idnits 2.17.1 draft-ietf-precis-problem-statement-06.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 document seems to lack a both a reference to RFC 2119 and the recommended RFC 2119 boilerplate, even if it appears to use RFC 2119 keywords. RFC 2119 keyword, line 986: '...le username, and SHOULD use the SASLpr...' RFC 2119 keyword, line 990: '...mpty string), the server MUST fail the...' RFC 2119 keyword, line 994: '... [SASLprep]), and both client and server SHOULD (*) use the...' RFC 2119 keyword, line 998: '...mpty string), the server MUST fail the...' RFC 2119 keyword, line 1000: '...s requirement to MUST. Currently, the...' (6 more instances...) Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year == Line 830 has weird spacing: '...is used iSCSI...' -- The document date (July 9, 2012) is 4310 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Missing Reference: 'SASL' is mentioned on line 1004, but not defined == Missing Reference: 'SASLprep' is mentioned on line 1157, but not defined == Missing Reference: 'StringPrep' is mentioned on line 1253, but not defined == Missing Reference: 'RFC3629' is mentioned on line 1145, but not defined == Missing Reference: 'Stringprep' is mentioned on line 1152, but not defined == Missing Reference: 'PR29' is mentioned on line 1168, but not defined == Missing Reference: 'UTF-8' is mentioned on line 1228, but not defined == Missing Reference: 'Unicode' is mentioned on line 1241, but not defined == Outdated reference: A later version (-09) exists of draft-iab-identifier-comparison-00 -- Obsolete informational reference (is this intentional?): RFC 3454 (Obsoleted by RFC 7564) -- Obsolete informational reference (is this intentional?): RFC 3490 (Obsoleted by RFC 5890, RFC 5891) -- Obsolete informational reference (is this intentional?): RFC 3491 (Obsoleted by RFC 5891) -- Obsolete informational reference (is this intentional?): RFC 3530 (Obsoleted by RFC 7530) -- Obsolete informational reference (is this intentional?): RFC 3920 (Obsoleted by RFC 6120) -- Obsolete informational reference (is this intentional?): RFC 4013 (Obsoleted by RFC 7613) -- Obsolete informational reference (is this intentional?): RFC 5661 (Obsoleted by RFC 8881) Summary: 1 error (**), 0 flaws (~~), 11 warnings (==), 8 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group M. Blanchet 3 Internet-Draft Viagenie 4 Intended status: Informational A. Sullivan 5 Expires: January 10, 2013 Dyn, Inc. 6 July 9, 2012 8 Stringprep Revision and PRECIS Problem Statement 9 draft-ietf-precis-problem-statement-06.txt 11 Abstract 13 If a protocol expects to compare two strings and is prepared only for 14 those strings to be ASCII, then using Unicode codepoints in those 15 string requires they be prepared somehow. Internationalizing Domain 16 Names in Applications (here called IDNA2003) defined and used 17 Stringprep and Nameprep. Other protocols subsequently defined 18 Stringprep profiles. A new approach different from Stringprep and 19 Nameprep is used for a revision of IDNA2003 (called IDNA2008). Other 20 Stringprep profiles need to be similarly updated or a replacement of 21 Stringprep needs to be designed. This document outlines the issues 22 to be faced by those designing a Stringprep replacement. 24 Status of this Memo 26 This Internet-Draft is submitted in full conformance with the 27 provisions of BCP 78 and BCP 79. 29 Internet-Drafts are working documents of the Internet Engineering 30 Task Force (IETF). Note that other groups may also distribute 31 working documents as Internet-Drafts. The list of current Internet- 32 Drafts is at http://datatracker.ietf.org/drafts/current/. 34 Internet-Drafts are draft documents valid for a maximum of six months 35 and may be updated, replaced, or obsoleted by other documents at any 36 time. It is inappropriate to use Internet-Drafts as reference 37 material or to cite them other than as "work in progress." 39 This Internet-Draft will expire on January 10, 2013. 41 Copyright Notice 43 Copyright (c) 2012 IETF Trust and the persons identified as the 44 document authors. All rights reserved. 46 This document is subject to BCP 78 and the IETF Trust's Legal 47 Provisions Relating to IETF Documents 48 (http://trustee.ietf.org/license-info) in effect on the date of 49 publication of this document. Please review these documents 50 carefully, as they describe your rights and restrictions with respect 51 to this document. Code Components extracted from this document must 52 include Simplified BSD License text as described in Section 4.e of 53 the Trust Legal Provisions and are provided without warranty as 54 described in the Simplified BSD License. 56 Table of Contents 58 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 59 2. Conventions . . . . . . . . . . . . . . . . . . . . . . . . . 7 60 3. Stringprep Profiles Limitations . . . . . . . . . . . . . . . 8 61 4. Major Topics for Consideration . . . . . . . . . . . . . . . . 10 62 4.1. Comparison . . . . . . . . . . . . . . . . . . . . . . . . 10 63 4.1.1. Types of Identifiers . . . . . . . . . . . . . . . . . 10 64 4.1.2. Effect of comparison . . . . . . . . . . . . . . . . . 10 65 4.2. Dealing with characters . . . . . . . . . . . . . . . . . 10 66 4.2.1. Case folding, case sensitivity, and case 67 preservation . . . . . . . . . . . . . . . . . . . . . 11 68 4.2.2. Stringprep and NFKC . . . . . . . . . . . . . . . . . 11 69 4.2.3. Character mapping . . . . . . . . . . . . . . . . . . 11 70 4.2.4. Prohibited characters . . . . . . . . . . . . . . . . 12 71 4.2.5. Internal structure, delimiters, and special 72 characters . . . . . . . . . . . . . . . . . . . . . . 12 73 4.2.6. Restrictions because of glyph similarity . . . . . . . 13 74 4.3. Where the data comes from and where it goes . . . . . . . 13 75 4.3.1. User input and the source of protocol elements . . . . 13 76 4.3.2. User output . . . . . . . . . . . . . . . . . . . . . 13 77 4.3.3. Operations . . . . . . . . . . . . . . . . . . . . . . 14 78 5. Considerations for Stringprep replacement . . . . . . . . . . 16 79 6. Security Considerations . . . . . . . . . . . . . . . . . . . 17 80 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 18 81 8. Discussion home for this draft . . . . . . . . . . . . . . . . 19 82 9. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 20 83 10. Informative References . . . . . . . . . . . . . . . . . . . . 21 84 Appendix A. Classification of Stringprep Profiles . . . . . . . . 25 85 Appendix B. Evaluation of Stringprep Profiles . . . . . . . . . . 26 86 B.1. iSCSI Stringprep Profiles: RFC3722, RFC3721, RFC3720 . . . 26 87 B.2. SMTP/POP3/ManageSieve Stringprep Profiles: 88 RFC4954,RFC5034,RFC 5804 . . . . . . . . . . . . . . . . . 28 89 B.3. IMAP Stringprep Profiles: RFC5738, RFC4314: Usernames . . 30 90 B.4. IMAP Stringprep Profiles: RFC5738: Passwords . . . . . . . 31 91 B.5. Anonymous SASL Stringprep Profiles: RFC4505 . . . . . . . 33 92 B.6. XMPP Stringprep Profiles: RFC3920 Nodeprep . . . . . . . . 35 93 B.7. XMPP Stringprep Profiles: RFC3920 Resourceprep . . . . . . 36 94 B.8. EAP Stringprep Profiles: RFC3748 . . . . . . . . . . . . . 37 95 Appendix C. Changes between versions . . . . . . . . . . . . . . 39 96 C.1. 00 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 97 C.2. 01 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 98 C.3. 02 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 99 C.4. 03 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 100 C.5. 04 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 101 C.6. 05 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 102 C.7. 06 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 103 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 41 105 1. Introduction 107 Internationalizing Domain Names in Applications (here called 108 IDNA2003) [RFC3490], [RFC3491], [RFC3492], [RFC3454] describes a 109 mechanism for encoding Unicode labels making up Internationalized 110 Domain Names (IDNs) as standard DNS labels. The labels were 111 processed using a method called Nameprep [RFC3491] and Punycode 112 [RFC3492]. That method was specific to IDNA2003, but is generalized 113 as Stringprep [RFC3454]. The general mechanism is used by other 114 protocols with similar needs, but with different constraints than 115 IDNA2003. 117 Stringprep defines a framework within which protocols define their 118 Stringprep profiles. Known IETF specifications using Stringprep are 119 listed below: 121 o The Nameprep profile [RFC3490] for use in Internationalized Domain 122 Names (IDNs); 124 o NFSv4 [RFC3530] and NFSv4.1 [RFC5661]; 126 o The iSCSI profile [RFC3722] for use in Internet Small Computer 127 Systems Interface (iSCSI) Names; 129 o EAP [RFC3748]; 131 o The Nodeprep and Resourceprep profiles [RFC3920] for use in the 132 Extensible Messaging and Presence Protocol (XMPP), and the XMPP to 133 CPIM mapping [RFC3922] (the latter of these relies on the former); 135 o The Policy MIB profile [RFC4011] for use in the Simple Network 136 Management Protocol (SNMP); 138 o The SASLprep profile [RFC4013] for use in the Simple 139 Authentication and Security Layer (SASL), and SASL itself 140 [RFC4422]; 142 o TLS [RFC4279]; 144 o IMAP4 using SASLprep [RFC4314]; 146 o The trace profile [RFC4505] for use with the SASL ANONYMOUS 147 mechanism; 149 o The LDAP profile [RFC4518] for use with LDAP [RFC4511] and its 150 authentication methods [RFC4513]; 152 o Plain SASL using SASLprep [RFC4616]; 154 o NNTP using SASLprep [RFC4643]; 156 o PKIX subject identification using LDAPprep [RFC4683]; 158 o Internet Application Protocol Collation Registry [RFC4790]; 160 o SMTP Auth using SASLprep [RFC4954]; 162 o POP3 Auth using SASLprep [RFC5034]; 164 o TLS SRP using SASLprep [RFC5054]; 166 o IRI and URI in XMPP [RFC5122]; 168 o PKIX CRL using LDAPprep [RFC5280]; 170 o IAX using Nameprep [RFC5456]; 172 o SASL SCRAM using SASLprep [RFC5802]; 174 o Remote management of Sieve using SASLprep [RFC5804]; 176 o The unicode-casemap Unicode Collation [RFC5051]. 178 However, a review (see [ietf78precis]) of these protocol 179 specifications found that they are very similar and can be grouped 180 into a short number of classes. Moreover, many reuse the same 181 Stringprep profile, such as the SASL one. 183 IDNA2003 was replaced because of some limitations described in 184 [RFC4690]. The new IDN specification, called IDNA2008 [RFC5890], 185 [RFC5891], [RFC5892], [RFC5893] was designed based on the 186 considerations found in [RFC5894]. One of the effects of IDNA2008 is 187 that Nameprep and Stringprep are not used at all. Instead, an 188 algorithm based on Unicode properties of codepoints is defined. That 189 algorithm generates a stable and complete table of the supported 190 Unicode codepoints for each Unicode version. This algorithm is based 191 on an inclusion-based approach, instead of the exclusion-based 192 approach of Stringprep/Nameprep. That is, IDNA2003 created an 193 explicit list of excluded or mapped-away characters; anything in 194 Unicode 3.2 that was not so listed could be assumed to be allowed 195 under the protocol. IDNA2008 begins instead from the assumption that 196 code points are disallowed, and then relies on Unicode properties to 197 derive whether a given code point actually is allowed in the 198 protocol. 200 This document lists the shortcomings and issues found by protocols 201 listed above that defined Stringprep profiles. It also lists the 202 requirements for any potential replacement of Stringprep. 204 2. Conventions 206 A single Unicode code point in this memo is denoted by "U+" followed 207 by four to six hexadecimal digits. Compare to [Unicode61], Appendix 208 A. 210 3. Stringprep Profiles Limitations 212 During IETF 77, a BOF discussed the current state of the protocols 213 that have defined Stringprep profiles [NEWPREP]. The main 214 conclusions from that discussion were as follows: 216 o Stringprep is bound to version 3.2 of Unicode. Stringprep has not 217 been updated to new versions of Unicode. Therefore, the protocols 218 using Stringprep are stuck to Unicode 3.2. 220 o The protocols need to be updated to support new versions of 221 Unicode. The protocols would like to not be bound to a specific 222 version of Unicode, but rather have better Unicode agility in the 223 way of IDNA2008. This is important partly because it is usually 224 impossible for an application to require Unicode 3.2; the 225 application gets whatever version of Unicode is available on the 226 host. 228 o The protocols require better bidirectional support (bidi) than 229 currently offered by Stringprep. 231 o If the protocols are updated to use a new version of Stringprep or 232 another framework, then backward compatibility is an important 233 requirement. For example, Stringprep is based on and profiles may 234 use NFKC [UAX15], while IDNA2008 mostly uses NFC [UAX15]. 236 o Identifiers are passed between protocols. For example, the same 237 username string of codepoints may be passed between SASL, XMPP, 238 LDAP and EAP. Therefore, common set of rules or classes of 239 strings are preferred over specific rules for each protocol. 240 Without real planning in advance, many stringprep profiles reuse 241 other profiles, so this goal was accomplished by accident with 242 Stringprep. 244 Protocols that use Stringprep profiles use strings for different 245 purposes: 247 o XMPP uses a different Stringprep profile for each part of the XMPP 248 address (JID): a localpart which is similar to a username and used 249 for authentication, a domainpart which is a domain name and a 250 resource part which is less restrictive than the localpart. 252 o iSCSI uses a Stringprep profile for the IQN, which is very similar 253 to (often is) a DNS domain name. 255 o SASL and LDAP uses a Stringprep profile for usernames. 257 o LDAP uses a set of Stringprep profiles. 259 The apparent judgement of the BOF attendees [NEWPREP] was that it 260 would be highly desirable to have a replacement of Stringprep, with 261 similar characteristics to IDNA2008. That replacement should be 262 defined so that the protocols could use internationalized strings 263 without a lot of specialized internationalization work, since 264 internationalization expertise is not available in the respective 265 protocols or working groups. Accordingly, the IESG formed the PRECIS 266 working group to undertake the task. 268 Notwithstanding the desire evident in [NEWPREP] and the chartering of 269 a working group, IDNA2008 may be a poor model for what other 270 protocols ought to do, because it is designed to support an old 271 protocol that is designed to operate on the scale of the entire 272 Internet. Moreover, IDNA2008 is intended to be deployed without any 273 change to the base DNS protocol. Other protocols may aim at 274 deployment in more local environments, or may have protocol version 275 negotiation built in. 277 4. Major Topics for Consideration 279 This section provides an overview of major topics that a Stringprep 280 replacement needs to address. The headings correspond roughly with 281 categories under which known Stringprep-using protocol RFCs have been 282 evaluated. For the details of those evaluations, see Appendix A. 284 4.1. Comparison 286 4.1.1. Types of Identifiers 288 Following [I-D.iab-identifier-comparison], it is possible to organize 289 identifiers into three classes in respect of how they may be compared 290 with one another: 292 Absolute Identifiers Identifiers that can be compared byte-by-byte 293 for equality. 295 Definite Identifiers Identifiers that have a well-defined comparison 296 algorithm on which all parties agree. 298 Indefinite Identifiers Identifiers that have no single comparison 299 algorithm on which all parties agree. 301 Definite Identifiers include cases like the comparison of Unicode 302 code points in different encodings: they do not match byte for byte, 303 but can all be converted to a single encoding which then does match 304 byte for byte. Indefinite Identifiers are sometimes algorithmically 305 comparable by well-specified subsets of parties. For more discussion 306 of these categories, see [I-D.iab-identifier-comparison]. 308 The section on treating the existing known cases, Appendix A uses the 309 categories above. 311 4.1.2. Effect of comparison 313 The three classes of comparison style outlined in Section 4.1.1 may 314 have different effects when applied. It is necessary to evaluate the 315 effects if a comparison results in a false positive, and what the 316 effects are if a comparison results in a false negative, especially 317 in terms of the consequences to security and usability. 319 4.2. Dealing with characters 321 This section outlines a range of issues having to do with characters 322 in the target protocols, and outlines the ways in which IDNA2008 323 might be a good analogy to other protocols, and ways in which it 324 might be a poor one. 326 4.2.1. Case folding, case sensitivity, and case preservation 328 In IDNA2003, labels are always mapped to lower case before the 329 Punycode transformation. In IDNA2008, there is no mapping at all: 330 input is either a valid U-label or it is not. At the same time, 331 upper-case characters are by definition not valid U-labels, because 332 they fall into the Unstable category (category B) of [RFC5892]. 334 If there are protocols that require upper and lower cases be 335 preserved, then the analogy with IDNA2008 will break down. 336 Accordingly, existing protocols are to be evaluated according to the 337 following criteria: 339 1. Does the protocol use case folding? For all blocks of code 340 points, or just for certain subsets? 342 2. Is the system or protocol case sensitive? 344 3. Does the system or protocol preserve case? 346 4.2.2. Stringprep and NFKC 348 Stringprep profiles may use normalization. If they do, they use NFKC 349 [UAX15] (most profiles do). It is not clear that NFKC is the right 350 normalization to use in all cases. In [UAX15], there is the 351 following observation regarding Normalization Forms KC and KD: "It is 352 best to think of these Normalization Forms as being like uppercase or 353 lowercase mappings: useful in certain contexts for identifying core 354 meanings, but also performing modifications to the text that may not 355 always be appropriate." In general, it can be said that NFKC is more 356 aggressive about finding matches between codepoints than NFC. For 357 things like the spelling of users' names, then, NFKC may not be the 358 best form to use. At the same time, one of the nice things about 359 NFKC is that it deals with the width of characters that are otherwise 360 similar, by canonicalizing half-width to full-width. This mapping 361 step can be crucial in practice. A replacement for stringprep 362 depends on analyzing the different use profiles and considering 363 whether NFKC or NFC is a better normalization for each profile. 365 For the purposes of evaluating an existing example of Stringprep use, 366 it is helpful to know whether it uses no normalization, NFKC, or NFC. 368 4.2.3. Character mapping 370 Along with the case mapping issues raised in Section 4.2.1, there is 371 the question of whether some characters are mapped either to other 372 characters or to nothing during Stringprep. [RFC3454], Section 3, 373 outlines a number of characters that are mapped to nothing, and also 374 permits Stringprep profiles to define their own mappings. 376 4.2.4. Prohibited characters 378 Along with case folding and other character mappings, many protocols 379 have characters that are simply disallowed. For example, control 380 characters and special characters such as "@" or "/" may be 381 prohibited in a protocol. 383 One of the primary changes of IDNA2008 is in the way it approaches 384 Unicode code points, using the new inclusion-based approach (see 385 Section 1). 387 Because of the default assumption in IDNA2008 that a code point is 388 not allowed by the protocol, it has more than one class of "allowed 389 by the protocol"; this is unlike IDNA2003. While some code points 390 are disallowed outright, some are allowed only in certain contexts. 391 The reasons for the context-dependent rules have to do with the way 392 some characters are used. For instance, the ZERO WIDTH JOINER and 393 ZERO WIDTH NON-JOINER (ZWJ, U+200D and ZWNJ, U+200C) are allowed with 394 contextual rules because they are required in some circumstances, yet 395 are considered punctuation by Unicode and would therefore be 396 DISALLOWED under the usual IDNA2008 derivation rules. The goal of 397 IDNA2008 is to provide the widest repertoire of code points possible 398 and consistent with the traditional DNS "LDH" (letters, digits, 399 hyphen; see [RFC0952]) rule, trusting to the operators of individual 400 zones to make sensible (and usually more restrictive) policies for 401 their zones. 403 4.2.5. Internal structure, delimiters, and special characters 405 IDNA2008 has a special problem with delimiters, because the delimiter 406 "character" in the DNS wire format is not really part of the data. 407 In DNS, labels are not separated exactly; instead, a label carries 408 with it an indicator that says how long the label is. When the label 409 is presented in presentation format as part of a fully qualified 410 domain name, the label separator FULL STOP, U+002E (.) is used to 411 break up the labels. But because that label separator does not 412 travel with the wire format of the domain name, there is no way to 413 encode a different, "internationalized" separator in IDNA2008. 415 Other protocols may include characters with similar special meaning 416 within the protocol. Common characters for these purposes include 417 FULL STOP, U+002E (.); COMMERCIAL AT, U+0040 (@); HYPHEN-MINUS, 418 U+002D (-); SOLIDUS, U+002F (/); and LOW LINE, U+005F (_). The mere 419 inclusion of such a character in the protocol is not enough for it to 420 be considered similar to another protocol using the same character; 421 instead, handling of the character must be taken into consideration 422 as well. 424 An important issue to tackle here is whether it is valuable to map to 425 or from these special characters as part of the Stringprep 426 replacement. In some locales, the analogue to FULL STOP, U+002E is 427 some other character, and users may expect to be able to substitute 428 their normal stop for FULL STOP, U+002E. At the same time, there are 429 predictability arguments in favour of treating identifiers with FULL 430 STOP, U+002E in them just the way they are treated under IDNA2008. 432 4.2.6. Restrictions because of glyph similarity 434 Homoglyphs are similarly (or identically) rendered glyphs of 435 different codepoints. For DNS names, homoglyphs may enable phishing. 436 If a protocol requires some visual comparison by end-users, then the 437 issue of homoglyphs are to be considered. In the DNS context, theses 438 issues are documented in [RFC5894] and [RFC4690]. IDNA2008 does not, 439 however, have a mechanism to deal with them, trusting to DNS zone 440 operators to enact sensible policies for the subset of Unicode they 441 wish to support, given their user community. A similar policy/ 442 protocol split may not be desirable in every protocol. 444 4.3. Where the data comes from and where it goes 446 4.3.1. User input and the source of protocol elements 448 Some protocol elements are provided by users, and others are not. 449 Those that are not may presumably be subject to greater restrictions, 450 whereas those that users provide likely need to permit the broadest 451 range of code points. The following questions are helpful: 453 1. Do users input the strings directly? 455 2. If so, how? (keyboard, stylus, voice, copy-paste, etc.) 457 3. Where do we place the dividing line between user interface and 458 protocol? (see [RFC5895]) 460 4.3.2. User output 462 Just as only some protocol elements are expected to be entered 463 directly by users, only some protocol elements are intended to be 464 consumed directly by users. It is important to know how users are 465 expected to be able to consume the protocol elements, because 466 different environments present different challenges. An element that 467 is only ever delivered as part of a vCard remains in machine-readable 468 format, so the problem of visual confusion is not a great one. Is 469 the protocol element published as part of a vCard, a web directory, 470 on a business card, or on "the side of a bus"? Do users use the 471 protocol element as an identifier (which means that they might enter 472 it again in some other context)? (See also Section 4.2.6.) 474 4.3.3. Operations 476 Some strings are useful as part of the protocol but are not used as 477 input to other operations (for instance, purely informative or 478 descriptive text). Other strings are used directly as input to other 479 operations (such as cryptographic hash functions), or are used 480 together with other strings to (such as concatenating a string with 481 some others to form a unique identifier). 483 4.3.3.1. String classes 485 Strings often have a similar function in different protocols. For 486 instance, many different protocols contain user identifiers or 487 passwords. A single profile for all such uses might be desirable. 489 Often, a string in a protocol is effectively a protocol element from 490 another protocol. For instance, different systems might use the same 491 credentials database for authentication. 493 4.3.3.2. Community Considerations 495 A Stringprep replacement that does anything more than just update 496 Stringprep to the latest version of Unicode will probably entail some 497 changes. It is important to identify the willingness of the 498 protocol-using community to accept backwards-incompatible changes. 499 By the same token, it is important to evaluate the desire of the 500 community for features not available under Stringprep. 502 4.3.3.3. Unicode Incompatible Changes 504 IDNA2008 uses an algorithm to derive the validity of a Unicode code 505 point for use under IDNA2008. It does this by using the properties 506 of each code point to test its validity. 508 This approach depends crucially on the idea that code points, once 509 valid for a protocol profile, will not later be made invalid. That 510 is not a guarantee currently provided by Unicode. Properties of code 511 points may change between versions of Unicode. Rarely, such a change 512 could cause a given code point to become invalid under a protocol 513 profile, even though the code point would be valid with an earlier 514 version of Unicode. This is not merely a theoretical possibility, 515 because it has occurred ([RFC6452]). 517 Accordingly, as in IDNA2008, a Stringprep replacement that intends to 518 be Unicode version agnostic will need to work out a mechanism to 519 address cases where incompatible changes occur because of new Unicode 520 versions. 522 5. Considerations for Stringprep replacement 524 The above suggests the following guidance for replacing Stringprep: 526 o A stringprep replacement should be defined. 528 o The replacement should take an approach similar to IDNA2008, (e.g. 529 by using codepoint properties instead of codepoint whitelisting) 530 in that it enables better Unicode agility. 532 o Protocols share similar characteristics of strings. Therefore, 533 defining internationalization preparation algorithms for the 534 smallest set of string classes may be sufficient for most cases, 535 providing coherence among a set of related protocols or protocols 536 where identifiers are exchanged. 538 o The sets of string classes need to be evaluated according to the 539 considerations that make up the headings in Section 4 541 o It is reasonable to limit scope to Unicode code points, and rule 542 the mapping of data from other character encodings outside the 543 scope of this effort. 545 o The replacement ought at least to provide guidance to applications 546 using the replacement on how to handle protocol incompatibilities 547 resulting from changes to Unicode. In an ideal world, the 548 stringprep replacement would handle the changes automatically, but 549 it appears that such automatic handling would require magic and 550 cannot be expected. 552 o Compatibility within each protocol between a technique that is 553 stringprep-based and the technique's replacement has to be 554 considered very carefully. 556 Existing deployments already depend on Stringprep profiles. 557 Therefore, a replacement must consider the effects of any new 558 strategy on existing deployments. By way of comparison, it is worth 559 noting that some characters were acceptable in IDNA labels under 560 IDNA2003, but are not protocol-valid under IDNA2008 (and conversely); 561 disagreement about what to do during the transition has resulted in 562 different approaches to mapping. Different implementers may make 563 different decisions about what to do in such cases; this could have 564 interoperability effects. It is necessary to trade better support 565 for different linguistic environments against the potential side 566 effects of backward incompatibility. 568 6. Security Considerations 570 This document merely states what problems are to be solved, and does 571 not define a protocol. There are undoubtedly security implications 572 of the particular results that will come from the work to be 573 completed. 575 7. IANA Considerations 577 This document has no actions for IANA. 579 8. Discussion home for this draft 581 Note: RFC-Editor, please remove this section before publication. 583 This document is intended to define the problem space discussed on 584 the precis@ietf.org mailing list. 586 9. Acknowledgements 588 This document is the product of the PRECIS IETF Working Group, and 589 participants in that Working Group were helpful in addressing issues 590 with the text. 592 Specific contributions came from David Black, Alan DeKok, Simon 593 Josefsson, Bill McQuillan, Alexey Melnikov, Peter Saint-Andre, Dave 594 Thaler, and Yoshiro Yoneya. 596 Dave Thaler provided the "buckets" insight in Section 4.1.1, central 597 to the organization of the problem. 599 Evaluations of Stringprep profiles that are included in Appendix B 600 were done by: David Black, Alexey Melnikov, Peter Saint-Andre, Dave 601 Thaler. 603 10. Informative References 605 [I-D.iab-identifier-comparison] 606 Thaler, D., "Issues in Identifier Comparison for Security 607 Purposes", draft-iab-identifier-comparison-00 (work in 608 progress), July 2011. 610 [NEWPREP] "Newprep BoF Meeting Minutes", March 2010. 612 [RFC0952] Harrenstien, K., Stahl, M., and E. Feinler, "DoD Internet 613 host table specification", RFC 952, October 1985. 615 [RFC3454] Hoffman, P. and M. Blanchet, "Preparation of 616 Internationalized Strings ("stringprep")", RFC 3454, 617 December 2002. 619 [RFC3490] Faltstrom, P., Hoffman, P., and A. Costello, 620 "Internationalizing Domain Names in Applications (IDNA)", 621 RFC 3490, March 2003. 623 [RFC3491] Hoffman, P. and M. Blanchet, "Nameprep: A Stringprep 624 Profile for Internationalized Domain Names (IDN)", 625 RFC 3491, March 2003. 627 [RFC3492] Costello, A., "Punycode: A Bootstring encoding of Unicode 628 for Internationalized Domain Names in Applications 629 (IDNA)", RFC 3492, March 2003. 631 [RFC3530] Shepler, S., Callaghan, B., Robinson, D., Thurlow, R., 632 Beame, C., Eisler, M., and D. Noveck, "Network File System 633 (NFS) version 4 Protocol", RFC 3530, April 2003. 635 [RFC3722] Bakke, M., "String Profile for Internet Small Computer 636 Systems Interface (iSCSI) Names", RFC 3722, April 2004. 638 [RFC3748] Aboba, B., Blunk, L., Vollbrecht, J., Carlson, J., and H. 639 Levkowetz, "Extensible Authentication Protocol (EAP)", 640 RFC 3748, June 2004. 642 [RFC3920] Saint-Andre, P., Ed., "Extensible Messaging and Presence 643 Protocol (XMPP): Core", RFC 3920, October 2004. 645 [RFC3922] Saint-Andre, P., "Mapping the Extensible Messaging and 646 Presence Protocol (XMPP) to Common Presence and Instant 647 Messaging (CPIM)", RFC 3922, October 2004. 649 [RFC4011] Waldbusser, S., Saperia, J., and T. Hongal, "Policy Based 650 Management MIB", RFC 4011, March 2005. 652 [RFC4013] Zeilenga, K., "SASLprep: Stringprep Profile for User Names 653 and Passwords", RFC 4013, February 2005. 655 [RFC4279] Eronen, P. and H. Tschofenig, "Pre-Shared Key Ciphersuites 656 for Transport Layer Security (TLS)", RFC 4279, 657 December 2005. 659 [RFC4314] Melnikov, A., "IMAP4 Access Control List (ACL) Extension", 660 RFC 4314, December 2005. 662 [RFC4422] Melnikov, A. and K. Zeilenga, "Simple Authentication and 663 Security Layer (SASL)", RFC 4422, June 2006. 665 [RFC4505] Zeilenga, K., "Anonymous Simple Authentication and 666 Security Layer (SASL) Mechanism", RFC 4505, June 2006. 668 [RFC4511] Sermersheim, J., "Lightweight Directory Access Protocol 669 (LDAP): The Protocol", RFC 4511, June 2006. 671 [RFC4513] Harrison, R., "Lightweight Directory Access Protocol 672 (LDAP): Authentication Methods and Security Mechanisms", 673 RFC 4513, June 2006. 675 [RFC4518] Zeilenga, K., "Lightweight Directory Access Protocol 676 (LDAP): Internationalized String Preparation", RFC 4518, 677 June 2006. 679 [RFC4616] Zeilenga, K., "The PLAIN Simple Authentication and 680 Security Layer (SASL) Mechanism", RFC 4616, August 2006. 682 [RFC4643] Vinocur, J. and K. Murchison, "Network News Transfer 683 Protocol (NNTP) Extension for Authentication", RFC 4643, 684 October 2006. 686 [RFC4683] Park, J., Lee, J., Lee, H., Park, S., and T. Polk, 687 "Internet X.509 Public Key Infrastructure Subject 688 Identification Method (SIM)", RFC 4683, October 2006. 690 [RFC4690] Klensin, J., Faltstrom, P., Karp, C., and IAB, "Review and 691 Recommendations for Internationalized Domain Names 692 (IDNs)", RFC 4690, September 2006. 694 [RFC4790] Newman, C., Duerst, M., and A. Gulbrandsen, "Internet 695 Application Protocol Collation Registry", RFC 4790, 696 March 2007. 698 [RFC4954] Siemborski, R. and A. Melnikov, "SMTP Service Extension 699 for Authentication", RFC 4954, July 2007. 701 [RFC5034] Siemborski, R. and A. Menon-Sen, "The Post Office Protocol 702 (POP3) Simple Authentication and Security Layer (SASL) 703 Authentication Mechanism", RFC 5034, July 2007. 705 [RFC5051] Crispin, M., "i;unicode-casemap - Simple Unicode Collation 706 Algorithm", RFC 5051, October 2007. 708 [RFC5054] Taylor, D., Wu, T., Mavrogiannopoulos, N., and T. Perrin, 709 "Using the Secure Remote Password (SRP) Protocol for TLS 710 Authentication", RFC 5054, November 2007. 712 [RFC5122] Saint-Andre, P., "Internationalized Resource Identifiers 713 (IRIs) and Uniform Resource Identifiers (URIs) for the 714 Extensible Messaging and Presence Protocol (XMPP)", 715 RFC 5122, February 2008. 717 [RFC5280] Cooper, D., Santesson, S., Farrell, S., Boeyen, S., 718 Housley, R., and W. Polk, "Internet X.509 Public Key 719 Infrastructure Certificate and Certificate Revocation List 720 (CRL) Profile", RFC 5280, May 2008. 722 [RFC5456] Spencer, M., Capouch, B., Guy, E., Miller, F., and K. 723 Shumard, "IAX: Inter-Asterisk eXchange Version 2", 724 RFC 5456, February 2010. 726 [RFC5661] Shepler, S., Eisler, M., and D. Noveck, "Network File 727 System (NFS) Version 4 Minor Version 1 Protocol", 728 RFC 5661, January 2010. 730 [RFC5802] Newman, C., Menon-Sen, A., Melnikov, A., and N. Williams, 731 "Salted Challenge Response Authentication Mechanism 732 (SCRAM) SASL and GSS-API Mechanisms", RFC 5802, July 2010. 734 [RFC5804] Melnikov, A. and T. Martin, "A Protocol for Remotely 735 Managing Sieve Scripts", RFC 5804, July 2010. 737 [RFC5890] Klensin, J., "Internationalized Domain Names for 738 Applications (IDNA): Definitions and Document Framework", 739 RFC 5890, August 2010. 741 [RFC5891] Klensin, J., "Internationalized Domain Names in 742 Applications (IDNA): Protocol", RFC 5891, August 2010. 744 [RFC5892] Faltstrom, P., "The Unicode Code Points and 745 Internationalized Domain Names for Applications (IDNA)", 746 RFC 5892, August 2010. 748 [RFC5893] Alvestrand, H. and C. Karp, "Right-to-Left Scripts for 749 Internationalized Domain Names for Applications (IDNA)", 750 RFC 5893, August 2010. 752 [RFC5894] Klensin, J., "Internationalized Domain Names for 753 Applications (IDNA): Background, Explanation, and 754 Rationale", RFC 5894, August 2010. 756 [RFC5895] Resnick, P. and P. Hoffman, "Mapping Characters for 757 Internationalized Domain Names in Applications (IDNA) 758 2008", RFC 5895, September 2010. 760 [RFC6452] Faltstrom, P. and P. Hoffman, "The Unicode Code Points and 761 Internationalized Domain Names for Applications (IDNA) - 762 Unicode 6.0", RFC 6452, November 2011. 764 [UAX15] "Unicode Standard Annex #15: Unicode Normalization Forms", 765 UAX 15, September 2009. 767 [Unicode61] 768 The Unicode Consortium. The Unicode Standard, Version 769 6.1, defined by:, "The Unicode Standard -- Version 6.1", 770 (Mountain View, CA: The Unicode Consortium, 2012. ISBN 771 978-1-936213-02-3), September 2009, 772 . 774 [ietf78precis] 775 Blanchet, M., "PRECIS Framework", Proceedings of the 776 Seventy-Eighth Internet Engineering Task 777 Force https://www.ietf.org/proceedings/78/, July 2010, 778 . 780 Appendix A. Classification of Stringprep Profiles 782 A number of the known cases of Stringprep use were evaluated during 783 the preparation of this document. The known cases are here described 784 in two ways. The types of identifiers the protocol uses is first 785 called out in the ID type column (from Section 4.1.1), using the 786 short forms "a" for Absolute, "d" for Definite, and "i" for 787 Indefinite. Next, there is a column that contains an "i" if the 788 protocol string comes from user input, an "o" if the protocol string 789 becomes user-facing output, "b" if both are true, and "n" if neither 790 is true. 792 +------+--------+-------+ 793 | RFC | IDtype | User? | 794 +------+--------+-------+ 795 | 3722 | a | o | 796 | | | | 797 | 3748 | - | - | 798 | | | | 799 | 3920 | a,d | b | 800 | | | | 801 | 4505 | a | i | 802 | | | | 803 | 4314 | a,d | b | 804 | | | | 805 | 4954 | a,d | b | 806 | | | | 807 | 5034 | a,d | b | 808 | | | | 809 | 5804 | a,d | b | 810 +------+--------+-------+ 812 Table 1 814 Appendix B. Evaluation of Stringprep Profiles 816 This section is a summary of evaluation of Stringprep profiles that 817 was done to get a good understanding of the usage of Stringprep. 818 This summary is by no means normative nor the actual evaluations 819 themselves. A template was used for reviewers to get a coherent view 820 of all evaluations. 822 B.1. iSCSI Stringprep Profiles: RFC3722, RFC3721, RFC3720 824 Description: An iSCSI session consists of an Initiator (i.e., host 825 or server that uses storage) communicating with a target (i.e., a 826 storage array or other system that provides storage). Both the 827 iSCSI initiator and target are named by iSCSI Names. The iSCSI 828 stringprep profile is used for iSCSI names. 830 How it is used iSCSI initiators and targets (see above). They can 831 also be used to identify SCSI ports (these are software entities 832 in the iSCSI protocol, not hardware ports), and iSCSI logical 833 units (storage volumes), although both are unusual in practice. 835 What entities create these identifiers? Generally a Human user (1) 836 configures an Automated system (2) that generates the names. 837 Advance configuration of the system is required due to the 838 embedded use of external unique identifier (from the DNS or IEEE). 840 How is the string input in the system? Keyboard and copy-paste are 841 common. Copy-paste is common because iSCSI names are long enough 842 to be problematic for humans to remember, causing use of email, 843 sneaker-net, text files, etc. to avoid mistype mistakes. 845 Where do we place the dividing line between user interface and 846 protocol? The iSCSI protocol requires that all internationalization 847 string preparation occur in the user interface. The iSCSI 848 protocol treats iSCSI names as opaque identifiers that are 849 compared byte-by-byte for equality. iSCSI names are generally not 850 checked for correct formatting by the protocol. 852 What entities enforce the rules? There are no iSCSI-specific 853 enforcement entities, although the use of unique identifier 854 information in the names relies on DNS registrars and the IEEE 855 Registration Authority. 857 Comparison Byte-by-byte 858 Case Folding, Sensitivity, Preservation Case folding is required for 859 the code blocks specified in RFC 3454, Table B.2. The overall 860 iSCSI naming system (UI + protocol) is case-insensitive. 862 What is the impact if the comparison results in a false positive? 863 Potential access to the wrong storage. - If the initiator has no 864 access to the wrong storage, an authentication failure is the 865 probable result. - If the initiator has access to the worng 866 storage, the resulting mis-identificaiton could result in use of 867 the wrong data and possible corruption of stored data. 869 What is the impact if the comparison results in a false negative? 870 Denial of authorized storage access. 872 What are the security impacts? iSCSI names are often used as the 873 authentication identities for storage systems. Comparison 874 problems could result in authentication problems, although note 875 that authentication failure ameliorates some of the false positive 876 cases. 878 Normalization NFKC, as specified by RFC 3454. 880 Mapping Yes, as specified by table B.1 in RFC 3454 882 Disallowed Characters Only the following characters are allowed: - 883 ASCII dash, dot, colon - ASCII lower case letters and digits - 884 Unicode lower case characters as specified by RFC 3454 All other 885 characters are disallowed. 887 Which other strings or identifiers are these most similar to? None - 888 iSCSI names are unique to iSCSI. 890 Are these strings or identifiers sometimes the same as strings or 891 identifiers from other protocols? No 893 Does the identifier have internal structure that needs to be 894 respected? Yes - ASCII dot, dash and colon are used for internal 895 name structure. These are not reserved characters in that they 896 can occur in the name in locations other than those used for 897 structuring purposes (e.g., only the first occurrence of a colon 898 character is structural, others are not). 900 How are users exposed to these strings? How are they published? 901 iSCSI names appear in server and storage system configuration 902 interfaces. They also appear in system logs. 904 Is the string / identifier used as input to other operations? 905 Effectively, no. The rarely used port and logical unit names 906 involve concatenation, which effectively extends a unique iSCSI 907 Name for a target to uniquely identify something within that 908 target. 910 How much tolerance for change from existing stringprep approach? 911 Good tolerance; the community would prefer that 912 internationalization experts solve internationalization problems 913 ;-). 915 How strong a desire for change (e.g., for Unicode agility)? Unicode 916 agility is desired in principle as long as nothing significant 917 breaks. 919 B.2. SMTP/POP3/ManageSieve Stringprep Profiles: RFC4954,RFC5034,RFC 920 5804 922 Description: Authorization identity (user identifier) exchanged 923 during SASL authentication: AUTH (SMTP/POP3) or AUTHENTICATE 924 (ManageSieve) command. 926 How It's Used: Used for proxy authorization, e.g. to [lawfully] 927 impersonate a particular user after a privileged authentication 929 Who Generates It: Typically generated by email system administrators 930 using some tools/conventions, sometimes from some backend 931 database. - In some setups human users can register own usernames 932 (e.g. webmail self registration) 934 User Input Methods: - Typed by user / selected from a list - Copy- 935 and-paste - Perhaps voice input - Can also be specified in 936 configuration files or on a command line 938 Enforcement: - Rules enforced by server / add-on service (e.g., 939 gateway service) on registration of account 941 Comparison Method: "Type 1" (byte-for-byte) or "type 2" (compare by 942 a common algorithm that everyone agrees on (e.g., normalize and 943 then compare the result byte-by-byte)) 945 Case Folding, Sensitivity, Preservation: Most likely case sensitive. 946 Exact requirements on case-sensitivity/case-preservation depend on 947 a specific implementation, e.g. an implementation might treat all 948 user identifiers as case insensitive (or case insensitive for US- 949 ASCII subset only). 951 Impact of Comparison: False positives: - an unauthorized user is 952 allowed email service access (login) False negatives: - an 953 authorized user is denied email service access 955 Normalization: NFKC (as per RFC 4013) 957 Mapping: (see Section 2 of RFC 4013 for the full list): Non ASCII 958 spaces are mapped to space, etc. 960 Disallowed Characters: (see Section 2 of RFC 4013 for the full 961 list): Unicode Control characters, etc. 963 String Classes: - simple username. See Section 2 of RFC 4013 for 964 details on restrictions. Note that some implementations allow 965 spaces in these. While implementations are not required to use a 966 specific format, an authorization identity frequently has the same 967 format as an email address (and EAI email address in the future), 968 or as a left hand side of an email address. Note: whatever is 969 recommended for SMTP/POP/ManageSieve authorization identity should 970 also be used for IMAP authorization identities, as IMAP/POP3/SMTP/ 971 ManageSieve are frequently implemented together. 973 Internal Structure: None 975 User Output: Unlikely, but possible. For example, if it is the same 976 as an email address. 978 Operations: - Sometimes concatenated with other data and then used 979 as input to a cryptographic hash function 981 How much tolerance for change from existing stringprep approach? Not 982 sure. 984 Background information: In RFC 5034, when describing the POP3 AUTH 985 command: The authorization identity generated by the SASL exchange 986 is a simple username, and SHOULD use the SASLprep profile (see 987 [RFC4013]) of the StringPrep algorithm (see [RFC3454]) to prepare 988 these names for matching. If preparation of the authorization 989 identity fails or results in an empty string (unless it was 990 transmitted as the empty string), the server MUST fail the 991 authentication. In RFC 4954, when describing the SMTP AUTH 992 command: The authorization identity generated by this [SASL] 993 exchange is a "simple username" (in the sense defined in 994 [SASLprep]), and both client and server SHOULD (*) use the 995 [SASLprep] profile of the [StringPrep] algorithm to prepare these 996 names for transmission or comparison. If preparation of the 997 authorization identity fails or results in an empty string (unless 998 it was transmitted as the empty string), the server MUST fail the 999 authentication. (*) Note: Future revision of this specification 1000 may change this requirement to MUST. Currently, the SHOULD is 1001 used in order to avoid breaking the majority of existing 1002 implementations. In RFC 5804, when describing the ManageSieve 1003 AUTHENTICATE command: The authorization identity generated by this 1004 [SASL] exchange is a "simple username" (in the sense defined in 1005 [SASLprep]), and both client and server MUST use the [SASLprep] 1006 profile of the [StringPrep] algorithm to prepare these names for 1007 transmission or comparison. If preparation of the authorization 1008 identity fails or results in an empty string (unless it was 1009 transmitted as the empty string), the server MUST fail the 1010 authentication. 1012 B.3. IMAP Stringprep Profiles: RFC5738, RFC4314: Usernames 1014 Evaluation Note These documents have 2 types of strings (usernames 1015 and passwords), so there are two separate templates. 1017 Description: "username" parameter to the IMAP LOGIN command, 1018 identifiers in IMAP ACL commands. Note that any valid username is 1019 also an IMAP ACL identifier, but IMAP ACL identifiers can include 1020 other things like name of group of users. 1022 How It's Used: Used for authentication (Usernames), or in IMAP 1023 Access Control Lists (Usernames or Group names) 1025 Who Generates It: - Typically generated by email system 1026 administrators using some tools/conventions, sometimes from some 1027 backend database. - In some setups human users can register own 1028 usernames (e.g. webmail self registration) 1030 User Input Methods: - Typed by user / selected from a list - Copy- 1031 and-paste - Perhaps voice input - Can also be specified in 1032 configuration files or on a command line 1034 Enforcement: - Rules enforced by server / add-on service (e.g., 1035 gateway service) on registration of account 1037 Comparison Method: Type 1" (byte-for-byte) or "type 2" (compare by a 1038 common algorithm that everyone agrees on (e.g., normalize and then 1039 compare the result byte-by-byte)) 1041 Case Folding, Sensitivity, Preservation: - Most likely case 1042 sensitive. Exact requirements on case-sensitivity/ 1043 case-preservation depend on a specific implementation, e.g. an 1044 implementation might treat all user identifiers as case 1045 insensitive (or case insensitive for US-ASCII subset only). 1047 Impact of Comparison: False positives: - an unauthorized user is 1048 allowed IMAP access (login) - improperly grant privileges (e.g., 1049 access to a specific mailbox, ability to manage ACLs for a 1050 mailbox) False negatives: - an authorized user is denied IMAP 1051 access - unable to use granted privileges (e.g., access to a 1052 specific mailbox, ability to manage ACLs for a mailbox) 1054 Normalization: NFKC (as per RFC 4013) 1056 Mapping: (see Section 2 of RFC 4013 for the full list): non ASCII 1057 spaces are mapped to space 1059 Disallowed Characters: (see Section 2 of RFC 4013 for the full 1060 list): Unicode Control characters, etc. 1062 String Classes: - simple username. See Section 2 of RFC 4013 for 1063 details on restrictions. Note that some implementations allow 1064 spaces in these. While IMAP implementations are not required to 1065 use a specific format, an IMAP username frequently has the same 1066 format as an email address (and EAI email address in the future), 1067 or as a left hand side of an email address. Note: whatever is 1068 recommended for IMAP username should also be used for ManageSieve, 1069 POP3 and SMTP authorization identities, as IMAP/POP3/SMTP/ 1070 ManageSieve are frequently implemented together. 1072 Internal Structure: None 1074 User Output: Unlikely, but possible. For example, if it is the same 1075 as an email address. - access control lists (e.g. in IMAP ACL 1076 extension), both when managing membership and listing membership 1077 of existing access control lists. - often show up as mailbox names 1078 (under Other Users IMAP namespace) 1080 Operations: - Sometimes concatenated with other data and then used 1081 as input to a cryptographic hash function 1083 How much tolerance for change from existing stringprep approach? Not 1084 sure. Non-ASCII IMAP usernames are currently prohibited by IMAP 1085 (RFC 3501). However they are allowed when used in IMAP ACL 1086 extension. 1088 B.4. IMAP Stringprep Profiles: RFC5738: Passwords 1090 Description: "Password" parameter to the IMAP LOGIN command 1091 How It's Used: Used for authentication (Passwords) 1093 Who Generates It: Either generated by email system administrators 1094 using some tools/conventions, or specified by the human user. 1096 User Input Methods: - Typed by user - Copy-and-paste - Perhaps voice 1097 input - Can also be specified in configuration files or on a 1098 command line 1100 Enforcement: Rules enforced by server / add-on service (e.g., 1101 gateway service or backend databse) on registration of account 1103 Comparison Method: "Type 1" (byte-for-byte) 1105 Case Folding, Sensitivity, Preservation: Most likely case sensitive. 1107 Impact of Comparison: False positives: - an unauthorized user is 1108 allowed IMAP access (login) False negatives: - an authorized user 1109 is denied IMAP access 1111 Normalization: NFKC (as per RFC 4013) 1113 Mapping: (see Section 2 of RFC 4013 for the full list): non ASCII 1114 spaces are mapped to space 1116 Disallowed Characters: (see Section 2 of RFC 4013 for the full 1117 list): Unicode Control characters, etc. 1119 String Classes: Currently defined as "simple username" (see Section 1120 2 of RFC 4013 for details on restrictions.), however this is 1121 likely to be a different class from usernames. Note that some 1122 implementations allow spaces in these. Password in all email 1123 related protocols should be treated in the same way. Same 1124 passwords are frequently shared with web, IM, etc. applications. 1126 Internal Structure: None 1128 User Output: - text of email messages (e.g. in "you forgot your 1129 password" email messages) - web page / directory - side of the bus 1130 / in ads -- possible 1132 Operations: Sometimes concatenated with other data and then used as 1133 input to a cryptographic hash function. Frequently stored as is, 1134 or hashed. 1136 How much tolerance for change from existing stringprep approach? Not 1137 sure. Non-ASCII IMAP passwords are currently prohibited by IMAP 1138 (RFC 3501), however they are likely to be in widespread use. 1140 Background information: RFC 5738 (IMAP INTERNATIONALIZATION): 5. 1141 UTF8=USER Capability If the "UTF8=USER" capability is advertised, 1142 that indicates the server accepts UTF-8 user names and passwords 1143 and applies SASLprep [RFC4013] to both arguments of the LOGIN 1144 command. The server MUST reject UTF-8 that fails to comply with 1145 the formal syntax in RFC 3629 [RFC3629] or if it encounters 1146 Unicode characters listed in Section 2.3 of SASLprep RFC 4013 1147 [RFC4013]. RFC 4314 (IMAP4 Access Control List (ACL) Extension): 1148 3. Access control management commands and responses Servers, when 1149 processing a command that has an identifier as a parameter (i.e., 1150 any of SETACL, DELETEACL, and LISTRIGHTS commands), SHOULD first 1151 prepare the received identifier using "SASLprep" profile 1152 [SASLprep] of the "stringprep" algorithm [Stringprep]. If the 1153 preparation of the identifier fails or results in an empty string, 1154 the server MUST refuse to perform the command with a BAD response. 1155 Note that Section 6 recommends additional identifier's 1156 verification steps. and in Section 6: This document relies on 1157 [SASLprep] to describe steps required to perform identifier 1158 canonicalization (preparation). The preparation algorithm in 1159 SASLprep was specifically designed such that its output is 1160 canonical, and it is well-formed. However, due to an anomaly 1161 [PR29] in the specification of Unicode normalization, canonical 1162 equivalence is not guaranteed for a select few character 1163 sequences. Identifiers prepared with SASLprep can be stored and 1164 returned by an ACL server. The anomaly affects ACL manipulation 1165 and evaluation of identifiers containing the selected character 1166 sequences. These sequences, however, do not appear in well-formed 1167 text. In order to address this problem, an ACL server MAY reject 1168 identifiers containing sequences described in [PR29] by sending 1169 the tagged BAD response. This is in addition to the requirement 1170 to reject identifiers that fail SASLprep preparation as described 1171 in Section 3. 1173 B.5. Anonymous SASL Stringprep Profiles: RFC4505 1175 Description: RFC 4505 defines a "trace" field: 1177 Comparison: this field is not intended for comparison (only used for 1178 logging) 1180 Case folding; case sensitivity, preserve case: No case folding/case 1181 sensitive 1183 Do users input the strings directly? Yes. Possibly entered in 1184 configuration UIs, or on a command line. Can also be stored in 1185 configuration files. The value can also be automatically 1186 generated by clients (e.g. a fixed string is used, or a user's 1187 email address). 1189 How users input strings? Keyboard/voice, stylus (pick from a list). 1190 Copy-paste - possibly. 1192 Normalization: None 1194 Disallowed Characters Control characters are disallowed. (See 1195 Section 3 of RFC 4505) 1197 Which other strings or identifiers are these most similar to? RFC 1198 4505 says that the trace "should take one of two forms: an 1199 Internet email address, or an opaque string that does not contain 1200 the '@' U+0040) character and that can be interpreted by the 1201 system administrator of the client's domain." In practice, this 1202 is a freeform text, so it belongs to a different class from "email 1203 address" or "username". 1205 Are these strings or identifiers sometimes the same as strings or 1206 identifiers from other protocols (e.g., does an IM system sometimes 1207 use the same credentials database for authentication as an email 1208 system)? Yes: see above. However there is no strong need to keep 1209 them consistent in the future. 1211 How are users exposed to these strings, how are they published? No. 1212 However, The value can be seen in server logs 1214 Impacts of false positives and false negatives: False positive: a 1215 user can be confused with another user. False negative: two 1216 distinct users are treated as the same user. But note that the 1217 trace field is not authenticated, so it can be easily falsified. 1219 Tolerance of changes in the community The community would be 1220 flexible. 1222 Delimiters No internal structure, but see comments above about 1223 frequent use of email addresses. 1225 Background information: The Anonymous Mechanism The mechanism 1226 consists of a single message from the client to the server. The 1227 client may include in this message trace information in the form 1228 of a string of [UTF-8]-encoded [Unicode] characters prepared in 1229 accordance with [StringPrep] and the "trace" stringprep profile 1230 defined in Section 3 of this document. The trace information, 1231 which has no semantical value, should take one of two forms: an 1232 Internet email address, or an opaque string that does not contain 1233 the '@' (U+0040) character and that can be interpreted by the 1234 system administrator of the client's domain. For privacy reasons, 1235 an Internet email address or other information identifying the 1236 user should only be used with permission from the user. 3. The 1237 "trace" Profile of "Stringprep" This section defines the "trace" 1238 profile of [StringPrep]. This profile is designed for use with 1239 the SASL ANONYMOUS Mechanism. Specifically, the client is to 1240 prepare the message production in accordance with this profile. 1241 The character repertoire of this profile is Unicode 3.2 [Unicode]. 1242 No mapping is required by this profile. No Unicode normalization 1243 is required by this profile. The list of unassigned code points 1244 for this profile is that provided in Appendix A of [StringPrep]. 1245 Unassigned code points are not prohibited. Characters from the 1246 following tables of [StringPrep] are prohibited: - C.2.1 (ASCII 1247 control characters) - C.2.2 (Non-ASCII control characters) - C.3 1248 (Private use characters) - C.4 (Non-character code points) - C.5 1249 (Surrogate codes) - C.6 (Inappropriate for plain text) - C.8 1250 (Change display properties are deprecated) - C.9 (Tagging 1251 characters) No additional characters are prohibited. This profile 1252 requires bidirectional character checking per Section 6 of 1253 [StringPrep]. 1255 B.6. XMPP Stringprep Profiles: RFC3920 Nodeprep 1257 Description: Localpart of JabberID ("JID"), as in: 1258 localpart@domainpart/resourcepart 1260 How It's Used: - Usernames (e.g., stpeter@jabber.org) - Chatroom 1261 names (e.g., precis@jabber.ietf.org) - Publish-subscribe nodes - 1262 Bot names 1264 Who Generates It: - Typically, end users via an XMPP client - 1265 Sometimes created in an automated fashion 1267 User Input Methods: - Typed by user - Copy-and-paste - Perhaps voice 1268 input - Clicking a URI/IRI 1270 Enforcement: - Rules enforced by server / add-on service (e.g., 1271 chatroom service) on registration of account, creation of room, 1272 etc. 1274 Comparison Method: "Type 2" (common algorithm) 1275 Case Folding, Sensitivity, Preservation: - Strings are always folded 1276 to lowercase - Case is not preserved 1278 Impact of Comparison: False positives: - unable to authenticate at 1279 server (or authenticate to wrong account) - add wrong person to 1280 buddy list - join the wrong chatroom - improperly grant privileges 1281 (e.g., chatroom admin) - subscribe to wrong pubsub node - interact 1282 with wrong bot - allow communication with blocked entity False 1283 negatives: - unable to authenticate - unable to add someone to 1284 buddy list - unable to join desired chatroom - unable to use 1285 granted privileges (e.g., chatroom admin) - unable to subscribe to 1286 desired pubsub node - unable to interact with desired bot - 1287 disallow communication with unblocked entity 1289 Normalization: NFKC 1291 Mapping: Spaces are mapped to nothing 1293 Disallowed Characters: ",&,',/,:,<,>,@ 1295 String Classes: - Often similar to generic username - Often similar 1296 to localpart of email address - Sometimes same as localpart of 1297 email address 1299 Internal Structure: None 1301 User Output: - vCard - email signature - web page / directory - text 1302 of message (e.g., in a chatroom) 1304 Operations: - Sometimes concatenated with other data and then used 1305 as input to a cryptographic hash function 1307 B.7. XMPP Stringprep Profiles: RFC3920 Resourceprep 1309 Description: - Resourcepart of JabberID ("JID"), as in: 1310 localpart@domainpart/resourcepart - Typically free-form text 1312 How It's Used: - Device / session names (e.g., 1313 stpeter@jabber.org/Home) - Nicknames (e.g., 1314 precis@jabber.ietf.org/StPeter) 1316 Who Generates It: - Often human users via an XMPP client - Often 1317 generated in an automated fashion by client or server 1319 User Input Methods: - Typed by user - Copy-and-paste - Perhaps voice 1320 input - Clicking a URI/IRI 1322 Enforcement: - Rules enforced by server / add-on service (e.g., 1323 chatroom service) on account login, joining a chatroom, etc. 1325 Comparison Method: "Type 2" (byte-for-byte) 1327 Case Folding, Sensitivity, Preservation: - Strings are never folded 1328 - Case is preserved 1330 Impact of Comparison: False positives: - interact with wrong device 1331 (e.g., for file transfer or voice call) - interact with wrong 1332 chatroom participant - improperly grant privileges (e.g., chatroom 1333 moderator) - allow communication with blocked entity False 1334 negatives: - unable to choose desired chatroom nick - unable to 1335 use granted privileges (e.g., chatroom moderator) - disallow 1336 communication with unblocked entity 1338 Normalization: NFKC 1340 Mapping: Spaces are mapped to nothing 1342 Disallowed Characters: None 1344 String Classes: Basically a free-form identifier 1346 Internal Structure: None 1348 User Output: - text of message (e.g., in a chatroom) - device names 1349 often not exposed to human users 1351 Operations: Sometimes concatenated with other data and then used as 1352 input to a cryptographic hash function 1354 B.8. EAP Stringprep Profiles: RFC3748 1356 Description: RFC 3748 section 5 references Stringprep, but the WG 1357 did not agree with the text (was added by IESG) and there are no 1358 known implementations that use Stringprep. The main problem with 1359 that text is that the use of strings is a per-method concept, not 1360 a generic EAP concept and so RFC 3748 itself does not really use 1361 Stringprep, but individual EAP methods could. As such, the 1362 answers to the template questions are mostly not applicable, but a 1363 few answers are universal across methods. The list of IANA 1364 registered EAP methods is at http://www.iana.org/assignments/ 1365 eap-numbers/eap-numbers.xml#eap-numbers-3 1367 Comparison Methods: n/a (per-method) 1369 Case Folding, Case Sensitivity, Case Preservation: n/a (per-method) 1371 Impact of comparison: A false positive results in unauthorized 1372 network access (and possibly theft of service if some else is 1373 billed). A false negative results in lack of authorized network 1374 access (no connectivity). 1376 User input: n/a (per-method) 1378 Normalization: n/a (per-method) 1380 Mapping: n/a (per-method) 1382 Disallowed characters: n/a (per-method) 1384 String classes: Although some EAP methods may use a syntax similar 1385 to other types of identifiers, EAP mandates that the actual values 1386 must not be assumed to be identifiers usable with anything else. 1388 Internal structure: n/a (per-method) 1390 User output: Identifiers are never human displayed except perhaps as 1391 they're typed by a human. 1393 Operations: n/a (per-method) 1395 Community considerations: There is no resistance to change for the 1396 base EAP protocol (as noted, the WG didn't want the existing 1397 text). However actual use of stringprep, if any, within specific 1398 EAP methods may have resistance. It is currently unknown whether 1399 any EAP methods use stringprep. 1401 Appendix C. Changes between versions 1403 Note to RFC Editor: This section should be removed prior to 1404 publication. 1406 C.1. 00 1408 First WG version. Based on 1409 draft-blanchet-precis-problem-statement-00. 1411 C.2. 01 1413 o Made clear that the document is talking only about Unicode code 1414 points, and not any particular encoding. 1416 o Substantially reorganized the document along the lines of the 1417 review template at . 1420 o Included specific questions for each topic for consideration. 1422 o Moved spot for individual protocol review to appendix. Not 1423 populated yet. 1425 C.3. 02 1427 o Cleared up details of comparison classes 1429 o Added a section on changes in Unicode 1431 C.4. 03 1433 o Aligned comparison discussion with identifier discussion from 1434 draft-iab-identifier-comparison-00 1436 o Added section on classes of strings ("Namey" and so on) 1438 C.5. 04 1440 Keepalive version 1442 C.6. 05 1444 o Changed classes of strings to align with framework doc 1446 o Altered table in Appendix A 1447 o Added all profiles evaluations from the wg wiki in appendix B 1449 C.7. 06 1451 o Respond to comments received in WGLC 1453 o Removed classes of strings (also from Appendix A) 1455 o Moved inclusion/exclusion distinction to Introduction 1457 o Fix some sentences to clarify terminology and add or fix 1458 references 1460 Authors' Addresses 1462 Marc Blanchet 1463 Viagenie 1464 246 Aberdeen 1465 Quebec, QC G1R 2E1 1466 Canada 1468 Email: Marc.Blanchet@viagenie.ca 1469 URI: http://viagenie.ca 1471 Andrew Sullivan 1472 Dyn, Inc. 1473 150 Dow St 1474 Manchester, NH 03101 1475 U.S.A. 1477 Email: asullivan@dyn.com