Network Working Group E. Omara Internet-Draft Google Intended status: Informational B. Beurdouche Expires: September 12, 2019 INRIA E. Rescorla Mozilla S. Inguva Twitter A. Kwon MIT A. Duric Wire March 11, 2019 The Messaging Layer Security (MLS) Architecture draft-ietf-mls-architecture-02 Abstract This document describes the architecture and requirements for the Messaging Layer Security (MLS) protocol. MLS provides a security layer for group messaging applications with from two to a large number of clients. It is meant to protect against eavesdropping, tampering, and message forgery. Status of This Memo This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet- Drafts is at https://datatracker.ietf.org/drafts/current/. Internet-Drafts are draft documents 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." This Internet-Draft will expire on September 12, 2019. Copyright Notice Copyright (c) 2019 IETF Trust and the persons identified as the document authors. All rights reserved. Omara, et al. Expires September 12, 2019 [Page 1] Internet-Draft MLS Architecture March 2019 This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://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 2. General Setting . . . . . . . . . . . . . . . . . . . . . . . 3 2.1. Group, Members and Clients . . . . . . . . . . . . . . . 5 2.2. Authentication Service . . . . . . . . . . . . . . . . . 6 2.3. Delivery Service . . . . . . . . . . . . . . . . . . . . 6 2.3.1. Key Storage . . . . . . . . . . . . . . . . . . . . . 7 2.3.2. Key Retrieval . . . . . . . . . . . . . . . . . . . . 7 2.3.3. Delivery of messages and attachments . . . . . . . . 8 2.3.4. Membership knowledge . . . . . . . . . . . . . . . . 8 2.3.5. Membership and offline members . . . . . . . . . . . 9 3. System Requirements . . . . . . . . . . . . . . . . . . . . . 9 3.1. Functional Requirements . . . . . . . . . . . . . . . . . 9 3.1.1. Asynchronous Usage . . . . . . . . . . . . . . . . . 9 3.1.2. Recovery After State Loss . . . . . . . . . . . . . . 9 3.1.3. Support for Multiple Devices . . . . . . . . . . . . 9 3.1.4. Extensibility / Pluggability . . . . . . . . . . . . 10 3.1.5. Privacy . . . . . . . . . . . . . . . . . . . . . . . 10 3.1.6. Federation . . . . . . . . . . . . . . . . . . . . . 10 3.1.7. Compatibility with future versions of MLS . . . . . . 10 3.2. Security Requirements . . . . . . . . . . . . . . . . . . 10 3.2.1. Connections between Clients and Servers (one-to-one) 11 3.2.2. Message Secrecy and Authentication . . . . . . . . . 11 4. Security Considerations . . . . . . . . . . . . . . . . . . . 13 4.1. Transport Security Links . . . . . . . . . . . . . . . . 13 4.2. Delivery Service Compromise . . . . . . . . . . . . . . . 14 4.3. Authentication Service Compromise . . . . . . . . . . . . 14 4.4. Client Compromise . . . . . . . . . . . . . . . . . . . . 14 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 15 6. Contributors . . . . . . . . . . . . . . . . . . . . . . . . 15 7. Informative References . . . . . . . . . . . . . . . . . . . 15 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 16 Omara, et al. Expires September 12, 2019 [Page 2] Internet-Draft MLS Architecture March 2019 1. Introduction RFC EDITOR: PLEASE REMOVE THE FOLLOWING PARAGRAPH The source for this draft is maintained in GitHub. Suggested changes should be submitted as pull requests at https://github.com/mlswg/mls- architecture. Instructions are on that page as well. Editorial changes can be managed in GitHub, but any substantive change should be discussed on the MLS mailing list. End-to-end security is a requirement for instant messaging systems and is commonly deployed in many such systems. In this context, "end-to-end" captures the notion that users of the system enjoy some level of security - with the precise level depending on the system design - even when the messaging service they are using performs unsatisfactorily. Messaging Layer Security (MLS) specifies an architecture (this document) and an abstract protocol [MLSPROTO] for providing end-to- end security in this setting. MLS is not intended as a full instant messaging protocol but rather is intended to be embedded in a concrete protocol such as XMPP [RFC6120]. In addition, it does not specify a complete wire encoding, but rather a set of abstract data structures which can then be mapped onto a variety of concrete encodings, such as TLS [I-D.ietf-tls-tls13], CBOR [RFC7049], and JSON [RFC7159]. Implementations which adopt compatible encodings will have some degree of interoperability at the message level, though they may have incompatible identity/authentication infrastructures. This document is intended to describe the overall messaging system architecture which the MLS protocol fits into, and the requirements which it is intended to fulfill. 2. General Setting A Group using a Messaging Service (MS) comprises a set of participants called Members where each member is typically expected to own multiple devices, called Clients. A group may be as small as two members (the simple case of person to person messaging) or as large as thousands. In order to communicate securely, clients initially use services at their disposal to obtain the necessary secrets and credentials required for security. The Messaging Service (MS) presents as two abstract services that allow clients to prepare for sending and receiving messages securely: o An Authentication Service (AS) which is responsible for maintaining user long term identities, issuing credentials which Omara, et al. Expires September 12, 2019 [Page 3] Internet-Draft MLS Architecture March 2019 allow them to authenticate each other, and potentially allowing users to discover each others long-term identity keys. o A Delivery Service (DS) which is responsible for receiving and redistributing messages between group members. In the case of group messaging, the delivery service may also be responsible for acting as a "broadcaster" where the sender sends a single message to a group which is then forwarded to each recipient in the group by the DS. The DS is also responsible for storing and delivering initial public key material required by clients in order to proceed with the group secret key establishment process. ---------------- -------------- | Authentication | | Delivery | | Service (AS) | | Service (DS) | ---------------- -------------- / | \ Group ********************************************************* * / | \ * * / | \ * * ---------- ---------- ---------- * * | Client 0 | | Client 1 | | Client N | * * ---------- ---------- ---------- * * ............................. ............ * * User 0 User 1 * * * ********************************************************* In many systems, the AS and the DS are actually operated by the same entity and may even be the same server. However, they are logically distinct and, in other systems, may be operated by different entities, hence we show them as being separate here. Other partitions are also possible, such as having a separate directory server. A typical group messaging scenario might look like this: 1. Alice, Bob and Charlie create accounts with a messaging service and obtain credentials from the AS. 2. Alice, Bob and Charlie authenticate to the DS and store some initial keying material which can be used to send encrypted messages to them for the first time. This keying material is authenticated with their long term credentials. 3. When Alice wants to send a message to Bob and Charlie, she contacts the DS and looks up their initial keying material. She Omara, et al. Expires September 12, 2019 [Page 4] Internet-Draft MLS Architecture March 2019 uses these keys to establish a new set of keys which she can use to send encrypted messages to Bob and Charlie. She then sends the encrypted message(s) to the DS, which forwards them to the recipients. 4. Bob and/or Charlie respond to Alice's message. Their messages might trigger a new key derivation step which allows the shared group key to be updated to provide post-compromise security Section 3.2.2.1. Clients may wish to do the following: o create a group by inviting a set of other clients; o add one or more clients to an existing group; o remove one or more members from an existing group; o join an existing group; o leave a group; o send a message to everyone in the group; o receive a message from someone in the group. At the cryptographic level, clients in groups (and by extension Members) are peers. For instance, any client can add another client to a group. This is in contrast to some designs in which there is a single group controller who can modify the group. MLS is compatible with having group administration restricted to certain users, but we assume that those restrictions are enforced by authentication and access control at the application layer. Thus, for instance, while it might be technically possible for any member to send a message adding a new client to a group, the group might have the policy that only certain members are allowed to make changes and thus other members can ignore or reject such a message from an unauthorized user. 2.1. Group, Members and Clients Informally, a group is a set of users who possibly use multiple endpoint devices to interact with the Messaging Service. These members will typically correspond to end-user devices such as phones, web clients or other devices running MLS, which are called clients. Each client owns at least one long term identity key pair that uniquely defines its identity to other clients or members a the Omara, et al. Expires September 12, 2019 [Page 5] Internet-Draft MLS Architecture March 2019 Group. Because a single user may operate multiple devices simultaneously (e.g., a desktop and a phone) or sequentially (e.g., replacing one phone with another), the formal definition of a group in MLS is the set of clients that has knowledge of the shared group secret established in the group key establishment phase of the protocol. Multiple user devices can be grouped, appearing as one virtual client to the rest of the group. In some messaging systems, clients belonging to the same user must all share the same identity key pair, but MLS does not assume this. The MLS architecture considers the more general case and allows for important use cases, such as a member adding a new client when all their existing clients are offline. MLS has been designed to provide similar security guarantees to all clients, for all group sizes, even when it reduces to only two clients. 2.2. Authentication Service The basic function of the Authentication Service (AS) is to provide a trusted mapping from user identities (usernames, phone numbers, etc.), to long-term identity keys, which may either be one per client or may be shared amongst the clients attached to a user. It typically acts as: o A certification authority, or similar service, which signs some sort of portable credential binding an identity to a key; o A directory server which provides the key for a given identity (presumably this connection is secured via some form of transport security such as TLS). By definition, the AS is invested with a large amount of trust. A malicious AS can impersonate - or allow an attacker to impersonate - any user of the system. This risk can be mitigated by publishing the binding between identities and keys in a public log such as Key Transparency (KT) [KeyTransparency]. It is possible to build a functional MLS system without any kind of public key logging, but such a system will necessarily be somewhat vulnerable to attack by a malicious or untrusted AS. 2.3. Delivery Service The Delivery Service (DS) is expected to play multiple roles in the Messaging Service architecture: Omara, et al. Expires September 12, 2019 [Page 6] Internet-Draft MLS Architecture March 2019 o To act as a directory service providing the initial keying material for clients to use. This allows a client to establish a shared key and send encrypted messages to other clients even if the other client is offline. o To route messages between clients and to act as a message broadcaster, taking in one message and forwarding it to multiple clients (also known as "server side fanout"). Depending on the level of trust given by the group to the Delivery Service, the functional and security guarantees provided by MLS may differ. 2.3.1. Key Storage Upon joining the system, each client stores its initial cryptographic key material with the DS. This key material represents the initial contribution that will be used in the establishment of the shared group secret. This initial keying material is authenticated using the client's identity key. Thus, the client stores: o A credential from the Authentication service attesting to the binding between the user and the client's identity key. o The client's initial keying material signed with the client's identity key. As noted above, users may own multiple clients, each with their own keying material, and thus there may be multiple entries stored by each user. The Delivery Service is also responsible for allowing users to add, remove or update their initial keying material and to ensure that the identifier for these keys are unique accross all keys stored on the DS. 2.3.2. Key Retrieval When a client wishes to establish a group and send an initial message to that group, it contacts the DS and retrieves the initial key material for each other client, verifies it using the identity key, and from those forms the group secret, which it can use for the encryption of messages. Omara, et al. Expires September 12, 2019 [Page 7] Internet-Draft MLS Architecture March 2019 2.3.3. Delivery of messages and attachments The DS's main responsibility is to ensure delivery of messages. Specifically, we assume that DSs provide: o Reliable delivery: when a message is provided to the DS, it is eventually delivered to all clients. o In-order delivery: messages are delivered to the group in the order they are received from a given client and in approximately the order in which they are sent by clients. The latter is an approximate guarantee because multiple clients may send messages at the same time and so the DS needs some latitude in enforcing ordering across clients. o Consistent ordering: the DS must ensure that all clients have the same view of message ordering for cryptographically relevant operations. This means that the DS MUST enforce global consistency of the ordering of these messages while MLS provides causal consistency of the application messages for each sender. Note that the DS may provide ordering guarantees by ensuring in-order delivery or by providing messages with some kind of sequence information and allowing clients to reorder on receipt. The MLS protocol itself can verify these properties. For instance, if the DS reorders messages from a client or provides different clients with inconsistent orderings, then clients can detect this misconduct. However, MLS need not provide mechanisms to recover from a misbehaving DS. Note that some forms of DS misbehavior are still possible and difficult to detect. For instance, a DS can simply refuse to relay messages to and from a given client. Without some sort of side information, other clients cannot generally distinguish this form of Denial of Service (DoS) attack. 2.3.4. Membership knowledge Group membership is itself sensitive information and MLS is designed so that neither the DS nor the AS need have static knowledge of which clients are in which group. However, they may learn this information through traffic analysis. For instance, in a server side fanout model, the DS learns that a given client is sending the same message to a set of other clients. In addition, there may be applications of MLS in which the group membership list is stored on some server associated with the MS. Omara, et al. Expires September 12, 2019 [Page 8] Internet-Draft MLS Architecture March 2019 2.3.5. Membership and offline members Because Forward Secrecy (FS) and Post-Compromise Security (PCS) rely on the deletion and replacement of keying material, any client which is persistently offline may still be holding old keying material and thus be a threat to both FS and PCS if it is later compromised. MLS does not inherently defend against this problem, but MLS-using systems can enforce some mechanism for doing so. Typically this will consist of evicting clients which are idle for too long, thus containing the threat of compromise. The precise details of such mechanisms are a matter of local policy and beyond the scope of this document. 3. System Requirements 3.1. Functional Requirements MLS is designed as a large scale group messaging protocol and hence aims to provide performance and safety to its users. Messaging systems that implement MLS provide support for conversations involving two or more members, and aim to scale to approximately 50,000 members, typically including many users using multiple devices. 3.1.1. Asynchronous Usage No operation in MLS requires two distinct users or clients to be online simultaneously. In particular, clients participating in conversations protected using MLS can update shared keys, add or remove new members, and send messages and attachments without waiting for another user's reply. Messaging systems that implement MLS provide a transport layer for delivering messages asynchronously and reliably. 3.1.2. Recovery After State Loss Conversation participants whose local MLS state is lost or corrupted can reinitialize their state and continue participating in the conversation. This may entail some level of message loss, but does not result in permanent exclusion from the group. 3.1.3. Support for Multiple Devices It is typically expected for users within Group to own different devices. Omara, et al. Expires September 12, 2019 [Page 9] Internet-Draft MLS Architecture March 2019 A new device can be added to a group by sharing of an existing client secrets or be considered as a new client by the protocol. This client will not gain access to the history even if it is owned by someone who owns another member of the Group. Restoring history is typically not allowed at the protocol level but applications can elect to provide such a mechanism outside of MLS. 3.1.4. Extensibility / Pluggability Messages that do not affect the group state can carry an arbitrary payload with the purpose of sharing that payload between group members. No assumptions are made about the format of the payload. 3.1.5. Privacy The protocol is designed in a way that limits the server-side (AS and DS) metadata footprint. The DS only persists data required for the delivery of messages and avoid Personally Identifiable Information (PII) or other sensitive metadata wherever possible. A Messaging Service provider that has control over both the AS and the DS, will not be able to correlate encrypted messages forwarded by the DS, with the initial public keys signed by the AS. 3.1.6. Federation The protocol aims to be compatible with federated environments. While this document does not specify all necessary mechanisms required for federation, multiple MLS implementations can interoperate to form federated systems if they use compatible wire encodings. 3.1.7. Compatibility with future versions of MLS It is important that multiple versions of MLS be able to coexist in the future. Thus, MLS offers a version negotiation mechanism; this mechanism prevents version downgrade attacks where an attacker would actively rewrite messages messages with a lower protocol version than the ones originally offered by the endpoints. When multiple versions of MLS are available, the negotiation protocol guarantees that the version agreed upon will be the highest version supported in common by the group. 3.2. Security Requirements Omara, et al. Expires September 12, 2019 [Page 10] Internet-Draft MLS Architecture March 2019 3.2.1. Connections between Clients and Servers (one-to-one) We assume that all transport connections are secured via some transport layer security mechanism such as TLS [I-D.ietf-tls-tls13]. However, as noted above, the security of MLS will generally survive compromise of the transport layer, so long as identity keys provided by the AS are authenticated at a minimum. 3.2.2. Message Secrecy and Authentication The trust establishment step of the MLS protocol is followed by a conversation protection step where encryption is used by clients to transmit authenticated messages to other clients through the DS. This ensures that the DS does not have access to the group's private content. MLS aims to provide secrecy, integrity and authentication for all messages. Message Secrecy in the context of MLS means that only intended recipients (current group members), can read any message sent to the group, even in the context of an active adversary as described in the threat model. Message Integrity and Authentication mean that an honest client can only accept a message if it was sent by a group member and that a client cannot send a message which other clients would accept as being from a different client. A corollary to this statement is that the AS and the DS cannot read the content of messages sent between members as they are not members of the group. MLS optionally provides additional protections regarding traffic analysis so as to reduce the ability of adversaries, to deduce the content of the messages depending on (for example) their size. One of these protections includes padding messages in order to produce ciphertexts of standard length. While this protection is highly recommended it is not mandatory as it can be costly in terms of performance for clients and the MS. Message content can be deniable if the signature keys are exchanged over a deniable channel prior to signing messages. 3.2.2.1. Forward and Post-Compromise Security MLS provides additional protection regarding secrecy of past messages and future messages. These cryptographic security properties are Forward Secrecy (FS) and Post-Compromise Security (PCS). Omara, et al. Expires September 12, 2019 [Page 11] Internet-Draft MLS Architecture March 2019 FS means that access to all encrypted traffic history combined with an access to all current keying material on clients will not defeat the secrecy properties of messages older than the oldest key of the compromised client. Note that this means that clients have the extremely important role of deleting appropriate keys as soon as they have been used with the expected message, otherwise the secrecy of the messages and the security for MLS is considerably weakened. PCS means that if a group member is compromised at some time T but subsequently performs an update at some time T', then all MLS guarantees apply to messages sent after time T'. For example, if an adversary learns all secrets known to Alice at time T, including both Alice's secrets and all shared group secrets, but Alice performs a key update at time T', which is not under the control of the adversary, then the adversary is unable to violate any of the MLS security properties after time T'. Both of these properties are satisfied even against compromised DSs and ASs. 3.2.2.2. Membership Changes MLS aims to provide agreement on group membership, meaning that all group members have agreed on the list of current group members. Some applications may wish to enforce ACLs to limit addition or removal of group members, to privileged clients or users. Others may wish to require authorization from the current group members or a subset thereof. Regardless, MLS does not allow addition or removal of group members without informing all other members. Once a client is part of a group, the set of devices controlled by the user can only be altered by an authorized member of the group. This authorization could depend on the application: some applications might want to allow certain other members of the group to add or remove devices on behalf of another member, while other applications might want a more strict policy and allow only the owner of the devices to add or remove them at the potential cost of weaker PCS guarantees. Members who are removed from a group do not enjoy special privileges: compromise of a removed group member does not affect the security of messages sent after their removal but might affect previous messages if the group secrets have not been deleted properly. Omara, et al. Expires September 12, 2019 [Page 12] Internet-Draft MLS Architecture March 2019 3.2.2.3. Security of Attachments The security properties expected for attachments in the MLS protocol are very similar to the ones expected from messages. The distinction between messages and attachments stems from the fact that the typical average time between the download of a message and the one from the attachments may be different. For many reasons (a typical reason being the lack of high bandwidth network connectivity), the lifetime of the cryptographic keys for attachments is usually higher than for messages, hence slightly weakening the PCS guarantees for attachments. 3.2.2.4. Denial of Service In general we do not consider Denial of Service (DoS) resistance to be the responsibility of the protocol. However, it should not be possible for anyone aside from the DS to perform a trivial DoS attack from which it is hard to recover. 3.2.2.5. Non-Repudiation vs Deniability As described in Section 4.4, MLS provides strong authentication within a group, such that a group member cannot send a message that appears to be from another group member. Additionally, some services require that a recipient be able to prove to the messaging service that a message was sent by a given client, in order to report abuse. MLS supports both of these use cases. In some deployments, these services are provided by mechanisms which allow the receiver to prove a message's origin to a third party (this if often called "non- repudiation"), but it should also be possible to operate MLS in a "deniable" mode where such proof is not possible. [[OPEN ISSUE: Exactly how to supply this is still a protocol question.]] 4. Security Considerations MLS adopts the Internet threat model [RFC3552] and therefore assumes that the attacker has complete control of the network. It is intended to provide the security services described in in the face of such attackers. In addition, these guarantees are intended to degrade gracefully in the presence of compromise of the transport security links as well as of both clients and elements of the messaging system, as described in the remainder of this section. 4.1. Transport Security Links [TODO: Mostly DoS, message suppression, and leakage of group membership.] Omara, et al. Expires September 12, 2019 [Page 13] Internet-Draft MLS Architecture March 2019 4.2. Delivery Service Compromise MLS is intended to provide strong guarantees in the face of compromise of the DS. Even a totally compromised DS should not be able to read messages or inject messages that will be acceptable to legitimate clients. It should also not be able to undetectably remove, reorder or replay messages. However, a DS can mount a variety of DoS attacks on the system, including total DoS attacks (where it simply refuses to forward any messages) and partial DoS attacks (where it refuses to forward messages to and from specific clients). As noted in Section 2.3.3, these attacks are only partially detectable by clients without an out-of-band channel. Ultimately, failure of the DS to provide reasonable service must be dealt with as a customer service matter, not via technology. Because the DS is responsible for providing the initial keying material to clients, it can provide stale keys. This does not inherently lead to compromise of the message stream, but does allow it to attack forward security to a limited extent. This threat can be mitigated by having initial keys expire. 4.3. Authentication Service Compromise A compromised AS is a serious matter, as the AS can provide incorrect or adversarial identities to clients. As noted in Section 2.2, mitigating this form of attack requires some sort of transparency/ logging mechanism. Without such a mechanism, MLS will only provide limited security against a compromised AS. 4.4. Client Compromise In general, MLS only provides limited protection against compromised clients. When the client secrets are compromised, then the attacker will obviously be able to decrypt any messages for groups in which the client is a member. It will also be able to send messages impersonating the compromised client until the client updates its keying material (see Section 3.2.2.1). MLS attempts to provide some security in the face of client compromise. In addition, a client cannot send a message to a group which appears to be from another client with a different identity. Note that if devices from the same user share keying material, then one will be able to impersonate another. Finally, clients should not be able to perform DoS attacks Section 3.2.2.4. Omara, et al. Expires September 12, 2019 [Page 14] Internet-Draft MLS Architecture March 2019 5. IANA Considerations This document makes no requests of IANA. 6. Contributors o Katriel Cohn-Gordon University of Oxford me@katriel.co.uk o Cas Cremers University of Oxford cas.cremers@cs.ox.ac.uk o Thyla van der Merwe Royal Holloway, University of London thyla.van.der@merwe.tech o Jon Millican Facebook jmillican@fb.com o Raphael Robert Wire raphael@wire.com 7. Informative References [I-D.ietf-tls-tls13] Rescorla, E., "The Transport Layer Security (TLS) Protocol Version 1.3", draft-ietf-tls-tls13-28 (work in progress), March 2018. [KeyTransparency] Google, ., "Key Transparency", n.d., . [MLSPROTO] Barnes, R., Millican, J., Omara, E., Cohn-Gordon, K., and R. Robert, "Messaging Layer Security Protocol", 2018. [RFC3552] Rescorla, E. and B. Korver, "Guidelines for Writing RFC Text on Security Considerations", BCP 72, RFC 3552, DOI 10.17487/RFC3552, July 2003, . Omara, et al. Expires September 12, 2019 [Page 15] Internet-Draft MLS Architecture March 2019 [RFC6120] Saint-Andre, P., "Extensible Messaging and Presence Protocol (XMPP): Core", RFC 6120, DOI 10.17487/RFC6120, March 2011, . [RFC7049] Bormann, C. and P. Hoffman, "Concise Binary Object Representation (CBOR)", RFC 7049, DOI 10.17487/RFC7049, October 2013, . [RFC7159] Bray, T., Ed., "The JavaScript Object Notation (JSON) Data Interchange Format", RFC 7159, DOI 10.17487/RFC7159, March 2014, . Authors' Addresses Emad Omara Google Email: emadomara@google.com Benjamin Beurdouche INRIA Email: benjamin.beurdouche@inria.fr Eric Rescorla Mozilla Email: ekr@rtfm.com Srinivas Inguva Twitter Email: singuva@twitter.com Albert Kwon MIT Email: kwonal@mit.edu Alan Duric Wire Email: alan@wire.com Omara, et al. Expires September 12, 2019 [Page 16]