair from a certificate. All of these have approximately the same structure as follows: TYPE, KEYID, STRING Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 5]
INTERNET DRAFT PEM and MIME July 1994 The TYPE field is a literal string, one for each of the possible identifiers. The KEYID field is used to distinguish between the multiple public keys that may be associated with the name form in the STRING field. In 3 of the identifiers its value is arbitrary, chosen by the owner of the key pair, except that it must be distinct from all the other KEYIDs used by the owner. Suggested values include a portion (low-order 16 or 32 bits) or all of the actual public key used. In the other 3 identifiers the value is still chosen by the owner of the public key and it must still be unique, but its value is chosen from a more restricted alphabet. The STRING field is the name form and has a different syntax according to the value of the TYPE field. The identifier used in each of the originator and recipient fields is described by the following grammar. The definition of the key identifier token is included here since it used by several of the identifiers below. <id> ::= <nameid> / <id-publickey> / <id-issuer> <nameid> ::= <id-email> / <id-string> / <id-dname> / <id-pgp> <keyid> ::= <encbin> ; a printably encoded non-null sequence of octets Each of the identifier name forms is described below. 4.2.1. Email Address The email address identifier has the following syntax. <id-email> ::= "EN" "," <keyid> "," <emailstr> CRLF 4.2.2. Arbitrary String The arbitrary string identifier has the following syntax. <id-string> ::= "STR" "," <keyid> "," <string> CRLF Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 6]
INTERNET DRAFT PEM and MIME July 1994 4.2.3. Distinguished Name The distinguished name identifier has the following syntax. <id-dname> ::= "DN" "," <keyid> "," <dnamestr> CRLF The actual form and syntax of the distinguished name is outside the scope of this specification. RFC1422 specifies one possible form based on a particular choice of a certification hierarchy for certificates. 4.2.4. PGP Public Key The PGP public key identifier has the following syntax. <id-pgp> ::= "PGP2" ",0x" <pgp-keyid> "," <string> CRLF <pgp-keyid> ::= ; a sequence from the following alphabet: {0-9, A-F} ; which is either exactly six or eight characters long 4.2.5. Public Key The public key identifier has the following syntax. This identifer, as compared to the others, has the unique property that the STRING element is optional and, when included, is not a string but rather one of four of the other identifiers. <id-publickey> ::= "PK" "," <publickey> [ "," <nameid> ] CRLF <publickey> ::= <encbin> ; a printably encoded, ASN.1 encoded ; subjectPublicKeyInfo In normal usage, the STRING element is expected to be absent. When present, it represents a mechanism by which an identifier (name form and key identifier) can be associated with a public key. Recipients of a public key identifier must take care to verify the accuracy of the purported association. If not, it may be possible for a malicious originator to assert an identifier that accords the originator unauthorized privileges. See Section 7.2 for more details. The object subjectPublicKeyInfo is imported from the X.500 Directory from the certificate object. It is currently the best choice for a Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 7]
INTERNET DRAFT PEM and MIME July 1994 general purpose public key encoding. 4.2.6. Issuer Name and Serial Number The issuer name and serial number identifier has the following syntax. <id-issuer> ::= "IS" "," <dnamestr> "," <serial> CRLF <serial> ::= 1*<hexchar> ; hex dump of the serial number of a certificate The <id-issuer> identifier is included for backward compatibility with the ID-ASymmetric fields defined in [3]. The older fields are easily converted to this new form by prefixing the old value with "IS," and replacing the field name with an appropriate new ID field. 5. Applying PEM Security Services to MIME Body Parts The next section describes the processing steps necessary to prepare a MIME body part for the application of PEM security services. The succeeding two sections describe the content of the multipart/signed and multipart/encrypted body parts resulting from the application of PEM security services to MIME body parts. 5.1. PEM Processing Steps The definition of the multipart/signed and multipart/encrypted body parts in [7] specifies three steps for creating both body parts. (1) The body part is to be protected is created according to a local convention. (2) The body part is prepared for protection according to the protocol parameter. (3) The prepared body part is protected according to the protocol parameter. This specification makes no changes to step one in the sequence. For step two, there is no preparation necessary for the encryption service. For the digital signature service, the body part must be canonicalized as described below. This specification makes no changes to step three in the sequence. Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 8]
INTERNET DRAFT PEM and MIME July 1994 Prior to the application of the digital signature service, the body part must be in a canonical form. Transforming the body part to be signed into a canonical form is a necessary and essential step in the digital signature process. The canonical form must satisfy the property that it is uniquely and unambiguously representable in both the originator and recipient's local environment. This is required in order to ensure that both the originator and recipient have the same data with which to calculate the digital signature; the originator needs to be able to include the digital signature value when transferring the body part, while the recipient needs to be able to compare a re-computed value with the received value. Further, the canonical form should satisfy the property that it is representable on as many different host computers as possible. By satisfying this property, signed data may be forwarded by recipients to additional recipients, who will also be able to verify the original signature. This service is called forwardable authentication. The canonicalization transformation is a two step process. First, the body part must be converted to canonical representation suitable for transport between originators and recipients. Second, the body part must have its line delimiters canonicalized prior to computing the digital signature and prior to each verification of the digital signature. The canonical representation of all body parts is specified to be 7bit, as defined by [2]. Since the headers of body parts are already required to be representable in 7bit, this step requires that if the data to be signed is not already 7bit it must be encoded with an appropriate MIME content transfer encoding. Note, since the MIME standard explicitly disallows nested content transfer encodings, i.e., the content types multipart and message may not themselves be encoded, body parts enclosed within, for example, a multipart content type, must be encoded in a 7bit representation. Any valid MIME encoding may be selected. The 7bit representation of the data must be transferred to the recipient. As may be required by MIME, an appropriate Content- Transfer-Encoding: header is included with the data. Upon receipt, a MIME implementation would verify the signature of the data prior to decoding the data and displaying it to the recipient. Representing all complex content types as 7bit transforms them into text-based content types. However, text-based content types present a unique problem. In particular, the line delimiter used on a text-based content type is specific to a local environment; different environments use the single character carriage-return (<CR>), the single character line-feed (<LF>), or the two character sequence "carriage-return line- Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 9]
INTERNET DRAFT PEM and MIME July 1994 feed (<CR><LF>)." The application of the digital signature service requires that the same line delimiter be used by both the originator and the recipient. This document specifies that the two character sequence "<CR><LF>" must be used as the line delimiter. Thus, the canonicalization transformation includes the transformation of the local line delimiter to the two character sequence "<CR><LF>". The transformation to the canonical line delimiter is only required for the purposes of computing the digital signature. Thus, originators must apply the canonical line delimiter transformation before computing the digital signature but must transfer the data without the canonical line delimiter transformation. Similarly, recipients must apply the canonical line delimiter transformation before computing the digital signature. NOTE: An originator can not transfer the content with the canonical line delimiter transformation intact because the transformation process is not idempotent. In particular, SMTP servers may themselves convert the canonical line delimiter to a local line delimiter, prior to the message being delivered to the user. Thus, a recipient has no way of knowing if the transformation is present or not. Thus, if the recipient applies the transformation to a content in which it is already present, the resulting content may have two line delimiters present, which would cause the verification of the signature to fail. IMPLEMENTORS NOTE: Implementors should be aware that the transformation to a canonical representation is a function that is available even in a minimally compliant MIME user agent. Further, the canonical line delimiter transformation required here is distinct from the same transformation included in that function. Specifically, the line delimiter transformation in the former case is performed prior to the application of the canonical representation while it is performed after the application of the canonical representation in the latter case. 5.2. Use of multipart/signed Content Type When this content type is used, the value of the required parameter "protocol" is "pem" and the value of the required parameter "hashalg" is Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 10]
INTERNET DRAFT PEM and MIME July 1994 one of the valid choices from [5], for example: Content-Type: multipart/signed; protocol="pem"; hashalg="md5"; boundary="Signed Message" --Signed Message Content-Type: text/plain This is some example text. --Signed Message Content-Type: application/signature <pemsig> --Signed Message-- where the <pemsig> token is defined as follows. <pemsig> ::= <version> ( 1*<origasymflds> ) <version> ::= "Version:" "5" CRLF <origasymflds> ::= <origid> <micinfo> <origid> ::= "Originator-ID:" <id> CRLF The token <id> is defined in Section 4.2. The only valid value for a Content-Transfer-Encoding: header, if included, is "7bit". 5.3. Use of multipart/encrypted Content Type When this content type is used, the value of the required parameter "protocol" is "pem", for example: Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 11]
INTERNET DRAFT PEM and MIME July 1994 Content-Type: multipart/encrypted; protocol="pem"; boundary="Encrypted Message" --Encrypted Message Content-Type: application/keys <pemkeys> --Encrypted Message Content-Type: application/octet-stream <encrypted data> --Encrypted Message-- where the <pemkeys> token is defined as follows. <pemkeys> ::= <version> <dekinfo> 1*<recipasymflds> <version> ::= "Version:" "5" CRLF <recipasymflds> ::= <recipid> <asymkeyinfo> <recipid> ::= "Recipient-ID:" <id> CRLF <asymkeyinfo> ::= "Key-Info" ":" <ikalgid> "," <asymencdek> CRLF The token <id> is defined in Section 4.2. 6. Removing PEM Security Services from PEM Body Parts This section describes the processing steps necessary to verify or decrypt the PEM security services that have been applied to MIME body parts. Outer layers of PEM security services must be processed prior to processing inner layers of PEM security services. Processing includes a user choosing to display a content without removing the PEM security services. The definition of the multipart/signed and multipart/encrypted body parts in [7] specifies three steps for receiving both body parts. (1) The protected body part and the control information body part are prepared for processing. Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 12]
INTERNET DRAFT PEM and MIME July 1994 (2) The prepared body parts are made available to the protection removal process. (3) The results of the protection removal process are made available to the user and processing continues with the unprotected body part, as returned by the protection removal process. For step one, the preparation for digitally signed and encrypted body parts is different, as described below. No changes are required to steps two and three in the sequence. For multipart/signed body parts, the control information is prepared by removing any content transfer encodings that may be present. The digitally signed body part is prepared by leaving the content transfer encodings intact and canonicalizing the line delimiters according to Step 2 of Section 5.1. Multipart/encrypted body parts are prepared by removing the content transfer encodings, if present, from both the control information and the encrypted body part. 7. Definition of New Content Types This document defines two new content types, the contents of which comprise a replacement mechanism for [6]. The first content type is application/key-request, which replaces the certification and CRL- retrieval request messages. The second content type is application/key-data, which replaces the certification reply message, the crl-storage request message, and the crl-retrieval reply message. There were no requirements for a crl-storage reply message and none are specified in this document. This document includes a specification for a public key and certificate request message, which were previously undefined. NOTE: RFC1424 has some descriptive text, especially for certification messages, that should probably be included. 7.1. application/key-request Content Type Definition (1) MIME type name: application (2) MIME subtype name: key-request Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 13]
INTERNET DRAFT PEM and MIME July 1994 (3) Required parameters: none (4) Optional parameters: none (5) Encoding considerations: quoted-printable is always sufficient (6) Security Considerations: none The content of this body part corresponds to the following production. <request> ::= <version> ( <subject> / <issuer> / <certification> ) <version> ::= "Version:" "5" CRLF <subject> ::= "Subject:" <id> CRLF <issuer> ::= "Issuer:" <id> CRLF <certification> ::= "Certification:" <encbin> CRLF This content type is used to provide for some of the requests described in [6]. The information in the body part is entirely independent of any other body part. As such, the application/key-request content type is an independent body part. The certification request, certificate-retrieval request and crl- retrieval request are provided for directly. If the content contains a Certification: field it requests certification of the self-signed certificate in the field value. If the content contains an Issuer: field it requests the certificate revocation list chain beginning with the issuer identified in the field value. If the content contains a Subject: field it requests either the public key of the subject or the certificate chain beginning with the subject identified in the field value, or both. The Subject: and Issuer: fields each contain a value of type <id>, which is defined in Section 4.2. The crl-storage request is provided for by the application/key-data content type described in the next section. In each case, the response is transmitted in an application/key-data content type. When returning public keys, certificate chains, and certificate revocation list chains, if there exists more than one, several application/key-data contents are to be returned in the reply Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 14]
INTERNET DRAFT PEM and MIME July 1994 message, one for each. 7.2. application/key-data Content Type Definition The principal objective of this content type is to convey cryptographic keying material from an originator to a recipient. However, no explicit provision is made for determining the authenticity or accuracy of the data being conveyed. In particular, when a public key and the identifier for its owner is conveyed, there is nothing to prevent an originator or any interloper along the path from an originator to a recipient from substituting alternate values for either the public key or the identifier, thus setting up the recipient to potentially send sensitive information that may be intercepted and disclosed inappropriately. It is incumbent upon a recipient to verify the authenticity and accuracy of the data received prior to its use. The problem is addressed by the use of certificates, since a certification hierarchy is a well-defined mechanism that conveniently supports the automatic verification of the data. Alternatively, the application/key-data body part could be digitally signed by the originator. In this way, if a recipient believes that correct originator's public key is available locally and if the recipient believes the originator would convey accurate data, then the key data received from the originator can be believed. NOTE: Insofar as a certificate represents a mechanism by which an issuer vouches for the binding between the name and public key it embodies, the signing of an application/key-data body part is a similar mechanism. (1) MIME type name: application (2) MIME subtype name: key-data (3) Required parameters: none (4) Optional parameters: none (5) Encoding considerations: quoted-printable is always sufficient. (6) Security Considerations: none The content of this body part corresponds to the following production. Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 15]
INTERNET DRAFT PEM and MIME July 1994 <certdata> ::= <version> ( <keydata> / <certchain> / <crlchain> ) <version> ::= "Version:" "5" CRLF <keydata> ::= "Key:" <id> "," <nameid> CRLF <certchain> ::= <cert> *( [ <crl> ] <cert> ) <crlchain> ::= 1*( <crl> [ <cert> ] ) <cert> ::= "Certificate:" <encbin> CRLF <crl> ::= "CRL:" <encbin> CRLF This content type is used to transfer public keys, certificate chains, or Certificate Revocation List (CRL) chains. The information in the body part is entirely independent of any other body part. (Note that the converse is not true: the validity of a protected body part cannot be determined without the proper public keys, certificates, or current CRL information.) As such, the application/key-data content type is an independent body part. The <keydata> production contains exactly one public key. It is used to bind a public key with its corresponding name form and key identifier. It is recommended that when responders are returning this information that the enclosing body part be digitally signed by the responder in order to protect the information. The <certchain> production contains one certificate chain. A certificate chain starts with a certificate and continues with the certificates of subsequent issuers. Each issuer certificate included must have issued the preceding certificate. For each issuer, a CRL may be supplied. A CRL in the chain belongs to the immediately following issuer. Therefore, it potentially contains the immediately preceding certificate. The <crlchain> production contains one certificate revocation list chain. The CRLs in the chain begin with the requested CRL and continue with the CRLs of subsequent issuers. The issuer of each CRL is presumed to have issued a certificate for the issuer of the preceding CRL. For each CRL, the issuer's certificate may be supplied. A certificate in the chain must belong to the issuer of the immediately preceding CRL. The relationship between a certificate and an immediately preceding CRL is the same in both <certchain> and <crlchain>. In a <certchain> the CRLs are optional. In a <crlchain> the certificates are optional. Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 16]
INTERNET DRAFT PEM and MIME July 1994 8. Examples NOTE: To be included upon completion of implementation. 9. Observations The use of the pre-submission and post-delivery algorithms to combine PEM and MIME capabilities exhibits several properties: (1) It allows privacy-enhancement of an arbitrary content, not just the body of an RFC822 message. (2) For a multipart or message content, it allows the user to specify different privacy enhancements to be applied to different components of the structure of the content. (3) It provides for messages containing several privacy enhanced contents, thereby removing the requirement for PEM software to be able to generate or interpret a single content which intermixes both unenhanced and enhanced components. The use of a MIME-capable user agent makes complex nesting of enhanced message body parts much easier. For example, the user can separately sign and encrypt a message. This motivates a complete separation of the confidentiality security service from the digital signature security service. That is, different key pairs could be used for the different services and could be protected separately. This means an employee's company could be given access to the (private) decryption key but not the (private) signature key, thereby granting the company the ability to decrypt messages addressed to the employee in emergencies without also granting the company the ability to sign messages as the employee. The use of two private keys requires the ability to maintain multiple certificates for each user. 10. Summary of Changes to PEM Specification This document updates the message encryption and signature procedures defined by [3] and obsoletes the key certification and related services defined by [6]. The changes are enumerated below. (1) The PEM specification currently requires that encryption services be applied only to message bodies that have been signed. By providing for each of the services separately, they may be applied Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 17]
INTERNET DRAFT PEM and MIME July 1994 recursively in any order according to the needs of the requesting application. (2) PEM implementations are currently restricted to processing only text-based electronic mail messages. In fact, the message text is required to be represented by the ASCII character set with "<CR><LF>" line delimiters. This restriction no longer applies. (3) MIME includes transfer encoding operations to ensure the unmodified transfer of body parts, which obviates these services in PEM. (4) PEM specifies a Proc-Type: header field to identify the type of processing that was performed on the message. This functionality is subsumed by the MIME Content-Type: headers. The Proc-Type: header also included a decimal number that was used to distinguish among incompatible encapsulated header field interpretations which may arise as changes are made to the PEM standard. This functionality is replaced by the Version: header specified in this document. (5) PEM specifies a Content-Domain: header, the purpose of which is to describe the type of the content which is represented within a PEM message's encapsulated text. This functionality is subsumed by the MIME Content-Type: headers. (6) The PEM specifications include a document that defines new types of PEM messages, specified by unique values used in the Proc-Type: header, to be used to request certificate and certificate revocation list information. This functionality is subsumed by two new content types specified in this document. (7) The header fields having to do with certificates (Originator- Certificate: and Issuer-Certificate:) and CRLs (CRL:) are relegated for use only in the application/key-data and application/key- request content types and are no longer allowed in the header portion of a PEM signed or encrypted message. (8) The grammar specified here explicitly separates the header fields that may appear for the encryption and signature security services. It is the intent of this document to specify a precise expression of the allowed header fields; there is no intent to reduce the functionality of combinations of encryption and signature security from those of [3]. Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 18]
INTERNET DRAFT PEM and MIME July 1994 (9) With the separation of the encryption and signature security services, there is no need for a MIC-Info: field in the headers associated with an encrypted message. (10) In [3], when asymmetric key management is used, an Originator-ID field is required in order to identify the private key used to sign the MIC argument in the MIC-Info: field. Because no MIC-Info: field is associated with the encryption security service under asymmetric key managment, there is no requirement in that case to include an Originator-ID field. These changes represent a departure in mechanism, not in effect, from those specified in [3] and [6]. The following technical changes to [3] and [4] are also specified by this document. (1) The grammar specified here explicitly excludes symmetric key management. Currently, there are no generally available implementations of symmetric key management nor are there any known plans for implementing it. As a result, the IETF standards process will require this feature to be dropped when the documents are promoted to draft standard status from proposed standard status. (2) This document requires all data that is to be digitally signed to be represented in 7bit form. (3) This document broadens the allowable name forms that users may use to identify their public keys. Users may use arbitrary strings and email addresses as their name. Further, users may distribute their public key directly in lieu of using certificates. In support of this change the Originator-ID-ASymmetric: and Recipient-ID- ASymmetric: fields are deprecated in favor of Originator-ID: and Recipient-ID: fields, respectively. 11. Collected Grammar The following is a summary of the grammar presented in this document. (1) Signature headers Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 19]
INTERNET DRAFT PEM and MIME July 1994 <pemsig> ::= <version> ( 1*<origasymflds> ) <version> ::= "Version:" "5" CRLF <origasymflds> ::= <origid> <micinfo> <origid> ::= "Originator-ID:" <id> CRLF (2) Encryption Headers <pemkeys> ::= <version> <dekinfo> 1*<recipasymflds> <version> ::= "Version:" "5" CRLF <recipasymflds> ::= <recipid> <asymkeyinfo> <recipid> ::= "Recipient-ID:" <id> CRLF <asymkeyinfo> ::= "Key-Info" ":" <ikalgid> "," <asymencdek> CRLF (3) Identifier Name Forms Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 20]
INTERNET DRAFT PEM and MIME July 1994 <id> ::= <nameid> / <id-publickey> / <id-issuer> <nameid> ::= <id-email> / <id-string> / <id-dname> / <id-pgp> <id-email> ::= "EN" "," <keyid> "," <emailstr> CRLF <id-string> ::= "STR" "," <keyid> "," <string> CRLF <id-dname> ::= "DN" "," <keyid> "," <dnamestr> CRLF <id-pgp> ::= "PGP2" ",0x" <pgp-keyid> "," <string> CRLF <id-publickey> ::= "PK" "," <publickey> [ "," <nameid> ] CRLF <id-issuer> ::= "IS" "," <dnamestr> "," <serial> CRLF <keyid> ::= <encbin> ; a printably encoded non-null sequence of octets <emailstr> ::= <addr-spec> / <route-addr> ; an electronic mail address as defined by ; these two tokens from RFC822 <string> ::= ; a non-null sequence of us-ascii characters <dnamestr> ::= <encbin> ; a printably encoded, ASN.1 encoded ; distinguished name <pgp-keyid> ::= ; a sequence from the following alphabet: {0-9, A-F} ; which is either exactly six or eight characters long <publickey> ::= <encbin> ; a printably encoded, ASN.1 encoded ; subjectPublicKeyInfo <serial> ::= 1*<hexchar> ; hex dump of the serial number of a certificate (4) Request Headers (certificate, certification, etc.) Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 21]
INTERNET DRAFT PEM and MIME July 1994 <request> ::= <version> ( <subject> / <issuer> / <certification> ) <version> ::= "Version:" "5" CRLF <subject> ::= "Subject:" <id> CRLF <issuer> ::= "Issuer:" <id> CRLF <certification> ::= "Certification:" <encbin> CRLF (5) Data Headers (certificate, certification revocation list) <certdata> ::= <certchain> / <crlchain> <certchain> ::= <version> <cert> *( [ <crl> ] <cert> ) <crlchain> ::= <version> 1*( <crl> [ <cert> ] ) <cert> ::= "Certificate:" <encbin> CRLF <crl> ::= "CRL:" <encbin> CRLF <version> ::= "Version:" "5" CRLF 12. Security Considerations NOTE: to be done 13. Acknowledgements David H. Crocker suggested the use of a multipart structure for MIME-PEM interaction. 14. References [1] David H. Crocker. Standard for the Format of ARPA Internet Text Messages. RFC 822, University of Delaware, August 1982. [2] Nathaniel Borenstein and Ned Freed. MIME (Multipurpose Internet Mail Extension) Part One: Mechanisms for Specifying and Describing the Format of Internet Message Bodies. RFC 1521, Bellcore and Innosoft, September 1993. Obsoletes RFC 1341. [3] John Linn. Privacy Enhancement for Internet Electronic Mail: Part I: Message Encryption and Authentication Procedures. RFC 1421, February 1993. Obsoletes RFC 1113. [4] Steve Kent. Privacy Enhancement for Internet Electronic Mail: Part II: Certificate-Based Key Management. RFC 1422, BBN Communications, February 1993. Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 22]
INTERNET DRAFT PEM and MIME July 1994 [5] David M. Balenson. Privacy Enhancement for Internet Electronic Mail: Part III: Algorithms, Modes, and Identifiers. RFC 1423, Trusted Information Systems, February 1993. [6] Burton S. Kaliski. Privacy Enhancement for Internet Electronic Mail: Part IV: Key Certification and Related Services. RFC 1424, RSA Laboratories, February 1993. [7] James Galvin, Sandy Murphy, Steve Crocker, and Ned Freed. Security Multiparts for MIME: Multipart/Signed and Multipart/Encrypted. RFC XXXX, Trusted Information Systems and Innosoft, XXXX 1994. 15. Authors' Addresses Steve Crocker email: crocker@tis.com James M. Galvin email: galvin@tis.com Sandra Murphy email: murphy@tis.com Trusted Information Systems 3060 Washington Road Glenwood, MD 21738 Tel: +1 301 854 6889 FAX: +1 301 854 5363 Ned Freed Innosoft International, Inc. 250 West First Street, Suite 240 Claremont, CA 91711 Tel: +1 909 624 7907 FAX: +1 909 621 5319 email: ned@innosoft.com Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 23]
INTERNET DRAFT PEM and MIME July 1994 16. Appendix: Imported Grammar The following productions are taken from [3]. The grammar presented in [3] remains the authoritative source for these productions; they are repeated here for the convenience of the reader. <dekinfo> ::= "DEK-Info" ":" <dekalgid> [ "," <dekparameters> ] CRLF <micinfo> ::= "MIC-Info" ":" <micalgid> "," <ikalgid> "," <asymsignmic> CRLF <encbin> ::= 1*<encbingrp> <encbingrp> ::= 4*4<encbinchar> <encbinchar> ::= ALPHA / DIGIT / "+" / "/" / "=" The following productions are taken from [5]. The grammar presented in [5] remains the authoritative source for these productions; they are repeated here for the convenience of the reader. <dekalgid> ::= "DES-CBC" <ikalgid> ::= "DES-EDE" / "DES-ECB" / "RSA" <micalgid> ::= "RSA-MD2" / "RSA-MD5" <dekparameters> ::= <DESCBCparameters> <DESCBCparameters> ::= <IV> <IV> ::= <hexchar16> <asymsignmic> ::= <RSAsignmic> <RSAsignmic> ::= <encbin> <asymencdek> ::= <RSAencdek> <RSAencdek> ::= <encbin> <hexchar16> ::= 16*16<hexchar> <hexchar> ::= DIGIT / "A" / "B" / "C" / "D" / "E" / "F" ; no lower case The following productions are taken from [1]. The grammar presented in [1] remains the authorative source for these productions; they are repeated here for the convenience of the reader. Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 24]
INTERNET DRAFT PEM and MIME July 1994 <addr-spec> ::= <local-part> "@" <domain> ; global address <local-part> ::= <word> *( "." <word> ) ; uninterpreted ; case-preserved <domain> ::= <sub-domain> *( "." <sub-domain> ) <sub-domain> ::= <domain-ref> / <domain-literal> <domain-ref> ::= <atom> ; symbolic reference <route-addr> ::= "<" [ <route> ] <addr-spec> ">" <route> ::= 1# ( "@" <domain> ) ":" ; path-relative <word> ::= <atom> / <quoted-string> <quoted-string> ::= """ *( <qtext> / <quoted-pair> ) """ <qtext> ::= (any <CHAR> excepting """, " and including <linear-white-space>) <quoted-pair> ::= " <linear-white-space> ::= 1*( [ CRLF ] <LWSP-char> ) ; semantics = SPACE ; CRLF => folding <LWSP-char> ::= SPACE / HTAB ; semantics = SPACE <atom> ::= 1*(any <CHAR> except <specials>, SPACE and <CTL>s) <CHAR> ::= <any ASCII character> <CTL> ::= <any ASCII control character and DEL> <specials> ::= "(" / ")" / "<" / ">" / "@" ; Must be in quoted- / "," / ";" / ":" / " / "." / "[" / "]" ; within a word. Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 25]
INTERNET DRAFT PEM and MIME July 1994 Table of Contents 1 Status of this Memo ............................................. 1 2 Abstract ........................................................ 1 3 Introduction .................................................... 2 4 Name Forms and Identifiers ...................................... 3 4.1 Name Forms .................................................... 4 4.1.1 Email Addresses ............................................. 4 4.1.2 Arbitrary Strings ........................................... 5 4.1.3 Distinguished Names ......................................... 5 4.2 Identifiers ................................................... 5 4.2.1 Email Address ............................................... 6 4.2.2 Arbitrary String ............................................ 6 4.2.3 Distinguished Name .......................................... 7 4.2.4 PGP Public Key .............................................. 7 4.2.5 Public Key .................................................. 7 4.2.6 Issuer Name and Serial Number ............................... 8 5 Applying PEM Security Services to MIME Body Parts ............... 8 5.1 PEM Processing Steps .......................................... 8 5.2 Use of multipart/signed Content Type .......................... 10 5.3 Use of multipart/encrypted Content Type ....................... 11 6 Removing PEM Security Services from PEM Body Parts .............. 12 7 Definition of New Content Types ................................. 13 7.1 application/key-request Content Type Definition ............... 13 7.2 application/key-data Content Type Definition .................. 15 8 Examples ........................................................ 17 9 Observations .................................................... 17 10 Summary of Changes to PEM Specification ........................ 17 11 Collected Grammar .............................................. 19 12 Security Considerations ........................................ 22 13 Acknowledgements ............................................... 22 14 References ..................................................... 22 15 Authors' Addresses ............................................. 23 16 Appendix: Imported Grammar ..................................... 24 Crocker/Freed/Galvin/Murphy Expires: January 1995 [Page 26]
