idnits 2.17.1 draft-ietf-tokbind-https-13.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- ** There is 1 instance of too long lines in the document, the longest one being 2 characters in excess of 72. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (April 12, 2018) is 2205 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Outdated reference: A later version (-14) exists of draft-ietf-tokbind-negotiation-10 == Outdated reference: A later version (-19) exists of draft-ietf-tokbind-protocol-16 ** Obsolete normative reference: RFC 2818 (Obsoleted by RFC 9110) ** Obsolete normative reference: RFC 5246 (Obsoleted by RFC 8446) ** Obsolete normative reference: RFC 7230 (Obsoleted by RFC 9110, RFC 9112) ** Obsolete normative reference: RFC 7231 (Obsoleted by RFC 9110) -- Obsolete informational reference (is this intentional?): RFC 7540 (Obsoleted by RFC 9113) Summary: 5 errors (**), 0 flaws (~~), 3 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Internet Engineering Task Force A. Popov 3 Internet-Draft M. Nystroem 4 Intended status: Standards Track Microsoft Corp. 5 Expires: October 14, 2018 D. Balfanz, Ed. 6 A. Langley 7 N. Harper 8 Google Inc. 9 J. Hodges 10 PayPal 11 April 12, 2018 13 Token Binding over HTTP 14 draft-ietf-tokbind-https-13 16 Abstract 18 This document describes a collection of mechanisms that allow HTTP 19 servers to cryptographically bind security tokens (such as cookies 20 and OAuth tokens) to TLS connections. 22 We describe both first-party and federated scenarios. In a first- 23 party scenario, an HTTP server is able to cryptographically bind the 24 security tokens it issues to a client, and which the client 25 subsequently returns to the server, to the TLS connection between the 26 client and server. Such bound security tokens are protected from 27 misuse since the server can generally detect if they are replayed 28 inappropriately, e.g., over other TLS connections. 30 Federated token bindings, on the other hand, allow servers to 31 cryptographically bind security tokens to a TLS connection that the 32 client has with a different server than the one issuing the token. 34 This Internet-Draft is a companion document to The Token Binding 35 Protocol. 37 Status of This Memo 39 This Internet-Draft is submitted in full conformance with the 40 provisions of BCP 78 and BCP 79. 42 Internet-Drafts are working documents of the Internet Engineering 43 Task Force (IETF). Note that other groups may also distribute 44 working documents as Internet-Drafts. The list of current Internet- 45 Drafts is at https://datatracker.ietf.org/drafts/current/. 47 Internet-Drafts are draft documents valid for a maximum of six months 48 and may be updated, replaced, or obsoleted by other documents at any 49 time. It is inappropriate to use Internet-Drafts as reference 50 material or to cite them other than as "work in progress." 52 This Internet-Draft will expire on October 14, 2018. 54 Copyright Notice 56 Copyright (c) 2018 IETF Trust and the persons identified as the 57 document authors. All rights reserved. 59 This document is subject to BCP 78 and the IETF Trust's Legal 60 Provisions Relating to IETF Documents 61 (https://trustee.ietf.org/license-info) in effect on the date of 62 publication of this document. Please review these documents 63 carefully, as they describe your rights and restrictions with respect 64 to this document. Code Components extracted from this document must 65 include Simplified BSD License text as described in Section 4.e of 66 the Trust Legal Provisions and are provided without warranty as 67 described in the Simplified BSD License. 69 Table of Contents 71 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 72 1.1. Requirements Language . . . . . . . . . . . . . . . . . . 4 73 2. The Sec-Token-Binding HTTP Request Header Field . . . . . . . 4 74 2.1. HTTPS Token Binding Key Pair Scoping . . . . . . . . . . 5 75 3. TLS Renegotiation . . . . . . . . . . . . . . . . . . . . . . 6 76 4. First-Party Use Cases . . . . . . . . . . . . . . . . . . . . 6 77 5. Federation Use Cases . . . . . . . . . . . . . . . . . . . . 6 78 5.1. Introduction . . . . . . . . . . . . . . . . . . . . . . 7 79 5.2. Overview . . . . . . . . . . . . . . . . . . . . . . . . 8 80 5.3. HTTP Redirects . . . . . . . . . . . . . . . . . . . . . 10 81 5.4. Negotiated Key Parameters . . . . . . . . . . . . . . . . 12 82 5.5. Federation Example . . . . . . . . . . . . . . . . . . . 12 83 6. Implementation Considerations . . . . . . . . . . . . . . . . 15 84 7. Security Considerations . . . . . . . . . . . . . . . . . . . 15 85 7.1. Security Token Replay . . . . . . . . . . . . . . . . . . 15 86 7.2. Triple Handshake Vulnerability in TLS 1.2 and Older TLS 87 Versions . . . . . . . . . . . . . . . . . . . . . . . . 16 88 7.3. Sensitivity of the Sec-Token-Binding Header . . . . . . . 16 89 7.4. Securing Federated Sign-On Protocols . . . . . . . . . . 17 90 8. Privacy Considerations . . . . . . . . . . . . . . . . . . . 19 91 8.1. Scoping of Token Binding Key Pairs . . . . . . . . . . . 19 92 8.2. Lifetime of Token Binding Key Pairs . . . . . . . . . . . 20 93 8.3. Correlation . . . . . . . . . . . . . . . . . . . . . . . 20 94 9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 21 95 10. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 22 96 11. References . . . . . . . . . . . . . . . . . . . . . . . . . 22 97 11.1. Normative References . . . . . . . . . . . . . . . . . . 22 98 11.2. Informative References . . . . . . . . . . . . . . . . . 23 99 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 24 101 1. Introduction 103 The Token Binding Protocol [I-D.ietf-tokbind-protocol] defines a 104 Token Binding ID for a TLS connection between a client and a server. 105 The Token Binding ID of a TLS connection is constructed using the 106 public key of a private-public key pair. The client proves 107 possession of the corresponding private key. This Token Binding key 108 pair is long-lived. I.e., subsequent TLS connections between the 109 same client and server have the same Token Binding ID, unless 110 specifically reset, e.g., by the user. When issuing a security token 111 (e.g., an HTTP cookie or an OAuth token [RFC6749]) to a client, the 112 server can include the Token Binding ID in the token, thus 113 cryptographically binding the token to TLS connections between that 114 particular client and server, and inoculating the token against abuse 115 (re-use, attempted impersonation, etc.) by attackers. 117 While the Token Binding Protocol [I-D.ietf-tokbind-protocol] defines 118 a message format for establishing a Token Binding ID, it does not 119 specify how this message is embedded in higher-level protocols. The 120 purpose of this specification is to define how TokenBindingMessages 121 are embedded in HTTP (both versions 1.1 [RFC7230] and 2 [RFC7540]). 122 Note that TokenBindingMessages are only defined if the underlying 123 transport uses TLS. This means that Token Binding over HTTP is only 124 defined when the HTTP protocol is layered on top of TLS (commonly 125 referred to as HTTPS [RFC2818]). 127 HTTP clients establish a Token Binding ID with a server by including 128 a special HTTP header field in HTTP requests. The HTTP header field 129 value is a base64url-encoded TokenBindingMessage. 131 TokenBindingMessages allow clients to establish multiple Token 132 Binding IDs with the server, by including multiple TokenBinding 133 structures in the TokenBindingMessage. By default, a client will 134 establish a provided Token Binding ID with the server, indicating a 135 Token Binding ID that the client will persistently use with the 136 server. Under certain conditions, the client can also include a 137 referred Token Binding ID in the TokenBindingMessage, indicating a 138 Token Binding ID that the client is using with a different server 139 than the one that the TokenBindingMessage is sent to. This is useful 140 in federation scenarios. 142 1.1. Requirements Language 144 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 145 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 146 document are to be interpreted as described in [RFC2119]. 148 2. The Sec-Token-Binding HTTP Request Header Field 150 Once a client and server have negotiated the Token Binding Protocol 151 with HTTP/1.1 or HTTP/2 (see [I-D.ietf-tokbind-protocol] and 152 [I-D.ietf-tokbind-negotiation]), clients MUST include a Sec-Token- 153 Binding header field in their HTTP requests, and MUST include only 154 one such header field per HTTP request. Also, The Sec-Token-Binding 155 header field MUST NOT be included in HTTP responses. The ABNF of the 156 Sec-Token-Binding header field is (in [RFC7230] style, see also 157 Section 8.3 of [RFC7231]): 159 Sec-Token-Binding = EncodedTokenBindingMessage 161 The header field name is Sec-Token-Binding and its single value, 162 EncodedTokenBindingMessage, is a base64url encoding of a single 163 TokenBindingMessage, as defined in [I-D.ietf-tokbind-protocol], using 164 the URL- and filename-safe character set described in Section 5 of 165 [RFC4648], with all trailing padding characters '=' omitted and 166 without the inclusion of any line breaks, whitespace, or other 167 additional characters. 169 For example: 171 Sec-Token-Binding: AIkAAgBBQFzK4_bhAqLDwRQxqJWte33d7hZ0hZWHwk-miKPg4E\ 172 9fcgs7gBPoz-9RfuDfN9WCw6keHEw1ZPQMGs9CxpuHm-YAQM_j\ 173 aOwwej6a-cQBGU7CJpUHOvXG4VvjNq8jDsvta9Y8_bPEPj25Gg\ 174 mKiPjhJEtZA6mJ_9SNifLvVBTi7fR9wSAAAA 176 If the server receives more than one Sec-Token-Binding header field 177 in an HTTP request, then the server MUST reject the message with a 178 400 (Bad Request) HTTP status code. Additionally, the Sec-Token- 179 Binding header field: 181 SHOULD NOT be stored by origin servers on PUT requests, 183 MAY be listed by a server in a Vary response header field, and, 185 MUST NOT be used in HTTP trailers. 187 The TokenBindingMessage MUST contain exactly one TokenBinding 188 structure with TokenBindingType of provided_token_binding, which MUST 189 be signed with the Token Binding private key used by the client for 190 connections between itself and the server that the HTTP request is 191 sent to (clients use different Token Binding key pairs for different 192 servers, see Section 2.1 below). The Token Binding ID established by 193 this TokenBinding is called a Provided Token Binding ID. 195 The TokenBindingMessage MAY also contain exactly one TokenBinding 196 structure with TokenBindingType of referred_token_binding, as 197 specified in Section 5.3. In addition to the latter, or rather than 198 the latter, the TokenBindingMessage MAY contain other TokenBinding 199 structures. This is use case-specific, and such use cases are 200 outside the scope of this specification. 202 A TokenBindingMessage is validated by the server as described in 203 Section 4.2. ("Server Processing Rules") of 204 [I-D.ietf-tokbind-protocol]. If validation fails and a Token Binding 205 is rejected, any associated bound tokens MUST also be rejected by the 206 server. HTTP requests containing invalid tokens MUST be rejected. 207 In this case, the server application MAY return HTTP status code 400 208 (Bad Request) or proceed with an application-specific invalid token 209 response (e.g., directing the client to re-authenticate and present a 210 different token), or terminate the connection. 212 In HTTP/2, the client SHOULD use Header Compression [RFC7541] to 213 avoid the overhead of repeating the same header field in subsequent 214 HTTP requests. 216 2.1. HTTPS Token Binding Key Pair Scoping 218 HTTPS is used in conjunction with various application protocols and 219 application contexts, in various ways. For example, general-purpose 220 Web browsing is one such HTTP-based application context. Within the 221 latter context, HTTP cookies [RFC6265] are typically utilized for 222 state management, including client authentication. A related, though 223 distinct, example of other HTTP-based application contexts is where 224 OAuth tokens [RFC6749] are utilized to manage authorization for 225 third-party application access to resources. The token scoping rules 226 of these two examples can differ: the scoping rules for cookies are 227 concisely specified in [RFC6265], whereas OAuth is a framework and 228 defines various token types with various scopings, some of which are 229 determined by the encompassing application. 231 The scoping of Token Binding key pairs generated by Web browsers for 232 use in first-party and federation use cases defined in this 233 specification (Section 5), and intended for binding HTTP cookies, 234 MUST be no wider than the granularity of "effective top-level domain 235 (public suffix) + 1" (eTLD+1). I.e., the scope of Token Binding key 236 pairs is no wider than the scope at which cookies can be set (see 238 [RFC6265]), but MAY be more narrow if cookies are scoped more 239 narrowly. 241 Key pairs used to bind other application tokens, such as OAuth tokens 242 or OpenID Connect ID Tokens, SHOULD generally adhere to the above 243 eTLD+1 scoping requirement for those tokens being employed in first- 244 party or federation scenarios. Applications other than Web browsers 245 MAY use different key pair scoping rules. See also Section 8.1, 246 below. 248 Scoping rules for other HTTP-based application contexts are outside 249 the scope of this specification. 251 3. TLS Renegotiation 253 Token Binding over HTTP/1.1 [RFC7230] can be performed in combination 254 with TLS renegotiation. In this case, renegotiation MUST only occur 255 between a client's HTTP request and the server's response, the client 256 MUST NOT send any pipelined requests, and the client MUST NOT 257 initiate renegotiation. (I.e., the client may only send a 258 renegotiation ClientHello in response to the server's HelloRequest.) 259 These conditions ensure that both the client and the server can 260 clearly identify which TLS Exported Keying Material value [RFC5705] 261 to use when generating or verifying the TokenBindingMessage. This 262 also prevents a TokenBindingMessage from being split across TLS 263 renegotiation boundaries. (I.e., due to TLS message fragmentation - 264 see Section 6.2.1 of [RFC5246].) 266 4. First-Party Use Cases 268 In a first-party use case (also known as a "same-site" use case), an 269 HTTP server issues a security token such as a cookie (or similar) to 270 a client, and expects the client to return the security token at a 271 later time, e.g., in order to authenticate. Binding the security 272 token to the TLS connection between client and server protects the 273 security token from misuse, since the server can detect if the 274 security token is replayed inappropriately, e.g., over other TLS 275 connections. 277 See Section 5 of [I-D.ietf-tokbind-protocol] for general guidance 278 regarding binding of security tokens and their subsequent validation. 280 5. Federation Use Cases 281 5.1. Introduction 283 For privacy reasons, clients use different Token Binding key pairs to 284 establish Provided Token Binding IDs with different servers. As a 285 result, a server cannot bind a security token (such as an OAuth token 286 or an OpenID Connect ID Token [OpenID.Core]) to a TLS connection that 287 the client has with a different server. This is, however, a common 288 requirement in federation scenarios: For example, an Identity 289 Provider may wish to issue an identity token to a client and 290 cryptographically bind that token to the TLS connection between the 291 client and a Relying Party. 293 In this section, we describe mechanisms to achieve this. The common 294 idea among these mechanisms is that a server (called the Token 295 Consumer in this document) signals to the client that it should 296 reveal the Provided Token Binding ID that is used between the client 297 and itself, to another server (called the Token Provider in this 298 document). Also common across the mechanisms is how the Token 299 Binding ID is revealed to the Token Provider: The client uses the 300 Token Binding Protocol [I-D.ietf-tokbind-protocol], and includes a 301 TokenBinding structure in the Sec-Token-Binding HTTP header field 302 defined above. What differs between the various mechanisms is how 303 the Token Consumer signals to the client that it should reveal the 304 Token Binding ID to the Token Provider. Below, we specify one such 305 mechanism, which is suitable for redirect-based interactions between 306 Token Consumers and Token Providers. 308 Client Token Consumer Token Provider 309 +--------+ +----+ +-----+ 310 | Client | | TC | | TP | 311 +--------+ +----+ +-----+ 312 | | | 313 | | | 314 | | | 315 | Client interacts w/TC | | 316 | using TokenBindingID TBID1: | | 317 | TBMSG[[provided_token_binding,| | 318 | TBID1, signature]] | | 319 |------------------------------>| | 320 | | | 321 | Client interacts w/TP | 322 | using TokenBindingID TBID2: | 323 | TBMSG[[provided_token_binding, | 324 | TBID2, signature]] | 325 |----------------------------------------------------->| 326 | | 327 | | | 328 | TC signals permission to | | 329 | reveal TBID1 to TP | | 330 |<------------------------------| | 331 | | | 332 | | 333 | Client interacts w/TP | 334 | using TokenBindingID TBID1 and TBID2: | 335 | TBMSG[[provided_token_binding, | 336 | TBID2, signature], | 337 | [referred_token_binding, | 338 | TBID1, signature]] | 339 |----------------------------------------------------->| 340 | | 341 | | | 342 | | | 344 5.2. Overview 346 In a Federated Sign-On protocol, an Identity Provider issues an 347 identity token to a client, which sends the identity token to a 348 Relying Party to authenticate itself. Examples of this include 349 OpenID Connect (in which the identity token is called an "ID Token") 350 and SAML (in which the identity token is a SAML assertion). 352 To better protect the security of the identity token, the Identity 353 Provider may wish to bind the identity token to the TLS connection 354 between the client and the Relying Party, thus ensuring that only 355 said client can use the identity token. The Relying Party will 356 compare the Token Binding ID (or a cryptographic hash of it) in the 357 identity token with the Token Binding ID (or a hash thereof) of the 358 TLS connection between this Relying Party and the client. 360 This is an example of a federation scenario, which more generally can 361 be described as follows: 363 o A Token Consumer causes the client to issue a token request to the 364 Token Provider. The goal is for the client to obtain a token and 365 then use it with the Token Consumer. 367 o The client delivers the token request to the Token Provider. 369 o The Token Provider issues the token. The token is issued for the 370 specific Token Consumer who requested it (thus preventing 371 malicious Token Consumers from using tokens with other Token 372 Consumers). The token is, however, typically a bearer token, 373 meaning that any client can use it with the Token Consumer, not 374 just the client to which it was issued. 376 o Therefore, in the previous step, the Token Provider may want to 377 include in the token the Token Binding ID (or a cryptographic hash 378 of it) that the client uses when communicating with the Token 379 Consumer, thus binding the token to the client's Token Binding key 380 pair. The client proves possession of the private key when 381 communicating with the Token Consumer through the Token Binding 382 Protocol [I-D.ietf-tokbind-protocol], and uses the corresponding 383 public key of this key pair as a component of the Token Binding 384 ID. Comparing the Token Binding ID from the token to the Token 385 Binding ID established with the client allows the Token Consumer 386 to verify that the token was sent to it by the legitimate client. 388 o To allow the Token Provider to include the Token Binding ID in the 389 token, the Token Binding ID (between client and Token Consumer) 390 must therefore be communicated to the Token Provider along with 391 the token request. Communicating a Token Binding ID involves 392 proving possession of a private key and is described in the Token 393 Binding Protocol [I-D.ietf-tokbind-protocol]. 395 The client will perform this last operation (proving possession of a 396 private key that corresponds to a Token Binding ID between the client 397 and the Token Consumer while delivering the token request to the 398 Token Provider) only if the Token Consumer requests the client to do 399 so. 401 Below, we specify how Token Consumers can signal this request in 402 redirect-based federation protocols. Note that this assumes that the 403 federated sign-on flow starts at the Token Consumer, or at the very 404 least, includes a redirect from the Token Consumer to the Token 405 Provider. It is outside the scope of this document to specify 406 similar mechanisms for flows that do not include such redirects. 408 5.3. HTTP Redirects 410 When a Token Consumer redirects the client to a Token Provider as a 411 means to deliver the token request, it SHOULD include an Include- 412 Referred-Token-Binding-ID HTTP response header field in its HTTP 413 response. The ABNF of the Include-Referred-Token-Binding-ID header 414 is (in [RFC7230] style, see also Section 8.3 of [RFC7231]): 416 Include-Referred-Token-Binding-ID = "true" 418 Where the header field name is "Include-Referred-Token-Binding-ID", 419 and the field-value of "true" is case-insensitive. For example: 421 Include-Referred-Token-Binding-ID: true 423 Including this response header field signals to the client that it 424 should reveal, to the Token Provider, the Token Binding ID used 425 between itself and the Token Consumer. In the absence of this 426 response header field, the client will not disclose any information 427 about the Token Binding used between the client and the Token 428 Consumer to the Token Provider. 430 As illustrated in Section 5.5, when a client receives this header 431 field, it should take the TokenBindingID of the provided TokenBinding 432 from the referrer and create a referred TokenBinding with it to 433 include in the TokenBindingMessage on the redirect request. In other 434 words, the Token Binding message in the redirect request to the Token 435 Provider now includes one provided binding and one referred binding, 436 the latter constructed from the binding between the client and the 437 Token Consumer. 439 When a client receives the Include-Referred-Token-Binding-ID header, 440 it includes the referred token binding even if both the Token 441 Provider and the Token Consumer fall under the same eTLD+1 and the 442 provided and referred token binding IDs are the same. Note that the 443 referred token binding is sent only on the request resulting from the 444 redirect and not on any subsequent requests to the Token Provider. 446 If the Include-Referred-Token-Binding-ID header field is received in 447 response to a request that did not include the Token-Binding header 448 field, the client MUST ignore the Include-Referred-Token-Binding-ID 449 header field. 451 This header field has only meaning if the HTTP status code is 301, 452 302, 303, 307 or 308, and MUST be ignored by the client for any other 453 status codes. If the client supports the Token Binding Protocol, and 454 has negotiated the Token Binding Protocol with both the Token 455 Consumer and the Token Provider, it already sends the Sec-Token- 456 Binding header field to the Token Provider with each HTTP request (as 457 described in Section 2 above). 459 The TokenBindingMessage SHOULD contain a TokenBinding with 460 TokenBindingType referred_token_binding. If included, this 461 TokenBinding MUST be signed with the Token Binding private key used 462 by the client for connections between itself and the Token Consumer 463 (more specifically, the server that issued the Include-Referred- 464 Token-Binding-ID response header field). The Token Binding ID 465 established by this TokenBinding is called a Referred Token Binding 466 ID. 468 As described above, the TokenBindingMessage MUST additionally contain 469 a Provided Token Binding ID, i.e., a TokenBinding structure with 470 TokenBindingType of provided_token_binding, which MUST be signed with 471 the Token Binding private key used by the client for connections 472 between itself and the Token Provider (more specifically, the server 473 that the token request is being sent to). 475 If, for some deployment-specific reason, the initial Token Provider 476 ("TP1") needs to redirect the client to another Token Provider 477 ("TP2"), rather than directly back to the Token Consumer, it can be 478 accommodated using the header fields defined in this specification in 479 the following fashion ("the redirect-chain approach"): 481 Initially, the client is redirected to TP1 by the Token Consumer 482 ("TC"), as described above. Upon receiving the client's request, 483 containing a TokenBindingMessage which contains both provided and 484 referred TokenBindings (for TP1 and TC, respectively), TP1 485 responds to the client with a redirect response containing the 486 Include-Referred-Token-Binding-ID header field and directing the 487 client to send a request to TP2. This causes the client to follow 488 the same pattern and send a request containing a 489 TokenBindingMessage which contains both provided and referred 490 TokenBindings (for TP2 and TP1, respectively) to TP2. Note that 491 this pattern can continue to further Token Providers. In this 492 case, TP2 issues a security token, bound to the client's 493 TokenBinding with TP1, and sends a redirect response to the client 494 pointing to TP1. TP1 in turn constructs a security token for the 495 Token Consumer, bound to the TC's referred TokenBinding which had 496 been conveyed earlier, and sends a redirect response pointing to 497 the TC, containing the bound security token, to the client. 499 The above is intended as only a non-normative example. Details are 500 specific to deployment contexts. Other approaches are possible, but 501 are outside the scope of this specification. 503 5.4. Negotiated Key Parameters 505 The TLS Extension for Token Binding Protocol Negotiation 506 [I-D.ietf-tokbind-negotiation] allows the server and client to 507 negotiate the parameters (signature algorithm, length) of the Token 508 Binding key pair. It is possible that the Token Binding ID used 509 between the client and the Token Consumer, and the Token Binding ID 510 used between the client and Token Provider, use different key 511 parameters. The client MUST use the key parameters negotiated with 512 the Token Consumer in the referred_token_binding TokenBinding of the 513 TokenBindingMessage, even if those key parameters are different from 514 the ones negotiated with the server that the header field is sent to. 516 Token Providers SHOULD support all the Token Binding key parameters 517 specified in [I-D.ietf-tokbind-protocol]. If a token provider does 518 not support the key parameters specified in the 519 referred_token_binding TokenBinding in the TokenBindingMessage, it 520 MUST NOT issue a bound token. 522 5.5. Federation Example 524 The diagram below shows a typical HTTP Redirect-based Web Browser SSO 525 Profile (no artifact, no callbacks), featuring binding of, e.g., a 526 TLS Token Binding ID into an OpenID Connect ID Token. 528 Legend: 530 +------------+------------------------------------------------------+ 531 | EKM: | TLS Exported Keying Material [RFC5705] | 532 | {EKMn}Ksm: | EKM for server "n", signed by private key of TBID | 533 | | "m", where "n" must represent server receiving the | 534 | | ETBMSG, if a conveyed TB's type is | 535 | | provided_token_binding, then m = n, else if TB's | 536 | | type is referred_token_binding, then m != n. E.g., | 537 | | see step 1b in diagram below. | 538 | ETBMSG: | "Sec-Token-Binding" HTTP header field conveying an | 539 | | EncodedTokenBindingMessage, in turn conveying | 540 | | TokenBinding (TB)struct(s), e.g.: ETBMSG[[TB]] or | 541 | | ETBMSG[[TB1],[TB2]] | 542 | ID Token: | the ID Token in OpenID Connect, it is the semantic | 543 | | equivalent of a SAML "authentication assertion". "ID | 544 | | Token w/TBIDn" denotes a "token bound" ID Token | 545 | | containing TBIDn. | 546 | Ks & Kp: | private (aka secret) key, and public key, | 547 | | respectively, of client-side Token Binding key pair | 548 | OIDC: | OpenID Connect | 549 | TB: | TokenBinding struct containing signed EKM, TBID, and | 550 | | TB type, e.g.: | 551 | | [{EKM1}Ks1,TBID1,provided_token_binding] | 552 | TBIDn: | Token Binding ID for client and server n's token- | 553 | | bound TLS association. TBIDn contains Kpn. | 554 +------------+------------------------------------------------------+ 556 Client, Token Consumer, Token Provider, 557 aka: aka: aka: 558 User Agent OpenID Client, OpenID Provider, 559 OIDC Relying Party, OIDC Provider, 560 SAML Relying Party SAML Identity Provider 561 [ server "1" ] [ server "2" ] 562 +--------+ +----+ +-----+ 563 | Client | | TC | | TP | 564 +--------+ +----+ +-----+ 565 | | | 566 | | | 567 | | | 568 | 0. Client interacts w/TC | | 569 | over HTTPS, establishes Ks1 & Kp1, TBID1 | 570 | ETBMSG[[{EKM1}Ks1,TBID1,provided_token_binding]] | 571 |------------------------------>| | 572 | | | 573 | | | 574 | | | 575 | 1a. OIDC ID Token request, aka| | 576 | "Authentication Request", conveyed with | 577 | HTTP response header field of: | 578 | Include-Referred-Token-Binding-ID:true | 579 | any security-relevant cookies | | 580 | should contain TBID1 | | 581 +<- - - - - - - - - - - - - - - - | | 582 . | (redirect to TP via 301, 302, | | 583 . | 303, 307, or 308) | | 584 . | | | 585 +------------------------------------------------------->| 586 | 1b. opens HTTPS w/TP, | 587 | establishes Ks2, Kp2, TBID2; | 588 | sends GET or POST with | 589 | ETBMSG[[{EKM2}Ks2,TBID2,provided_token_binding], | 590 | [{EKM2}Ks1,TBID1,referred_token_binding]] | 591 | as well as the ID Token request | 592 | | | 593 | | | 594 | | | 595 | 2. user authentication (if applicable, | 596 | methods vary, particulars are out of scope) | 597 |<====================================================>| 598 | (TP generates ID Token for TC containing TBID1, may | 599 | also set cookie(s) containing TBID2 and/or TBID1, | 600 | details vary, particulars are out of scope) | 601 | | | 602 | | | 603 | | | 604 | 3a. ID Token containing Kp1, issued for TC, | 605 | conveyed via OIDC "Authentication Response" | 606 +<- - - - - - - - - - - - - - - - - - - - - - - - - - - -| 607 . | (redirect to TC) | | 608 . | | | 609 . | | | 610 +-------------------------------->| | 611 | 3b. HTTPS GET or POST with | 612 | ETBMSG[[{EKM1}Ks1,TBID1,provided_token_binding]] | 613 | conveying Authn Response containing | 614 | ID Token w/TBID1, issued for TC | 615 | | | 616 | | | 617 | | | 618 | 4. user is signed-on, any security-relevant cookie(s)| 619 | that are set SHOULD contain TBID1 | 620 |<------------------------------| | 621 | | | 622 | | | 624 6. Implementation Considerations 626 HTTPS-based applications may have multi-party use cases other than, 627 or in addition to, the HTTP redirect-based signaling-and-conveyance 628 of referred token bindings, as presented above in Section 5.3. 630 Thus, generic Token Binding implementations intended to support any 631 HTTPS-based client-side application (e.g., so-called "native 632 applications"), should provide means for applications to have Token 633 Binding messages, containing Token Binding IDs of various 634 application-specified Token Binding types and for application- 635 specified TLS connections, conveyed over an application-specified 636 HTTPS connection, i.e., within the TokenBindingMessage conveyed by 637 the Sec-Token-Binding header field. 639 However, such implementations MUST only convey Token Binding IDs to 640 servers if signaled to do so by an application. For example, a 641 server can return an Include-Referred-Token-Binding-ID HTTP response 642 header field to a Web browser, thus signaling to the Token Binding 643 implementation in the Web browser that the server intends to convey 644 the Web browser's Token Binding ID to another server. Other 645 signaling mechanisms are possible, and are specific to the 646 application layer protocol, but are outside the scope of this 647 specification. 649 NOTE: See Section 8 ("Privacy Considerations"), for privacy guidance 650 regarding the use of this functionality. 652 7. Security Considerations 654 7.1. Security Token Replay 656 The goal of the Federated Token Binding mechanisms is to prevent 657 attackers from exporting and replaying tokens used in protocols 658 between the client and Token Consumer, thereby impersonating 659 legitimate users and gaining access to protected resources. Although 660 bound tokens can still be replayed by any malware present in clients 661 (which may be undetectable by a server), in order to export bound 662 tokens to other machines and successfully replay them, attackers also 663 need to export the corresponding Token Binding private keys. Token 664 Binding private keys are therefore high-value assets and SHOULD be 665 strongly protected, ideally by generating them in a hardware security 666 module that prevents key export. 668 7.2. Triple Handshake Vulnerability in TLS 1.2 and Older TLS Versions 670 The Token Binding protocol relies on the exported key material (EKM) 671 value [RFC5705] to associate a TLS connection with a TLS Token 672 Binding. The triple handshake attack [TRIPLE-HS] is a known 673 vulnerability in TLS 1.2 and older TLS versions, allowing the 674 attacker to synchronize keying material between TLS connections. The 675 attacker can then successfully replay bound tokens. For this reason, 676 the Token Binding protocol MUST NOT be negotiated with these TLS 677 versions, unless the Extended Master Secret [RFC7627] and 678 Renegotiation Indication [RFC5746] TLS extensions have also been 679 negotiated. 681 7.3. Sensitivity of the Sec-Token-Binding Header 683 The purpose of the Token Binding protocol is to convince the server 684 that the client that initiated the TLS connection controls a certain 685 key pair. For the server to correctly draw this conclusion after 686 processing the Sec-Token-Binding header field, certain secrecy and 687 integrity requirements must be met. 689 For example, the client's Token Binding private key must be kept 690 secret by the client. If the private key is not secret, then another 691 actor in the system could create a valid Token Binding header field, 692 impersonating the client. This can render the main purpose of the 693 protocol - to bind bearer tokens to certain clients - moot. 694 Consider, for example, an attacker who obtained (perhaps through a 695 network intrusion) an authentication cookie that a client uses with a 696 certain server. Consider further that the server bound that cookie 697 to the client's Token Binding ID precisely to thwart misuse of the 698 cookie. If the attacker were to come into possession of the client's 699 private key, he could then establish a TLS connection with the server 700 and craft a Sec-Token-Binding header field that matches the binding 701 present in the cookie, thus successfully authenticating as the 702 client, and gaining access to the client's data at the server. The 703 Token Binding protocol, in this case, did not successfully bind the 704 cookie to the client. 706 Likewise, we need integrity protection of the Sec-Token-Binding 707 header field. A client should not be tricked into sending a Sec- 708 Token-Binding header field to a server that contains Token Binding 709 messages about key pairs that the client does not control. Consider 710 an attacker A that somehow has knowledge of the exported keying 711 material (EKM) for a TLS connection between a client C and a server 712 S. (While that is somewhat unlikely, it is also not entirely out of 713 the question, since the client might not treat the EKM as a secret - 714 after all, a pre-image-resistant hash function has been applied to 715 the TLS master secret, making it impossible for someone knowing the 716 EKM to recover the TLS master secret. Such considerations might lead 717 some clients to not treat the EKM as a secret.) Such an attacker A 718 could craft a Sec-Token-Binding header field with A's key pair over 719 C's EKM. If the attacker could now trick C into sending such a 720 header field to S, it would appear to S as if C controls a certain 721 key pair, when in fact it does not (the attacker A controls the key 722 pair). 724 If A has a pre-existing relationship with S (perhaps has an account 725 on S), it now appears to the server S as if A is connecting to it, 726 even though it is really C. (If the server S does not simply use 727 Token Binding IDs to identify clients, but also uses bound 728 authentication cookies, then A would also have to trick C into 729 sending one of A's cookies to S, which it can do through a variety of 730 means - inserting cookies through Javascript APIs, setting cookies 731 through related-domain attacks, etc.) In other words, A tricked C 732 into logging into A's account on S. This could lead to a loss of 733 privacy for C, since A presumably has some other way to also access 734 the account, and can thus indirectly observe A's behavior (for 735 example, if S has a feature that lets account holders see their 736 activity history on S). 738 Therefore, we need to protect the integrity of the Sec-Token-Binding 739 header field. One eTLD+1 should not be able to set the Sec-Token- 740 Binding header field (through a DOM API or otherwise) that the User 741 Agent uses with another eTLD+1. Employing the "Sec-" header field 742 prefix helps to meet this requirement by denoting the header field 743 name to be a "forbidden header name", see [fetch-spec]. 745 7.4. Securing Federated Sign-On Protocols 747 As explained above, in a federated sign-in scenario, a client will 748 prove possession of two different Token Binding private keys to a 749 Token Provider: One private key corresponds to the "provided" Token 750 Binding ID (which the client normally uses with the Token Provider), 751 and the other is the Token Binding private key corresponding to the 752 "referred" Token Binding ID (which the client normally uses with the 753 Token Consumer). The Token Provider is expected to issue a token 754 that is bound to the referred Token Binding ID. 756 Both proofs (that of the provided Token Binding private key and that 757 of the referred Token Binding private key) are necessary. To show 758 this, consider the following scenario: 760 o The client has an authentication token with the Token Provider 761 that is bound to the client's Token Binding ID used with that 762 Token Provider. 764 o The client wants to establish a secure (i.e., free of men-in-the- 765 middle) authenticated session with the Token Consumer, but has not 766 done so yet (in other words, we are about to run the federated 767 sign-on protocol). 769 o A man-in-the-middle is allowed to intercept the connection between 770 client and Token Consumer or between Client and Token Provider (or 771 both). 773 The goal is to detect the presence of the man-in-the-middle in these 774 scenarios. 776 First, consider a man-in-the-middle between the client and the Token 777 Provider. Recall that we assume that the client possesses a bound 778 authentication token (e.g., cookie) for the Token Provider. The man- 779 in-the-middle can intercept and modify any message sent by the client 780 to the Token Provider, and any message sent by the Token Provider to 781 the client. (This means, among other things, that the man-in-the- 782 middle controls the Javascript running at the client in the origin of 783 the Token Provider.) It is not, however, in possession of the 784 client's Token Binding private key. Therefore, it can either choose 785 to replace the Token Binding ID in requests from the client to the 786 Token Provider, and create a Sec-Token-Binding header field that 787 matches the TLS connection between the man-in-the-middle and the 788 Token Provider, or it can choose to leave the Sec-Token-Binding 789 header field unchanged. If it chooses the latter, the signature in 790 the Token Binding message (created by the original client on the 791 exported keying material (EKM) for the connection between client and 792 man-in-the-middle) will not match a signature on the EKM between man- 793 in-the-middle and the Token Provider. If it chooses the former (and 794 creates its own signature, using its own Token Binding private key, 795 over the EKM for the connection between itself, the man-in-the- 796 middle, and Token Provider), then the Token Binding message will 797 match the connection between man-in-the-middle and Token Provider, 798 but the Token Binding ID in the message will not match the Token 799 Binding ID that the client's authentication token is bound to. 800 Either way, the man-in-the-middle is detected by the Token Provider, 801 but only if the proof of possession of the provided Token Binding 802 private key is required in the protocol (as is done above). 804 Next, consider the presence of a man-in-the-middle between client and 805 Token Consumer. That man-in-the-middle can intercept and modify any 806 message sent by the client to the Token Consumer and any message sent 807 by the Token Consumer to the client. The Token Consumer is the party 808 that redirects the client to the Token Provider. In this case, the 809 man-in-the-middle controls the redirect URL and can tamper with any 810 redirect URL issued by the Token Consumer (as well as with any 811 Javascript running in the origin of the Token Consumer). The goal of 812 the man-in-the-middle is to trick the Token Provider into issuing a 813 token bound to its Token Binding ID, not to the Token Binding ID of 814 the legitimate client. To thwart this goal of the man-in-the-middle, 815 the client's referred Token Binding ID must be communicated to the 816 Token Producer in a manner that cannot be affected by the man-in-the- 817 middle (who, as we recall, can modify redirect URLs and Javascript at 818 the client). Including the referred Token Binding structure in the 819 Sec-Token-Binding header field (as opposed to, say, including the 820 referred Token Binding ID in an application-level message as part of 821 the redirect URL) is one way to assure that the man-in-the-middle 822 between client and Token Consumer cannot affect the communication of 823 the referred Token Binding ID to the Token Provider. 825 Therefore, the Sec-Token-Binding header field in the federated sign- 826 on use case contains both: a proof of possession of the provided 827 Token Binding key, as well as a proof of possession of the referred 828 Token Binding key. 830 Note that the presence of Token Binding does not relieve the Token 831 Provider and Token Consumer from performing various checks to ensure 832 the security of clients during federated sign-on protocols. These 833 include the following: 835 o The Token Provider should not issue tokens to Token Consumers that 836 have been shown to act maliciously. To aid in this, the 837 federation protocol should identify the Token Consumer to the 838 Token Provider (e.g., through OAuth client IDs or similar 839 mechanisms), and the Token Provider should ensure that tokens are 840 indeed issued to the Token Consumer identified in the token 841 request (e.g., by verifying that the redirect URI is associated 842 with the OAuth client ID.) 844 o The Token Consumer should verify that the tokens were issued for 845 it, and not some other token consumer. To aid in this, the 846 federation protocol should include an audience parameter in the 847 token response, or apply equivalent mechanisms (the implicit OAuth 848 flow requires Token Consumers to identify themselves when they 849 exchange OAuth authorization codes for OAuth refresh tokens, 850 leaving it up to the Token Provider to verify that the OAuth 851 authorization was delivered to the correct Token Consumer). 853 8. Privacy Considerations 855 8.1. Scoping of Token Binding Key Pairs 857 Clients use different Token Binding key pairs for different servers, 858 so as to not allow Token Binding to become a tracking tool across 859 different servers. However, the scoping of the Token Binding key 860 pairs to servers varies according to the scoping rules of the 861 application protocol (Section 4.1 of [I-D.ietf-tokbind-protocol]). 863 In the case of HTTP cookies, servers may use Token Binding to secure 864 their cookies. These cookies can be attached to any sub-domain of 865 effective top-level domains (eTLDs), and clients therefore should use 866 the same Token Binding key pair across such subdomains. This will 867 ensure that any server capable of receiving the cookie will see the 868 same Token Binding ID from the client, and thus be able to verify the 869 token binding of the cookie. See Section 2.1, above. 871 If the client application is not a Web browser, it may have 872 additional knowledge about the relationship between different 873 servers. For example, the client application might be aware of the 874 fact that two servers play the role of Relying Party and Identity 875 Provider in a federated sign-on protocol, and that they therefore 876 share the identity of the user. In such cases, it is permissible to 877 use different Token Binding key pair scoping rules, such as using the 878 same Token Binding key pair for both the Relying Party and the 879 Identity Provider. Absent such special knowledge, conservative key- 880 scoping rules should be used, assuring that clients use different 881 Token Binding key pairs with different servers. 883 8.2. Lifetime of Token Binding Key Pairs 885 Token Binding key pairs do not have an expiration time. This means 886 that they can potentially be used by a server to track a user for an 887 extended period of time (similar to a long-lived cookie). HTTPS 888 clients such as Web user agents should therefore provide a user 889 interface for discarding Token Binding key pairs (similar to the 890 affordances provided to delete cookies). 892 If a user agent provides modes such as private browsing mode in which 893 the user is promised that browsing state such as cookies are 894 discarded after the session is over, the user agent should also 895 discard Token Binding key pairs from such modes after the session is 896 over. Generally speaking, users should be given the same level of 897 control over lifetime of Token Binding key pairs as they have over 898 cookies or other potential tracking mechanisms. 900 8.3. Correlation 902 An application's various communicating endpoints that receive Token 903 Binding IDs for TLS connections other than their own, obtain 904 information about the application's other TLS connections. (In this 905 context, "an application" is a combination of client-side and server- 906 side components, communicating over HTTPS, where the client side may 907 be either or both Web browser-based or native application-based.) 908 These other Token Binding IDs can serve as correlation handles for 909 the endpoints of the other connections. If the receiving endpoints 910 are otherwise aware of these other connections, then no additional 911 information is being exposed. For instance, if in a redirect-based 912 federation protocol, the Identity Provider and Relying Party already 913 possess URLs for one another, also having Token Binding IDs for these 914 connections does not provide additional correlation information. If 915 not, then, by providing the other Token Binding IDs, additional 916 information is exposed that can be used to correlate the other 917 endpoints. In such cases, a privacy analysis of enabled correlations 918 and their potential privacy impacts should be performed as part of 919 the application design decisions of how, and whether, to utilize 920 Token Binding. 922 Also, Token Binding implementations must take care to only reveal 923 Token Binding IDs to other endpoints if the application associated 924 with a Token Binding ID signals to do so, see Section 6 925 ("Implementation Considerations"). 927 Finally, care should be taken to ensure that unrelated applications 928 do not obtain information about each other's Token Bindings. For 929 instance, a Token Binding implementation shared between multiple 930 applications on a given system should prevent unrelated applications 931 from obtaining each other's Token Binding information. This may be 932 accomplished by using techniques such as application isolation and 933 key segregation, depending upon system capabilities. 935 9. IANA Considerations 937 Below are the Internet Assigned Numbers Authority (IANA) Permanent 938 Message Header Field registration information per [RFC3864]. 940 Header field name: Sec-Token-Binding 941 Applicable protocol: HTTP 942 Status: standard 943 Author/Change controller: IETF 944 Specification document(s): this one 946 Header field name: Include-Referred-Token-Binding-ID 947 Applicable protocol: HTTP 948 Status: standard 949 Author/Change controller: IETF 950 Specification document(s): this one 952 10. Acknowledgements 954 This document incorporates comments and suggestions offered by Eric 955 Rescorla, Gabriel Montenegro, Martin Thomson, Vinod Anupam, Anthony 956 Nadalin, Michael B. Jones, Bill Cox, Brian Campbell, and others. 958 This document was produced under the chairmanship of John Bradley and 959 Leif Johansson. The area directors included Eric Rescorla, Kathleen 960 Moriarty and Stephen Farrell. 962 11. References 964 11.1. Normative References 966 [fetch-spec] 967 WhatWG, "Fetch", Living Standard , 968 . 970 [I-D.ietf-tokbind-negotiation] 971 Popov, A., Nystrom, M., Balfanz, D., and A. Langley, 972 "Transport Layer Security (TLS) Extension for Token 973 Binding Protocol Negotiation", draft-ietf-tokbind- 974 negotiation-10 (work in progress), October 2017. 976 [I-D.ietf-tokbind-protocol] 977 Popov, A., Nystrom, M., Balfanz, D., Langley, A., and J. 978 Hodges, "The Token Binding Protocol Version 1.0", draft- 979 ietf-tokbind-protocol-16 (work in progress), October 2017. 981 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 982 Requirement Levels", BCP 14, RFC 2119, 983 DOI 10.17487/RFC2119, March 1997, 984 . 986 [RFC2818] Rescorla, E., "HTTP Over TLS", RFC 2818, 987 DOI 10.17487/RFC2818, May 2000, 988 . 990 [RFC3864] Klyne, G., Nottingham, M., and J. Mogul, "Registration 991 Procedures for Message Header Fields", BCP 90, RFC 3864, 992 DOI 10.17487/RFC3864, September 2004, 993 . 995 [RFC4648] Josefsson, S., "The Base16, Base32, and Base64 Data 996 Encodings", RFC 4648, DOI 10.17487/RFC4648, October 2006, 997 . 999 [RFC5246] Dierks, T. and E. Rescorla, "The Transport Layer Security 1000 (TLS) Protocol Version 1.2", RFC 5246, 1001 DOI 10.17487/RFC5246, August 2008, 1002 . 1004 [RFC5705] Rescorla, E., "Keying Material Exporters for Transport 1005 Layer Security (TLS)", RFC 5705, DOI 10.17487/RFC5705, 1006 March 2010, . 1008 [RFC6265] Barth, A., "HTTP State Management Mechanism", RFC 6265, 1009 DOI 10.17487/RFC6265, April 2011, 1010 . 1012 [RFC7230] Fielding, R., Ed. and J. Reschke, Ed., "Hypertext Transfer 1013 Protocol (HTTP/1.1): Message Syntax and Routing", 1014 RFC 7230, DOI 10.17487/RFC7230, June 2014, 1015 . 1017 [RFC7231] Fielding, R., Ed. and J. Reschke, Ed., "Hypertext Transfer 1018 Protocol (HTTP/1.1): Semantics and Content", RFC 7231, 1019 DOI 10.17487/RFC7231, June 2014, 1020 . 1022 [RFC7541] Peon, R. and H. Ruellan, "HPACK: Header Compression for 1023 HTTP/2", RFC 7541, DOI 10.17487/RFC7541, May 2015, 1024 . 1026 11.2. Informative References 1028 [OpenID.Core] 1029 Sakimura, N., Bradley, J., Jones, M., de Medeiros, B., and 1030 C. Mortimore, "OpenID Connect Core 1.0", August 2015, 1031 . 1033 [RFC5746] Rescorla, E., Ray, M., Dispensa, S., and N. Oskov, 1034 "Transport Layer Security (TLS) Renegotiation Indication 1035 Extension", RFC 5746, DOI 10.17487/RFC5746, February 2010, 1036 . 1038 [RFC6749] Hardt, D., Ed., "The OAuth 2.0 Authorization Framework", 1039 RFC 6749, DOI 10.17487/RFC6749, October 2012, 1040 . 1042 [RFC7540] Belshe, M., Peon, R., and M. Thomson, Ed., "Hypertext 1043 Transfer Protocol Version 2 (HTTP/2)", RFC 7540, 1044 DOI 10.17487/RFC7540, May 2015, 1045 . 1047 [RFC7627] Bhargavan, K., Ed., Delignat-Lavaud, A., Pironti, A., 1048 Langley, A., and M. Ray, "Transport Layer Security (TLS) 1049 Session Hash and Extended Master Secret Extension", 1050 RFC 7627, DOI 10.17487/RFC7627, September 2015, 1051 . 1053 [TRIPLE-HS] 1054 Bhargavan, K., Delignat-Lavaud, A., Fournet, C., Pironti, 1055 A., and P. Strub, "Triple Handshakes and Cookie Cutters: 1056 Breaking and Fixing Authentication over TLS. IEEE 1057 Symposium on Security and Privacy", 2014. 1059 Authors' Addresses 1061 Andrei Popov 1062 Microsoft Corp. 1063 USA 1065 Email: andreipo@microsoft.com 1067 Magnus Nystroem 1068 Microsoft Corp. 1069 USA 1071 Email: mnystrom@microsoft.com 1073 Dirk Balfanz (editor) 1074 Google Inc. 1075 USA 1077 Email: balfanz@google.com 1079 Adam Langley 1080 Google Inc. 1081 USA 1083 Email: agl@google.com 1085 Nick Harper 1086 Google Inc. 1087 USA 1089 Email: nharper@google.com 1090 Jeff Hodges 1091 PayPal 1092 USA 1094 Email: Jeff.Hodges@paypal.com