idnits 2.17.00 (12 Aug 2021) /tmp/idnits16989/draft-ietf-idnabis-rationale-16.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** You're using the IETF Trust Provisions' Section 6.b License Notice from 12 Sep 2009 rather than the newer Notice from 28 Dec 2009. (See https://trustee.ietf.org/license-info/) 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 86: '... CONTEXTUAL RULE REQUIRED . . . . . . ...' RFC 2119 keyword, line 476: '... CONTEXTUAL RULE REQUIRED, DISALLOWED ...' RFC 2119 keyword, line 502: '...3.1.2. CONTEXTUAL RULE REQUIRED...' RFC 2119 keyword, line 1639: '... CONTEXTUAL RULE REQUIRED categories c...' RFC 2119 keyword, line 2144: '... MUST, SHOULD, or MAY) and small ...' Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document seems to contain a disclaimer for pre-RFC5378 work, and may have content which was first submitted before 10 November 2008. The disclaimer is necessary when there are original authors that you have been unable to contact, or if some do not wish to grant the BCP78 rights to the IETF Trust. If you are able to get all authors (current and original) to grant those rights, you can and should remove the disclaimer; otherwise, the disclaimer is needed and you can ignore this comment. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (January 7, 2010) is 4516 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Unused Reference: 'ASCII' is defined on line 1891, but no explicit reference was found in the text == Unused Reference: 'Unicode-UAX15' is defined on line 1932, but no explicit reference was found in the text == Unused Reference: 'BIG5' is defined on line 1948, but no explicit reference was found in the text == Unused Reference: 'GB18030' is defined on line 1957, but no explicit reference was found in the text == Unused Reference: 'RFC0810' is defined on line 1967, but no explicit reference was found in the text == Unused Reference: 'RFC2673' is defined on line 1995, but no explicit reference was found in the text == Unused Reference: 'RFC4690' is defined on line 2025, but no explicit reference was found in the text == Unused Reference: 'Unicode-UTR36' is defined on line 2043, but no explicit reference was found in the text == Outdated reference: draft-ietf-idnabis-defs has been published as RFC 5890 == Outdated reference: draft-ietf-idnabis-tables has been published as RFC 5892 ** Obsolete normative reference: RFC 3490 (Obsoleted by RFC 5890, RFC 5891) -- No information found for draft-ietf-idnabis-mapping - is the name correct? -- Obsolete informational reference (is this intentional?): RFC 810 (Obsoleted by RFC 952) -- Obsolete informational reference (is this intentional?): RFC 2671 (Obsoleted by RFC 6891) -- Obsolete informational reference (is this intentional?): RFC 2673 (Obsoleted by RFC 6891) -- Obsolete informational reference (is this intentional?): RFC 3454 (Obsoleted by RFC 7564) -- Obsolete informational reference (is this intentional?): RFC 3491 (Obsoleted by RFC 5891) Summary: 3 errors (**), 0 flaws (~~), 11 warnings (==), 8 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group J. Klensin 3 Internet-Draft January 7, 2010 4 Intended status: Informational 5 Expires: July 11, 2010 7 Internationalized Domain Names for Applications (IDNA): Background, 8 Explanation, and Rationale 9 draft-ietf-idnabis-rationale-16.txt 11 Abstract 13 Several years have passed since the original protocol for 14 Internationalized Domain Names (IDNs) was completed and deployed. 15 During that time, a number of issues have arisen, including the need 16 to update the system to deal with newer versions of Unicode. Some of 17 these issues require tuning of the existing protocols and the tables 18 on which they depend. This document provides an overview of a 19 revised system and provides explanatory material for its components. 21 Status of this Memo 23 This Internet-Draft is submitted to IETF in full conformance with the 24 provisions of BCP 78 and BCP 79. 26 Internet-Drafts are working documents of the Internet Engineering 27 Task Force (IETF), its areas, and its working groups. Note that 28 other groups may also distribute working documents as Internet- 29 Drafts. 31 Internet-Drafts are draft documents valid for a maximum of six months 32 and may be updated, replaced, or obsoleted by other documents at any 33 time. It is inappropriate to use Internet-Drafts as reference 34 material or to cite them other than as "work in progress." 36 The list of current Internet-Drafts can be accessed at 37 http://www.ietf.org/ietf/1id-abstracts.txt. 39 The list of Internet-Draft Shadow Directories can be accessed at 40 http://www.ietf.org/shadow.html. 42 This Internet-Draft will expire on July 11, 2010. 44 Copyright Notice 46 Copyright (c) 2010 IETF Trust and the persons identified as the 47 document authors. All rights reserved. 49 This document is subject to BCP 78 and the IETF Trust's Legal 50 Provisions Relating to IETF Documents 51 (http://trustee.ietf.org/license-info) in effect on the date of 52 publication of this document. Please review these documents 53 carefully, as they describe your rights and restrictions with respect 54 to this document. Code Components extracted from this document must 55 include Simplified BSD License text as described in Section 4.e of 56 the Trust Legal Provisions and are provided without warranty as 57 described in the BSD License. 59 This document may contain material from IETF Documents or IETF 60 Contributions published or made publicly available before November 61 10, 2008. The person(s) controlling the copyright in some of this 62 material may not have granted the IETF Trust the right to allow 63 modifications of such material outside the IETF Standards Process. 64 Without obtaining an adequate license from the person(s) controlling 65 the copyright in such materials, this document may not be modified 66 outside the IETF Standards Process, and derivative works of it may 67 not be created outside the IETF Standards Process, except to format 68 it for publication as an RFC or to translate it into languages other 69 than English. 71 Table of Contents 73 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 74 1.1. Context and Overview . . . . . . . . . . . . . . . . . . . 4 75 1.2. Discussion Forum . . . . . . . . . . . . . . . . . . . . . 5 76 1.3. Terminology . . . . . . . . . . . . . . . . . . . . . . . 5 77 1.3.1. DNS "Name" Terminology . . . . . . . . . . . . . . . . 5 78 1.3.2. New Terminology and Restrictions . . . . . . . . . . . 6 79 1.4. Objectives . . . . . . . . . . . . . . . . . . . . . . . . 6 80 1.5. Applicability and Function of IDNA . . . . . . . . . . . . 7 81 1.6. Comprehensibility of IDNA Mechanisms and Processing . . . 8 82 2. Processing in IDNA2008 . . . . . . . . . . . . . . . . . . . . 9 83 3. Permitted Characters: An Inclusion List . . . . . . . . . . . 9 84 3.1. A Tiered Model of Permitted Characters and Labels . . . . 10 85 3.1.1. PROTOCOL-VALID . . . . . . . . . . . . . . . . . . . . 10 86 3.1.2. CONTEXTUAL RULE REQUIRED . . . . . . . . . . . . . . . 11 87 3.1.2.1. Contextual Restrictions . . . . . . . . . . . . . 11 88 3.1.2.2. Rules and Their Application . . . . . . . . . . . 12 89 3.1.3. DISALLOWED . . . . . . . . . . . . . . . . . . . . . . 12 90 3.1.4. UNASSIGNED . . . . . . . . . . . . . . . . . . . . . . 13 91 3.2. Registration Policy . . . . . . . . . . . . . . . . . . . 13 92 3.3. Layered Restrictions: Tables, Context, Registration, 93 Applications . . . . . . . . . . . . . . . . . . . . . . . 14 94 4. Issues that Constrain Possible Solutions . . . . . . . . . . . 15 95 4.1. Display and Network Order . . . . . . . . . . . . . . . . 15 96 4.2. Entry and Display in Applications . . . . . . . . . . . . 16 97 4.3. Linguistic Expectations: Ligatures, Digraphs, and 98 Alternate Character Forms . . . . . . . . . . . . . . . . 18 99 4.4. Case Mapping and Related Issues . . . . . . . . . . . . . 20 100 4.5. Right to Left Text . . . . . . . . . . . . . . . . . . . . 21 101 5. IDNs and the Robustness Principle . . . . . . . . . . . . . . 21 102 6. Front-end and User Interface Processing for Lookup . . . . . . 22 103 7. Migration from IDNA2003 and Unicode Version Synchronization . 24 104 7.1. Design Criteria . . . . . . . . . . . . . . . . . . . . . 24 105 7.1.1. Summary and Discussion of IDNA Validity Criteria . . . 24 106 7.1.2. Labels in Registration . . . . . . . . . . . . . . . . 25 107 7.1.3. Labels in Lookup . . . . . . . . . . . . . . . . . . . 26 108 7.2. Changes in Character Interpretations . . . . . . . . . . . 28 109 7.2.1. Character Changes: Eszett and Final Sigma . . . . . . 28 110 7.2.2. Character Changes: Zero-Width Joiner and Non-Joiner . 28 111 7.2.3. Character Changes and the Need for Transition . . . . 28 112 7.2.4. Transition Strategies . . . . . . . . . . . . . . . . 29 113 7.3. Elimination of Character Mapping . . . . . . . . . . . . . 30 114 7.4. The Question of Prefix Changes . . . . . . . . . . . . . . 30 115 7.4.1. Conditions Requiring a Prefix Change . . . . . . . . . 31 116 7.4.2. Conditions Not Requiring a Prefix Change . . . . . . . 31 117 7.4.3. Implications of Prefix Changes . . . . . . . . . . . . 32 118 7.5. Stringprep Changes and Compatibility . . . . . . . . . . . 32 119 7.6. The Symbol Question . . . . . . . . . . . . . . . . . . . 33 120 7.7. Migration Between Unicode Versions: Unassigned Code 121 Points . . . . . . . . . . . . . . . . . . . . . . . . . . 34 122 7.8. Other Compatibility Issues . . . . . . . . . . . . . . . . 36 123 8. Name Server Considerations . . . . . . . . . . . . . . . . . . 36 124 8.1. Processing Non-ASCII Strings . . . . . . . . . . . . . . . 36 125 8.2. Root and other DNS Server Considerations . . . . . . . . . 37 126 9. Internationalization Considerations . . . . . . . . . . . . . 37 127 10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 38 128 10.1. IDNA Character Registry . . . . . . . . . . . . . . . . . 38 129 10.2. IDNA Context Registry . . . . . . . . . . . . . . . . . . 38 130 10.3. IANA Repository of IDN Practices of TLDs . . . . . . . . . 38 131 11. Security Considerations . . . . . . . . . . . . . . . . . . . 38 132 11.1. General Security Issues with IDNA . . . . . . . . . . . . 38 133 12. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 39 134 13. Contributors . . . . . . . . . . . . . . . . . . . . . . . . . 39 135 14. References . . . . . . . . . . . . . . . . . . . . . . . . . . 40 136 14.1. Normative References . . . . . . . . . . . . . . . . . . . 40 137 14.2. Informative References . . . . . . . . . . . . . . . . . . 41 138 Appendix A. Change Log . . . . . . . . . . . . . . . . . . . . . 43 139 A.1. Changes between Version -00 and Version -01 of 140 draft-ietf-idnabis-rationale . . . . . . . . . . . . . . . 43 141 A.2. Version -02 . . . . . . . . . . . . . . . . . . . . . . . 44 142 A.3. Version -03 . . . . . . . . . . . . . . . . . . . . . . . 44 143 A.4. Version -04 . . . . . . . . . . . . . . . . . . . . . . . 45 144 A.5. Version -05 . . . . . . . . . . . . . . . . . . . . . . . 45 145 A.6. Version -06 . . . . . . . . . . . . . . . . . . . . . . . 45 146 A.7. Version -07 . . . . . . . . . . . . . . . . . . . . . . . 46 147 A.8. Version -08 . . . . . . . . . . . . . . . . . . . . . . . 46 148 A.9. Version -09 . . . . . . . . . . . . . . . . . . . . . . . 46 149 A.10. Version -10 . . . . . . . . . . . . . . . . . . . . . . . 47 150 A.11. Version -11 . . . . . . . . . . . . . . . . . . . . . . . 47 151 A.12. Version -12 . . . . . . . . . . . . . . . . . . . . . . . 48 152 A.13. Version -13 . . . . . . . . . . . . . . . . . . . . . . . 48 153 A.14. Version -14 . . . . . . . . . . . . . . . . . . . . . . . 48 154 A.15. Version -15 . . . . . . . . . . . . . . . . . . . . . . . 49 155 A.16. Version -16 . . . . . . . . . . . . . . . . . . . . . . . 49 156 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . . 49 158 1. Introduction 160 1.1. Context and Overview 162 Internationalized Domain Names in Applications (IDNA) is a collection 163 of standards that allow client applications to convert some Unicode 164 mnemonics to an ASCII-compatible encoding form ("ACE") which is a 165 valid DNS label containing only letters, digits, and hyphens. The 166 specific form of ACE label used by IDNA is called an "A-label". A 167 client can look up an exact A-label in the existing DNS, so A-labels 168 do not require any extensions to DNS, upgrades of DNS servers or 169 updates to low-level client libraries. An A-label is recognizable 170 from the prefix "xn--" before the characters produced by the Punycode 171 algorithm [RFC3492], thus a user application can identify an A-label 172 and convert it into Unicode (or some local coded character set) for 173 display. 175 On the registry side, IDNA allows a registry to offer 176 Internationalized Domain Names (IDNs) for registration as A-labels. 177 A registry may offer any subset of valid IDNs, and may apply any 178 restrictions or bundling (grouping of similar labels together in one 179 registration) appropriate for the context of that registry. 180 Registration of labels is sometimes discussed separately from lookup, 181 and is subject to a few specific requirements that do not apply to 182 lookup. 184 DNS clients and registries are subject to some differences in 185 requirements for handling IDNs. In particular, registries are urged 186 to register only exact, valid A-labels, while clients might do some 187 mapping to get from otherwise-invalid user input to a valid A-label. 189 The first version of IDNA was published in 2003 and is referred to 190 here as IDNA2003 to contrast it with the current version, which is 191 known as IDNA2008 (after the year in which IETF work started on it). 192 IDNA2003 consists of four documents: the IDNA base specification 193 [RFC3490], Nameprep [RFC3491], Punycode [RFC3492], and Stringprep 194 [RFC3454]. The current set of documents, IDNA2008, are not dependent 195 on any of the IDNA2003 specifications other than the one for Punycode 196 encoding. References to "IDNA2008", "these specifications", or 197 "these documents" are to the entire IDNA2008 set listed in 198 [IDNA2008-Defs]. The characters that are valid in A-labels are 199 identified from rules listed in the Tables document 200 [IDNA2008-Tables], but validity can be derived from the Unicode 201 properties of those characters with a very few exceptions. 203 Traditionally, DNS labels are matched case-insensitively 204 [RFC1034][RFC1035]. That convention was preserved in IDNA2003 by a 205 case-folding operation that generally maps capital letters into 206 lower-case ones. However, if case rules are enforced from one 207 language, another language sometimes loses the ability to treat two 208 characters separately. Case-insensitivity is treated slightly 209 differently in IDNA2008. 211 IDNA2003 used Unicode version 3.2 only. In order to keep up with new 212 characters added in new versions of UNICODE, IDNA2008 decouples its 213 rules from any particular version of UNICODE. Instead, the 214 attributes of new characters in Unicode, supplemented by a small 215 number of exception cases, determine how and whether the characters 216 can be used in IDNA labels. 218 This document provides informational context for IDNA2008, including 219 terminology, background, and policy discussions. It contains no 220 normative material; specifications for conformance to the IDNA2008 221 protocols appears entirely in the other documents in the series. 223 1.2. Discussion Forum 225 [[ RFC Editor: please remove this section. ]] 227 IDNA2008 is being discussed in the IETF "idnabis" Working Group and 228 on the mailing list idna-update@alvestrand.no 230 1.3. Terminology 232 Terminology for IDNA2008 appears in [IDNA2008-Defs]. That document 233 also contains a roadmap to the IDNA2008 document collection. No 234 attempt should be made to understand this document without the 235 definitions and concepts that appear there. 237 1.3.1. DNS "Name" Terminology 239 In the context of IDNs, the DNS term "name" has introduced some 240 confusion as people speak of DNS labels in terms of the words or 241 phrases of various natural languages. Historically, many of the 242 "names" in the DNS have been mnemonics to identify some particular 243 concept, object, or organization. They are typically rooted in some 244 language because most people think in language-based ways. But, 245 because they are mnemonics, they need not obey the orthographic 246 conventions of any language: it is not a requirement that it be 247 possible for them to be "words". 249 This distinction is important because the reasonable goal of an IDN 250 effort is not to be able to write the great Klingon (or language of 251 one's choice) novel in DNS labels but to be able to form a usefully 252 broad range of mnemonics in ways that are as natural as possible in a 253 very broad range of scripts. 255 1.3.2. New Terminology and Restrictions 257 IDNA2008 introduces new terminology. Precise definitions are 258 provided (in [IDNA2008-Defs]), for the terms "U-label", "A-Label", 259 LDH-label (to which all valid pre-IDNA host names conformed), 260 Reserved-LDH-label (R-LDH-label), XN-label, Fake-A-Label, and Non- 261 Reserved-LDH-label (NR-LDH-label). 263 In addition, the term "putative label" has been adopted to refer to a 264 label that may appear to meet certain definitional constraints but 265 has not yet been sufficiently tested for validity. 267 These definitions are also illustrated in Figure 1 of the Definitions 268 Document [IDNA2008-Defs]. R-LDH-labels contain "--" in the third and 269 fourth character from the beginning of the label. In IDNA-aware 270 applications, only a subset of these reserved labels is permitted to 271 be used, namely the A-label subset. A-labels are a subset of the 272 R-LDH-labels that begin with the case-insensitive string "xn--". 273 Labels that bear this prefix but which are not otherwise valid fall 274 into the "Fake A-label" category. The non-reserved labels (NR-LDH- 275 labels) are implicitly valid since they do not bear any resemblance 276 to the labels specified by IDNA. 278 The creation of the Reserved-LDH category is required for three 279 reasons: 281 o to prevent confusion with pre-IDNA coding forms; 283 o to permit future extensions that would require changing the 284 prefix, no matter how unlikely those might be (see Section 7.4); 285 and 287 o to reduce the opportunities for attacks via the Punycode encoding 288 algorithm itself. 290 As with other documents in the IDNA2008 set, this document uses the 291 term "registry" to describe any zone in the DNS. That term, and the 292 terms "zone" or "zone administration", are interchangeable. 294 1.4. Objectives 296 These are the main objectives in revising IDNA. 298 o Use a more recent version of Unicode, and allow IDNA to be 299 independent of Unicode versions, so that IDNA2008 need not be 300 updated for implementations to adopt codepoints from new Unicode 301 versions. 303 o Fix a very small number of code-point categorizations that have 304 turned out to cause problems in the communities that use those 305 code-points. 307 o Reduce the dependency on mapping, in order that the pre-mapped 308 forms (which are not valid IDNA labels) tend to appear less often 309 in various contexts, in favor of valid A-labels. 311 o Fix some details in the bidirectional codepoint handling 312 algorithms. 314 1.5. Applicability and Function of IDNA 316 The IDNA specification solves the problem of extending the repertoire 317 of characters that can be used in domain names to include a large 318 subset of the Unicode repertoire. 320 IDNA does not extend DNS. Instead, the applications (and, by 321 implication, the users) continue to see an exact-match lookup 322 service. Either there is a single exactly-matching name (subject to 323 the base DNS requirement of case-insensitive ASCII matching) or there 324 is no match. This model has served the existing applications well, 325 but it requires, with or without internationalized domain names, that 326 users know the exact spelling of the domain names that are to be 327 typed into applications such as web browsers and mail user agents. 328 The introduction of the larger repertoire of characters potentially 329 makes the set of misspellings larger, especially given that in some 330 cases the same appearance, for example on a business card, might 331 visually match several Unicode code points or several sequences of 332 code points. 334 The IDNA standard does not require any applications to conform to it, 335 nor does it retroactively change those applications. An application 336 can elect to use IDNA in order to support IDN while maintaining 337 interoperability with existing infrastructure. If an application 338 wants to use non-ASCII characters in public DNS domain names, IDNA is 339 the only currently-defined option. Adding IDNA support to an 340 existing application entails changes to the application only, and 341 leaves room for flexibility in front-end processing and more 342 specifically in the user interface (see Section 6). 344 A great deal of the discussion of IDN solutions has focused on 345 transition issues and how IDNs will work in a world where not all of 346 the components have been updated. Proposals that were not chosen by 347 the original IDN Working Group would have depended on updating of 348 user applications, DNS resolvers, and DNS servers in order for a user 349 to apply an internationalized domain name in any form or coding 350 acceptable under that method. While processing must be performed 351 prior to or after access to the DNS, IDNA requires no changes to the 352 DNS protocol, any DNS servers, or the resolvers on users' computers. 354 IDNA allows the graceful introduction of IDNs not only by avoiding 355 upgrades to existing infrastructure (such as DNS servers and mail 356 transport agents), but also by allowing some limited use of IDNs in 357 applications by using the ASCII-encoded representation of the labels 358 containing non-ASCII characters. While such names are user- 359 unfriendly to read and type, and hence not optimal for user input, 360 they can be used as a last resort to allow rudimentary IDN usage. 361 For example, they might be the best choice for display if it were 362 known that relevant fonts were not available on the user's computer. 363 In order to allow user-friendly input and output of the IDNs and 364 acceptance of some characters as equivalent to those to be processed 365 according to the protocol, the applications need to be modified to 366 conform to this specification. 368 This version of IDNA uses the Unicode character repertoire, for 369 continuity with the original version of IDNA. 371 1.6. Comprehensibility of IDNA Mechanisms and Processing 373 One goal of IDNA2008, which is aided by the main goal of reducing the 374 dependency on mapping, is to improve the general understanding of how 375 IDNA works and what characters are permitted and what happens to 376 them. Comprehensibility and predictability to users and registrants 377 are important design goals for this effort. End-user applications 378 have an important role to play in increasing this comprehensibility. 380 Any system that tries to handle international characters encounters 381 some common problems. For example, a UI cannot display a character 382 if no font for that character is available. In some cases, 383 internationalization enables effective localization while maintaining 384 some global uniformity but losing some universality. 386 It is difficult to even make suggestions for end-user applications to 387 cope when characters and fonts are not available. Because display 388 functions are rarely controlled by the types of applications that 389 would call upon IDNA, such suggestions will rarely be very effective. 391 Converting between local character sets and normalized Unicode, if 392 needed, is part of this set of user agent issues. This conversion 393 introduces complexity in a system that is not Unicode-native. If a 394 label is converted to a local character set that does not have all 395 the needed characters, or that uses different character-coding 396 principles, the user agent may have to add special logic to avoid or 397 reduce loss of information. 399 The major difficulty may lie in accurately identifying the incoming 400 character set and applying the correct conversion routine. Even more 401 difficult, the local character coding system could be based on 402 conceptually different assumptions than those used by Unicode (e.g., 403 choice of font encodings used for publications in some Indic 404 scripts). Those differences may not easily yield unambiguous 405 conversions or interpretations even if each coding system is 406 internally consistent and adequate to represent the local language 407 and script. 409 IDNA2008 shifts responsibility for character mapping and other 410 adjustments from the protocol (where it was located in IDNA2003) to 411 pre-processing before invoking IDNA itself. The intent is that this 412 change will lead to greater usage of fully-valid A-Labels or U-labels 413 in display, transit and storage, which should aid comprehensibility 414 and predictability. A careful look at pre-processing raises issues 415 about what that pre-processing should do and at what point pre- 416 processing becomes harmful, how universally consistent pre-processing 417 algorithms can be, and how to be compatible with labels prepared in a 418 IDNA2003 context. Those issues are discussed in Section 6 and in the 419 separate document [IDNA2008-Mapping]. 421 2. Processing in IDNA2008 423 IDNA2008 separates Domain Name Registration and Lookup in the 424 protocol specification. Although most steps in the two processes are 425 similar, the separation reflects current practice in which per- 426 registry (DNS zone) restrictions and special processing are applied 427 at registration time but not during lookup. Another significant 428 benefit is that separation facilitates incremental addition of 429 permitted character groups to avoid freezing on one particular 430 version of Unicode. 432 The actual registration and lookup protocols for IDNA2008 are 433 specified in [IDNA2008-Protocol]. 435 3. Permitted Characters: An Inclusion List 437 IDNA2008 adopts the inclusion model. A code-point is assumed to be 438 invalid for IDN use unless it is included as part of a Unicode 439 property-based rule or, in rare cases, included individually by an 440 exception. When an implementation moves to a new version of Unicode, 441 the rules may indicate new valid code-points. 443 This section provides an overview of the model used to establish the 444 algorithm and character lists of [IDNA2008-Tables] and describes the 445 names and applicability of the categories used there. Note that the 446 inclusion of a character in the first category group (Section 3.1.1) 447 does not imply that it can be used indiscriminately; some characters 448 are associated with contextual rules that must be applied as well. 450 The information given in this section is provided to make the rules, 451 tables, and protocol easier to understand. The normative generating 452 rules that correspond to this informal discussion appear in 453 [IDNA2008-Tables] and the rules that actually determine what labels 454 can be registered or looked up are in [IDNA2008-Protocol]. 456 3.1. A Tiered Model of Permitted Characters and Labels 458 Moving to an inclusion model involves a new specification for the 459 list of characters that are permitted in IDNs. In IDNA2003, 460 character validity is independent of context and fixed forever (or 461 until the standard is replaced). However, globally context- 462 independent rules have proved to be impractical because some 463 characters, especially those that are called "Join_Controls" in 464 Unicode, are needed to make reasonable use of some scripts but have 465 no visible effect in others. IDNA2003 prohibited those types of 466 characters entirely by discarding them. We now have a consensus that 467 under some conditions, these "joiner" characters are legitimately 468 needed to allow useful mnemonics for some languages and scripts. In 469 general, context-dependent rules help deal with characters (generally 470 characters that would otherwise be prohibited entirely) that are used 471 differently or perceived differently across different scripts, and 472 allow the standard to be applied more appropriately in cases where a 473 string is not universally handled the same way. 475 IDNA2008 divides all possible Unicode code-points into four 476 categories: PROTOCOL-VALID, CONTEXTUAL RULE REQUIRED, DISALLOWED and 477 UNASSIGNED. 479 3.1.1. PROTOCOL-VALID 481 Characters identified as "PROTOCOL-VALID" (often abbreviated 482 "PVALID") are permitted in IDNs. Their use may be restricted by 483 rules about the context in which they appear or by other rules that 484 apply to the entire label in which they are to be embedded. For 485 example, any label that contains a character in this category that 486 has a "right-to-left" property must be used in context with the 487 "Bidi" rules (see [IDNA2008-Bidi]). 489 The term "PROTOCOL-VALID" is used to stress the fact that the 490 presence of a character in this category does not imply that a given 491 registry need accept registrations containing any of the characters 492 in the category. Registries are still expected to apply judgment 493 about labels they will accept and to maintain rules consistent with 494 those judgments (see [IDNA2008-Protocol] and Section 3.3). 496 Characters that are placed in the "PROTOCOL-VALID" category are 497 expected to never be removed from it or reclassified. While 498 theoretically characters could be removed from Unicode, such removal 499 would be inconsistent with the Unicode stability principles (see 500 [Unicode51], Appendix F) and hence should never occur. 502 3.1.2. CONTEXTUAL RULE REQUIRED 504 Some characters may be unsuitable for general use in IDNs but 505 necessary for the plausible support of some scripts. The two most 506 commonly-cited examples are the zero-width joiner and non-joiner 507 characters (ZWJ, U+200D and ZWNJ, U+200C) but other characters may 508 require special treatment because they would otherwise be DISALLOWED 509 (typically because Unicode considers them punctuation or special 510 symbols) but need to be permitted in limited contexts. Other 511 characters are given this special treatment because they pose 512 exceptional danger of being used to produce misleading labels or to 513 cause unacceptable ambiguity in label matching and interpretation. 515 3.1.2.1. Contextual Restrictions 517 Characters with contextual restrictions are identified as "CONTEXTUAL 518 RULE REQUIRED" and associated with a rule. The rule defines whether 519 the character is valid in a particular string, and also whether the 520 rule itself is to be applied on lookup as well as registration. 522 A distinction is made between characters that indicate or prohibit 523 joining and ones similar to them (known as "CONTEXT-JOINER" or 524 "CONTEXTJ") and other characters requiring contextual treatment 525 ("CONTEXT-OTHER" or "CONTEXTO"). Only the former require full 526 testing at lookup time. 528 It is important to note that these contextual rules cannot prevent 529 all uses of the relevant characters that might be confusing or 530 problematic. What they are expected to do is to confine 531 applicability of the characters to scripts (and narrower contexts) 532 where zone administrators are knowledgeable enough about the use of 533 those characters to be prepared to deal with them appropriately. 535 For example, a registry dealing with an Indic script that requires 536 ZWJ and/or ZWNJ as part of the writing system is expected to 537 understand where the characters have visible effect and where they do 538 not and to make registration rules accordingly. By contrast, a 539 registry dealing primarily with Latin or Cyrillic script might not be 540 actively aware that the characters exist, much less about the 541 consequences of embedding them in labels drawn from those scripts and 542 therefore should avoid accepting registrations containing those 543 characters, at least in Latin or Cyrillic-script labels. 545 3.1.2.2. Rules and Their Application 547 Rules have descriptions such as "Must follow a character from Script 548 XYZ", "Must occur only if the entire label is in Script ABC", or 549 "Must occur only if the previous and subsequent characters have the 550 DFG property". The actual rules may be DEFINED or NULL. If present, 551 they may have values of "True" (character may be used in any position 552 in any label), "False" (character may not be used in any label), or 553 may be a set of procedural rules that specify the context in which 554 the character is permitted. 556 Because it is easier to identify these characters than to know that 557 they are actually needed in IDNs or how to establish exactly the 558 right rules for each one, a rule may have a null value in a given 559 version of the tables. Characters associated with null rules are not 560 permitted to appear in putative labels for either registration or 561 lookup. Of course, a later version of the tables might contain a 562 non-null rule. 564 The actual rules and their descriptions are in Sections 2 and 3 of 565 [IDNA2008-Tables]. That document also specifies the creation of a 566 registry for future rules. 568 3.1.3. DISALLOWED 570 Some characters are inappropriate for use in IDNs and are thus 571 excluded for both registration and lookup (i.e., IDNA-conforming 572 applications performing name lookup should verify that these 573 characters are absent; if they are present, the label strings should 574 be rejected rather than converted to A-labels and looked up. Some of 575 these characters are problematic for use in IDNs (such as the 576 FRACTION SLASH character, U+2044), while some of them (such as the 577 various HEART symbols, e.g., U+2665, U+2661, and U+2765, see 578 Section 7.6) simply fall outside the conventions for typical 579 identifiers (basically letters and numbers). 581 Of course, this category would include code points that had been 582 removed entirely from Unicode should such removals ever occur. 584 Characters that are placed in the "DISALLOWED" category are expected 585 to never be removed from it or reclassified. If a character is 586 classified as "DISALLOWED" in error and the error is sufficiently 587 problematic, the only recourse would be either to introduce a new 588 code point into Unicode and classify it as "PROTOCOL-VALID" or for 589 the IETF to accept the considerable costs of an incompatible change 590 and replace the relevant RFC with one containing appropriate 591 exceptions. 593 There is provision for exception cases but, in general, characters 594 are placed into "DISALLOWED" if they fall into one or more of the 595 following groups: 597 o The character is a compatibility equivalent for another character. 598 In slightly more precise Unicode terms, application of 599 normalization method NFKC to the character yields some other 600 character. 602 o The character is an upper-case form or some other form that is 603 mapped to another character by Unicode casefolding. 605 o The character is a symbol or punctuation form or, more generally, 606 something that is not a letter, digit, or a mark that is used to 607 form a letter or digit. 609 3.1.4. UNASSIGNED 611 For convenience in processing and table-building, code points that do 612 not have assigned values in a given version of Unicode are treated as 613 belonging to a special UNASSIGNED category. Such code points are 614 prohibited in labels to be registered or looked up. The category 615 differs from DISALLOWED in that code points are moved out of it by 616 the simple expedient of being assigned in a later version of Unicode 617 (at which point, they are classified into one of the other categories 618 as appropriate). 620 The rationale for restricting the processing of UNASSIGNED characters 621 is simply that the properties of such code points cannot be 622 completely known until actual characters are assigned to them. For 623 example, assume that an UNASSIGNED code point were included in a 624 label to be looked up. Assume that the code point was later assigned 625 to a character that required some set of contextual rules. With that 626 combination, un-updated instances of IDNA-aware software might permit 627 lookup of labels containing the previously-unassigned characters 628 while updated versions of the software might restrict use of the same 629 label in lookup, depending on the contextual rules. It should be 630 clear that under no circumstance should an UNASSIGNED character be 631 permitted in a label to be registered as part of a domain name. 633 3.2. Registration Policy 635 While these recommendations cannot and should not define registry 636 policies, registries should develop and apply additional restrictions 637 as needed to reduce confusion and other problems. For example, it is 638 generally believed that labels containing characters from more than 639 one script are a bad practice although there may be some important 640 exceptions to that principle. Some registries may choose to restrict 641 registrations to characters drawn from a very small number of 642 scripts. For many scripts, the use of variant techniques such as 643 those as described in RFC 3743 [RFC3743] and RFC 4290 [RFC4290], and 644 illustrated for Chinese by the tables described in RFC 4713 [RFC4713] 645 may be helpful in reducing problems that might be perceived by users. 647 In general, users will benefit if registries only permit characters 648 from scripts that are well-understood by the registry or its 649 advisers. If a registry decides to reduce opportunities for 650 confusion by constructing policies that disallow characters used in 651 historic writing systems or characters whose use is restricted to 652 specialized, highly technical contexts, some relevant information may 653 be found in Section 2.4 "Specific Character Adjustments", Table 4 654 "Candidate Characters for Exclusion from Identifiers" of 655 [Unicode-UAX31] and Section 3.1. "General Security Profile for 656 Identifiers" in [Unicode-Security]. 658 The requirement (in Section 4.1 of [IDNA2008-Protocol]) that 659 registration procedures use only U-labels and/or A-labels is intended 660 to ensure that registrants are fully aware of exactly what is being 661 registered as well as encouraging use of those canonical forms. That 662 provision should not be interpreted as requiring that registrants 663 need to provide characters in a particular code sequence. Registrant 664 input conventions and management are part of registrant-registrar 665 interactions and relationships between registries and registrars and 666 are outside the scope of these standards. 668 It is worth stressing that these principles of policy development and 669 application apply at all levels of the DNS, not only, e.g., TLD or 670 SLD registrations. Even a trivial, "anything is permitted that is 671 valid under the protocol" policy is helpful in that it helps users 672 and application developers know what to expect. 674 3.3. Layered Restrictions: Tables, Context, Registration, Applications 676 The character rules in IDNA2008 are based on the realization that 677 there is no single magic bullet for any of the security, 678 confusability, or other issues associated with IDNs. Instead, the 679 specifications define a variety of approaches. The character tables 680 are the first mechanism, protocol rules about how those characters 681 are applied or restricted in context are the second, and those two in 682 combination constitute the limits of what can be done in the 683 protocol. As discussed in the previous section (Section 3.2), 684 registries are expected to restrict what they permit to be 685 registered, devising and using rules that are designed to optimize 686 the balance between confusion and risk on the one hand and maximum 687 expressiveness in mnemonics on the other. 689 In addition, there is an important role for user agents in warning 690 against label forms that appear problematic given their knowledge of 691 local contexts and conventions. Of course, no approach based on 692 naming or identifiers alone can protect against all threats. 694 4. Issues that Constrain Possible Solutions 696 4.1. Display and Network Order 698 Domain names are always transmitted in network order (the order in 699 which the code points are sent in protocols), but may have a 700 different display order (the order in which the code points are 701 displayed on a screen or paper). When a domain name contains 702 characters that are normally written right to left, display order may 703 be affected although network order is not. It gets even more 704 complicated if left to right and right to left labels are adjacent to 705 each other within a domain name. The decision about the display 706 order is ultimately under the control of user agents --including Web 707 browsers, mail clients, hosted Web applications and many more -- 708 which may be highly localized. Should a domain name abc.def, in 709 which both labels are represented in scripts that are written right 710 to left, be displayed as fed.cba or cba.fed? Applications that are 711 in deployment today are already diverse, and one can find examples of 712 either choice. 714 The picture changes once again when an IDN appears in a 715 Internationalized Resource Identifier (IRI) [RFC3987]. An IRI or 716 Internationalized Email address contains elements other than the 717 domain name. For example, IRIs contain protocol identifiers and 718 field delimiter syntax such as "http://" or "mailto:" while email 719 addresses contain the "@" to separate local parts from domain names. 720 An IRI in network order begins with "http://" followed by domain 721 labels in network order, thus "http://abc.def". 723 User agents are not required to display and allow input of IRIs 724 directly but often do so. Implementors have to choose whether the 725 overall direction of these strings will always be left to right (or 726 right to left) for an IRI or email address. The natural order for a 727 user typing a domain name on a right to left system is fed.cba. 728 Should the R2L user agent reverse the entire domain name each time a 729 domain name is typed? Does this change if the user types "http://" 730 right before typing a domain name, thus implying that the user is 731 beginning at the beginning of the network order IRI? Experience in 732 the 1980s and 1990s with mixing systems in which domain name labels 733 were read in network order (left to right) and those in which those 734 labels were read right to left would predict a great deal of 735 confusion. 737 If each implementation of each application makes its own decisions on 738 these issues, users will develop heuristics that will sometimes fail 739 when switching applications. However, while some display order 740 conventions, voluntarily adopted, would be desirable to reduce 741 confusion, such suggestions are beyond the scope of these 742 specifications. 744 4.2. Entry and Display in Applications 746 Applications can accept and display domain names using any character 747 set or character coding system. The IDNA protocol does not 748 necessarily affect the interface between users and applications. An 749 IDNA-aware application can accept and display internationalized 750 domain names in two formats: the internationalized character set(s) 751 supported by the application (i.e., an appropriate local 752 representation of a U-label), and as an A-label. Applications may 753 allow the display of A-labels, but are encouraged to not do so except 754 as an interface for special purposes, possibly for debugging, or to 755 cope with display limitations. In general, they should allow, but 756 not encourage, user input of A-labels. A-labels are opaque, ugly, 757 and malicious variations on them are not easily detected by users. 758 Where possible, they should thus only be exposed when they are 759 absolutely needed. Because IDN labels can be rendered either as 760 A-labels or U-labels, the application may reasonably have an option 761 for the user to select the preferred method of display. Rendering 762 the U-label should normally be the default. 764 Domain names are often stored and transported in many places. For 765 example, they are part of documents such as mail messages and web 766 pages. They are transported in many parts of many protocols, such as 767 both the control commands of SMTP and associated message body parts, 768 and in the headers and the body content in HTTP. It is important to 769 remember that domain names appear both in domain name slots and in 770 the content that is passed over protocols and it would be helpful if 771 protocols explicitly define what their domain name slots are. 773 In protocols and document formats that define how to handle 774 specification or negotiation of charsets, labels can be encoded in 775 any charset allowed by the protocol or document format. If a 776 protocol or document format only allows one charset, the labels must 777 be given in that charset. Of course, not all charsets can properly 778 represent all labels. If a U-label cannot be displayed in its 779 entirety, the only choice (without loss of information) may be to 780 display the A-label. 782 Where a protocol or document format allows IDNs, labels should be in 783 whatever character encoding and escape mechanism the protocol or 784 document format uses at that place. This provision is intended to 785 prevent situations in which, e.g., UTF-8 domain names appear embedded 786 in text that is otherwise in some other character coding. 788 All protocols that use domain name slots (See Section 2.3.1.6 in 789 [IDNA2008-Defs]) already have the capacity for handling domain names 790 in the ASCII charset. Thus, A-labels can inherently be handled by 791 those protocols. 793 IDNA2008 does not specify required mappings between one character or 794 code point and others. An extended discussion of mapping issues 795 occurs in Section 6 and specific recommendations appear in 796 [IDNA2008-Mapping]. In general, IDNA2008 prohibits characters that 797 would be mapped to others by normalization or other rules. As 798 examples, while mathematical characters based on Latin ones are 799 accepted as input to IDNA2003, they are prohibited in IDNA2008. 800 Similarly, upper-case characters, double-width characters, and other 801 variations are prohibited as IDNA input although mapping them as 802 needed in user interfaces is strongly encouraged. 804 Since the rules in [IDNA2008-Tables] have the effect that only 805 strings that are not transformed by NFKC are valid, if an application 806 chooses to perform NFKC normalization before lookup, that operation 807 is safe since this will never make the application unable to look up 808 any valid string. However, as discussed above, the application 809 cannot guarantee that any other application will perform that 810 mapping, so it should be used only with caution and for informed 811 users. 813 In many cases these prohibitions should have no effect on what the 814 user can type as input to the lookup process. It is perfectly 815 reasonable for systems that support user interfaces to perform some 816 character mapping that is appropriate to the local environment. This 817 would normally be done prior to actual invocation of IDNA. At least 818 conceptually, the mapping would be part of the Unicode conversions 819 discussed above and in [IDNA2008-Protocol]. However, those changes 820 will be local ones only -- local to environments in which users will 821 clearly understand that the character forms are equivalent. For use 822 in interchange among systems, it appears to be much more important 823 that U-labels and A-labels can be mapped back and forth without loss 824 of information. 826 One specific, and very important, instance of this strategy arises 827 with case-folding. In the ASCII-only DNS, names are looked up and 828 matched in a case-independent way, but no actual case-folding occurs. 829 Names can be placed in the DNS in either upper or lower case form (or 830 any mixture of them) and that form is preserved, returned in queries, 831 and so on. IDNA2003 approximated that behavior for non-ASCII strings 832 by performing case-folding at registration time (resulting in only 833 lower-case IDNs in the DNS) and when names were looked up. 835 As suggested earlier in this section, it appears to be desirable to 836 do as little character mapping as possible as long as Unicode works 837 correctly (e.g., NFC mapping to resolve different codings for the 838 same character is still necessary although the specifications require 839 that it be performed prior to invoking the protocol) in order to make 840 the mapping between A-labels and U-labels idempotent. Case-mapping 841 is not an exception to this principle. If only lower case characters 842 can be registered in the DNS (i.e., be present in a U-label), then 843 IDNA2008 should prohibit upper-case characters as input even though 844 user interfaces to applications should probably map those characters. 845 Some other considerations reinforce this conclusion. For example, in 846 ASCII case-mapping for individual characters, uppercase(character) 847 must be equal to uppercase(lowercase(character)). That may not be 848 true with IDNs. In some scripts that use case distinctions, there 849 are a few characters that do not have counterparts in one case or the 850 other. The relationship between upper case and lower case may even 851 be language-dependent, with different languages (or even the same 852 language in different areas) expecting different mappings. User 853 agents can meet the expectations of users who are accustomed to the 854 case-insensitive DNS environment by performing case folding prior to 855 IDNA processing, but the IDNA procedures themselves should neither 856 require such mapping nor expect them when they are not natural to the 857 localized environment. 859 4.3. Linguistic Expectations: Ligatures, Digraphs, and Alternate 860 Character Forms 862 Users have expectations about character matching or equivalence that 863 are based on their own languages and the orthography of those 864 languages. These expectations may not always be met in a global 865 system, especially if multiple languages are written using the same 866 script but using different conventions. Some examples: 868 o A Norwegian user might expect a label with the ae-ligature to be 869 treated as the same label as one using the Swedish spelling with 870 a-diaeresis even though applying that mapping to English would be 871 astonishing to users. 873 o A user in German might expect a label with an o-umlaut and a label 874 that had "oe" substituted, but was otherwise the same, treated as 875 equivalent even though that substitution would be a clear error in 876 Swedish. 878 o A Chinese user might expect automatic matching of Simplified and 879 Traditional Chinese characters, but applying that matching for 880 Korean or Japanese text would create considerable confusion. 882 o An English user might expect "theater" and "theatre" to match. 884 A number of languages use alphabetic scripts in which single phonemes 885 are written using two characters, termed a "digraph", for example, 886 the "ph" in "pharmacy" and "telephone". (Such characters can also 887 appear consecutively without forming a digraph, as in "tophat".) 888 Certain digraphs may be indicated typographically by setting the two 889 characters closer together than they would be if used consecutively 890 to represent different phonemes. Some digraphs are fully joined as 891 ligatures. For example, the word "encyclopaedia" is sometimes set 892 with a U+00E6 LATIN SMALL LIGATURE AE. When ligature and digraph 893 forms have the same interpretation across all languages that use a 894 given script, application of Unicode normalization generally resolves 895 the differences and causes them to match. When they have different 896 interpretations, matching must utilize other methods, presumably 897 chosen at the registry level, or users must be educated to understand 898 that matching will not occur. 900 The nature of the problem can be illustrated by many words in the 901 Norwegian language, where the "ae" ligature is the 27th letter of a 902 29-letter extended Latin alphabet. It is equivalent to the 28th 903 letter of the Swedish alphabet (also containing 29 letters), U+00E4 904 LATIN SMALL LETTER A WITH DIAERESIS, for which an "ae" cannot be 905 substituted according to current orthographic standards. That 906 character (U+00E4) is also part of the German alphabet where, unlike 907 in the Nordic languages, the two-character sequence "ae" is usually 908 treated as a fully acceptable alternate orthography for the "umlauted 909 a" character. The inverse is however not true, and those two 910 characters cannot necessarily be combined into an "umlauted a". This 911 also applies to another German character, the "umlauted o" (U+00F6 912 LATIN SMALL LETTER O WITH DIAERESIS) which, for example, cannot be 913 used for writing the name of the author "Goethe". It is also a 914 letter in the Swedish alphabet where, like the "a with diaeresis", it 915 cannot be correctly represented as "oe" and in the Norwegian 916 alphabet, where it is represented, not as "o with diaeresis", but as 917 "slashed o", U+00F8. 919 Some of the ligatures that have explicit code points in Unicode were 920 given special handling in IDNA2003 and now pose additional problems 921 in transition. See Section 7.2. 923 Additional cases with alphabets written right to left are described 924 in Section 4.5. 926 Matching and comparison algorithm selection often requires 927 information about the language being used, context, or both -- 928 information that is not available to IDNA or the DNS. Consequently, 929 IDNA2008 makes no attempt to treat combined characters in any special 930 way. A registry that is aware of the language context in which 931 labels are to be registered, and where that language sometimes (or 932 always) treats the two- character sequences as equivalent to the 933 combined form, should give serious consideration to applying a 934 "variant" model [RFC3743][RFC4290], or to prohibiting registration of 935 one of the forms entirely, to reduce the opportunities for user 936 confusion and fraud that would result from the related strings being 937 registered to different parties. 939 4.4. Case Mapping and Related Issues 941 In the DNS, ASCII letters are stored with their case preserved. 942 Matching during the query process is case-independent, but none of 943 the information that might be represented by choices of case has been 944 lost. That model has been accidentally helpful because, as people 945 have created DNS labels by catenating words (or parts of words) to 946 form labels, case has often been used to distinguish among components 947 and make the labels more memorable. 949 Since DNS servers do not get involved in parsing IDNs, they cannot do 950 case-independent matching. Thus, keeping the cases separate in 951 lookup or registration, and doing matching at the server, is not 952 feasible with IDNA or any similar approach. Case-matching must be 953 done, if desired, by IDN clients even though it wasn't done by ASCII- 954 only DNS clients. That situation was recognized in IDNA2003 and 955 nothing in IDNA2008 fundamentally changes it or could do so. In 956 IDNA2003, all characters are case-folded and mapped by clients in a 957 standardized step. 959 Some characters do not have upper case forms. For example the 960 Unicode case folding operation maps Greek Final Form Sigma (U+03C2) 961 to the medial form (U+03C3) and maps Eszett (German Sharp S, U+00DF) 962 to "ss". Neither of these mappings is reversible because the upper 963 case of U+03C3 is the Upper Case Sigma (U+03A3) and "ss" is an ASCII 964 string. IDNA2008 permits, at the risk of some incompatibility, 965 slightly more flexibility in this area by avoiding case folding and 966 treating these characters as themselves. Approaches to handling one- 967 way mappings are discussed in Section 7.2. 969 Because IDNA2003 maps Final Sigma and Eszett to other characters, and 970 the reverse mapping is never possible, neither Final Sigma nor Eszett 971 can be represented in the ACE form of IDNA2003 IDN nor in the native 972 character (U-label) form derived from it. With IDNA2008, both 973 characters can be used in an IDN and so the A-label used for lookup 974 for any U-label containing those characters, is now different. See 975 Section 7.1 for a discussion of what kinds of changes might require 976 the IDNA prefix to change; after extended discussions, the WG came to 977 consensus that the change for these characters did not justify a 978 prefix change. 980 4.5. Right to Left Text 982 In order to be sure that the directionality of right to left text is 983 unambiguous, IDNA2003 required that any label in which right to left 984 characters appear both starts and ends with them and that it not 985 include any characters with strong left to right properties (that 986 excludes other alphabetic characters but permits European digits). 987 Any other string that contains a right to left character and does not 988 meet those requirements is rejected. This is one of the few places 989 where the IDNA algorithms (both in IDNA2003 and in IDAN2008) examine 990 an entire label, not just individual characters. The algorithmic 991 model used in IDNA2003 rejects the label when the final character in 992 a right to left string requires a combining mark in order to be 993 correctly represented. 995 That prohibition is not acceptable for writing systems for languages 996 written with consonantal alphabets to which diacritical vocalic 997 systems are applied, and for languages with orthographies derived 998 from them where the combining marks may have different functionality. 999 In both cases the combining marks can be essential components of the 1000 orthography. Examples of this are Yiddish, written with an extended 1001 Hebrew script, and Dhivehi (the official language of Maldives) which 1002 is written in the Thaana script (which is, in turn, derived from the 1003 Arabic script). IDNA2008 removes the restriction on final combining 1004 characters with a new set of rules for right to left scripts and 1005 their characters. Those new rules are specified in [IDNA2008-Bidi]. 1007 5. IDNs and the Robustness Principle 1009 The "Robustness Principle" is often stated as "Be conservative about 1010 what you send and liberal in what you accept" (See, e.g., Section 1011 1.2.2 of the applications-layer Host Requirements specification 1012 [RFC1123]) This principle applies to IDNA. In applying the principle 1013 to registries as the source ("sender") of all registered and useful 1014 IDNs, registries are responsible for being conservative about what 1015 they register and put out in the Internet. For IDNs to work well, 1016 zone administrators (registries) must have and require sensible 1017 policies about what is registered -- conservative policies -- and 1018 implement and enforce them. 1020 Conversely, lookup applications are expected to reject labels that 1021 clearly violate global (protocol) rules (no one has ever seriously 1022 claimed that being liberal in what is accepted requires being 1023 stupid). However, once one gets past such global rules and deals 1024 with anything sensitive to script or locale, it is necessary to 1025 assume that garbage has not been placed into the DNS, i.e., one must 1026 be liberal about what one is willing to look up in the DNS rather 1027 than guessing about whether it should have been permitted to be 1028 registered. 1030 If a string cannot be successfully found in the DNS after the lookup 1031 processing described here, it makes no difference whether it simply 1032 wasn't registered or was prohibited by some rule at the registry. 1033 Application implementors should be aware that where DNS wildcards are 1034 used, the ability to successfully resolve a name does not guarantee 1035 that it was actually registered. 1037 6. Front-end and User Interface Processing for Lookup 1039 Domain names may be identified and processed in many contexts. They 1040 may be typed in by users either by themselves or embedded in an 1041 identifier such as email addresses, URIs, or IRIs. They may occur in 1042 running text or be processed by one system after being provided in 1043 another. Systems may try to normalize URLs to determine (or guess) 1044 whether a reference is valid or two references point to the same 1045 object without actually looking the objects up (comparison without 1046 lookup is necessary for URI types that are not intended to be 1047 resolved). Some of these goals may be more easily and reliably 1048 satisfied than others. While there are strong arguments for any 1049 domain name that is placed "on the wire" -- transmitted between 1050 systems -- to be in the zero-ambiguity forms of A-labels, it is 1051 inevitable that programs that process domain names will encounter 1052 U-labels or variant forms. 1054 An application that implements the IDNA protocol [IDNA2008-Protocol] 1055 will always take any user input and convert it to a set of Unicode 1056 code points. That user input may be acquired by any of several 1057 different input methods, all with differing conversion processes to 1058 be taken into consideration (e.g., typed on a keyboard, written by 1059 hand onto some sort of digitizer, spoken into a microphone and 1060 interpreted by a speech-to-text engine, etc.). The process of taking 1061 any particular user input and mapping it into a Unicode code point 1062 may be a simple one: If a user strikes the "A" key on a US English 1063 keyboard, without any modifiers such as the "Shift" key held down, in 1064 order to draw a Latin small letter A ("a"), many (perhaps most) 1065 modern operating system input methods will produce to the calling 1066 application the code point U+0061, encoded in a single octet. 1068 Sometimes the process is somewhat more complicated: a user might 1069 strike a particular set of keys to represent a combining macron 1070 followed by striking the "A" key in order to draw a Latin small 1071 letter A with a macron above it. Depending on the operating system, 1072 the input method chosen by the user, and even the parameters with 1073 which the application communicates with the input method, the result 1074 might be the code point U+0101 (encoded as two octets in UTF-8 or 1075 UTF-16, four octets in UTF-32, etc.), the code point U+0061 followed 1076 by the code point U+0304 (again, encoded in three or more octets, 1077 depending upon the encoding used) or even the code point U+FF41 1078 followed by the code point U+0304 (and encoded in some form). And 1079 these examples leave aside the issue of operating systems and input 1080 methods that do not use Unicode code points for their character set. 1082 In every case, applications (with the help of the operating systems 1083 on which they run and the input methods used) need to perform a 1084 mapping from user input into Unicode code points. 1086 The original version of the IDNA protocol [RFC3490] used a model 1087 whereby input was taken from the user, mapped (via whatever input 1088 method mechanisms were used) to a set of Unicode code points, and 1089 then further mapped to a set of Unicode code points using the 1090 Nameprep profile specified in [RFC3491]. In this procedure, there 1091 are two separate mapping steps: First, a mapping done by the input 1092 method (which might be controlled by the operating system, the 1093 application, or some combination) and then a second mapping performed 1094 by the Nameprep portion of the IDNA protocol. The mapping done in 1095 Nameprep includes a particular mapping table to re-map some 1096 characters to other characters, a particular normalization, and a set 1097 of prohibited characters. 1099 Note that the result of the two step mapping process means that the 1100 mapping chosen by the operating system or application in the first 1101 step might differ significantly from the mapping supplied by the 1102 Nameprep profile in the second step. This has advantages and 1103 disadvantages. Of course, the second mapping regularizes what gets 1104 looked up in the DNS, making for better interoperability between 1105 implementations which use the Nameprep mapping. However, the 1106 application or operating system may choose mappings in their input 1107 methods, which when passed through the second (Nameprep) mapping 1108 result in characters that are "surprising" to the end user. 1110 The other important feature of the original version of the IDNA 1111 protocol is that, with very few exceptions, it assumes that any set 1112 of Unicode code points provided to the Nameprep mapping can be mapped 1113 into a string of Unicode code points that are "sensible", even if 1114 that means mapping some code points to nothing (that is, removing the 1115 code points from the string). This allowed maximum flexibility in 1116 input strings. 1118 The present version of IDNA differs significantly in approach from 1119 the original version. First and foremost, it does not provide 1120 explicit mapping instructions. Instead, it assumes that the 1121 application (perhaps via an operating system input method) will do 1122 whatever mapping it requires to convert input into Unicode code 1123 points. This has the advantage of giving flexibility to the 1124 application to choose a mapping that is suitable for its user given 1125 specific user requirements, and avoids the two-step mapping of the 1126 original protocol. Instead of a mapping, the current version of IDNA 1127 provides a set of categories that can be used to specify the valid 1128 code points allowed in a domain name. 1130 In principle, an application ought to take user input of a domain 1131 name and convert it to the set of Unicode code points that represent 1132 the domain name the user intends. As a practical matter, of course, 1133 determining user intent is a tricky business, so an application needs 1134 to choose a reasonable mapping from user input. That may differ 1135 based on the particular circumstances of a user, depending on locale, 1136 language, type of input method, etc. It is up to the application to 1137 make a reasonable choice. 1139 7. Migration from IDNA2003 and Unicode Version Synchronization 1141 7.1. Design Criteria 1143 As mentioned above and in RFC 4690, two key goals of the IDNA2008 1144 design are 1146 o to enable applications to be agnostic about whether they are being 1147 run in environments supporting any Unicode version from 3.2 1148 onward, 1150 o to permit incrementally adding new characters, character groups, 1151 scripts, and other character collections as they are incorporated 1152 into Unicode, doing so without disruption and, in the long term, 1153 without "heavy" processes (an IETF consensus process is required 1154 by the IDNA2008 specifications and is expected to be required and 1155 used until significant experience accumulates with IDNA operations 1156 and new versions of Unicode). 1158 7.1.1. Summary and Discussion of IDNA Validity Criteria 1160 The general criteria for a label to be considered valid under IDNA 1161 are (the actual rules are rigorously defined in [IDNA2008-Protocol] 1162 and [IDNA2008-Tables]): 1164 o The characters are "letters", marks needed to form letters, 1165 numerals, or other code points used to write words in some 1166 language. Symbols, drawing characters, and various notational 1167 characters are intended to be permanently excluded. There is no 1168 evidence that they are important enough to Internet operations or 1169 internationalization to justify expansion of domain names beyond 1170 the general principle of "letters, digits, and hyphen". 1171 (Additional discussion and rationale for the symbol decision 1172 appears in Section 7.6). 1174 o Other than in very exceptional cases, e.g., where they are needed 1175 to write substantially any word of a given language, punctuation 1176 characters are excluded. The fact that a word exists is not proof 1177 that it should be usable in a DNS label and DNS labels are not 1178 expected to be usable for multiple-word phrases (although they are 1179 certainly not prohibited if the conventions and orthography of a 1180 particular language cause that to be possible). 1182 o Characters that are unassigned (have no character assignment at 1183 all) in the version of Unicode being used by the registry or 1184 application are not permitted, even on lookup. The issues 1185 involved in this decision are discussed in Section 7.7. 1187 o Any character that is mapped to another character by a current 1188 version of NFKC is prohibited as input to IDNA (for either 1189 registration or lookup). With a few exceptions, this principle 1190 excludes any character mapped to another by Nameprep [RFC3491]. 1192 The principles above drive the design of rules that are specified 1193 exactly in [IDNA2008-Tables]. Those rules identify the characters 1194 that are valid under IDNA. The rules themselves are normative, and 1195 the tables are derived from them, rather than vice versa. 1197 7.1.2. Labels in Registration 1199 Any label registered in a DNS zone must be validated -- i.e., the 1200 criteria for that label must be met -- in order for applications to 1201 work as intended. This principle is not new. For example, since the 1202 DNS was first deployed, zone administrators have been expected to 1203 verify that names meet "hostname" requirements [RFC0952] where those 1204 requirements are imposed by the expected applications. Other 1205 applications contexts, such as the later addition of special service 1206 location formats [RFC2782] imposed new requirements on zone 1207 administrators. For zones that will contain IDNs, support for 1208 Unicode version-independence requires restrictions on all strings 1209 placed in the zone. In particular, for such zones (the exact rules 1210 appear in the Protocol Document, Section 4 [IDNA2008-Protocol]): 1212 o Any label that appears to be an A-label, i.e., any label that 1213 starts in "xn--", must be valid under IDNA, i.e., they must be 1214 valid A-labels, as discussed in Section 2 above. 1216 o The Unicode tables (i.e., tables of code points, character 1217 classes, and properties) and IDNA tables (i.e., tables of 1218 contextual rules such as those that appear in the Tables 1219 document), must be consistent on the systems performing or 1220 validating labels to be registered. Note that this does not 1221 require that tables reflect the latest version of Unicode, only 1222 that all tables used on a given system are consistent with each 1223 other. 1225 Under this model, registry tables will need to be updated (both the 1226 Unicode-associated tables and the tables of permitted IDN characters) 1227 to enable a new script or other set of new characters. The registry 1228 will not be affected by newer versions of Unicode, or newly- 1229 authorized characters, until and unless it wishes to support them. 1230 The zone administrator is responsible for verifying validity for IDNA 1231 as well as its local policies -- a more extensive set of checks than 1232 are required for looking up the labels. Systems looking up or 1233 resolving DNS labels, especially IDN DNS labels, must be able to 1234 assume that applicable registration rules were followed for names 1235 entered into the DNS. 1237 7.1.3. Labels in Lookup 1239 Any application processing a label through IDNA so it can be looked 1240 up in a DNS zone is required to (the exact rules appear in the 1241 Protocol Document, Section 5 [IDNA2008-Protocol]) 1243 o Maintain IDNA and Unicode tables that are consistent with regard 1244 to versions, i.e., unless the application actually executes the 1245 classification rules in [IDNA2008-Tables], its IDNA tables must be 1246 derived from the version of Unicode that is supported more 1247 generally on the system. As with registration, the tables need 1248 not reflect the latest version of Unicode but they must be 1249 consistent. 1251 o Validate the characters in labels to be looked up only to the 1252 extent of determining that the U-label does not contain 1253 "DISALLOWED" code points or code points that are unassigned in its 1254 version of Unicode. 1256 o Validate the label itself for conformance with a small number of 1257 whole-label rules. In particular, it must verify that 1258 * there are no leading combining marks, 1260 * the "bidi" conditions are met if right to left characters 1261 appear, 1263 * any required contextual rules are available, and 1265 * any contextual rules that are associated with Joiner Controls 1266 (and "CONTEXTJ" characters more generally) are tested. 1268 o Do not reject labels based on other contextual rules about 1269 characters, including mixed-script label prohibitions. Such rules 1270 may be used to influence presentation decisions in the user 1271 interface, but not to avoid looking up domain names. 1273 To further clarify the rules about handling characters that require 1274 contextual rules, note that one can have a context-required character 1275 (i.e., one that requires a rule), but no rule. In that case, the 1276 character is treated the same way DISALLOWED characters are treated, 1277 until and unless a rule is supplied. That state is more or less 1278 equivalent to "the idea of permitting this character is accepted in 1279 principle, but it won't be permitted in practice until consensus is 1280 reached on a safe way to use it". 1282 The ability to add a rule more or less exempts these characters from 1283 the prohibition against reclassifying characters from DISALLOWED to 1284 PVALID. 1286 And, obviously, "no rule" is different from "have a rule, but the 1287 test either succeeds or fails". 1289 Lookup applications that follow these rules, rather than having their 1290 own criteria for rejecting lookup attempts, are not sensitive to 1291 version incompatibilities with the particular zone registry 1292 associated with the domain name except for labels containing 1293 characters recently added to Unicode. 1295 An application or client that processes names according to this 1296 protocol and then resolves them in the DNS will be able to locate any 1297 name that is registered, as long as those registrations are valid 1298 under IDNA and its version of the IDNA tables is sufficiently up-to- 1299 date to interpret all of the characters in the label. Messages to 1300 users should distinguish between "label contains an unallocated code 1301 point" and other types of lookup failures. A failure on the basis of 1302 an old version of Unicode may lead the user to a desire to upgrade to 1303 a newer version, but will have no other ill effects (this is 1304 consistent with behavior in the transition to the DNS when some hosts 1305 could not yet handle some forms of names or record types). 1307 7.2. Changes in Character Interpretations 1309 As a consequence of the elimination of mapping, the current version 1310 of IDNA changes the interpretation of a few characters relative to 1311 its predecessors. This subsection outlines the issues and discusses 1312 possible transition strategies. 1314 7.2.1. Character Changes: Eszett and Final Sigma 1316 In those scripts that make case distinctions, there are a few 1317 characters for which an obvious and unique upper case character has 1318 not historically been available to match a lower case one or vice 1319 versa. For those characters, the mappings used in constructing the 1320 Stringprep tables for IDNA2003, performed using the Unicode CaseFold 1321 operation (See Section 5.8 of the Unicode Standard [Unicode51]), 1322 generate different characters or sets of characters. Those 1323 operations are not reversible and lose even more information than 1324 traditional upper case or lower case transformations, but are more 1325 useful than those transformations for comparison purposes. Two 1326 notable characters of this type are the German character Eszett 1327 (Sharp S, U+00DF) and the Greek Final Form Sigma (U+03C2). The 1328 former is case-folded to the ASCII string "ss", the latter to a 1329 medial (Lower Case) Sigma (U+03C3). 1331 7.2.2. Character Changes: Zero-Width Joiner and Non-Joiner 1333 IDNA2003 mapped both Zero-Width Joiner (ZWJ, U+200D) and Zero-Width 1334 Non-Joiner (ZWNJ, U+200C) to nothing, effectively dropping these 1335 characters from any label in which they appeared and treating strings 1336 containing them as identical to strings that did not. As discussed 1337 in Section 3.1.2 above, those characters are essential for writing 1338 many reasonable mnemonics for certain scripts. However, treating 1339 them as valid in the current version of IDNA, even with contextual 1340 restrictions, raises approximately the same problem as exists with 1341 Eszett and Final Sigma: strings that were valid under IDNA2003 have 1342 different interpretations as labels, and different A-labels, than the 1343 same strings under this newer version. 1345 7.2.3. Character Changes and the Need for Transition 1347 The decision to eliminate mandatory and standardized mappings, 1348 including case folding, from the IDNA2008 protocol in order to make 1349 A-labels and U-labels idempotent made these characters problematic. 1350 If they were to be disallowed, important words and mnemonics could 1351 not be written in orthographically reasonable ways. If they were to 1352 be permitted as distinct characters, there would be no information 1353 loss and registries would have more flexibility, but IDNA2003 and 1354 IDNA2008 lookups might result in different A-labels. 1356 With the understanding that there would be incompatibility either way 1357 but a judgment that the incompatibility was not significant enough to 1358 justify a prefix change, the WG concluded that Eszett and Final Form 1359 Sigma should be treated as distinct and Protocol-Valid characters. 1361 Since these characters are interpreted in different ways under the 1362 older and newer versions of IDNA, transition strategies and policies 1363 will be necessary. Some actions can reasonably be taken by 1364 applications client programs (those that perform lookup operations or 1365 cause them to be performed) but, because of the diversity of 1366 situations and uses of the DNS, much of the responsibility will need 1367 to fall on registries. 1369 Registries, especially those maintaining zones for third parties, 1370 must decide how to introduce a new service in a way that does not 1371 create confusion or significantly weaken or invalidate existing 1372 identifiers. This is not a new problem; registries were faced with 1373 similar issues when IDNs were introduced (potentially, and especially 1374 for Latin-based scripts, in conflict with existing labels that had 1375 been rendered in ASCII character by applying more or less 1376 standardized conventions) and when other new forms of strings have 1377 been permitted as labels. 1379 7.2.4. Transition Strategies 1381 There are several approaches to the introduction of new characters or 1382 changes in interpretation of existing characters from their mapped 1383 forms in the earlier version of IDNA. The transition issue is 1384 complicated because the forms of these labels after 1385 ToUnicode(ToASCII()) translation in IDNA2003 not only remain valid 1386 but do not provide strong indications of what the registrant 1387 intended: a string containing "ss" could have simply been intended to 1388 be that string or could have been intended to contain an Eszett, a 1389 string containing lower-case Sigma could have been intended to 1390 contain Final Sigma (one might make heuristic guesses based on 1391 position in a string, but the long tradition of forming labels by 1392 concatenating words makes such heuristics unreliable), and strings 1393 that do not contain ZWJ or ZWNJ might have been intended to contain 1394 them. Without any preference or claim to completeness, some of 1395 these, all of which have been used by registries in the past for 1396 similar transitions, are: 1398 1. Do not permit use of the newly-available character at the 1399 registry level. This might cause lookup failures if a domain 1400 name were to be written with the expectation of the IDNA2003 1401 mapping behavior, but would eliminate any possibility of false 1402 matches. 1404 2. Hold a "sunrise"-like arrangement in which holders of labels 1405 containing "ss" in the Eszett case, Lower Case Sigma in that 1406 case, or that might have contained ZWJ or ZWNJ in context, are 1407 given priority (and perhaps other benefits) for registering the 1408 corresponding string containing Eszett, Final Sigma, or the 1409 appropriate Zero-width character respectively. 1411 3. Adopt some sort of "variant" approach in which registrants obtain 1412 labels with both character forms. 1414 4. Adopt a different form of "variant" approach in which 1415 registration of additional strings that would produce the same 1416 A-label if interpreted according to IDNA2003 is either not 1417 permitted at all or permitted only by the registrant who already 1418 has one of the names. 1420 5. Ignore the issue and assume that the marketplace or other 1421 mechanisms will sort things out. 1423 In any event, a registry (at any level of the DNS tree) that chooses 1424 to permit labels to be registered that contains these characters, or 1425 considers doing so, will have to address the relationship with 1426 existing, possibly-conflicting, labels in some way, just as 1427 registries that already had a considerable number of labels did when 1428 IDNs were first introduced. 1430 7.3. Elimination of Character Mapping 1432 As discussed at length in Section 6, IDNA2003, via Nameprep (see 1433 Section 7.5), mapped many characters into related ones. Those 1434 mappings no longer exist as requirements in IDNA2008. These 1435 specifications strongly prefer that only A-labels or U-labels be used 1436 in protocol contexts and as much as practical more generally. 1437 IDNA2008 does anticipate situations in which some mapping at the time 1438 of user input into lookup applications is appropriate and desirable. 1439 The issues are discussed in Section 6 and specific recommendations 1440 are made in [IDNA2008-Mapping]. 1442 7.4. The Question of Prefix Changes 1444 The conditions that would have required a change in the IDNA ACE 1445 prefix ("xn--" for the version of IDNA specified in [RFC3490]) were 1446 of great concern to the community. A prefix change would have 1447 clearly been necessary if the algorithms were modified in a manner 1448 that would have created serious ambiguities during subsequent 1449 transition in registrations. This section summarizes the working 1450 group's conclusions about the conditions under which a change in the 1451 prefix would have been necessary and the implications of such a 1452 change. 1454 7.4.1. Conditions Requiring a Prefix Change 1456 An IDN prefix change would have been needed if a given string would 1457 be looked up or otherwise interpreted differently depending on the 1458 version of the protocol or tables being used. This IDNA upgrade 1459 would have required a prefix change if, and only if, one of the 1460 following four conditions were met: 1462 1. The conversion of an A-label to Unicode (i.e., a U-label) would 1463 have yielded one string under IDNA2003 (RFC3490) and a different 1464 string under IDNA2008. 1466 2. In a significant number of cases, an input string that was valid 1467 under IDNA2003 and also valid under IDNA2008 would have yielded 1468 two different A-labels with the different versions. This 1469 condition is believed to be essentially equivalent to the one 1470 above except for a very small number of edge cases that were not 1471 found to justify a prefix change (See Section 7.2). 1473 Note that if the input string was valid under one version and not 1474 valid under the other, this condition would not apply. See the 1475 first item in Section 7.4.2, below. 1477 3. A fundamental change was made to the semantics of the string that 1478 would be inserted in the DNS, e.g., if a decision were made to 1479 try to include language or script information in the encoding in 1480 addition to the string itself. 1482 4. A sufficiently large number of characters were added to Unicode 1483 so that the Punycode mechanism for block offsets would no longer 1484 reference the higher-numbered planes and blocks. This condition 1485 is unlikely even in the long term and certain not to arise in the 1486 next several years. 1488 7.4.2. Conditions Not Requiring a Prefix Change 1490 As a result of the principles described above, none of the following 1491 changes required a new prefix: 1493 1. Prohibition of some characters as input to IDNA. Such a 1494 prohibition might make names that were previously registered 1495 inaccessible, but did not change those names. 1497 2. Adjustments in IDNA tables or actions, including normalization 1498 definitions, that affected characters that were already invalid 1499 under IDNA2003. 1501 3. Changes in the style of the IDNA definition that did not alter 1502 the actions performed by IDNA. 1504 7.4.3. Implications of Prefix Changes 1506 While it might have been possible to make a prefix change, the costs 1507 of such a change are considerable. Registries could not have 1508 converted all IDNA2003 ("xn--") registrations to a new form at the 1509 same time and synchronize that change with applications supporting 1510 lookup. Unless all existing registrations were simply to be declared 1511 invalid (and perhaps even then) systems that needed to support both 1512 labels with old prefixes and labels with new ones would be required 1513 to first process a putative label under the IDNA2008 rules and try to 1514 look it up and then, if it were not found, would be required to 1515 process the label under IDNA2003 rules and look it up again. That 1516 process would probably have significantly slowed down all processing 1517 that involved IDNs in the DNS especially since a fully-qualified name 1518 might contain a mixture of labels that were registered with the old 1519 and new prefixes. That would have made DNS caching very difficult. 1520 In addition, looking up the same input string as two separate 1521 A-labels would have created some potential for confusion and attacks, 1522 since the labels could map to different targets and then resolve to 1523 different entries in the DNS. 1525 Consequently, a prefix change should have been, and was, avoided if 1526 at all possible, even if it means accepting some IDNA2003 decisions 1527 about character distinctions as irreversible and/or giving special 1528 treatment to edge cases. 1530 7.5. Stringprep Changes and Compatibility 1532 The Nameprep [RFC3491] specification, a key part of IDNA2003, is a 1533 profile of Stringprep [RFC3454]. While Nameprep is a Stringprep 1534 profile specific to IDNA, Stringprep is used by a number of other 1535 protocols. Were Stringprep to have been modified by IDNA2008, those 1536 changes to improve the handling of IDNs could cause problems for non- 1537 DNS uses, most notably if they affected identification and 1538 authentication protocols. Several elements of IDNA2008 give 1539 interpretations to strings prohibited under IDNA2003 or prohibit 1540 strings that IDNA2003 permitted. Those elements include the proposed 1541 new inclusion tables [IDNA2008-Tables], the reduction in the number 1542 of characters permitted as input for registration or lookup 1543 (Section 3), and even the proposed changes in handling of right to 1544 left strings [IDNA2008-Bidi]. IDNA2008 does not use Nameprep or 1545 Stringprep at all, so there are no side-effect changes to other 1546 protocols. 1548 It is particularly important to keep IDNA processing separate from 1549 processing for various security protocols because some of the 1550 constraints that are necessary for smooth and comprehensible use of 1551 IDNs may be unwanted or undesirable in other contexts. For example, 1552 the criteria for good passwords or passphrases are very different 1553 from those for desirable IDNs: passwords should be hard to guess, 1554 while domain names should normally be easily memorable. Similarly, 1555 internationalized SCSI identifiers and other protocol components are 1556 likely to have different requirements than IDNs. 1558 7.6. The Symbol Question 1560 One of the major differences between this specification and the 1561 original version of IDNA is that the original version permitted non- 1562 letter symbols of various sorts, including punctuation and line- 1563 drawing symbols, in the protocol. They were always discouraged in 1564 practice. In particular, both the "IESG Statement" about IDNA and 1565 all versions of the ICANN Guidelines specify that only language 1566 characters be used in labels. This specification disallows symbols 1567 entirely. There are several reasons for this, which include: 1569 1. As discussed elsewhere, the original IDNA specification assumed 1570 that as many Unicode characters as possible should be permitted, 1571 directly or via mapping to other characters, in IDNs. This 1572 specification operates on an inclusion model, extrapolating from 1573 the original "hostname" rules (LDH, see [IDNA2008-Defs]) -- which 1574 have served the Internet very well -- to a Unicode base rather 1575 than an ASCII base. 1577 2. Symbol names are more problematic than letters because there may 1578 be no general agreement on whether a particular glyph matches a 1579 symbol; there are no uniform conventions for naming; variations 1580 such as outline, solid, and shaded forms may or may not exist; 1581 and so on. As just one example, consider a "heart" symbol as it 1582 might appear in a logo that might be read as "I love...". While 1583 the user might read such a logo as "I love..." or "I heart...", 1584 considerable knowledge of the coding distinctions made in Unicode 1585 is needed to know that there is more than one "heart" character 1586 (e.g., U+2665, U+2661, and U+2765) and how to describe it. These 1587 issues are of particular importance if strings are expected to be 1588 understood or transcribed by the listener after being read out 1589 loud. 1591 3. Design of a screen reader used by blind Internet users who must 1592 listen to renderings of IDN domain names and possibly reproduce 1593 them on the keyboard becomes considerably more complicated when 1594 the names of characters are not obvious and intuitive to anyone 1595 familiar with the language in question. 1597 4. As a simplified example of this, assume one wanted to use a 1598 "heart" or "star" symbol in a label. This is problematic because 1599 those names are ambiguous in the Unicode system of naming (the 1600 actual Unicode names require far more qualification). A user or 1601 would-be registrant has no way to know -- absent careful study of 1602 the code tables -- whether it is ambiguous (e.g., where there are 1603 multiple "heart" characters) or not. Conversely, the user seeing 1604 the hypothetical label doesn't know whether to read it -- try to 1605 transmit it to a colleague by voice -- as "heart", as "love", as 1606 "black heart", or as any of the other examples below. 1608 5. The actual situation is even worse than this. There is no 1609 possible way for a normal, casual, user to tell the difference 1610 between the hearts of U+2665 and U+2765 and the stars of U+2606 1611 and U+2729 without somehow knowing to look for a distinction. We 1612 have a white heart (U+2661) and few black hearts. Consequently, 1613 describing a label as containing a heart is hopelessly ambiguous: 1614 we can only know that it contains one of several characters that 1615 look like hearts or have "heart" in their names. In cities where 1616 "Square" is a popular part of a location name, one might well 1617 want to use a square symbol in a label as well and there are far 1618 more squares of various flavors in Unicode than there are hearts 1619 or stars. 1621 The consequence of these ambiguities is that symbols are a very poor 1622 basis for reliable communication. Consistent with this conclusion, 1623 the Unicode standard recommends that strings used in identifiers not 1624 contain symbols or punctuation [Unicode-UAX31]. Of course, these 1625 difficulties with symbols do not arise with actual pictographic 1626 languages and scripts which would be treated like any other language 1627 characters; the two should not be confused. 1629 7.7. Migration Between Unicode Versions: Unassigned Code Points 1631 In IDNA2003, labels containing unassigned code points are looked up 1632 on the assumption that, if they appear in labels and can be mapped 1633 and then resolved, the relevant standards must have changed and the 1634 registry has properly allocated only assigned values. 1636 In the IDNA2008 protocol, strings containing unassigned code points 1637 must not be either looked up or registered. In summary, the status 1638 of an unassigned character with regard to the DISALLOWED, PROTOCOL- 1639 VALID, and CONTEXTUAL RULE REQUIRED categories cannot be evaluated 1640 until a character is actually assigned and known. There are several 1641 reasons for this, with the most important ones being: 1643 o Tests involving the context of characters (e.g., some characters 1644 being permitted only adjacent to others of specific types) and 1645 integrity tests on complete labels are needed. Unassigned code 1646 points cannot be permitted because one cannot determine whether 1647 particular code points will require contextual rules (and what 1648 those rules should be) before characters are assigned to them and 1649 the properties of those characters fully understood. 1651 o It cannot be known in advance, and with sufficient reliability, 1652 whether a newly-assigned code point will be associated with a 1653 character that would be disallowed by the rules in 1654 [IDNA2008-Tables] (such as a compatibility character). In 1655 IDNA2003, since there is no direct dependency on NFKC (many of the 1656 entries in Stringprep's tables are based on NFKC, but IDNA2003 1657 depends only on Stringprep), allocation of a compatibility 1658 character might produce some odd situations, but it would not be a 1659 problem. In IDNA2008, where compatibility characters are 1660 DISALLOWED unless character-specific exceptions are made, 1661 permitting strings containing unassigned characters to be looked 1662 up would violate the principle that characters in DISALLOWED are 1663 not looked up. 1665 o The Unicode Standard specifies that an unassigned code point 1666 normalizes (and, where relevant, case folds) to itself. If the 1667 code point is later assigned to a character, and particularly if 1668 the newly-assigned code point has a combining class that 1669 determines its placement relative to other combining characters, 1670 it could normalize to some other code point or sequence. 1672 It is possible to argue that the issues above are not important and 1673 that, as a consequence, it is better to retain the principle of 1674 looking up labels even if they contain unassigned characters because 1675 all of the important scripts and characters have been coded as of 1676 Unicode 5.1 and hence unassigned code points will be assigned only to 1677 obscure characters or archaic scripts. Unfortunately, that does not 1678 appear to be a safe assumption for at least two reasons. First, much 1679 the same claim of completeness has been made for earlier versions of 1680 Unicode. The reality is that a script that is obscure to much of the 1681 world may still be very important to those who use it. Cultural and 1682 linguistic preservation principles make it inappropriate to declare 1683 the script of no importance in IDNs. Second, we already have 1684 counterexamples in, e.g., the relationships associated with new Han 1685 characters being added (whether in the BMP or in Unicode Plane 2). 1687 Independent of the technical transition issues identified above, it 1688 can be observed that any addition of characters to an existing script 1689 to make it easier to use or to better accommodate particular 1690 languages may lead to transition issues. Such additions may change 1691 the preferred form for writing a particular string, changes that may 1692 be reflected, e.g., in keyboard transition modules that would 1693 necessarily be different from those for earlier versions of Unicode 1694 where the newer characters may not exist. This creates an inherent 1695 transition problem because attempts to access labels may use either 1696 the old or the new conventions, requiring registry action whether the 1697 older conventions were used in labels or not. The need to consider 1698 transition mechanisms is inherent to evolution of Unicode to better 1699 accommodate writing systems and is independent of how IDNs are 1700 represented in the DNS or how transitions among versions of those 1701 mechanisms occur. The requirement for transitions of this type is 1702 illustrated by the addition of Malayalam Chillu in Unicode 5.1.0. 1704 7.8. Other Compatibility Issues 1706 The 2003 IDNA model includes several odd artifacts of the context in 1707 which it was developed. Many, if not all, of these are potential 1708 avenues for exploits, especially if the registration process permits 1709 "source" names (names that have not been processed through IDNA and 1710 Nameprep) to be registered. As one example, since the character 1711 Eszett, used in German, is mapped by IDNA2003 into the sequence "ss" 1712 rather than being retained as itself or prohibited, a string 1713 containing that character but that is otherwise in ASCII is not 1714 really an IDN (in the U-label sense defined above) at all. After 1715 Nameprep maps the Eszett out, the result is an ASCII string and so 1716 does not get an xn-- prefix, but the string that can be displayed to 1717 a user appears to be an IDN. The newer version of the protocol 1718 eliminates this artifact. A character is either permitted as itself 1719 or it is prohibited; special cases that make sense only in a 1720 particular linguistic or cultural context can be dealt with as 1721 localization matters where appropriate. 1723 8. Name Server Considerations 1725 8.1. Processing Non-ASCII Strings 1727 Existing DNS servers do not know the IDNA rules for handling non- 1728 ASCII forms of IDNs, and therefore need to be shielded from them. 1729 All existing channels through which names can enter a DNS server 1730 database (for example, master files (as described in RFC 1034) and 1731 DNS update messages [RFC2136]) are IDN-unaware because they predate 1732 IDNA. Other sections of this document provide the needed shielding 1733 by ensuring that internationalized domain names entering DNS server 1734 databases through such channels have already been converted to their 1735 equivalent ASCII A-label forms. 1737 Because of the distinction made between the algorithms for 1738 Registration and Lookup in [IDNA2008-Protocol] (a domain name 1739 containing only ASCII codepoints cannot be converted to an A-label), 1740 there cannot be more than one A-label form for any given U-label. 1742 As specified in RFC 2181 [RFC2181], the DNS protocol explicitly 1743 allows domain labels to contain octets beyond the ASCII range 1744 (0000..007F), and this document does not change that. However, 1745 although the interpretation of octets 0080..00FF is well-defined in 1746 the DNS, many application protocols support only ASCII labels and 1747 there is no defined interpretation of these non-ASCII octets as 1748 characters and, in particular, no interpretation of case-independent 1749 matching for them (see, e.g., [RFC4343]). If labels containing these 1750 octets are returned to applications, unpredictable behavior could 1751 result. The A-label form, which cannot contain those characters, is 1752 the only standard representation for internationalized labels in the 1753 DNS protocol. 1755 8.2. Root and other DNS Server Considerations 1757 IDNs in A-label form will generally be somewhat longer than current 1758 domain names, so the bandwidth needed by the root servers is likely 1759 to go up by a small amount. Also, queries and responses for IDNs 1760 will probably be somewhat longer than typical queries historically, 1761 so EDNS0 [RFC2671] support may be more important (otherwise, queries 1762 and responses may be forced to go to TCP instead of UDP). 1764 9. Internationalization Considerations 1766 DNS labels and fully-qualified domain names provide mnemonics that 1767 assist in identifying and referring to resources on the Internet. 1768 IDNs expand the range of those mnemonics to include those based on 1769 languages and character sets other than Western European and Roman- 1770 derived ones. But domain "names" are not, in general, words in any 1771 language. The recommendations of the IETF policy on character sets 1772 and languages, (BCP 18 [RFC2277]) are applicable to situations in 1773 which language identification is used to provide language-specific 1774 contexts. The DNS is, by contrast, global and international and 1775 ultimately has nothing to do with languages. Adding languages (or 1776 similar context) to IDNs generally, or to DNS matching in particular, 1777 would imply context dependent matching in DNS, which would be a very 1778 significant change to the DNS protocol itself. It would also imply 1779 that users would need to identify the language associated with a 1780 particular label in order to look that label up. That knowledge is 1781 generally not available because many labels are not words in any 1782 language and some may be words in more than one. 1784 10. IANA Considerations 1786 This section gives an overview of IANA registries required for IDNA. 1787 The actual definitions of, and specifications for, the first two, 1788 which must be newly-created for IDNA2008, appear in 1789 [IDNA2008-Tables]. This document describes the registries but does 1790 not specify any IANA actions. 1792 10.1. IDNA Character Registry 1794 The distinction among the major categories "UNASSIGNED", 1795 "DISALLOWED", "PROTOCOL-VALID", and "CONTEXTUAL RULE REQUIRED" is 1796 made by special categories and rules that are integral elements of 1797 [IDNA2008-Tables]. While not normative, an IANA registry of 1798 characters and scripts and their categories, updated for each new 1799 version of Unicode and the characters it contains, will be convenient 1800 for programming and validation purposes. The details of this 1801 registry are specified in [IDNA2008-Tables]. 1803 10.2. IDNA Context Registry 1805 IANA will create and maintain a list of approved contextual rules for 1806 characters that are defined in the IDNA Character Registry list as 1807 requiring a Contextual Rule (i.e., the types of rule described in 1808 Section 3.1.2). The details for those rules appear in 1809 [IDNA2008-Tables]. 1811 10.3. IANA Repository of IDN Practices of TLDs 1813 This registry, historically described as the "IANA Language Character 1814 Set Registry" or "IANA Script Registry" (both somewhat misleading 1815 terms) is maintained by IANA at the request of ICANN. It is used to 1816 provide a central documentation repository of the IDN policies used 1817 by top level domain (TLD) registries who volunteer to contribute to 1818 it and is used in conjunction with ICANN Guidelines for IDN use. 1820 It is not an IETF-managed registry and, while the protocol changes 1821 specified here may call for some revisions to the tables, IDNA2008 1822 has no direct effect on that registry and no IANA action is required 1823 as a result. 1825 11. Security Considerations 1827 11.1. General Security Issues with IDNA 1829 This document is purely explanatory and informational and 1830 consequently introduces no new security issues. It would, of course, 1831 be a poor idea for someone to try to implement from it; such an 1832 attempt would almost certainly lead to interoperability problems and 1833 might lead to security ones. A discussion of security issues with 1834 IDNA, including some relevant history, appears in [IDNA2008-Defs]. 1836 12. Acknowledgments 1838 The editor and contributors would like to express their thanks to 1839 those who contributed significant early (pre-WG) review comments, 1840 sometimes accompanied by text, Paul Hoffman, Simon Josefsson, and Sam 1841 Weiler. In addition, some specific ideas were incorporated from 1842 suggestions, text, or comments about sections that were unclear 1843 supplied by Vint Cerf, Frank Ellerman, Michael Everson, Asmus 1844 Freytag, Erik van der Poel, Michel Suignard, and Ken Whistler. 1845 Thanks are also due to Vint Cerf, Lisa Dusseault, Debbie Garside, and 1846 Jefsey Morfin for conversations that led to considerable improvements 1847 in the content of this document and to several others, including Ben 1848 Campbell, Martin Duerst, Subramanian Moonesamy, Peter Saint-Andre, 1849 and Dan Winship, for catching specific errors and recommending 1850 corrections. 1852 A meeting was held on 30 January 2008 to attempt to reconcile 1853 differences in perspective and terminology about this set of 1854 specifications between the design team and members of the Unicode 1855 Technical Consortium. The discussions at and subsequent to that 1856 meeting were very helpful in focusing the issues and in refining the 1857 specifications. The active participants at that meeting were (in 1858 alphabetic order as usual) Harald Alvestrand, Vint Cerf, Tina Dam, 1859 Mark Davis, Lisa Dusseault, Patrik Faltstrom (by telephone), Cary 1860 Karp, John Klensin, Warren Kumari, Lisa Moore, Erik van der Poel, 1861 Michel Suignard, and Ken Whistler. We express our thanks to Google 1862 for support of that meeting and to the participants for their 1863 contributions. 1865 Useful comments and text on the WG versions of the draft were 1866 received from many participants in the IETF "IDNABIS" WG and a number 1867 of document changes resulted from mailing list discussions made by 1868 that group. Marcos Sanz provided specific analysis and suggestions 1869 that were exceptionally helpful in refining the text, as did Vint 1870 Cerf, Martin Duerst, Andrew Sullivan, and Ken Whistler. Lisa 1871 Dusseault provided extensive editorial suggestions during the spring 1872 of 2009, most of which were incorporated. 1874 13. Contributors 1876 While the listed editor held the pen, the core of this document and 1877 the initial WG version represents the joint work and conclusions of 1878 an ad hoc design team consisting of the editor and, in alphabetic 1879 order, Harald Alvestrand, Tina Dam, Patrik Faltstrom, and Cary Karp. 1880 Considerable material describing mapping principles has been 1881 incorporated from a draft of [IDNA2008-Mapping] by Pete Resnick and 1882 Paul Hoffman. In addition, there were many specific contributions 1883 and helpful comments from those listed in the Acknowledgments section 1884 and others who have contributed to the development and use of the 1885 IDNA protocols. 1887 14. References 1889 14.1. Normative References 1891 [ASCII] American National Standards Institute (formerly United 1892 States of America Standards Institute), "USA Code for 1893 Information Interchange", ANSI X3.4-1968, 1968. 1895 ANSI X3.4-1968 has been replaced by newer versions with 1896 slight modifications, but the 1968 version remains 1897 definitive for the Internet. 1899 [IDNA2008-Bidi] 1900 Alvestrand, H. and C. Karp, "An updated IDNA criterion for 1901 right to left scripts", August 2009, . 1904 [IDNA2008-Defs] 1905 Klensin, J., "Internationalized Domain Names for 1906 Applications (IDNA): Definitions and Document Framework", 1907 August 2009, . 1910 [IDNA2008-Protocol] 1911 Klensin, J., "Internationalized Domain Names in 1912 Applications (IDNA): Protocol", August 2009, . 1915 [IDNA2008-Tables] 1916 Faltstrom, P., "The Unicode Code Points and IDNA", 1917 August 2009, . 1920 A version of this document is available in HTML format at 1921 http://stupid.domain.name/idnabis/ 1922 draft-ietf-idnabis-tables-06.html 1924 [RFC3490] Faltstrom, P., Hoffman, P., and A. Costello, 1925 "Internationalizing Domain Names in Applications (IDNA)", 1926 RFC 3490, March 2003. 1928 [RFC3492] Costello, A., "Punycode: A Bootstring encoding of Unicode 1929 for Internationalized Domain Names in Applications 1930 (IDNA)", RFC 3492, March 2003. 1932 [Unicode-UAX15] 1933 The Unicode Consortium, "Unicode Standard Annex #15: 1934 Unicode Normalization Forms", March 2008, 1935 . 1937 [Unicode51] 1938 The Unicode Consortium, "The Unicode Standard, Version 1939 5.1.0", 2008. 1941 defined by: The Unicode Standard, Version 5.0, Boston, MA, 1942 Addison-Wesley, 2007, ISBN 0-321-48091-0, as amended by 1943 Unicode 5.1.0 1944 (http://www.unicode.org/versions/Unicode5.1.0/). 1946 14.2. Informative References 1948 [BIG5] Institute for Information Industry of Taiwan, "Computer 1949 Chinese Glyph and Character Code Mapping Table, Technical 1950 Report C-26", 1984. 1952 There are several forms and variations and a closely- 1953 related standard, CNS 11643. See the discussion in 1954 Chapter 3 of Lunde, K., CJKV Information Processing, 1955 O'Reilly & Associates, 1999 1957 [GB18030] "Chinese National Standard GB 18030-2000: Information 1958 Technology -- Chinese ideograms coded character set for 1959 information interchange -- Extension for the basic set.", 1960 2000. 1962 [IDNA2008-Mapping] 1963 Resnick, P., "Mapping Characters in IDNA", August 2009, . 1967 [RFC0810] Feinler, E., Harrenstien, K., Su, Z., and V. White, "DoD 1968 Internet host table specification", RFC 810, March 1982. 1970 [RFC0952] Harrenstien, K., Stahl, M., and E. Feinler, "DoD Internet 1971 host table specification", RFC 952, October 1985. 1973 [RFC1034] Mockapetris, P., "Domain names - concepts and facilities", 1974 STD 13, RFC 1034, November 1987. 1976 [RFC1035] Mockapetris, P., "Domain names - implementation and 1977 specification", STD 13, RFC 1035, November 1987. 1979 [RFC1123] Braden, R., "Requirements for Internet Hosts - Application 1980 and Support", STD 3, RFC 1123, October 1989. 1982 [RFC2136] Vixie, P., Thomson, S., Rekhter, Y., and J. Bound, 1983 "Dynamic Updates in the Domain Name System (DNS UPDATE)", 1984 RFC 2136, April 1997. 1986 [RFC2181] Elz, R. and R. Bush, "Clarifications to the DNS 1987 Specification", RFC 2181, July 1997. 1989 [RFC2277] Alvestrand, H., "IETF Policy on Character Sets and 1990 Languages", BCP 18, RFC 2277, January 1998. 1992 [RFC2671] Vixie, P., "Extension Mechanisms for DNS (EDNS0)", 1993 RFC 2671, August 1999. 1995 [RFC2673] Crawford, M., "Binary Labels in the Domain Name System", 1996 RFC 2673, August 1999. 1998 [RFC2782] Gulbrandsen, A., Vixie, P., and L. Esibov, "A DNS RR for 1999 specifying the location of services (DNS SRV)", RFC 2782, 2000 February 2000. 2002 [RFC3454] Hoffman, P. and M. Blanchet, "Preparation of 2003 Internationalized Strings ("stringprep")", RFC 3454, 2004 December 2002. 2006 [RFC3491] Hoffman, P. and M. Blanchet, "Nameprep: A Stringprep 2007 Profile for Internationalized Domain Names (IDN)", 2008 RFC 3491, March 2003. 2010 [RFC3743] Konishi, K., Huang, K., Qian, H., and Y. Ko, "Joint 2011 Engineering Team (JET) Guidelines for Internationalized 2012 Domain Names (IDN) Registration and Administration for 2013 Chinese, Japanese, and Korean", RFC 3743, April 2004. 2015 [RFC3987] Duerst, M. and M. Suignard, "Internationalized Resource 2016 Identifiers (IRIs)", RFC 3987, January 2005. 2018 [RFC4290] Klensin, J., "Suggested Practices for Registration of 2019 Internationalized Domain Names (IDN)", RFC 4290, 2020 December 2005. 2022 [RFC4343] Eastlake, D., "Domain Name System (DNS) Case Insensitivity 2023 Clarification", RFC 4343, January 2006. 2025 [RFC4690] Klensin, J., Faltstrom, P., Karp, C., and IAB, "Review and 2026 Recommendations for Internationalized Domain Names 2027 (IDNs)", RFC 4690, September 2006. 2029 [RFC4713] Lee, X., Mao, W., Chen, E., Hsu, N., and J. Klensin, 2030 "Registration and Administration Recommendations for 2031 Chinese Domain Names", RFC 4713, October 2006. 2033 [Unicode-Security] 2034 The Unicode Consortium, "Unicode Technical Standard #39: 2035 Unicode Security Mechanisms", August 2008, 2036 . 2038 [Unicode-UAX31] 2039 The Unicode Consortium, "Unicode Standard Annex #31: 2040 Unicode Identifier and Pattern Syntax", March 2008, 2041 . 2043 [Unicode-UTR36] 2044 The Unicode Consortium, "Unicode Technical Report #36: 2045 Unicode Security Considerations", July 2008, 2046 . 2048 Appendix A. Change Log 2050 [[ RFC Editor: Please remove this appendix. ]] 2052 A.1. Changes between Version -00 and Version -01 of 2053 draft-ietf-idnabis-rationale 2055 o Clarified the U-label definition to note that U-labels must 2056 contain at least one non-ASCII character. Also clarified the 2057 relationship among label types. 2059 o Rewrote the discussion of Labels in Registration (Section 7.1.2) 2060 and related text about IDNA-validity (in the "Defs" document as of 2061 -04 of this one) to narrow its focus and remove more general 2062 restrictions. Added a temporary note in line to explain the 2063 situation. 2065 o Changed the "IDNA uses Unicode" statement to focus on 2066 compatibility with IDNA2003 and avoid more general or 2067 controversial assertions. 2069 o Added a discussion of examples to Section 7.1 2071 o Made a number of other small editorial changes and corrections 2072 suggested by Mark Davis. 2074 o Added several more discussion anchors and notes and expanded or 2075 updated some existing ones. 2077 A.2. Version -02 2079 o Trimmed change log, removing information about pre-WG drafts. 2081 o Adjusted discussion of Contextual Rules to match the new location 2082 of the tables and some conceptual material. 2084 o Rewrote the material on preprocessing somewhat. 2086 o Moved the material about relationships with IDNA2003 to be part of 2087 a single section on transitions. 2089 o Removed several placeholders and made editorial changes in 2090 accordance with decisions made at IETF 72 in Dublin and not 2091 disputed on the mailing list. 2093 A.3. Version -03 2095 This special update to the Rationale document is intended to try to 2096 get the discussion of what is normative or not under control. While 2097 the IETF does not normally annotate individual sections of documents 2098 with whether they are normative or not, concerns that we don't know 2099 which is which, claims that some material is normative that would be 2100 problematic if so classified, etc., argue that we should at least be 2101 able to have a clear discussion on the subject. 2103 Two annotations have been applied to sections that might reasonably 2104 be considered normative. One annotation is based on the list of 2105 sections in Mark Davis's note of 29 September (http:// 2106 www.alvestrand.no/pipermail/idna-update/2008-September/002667.html). 2107 The other is based on an elaboration of John Klensin's response on 7 2108 October (http://www.alvestrand.no/pipermail/idna-update/2008-October/ 2109 002691.html). These should just be considered two suggestions to 2110 illuminate and, one hopes, advance the Working Group's discussions. 2112 Some additional editorial changes have been made, but they are 2113 basically trivial. In the editor's judgment, it is not possible to 2114 make significantly more progress with this document until the matter 2115 of document organization is settled. 2117 A.4. Version -04 2119 o Definitional and other normative material moved to new document 2120 (draft-ietf-idnabis-defs). Version -03 annotations removed. 2122 o Material on differences between IDNA2003 and IDNA2008 moved to an 2123 appendix in Protocol. 2125 o Material left over from the origins of this document as a 2126 preliminary proposal has been removed or rewritten. 2128 o Changes made to reflect consensus call results, including removing 2129 several placeholder notes for discussion. 2131 o Added more material, including discussion of historic scripts, to 2132 Section 3.2 on registration policies. 2134 o Added a new section (Section 7.2) to contain specific discussion 2135 of handling of characters that are interpreted differently in 2136 input to IDNA2003 and 2008. 2138 o Some material, including this section/appendix, rearranged. 2140 A.5. Version -05 2142 o Many small editorial changes, including changes to eliminate the 2143 last vestiges of what appeared to be 2119 language (upper-case 2144 MUST, SHOULD, or MAY) and small adjustments to terminology. 2146 A.6. Version -06 2148 o Removed Security Considerations material and pointed to Defs, 2149 where it now appears as of version 05. 2151 o Started changing uses of "IDNA2008" in running text to "in these 2152 specifications" or the equivalent. These documents are titled 2153 simply "IDNA"; once they are standardized, "the current version" 2154 may be a more appropriate reference than one containing a year. 2155 As discussed on the mailing list, we can and should discuss how to 2156 refer to these documents at an appropriate time (e.g., when we 2157 know when we will be finished) but, in the interim, it seems 2158 appropriate to simply start getting rid of the version-specific 2159 terminology where it can naturally be removed. 2161 o Additional discussion of mappings, etc., especially for case- 2162 sensitivity. 2164 o Clarified relationship to base DNS specifications. 2166 o Consolidated discussion of lookup of unassigned characters. 2168 o More editorial fine-tuning. 2170 A.7. Version -07 2172 o Revised terminology by adding terms: NR-LDH-label, Invalid-A-label 2173 (or False-A-label), R-LDH-label, valid IDNA-label in 2174 Section 1.3.2. 2176 o Moved the "name server considerations" material to this document 2177 from Protocol because it is non-normative and not part of the 2178 protocol itself. 2180 o To improve clarity, redid discussion of the reasons why looking up 2181 unassigned code points is prohibited. 2183 o Editorial and other non-substantive corrections to reflect earlier 2184 errors as well as new definitions and terminology. 2186 A.8. Version -08 2188 o Slight revision to "contextual" discussion (Section 3.1.2) and 2189 moving it to a separate subsection, rather than under "PVALID", 2190 for better parallelism with Tables. Also reflected Mark's 2191 comments about the limitations of the approach. 2193 o Added placeholder notes as reminders of where references to the 2194 other documents need Section numbers. More of these will be added 2195 as needed (feel free to identify relevant places), but the actual 2196 section numbers will not be inserted until the documents are 2197 completely stable, i.e., on their way to the RFC Editor. 2199 A.9. Version -09 2201 o Small editorial changes to clarify transition possibilities. 2203 o Small clarification to the description of DNS "exact match". 2205 o Added discussion of adding characters to an existing script to the 2206 discussion of unassigned code point transitions in Section 7.7. 2208 o Tightened up the discussion of non-ASCII string processing 2209 (Section 8.1) slightly. 2211 o Removed some placeholders and comments that have been around long 2212 enough to be considered acceptable or that no longer seem 2213 necessary for other reasons. 2215 A.10. Version -10 2217 o Extensive editorial improvements, mostly due to suggestions from 2218 Lisa Dusseault. 2220 o Changes required for the new "mapping" approach and document have, 2221 in general, not been incorporated despite several suggestions. 2222 The editor intends to wait until the mapping model is stable, or 2223 at least until -11 of this document, before trying to incorporate 2224 those suggestions. 2226 A.11. Version -11 2228 o Several placeholders for additional material or editing have been 2229 removed since no comments have been received. 2231 o Updated references. 2233 o Corrected an apparent patching error in Section 1.6 and another 2234 one in Section 4.3. 2236 o Adjusted several sections that had not properly reflected removal 2237 of the material that is now in the Definitions document and 2238 removed an unnecessary one. 2240 o New material added to Section 3.2 about registration policy issues 2241 to reflect discussions on the mailing list. 2243 o Incorporated mapping material from the former "Architectural 2244 Principles" of version -01 of the Mapping draft into Section 6 and 2245 removed most of the prior mapping material and explanations. 2247 o Eliminated the former Section 7.3 ("More Flexibility in User 2248 Agents"), moving its material into Section 4.2. The replacement 2249 section is basically a placeholder to retain the mapping issues as 2250 one of the migration topics. Note that this item and the previous 2251 one involve considerable text, so people should check things 2252 carefully. 2254 o Corrected several typographical and editorial errors that don't 2255 fall into any of the above categories. 2257 A.12. Version -12 2259 o Got rid of the term "IDNA-valid". It no longer appears in 2260 Definitions and we didn't really need the extra term. Where the 2261 concept was needed, the text now says "valid under IDNA" or 2262 equivalent. 2264 o Adjusted Acknowledgments to remove Mark Davis's name, per his 2265 request and advice from IETF Trust Counsel. 2267 o Incorporated other changes from WG Last Call. 2269 o Small typographical and editorial corrections. 2271 A.13. Version -13 2273 o Substituted in Section numbers to references to other IDNA2008 2274 documents. 2276 A.14. Version -14 2278 This is the version of the document produced to reflect comments on 2279 IETF Last Call. For the convenience of those who made comments and 2280 of the IESG in evaluating them, this section therefore identifies 2281 non-editorial changes made in response to Last Call comments in 2282 somewhat more detail than may be usual. 2284 o Removed the discussion of DNSSEC after extensive discussion on the 2285 IETF and IDNABIS lists. 2287 o Modified the discussion of prefix changes to make it clear that 2288 the decisions have been made, rather than still representing open 2289 issues. (Dan Winship review, 20091013) 2291 o Suggested explicit identification of domain name slots in 2292 protocols that use IDNA. Peter Saint-Andre, 20091019. 2294 o Several other clarifications as suggested by Peter Saint-Andre, 2295 20091019. 2297 o Several minor editorial corrections per suggestions in Ben 2298 Campbell's Gen-ART review 20091013. 2300 o Typo corrections. 2302 A.15. Version -15 2304 o Rewrote and expanded the "transition" material of Section 7.2. 2306 A.16. Version -16 2308 This version contains changes made at IESG request during their 2309 review. Some additional comments were logged during or immediately 2310 before the 7 January 2010 teleconference, so this is not the final 2311 I-D version. 2313 o Altered use of "these documents" and "these specifications" back 2314 to "IDNA2008", undoing the change made in Appendix A.6. The 2315 convolutions became ambiguous in places. 2317 o Added a sentence to the Introduction to make the non-normative 2318 status of this document even more clear and added references to 2319 7.1.2 and 7.1.3 to point to the more formal definitions. 2321 Author's Address 2323 John C Klensin 2324 1770 Massachusetts Ave, Ste 322 2325 Cambridge, MA 02140 2326 USA 2328 Phone: +1 617 245 1457 2329 Email: john+ietf@jck.com