RFC 5804
A Protocol for Remotely Managing Sieve Scripts
Internet Engineering Task Force (IETF) A. Melnikov, Ed. Request for Comments: 5804 Isode Limited Category: Standards Track T. Martin ISSN: 2070-1721 BeThereBeSquare, Inc. July 2010 A Protocol for Remotely Managing Sieve ScriptsAbstract
Sieve scripts allow users to filter incoming email. Message stores are commonly sealed servers so users cannot log into them, yet users must be able to update their scripts on them. This document describes a protocol "ManageSieve" for securely managing Sieve scripts on a remote server. This protocol allows a user to have multiple scripts, and also alerts a user to syntactically flawed scripts. Status of This Memo This is an Internet Standards Track document. This document is a product of the Internet Engineering Task Force (IETF). It represents the consensus of the IETF community. It has received public review and has been approved for publication by the Internet Engineering Steering Group (IESG). Further information on Internet Standards is available in Section 2 of RFC 5741. Information about the current status of this document, any errata, and how to provide feedback on it may be obtained at http://www.rfc-editor.org/info/rfc5804. Copyright Notice Copyright (c) 2010 IETF Trust and the persons identified as the document authors. All rights reserved. This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.
Table of Contents
1. Introduction ....................................................3 1.1. Commands and Responses .....................................3 1.2. Syntax .....................................................3 1.3. Response Codes .............................................3 1.4. Active Script ..............................................6 1.5. Quotas .....................................................6 1.6. Script Names ...............................................6 1.7. Capabilities ...............................................7 1.8. Transport ..................................................9 1.9. Conventions Used in This Document .........................10 2. Commands .......................................................10 2.1. AUTHENTICATE Command ......................................11 2.1.1. Use of SASL PLAIN Mechanism over TLS ...............16 2.2. STARTTLS Command ..........................................16 2.2.1. Server Identity Check ..............................17 2.3. LOGOUT Command ............................................20 2.4. CAPABILITY Command ........................................20 2.5. HAVESPACE Command .........................................20 2.6. PUTSCRIPT Command .........................................21 2.7. LISTSCRIPTS Command .......................................23 2.8. SETACTIVE Command .........................................24 2.9. GETSCRIPT Command .........................................25 2.10. DELETESCRIPT Command .....................................25 2.11. RENAMESCRIPT Command .....................................26 2.12. CHECKSCRIPT Command ......................................27 2.13. NOOP Command .............................................28 2.14. Recommended Extensions ...................................28 2.14.1. UNAUTHENTICATE Command ............................28 3. Sieve URL Scheme ...............................................29 4. Formal Syntax ..................................................31 5. Security Considerations ........................................37 6. IANA Considerations ............................................38 6.1. ManageSieve Capability Registration Template ..............39 6.2. Registration of Initial ManageSieve Capabilities ..........39 6.3. ManageSieve Response Code Registration Template ...........41 6.4. Registration of Initial ManageSieve Response Codes ........41 7. Internationalization Considerations ............................46 8. Acknowledgements ...............................................46 9. References .....................................................47 9.1. Normative References ......................................47 9.2. Informative References ....................................48
1. Introduction
1.1. Commands and Responses
A ManageSieve connection consists of the establishment of a client/ server network connection, an initial greeting from the server, and client/server interactions. These client/server interactions consist of a client command, server data, and a server completion result response. All interactions transmitted by client and server are in the form of lines, that is, strings that end with a CRLF. The protocol receiver of a ManageSieve client or server is either reading a line or reading a sequence of octets with a known count followed by a line.1.2. Syntax
ManageSieve is a line-oriented protocol much like [IMAP] or [ACAP], which runs over TCP. There are three data types: atoms, numbers and strings. Strings may be quoted or literal. See [ACAP] for detailed descriptions of these types. Each command consists of an atom (the command name) followed by zero or more strings and numbers terminated by CRLF. All client queries are replied to with either an OK, NO, or BYE response. Each response may be followed by a response code (see Section 1.3) and by a string consisting of human-readable text in the local language (as returned by the LANGUAGE capability; see Section 1.7), encoded in UTF-8 [UTF-8]. The contents of the string SHOULD be shown to the user ,and implementations MUST NOT attempt to parse the message for meaning. The BYE response SHOULD be used if the server wishes to close the connection. A server may wish to do this because the client was idle for too long or there were too many failed authentication attempts. This response can be issued at any time and should be immediately followed by a server hang-up of the connection. If a server has an inactivity timeout resulting in client autologout, it MUST be no less than 30 minutes after successful authentication. The inactivity timeout MAY be less before authentication.1.3. Response Codes
An OK, NO, or BYE response from the server MAY contain a response code to describe the event in a more detailed machine-parsable fashion. A response code consists of data inside parentheses in the form of an atom, possibly followed by a space and arguments.
Response codes are defined when there is a specific action that a client can take based upon the additional information. In order to support future extension, the response code is represented as a slash-separated (Solidus, %x2F) hierarchy with each level of hierarchy representing increasing detail about the error. Response codes MUST NOT start with the Solidus character. Clients MUST tolerate additional hierarchical response code detail that they don't understand. For example, if the client supports the "QUOTA" response code, but doesn't understand the "QUOTA/MAXSCRIPTS" response code, it should treat "QUOTA/MAXSCRIPTS" as "QUOTA". Client implementations MUST tolerate (ignore) response codes that they do not recognize. The currently defined response codes are the following: AUTH-TOO-WEAK This response code is returned in the NO or BYE response from an AUTHENTICATE command. It indicates that site security policy forbids the use of the requested mechanism for the specified authentication identity. ENCRYPT-NEEDED This response code is returned in the NO or BYE response from an AUTHENTICATE command. It indicates that site security policy requires the use of a strong encryption mechanism for the specified authentication identity and mechanism. QUOTA If this response code is returned in the NO/BYE response, it means that the command would have placed the user above the site-defined quota constraints. If this response code is returned in the OK response, it can mean that the user's storage is near its quota, or it can mean that the account exceeded its quota but that the condition is being allowed by the server (the server supports so-called soft quotas). The QUOTA response code has two more detailed variants: "QUOTA/MAXSCRIPTS" (the maximum number of per-user scripts) and "QUOTA/MAXSIZE" (the maximum script size). REFERRAL This response code may be returned with a BYE result from any command, and includes a mandatory parameter that indicates what server to access to manage this user's Sieve scripts. The server will be specified by a Sieve URL (see Section 3). The scriptname
portion of the URL MUST NOT be specified. The client should authenticate to the specified server and use it for all further commands in the current session. SASL This response code can occur in the OK response to a successful AUTHENTICATE command and includes the optional final server response data from the server as specified by [SASL]. TRANSITION-NEEDED This response code occurs in a NO response of an AUTHENTICATE command. It indicates that the user name is valid, but the entry in the authentication database needs to be updated in order to permit authentication with the specified mechanism. This is typically done by establishing a secure channel using TLS, verifying server identity as specified in Section 2.2.1, and finally authenticating once using the [PLAIN] authentication mechanism. The selected mechanism SHOULD then work for authentications in subsequent sessions. This condition can happen if a user has an entry in a system authentication database such as Unix /etc/passwd, but does not have credentials suitable for use by the specified mechanism. TRYLATER A command failed due to a temporary server failure. The client MAY continue using local information and try the command later. This response code only makes sense when returned in a NO/BYE response. ACTIVE A command failed because it is not allowed on the active script, for example, DELETESCRIPT on the active script. This response code only makes sense when returned in a NO/BYE response. NONEXISTENT A command failed because the referenced script name doesn't exist. This response code only makes sense when returned in a NO/BYE response. ALREADYEXISTS A command failed because the referenced script name already exists. This response code only makes sense when returned in a NO/BYE response.
TAG This response code name is followed by a string specified in the command. See Section 2.13 for a possible use case. WARNINGS This response code MAY be returned by the server in the OK response (but it might be returned with the NO/BYE response as well) and signals the client that even though the script is syntactically valid, it might contain errors not intended by the script writer. This response code is typically returned in response to PUTSCRIPT and/or CHECKSCRIPT commands. A client seeing such response code SHOULD present the returned warning text to the user.1.4. Active Script
A user may have multiple Sieve scripts on the server, yet only one script may be used for filtering of incoming messages. This is the active script. Users may have zero or one active script and MUST use the SETACTIVE command described below for changing the active script or disabling Sieve processing. For example, users may have an everyday script they normally use and a special script they use when they go on vacation. Users can change which script is being used without having to download and upload a script stored somewhere else.1.5. Quotas
Servers SHOULD impose quotas to prevent malicious users from overflowing available storage. If a command would place a user over a quota setting, servers that impose such quotas MUST reply with a NO response containing the QUOTA response code. Client implementations MUST be able to handle commands failing because of quota restrictions.1.6. Script Names
A Sieve script name is a sequence of Unicode characters encoded in UTF-8 [UTF-8]. A script name MUST comply with Net-Unicode Definition (Section 2 of [NET-UNICODE]), with the additional restriction of prohibiting the following Unicode characters: o 0000-001F; [CONTROL CHARACTERS] o 007F; DELETE o 0080-009F; [CONTROL CHARACTERS]
o 2028; LINE SEPARATOR o 2029; PARAGRAPH SEPARATOR Sieve script names MUST be at least one octet (and hence Unicode character) long. Zero octets script name has a special meaning (see Section 2.8). Servers MUST allow names of up to 128 Unicode characters in length (which can take up to 512 bytes when encoded in UTF-8, not counting the terminating NUL), and MAY allow longer names. A server that receives a script name longer than its internal limit MUST reject the corresponding operation, in particular it MUST NOT truncate the script name.1.7. Capabilities
Server capabilities are sent automatically by the server upon a client connection, or after successful STARTTLS and AUTHENTICATE (which establishes a Simple Authentication and Security Layer (SASL)) commands. Capabilities may change immediately after a successfully completed STARTTLS command, and/or immediately after a successfully completed AUTHENTICATE command, and/or after a successfully completed UNAUTHENTICATE command (see Section 2.14.1). Capabilities MUST remain static at all other times. Clients MAY request the capabilities at a later time by issuing the CAPABILITY command described later. The capabilities consist of a series of lines each with one or two strings. The first string is the name of the capability, which is case-insensitive. The second optional string is the value associated with that capability. Order of capabilities is arbitrary, but each capability name can appear at most once. The following capabilities are defined in this document: IMPLEMENTATION - Name of implementation and version. This capability MUST always be returned by the server. SASL - List of SASL mechanisms supported by the server, each separated by a space. This list can be empty if and only if STARTTLS is also advertised. This means that the client must negotiate TLS encryption with STARTTLS first, at which point the SASL capability will list a non-empty list of SASL mechanisms. SIEVE - List of space-separated Sieve extensions (as listed in Sieve "require" action [SIEVE]) supported by the Sieve engine. This capability MUST always be returned by the server.
STARTTLS - If TLS [TLS] is supported by this implementation. Before advertising this capability a server MUST verify to the best of its ability that TLS can be successfully negotiated by a client with common cipher suites. Specifically, a server should verify that a server certificate has been installed and that the TLS subsystem has successfully initialized. This capability SHOULD NOT be advertised once STARTTLS or AUTHENTICATE command completes successfully. Client and server implementations MUST implement the STARTTLS extension. MAXREDIRECTS - Specifies the limit on the number of Sieve "redirect" actions a script can perform during a single evaluation. Note that this is different from the total number of "redirect" actions a script can contain. The value is a non-negative number represented as a ManageSieve string. NOTIFY - A space-separated list of URI schema parts for supported notification methods. This capability MUST be specified if the Sieve implementation supports the "enotify" extension [NOTIFY]. LANGUAGE - The language (<Language-Tag> from [RFC5646]) currently used for human-readable error messages. If this capability is not returned, the "i-default" [RFC2277] language is assumed. Note that the current language MAY be per-user configurable (i.e., it MAY change after authentication). OWNER - The canonical name of the logged-in user (SASL "authorization identity") encoded in UTF-8. This capability MUST NOT be returned in unauthenticated state and SHOULD be returned once the AUTHENTICATE command succeeds. VERSION - This capability MUST be returned by servers compliant with this document or its successor. For servers compliant with this document, the capability value is the string "1.0". Lack of this capability means that the server predates this specification and thus doesn't support the following commands: RENAMESCRIPT, CHECKSCRIPT, and NOOP. Section 2.14 defines some additional ManageSieve extensions and their respective capabilities. A server implementation MUST return SIEVE, IMPLEMENTATION, and VERSION capabilities. A client implementation MUST ignore any listed capabilities that it does not understand.
Example:
S: "IMPlemENTATION" "Example1 ManageSieved v001"
S: "SASl" "DIGEST-MD5 GSSAPI"
S: "SIeVE" "fileinto vacation"
S: "StaRTTLS"
S: "NOTIFY" "xmpp mailto"
S: "MAXREdIRECTS" "5"
S: "VERSION" "1.0"
S: OK
After successful authentication, this might look like this:
Example:
S: "IMPlemENTATION" "Example1 ManageSieved v001"
S: "SASl" "DIGEST-MD5 GSSAPI"
S: "SIeVE" "fileinto vacation"
S: "NOTIFY" "xmpp mailto"
S: "OWNER" "alexey@example.com"
S: "MAXREdIRECTS" "5"
S: "VERSION" "1.0"
S: OK
1.8. Transport
The ManageSieve protocol assumes a reliable data stream such as that
provided by TCP. When TCP is used, a ManageSieve server typically
listens on port 4190.
Before opening the TCP connection, the ManageSieve client first MUST
resolve the Domain Name System (DNS) hostname associated with the
receiving entity and determine the appropriate TCP port for
communication with the receiving entity. The process is as follows:
1. Attempt to resolve the hostname using a [DNS-SRV] Service of
"sieve" and a Proto of "tcp" for the target domain (e.g.,
"example.net"), resulting in resource records such as
"_sieve._tcp.example.net.". The result of the SRV lookup, if
successful, will be one or more combinations of a port and
hostname; the ManageSieve client MUST resolve the returned
hostnames to IPv4/IPv6 addresses according to returned SRV record
weight. IP addresses from the first successfully resolved
hostname (with the corresponding port number returned by SRV
lookup) are used to connect to the server. If connection using
one of the IP addresses fails, the next resolved IP address is
used to connect. If connection to all resolved IP addresses
fails, then the resolution/connect is repeated for the next
hostname returned by SRV lookup.
2. If the SRV lookup fails, the fallback SHOULD be a normal IPv4 or
IPv6 address record resolution to determine the IP address, where
the port used is the default ManageSieve port of 4190.
1.9. Conventions Used in This Document
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in [KEYWORDS].
In examples, "C:" and "S:" indicate lines sent by the client and
server respectively. Line breaks that do not start a new "C:" or
"S:" exist for editorial reasons.
Examples of authentication in this document are using DIGEST-MD5
[DIGEST-MD5] and GSSAPI [GSSAPI] SASL mechanisms.
2. Commands
This section and its subsections describe valid ManageSieve commands.
Upon initial connection to the server, the client's session is in
non-authenticated state. Prior to successful authentication, only
the AUTHENTICATE, CAPABILITY, STARTTLS, LOGOUT, and NOOP (see Section
2.13) commands are valid. ManageSieve extensions MAY define other
commands that are valid in non-authenticated state. Servers MUST
reject all other commands with a NO response. Clients may pipeline
commands (send more than one command at a time without waiting for
completion of the first command). However, a group of commands sent
together MUST NOT have an AUTHENTICATE (*), a STARTTLS, or a
HAVESPACE command anywhere but the last command in the list.
(*) - The only exception to this rule is when the AUTHENTICATE
command contains an initial response for a SASL mechanism that allows
clients to send data first, the mechanism is known to complete in one
round trip, and the mechanism doesn't negotiate a SASL security
layer. Two examples of such SASL mechanisms are PLAIN [PLAIN] and
EXTERNAL [SASL].
2.1. AUTHENTICATE Command
Arguments: String - mechanism String - initial data (optional) The AUTHENTICATE command indicates a SASL [SASL] authentication mechanism to the server. If the server supports the requested authentication mechanism, it performs an authentication protocol exchange to identify and authenticate the user. Optionally, it also negotiates a security layer for subsequent protocol interactions. If the requested authentication mechanism is not supported, the server rejects the AUTHENTICATE command by sending the NO response. The authentication protocol exchange consists of a series of server challenges and client responses that are specific to the selected authentication mechanism. A server challenge consists of a string (quoted or literal) followed by a CRLF. The contents of the string is a base-64 encoding [BASE64] of the SASL data. A client response consists of a string (quoted or literal) with the base-64 encoding of the SASL data followed by a CRLF. If the client wishes to cancel the authentication exchange, it issues a string containing a single "*". If the server receives such a response, it MUST reject the AUTHENTICATE command by sending a NO reply. Note that an empty challenge/response is sent as an empty string. If the mechanism dictates that the final response is sent by the server, this data MAY be placed within the data portion of the SASL response code to save a round trip. The optional initial-response argument to the AUTHENTICATE command is used to save a round trip when using authentication mechanisms that are defined to send no data in the initial challenge. When the initial-response argument is used with such a mechanism, the initial empty challenge is not sent to the client and the server uses the data in the initial-response argument as if it were sent in response to the empty challenge. If the initial-response argument to the AUTHENTICATE command is used with a mechanism that sends data in the initial challenge, the server MUST reject the AUTHENTICATE command by sending the NO response. The service name specified by this protocol's profile of SASL is "sieve". Reauthentication is not supported by ManageSieve protocol's profile of SASL. That is, after a successfully completed AUTHENTICATE command, no more AUTHENTICATE commands may be issued in the same session. After a successful AUTHENTICATE command completes, a server MUST reject any further AUTHENTICATE commands with a NO reply.
However, note that a server may implement the UNAUTHENTICATE extension described in Section 2.14.1. If a security layer is negotiated through the SASL authentication exchange, it takes effect immediately following the CRLF that concludes the successful authentication exchange for the client, and the CRLF of the OK response for the server. When a security layer takes effect, the ManageSieve protocol is reset to the initial state (the state in ManageSieve after a client has connected to the server). The server MUST discard any knowledge obtained from the client that was not obtained from the SASL (or TLS) negotiation itself. Likewise, the client MUST discard any knowledge obtained from the server, such as the list of ManageSieve extensions, that was not obtained from the SASL (and/or TLS) negotiation itself. (Note that a client MAY compare the advertised SASL mechanisms before and after authentication in order to detect an active down- negotiation attack. See below.) Once a SASL security layer is established, the server MUST re-issue the capability results, followed by an OK response. This is necessary to protect against man-in-the-middle attacks that alter the capabilities list prior to SASL negotiation. The capability results MUST include all SASL mechanisms the server was capable of negotiating with that client. This is done in order to allow the client to detect an active down-negotiation attack. If a user- oriented client detects such a down-negotiation attack, it SHOULD either notify the user (it MAY give the user the opportunity to continue with the ManageSieve session in this case) or close the transport connection and indicate that a down-negotiation attack might be in progress. If an automated client detects a down- negotiation attack, it SHOULD return or log an error indicating that a possible attack might be in progress and/or SHOULD close the transport connection. When both [TLS] and SASL security layers are in effect, the TLS encoding MUST be applied (when sending data) after the SASL encoding. Server implementations SHOULD support SASL proxy authentication so that an administrator can administer a user's scripts. Proxy authentication is when a user authenticates as herself/himself but requests the server to act (authorize) as another user. The authorization identity generated by this [SASL] exchange is a "simple username" (in the sense defined in [SASLprep]), and both client and server MUST use the [SASLprep] profile of the [StringPrep] algorithm to prepare these names for transmission or comparison. If preparation of the authorization identity fails or results in an
empty string (unless it was transmitted as the empty string), the server MUST fail the authentication. If an AUTHENTICATE command fails with a NO response, the client MAY try another authentication mechanism by issuing another AUTHENTICATE command. In other words, the client may request authentication types in decreasing order of preference. Note that a failed (NO) response to the AUTHENTICATE command may contain one of the following response codes: AUTH-TOO-WEAK, ENCRYPT- NEEDED, or TRANSITION-NEEDED. See Section 1.3 for detailed description of the relevant conditions. To ensure interoperability, both client and server implementations of the ManageSieve protocol MUST implement the SCRAM-SHA-1 [SCRAM] SASL mechanism, as well as [PLAIN] over [TLS]. Note: use of PLAIN over TLS reflects current use of PLAIN over TLS in other email-related protocols; however, a longer-term goal is to migrate email-related protocols from using PLAIN over TLS to SCRAM- SHA-1 mechanism. Examples (Note that long lines are folded for readability and are not part of protocol exchange): S: "IMPLEMENTATION" "Example1 ManageSieved v001" S: "SASL" "DIGEST-MD5 GSSAPI" S: "SIEVE" "fileinto vacation" S: "STARTTLS" S: "VERSION" "1.0" S: OK C: Authenticate "DIGEST-MD5" S: "cmVhbG09ImVsd29vZC5pbm5vc29mdC5leGFtcGxlLmNvbSIsbm9uY2U9Ik 9BNk1HOXRFUUdtMmhoIixxb3A9ImF1dGgiLGFsZ29yaXRobT1tZDUtc2Vz cyxjaGFyc2V0PXV0Zi04" C: "Y2hhcnNldD11dGYtOCx1c2VybmFtZT0iY2hyaXMiLHJlYWxtPSJlbHdvb2 QuaW5ub3NvZnQuZXhhbXBsZS5jb20iLG5vbmNlPSJPQTZNRzl0RVFHbTJo aCIsbmM9MDAwMDAwMDEsY25vbmNlPSJPQTZNSFhoNlZxVHJSayIsZGlnZX N0LXVyaT0ic2lldmUvZWx3b29kLmlubm9zb2Z0LmV4YW1wbGUuY29tIixy ZXNwb25zZT1kMzg4ZGFkOTBkNGJiZDc2MGExNTIzMjFmMjE0M2FmNyxxb3 A9YXV0aA==" S: OK (SASL "cnNwYXV0aD1lYTQwZjYwMzM1YzQyN2I1NTI3Yjg0ZGJhYmNkZ mZmZA==")
A slightly different variant of the same authentication exchange is:
S: "IMPLEMENTATION" "Example1 ManageSieved v001"
S: "SASL" "DIGEST-MD5 GSSAPI"
S: "SIEVE" "fileinto vacation"
S: "VERSION" "1.0"
S: "STARTTLS"
S: OK
C: Authenticate "DIGEST-MD5"
S: {136}
S: cmVhbG09ImVsd29vZC5pbm5vc29mdC5leGFtcGxlLmNvbSIsbm9uY2U9Ik
9BNk1HOXRFUUdtMmhoIixxb3A9ImF1dGgiLGFsZ29yaXRobT1tZDUtc2Vz
cyxjaGFyc2V0PXV0Zi04
C: {300+}
C: Y2hhcnNldD11dGYtOCx1c2VybmFtZT0iY2hyaXMiLHJlYWxtPSJlbHdvb2
QuaW5ub3NvZnQuZXhhbXBsZS5jb20iLG5vbmNlPSJPQTZNRzl0RVFHbTJo
aCIsbmM9MDAwMDAwMDEsY25vbmNlPSJPQTZNSFhoNlZxVHJSayIsZGlnZX
N0LXVyaT0ic2lldmUvZWx3b29kLmlubm9zb2Z0LmV4YW1wbGUuY29tIixy
ZXNwb25zZT1kMzg4ZGFkOTBkNGJiZDc2MGExNTIzMjFmMjE0M2FmNyxxb3
A9YXV0aA==
S: {56}
S: cnNwYXV0aD1lYTQwZjYwMzM1YzQyN2I1NTI3Yjg0ZGJhYmNkZmZmZA==
C: ""
S: OK
Another example demonstrating use of SASL PLAIN mechanism under TLS
follows. This example also demonstrate use of SASL "initial
response" (the second parameter to the Authenticate command):
S: "IMPLEMENTATION" "Example1 ManageSieved v001"
S: "VERSION" "1.0"
S: "SASL" ""
S: "SIEVE" "fileinto vacation"
S: "STARTTLS"
S: OK
C: STARTTLS
S: OK
<TLS negotiation, further commands are under TLS layer>
S: "IMPLEMENTATION" "Example1 ManageSieved v001"
S: "VERSION" "1.0"
S: "SASL" "PLAIN"
S: "SIEVE" "fileinto vacation"
S: OK
C: Authenticate "PLAIN" "QJIrweAPyo6Q1T9xu"
S: NO
C: Authenticate "PLAIN" "QJIrweAPyo6Q1T9xz"
S: NO
C: Authenticate "PLAIN" "QJIrweAPyo6Q1T9xy"
S: BYE "Too many failed authentication attempts"
<Server closes connection>
The following example demonstrates use of SASL "initial response".
It also demonstrates that an empty response can be sent as a literal
and that negotiating a SASL security layer results in the server
re-issuing server capabilities:
C: AUTHENTICATE "GSSAPI" {1488+}
C: YIIE[...1480 octets here ...]dA==
S: {208}
S: YIGZBgkqhkiG9xIBAgICAG+BiTCBhqADAgEFoQMCAQ+iejB4oAMCARKic
[...114 octets here ...]
/yzpAy9p+Y0LanLskOTvMc0MnjgAa4YEr3eJ6
C: {0+}
C:
S: {44}
S: BQQF/wAMAAwAAAAAYRGFAo6W0vIHti8i1UXODgEAEAA=
C: {44+}
C: BQQE/wAMAAwAAAAAIsT1iv9UkZApw471iXt6cwEAAAE=
S: OK
<Further commands/responses are under SASL security layer>
S: "IMPLEMENTATION" "Example1 ManageSieved v001"
S: "VERSION" "1.0"
S: "SASL" "PLAIN DIGEST-MD5 GSSAPI"
S: "SIEVE" "fileinto vacation"
S: "LANGUAGE" "ru"
S: "MAXREDIRECTS" "3"
S: ok
2.1.1. Use of SASL PLAIN Mechanism over TLS
This section is normative for ManageSieve client implementations that
support SASL [PLAIN] over [TLS].
If a ManageSieve client is willing to use SASL PLAIN over TLS to
authenticate to the ManageSieve server, the client MUST verify the
server identity (see Section 2.2.1). If the server identity can't be
verified (e.g., the server has not provided any certificate, or if
the certificate verification fails), the client MUST NOT attempt to
authenticate using the SASL PLAIN mechanism.
2.2. STARTTLS Command
Support for STARTTLS command in servers is optional. Its
availability is advertised with "STARTTLS" capability as described in
Section 1.7.
The STARTTLS command requests commencement of a TLS [TLS]
negotiation. The negotiation begins immediately after the CRLF in
the OK response. After a client issues a STARTTLS command, it MUST
NOT issue further commands until a server response is seen and the TLS negotiation is complete. The STARTTLS command is only valid in non-authenticated state. The server remains in non-authenticated state, even if client credentials are supplied during the TLS negotiation. The SASL [SASL] EXTERNAL mechanism MAY be used to authenticate once TLS client credentials are successfully exchanged, but servers supporting the STARTTLS command are not required to support the EXTERNAL mechanism. After the TLS layer is established, the server MUST re-issue the capability results, followed by an OK response. This is necessary to protect against man-in-the-middle attacks that alter the capabilities list prior to STARTTLS. This capability result MUST NOT include the STARTTLS capability. The client MUST discard cached capability information and replace it with the new information. The server MAY advertise different capabilities after STARTTLS. Example: C: StartTls S: oK <TLS negotiation, further commands are under TLS layer> S: "IMPLEMENTATION" "Example1 ManageSieved v001" S: "SASL" "PLAIN DIGEST-MD5 GSSAPI" S: "SIEVE" "fileinto vacation" S: "VERSION" "1.0" S: "LANGUAGE" "fr" S: ok2.2.1. Server Identity Check
During the TLS negotiation, the ManageSieve client MUST check its understanding of the server hostname/IP address against the server's identity as presented in the server Certificate message, in order to prevent man-in-the-middle attacks. In this section, the client's understanding of the server's identity is called the "reference identity". Checking is performed according to the following rules: o If the reference identity is a hostname: 1. If a subjectAltName extension of the SRVName [X509-SRV], dNSName [X509] (in that order of preference) type is present in the server's certificate, then it SHOULD be used as the
source of the server's identity. Matching is performed as
described in Section 2.2.1.1, with the exception that no
wildcard matching is allowed for SRVName type. If the
certificate contains multiple names (e.g., more than one
dNSName field), then a match with any one of the fields is
considered acceptable.
2. The client MAY use other types of subjectAltName for
performing comparison.
3. The server's identity MAY also be verified by comparing the
reference identity to the Common Name (CN) [RFC4519] value in
the leaf Relative Distinguished Name (RDN) of the subjectName
field of the server's certificate. This comparison is
performed using the rules for comparison of DNS names in
Section 2.2.1.1, below. Although the use of the Common Name
value is existing practice, it is deprecated, and
Certification Authorities are encouraged to provide
subjectAltName values instead. Note that the TLS
implementation may represent DNs in certificates according to
X.500 or other conventions. For example, some X.500
implementations order the RDNs in a DN using a left-to-right
(most significant to least significant) convention instead of
LDAP's right-to-left convention.
o When the reference identity is an IP address, the iPAddress
subjectAltName SHOULD be used by the client for comparison. The
comparison is performed as described in Section 2.2.1.2.
If the server identity check fails, user-oriented clients SHOULD
either notify the user (clients MAY give the user the opportunity to
continue with the ManageSieve session in this case) or close the
transport connection and indicate that the server's identity is
suspect. Automated clients SHOULD return or log an error indicating
that the server's identity is suspect and/or SHOULD close the
transport connection. Automated clients MAY provide a configuration
setting that disables this check, but MUST provide a setting that
enables it.
Beyond the server identity check described in this section, clients
should be prepared to do further checking to ensure that the server
is authorized to provide the service it is requested to provide. The
client may need to make use of local policy information in making
this determination.
2.2.1.1. Comparison of DNS Names
If the reference identity is an internationalized domain name, conforming implementations MUST convert it to the ASCII Compatible Encoding (ACE) format as specified in Section 4 of RFC 3490 [RFC3490] before comparison with subjectAltName values of type dNSName. Specifically, conforming implementations MUST perform the conversion operation specified in Section 4 of [RFC3490] as follows: o in step 1, the domain name SHALL be considered a "stored string"; o in step 3, set the flag called "UseSTD3ASCIIRules"; o in step 4, process each label with the "ToASCII" operation; and o in step 5, change all label separators to U+002E (full stop). After performing the "to-ASCII" conversion, the DNS labels and names MUST be compared for equality according to the rules specified in Section 3 of [RFC3490]; i.e., once all label separators are replaced with U+002E (dot) they are compared in the case-insensitive manner. The '*' (ASCII 42) wildcard character is allowed in subjectAltName values of type dNSName, and then only as the left-most (least significant) DNS label in that value. This wildcard matches any left-most DNS label in the server name. That is, the subject *.example.com matches the server names a.example.com and b.example.com, but does not match example.com or a.b.example.com.2.2.1.2. Comparison of IP Addresses
When the reference identity is an IP address, the identity MUST be converted to the "network byte order" octet string representation [RFC791][RFC2460]. For IP Version 4, as specified in RFC 791, the octet string will contain exactly four octets. For IP Version 6, as specified in RFC 2460, the octet string will contain exactly sixteen octets. This octet string is then compared against subjectAltName values of type iPAddress. A match occurs if the reference identity octet string and value octet strings are identical.2.2.1.3. Comparison of Other subjectName Types
Client implementations MAY support matching against subjectAltName values of other types as described in other documents.
2.3. LOGOUT Command
The client sends the LOGOUT command when it is finished with a connection and wishes to terminate it. The server MUST reply with an OK response. The server MUST ignore commands issued by the client after the LOGOUT command. The client SHOULD wait for the OK response before closing the connection. This avoids the TCP connection going into the TIME_WAIT state on the server. In order to avoid going into the TIME_WAIT TCP state, the server MAY wait for a short while for the client to close the TCP connection first. Whether or not the server waits for the client to close the connection, it MUST then close the connection itself. Example: C: Logout S: Ok <connection is terminated>2.4. CAPABILITY Command
The CAPABILITY command requests the server capabilities as described earlier in this document. It has no parameters. Example: C: CAPABILITY S: "IMPLEMENTATION" "Example1 ManageSieved v001" S: "VERSION" "1.0" S: "SASL" "PLAIN SCRAM-SHA-1 GSSAPI" S: "SIEVE" "fileinto vacation" S: "STARTTLS" S: OK2.5. HAVESPACE Command
Arguments: String - name Number - script size The HAVESPACE command is used to query the server for available space. Clients specify the name they wish to save the script as and its size in octets. Both parameters can be used by the server to see if the script with the specified name and size is within a user's quota(s). For example, the server MAY use the script name to check if a script would be replaced or a new one would be created. Servers respond with a NO if storing a script with that name and size would
fail or OK otherwise. Clients SHOULD issue this command before attempting to place a script on the server. Note that the OK response from the HAVESPACE command does not constitute a guarantee of success as server disk space conditions could change between the client issuing the HAVESPACE and the client issuing the PUTSCRIPT commands. A QUOTA response code (see Section 1.3) remains a possible (albeit unlikely) response to a subsequent PUTSCRIPT with the same name and size. Example: C: HAVESPACE "myscript" 999999 S: NO (QUOTA/MAXSIZE) "Quota exceeded" C: HAVESPACE "foobar" 435 S: OK2.6. PUTSCRIPT Command
Arguments: String - Script name String - Script content The PUTSCRIPT command is used by the client to submit a Sieve script to the server. If the script already exists, upon success the old script will be overwritten. The old script MUST NOT be overwritten if PUTSCRIPT fails in any way. A script of zero length SHOULD be disallowed. This command places the script on the server. It does not affect whether the script is processed on incoming mail, unless it replaces the script that is already active. The SETACTIVE command is used to mark a script as active. When submitting large scripts, clients SHOULD use the HAVESPACE command beforehand to query if the server is willing to accept a script of that size. The server MUST check the submitted script for validity, which includes checking that the script complies with the Sieve grammar [SIEVE] and that all Sieve extensions mentioned in the script's "require" statement(s) are supported by the Sieve interpreter. (Note that if the Sieve interpreter supports the Sieve "ihave" extension [I-HAVE], any unrecognized/unsupported extension mentioned in the "ihave" test MUST NOT cause the validation failure.) Other checks such as validating the supplied command arguments for each command MAY be performed. Essentially, the performed validation SHOULD be
the same as performed when compiling the script for execution. Implementations that use a binary representation to store compiled scripts can extend the validation to a full compilation, in order to avoid validating uploaded scripts multiple times. If the script fails the validation, the server MUST reply with a NO response. Any script that fails the validity test MUST NOT be stored on the server. The message given with a NO response MUST be human readable and SHOULD contain a specific error message giving the line number of the first error. Implementors should strive to produce helpful error messages similar to those given by programming language compilers. Client implementations should note that this may be a multiline literal string with more than one error message separated by CRLFs. The human-readable message is in the language returned in the latest LANGUAGE capability (or in "i-default"; see Section 1.7), encoded in UTF-8 [UTF-8]. An OK response MAY contain the WARNINGS response code. In such a case the human-readable message that follows the OK response SHOULD contain a specific warning message (or messages) giving the line number(s) in the script that might contain errors not intended by the script writer. The human-readable message is in the language returned in the latest LANGUAGE capability (or in "i-default"; see Section 1.7), encoded in UTF-8 [UTF-8]. A client seeing such a response code SHOULD present the message to the user.
Examples:
C: Putscript "foo" {31+}
C: #comment
C: InvalidSieveCommand
C:
S: NO "line 2: Syntax error"
C: Putscript "mysievescript" {110+}
C: require ["fileinto"];
C:
C: if envelope :contains "to" "tmartin+sent" {
C: fileinto "INBOX.sent";
C: }
S: OK
C: Putscript "myforwards" {190+}
C: redirect "111@example.net";
C:
C: if size :under 10k {
C: redirect "mobile@cell.example.com";
C: }
C:
C: if envelope :contains "to" "tmartin+lists" {
C: redirect "lists@groups.example.com";
C: }
S: OK (WARNINGS) "line 8: server redirect action
limit is 2, this redirect might be ignored"
2.7. LISTSCRIPTS Command
This command lists the scripts the user has on the server. Upon
success, a list of CRLF-separated script names (each represented as a
quoted or literal string) is returned followed by an OK response. If
there exists an active script, the atom ACTIVE is appended to the
corresponding script name. The atom ACTIVE MUST NOT appear on more
than one response line.
Example:
C: Listscripts
S: "summer_script"
S: "vacation_script"
S: {13}
S: clever"script
S: "main_script" ACTIVE
S: OK
C: listscripts
S: "summer_script"
S: "main_script" active
S: OK
2.8. SETACTIVE Command
Arguments: String - script name
This command sets a script active. If the script name is the empty
string (i.e., ""), then any active script is disabled. Disabling an
active script when there is no script active is not an error and MUST
result in an OK reply.
If the script does not exist on the server, then the server MUST
reply with a NO response. Such a reply SHOULD contain the
NONEXISTENT response code.
Examples:
C: Setactive "vacationscript"
S: Ok
C: Setactive ""
S: Ok
C: Setactive "baz"
S: No (NONEXISTENT) "There is no script by that name"
C: Setactive "baz"
S: No (NONEXISTENT) {31}
S: There is no script by that name
2.9. GETSCRIPT Command
Arguments: String - script name This command gets the contents of the specified script. If the script does not exist, the server MUST reply with a NO response. Such a reply SHOULD contain the NONEXISTENT response code. Upon success, a string with the contents of the script is returned followed by an OK response. Example: C: Getscript "myscript" S: {54} S: #this is my wonderful script S: reject "I reject all"; S: S: OK2.10. DELETESCRIPT Command
Arguments: String - script name This command is used to delete a user's Sieve script. Servers MUST reply with a NO response if the script does not exist. Such responses SHOULD include the NONEXISTENT response code. The server MUST NOT allow the client to delete an active script, so the server MUST reply with a NO response if attempted. Such a response SHOULD contain the ACTIVE response code. If a client wishes to delete an active script, it should use the SETACTIVE command to disable the script first. Example: C: Deletescript "foo" S: Ok C: Deletescript "baz" S: No (ACTIVE) "You may not delete an active script"
2.11. RENAMESCRIPT Command
Arguments: String - Old Script name String - New Script name This command is used to rename a user's Sieve script. Servers MUST reply with a NO response if the old script does not exist (in which case the NONEXISTENT response code SHOULD be included), or a script with the new name already exists (in which case the ALREADYEXISTS response code SHOULD be included). Renaming the active script is allowed; the renamed script remains active. Example: C: Renamescript "foo" "bar" S: Ok C: Renamescript "baz" "bar" S: No "bar already exists" If the server doesn't support the RENAMESCRIPT command, the client can emulate it by performing the following steps: 1. List available scripts with LISTSCRIPTS. If the script with the new script name exists, then the client should ask the user whether to abort the operation, to replace the script (by issuing the DELETESCRIPT <newname> after that), or to choose a different name. 2. Download the old script with GETSCRIPT <oldname>. 3. Upload the old script with the new name: PUTSCRIPT <newname>. 4. If the old script was active (as reported by LISTSCRIPTS in step 1), then make the new script active: SETACTIVE <newname>. 5. Delete the old script: DELETESCRIPT <oldname>. Note that these steps don't describe how to handle various other error conditions (for example, NO response containing QUOTA response code in step 3). Error handling is left as an exercise for the reader.
2.12. CHECKSCRIPT Command
Arguments: String - Script content The CHECKSCRIPT command is used by the client to verify Sieve script validity without storing the script on the server. The server MUST check the submitted script for syntactic validity, which includes checking that all Sieve extensions mentioned in Sieve script "require" statement(s) are supported by the Sieve interpreter. (Note that if the Sieve interpreter supports the Sieve "ihave" extension [I-HAVE], any unrecognized/unsupported extension mentioned in the "ihave" test MUST NOT cause the syntactic validation failure.) If the script fails this test, the server MUST reply with a NO response. The message given with a NO response MUST be human readable and SHOULD contain a specific error message giving the line number of the first error. Implementors should strive to produce helpful error messages similar to those given by programming language compilers. Client implementations should note that this may be a multiline literal string with more than one error message separated by CRLFs. The human-readable message is in the language returned in the latest LANGUAGE capability (or in "i-default"; see Section 1.7), encoded in UTF-8 [UTF-8]. Examples: C: CheckScript {31+} C: #comment C: InvalidSieveCommand C: S: NO "line 2: Syntax error" A ManageSieve server supporting this command MUST NOT check if the script will put the current user over its quota limit. An OK response MAY contain the WARNINGS response code. In such a case, the human-readable message that follows the OK response SHOULD contain a specific warning message (or messages) giving the line number(s) in the script that might contain errors not intended by the script writer. The human-readable message is in the language returned in the latest LANGUAGE capability (or in "i-default"; see Section 1.7), encoded in UTF-8 [UTF-8]. A client seeing such a response code SHOULD present the message to the user.
2.13. NOOP Command
Arguments: String - tag to echo back (optional) The NOOP command does nothing, beyond returning a response to the client. It may be used by clients for protocol re-synchronization or to reset any inactivity auto-logout timer on the server. The response to the NOOP command is always OK, followed by the TAG response code together with the supplied string. If no string was supplied in the NOOP command, the TAG response code MUST NOT be included. Examples: C: NOOP S: OK "NOOP completed" C: NOOP "STARTTLS-SYNC-42" S: OK (TAG {16} S: STARTTLS-SYNC-42) "Done"2.14. Recommended Extensions
The UNAUTHENTICATE extension (advertised as the "UNAUTHENTICATE" capability with no parameters) defines a new UNAUTHENTICATE command, which allows a client to return the server to non-authenticated state. Support for this extension is RECOMMENDED.2.14.1. UNAUTHENTICATE Command
The UNAUTHENTICATE command returns the server to the non-authenticated state. It doesn't affect any previously established TLS [TLS] or SASL (Section 2.1) security layer. The UNAUTHENTICATE command is only valid in authenticated state. If issued in a wrong state, the server MUST reject it with a NO response. The UNAUTHENTICATE command has no parameters. When issued in the authenticated state, the UNAUTHENTICATE command MUST NOT fail (i.e., it must never return anything other than OK or BYE).
3. Sieve URL Scheme
URI scheme name: sieve Status: permanent URI scheme syntax: Described using ABNF [ABNF]. Some ABNF productions not defined below are from [URI-GEN]. sieveurl = sieveurl-server / sieveurl-list-scripts / sieveurl-script sieveurl-server = "sieve://" authority sieveurl-list-scripts = "sieve://" authority ["/"] sieveurl-script = "sieve://" authority "/" [owner "/"] scriptname authority = <defined in [URI-GEN]> owner = *ochar ;; %-encoded version of [SASL] authorization ;; identity (script owner) or "userid". ;; ;; Empty owner is used to reference ;; global scripts. ;; ;; Note that ASCII characters such as " ", ";", ;; "&", "=", "/" and "?" must be %-encoded ;; as per rule specified in [URI-GEN]. scriptname = 1*ochar ;; %-encoded version of UTF-8 representation ;; of the script name. ;; Note that ASCII characters such as " ", ";", ;; "&", "=", "/" and "?" must be %-encoded ;; as per rule specified in [URI-GEN]. ochar = unreserved / pct-encoded / sub-delims-sh / ":" / "@" ;; Same as [URI-GEN] 'pchar', ;; but without ";", "&" and "=". unreserved = <defined in [URI-GEN]> pct-encoded = <defined in [URI-GEN]>
sub-delims-sh = "!" / "$" / "'" / "(" / ")" /
"*" / "+" / ","
;; Same as [URI-GEN] sub-delims,
;; but without ";", "&" and "=".
URI scheme semantics:
A Sieve URL identifies a Sieve server or a Sieve script on a Sieve
server. The latter form is associated with the application/sieve
MIME type defined in [SIEVE]. There is no MIME type associated
with the former form of Sieve URI.
The server form is used in the REFERRAL response code (see Section
1.3) in order to designate another server where the client should
perform its operations.
The script form allows to retrieve (GETSCRIPT), update
(PUTSCRIPT), delete (DELETESCRIPT), or activate (SETACTIVE) the
named script; however, the most typical action would be to
retrieve the script. If the script name is empty (omitted), the
URI requests that the client lists available scripts using the
LISTSCRIPTS command.
Encoding considerations:
The script name and/or the owner, if present, is in UTF-8. Non--
US-ASCII UTF-8 octets MUST be percent-encoded as described in
[URI-GEN]. US-ASCII characters such as " " (space), ";", "&",
"=", "/" and "?" MUST be %-encoded as described in [URI-GEN].
Note that "&" and "?" are in this list in order to allow for
future extensions.
Note that the empty owner (e.g., sieve://example.com//script) is
different from the missing owner (e.g.,
sieve://example.com/script) and is reserved for referencing global
scripts.
The user name (in the "authority" part), if present, is in UTF-8.
Non-US-ASCII UTF-8 octets MUST be percent-encoded as described in
[URI-GEN].
Applications/protocols that use this URI scheme name:
ManageSieve [RFC5804] clients and servers. Clients that can store
user preferences in protocols such as [LDAP] or [ACAP].
Interoperability considerations: None.
Security considerations: The <scriptname> part of a ManageSieve URL might potentially disclose some confidential information about the author of the script or, depending on a ManageSieve implementation, about configuration of the mail system. The latter might be used to prepare for a more complex attack on the mail system. Clients resolving ManageSieve URLs that wish to achieve data confidentiality and/or integrity SHOULD use the STARTTLS command (if supported by the server) before starting authentication, or use a SASL mechanism, such as GSSAPI, that provides a confidentiality security layer. Contact: Alexey Melnikov <alexey.melnikov@isode.com> Author/Change controller: IESG. References: This document and RFC 5228 [SIEVE].4. Formal Syntax
The following syntax specification uses the Augmented Backus-Naur Form (BNF) notation as specified in [ABNF]. This uses the ABNF core rules as specified in Appendix A of the ABNF specification [ABNF]. "UTF8-2", "UTF8-3", and "UTF8-4" non-terminal are defined in [UTF-8]. Except as noted otherwise, all alphabetic characters are case- insensitive. The use of upper- or lowercase characters to define token strings is for editorial clarity only. Implementations MUST accept these strings in a case-insensitive fashion. SAFE-CHAR = %x01-09 / %x0B-0C / %x0E-21 / %x23-5B / %x5D-7F ;; any TEXT-CHAR except QUOTED-SPECIALS QUOTED-CHAR = SAFE-UTF8-CHAR / "\" QUOTED-SPECIALS QUOTED-SPECIALS = DQUOTE / "\" SAFE-UTF8-CHAR = SAFE-CHAR / UTF8-2 / UTF8-3 / UTF8-4 ;; <UTF8-2>, <UTF8-3>, and <UTF8-4> ;; are defined in [UTF-8]. ATOM-CHAR = "!" / %x23-27 / %x2A-5B / %x5D-7A / %x7C-7E ;; Any CHAR except ATOM-SPECIALS ATOM-SPECIALS = "(" / ")" / "{" / SP / CTL / QUOTED-SPECIALS
NZDIGIT = %x31-39
;; 1-9
atom = 1*1024ATOM-CHAR
iana-token = atom
;; MUST be registered with IANA
auth-type = DQUOTE auth-type-name DQUOTE
auth-type-name = iana-token
;; as defined in SASL [SASL]
command = (command-any / command-auth /
command-nonauth) CRLF
;; Modal based on state
command-any = command-capability / command-logout /
command-noop
;; Valid in all states
command-auth = command-getscript / command-setactive /
command-listscripts / command-deletescript /
command-putscript / command-checkscript /
command-havespace /
command-renamescript /
command-unauthenticate
;; Valid only in Authenticated state
command-nonauth = command-authenticate / command-starttls
;; Valid only when in Non-Authenticated
;; state
command-authenticate = "AUTHENTICATE" SP auth-type [SP string]
*(CRLF string)
command-capability = "CAPABILITY"
command-deletescript = "DELETESCRIPT" SP sieve-name
command-getscript = "GETSCRIPT" SP sieve-name
command-havespace = "HAVESPACE" SP sieve-name SP number
command-listscripts = "LISTSCRIPTS"
command-noop = "NOOP" [SP string]
command-logout = "LOGOUT"
command-putscript = "PUTSCRIPT" SP sieve-name SP sieve-script
command-checkscript = "CHECKSCRIPT" SP sieve-script
sieve-script = string
command-renamescript = "RENAMESCRIPT" SP old-sieve-name SP
new-sieve-name
old-sieve-name = sieve-name
new-sieve-name = sieve-name
command-setactive = "SETACTIVE" SP active-sieve-name
command-starttls = "STARTTLS"
command-unauthenticate= "UNAUTHENTICATE"
extend-token = atom
;; MUST be defined by a Standards Track or
;; IESG-approved experimental protocol
;; extension
extension-data = extension-item *(SP extension-item)
extension-item = extend-token / string / number /
"(" [extension-data] ")"
literal-c2s = "{" number "+}" CRLF *OCTET
;; The number represents the number of
;; octets.
;; This type of literal can only be sent
;; from the client to the server.
literal-s2c = "{" number "}" CRLF *OCTET
;; Almost identical to literal-c2s,
;; but with no '+' character.
;; The number represents the number of
;; octets.
;; This type of literal can only be sent
;; from the server to the client.
number = (NZDIGIT *DIGIT) / "0"
;; A 32-bit unsigned number
;; with no extra leading zeros.
;; (0 <= n < 4,294,967,296)
number-str = string
;; <number> encoded as a <string>.
quoted = DQUOTE *1024QUOTED-CHAR DQUOTE
;; limited to 1024 octets between the <">s
resp-code = "AUTH-TOO-WEAK" / "ENCRYPT-NEEDED" / "QUOTA"
["/" ("MAXSCRIPTS" / "MAXSIZE")] /
resp-code-sasl /
resp-code-referral /
"TRANSITION-NEEDED" / "TRYLATER" /
"ACTIVE" / "NONEXISTENT" /
"ALREADYEXISTS" / "WARNINGS" /
"TAG" SP string /
resp-code-ext
resp-code-referral = "REFERRAL" SP sieveurl
resp-code-sasl = "SASL" SP string
resp-code-name = iana-token
;; The response code name is hierarchical,
;; separated by '/'.
;; The response code name MUST NOT start
;; with '/'.
resp-code-ext = resp-code-name [SP extension-data]
;; unknown response codes MUST be tolerated
;; by the client.
response = response-authenticate /
response-logout /
response-getscript /
response-setactive /
response-listscripts /
response-deletescript /
response-putscript /
response-checkscript /
response-capability /
response-havespace /
response-starttls /
response-renamescript /
response-noop /
response-unauthenticate
response-authenticate = *(string CRLF)
((response-ok [response-capability]) /
response-nobye)
;; <response-capability> is REQUIRED if a
;; SASL security layer was negotiated and
;; MUST be omitted otherwise.
response-capability = *(single-capability) response-oknobye
single-capability = capability-name [SP string] CRLF
capability-name = string
;; Note that literal-s2c is allowed.
initial-capabilities = DQUOTE "IMPLEMENTATION" DQUOTE SP string /
DQUOTE "SASL" DQUOTE SP sasl-mechs /
DQUOTE "SIEVE" DQUOTE SP sieve-extensions /
DQUOTE "MAXREDIRECTS" DQUOTE SP number-str /
DQUOTE "NOTIFY" DQUOTE SP notify-mechs /
DQUOTE "STARTTLS" DQUOTE /
DQUOTE "LANGUAGE" DQUOTE SP language /
DQUOTE "VERSION" DQUOTE SP version /
DQUOTE "OWNER" DQUOTE SP string
;; Each capability conforms to
;; the syntax for single-capability.
;; Also, note that the capability name
;; can be returned as either literal-s2c
;; or quoted, even though only "quoted"
;; string is shown above.
version = ( DQUOTE "1.0" DQUOTE ) / version-ext
version-ext = DQUOTE ver-major "." ver-minor DQUOTE
; Future versions specified in updates
; to this document. An increment to
; the ver-major means a backward-incompatible
; change to the protocol, e.g., "3.5" (ver-major "3")
; is not backward-compatible with any "2.X" version.
; Any version "Z.W" MUST be backward compatible
; with any version "Z.Q", where Q < W.
; For example, version "2.4" is backward compatible
; with version "2.0", "2.1", "2.2", and "2.3".
ver-major = number
ver-minor = number
sasl-mechs = string
; Space-separated list of SASL mechanisms,
; each SASL mechanism name complies with rules
; specified in [SASL].
; Can be empty.
sieve-extensions = string
; Space-separated list of supported SIEVE extensions.
; Can be empty.
language = string
; Contains <Language-Tag> from [RFC5646].
notify-mechs = string
; Space-separated list of URI schema parts
; for supported notification [NOTIFY] methods.
; MUST NOT be empty.
response-deletescript = response-oknobye
response-getscript = (sieve-script CRLF response-ok) /
response-nobye
response-havespace = response-oknobye
response-listscripts = *(sieve-name [SP "ACTIVE"] CRLF)
response-oknobye
;; ACTIVE may only occur with one sieve-name
response-logout = response-oknobye
response-unauthenticate= response-oknobye
;; "NO" response can only be returned when
;; the command is issued in a wrong state
;; or has a wrong number of parameters
response-ok = "OK" [SP "(" resp-code ")"]
[SP string] CRLF
;; The string contains human-readable text
;; encoded as UTF-8.
response-nobye = ("NO" / "BYE") [SP "(" resp-code ")"]
[SP string] CRLF
;; The string contains human-readable text
;; encoded as UTF-8.
response-oknobye = response-ok / response-nobye
response-noop = response-ok
response-putscript = response-oknobye
response-checkscript = response-oknobye
response-renamescript = response-oknobye
response-setactive = response-oknobye
response-starttls = (response-ok response-capability) /
response-nobye
sieve-name = string
;; See Section 1.6 for the full list of
;; prohibited characters.
;; Empty string is not allowed.
active-sieve-name = string
;; See Section 1.6 for the full list of
;; prohibited characters.
;; This is similar to <sieve-name>, but
;; empty string is allowed and has a special
;; meaning.
string = quoted / literal-c2s / literal-s2c
;; literal-c2s is only allowed when sent
;; from the client to the server.
;; literal-s2c is only allowed when sent
;; from the server to the client.
;; quoted is allowed in either direction.
5. Security Considerations
The AUTHENTICATE command uses SASL [SASL] to provide authentication
and authorization services. Integrity and privacy services can be
provided by [SASL] and/or [TLS]. When a SASL mechanism is used, the
security considerations for that mechanism apply.
This protocol's transactions are susceptible to passive observers or
man-in-the-middle attacks that alter the data, unless the optional
encryption and integrity services of the SASL (via the AUTHENTICATE
command) and/or [TLS] (via the STARTTLS command) are enabled, or an
external security mechanism is used for protection. It may be useful
to allow configuration of both clients and servers to refuse to
transfer sensitive information in the absence of strong encryption.
If an implementation supports SASL mechanisms that are vulnerable to passive eavesdropping attacks (such as [PLAIN]), then the implementation MUST support at least one configuration where these SASL mechanisms are not advertised or used without the presence of an external security layer such as [TLS]. Some response codes returned on failed AUTHENTICATE command may disclose whether or not the username is valid (e.g., TRANSITION- NEEDED), so server implementations SHOULD provide the ability to disable these features (or make them not conditional on a per-user basis) for sites concerned about such disclosure. In the case of ENCRYPT-NEEDED, if it is applied to all identities then no extra information is disclosed, but if it is applied on a per-user basis it can disclose information. A compromised or malicious server can use the TRANSITION-NEEDED response code to force the client that is configured to use a mechanism that does not disclose the user's password to the server (e.g., Kerberos), to send the bare password to the server. Clients SHOULD have the ability to disable the password transition feature, or disclose that risk to the user and offer the user an option of how to proceed.6. IANA Considerations
IANA has reserved TCP port number 4190 for use with the ManageSieve protocol described in this document. IANA has registered the "sieve" URI scheme defined in Section 3 of this document. IANA has registered "sieve" in the "GSSAPI/Kerberos/SASL Service Names" registry. IANA has created a new registry for ManageSieve capabilities. The registration template for ManageSieve capabilities is specified in Section 6.1. ManageSieve protocol capabilities MUST be specified in a Standards-Track or IESG-approved Experimental RFC. IANA has created a new registry for ManageSieve response codes. The registration template for ManageSieve response codes is specified in Section 6.3. ManageSieve protocol response codes MUST be specified in a Standards-Track or IESG-approved Experimental RFC.
6.1. ManageSieve Capability Registration Template
To: iana@iana.org Subject: ManageSieve Capability Registration Please register the following ManageSieve capability: Capability name: Description: Relevant publications: Person & email address to contact for further information: Author/Change controller:6.2. Registration of Initial ManageSieve Capabilities
To: iana@iana.org Subject: ManageSieve Capability Registration Please register the following ManageSieve capabilities: Capability name: IMPLEMENTATION Description: Its value contains the name of the server implementation and its version. Relevant publications: this RFC, Section 1.7. Person & email address to contact for further information: Alexey Melnikov <alexey.melnikov@isode.com> Author/Change controller: IESG. Capability name: SASL Description: Its value contains a space-separated list of SASL mechanisms supported by the server. Relevant publications: this RFC, Sections 1.7 and 2.1. Person & email address to contact for further information: Alexey Melnikov <alexey.melnikov@isode.com> Author/Change controller: IESG. Capability name: SIEVE Description: Its value contains a space-separated list of supported SIEVE extensions. Relevant publications: this RFC, Section 1.7. Also [SIEVE]. Person & email address to contact for further information: Alexey Melnikov <alexey.melnikov@isode.com> Author/Change controller: IESG.
Capability name: STARTTLS
Description: This capability is returned if the server supports TLS
(STARTTLS command).
Relevant publications: this RFC, Sections 1.7 and 2.2.
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Capability name: NOTIFY
Description: This capability is returned if the server supports the
'enotify' [NOTIFY] Sieve extension.
Relevant publications: this RFC, Section 1.7.
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Capability name: MAXREDIRECTS
Description: This capability returns the limit on the number of
Sieve "redirect" actions a script can perform during a
single evaluation. The value is a non-negative number
represented as a ManageSieve string.
Relevant publications: this RFC, Section 1.7.
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Capability name: LANGUAGE
Description: The language (<Language-Tag> from [RFC5646]) currently
used for human-readable error messages.
Relevant publications: this RFC, Section 1.7.
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Capability name: OWNER
Description: Its value contains the UTF-8-encoded name of the
currently logged-in user ("authorization identity"
according to RFC 4422).
Relevant publications: this RFC, Section 1.7.
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Capability name: VERSION
Description: This capability is returned if the server is compliant
with RFC 5804; i.e., that it supports RENAMESCRIPT,
CHECKSCRIPT, and NOOP commands.
Relevant publications: this RFC, Sections 2.11, 2.12, and 2.13.
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
6.3. ManageSieve Response Code Registration Template
To: iana@iana.org
Subject: ManageSieve Response Code Registration
Please register the following ManageSieve response code:
Response Code:
Arguments (use ABNF to specify syntax, or the word NONE if none
can be specified):
Purpose:
Published Specification(s):
Person & email address to contact for further information:
Author/Change controller:
6.4. Registration of Initial ManageSieve Response Codes
To: iana@iana.org
Subject: ManageSieve Response Code Registration
Please register the following ManageSieve response codes:
Response Code: AUTH-TOO-WEAK
Arguments (use ABNF to specify syntax, or the word NONE if none can
be specified): NONE
Purpose: This response code is returned in the NO response from
an AUTHENTICATE command. It indicates that site
security policy forbids the use of the requested
mechanism for the specified authentication identity.
Published Specification(s): [RFC5804]
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Response Code: ENCRYPT-NEEDED
Arguments (use ABNF to specify syntax, or the word NONE if none can
be specified): NONE
Purpose: This response code is returned in the NO response from
an AUTHENTICATE command. It indicates that site
security policy requires the use of a strong
encryption mechanism for the specified authentication
identity and mechanism.
Published Specification(s): [RFC5804]
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Response Code: QUOTA
Arguments (use ABNF to specify syntax, or the word NONE if none can
be specified): NONE
Purpose: If this response code is returned in the NO/BYE
response, it means that the command would have placed
the user above the site-defined quota constraints. If
this response code is returned in the OK response, it
can mean that the user is near its quota or that the
user exceeded its quota, but the server supports soft
quotas.
Published Specification(s): [RFC5804]
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Response Code: QUOTA/MAXSCRIPTS
Arguments (use ABNF to specify syntax, or the word NONE if none can
be specified): NONE
Purpose: If this response code is returned in the NO/BYE
response, it means that the command would have placed
the user above the site-defined limit on the number of
Sieve scripts. If this response code is returned in
the OK response, it can mean that the user is near its
quota or that the user exceeded its quota, but the
server supports soft quotas. This response code is a
more specific version of the QUOTA response code.
Published Specification(s): [RFC5804]
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Response Code: QUOTA/MAXSIZE
Arguments (use ABNF to specify syntax, or the word NONE if none can
be specified): NONE
Purpose: If this response code is returned in the NO/BYE
response, it means that the command would have placed
the user above the site-defined maximum script size.
If this response code is returned in the OK response,
it can mean that the user is near its quota or that
the user exceeded its quota, but the server supports
soft quotas. This response code is a more specific
version of the QUOTA response code.
Published Specification(s): [RFC5804]
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Response Code: REFERRAL
Arguments (use ABNF to specify syntax, or the word NONE if none can
be specified): <sieveurl>
Purpose: This response code may be returned with a BYE result
from any command, and includes a mandatory parameter
that indicates what server to access to manage this
user's Sieve scripts. The server will be specified by
a Sieve URL (see Section 3). The scriptname portion
of the URL MUST NOT be specified. The client should
authenticate to the specified server and use it for
all further commands in the current session.
Published Specification(s): [RFC5804]
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Response Code: SASL
Arguments (use ABNF to specify syntax, or the word NONE if none can
be specified): <string>
Purpose: This response code can occur in the OK response to a
successful AUTHENTICATE command and includes the
optional final server response data from the server as
specified by [SASL].
Published Specification(s): [RFC5804]
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Response Code: TRANSITION-NEEDED
Arguments (use ABNF to specify syntax, or the word NONE if none can
be specified): NONE
Purpose: This response code occurs in a NO response of an
AUTHENTICATE command. It indicates that the user name
is valid, but the entry in the authentication database
needs to be updated in order to permit authentication
with the specified mechanism. This is typically done
by establishing a secure channel using TLS, followed
by authenticating once using the [PLAIN]
authentication mechanism. The selected mechanism
SHOULD then work for authentications in subsequent
sessions.
Published Specification(s): [RFC5804]
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Response Code: TRYLATER
Arguments (use ABNF to specify syntax, or the word NONE if none can
be specified): NONE
Purpose: A command failed due to a temporary server failure.
The client MAY continue using local information and
try the command later. This response code only make
sense when returned in a NO/BYE response.
Published Specification(s): [RFC5804]
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Response Code: ACTIVE
Arguments (use ABNF to specify syntax, or the word NONE if none can
be specified): NONE
Purpose: A command failed because it is not allowed on the
active script, for example, DELETESCRIPT on the active
script. This response code only makes sense when
returned in a NO/BYE response.
Published Specification(s): [RFC5804]
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Response Code: NONEXISTENT
Arguments (use ABNF to specify syntax, or the word NONE if none can
be specified): NONE
Purpose: A command failed because the referenced script name
doesn't exist. This response code only makes sense
when returned in a NO/BYE response.
Published Specification(s): [RFC5804]
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Response Code: ALREADYEXISTS
Arguments (use ABNF to specify syntax, or the word NONE if none can
be specified): NONE
Purpose: A command failed because the referenced script name
already exists. This response code only makes sense
when returned in a NO/BYE response.
Published Specification(s): [RFC5804]
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Response Code: WARNINGS
Arguments (use ABNF to specify syntax, or the word NONE if none can
be specified): NONE
Purpose: This response code MAY be returned by the server in
the OK response (but it might be returned with the NO/
BYE response as well) and signals the client that even
though the script is syntactically valid, it might
contain errors not intended by the script writer.
Published Specification(s): [RFC5804]
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
Response Code: TAG
Arguments (use ABNF to specify syntax, or the word NONE if none can
be specified): string
Purpose: This response code name is followed by a string
specified in the command that caused this response.
It is typically used for client state synchronization.
Published Specification(s): [RFC5804]
Person & email address to contact for further information:
Alexey Melnikov <alexey.melnikov@isode.com>
Author/Change controller: IESG.
7. Internationalization Considerations
The LANGUAGE capability (see Section 1.7) allows a client to discover the current language used in all human-readable responses that might be returned at the end of any OK/NO/BYE response. Human-readable text in OK responses typically doesn't need to be shown to the user, unless it is returned in response to a PUTSCRIPT or CHECKSCRIPT command that also contains the WARNINGS response code (Section 1.3). Human-readable text from NO/BYE responses is intended be shown to the user, unless the client can automatically handle failure of the command that caused such a response. Clients SHOULD use response codes (Section 1.3) for automatic error handling. Response codes MAY also be used by the client to present error messages in a language understood by the user, for example, if the LANGUAGE capability doesn't return a language understood by the user. Note that the human-readable text from OK (WARNINGS) or NO/BYE responses for PUTSCRIPT/CHECKSCRIPT commands is intended for advanced users that understand Sieve language. Such advanced users are often sophisticated enough to be able to handle whatever language the server is using, even if it is not their preferred language, and will want to see error/warning text no matter what language the server puts it in. A client that generates Sieve script automatically, for example, if the script is generated without user intervention or from a UI that presents an abstract list of conditions and corresponding actions, SHOULD NOT present warning/error messages to the user, because the user might not even be aware that the client is using Sieve underneath. However, if the client has a debugging mode, such warnings/errors SHOULD be available in the debugging mode. Note that this document doesn't provide a way to modify the currently used language. It is expected that a future extension will address that.8. Acknowledgements
Thanks to Simon Josefsson, Larry Greenfield, Allen Johnson, Chris Newman, Lyndon Nerenberg, Tim Showalter, Sarah Robeson, Walter Wong, Barry Leiba, Arnt Gulbrandsen, Stephan Bosch, Ken Murchison, Phil Pennock, Ned Freed, Jeffrey Hutzelman, Mark E. Mallett, Dilyan Palauzov, Dave Cridland, Aaron Stone, Robert Burrell Donkin, Patrick Ben Koetter, Bjoern Hoehrmann, Martin Duerst, Pasi Eronen, Magnus Westerlund, Tim Polk, and Julien Coloos for help with this document. Special thank you to Phil Pennock for providing text for the NOOP command, as well as finding various bugs in the document.
9. References
9.1. Normative References
[ABNF] Crocker, D. and P. Overell, "Augmented BNF for Syntax Specifications: ABNF", STD 68, RFC 5234, January 2008. [ACAP] Newman, C. and J. Myers, "ACAP -- Application Configuration Access Protocol", RFC 2244, November 1997. [BASE64] Josefsson, S., "The Base16, Base32, and Base64 Data Encodings", RFC 4648, October 2006. [DNS-SRV] Gulbrandsen, A., Vixie, P., and L. Esibov, "A DNS RR for specifying the location of services (DNS SRV)", RFC 2782, February 2000. [KEYWORDS] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. [NET-UNICODE] Klensin, J. and M. Padlipsky, "Unicode Format for Network Interchange", RFC 5198, March 2008. [NOTIFY] Melnikov, A., Leiba, B., Segmuller, W., and T. Martin, "Sieve Email Filtering: Extension for Notifications", RFC 5435, January 2009. [RFC2277] Alvestrand, H., "IETF Policy on Character Sets and Languages", BCP 18, RFC 2277, January 1998. [RFC2460] Deering, S. and R. Hinden, "Internet Protocol, Version 6 (IPv6) Specification", RFC 2460, December 1998. [RFC3490] Faltstrom, P., Hoffman, P., and A. Costello, "Internationalizing Domain Names in Applications (IDNA)", RFC 3490, March 2003. [RFC4519] Sciberras, A., "Lightweight Directory Access Protocol (LDAP): Schema for User Applications", RFC 4519, June 2006. [RFC5646] Phillips, A. and M. Davis, "Tags for Identifying Languages", BCP 47, RFC 5646, September 2009. [RFC791] Postel, J., "Internet Protocol", STD 5, RFC 791, September 1981.
[SASL] Melnikov, A. and K. Zeilenga, "Simple Authentication and Security Layer (SASL)", RFC 4422, June 2006. [SASLprep] Zeilenga, K., "SASLprep: Stringprep Profile for User Names and Passwords", RFC 4013, February 2005. [SCRAM] Menon-Sen, A., Melnikov, A., Newman, C., and N. Williams, "Salted Challenge Response Authentication Mechanism (SCRAM) SASL and GSS-API Mechanisms", RFC 5802, July 2010. [SIEVE] Guenther, P. and T. Showalter, "Sieve: An Email Filtering Language", RFC 5228, January 2008. [StringPrep] Hoffman, P. and M. Blanchet, "Preparation of Internationalized Strings ("stringprep")", RFC 3454, December 2002. [TLS] Dierks, T. and E. Rescorla, "The Transport Layer Security (TLS) Protocol Version 1.2", RFC 5246, August 2008. [URI-GEN] Berners-Lee, T., Fielding, R., and L. Masinter, "Uniform Resource Identifier (URI): Generic Syntax", STD 66, RFC 3986, January 2005. [UTF-8] Yergeau, F., "UTF-8, a transformation format of ISO 10646", STD 63, RFC 3629, November 2003. [X509] Cooper, D., Santesson, S., Farrell, S., Boeyen, S., Housley, R., and W. Polk, "Internet X.509 Public Key Infrastructure Certificate and Certificate Revocation List (CRL) Profile", RFC 5280, May 2008. [X509-SRV] Santesson, S., "Internet X.509 Public Key Infrastructure Subject Alternative Name for Expression of Service Name", RFC 4985, August 2007.9.2. Informative References
[DIGEST-MD5] Leach, P. and C. Newman, "Using Digest Authentication as a SASL Mechanism", RFC 2831, May 2000. [GSSAPI] Melnikov, A., "The Kerberos V5 ("GSSAPI") Simple Authentication and Security Layer (SASL) Mechanism", RFC 4752, November 2006.
[I-HAVE] Freed, N., "Sieve Email Filtering: Ihave Extension", RFC 5463, March 2009. [IMAP] Crispin, M., "INTERNET MESSAGE ACCESS PROTOCOL - VERSION 4rev1", RFC 3501, March 2003. [LDAP] Zeilenga, K., "Lightweight Directory Access Protocol (LDAP): Technical Specification Road Map", RFC 4510, June 2006. [PLAIN] Zeilenga, K., "The PLAIN Simple Authentication and Security Layer (SASL) Mechanism", RFC 4616, August 2006.Authors' Addresses
Alexey Melnikov (editor) Isode Limited 5 Castle Business Village 36 Station Road Hampton, Middlesex TW12 2BX UK EMail: Alexey.Melnikov@isode.com Tim Martin BeThereBeSquare, Inc. 672 Haight st. San Francisco, CA 94117 USA Phone: +1 510 260-4175 EMail: timmartin@alumni.cmu.edu