idnits 2.17.1 draft-ietf-uta-email-deep-07.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 : ---------------------------------------------------------------------------- -- The draft header indicates that this document updates RFC5068, but the abstract doesn't seem to mention this, which it should. -- The draft header indicates that this document updates RFC2595, but the abstract doesn't seem to mention this, which it should. -- The draft header indicates that this document updates RFC3501, but the abstract doesn't seem to mention this, which it should. -- The draft header indicates that this document updates RFC1939, but the abstract doesn't seem to mention this, which it should. -- The draft header indicates that this document updates RFC3464, but the abstract doesn't seem to mention this, which it should. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year (Using the creation date from RFC1939, updated by this document, for RFC5378 checks: 1995-05-15) -- The document seems to lack a disclaimer for pre-RFC5378 work, but may have content which was first submitted before 10 November 2008. If you have contacted all the original authors and they are all willing to grant the BCP78 rights to the IETF Trust, then this is fine, and you can ignore this comment. If not, you may need to add the pre-RFC5378 disclaimer. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (July 3, 2017) is 2489 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) ** Obsolete normative reference: RFC 3501 (Obsoleted by RFC 9051) ** Obsolete normative reference: RFC 5246 (Obsoleted by RFC 8446) ** Obsolete normative reference: RFC 7525 (Obsoleted by RFC 9325) -- Obsolete informational reference (is this intentional?): RFC 6125 (Obsoleted by RFC 9525) Summary: 3 errors (**), 0 flaws (~~), 1 warning (==), 8 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group K. Moore 3 Internet-Draft Network Heretics 4 Updates: 1939, 2595, 3464, 3501, 5068, C. Newman 5 6186, 6409 (if approved) Oracle 6 Intended status: Standards Track July 3, 2017 7 Expires: January 4, 2018 9 Cleartext Considered Obsolete: Best Current Practices for Use of TLS for 10 Email Submission and Access 11 draft-ietf-uta-email-deep-07 13 Abstract 15 This specification outlines best current practices for use of 16 Transport Layer Security (TLS) to provide confidentiality of email 17 traffic between a mail user agent (MUA) and a mail submission or mail 18 access server. 20 Status of This Memo 22 This Internet-Draft is submitted in full conformance with the 23 provisions of BCP 78 and BCP 79. 25 Internet-Drafts are working documents of the Internet Engineering 26 Task Force (IETF). Note that other groups may also distribute 27 working documents as Internet-Drafts. The list of current Internet- 28 Drafts is at http://datatracker.ietf.org/drafts/current/. 30 Internet-Drafts are draft documents valid for a maximum of six months 31 and may be updated, replaced, or obsoleted by other documents at any 32 time. It is inappropriate to use Internet-Drafts as reference 33 material or to cite them other than as "work in progress." 35 This Internet-Draft will expire on January 4, 2018. 37 Copyright Notice 39 Copyright (c) 2017 IETF Trust and the persons identified as the 40 document authors. All rights reserved. 42 This document is subject to BCP 78 and the IETF Trust's Legal 43 Provisions Relating to IETF Documents 44 (http://trustee.ietf.org/license-info) in effect on the date of 45 publication of this document. Please review these documents 46 carefully, as they describe your rights and restrictions with respect 47 to this document. Code Components extracted from this document must 48 include Simplified BSD License text as described in Section 4.e of 49 the Trust Legal Provisions and are provided without warranty as 50 described in the Simplified BSD License. 52 Table of Contents 54 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 55 2. Conventions and Terminology Used in This Document . . . . . . 3 56 3. Implicit TLS . . . . . . . . . . . . . . . . . . . . . . . . 4 57 3.1. Implicit TLS for POP . . . . . . . . . . . . . . . . . . 5 58 3.2. Implicit TLS for IMAP . . . . . . . . . . . . . . . . . . 5 59 3.3. Implicit TLS for SMTP Submission . . . . . . . . . . . . 5 60 3.4. Implicit TLS Connection Closure for POP, IMAP and SMTP 61 Submission . . . . . . . . . . . . . . . . . . . . . . . 6 62 4. Best Current Practices for Use of TLS by Mail Access 63 Services and Message Submission Services . . . . . . . . . . 6 64 4.1. Deprecation of Services Using Cleartext and TLS Versions 65 < 1.1 . . . . . . . . . . . . . . . . . . . . . . . . . . 7 66 4.2. Mail Server Use of Client Certificate Authentication . . 8 67 4.3. Recording TLS Cipher Suite in Received Header . . . . . . 9 68 4.4. TLS Server Certificate Requirements . . . . . . . . . . . 9 69 4.5. Recommended DNS records for mail protocol servers . . . . 9 70 4.5.1. MX records . . . . . . . . . . . . . . . . . . . . . 10 71 4.5.2. SRV records . . . . . . . . . . . . . . . . . . . . . 10 72 4.5.3. DNSSEC . . . . . . . . . . . . . . . . . . . . . . . 10 73 4.5.4. TLSA records . . . . . . . . . . . . . . . . . . . . 10 74 4.6. Changes to Internet Facing Servers . . . . . . . . . . . 10 75 5. Best Current Practices for use of TLS by Mail User Agents . . 10 76 5.1. Use of SRV records in Establishing Configuration . . . . 12 77 5.2. Minimum Confidentiality Level . . . . . . . . . . . . . . 13 78 5.3. Certificiate Validation . . . . . . . . . . . . . . . . . 14 79 5.4. Certificate Pinning . . . . . . . . . . . . . . . . . . . 14 80 5.5. Client Certificate Authentication . . . . . . . . . . . . 14 81 6. Considerations related to Anti-Virus/Anti-Spam Software and 82 Services . . . . . . . . . . . . . . . . . . . . . . . . . . 15 83 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 15 84 7.1. POP3S Port Registration Update . . . . . . . . . . . . . 15 85 7.2. IMAPS Port Registration Update . . . . . . . . . . . . . 16 86 7.3. Submissions Port Registration . . . . . . . . . . . . . . 16 87 8. Security Considerations . . . . . . . . . . . . . . . . . . . 17 88 9. References . . . . . . . . . . . . . . . . . . . . . . . . . 17 89 9.1. Normative References . . . . . . . . . . . . . . . . . . 17 90 9.2. Informative References . . . . . . . . . . . . . . . . . 19 91 Appendix A. Design Considerations . . . . . . . . . . . . . . . 20 92 Appendix B. Change Log . . . . . . . . . . . . . . . . . . . . . 21 93 Appendix C. Acknowledgements . . . . . . . . . . . . . . . . . . 26 94 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 26 96 1. Introduction 98 Software that provides email service via Internet Message Access 99 Protocol (IMAP) [RFC3501], Post Office Protocol (POP) [RFC1939] and/ 100 or Simple Mail Transfer Protocol (SMTP) Submission [RFC6409] usually 101 has Transport Layer Security (TLS) [RFC5246] support but often does 102 not use it in a way that maximizes end-user confidentiality. This 103 specification describes best current practices for use of TLS in 104 interactions between Mail User Agents and Mail Access Services, and 105 between Mail User Agents and Mail Submission Services. 107 In brief, this memo now recommends that: 109 o TLS version 1.1 or greater be used for all traffic between mail 110 user agents (MUAs) and mail submission servers, and also between 111 MUAs and mail access servers. 113 o MUAs and mail service providers discourage use of cleartext 114 protocols for mail access and mail submission, and deprecate use 115 of cleartext protocols for these purposes as soon as practicable. 117 o Use of "Implicit TLS" on ports reserved for that purpose, in 118 preference to STARTTLS on a port that otherwise supports 119 cleartext. 121 This memo does not address use of TLS with SMTP for message relay 122 (where Message Submission [RFC6409] does not apply). Improved use of 123 TLS with SMTP for message relay requires a different approach. One 124 approach to address that topic is described in [RFC7672]. 126 The recommendations in this memo do not replace the functionality of, 127 and are not intended as a substitute for, end-to-end encryption of 128 electronic mail. 130 2. Conventions and Terminology Used in This Document 132 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 133 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 134 document are to be interpreted as described in [RFC2119]. 136 The term "Implicit TLS" refers to the automatic negotiation of TLS 137 whenever a TCP connection is made on a particular TCP port that is 138 used exclusively by that server for TLS connections. The term 139 "Implicit TLS" is intended to contrast with use of STARTTLS and 140 similar commands in POP, IMAP, SMTP message submission, and other 141 protocols, that are used by client and server to explicitly negotiate 142 TLS on an established cleartext TCP connection. 144 The term "Mail Access Services" includes POP, IMAP and any other 145 protocol used to access or modify received messages, or to access or 146 modify a mail user's account configuration. 148 "Mail Submission Service" refers to the use of the protocol specified 149 in [RFC6409] (or one of its predecessors or successors) for 150 submission of outgoing messages for delivery to recipients. 152 The term "Mail Service Provider" (MSP) refers to a provider of Mail 153 Access Services and/or Mail Submission Services. 155 The term "Mail Account" refers to a user's identity with a Mail 156 Service Provider, that user's authentication credentials, any user 157 email that is stored by the MSP, and any other per-user configuration 158 information maintained by the MSP (for example, spam filtering 159 instructions). Most Mail User Agents (MUAs) support the ability to 160 access multiple Mail Accounts. 162 For each account that an MUA accesses on its user's behalf, it must 163 have the server names, ports, authentication credentials, and other 164 configuration information specified by the user. This information 165 which is used by the MUA is referred to as "Mail Account 166 Configuration" 168 This specification expresses syntax using the Augmented Backus-Naur 169 Form (ABNF) as described in [RFC5234], including the core rules in 170 Appendix B and rules from [RFC5322]. 172 3. Implicit TLS 174 Previous standards for use of email protocols with TLS used the 175 STARTTLS mechanism: [RFC2595], [RFC3207], and [RFC3501]. With 176 STARTTLS, the client establishes a cleartext application session and 177 determines whether to issue a STARTTLS command based on server 178 capabilities and client configuration. If the client issues a 179 STARTTLS command, a TLS handshake follows that can upgrade the 180 connection. While this mechanism has been deployed, an alternate 181 mechanism where TLS is negotiated immediately at connection start on 182 a separate port (referred to in this document as "Implicit TLS") has 183 been deployed more successfully. To encourage more widespread use of 184 TLS, and to encourage a greater consistency for how TLS is used, this 185 specification now recommends use of Implicit TLS for POP, IMAP, SMTP 186 Submission, and all other protocols used between a Mail User Agent 187 and a mail service. 189 3.1. Implicit TLS for POP 191 When a TCP connection is established for the "pop3s" service (default 192 port 995), a TLS handshake begins immediately. Clients MUST 193 implement the certificate validation mechanism described in 194 [RFC7817]. Once the TLS session is established, POP3 [RFC1939] 195 protocol messages are exchanged as TLS application data for the 196 remainder of the TCP connection. After the server sends a +OK 197 greeting, the server and client MUST enter AUTHORIZATION state, even 198 if client credentials were supplied during the TLS handshake. 200 See Section 5.5 and Section 4.2 for additional information on client 201 certificate authentication. See Section 7.1 for port registration 202 information. 204 3.2. Implicit TLS for IMAP 206 When a TCP connection is established for the "imaps" service (default 207 port 993), a TLS handshake begins immediately. Clients MUST 208 implement the certificate validation mechanism described in [RFC3501] 209 and SHOULD implement the certificate validation mechanism described 210 in [RFC7817]. Once the TLS session is established, IMAP [RFC3501] 211 protocol messages are exchanged as TLS application data for the 212 remainder of the TCP connection. If client credentials were provided 213 during the TLS handshake that the server finds acceptable, the server 214 MAY issue a PREAUTH greeting in which case both the server and client 215 enter AUTHENTICATED state. If the server issues an OK greeting then 216 both server and client enter NOT AUTHENTICATED state. 218 See Section 5.5 and Section 4.2 for additional information on client 219 certificate authentication. See Section 7.1 for port registration 220 information. See Section 7.2 for port registration information. 222 3.3. Implicit TLS for SMTP Submission 224 When a TCP connection is established for the "submissions" service 225 (default port 465), a TLS handshake begins immediately. Clients MUST 226 implement the certificate validation mechanism described in 227 [RFC7817]. Once a TLS session is established, message submission 228 protocol data [RFC6409] is exchanged as TLS application data for the 229 remainder of the TCP connection. (Note: the "submissions" service 230 name is defined in section 10.3 of this document, and follows the 231 usual convention that the name of a service layered on top of 232 Implicit TLS consists of the name of the service as used without TLS, 233 with an "s" appended.) 235 The STARTTLS mechanism on port 587 is relatively widely deployed due 236 to the situation with port 465 (discussed in Section 7.3). This 237 differs from IMAP and POP services where Implicit TLS is more widely 238 deployed on servers than STARTTLS. It is desirable to migrate core 239 protocols used by MUA software to Implicit TLS over time for 240 consistency as well as the additional reasons discussed in 241 Appendix A. However, to maximize use of encryption for submission it 242 is desirable to support both mechanisms for Message Submission over 243 TLS for a transition period of several years. As a result, clients 244 and servers SHOULD implement both STARTTLS on port 587 and Implicit 245 TLS on port 465 for this transition period. Note that there is no 246 significant difference between the security properties of STARTTLS on 247 port 587 and Implicit TLS on port 465 if the implementations are 248 correct and both client and server are configured to require 249 successful negotiation of TLS prior to message submission. 251 Note that the "submissions" port provides access to a Mail Submission 252 Agent (MSA) as defined in [RFC6409] so requirements and 253 recommendations for MSAs in that document apply to the submissions 254 port, including the requirement to implement SMTP AUTH [RFC4954]. 256 See Section 5.5 and Section 4.2 for additional information on client 257 certificate authentication. See Section 7.3 for port registration 258 information. 260 3.4. Implicit TLS Connection Closure for POP, IMAP and SMTP Submission 262 When a client or server wishes to close the connection, it SHOULD 263 initiate the exchange of TLS close alerts before TCP connection 264 termination. The client MAY, after sending a TLS close alert, 265 gracefully close the TCP connection without waiting for a TLS 266 response from the server. 268 4. Best Current Practices for Use of TLS by Mail Access Services and 269 Message Submission Services 271 The following practices are recommended for Mail Access Services and 272 Mail Submission Services: 274 o Mail Service Providers (MSPs) which support POP, IMAP, and/or 275 Message Submission, SHOULD provide and support instances of those 276 services which use Implicit TLS. (See Section 3.) 278 o For compatibility with existing MUAs and existing MUA 279 configurations, MSPs SHOULD also, in the near term, provide 280 instances of these services which support STARTTLS. This will 281 permit legacy MUAs to discover new availability of TLS capability 282 on servers, and may increase use of TLS by such MUAs. However, 283 servers SHOULD NOT advertise STARTTLS if use of the STARTTLS 284 command by a client is likely to fail (for example, if the server 285 has no server certificate configured.) 287 o MSPs SHOULD advertise their Mail Access Services and Mail 288 Submission Services using DNS SRV records according to [RFC6186]. 289 Services supporting TLS SHOULD be advertised in preference to 290 cleartext services (if offered). In addition, services using 291 Implicit TLS SHOULD be advertised in preference to services 292 supporting STARTTLS (if offered). (See also Section 4.5.) 294 o MSPs SHOULD deprecate use of cleartext Mail Access Services and 295 Mail Submission Services as soon as practicable. (See 296 Section 4.1.) 298 o MSPs that provide mail submission as a service, SHOULD support 299 Mail Submission services using Implicit TLS. 301 o MSPs currently supporting such use of cleartext SMTP (on port 25) 302 as a means of message submission by their users (whether or not 303 requiring authentication) SHOULD transition their users to using 304 TLS (either Implicit TLS or STARTTLS) as soon as practicable. 306 o Mail services SHOULD support TLS 1.2 or later. 308 o All Mail services SHOULD implement the recommended TLS cipher 309 suites described in [RFC7525] or a future BCP or standards track 310 revision of that document. 312 o Mail services currently supporting SSL 2.x, SSL 3.0, or TLS 1.0 313 SHOULD transition their users to later versions of TLS, and 314 discontinue support for those versions of SSL and TLS, as soon as 315 practicable. 317 o Mail Submission Servers accepting mail using TLS SHOULD include 318 the TLS ciphersuite of the session in which the mail was received, 319 in the Received field of the outgoing message. (See Section 4.3.) 321 o All Mail services implementing TLS SHOULD log TLS cipher 322 information along with any connection or authentication logs that 323 they maintain. 325 Additional considerations and details appear below. 327 4.1. Deprecation of Services Using Cleartext and TLS Versions < 1.1 329 The specific means employed for deprecation of cleartext Mail Access 330 Services and Mail Submission Services this MAY vary from one MSP to 331 the next in light of their user communities' needs and constraints. 333 For example, an MSP MAY implement a gradual transition in which, over 334 time, more and more users are forbidden to authenticate to cleartext 335 instances of these services, thus encouraging those users to migrate 336 to Implicit TLS. Access to cleartext services should eventually be 337 either disabled, or limited strictly for use by legacy systems which 338 cannot be upgraded. 340 After a user's ability to authenticate to a service using cleartext 341 is revoked, the server denying such access MUST NOT provide any 342 indication of whether the user's authentication credentials were 343 valid. An attempt to authenticate as such a user using either 344 invalid credentials or valid credentials MUST both result in the same 345 indication of access being denied. 347 Also, users authenticating with passwords SHOULD be required to 348 change those passwords when migrating from cleartext to TLS, since 349 the old passwords were likely to have been compromised. 351 Transition of users from SSL or TLS 1.0 to later versions of TLS MAY 352 be accomplished by a means similar to that described above. There 353 are multiple ways to accomplish this. One way is for the server to 354 refuse a ClientHello message from any client sending a protocol 355 version number corresponding to any version of SSL or TLS 1.0. 356 Another way is for the server to accept ClientHello messages from 357 some client versions that it does not wish to support, but later 358 refuse to allow the user to authenticate. The latter method may 359 provide a better indication to the user of the reason for the failure 360 but (depending on the protocol and method of authentication used) may 361 also risk exposure of the user's password over an channel which is 362 known to not provide adequate confidentiality. 364 It is RECOMMENDED that new users be required to use TLS version 1.1 365 or greater from the start. However an MSP may find it necessary to 366 make exceptions to accommodate some legacy systems which support only 367 earlier versions of TLS, or only cleartext. 369 4.2. Mail Server Use of Client Certificate Authentication 371 Mail servers MAY implement client certificate authentication on the 372 Implicit TLS port. Servers MUST NOT request a client certificate 373 during the TLS handshake unless the server is configured to accept 374 some client certificates as sufficient for authentication and the 375 server has the ability to determine a mail server authorization 376 identity matching such certificates. How to make this determination 377 is presently implementation specific. 379 If the server accepts the client's certificate as sufficient for 380 authorization, it MUST enable the SASL EXTERNAL [RFC4422] mechanism. 382 An IMAPS server MAY issue a PREAUTH greeting instead of enabling SASL 383 EXTERNAL. 385 4.3. Recording TLS Cipher Suite in Received Header 387 The ESMTPS transmission type [RFC3848] provides trace information 388 that can indicate TLS was used when transferring mail. However, TLS 389 usage by itself is not a guarantee of confidentiality or security. 390 The TLS cipher suite provides additional information about the level 391 of security made available for a connection. This defines a new SMTP 392 "tls" Received header additional-registered-clause that is used to 393 record the TLS cipher suite that was negotiated for the connection. 394 The value included in this additional clause SHOULD be the registered 395 cipher suite name (e.g., TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) 396 included in the TLS cipher suite registry. In the event the 397 implementation does not know the name of the cipher suite (a 398 situation that should be remedied promptly), a four-digit hexadecimal 399 cipher suite identifier MAY be used. The ABNF for the field follows: 401 tls-cipher-clause = CFWS "tls" FWS tls-cipher 403 tls-cipher = tls-cipher-suite-name / tls-cipher-suite-hex 405 tls-cipher-name = ALPHA *(ALPHA / DIGIT / "_") 406 ; as registered in IANA cipher suite registry 408 tls-cipher-hex = "0x" 4HEXDIG 410 4.4. TLS Server Certificate Requirements 412 MSPs MUST maintain valid server certificates for all servers. See 413 [RFC7817] for the recommendations and requirements necessary to 414 achieve this. 416 If a protocol server provides service for more than one mail domain, 417 it MAY use a separate IP address for each domain and/or a server 418 certificate that advertises multiple domains. This will generally be 419 necessary unless and until it is acceptable to impose the constraint 420 that the server and all clients support the Server Name Indication 421 extension to TLS [RFC6066]. For more discussion of this problem, see 422 section 5.1 of [RFC7817]. 424 4.5. Recommended DNS records for mail protocol servers 426 This section discusses not only the DNS records that are recommended, 427 but also implications of DNS records for server configuration and TLS 428 server certificates. 430 4.5.1. MX records 432 It is recommended that MSPs advertise MX records for handling of 433 inbound mail (instead of relying entirely on A or AAAA records), and 434 that those MX records be signed using DNSSEC. This is mentioned here 435 only for completeness, as handling of inbound mail is out of scope 436 for this document. 438 4.5.2. SRV records 440 MSPs SHOULD advertise SRV records to aid MUAs in determination of 441 proper configuration of servers, per the instructions in [RFC6186]. 443 MSPs SHOULD advertise servers that support Implicit TLS in preference 444 to those which support cleartext and/or STARTTLS operation. 446 4.5.3. DNSSEC 448 All DNS records advertised by an MSP as a means of aiding clients in 449 communicating with the MSP's servers, SHOULD be signed using DNSSEC. 451 4.5.4. TLSA records 453 MSPs SHOULD advertise TLSA records to provide an additional trust 454 anchor for public keys used in TLS server certificates. However, 455 TLSA records MUST NOT be advertised unless they are signed using 456 DNSSEC. 458 4.6. Changes to Internet Facing Servers 460 When an MSP changes the Internet Facing Servers providing mail access 461 and mail submission services, including SMTP-based spam/virus 462 filters, it is generally necessary to support the same and/or a newer 463 version of TLS and the same security directives that were previously 464 advertised. 466 5. Best Current Practices for use of TLS by Mail User Agents 468 It is recommended that Mail User Agents implement the following 469 practices: 471 o MUAs SHOULD be capable of using DNS SRV records to discover Mail 472 Access Services and Mail Submission Services that are advertised 473 by a MSP for an account being configured. Other means of 474 discovering server configuration information (e.g. a database 475 maintained by the MUA vendor) MAY also be supported. (See 476 Section 5.1 for more information.) 478 o MUAs SHOULD be configurable to require a minimum level of 479 confidentiality for any particular Mail Account, and refuse to 480 exchange information via any service associated with that Mail 481 Account if the session does not provide that minimum level of 482 confidentiality. (See Section 5.2.) 484 o MUAs MUST NOT consider a session as meeting a minimum level of 485 confidentiality if the server's TLS certificate cannot be 486 validated. (See Section 5.3.) 488 o MUAs MAY impose other minimum confidentiality requirements in the 489 future, e.g. in order to discourage use of TLS versions or 490 cryptographic algorithms in which weaknesses have been discovered. 492 o MUAs SHOULD provide a prominent visual indication of the level of 493 confidentiality associated with an account configuration (for 494 example, indications such as "lock" icons or changed background 495 colors similar to those used by some browsers), at appropriate 496 times and locations in order to inform the user of the 497 confidentiality of the communications associated with that 498 account. For example, this might be done whenever (a) prompting 499 the user for authentication credentials, (b) the user is composing 500 mail that will be sent to a particular submission server, (c) a 501 list of accounts is displayed (particularly if the user can select 502 from that list to read mail), or (d) the user is requesting to 503 view or update any configuration data that will be stored on a 504 remote server. 506 o MUAs SHOULD implement TLS 1.2 or later. Earlier TLS and SSL 507 versions MAY also be supported so long as the MUA requires at 508 least TLS 1.1 when accessing accounts that are configured to 509 impose minimum confidentiality requirements. 511 o All MUAs SHOULD implement the recommended TLS cipher suites 512 described in [RFC7525] or a future BCP or standards track revision 513 of that document. 515 o MUAs that are configured to not require minimum confidentiality 516 for one or more accounts SHOULD detect when TLS becomes available 517 on those accounts, and offer to upgrade the account to impose 518 minimum confidentiality requirements. 520 Additional considerations and details appear below. 522 5.1. Use of SRV records in Establishing Configuration 524 This section updates [RFC6186] by changing the preference rules and 525 adding a new SRV service label _submissions._tcp to refer to Message 526 Submission with Implicit TLS. 528 User-configurable MUAs SHOULD support use of [RFC6186] for account 529 setup. However, when using configuration information obtained by 530 this method, MUAs SHOULD ignore advertised services that do not 531 satisfy minimum confidentiality requirements, unless the user has 532 explicitly requested reduced confidentiality. This will have the 533 effect of causing the MUA to default to ignoring advertised 534 configurations that do not support TLS, even when those advertised 535 configurations have a higher priority than other advertised 536 configurations. 538 When using [RFC6186] configuration information, Mail User Agents 539 SHOULD NOT automatically establish new configurations that do not 540 require TLS for all servers, unless there are no advertised 541 configurations using TLS. If such a configuration is chosen, prior 542 to attempting to authenticate to the server or use the server for 543 message submission, the MUA SHOULD warn the user that traffic to that 544 server will not be encrypted and that it will therefore likely be 545 intercepted by unauthorized parties. The specific wording is to be 546 determined by the implementation, but it should adequately capture 547 the sense of risk given the widespread incidence of mass surveillance 548 of email traffic. 550 Similarly, a MUA MUST NOT attempt to "test" a particular mail account 551 configuration by submitting the user's authentication credentials to 552 a server, unless a TLS session meeting minimum confidentiality levels 553 has been established with that server. If minimum confidentiality 554 requirements have not been satisfied, the MUA must explicitly warn 555 the user that his password may be exposed to attackers before testing 556 the new configuration. 558 When establishing a new configuration for connecting to an IMAP, POP, 559 or SMTP Submission server, an MUA SHOULD NOT blindly trust SRV 560 records unless they are signed by DNSSEC and have a valid signature. 561 Instead, the MUA SHOULD warn the user that the DNS-advertised 562 mechanism for connecting to the server is not authenticated, and 563 request the user to manually verify the connection details by 564 reference to his or her mail service provider's documentation. 566 Similarly, an MUA MUST NOT consult SRV records to determine which 567 servers to use on every connection attempt, unless those SRV records 568 are signed by DNSSEC and have a valid signature. However, an MUA MAY 569 consult SRV records from time to time to determine if an MSP's server 570 configuration has changed, and alert the user if it appears that this 571 has happened. This can also serve as a means to encourage users to 572 upgrade their configurations to require TLS if and when their MSPs 573 support it. 575 5.2. Minimum Confidentiality Level 577 MUAs SHOULD, by default, require a minimum level of confidentiality 578 for services accessed by each account. For MUAs supporting the 579 ability to access multiple mail accounts, this requirement SHOULD be 580 configurable on a per-account basis. 582 The default minimum expected level of confidentiality for all new 583 accounts SHOULD be at least use of TLS version 1.1 or greater, and 584 successful validation of the server's certificate. (Future revisions 585 to this specification may raise these requirements or impose 586 additional requirements to address newly-discovered weaknesses in 587 protocols or cryptographic algorithms.) 589 MUAs MAY permit the user to disable this minimum confidentiality 590 requirement during initial account configuration, or subsequently 591 editing an account configuration, but MUST warn users that such a 592 configuration will not assure privacy for either passwords or 593 messages. 595 An MUA which is configured to require a minimum level of 596 confidentiality for a mail account MUST NOT attempt to perform any 597 operation other than capability discovery, or STARTTLS for servers 598 not using Implicit TLS, unless the minimum level of confidentiality 599 is provided by that connection. 601 MUAs SHOULD NOT allow users to "click through" to access or send mail 602 via an connection, or to authenticate to any service using a 603 password, if that account is configured to impose minimum 604 confidentiality requirements and that connection does not meet all of 605 those requirements. Experience indicates that users presented with 606 such an option often "click through" without understanding the risks 607 that they're accepting by doing so. Furthermore, users who 608 frequently find the need to "click through" to use an insecure 609 connection may become conditioned to do so as a matter of habit, 610 before considering whether the risks are reasonable in each specific 611 instance. 613 An MUA which is not configured to require a minimum level of 614 confidentiality for a mail account SHOULD still attempt to connect to 615 the services associated with that account using the most secure means 616 available, e.g. by using Implicit TLS or STARTTLS. 618 5.3. Certificiate Validation 620 MUAs MUST validate TLS server certificates according to [RFC7817] and 621 PKIX [RFC5280]. 623 MUAs MAY also support DANE [RFC6698] as a means of validating server 624 certificates in order to meet minimum confidentiality requirements. 626 MUAs MAY support use of certificate pinning but MUST NOT consider a 627 connection in which the server's authenticity relies on certificate 628 pinning, as providing the minimum level of confidentiality. (See 629 Section 5.4.) 631 5.4. Certificate Pinning 633 During account setup, the MUA will identify servers that provide 634 account services such as mail access and mail submission (the 635 previous section describes one way to do this). The certificates for 636 these servers are verified using the rules described in [RFC7817] and 637 PKIX [RFC5280]. In the event the certificate does not validate due 638 to an expired certificate, lack of appropriate chain of trust, or 639 lack of identifier match, the MUA MAY offer to create a persistent 640 binding between that certificate and the saved host name for the 641 server, for use when accessing that account's servers. This is 642 called certificate pinning. 644 Certificate pinning is only appropriate during mail account setup and 645 MUST NOT be offered as an option in response to a failed certificate 646 validation for an existing mail account. An MUA that allows 647 certificate pinning MUST NOT allow a certificate pinned for one 648 account to validate connections for other accounts. 650 A pinned certificate is subject to a man-in-the-middle attack at 651 account setup time, and lacks a mechanism to revoke or securely 652 refresh the certificate. Note also that a man-in-the-middle attack 653 at account setup time will expose the user's password to the attacker 654 (if a password is used). Therefore use of a pinned certificate does 655 not meet the requirement for a minimum confidentiality level, and an 656 MUA MUST NOT indicate to the user that the such confidentiality is 657 provided. Additional advice on certificate pinning is present in 658 [RFC6125]. 660 5.5. Client Certificate Authentication 662 MUAs MAY implement client certificate authentication on the Implicit 663 TLS port. An MUA MUST NOT provide a client certificate during the 664 TLS handshake unless the server requests one and the client has 665 determined the certificate can be safely used with that specific 666 server, OR the client has been explicitly configured by the user to 667 use that particular certificate with that server. How to make this 668 determination is presently implementation specific. 670 A client supporting client certificate authentication with Implicit 671 TLS MUST implement the SASL EXTERNAL [RFC4422] mechanism using the 672 appropriate authentication command (AUTH for POP3 [RFC5034], AUTH for 673 SMTP Submission [RFC4954], AUTHENTICATE for IMAP [RFC3501]). 675 6. Considerations related to Anti-Virus/Anti-Spam Software and Services 677 There are multiple ways to connect an Anti-Virus and/or Anti-Spam 678 (AVAS) service to a mail server. Some mechanisms, such as the de- 679 facto milter protocol, are out of scope for this specification. 680 However, some services use an SMTP relay proxy that intercepts mail 681 at the application layer to perform a scan and proxy or forward to 682 another MTA. Deploying AVAS services in this way can cause many 683 problems [RFC2979] including direct interference with this 684 specification, and other forms of confidentiality or security 685 reduction. An AVAS product or service is considered compatible with 686 this specification if all IMAP, POP and SMTP-related software 687 (including proxies) it includes are compliant with this 688 specification. 690 Note that end-to-end email encryption prevents AVAS software and 691 services from using email content as part of a spam or virus 692 assessment. Furthermore, while a minimum confidentiality level can 693 prevent a man-in-the-middle from introducing spam or virus content 694 between the MUA and Submission server, it does not prevent other 695 forms of client or account compromise. Use of AVAS services for 696 submitted email therefore remains necessary. 698 7. IANA Considerations 700 7.1. POP3S Port Registration Update 702 IANA is asked to update the registration of the TCP well-known port 703 995 using the following template ([RFC6335]): 705 Service Name: pop3s 706 Transport Protocol: TCP 707 Assignee: IETF 708 Contact: IESG 709 Description: POP3 over TLS protocol 710 Reference: RFC XXXX (this document once published) 711 Port Number: 995 713 7.2. IMAPS Port Registration Update 715 IANA is asked to update the registration of the TCP well-known port 716 993 using the following template ([RFC6335]): 718 Service Name: imaps 719 Transport Protocol: TCP 720 Assignee: IETF 721 Contact: IESG 722 Description: IMAP over TLS protocol 723 Reference: RFC XXXX (this document once published) 724 Port Number: 993 726 7.3. Submissions Port Registration 728 IANA is asked to assign an alternate usage of port 465 in addition to 729 the current assignment using the following template ([RFC6335]): 731 Service Name: submissions 732 Transport Protocol: TCP 733 Assignee: IETF 734 Contact: IESG 735 Description: Message Submission over TLS protocol 736 Reference: RFC XXXX (this document once published) 737 Port Number: 465 739 This is a one-time procedural exception to the rules in RFC 6335. 740 This requires explicit IESG approval and does not set a precedent. 742 Historically, port 465 was briefly registered as the "smtps" port. 743 This registration made no sense as the SMTP transport MX 744 infrastructure has no way to specify a port, so port 25 is always 745 used. As a result, the registration was revoked and was subsequently 746 reassigned to a different service. In hindsight, the "smtps" 747 registration should have been renamed or reserved rather than 748 revoked. Unfortunately, some widely deployed mail software 749 interpreted "smtps" as "submissions" [RFC6409] and used that port for 750 email submission by default when an end-user requests security during 751 account setup. If a new port is assigned for the submissions 752 service, email software will either continue with unregistered use of 753 port 465 (leaving the port registry inaccurate relative to de-facto 754 practice and wasting a well-known port), or confusion between the de- 755 facto and registered ports will cause harmful interoperability 756 problems that will deter use of TLS for message submission. The 757 authors believe both of these outcomes are less desirable than a wart 758 in the registry documenting real-world usage of a port for two 759 purposes. Although STARTTLS-on-port-587 has deployed, it has not 760 replaced deployed use of Implicit TLS submission on port 465. 762 8. Security Considerations 764 This entire document is about security considerations. In general, 765 this is targeted to improve mail confidentiality and to mitigate 766 threats external to the email system such as network-level snooping 767 or interception; this is not intended to mitigate active attackers 768 who have compromised service provider systems. 770 It could be argued that sharing the name and version of the client 771 software with the server has privacy implications. Although 772 providing this information is not required, it is encouraged so that 773 mail service providers can more effectively inform end-users running 774 old clients that they need to upgrade to protect their security, or 775 know which clients to use in a test deployment prior to upgrading a 776 server to have higher security requirements. 778 9. References 780 9.1. Normative References 782 [RFC1939] Myers, J. and M. Rose, "Post Office Protocol - Version 3", 783 STD 53, RFC 1939, DOI 10.17487/RFC1939, May 1996, 784 . 786 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 787 Requirement Levels", BCP 14, RFC 2119, 788 DOI 10.17487/RFC2119, March 1997, 789 . 791 [RFC3207] Hoffman, P., "SMTP Service Extension for Secure SMTP over 792 Transport Layer Security", RFC 3207, DOI 10.17487/RFC3207, 793 February 2002, . 795 [RFC3501] Crispin, M., "INTERNET MESSAGE ACCESS PROTOCOL - VERSION 796 4rev1", RFC 3501, DOI 10.17487/RFC3501, March 2003, 797 . 799 [RFC5034] Siemborski, R. and A. Menon-Sen, "The Post Office Protocol 800 (POP3) Simple Authentication and Security Layer (SASL) 801 Authentication Mechanism", RFC 5034, DOI 10.17487/RFC5034, 802 July 2007, . 804 [RFC5068] Hutzler, C., Crocker, D., Resnick, P., Allman, E., and T. 805 Finch, "Email Submission Operations: Access and 806 Accountability Requirements", BCP 134, RFC 5068, 807 DOI 10.17487/RFC5068, November 2007, 808 . 810 [RFC5234] Crocker, D., Ed. and P. Overell, "Augmented BNF for Syntax 811 Specifications: ABNF", STD 68, RFC 5234, 812 DOI 10.17487/RFC5234, January 2008, 813 . 815 [RFC5246] Dierks, T. and E. Rescorla, "The Transport Layer Security 816 (TLS) Protocol Version 1.2", RFC 5246, 817 DOI 10.17487/RFC5246, August 2008, 818 . 820 [RFC5280] Cooper, D., Santesson, S., Farrell, S., Boeyen, S., 821 Housley, R., and W. Polk, "Internet X.509 Public Key 822 Infrastructure Certificate and Certificate Revocation List 823 (CRL) Profile", RFC 5280, DOI 10.17487/RFC5280, May 2008, 824 . 826 [RFC5322] Resnick, P., Ed., "Internet Message Format", RFC 5322, 827 DOI 10.17487/RFC5322, October 2008, 828 . 830 [RFC6186] Daboo, C., "Use of SRV Records for Locating Email 831 Submission/Access Services", RFC 6186, 832 DOI 10.17487/RFC6186, March 2011, 833 . 835 [RFC6409] Gellens, R. and J. Klensin, "Message Submission for Mail", 836 STD 72, RFC 6409, DOI 10.17487/RFC6409, November 2011, 837 . 839 [RFC7525] Sheffer, Y., Holz, R., and P. Saint-Andre, 840 "Recommendations for Secure Use of Transport Layer 841 Security (TLS) and Datagram Transport Layer Security 842 (DTLS)", BCP 195, RFC 7525, DOI 10.17487/RFC7525, May 843 2015, . 845 [RFC7672] Dukhovni, V. and W. Hardaker, "SMTP Security via 846 Opportunistic DNS-Based Authentication of Named Entities 847 (DANE) Transport Layer Security (TLS)", RFC 7672, 848 DOI 10.17487/RFC7672, October 2015, 849 . 851 [RFC7817] Melnikov, A., "Updated Transport Layer Security (TLS) 852 Server Identity Check Procedure for Email-Related 853 Protocols", RFC 7817, DOI 10.17487/RFC7817, March 2016, 854 . 856 9.2. Informative References 858 [RFC2595] Newman, C., "Using TLS with IMAP, POP3 and ACAP", 859 RFC 2595, DOI 10.17487/RFC2595, June 1999, 860 . 862 [RFC2979] Freed, N., "Behavior of and Requirements for Internet 863 Firewalls", RFC 2979, DOI 10.17487/RFC2979, October 2000, 864 . 866 [RFC3848] Newman, C., "ESMTP and LMTP Transmission Types 867 Registration", RFC 3848, DOI 10.17487/RFC3848, July 2004, 868 . 870 [RFC4422] Melnikov, A., Ed. and K. Zeilenga, Ed., "Simple 871 Authentication and Security Layer (SASL)", RFC 4422, 872 DOI 10.17487/RFC4422, June 2006, 873 . 875 [RFC4954] Siemborski, R., Ed. and A. Melnikov, Ed., "SMTP Service 876 Extension for Authentication", RFC 4954, 877 DOI 10.17487/RFC4954, July 2007, 878 . 880 [RFC6066] Eastlake 3rd, D., "Transport Layer Security (TLS) 881 Extensions: Extension Definitions", RFC 6066, 882 DOI 10.17487/RFC6066, January 2011, 883 . 885 [RFC6125] Saint-Andre, P. and J. Hodges, "Representation and 886 Verification of Domain-Based Application Service Identity 887 within Internet Public Key Infrastructure Using X.509 888 (PKIX) Certificates in the Context of Transport Layer 889 Security (TLS)", RFC 6125, DOI 10.17487/RFC6125, March 890 2011, . 892 [RFC6335] Cotton, M., Eggert, L., Touch, J., Westerlund, M., and S. 893 Cheshire, "Internet Assigned Numbers Authority (IANA) 894 Procedures for the Management of the Service Name and 895 Transport Protocol Port Number Registry", BCP 165, 896 RFC 6335, DOI 10.17487/RFC6335, August 2011, 897 . 899 [RFC6698] Hoffman, P. and J. Schlyter, "The DNS-Based Authentication 900 of Named Entities (DANE) Transport Layer Security (TLS) 901 Protocol: TLSA", RFC 6698, DOI 10.17487/RFC6698, August 902 2012, . 904 Appendix A. Design Considerations 906 This section is not normative. 908 The first version of this was written independently from draft-moore- 909 email-tls-00.txt; subsequent versions merge ideas from both drafts. 911 One author of this document was also the author of RFC 2595 that 912 became the standard for TLS usage with POP and IMAP, and the other 913 author was perhaps the first to propose that idea. In hindsight both 914 authors now believe that that approach was a mistake. At this point 915 the authors believe that while anything that makes it easier to 916 deploy TLS is good, the desirable end state is that these protocols 917 always use TLS, leaving no need for a separate port for cleartext 918 operation except to support legacy clients while they continue to be 919 used. The separate port model for TLS is inherently simpler to 920 implement, debug and deploy. It also enables a "generic TLS load- 921 balancer" that accepts secure client connections for arbitrary foo- 922 over-TLS protocols and forwards them to a server that may or may not 923 support TLS. Such load-balancers cause many problems because they 924 violate the end-to-end principle and the server loses the ability to 925 log security-relevant information about the client unless the 926 protocol is designed to forward that information (as this 927 specification does for the cipher suite). However, they can result 928 in TLS deployment where it would not otherwise happen which is a 929 sufficiently important goal that it overrides the problems. 931 Although STARTTLS appears only slightly more complex than separate- 932 port TLS, we again learned the lesson that complexity is the enemy of 933 security in the form of the STARTTLS command injection vulnerability 934 (CERT vulnerability ID #555316). Although there's nothing inherently 935 wrong with STARTTLS, the fact it resulted in a common implementation 936 error (made independently by multiple implementers) suggests it is a 937 less secure architecture than Implicit TLS. 939 Section 7 of RFC 2595 critiques the separate-port approach to TLS. 940 The first bullet was a correct critique. There are proposals in the 941 http community to address that, and use of SRV records as described 942 in RFC 6186 resolves that critique for email. The second bullet is 943 correct as well, but not very important because useful deployment of 944 security layers other than TLS in email is small enough to be 945 effectively irrelevant. The third bullet is incorrect because it 946 misses the desirable option of "use and latch-on TLS if available". 947 The fourth bullet may be correct, but is not a problem yet with 948 current port consumption rates. The fundamental error was 949 prioritizing a perceived better design based on a mostly valid 950 critique over real-world deployability. But getting security and 951 confidentiality facilities actually deployed is so important it 952 should trump design purity considerations. 954 Port 465 is presently used for two purposes: for submissions by a 955 large number of clients and service providers and for the "urd" 956 protocol by one vendor. Actually documenting this current state is 957 controversial as discussed in the IANA considerations section. 958 However, there is no good alternative. Registering a new port for 959 submissions when port 465 is widely used for that purpose already 960 will just create interoperability problems. Registering a port 961 that's only used if advertised by an SRV record (RFC 6186) would not 962 create interoperability problems but would require all client and 963 server deployments and software to change significantly which is 964 contrary to the goal of promoting more TLS use. Encouraging use of 965 STARTTLS on port 587 would not create interoperability problems, but 966 is unlikely to have impact on current undocumented use of port 465 967 and makes the guidance in this document less consistent. The 968 remaining option is to document the current state of the world and 969 support future use of port 465 for submission as this increases 970 consistency and ease-of-deployment for TLS email submission. 972 Appendix B. Change Log 974 Changes since draft-ietf-uta-email-deep-06: 976 o On the recommendation of one of the co-chairs and some working 977 group members, rewrote document with the intended status of BCP. 978 This involved removing a great deal of text that consisted 979 essentially of new protocol specification, especially the STS 980 features, on the theory that a BCP should base its recommendations 981 on current practice, and that new protocol features should be 982 subject to the interoperability test requirements associated with 983 normal standards-track documents. 985 Changes since draft-ietf-uta-email-deep-05: 987 o Clarify throughout that the confidentiality assurance level 988 associated with a mail account is a minimum level; attempt to 989 distinguish this from the current confidentiality level provided 990 by a connection between client and server. 992 o Change naming for confidentiality assurance levels: instead of 993 "high" or "no" confidence, assign numbers 1 and 0 to them 994 respectively. This because it seems likely that in the not-too- 995 distant future, what was defined in -05 as "high" confidence will 996 be considered insufficient, and calling that "high" confidence 997 will become misleading. For example, relying entirely on a list 998 of trusted CAs to validate server certificates from arbitrary 999 parties, appears to be less and less reliable in practice at 1000 thwarting MITM attacks. 1002 o Clarify that if some services associated with a mail account don't 1003 meet the minimum confidentiality assurance level assigned to that 1004 account, other services that do meet that minimum confidentiality 1005 assurance level may continue to be used. 1007 o Clarify that successful negotiation of at least TLS version 1.1 is 1008 required as a condition of meeting confidentiality assurance level 1009 1. 1011 o Clarify that validation of a server certificate using either DANE 1012 or PKIX is sufficient to meet the certificate validation 1013 requirement of confidentiality assurance level 1. 1015 o Clarify that minimum confidentiality assurance levels are separate 1016 from security directives, and that the requirements of both 1017 mechanisms must be met. 1019 o Explicitly cite an example that a security directive of tls- 1020 version=1.2 won't be saved if the currently negotiated tls-version 1021 is 1.1. (This example already appeared a bit later in the text, 1022 but for author KM it seemed to make the mechanism clearer to use 1023 this example earlier.) 1025 o Clarify some protocol examples as to whether PKIX or DANE was used 1026 to verify a server's certificate. 1028 o Remove most references to DEEP as the conversion from DEEP to MUA- 1029 STS seemed incomplete, but kept the DEEP command for use in POP3 1030 on the assumption that author CN wanted it that way. 1032 o Removed most references to "latch" and derivative words. 1034 o Added pkix+dane as a value for the tls-cert directive, to indicate 1035 (from a server) that both PKIX and DANE validation will be 1036 supported, or (from a client) that both PKIX and DANE were used to 1037 validate a certificate. Also clarified what each of any, pkix, 1038 dane, and pkix+dane mean when advertised by a server and in 1039 particular that tls-cert=any provides no assurance of future PKIX 1040 verifiability in contrast to tls-cert=pkix or tls-cert=pkix+dane. 1041 It seemed important to support the ability to evolve to using 1042 multiple trust anchors for certificate validation, but also to 1043 allow servers to have the option to migrate from PKIX to DANE if 1044 that made sense for them. This change seemed less disruptive than 1045 either defining additional directives, or allowing multiple 1046 instances of the same directive with different values to appear in 1047 the same advertisement. 1049 o Clarify interaction of this specification with anti-virus / anti- 1050 spam mechanisms. 1052 Changes since draft-ietf-uta-email-deep-04: 1054 o Swap sections 5.1 and 5.3 ("Email Security Tags" and "Server DEEP 1055 Status") as that order may aid understanding of the model. Also 1056 rewrote parts of these two sections to try to make the model 1057 clearer. 1059 o Add text about versioning of security tags to make the model 1060 clearer. 1062 o Add example of security tag upgrade. 1064 o Convert remaining mention of TLS 1.0 to TLS 1.1. 1066 o Change document title from DEEP to MUA STS to align with SMTP 1067 relay STS. 1069 * Slight updates to abstract and introductions. 1071 * Rename security latches/tags to security directives. 1073 * Rename server DEEP status to STS policy. 1075 * Change syntax to use directive-style HSTS syntax. 1077 o Make HSTS reference normative. 1079 o Remove SMTP DSN header as that belongs in SMTP relay STS document. 1081 Changes since draft-ietf-uta-email-deep-03: 1083 o Add more references to ietf-uta-email-tls-certs in implementation 1084 requirements section. 1086 o Replace primary reference to RFC 6125 with ietf-uta-email-tls- 1087 certs, so move RFC 6125 to informative list for this 1088 specification. 1090 Changes since draft-ietf-uta-email-deep-02: 1092 o Make reference to design considerations explicit rather than 1093 "elsewhere in this document". 1095 o Change provider requirement so SMTP submission services are 1096 separate from SMTP MTA services as opposed to the previous 1097 phrasing that required the servers be separate (which is too 1098 restrictive). 1100 o Update DANE SMTP reference 1102 Changes since draft-ietf-uta-email-deep-01: 1104 o Change text in tls11 and tls12 registrations to clarify 1105 certificate rules, including additional PKIX and DANE references. 1107 o Change from tls10 to tls11 (including reference) as the minimum. 1109 o Fix typo in example 5. 1111 o Remove open issues section; enough time has passed so not worth 1112 waiting for more input. 1114 Changes since draft-ietf-uta-email-deep-00: 1116 o Update and clarify abstract 1118 o use term confidentiality instead of privacy in most cases. 1120 o update open issues to request input for missing text. 1122 o move certificate pinning sub-section to account setup section and 1123 attempt to define it more precisely. 1125 o Add note about end-to-end encryption in AVAS section. 1127 o swap order of DNSSEC and TLSA sub-sections. 1129 o change meaning of 'tls10' and 'tls12' latches to require 1130 certificate validation. 1132 o Replace cipher suite advice with reference to RFC 7525. Change 1133 examples to use TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 as cipher 1134 suite. 1136 o Add text to update IMAP, POP3 and Message Submission standards 1137 with newer TLS advice. 1139 o Add clearer text in introduction that this does not cover SMTP 1140 relay. 1142 o Update references to uta-tls-certs. 1144 o Add paragraph to Implicit TLS for SMTP Submission section 1145 recommending that STARTTLS also be implemented. 1147 Changes since draft-newman-email-deep-02: 1149 o Changed "privacy assurance" to "confidentiality assurance" 1151 o Changed "low privacy assurance" to "no confidentiality assurance" 1153 o Attempt to improve definition of confidentiality assurance level. 1155 o Add SHOULD indicate when MUA is showing list of mail accounts. 1157 o Add SHOULD NOT latch tls10, tls12 tags until TLS negotiated. 1159 o Removed sentence about deleting and re-creating the account in 1160 latch failure section. 1162 o Remove use of word "fallback" with respect to TLS version 1163 negotiation. 1165 o Added bullet about changes to Internet facing servers to MSP 1166 section. 1168 o minor wording improvements based on feedback 1170 Changes since -01: 1172 o Updated abstract, introduction and document structure to focus 1173 more on mail user agent privacy assurance. 1175 o Added email account privacy section, also moving section on 1176 account setup using SRV records to that section. 1178 o Finished writing IANA considerations section 1180 o Remove provisional concept and instead have server explicitly list 1181 security tags clients should latch. 1183 o Added note that rules for the submissions port follow the same 1184 rules as those for the submit port. 1186 o Reference and update advice in [RFC5068]. 1188 o Fixed typo in Client Certificate Authentication section. 1190 o Removed tls-pfs security latch and all mention of perfect forward 1191 secrecy as it was controversial. 1193 o Added reference to HSTS. 1195 Changes since -00: 1197 o Rewrote introduction to merge ideas from draft-moore-email-tls-00. 1199 o Added Implicit TLS section, Account configuration section and IANA 1200 port registration updates based on draft-moore-email-tls-00. 1202 o Add protocol details necessary to standardize implicit TLS for 1203 POP/IMAP/submission, using ideas from draft-melnikov-pop3-over- 1204 tls. 1206 o Reduce initial set of security tags based on feedback. 1208 o Add deep status concept to allow a window for software updates to 1209 be backed out before latches make that problematic, as well as to 1210 provide service providers with a mechanism they can use to assist 1211 customers in the event of a privacy failure. 1213 o Add DNS SRV section from draft-moore-email-tls-00. 1215 o Write most of the missing IANA considerations section. 1217 o Rewrite most of implementation requirements section based more on 1218 draft-moore-email-tls-00. Remove new cipher requirements for now 1219 because those may be dealt with elsewhere. 1221 Appendix C. Acknowledgements 1223 Thanks to Ned Freed for discussion of the initial latch concepts in 1224 this document. Thanks to Alexey Melnikov for draft-melnikov-pop3- 1225 over-tls-02, which was the basis of the POP3 Implicit TLS text. 1226 Thanks to Russ Housley, Alexey Melnikov and Dan Newman for review 1227 feedback. Thanks to Paul Hoffman for interesting feedback in initial 1228 conversations about this idea. 1230 Authors' Addresses 1232 Keith Moore 1233 Network Heretics 1234 PO Box 1934 1235 Knoxville, TN 37901 1236 US 1238 Email: moore@network-heretics.com 1239 Chris Newman 1240 Oracle 1241 440 E. Huntington Dr., Suite 400 1242 Arcadia, CA 91006 1243 US 1245 Email: chris.newman@oracle.com