idnits 2.17.00 (12 Aug 2021) /tmp/idnits1061/draft-ietf-idnabis-rationale-12.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** The document seems to lack a License Notice according IETF Trust Provisions of 28 Dec 2009, Section 6.b.i or Provisions of 12 Sep 2009 Section 6.b -- however, there's a paragraph with a matching beginning. Boilerplate error? (You're using the IETF Trust Provisions' Section 6.b License Notice from 12 Feb 2009 rather than one of the newer Notices. 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 80: '... CONTEXTUAL RULE REQUIRED . . . . . . ...' RFC 2119 keyword, line 460: '... CONTEXTUAL RULE REQUIRED, DISALLOWED ...' RFC 2119 keyword, line 486: '...3.1.2. CONTEXTUAL RULE REQUIRED...' RFC 2119 keyword, line 1559: '...nd CONTEXTUAL RULE REQUIRED categories...' RFC 2119 keyword, line 1688: '...esented directly in ASCII MUST use the...' (2 more instances...) 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 (September 10, 2009) is 4635 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Unused Reference: 'ASCII' is defined on line 1835, but no explicit reference was found in the text == Unused Reference: 'Unicode-UAX15' is defined on line 1876, but no explicit reference was found in the text == Unused Reference: 'BIG5' is defined on line 1892, but no explicit reference was found in the text == Unused Reference: 'GB18030' is defined on line 1901, but no explicit reference was found in the text == Unused Reference: 'RFC0810' is defined on line 1911, but no explicit reference was found in the text == Unused Reference: 'RFC2673' is defined on line 1942, but no explicit reference was found in the text == Unused Reference: 'RFC4690' is defined on line 1972, but no explicit reference was found in the text == Unused Reference: 'Unicode-UTR36' is defined on line 1990, 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 2535 (Obsoleted by RFC 4033, RFC 4034, RFC 4035) -- 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 (==), 9 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group J. Klensin 3 Internet-Draft September 10, 2009 4 Intended status: Informational 5 Expires: March 14, 2010 7 Internationalized Domain Names for Applications (IDNA): Background, 8 Explanation, and Rationale 9 draft-ietf-idnabis-rationale-12.txt 11 Status of this Memo 13 This Internet-Draft is submitted to IETF in full conformance with the 14 provisions of BCP 78 and BCP 79. This document may contain material 15 from IETF Documents or IETF Contributions published or made publicly 16 available before November 10, 2008. The person(s) controlling the 17 copyright in some of this material may not have granted the IETF 18 Trust the right to allow modifications of such material outside the 19 IETF Standards Process. Without obtaining an adequate license from 20 the person(s) controlling the copyright in such materials, this 21 document may not be modified outside the IETF Standards Process, and 22 derivative works of it may not be created outside the IETF Standards 23 Process, except to format it for publication as an RFC or to 24 translate it into languages other than English. 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 March 14, 2010. 44 Copyright Notice 46 Copyright (c) 2009 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 in effect on the date of 51 publication of this document (http://trustee.ietf.org/license-info). 52 Please review these documents carefully, as they describe your rights 53 and restrictions with respect to this document. 55 Abstract 57 Several years have passed since the original protocol for 58 Internationalized Domain Names (IDNs) was completed and deployed. 59 During that time, a number of issues have arisen, including the need 60 to update the system to deal with newer versions of Unicode. Some of 61 these issues require tuning of the existing protocols and the tables 62 on which they depend. This document provides an overview of a 63 revised system and provides explanatory material for its components. 65 Table of Contents 67 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 68 1.1. Context and Overview . . . . . . . . . . . . . . . . . . . 4 69 1.2. Discussion Forum . . . . . . . . . . . . . . . . . . . . . 5 70 1.3. Terminology . . . . . . . . . . . . . . . . . . . . . . . 5 71 1.3.1. DNS "Name" Terminology . . . . . . . . . . . . . . . . 5 72 1.3.2. New Terminology and Restrictions . . . . . . . . . . . 6 73 1.4. Objectives . . . . . . . . . . . . . . . . . . . . . . . . 6 74 1.5. Applicability and Function of IDNA . . . . . . . . . . . . 7 75 1.6. Comprehensibility of IDNA Mechanisms and Processing . . . 8 76 2. Processing in IDNA2008 . . . . . . . . . . . . . . . . . . . . 9 77 3. Permitted Characters: An Inclusion List . . . . . . . . . . . 9 78 3.1. A Tiered Model of Permitted Characters and Labels . . . . 10 79 3.1.1. PROTOCOL-VALID . . . . . . . . . . . . . . . . . . . . 10 80 3.1.2. CONTEXTUAL RULE REQUIRED . . . . . . . . . . . . . . . 11 81 3.1.2.2. Rules and Their Application . . . . . . . . . . . 12 82 3.1.3. DISALLOWED . . . . . . . . . . . . . . . . . . . . . . 12 83 3.1.4. UNASSIGNED . . . . . . . . . . . . . . . . . . . . . . 13 84 3.2. Registration Policy . . . . . . . . . . . . . . . . . . . 14 85 3.3. Layered Restrictions: Tables, Context, Registration, 86 Applications . . . . . . . . . . . . . . . . . . . . . . . 14 87 4. Issues that Constrain Possible Solutions . . . . . . . . . . . 15 88 4.1. Display and Network Order . . . . . . . . . . . . . . . . 15 89 4.2. Entry and Display in Applications . . . . . . . . . . . . 16 90 4.3. Linguistic Expectations: Ligatures, Digraphs, and 91 Alternate Character Forms . . . . . . . . . . . . . . . . 18 92 4.4. Case Mapping and Related Issues . . . . . . . . . . . . . 20 93 4.5. Right to Left Text . . . . . . . . . . . . . . . . . . . . 21 94 5. IDNs and the Robustness Principle . . . . . . . . . . . . . . 21 95 6. Front-end and User Interface Processing for Lookup . . . . . . 22 96 7. Migration from IDNA2003 and Unicode Version Synchronization . 24 97 7.1. Design Criteria . . . . . . . . . . . . . . . . . . . . . 24 98 7.1.1. Summary and Discussion of IDNA Validity Criteria . . . 25 99 7.1.2. Labels in Registration . . . . . . . . . . . . . . . . 25 100 7.1.3. Labels in Lookup . . . . . . . . . . . . . . . . . . . 26 101 7.2. Changes in Character Interpretations . . . . . . . . . . . 28 102 7.3. Character Mapping . . . . . . . . . . . . . . . . . . . . 29 103 7.4. The Question of Prefix Changes . . . . . . . . . . . . . . 29 104 7.4.1. Conditions Requiring a Prefix Change . . . . . . . . . 29 105 7.4.2. Conditions Not Requiring a Prefix Change . . . . . . . 30 106 7.4.3. Implications of Prefix Changes . . . . . . . . . . . . 30 107 7.5. Stringprep Changes and Compatibility . . . . . . . . . . . 31 108 7.6. The Symbol Question . . . . . . . . . . . . . . . . . . . 32 109 7.7. Migration Between Unicode Versions: Unassigned Code 110 Points . . . . . . . . . . . . . . . . . . . . . . . . . . 33 111 7.8. Other Compatibility Issues . . . . . . . . . . . . . . . . 35 112 8. Name Server Considerations . . . . . . . . . . . . . . . . . . 35 113 8.1. Processing Non-ASCII Strings . . . . . . . . . . . . . . . 35 114 8.2. DNSSEC Authentication of IDN Domain Names . . . . . . . . 36 115 8.3. Root and other DNS Server Considerations . . . . . . . . . 36 116 9. Internationalization Considerations . . . . . . . . . . . . . 36 117 10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 37 118 10.1. IDNA Character Registry . . . . . . . . . . . . . . . . . 37 119 10.2. IDNA Context Registry . . . . . . . . . . . . . . . . . . 37 120 10.3. IANA Repository of IDN Practices of TLDs . . . . . . . . . 37 121 11. Security Considerations . . . . . . . . . . . . . . . . . . . 38 122 11.1. General Security Issues with IDNA . . . . . . . . . . . . 38 123 12. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 38 124 13. Contributors . . . . . . . . . . . . . . . . . . . . . . . . . 39 125 14. References . . . . . . . . . . . . . . . . . . . . . . . . . . 39 126 14.1. Normative References . . . . . . . . . . . . . . . . . . . 39 127 14.2. Informative References . . . . . . . . . . . . . . . . . . 40 128 Appendix A. Change Log . . . . . . . . . . . . . . . . . . . . . 42 129 A.1. Changes between Version -00 and Version -01 of 130 draft-ietf-idnabis-rationale . . . . . . . . . . . . . . . 43 131 A.2. Version -02 . . . . . . . . . . . . . . . . . . . . . . . 43 132 A.3. Version -03 . . . . . . . . . . . . . . . . . . . . . . . 43 133 A.4. Version -04 . . . . . . . . . . . . . . . . . . . . . . . 44 134 A.5. Version -05 . . . . . . . . . . . . . . . . . . . . . . . 44 135 A.6. Version -06 . . . . . . . . . . . . . . . . . . . . . . . 45 136 A.7. Version -07 . . . . . . . . . . . . . . . . . . . . . . . 45 137 A.8. Version -08 . . . . . . . . . . . . . . . . . . . . . . . 45 138 A.9. Version -09 . . . . . . . . . . . . . . . . . . . . . . . 46 139 A.10. Version -10 . . . . . . . . . . . . . . . . . . . . . . . 46 140 A.11. Version -11 . . . . . . . . . . . . . . . . . . . . . . . 46 141 A.12. Version -12 . . . . . . . . . . . . . . . . . . . . . . . 47 142 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . . 47 144 1. Introduction 146 1.1. Context and Overview 148 Internationalized Domain Names in Applications (IDNA) is a collection 149 of standards that allow client applications to convert some Unicode 150 mnemonics to an ASCII-compatible encoding form ("ACE") which is a 151 valid DNS label containing only letters, digits, and hyphens. The 152 specific form of ACE label used by IDNA is called an "A-label". A 153 client can look up an exact A-label in the existing DNS, so A-labels 154 do not require any extensions to DNS, upgrades of DNS servers or 155 updates to low-level client libraries. An A-label is recognizable 156 from the prefix "xn--" before the characters produced by the Punycode 157 algorithm [RFC3492], thus a user application can identify an A-label 158 and convert it into Unicode (or some local coded character set) for 159 display. 161 On the registry side, IDNA allows a registry to offer 162 Internationalized Domain Names (IDNs) for registration as A-labels. 163 A registry may offer any subset of valid IDNs, and may apply any 164 restrictions or bundling (grouping of similar labels together in one 165 registration) appropriate for the context of that registry. 166 Registration of labels is sometimes discussed separately from lookup, 167 and is subject to a few specific requirements that do not apply to 168 lookup. 170 DNS clients and registries are subject to some differences in 171 requirements for handling IDNs. In particular, registries are urged 172 to register only exact, valid A-labels, while clients might do some 173 mapping to get from otherwise-invalid user input to a valid A-label. 175 The first version of IDNA was published in 2003 and is referred to 176 here as IDNA2003 to contrast it with the current version, which is 177 known as IDNA2008 (after the year in which IETF work started on it). 178 IDNA2003 consists of four documents: the IDNA base specification 179 [RFC3490], Nameprep [RFC3491], Punycode [RFC3492], and Stringprep 180 [RFC3454]. The current set of documents, IDNA2008, are not dependent 181 on any of the IDNA2003 specifications other than the one for Punycode 182 encoding. References to "these specifications" or "these documents" 183 are to the entire IDNA2008 set listed in [IDNA2008-Defs]. The 184 characters that are valid in A-labels are identified from rules 185 listed in the Tables document [IDNA2008-Tables], but validity can be 186 derived from the Unicode properties of those characters with a very 187 few exceptions. 189 Traditionally, DNS labels are matched case-insensitively 190 [RFC1034][RFC1035]. That convention was preserved in IDNA2003 by a 191 case-folding operation that generally maps capital letters into 192 lower-case ones. However, if case rules are enforced from one 193 language, another language sometimes loses the ability to treat two 194 characters separately. Case-insensitivity is treated slightly 195 differently in IDNA2008. 197 IDNA2003 used Unicode version 3.2 only. In order to keep up with new 198 characters added in new versions of UNICODE, IDNA2008 decouples its 199 rules from any particular version of UNICODE. Instead, the 200 attributes of new characters in Unicode, supplemented by a small 201 number of exception cases, determine how and whether the characters 202 can be used in IDNA labels. 204 This document provides informational context for IDNA2008, including 205 terminology, background, and policy discussions. 207 1.2. Discussion Forum 209 [[ RFC Editor: please remove this section. ]] 211 IDNA2008 is being discussed in the IETF "idnabis" Working Group and 212 on the mailing list idna-update@alvestrand.no 214 1.3. Terminology 216 Terminology for IDNA2008 appears in [IDNA2008-Defs]. That document 217 also contains a roadmap to the IDNA2008 document collection. No 218 attempt should be made to understand this document without the 219 definitions and concepts that appear there. 221 1.3.1. DNS "Name" Terminology 223 In the context of IDNs, the DNS term "name" has introduced some 224 confusion as people speak of DNS labels in terms of the words or 225 phrases of various natural languages. Historically, many of the 226 "names" in the DNS have been mnemonics to identify some particular 227 concept, object, or organization. They are typically rooted in some 228 language because most people think in language-based ways. But, 229 because they are mnemonics, they need not obey the orthographic 230 conventions of any language: it is not a requirement that it be 231 possible for them to be "words". 233 This distinction is important because the reasonable goal of an IDN 234 effort is not to be able to write the great Klingon (or language of 235 one's choice) novel in DNS labels but to be able to form a usefully 236 broad range of mnemonics in ways that are as natural as possible in a 237 very broad range of scripts. 239 1.3.2. New Terminology and Restrictions 241 These documents introduce new terminology, and precise definitions 242 (in [IDNA2008-Defs]), for the terms "U-label", "A-Label", LDH-label 243 (to which all valid pre-IDNA host names conformed), Reserved-LDH- 244 label (R-LDH-label), XN-label, Fake-A-Label, and Non-Reserved-LDH- 245 label (NR-LDH-label). 247 In addition, the term "putative label" has been adopted to refer to a 248 label that may appear to meet certain definitional constraints but 249 has not yet been sufficiently tested for validity. 251 These definitions are also illustrated in Figure 1 of the Definitions 252 Document [IDNA2008-Defs]. R-LDH-labels contain "--" in the third and 253 fourth character from the beginning of the label. In IDNA-aware 254 applications, only a subset of these reserved labels is permitted to 255 be used, namely the A-label subset. A-labels are a subset of the 256 R-LDH-labels that begin with the case-insensitive string "xn--". 257 Labels that bear this prefix but which are not otherwise valid fall 258 into the "Fake-A-label" category. The non-reserved labels (NR-LDH- 259 labels) are implicitly valid since they do not trigger any 260 resemblance to IDNA-landr NR-LDH-labels. 262 The creation of the Reserved-LDH category is required for three 263 reasons: 265 o to prevent confusion with pre-IDNA coding forms; 267 o to permit future extensions that would require changing the 268 prefix, no matter how unlikely those might be (see Section 7.4); 269 and 271 o to reduce the opportunities for attacks via the Punycode encoding 272 algorithm itself. 274 As with other documents in the IDNA2008 set, this document uses the 275 term "registry" to describe any zone in the DNS. That term, and the 276 terms "zone" or "zone administration", are interchangeable. 278 1.4. Objectives 280 These are the main objectives in revising IDNA. 282 o Use a more recent version of Unicode, and allow IDNA to be 283 independent of Unicode versions, so that IDNA2008 need not be 284 updated for implementations to adopt codepoints from new Unicode 285 versions. 287 o Fix a very small number of code-point categorizations that have 288 turned out to cause problems in the communities that use those 289 code-points. 291 o Reduce the dependency on mapping, in order that the pre-mapped 292 forms (which are not valid IDNA labels) tend to appear less often 293 in various contexts, in favor of valid A-labels. 295 o Fix some details in the bidirectional codepoint handling 296 algorithms. 298 1.5. Applicability and Function of IDNA 300 The IDNA specification solves the problem of extending the repertoire 301 of characters that can be used in domain names to include a large 302 subset of the Unicode repertoire. 304 IDNA does not extend DNS. Instead, the applications (and, by 305 implication, the users) continue to see an exact-match lookup 306 service. Either there is a single exactly-matching (subject to the 307 base DNS requirement of case-insensitive ASCII matching) name or 308 there is no match. This model has served the existing applications 309 well, but it requires, with or without internationalized domain 310 names, that users know the exact spelling of the domain names that 311 are to be typed into applications such as web browsers and mail user 312 agents. The introduction of the larger repertoire of characters 313 potentially makes the set of misspellings larger, especially given 314 that in some cases the same appearance, for example on a business 315 card, might visually match several Unicode code points or several 316 sequences of code points. 318 The IDNA standard does not require any applications to conform to it, 319 nor does it retroactively change those applications. An application 320 can elect to use IDNA in order to support IDN while maintaining 321 interoperability with existing infrastructure. If an application 322 wants to use non-ASCII characters in public DNS domain names, IDNA is 323 the only currently-defined option. Adding IDNA support to an 324 existing application entails changes to the application only, and 325 leaves room for flexibility in front-end processing and more 326 specifically in the user interface (see Section 6). 328 A great deal of the discussion of IDN solutions has focused on 329 transition issues and how IDNs will work in a world where not all of 330 the components have been updated. Proposals that were not chosen by 331 the original IDN Working Group would have depended on updating of 332 user applications, DNS resolvers, and DNS servers in order for a user 333 to apply an internationalized domain name in any form or coding 334 acceptable under that method. While processing must be performed 335 prior to or after access to the DNS, IDNA requires no changes to the 336 DNS protocol or any DNS servers or the resolvers on user's computers. 338 IDNA allows the graceful introduction of IDNs not only by avoiding 339 upgrades to existing infrastructure (such as DNS servers and mail 340 transport agents), but also by allowing some limited use of IDNs in 341 applications by using the ASCII-encoded representation of the labels 342 containing non-ASCII characters. While such names are user- 343 unfriendly to read and type, and hence not optimal for user input, 344 they can be used as a last resort to allow rudimentary IDN usage. 345 For example, they might be the best choice for display if it were 346 known that relevant fonts were not available on the user's computer. 347 In order to allow user-friendly input and output of the IDNs and 348 acceptance of some characters as equivalent to those to be processed 349 according to the protocol, the applications need to be modified to 350 conform to this specification. 352 This version of IDNA uses the Unicode character repertoire, for 353 continuity with the original version of IDNA. 355 1.6. Comprehensibility of IDNA Mechanisms and Processing 357 One goal of IDNA2008, which is aided by the main goal of reducing the 358 dependency on mapping, is to improve the general understanding of how 359 IDNA works and what characters are permitted and what happens to 360 them. Comprehensibility and predictability to users and registrants 361 are important design goals for this effort. End-user applications 362 have an important role to play in increasing this comprehensibility. 364 Any system that tries to handle international characters encounters 365 some common problems. For example, a UI cannot display a character 366 if no font for that character is available. In some cases, 367 internationalization enables effective localization while maintaining 368 some global uniformity but losing some universality. 370 It is difficult to even make suggestions for end-user applications to 371 cope when characters and fonts are not available. Because display 372 functions are rarely controlled by the types of applications that 373 would call upon IDNA, such suggestions will rarely be very effective. 375 Converting between local character sets and normalized Unicode, if 376 needed, is part of this set of user agent issues. This conversion 377 introduces complexity in a system that is not Unicode-native. If a 378 label is converted to a local character set that does not have all 379 the needed characters, or that uses different character-coding 380 principles, the user agent may have to add special logic to avoid or 381 reduce loss of information. 383 The major difficulty may lie in accurately identifying the incoming 384 character set and applying the correct conversion routine. Even more 385 difficult, the local character coding system could be based on 386 conceptually different assumptions than those used by Unicode (e.g., 387 choice of font encodings used for publications in some Indic 388 scripts). Those differences may not easily yield unambiguous 389 conversions or interpretations even if each coding system is 390 internally consistent and adequate to represent the local language 391 and script. 393 IDNA2008 shifts responsibility for character mapping and other 394 adjustments from the protocol (where it was located in IDNA2003) to 395 pre-processing before invoking IDNA itself. The intent is that this 396 change will lead to greater usage of fully-valid A-Labels or U-labels 397 in display, transit and storage, which should aid comprehensibility 398 and predictability. A careful look at pre-processing raises issues 399 about what that pre-processing should do and at what point pre- 400 processing becomes harmful, how universally consistent pre-processing 401 algorithms can be, and how to be compatible with labels prepared in a 402 IDNA2003 context. Those issues are discussed in Section 6 and in the 403 separate document [IDNA2008-Mapping]. 405 2. Processing in IDNA2008 407 These specifications separate Domain Name Registration and Lookup in 408 the protocol specification. Although most steps in the two processes 409 are similar, the separation reflects current practice in which per- 410 registry (DNS zone) restrictions and special processing are applied 411 at registration time but not during lookup. Another significant 412 benefit is that separation facilitates incremental addition of 413 permitted character groups to avoid freezing on one particular 414 version of Unicode. 416 The actual registration and lookup protocols for IDNA2008 are 417 specified in [IDNA2008-Protocol]. 419 3. Permitted Characters: An Inclusion List 421 IDNA2008 adopts the inclusion model. A code-point is assumed to be 422 invalid for IDN use unless it is included as part of a Unicode 423 property-based rule or, in rare cases, included individually by an 424 exception. When an implementation moves to a new version of Unicode, 425 the rules may indicate new valid code-points. 427 This section provides an overview of the model used to establish the 428 algorithm and character lists of [IDNA2008-Tables] and describes the 429 names and applicability of the categories used there. Note that the 430 inclusion of a character in the first category group (Section 3.1.1) 431 does not imply that it can be used indiscriminately; some characters 432 are associated with contextual rules that must be applied as well. 434 The information given in this section is provided to make the rules, 435 tables, and protocol easier to understand. The normative generating 436 rules that correspond to this informal discussion appear in 437 [IDNA2008-Tables] and the rules that actually determine what labels 438 can be registered or looked up are in [IDNA2008-Protocol]. 440 3.1. A Tiered Model of Permitted Characters and Labels 442 Moving to an inclusion model involves a new specification for the 443 list of characters that are permitted in IDNs. In IDNA2003, 444 character validity is independent of context and fixed forever (or 445 until the standard is replaced). However, globally context- 446 independent rules have proved to be impractical because some 447 characters, especially those that are called "Join_Controls" in 448 Unicode, are needed to make reasonable use of some scripts but have 449 no visible effect in others. IDNA2003 prohibited those types of 450 characters entirely by discarding them. We now have a consensus that 451 under some conditions, these "joiner" characters are legitimately 452 needed to allow useful mnemonics for some languages and scripts. In 453 general, context-dependent rules help deal with characters (generally 454 characters that would otherwise be prohibited entirely) that are used 455 differently or perceived differently across different scripts, and 456 allow the standard to be applied more appropriately in cases where a 457 string is not universally handled the same way. 459 IDNA2008 divides all possible Unicode code-points into four 460 categories: PROTOCOL-VALID, CONTEXTUAL RULE REQUIRED, DISALLOWED and 461 UNASSIGNED. 463 3.1.1. PROTOCOL-VALID 465 Characters identified as "PROTOCOL-VALID" (often abbreviated 466 "PVALID") are permitted in IDNs. Their use may be restricted by 467 rules about the context in which they appear or by other rules that 468 apply to the entire label in which they are to be embedded. For 469 example, any label that contains a character in this category that 470 has a "right-to-left" property must be used in context with the 471 "Bidi" rules (see [IDNA2008-Bidi]). 473 The term "PROTOCOL-VALID" is used to stress the fact that the 474 presence of a character in this category does not imply that a given 475 registry need accept registrations containing any of the characters 476 in the category. Registries are still expected to apply judgment 477 about labels they will accept and to maintain rules consistent with 478 those judgments (see [IDNA2008-Protocol] and Section 3.3). 480 Characters that are placed in the "PROTOCOL-VALID" category are 481 expected to never be removed from it or reclassified. While 482 theoretically characters could be removed from Unicode, such removal 483 would be inconsistent with the Unicode stability principles (see 484 [Unicode51], Appendix F) and hence should never occur. 486 3.1.2. CONTEXTUAL RULE REQUIRED 488 Some characters may be unsuitable for general use in IDNs but 489 necessary for the plausible support of some scripts. The two most 490 commonly-cited examples are the zero-width joiner and non-joiner 491 characters (ZWJ, U+200D and ZWNJ, U+200C) but other characters may 492 require special treatment because they would otherwise be DISALLOWED 493 (typically because Unicode considers them punctuation or special 494 symbols) but need to be permitted in limited contexts. Other 495 characters are given this special treatment because they pose 496 exceptional danger of being used to produce misleading labels or to 497 cause unacceptable ambiguity in label matching and interpretation. 499 3.1.2.1. Contextual Restrictions 501 Characters with contextual restrictions are identified as "CONTEXTUAL 502 RULE REQUIRED" and associated with a rule. The rule defines whether 503 the character is valid in a particular string, and also whether the 504 rule itself is to be applied on lookup as well as registration. 506 A distinction is made between characters that indicate or prohibit 507 joining and ones similar to them (known as "CONTEXT-JOINER" or 508 "CONTEXTJ") and other characters requiring contextual treatment 509 ("CONTEXT-OTHER" or "CONTEXTO"). Only the former require full 510 testing at lookup time. 512 It is important to note that these contextual rules cannot prevent 513 all uses of the relevant characters that might be confusing or 514 problematic. What they are expected do is to confine applicability 515 of the characters to scripts (and narrower contexts) where zone 516 administrators are knowledgeable enough about the use of those 517 characters to be prepared to deal with them appropriately. 519 For example, a registry dealing with an Indic script that requires 520 ZWJ and/or ZWNJ as part of the writing system is expected to 521 understand where the characters have visible effect and where they do 522 not and to make registration rules accordingly. By contrast, a 523 registry dealing primarily with Latin or Cyrillic script might not be 524 actively aware that the characters exist, much less about the 525 consequences of embedding them in labels drawn from those scripts and 526 therefore should avoid accepting registrations containing those 527 characters, at least in Latin or Cyrillic-script labels. 529 3.1.2.2. Rules and Their Application 531 Rules have descriptions such as "Must follow a character from Script 532 XYZ", "Must occur only if the entire label is in Script ABC", or 533 "Must occur only if the previous and subsequent characters have the 534 DFG property". The actual rules may be DEFINED or NULL. If present, 535 they may have values of "True" (character may be used in any position 536 in any label), "False" (character may not be used in any label), or 537 may be a set of procedural rules that specify the context in which 538 the character is permitted. 540 Examples of descriptions of typical rules, stated informally and in 541 English, include "Must follow a character from Script XYZ", "Must 542 occur only if the entire label is in Script ABC", "Must occur only if 543 the previous and subsequent characters have the DFG property". 545 Because it is easier to identify these characters than to know that 546 they are actually needed in IDNs or how to establish exactly the 547 right rules for each one, a rule may have a null value in a given 548 version of the tables. Characters associated with null rules are not 549 permitted to appear in putative labels for either registration or 550 lookup. Of course, a later version of the tables might contain a 551 non-null rule. 553 The actual rules and their descriptions are in [IDNA2008-Tables]. 554 [[anchor9: ??? Section number would be good here.]] That document 555 also specifies the creation of a registry for future rules. 557 3.1.3. DISALLOWED 559 Some characters are inappropriate for use in IDNs and are thus 560 excluded for both registration and lookup (i.e., IDNA-conforming 561 applications performing name lookup should verify that these 562 characters are absent; if they are present, the label strings should 563 be rejected rather than converted to A-labels and looked up. Some of 564 these characters are problematic for use in IDNs (such as the 565 FRACTION SLASH character, U+2044), while some of them (such as the 566 various HEART symbols, e.g., U+2665, U+2661, and U+2765, see 567 Section 7.6) simply fall outside the conventions for typical 568 identifiers (basically letters and numbers). 570 Of course, this category would include code points that had been 571 removed entirely from Unicode should such removals ever occur. 573 Characters that are placed in the "DISALLOWED" category are expected 574 to never be removed from it or reclassified. If a character is 575 classified as "DISALLOWED" in error and the error is sufficiently 576 problematic, the only recourse would be either to introduce a new 577 code point into Unicode and classify it as "PROTOCOL-VALID" or for 578 the IETF to accept the considerable costs of an incompatible change 579 and replace the relevant RFC with one containing appropriate 580 exceptions. 582 There is provision for exception cases but, in general, characters 583 are placed into "DISALLOWED" if they fall into one or more of the 584 following groups: 586 o The character is a compatibility equivalent for another character. 587 In slightly more precise Unicode terms, application of 588 normalization method NFKC to the character yields some other 589 character. 591 o The character is an upper-case form or some other form that is 592 mapped to another character by Unicode casefolding. 594 o The character is a symbol or punctuation form or, more generally, 595 something that is not a letter, digit, or a mark that is used to 596 form a letter or digit. 598 3.1.4. UNASSIGNED 600 For convenience in processing and table-building, code points that do 601 not have assigned values in a given version of Unicode are treated as 602 belonging to a special UNASSIGNED category. Such code points are 603 prohibited in labels to be registered or looked up. The category 604 differs from DISALLOWED in that code points are moved out of it by 605 the simple expedient of being assigned in a later version of Unicode 606 (at which point, they are classified into one of the other categories 607 as appropriate). 609 The rationale for restricting the processing of UNASSIGNED characters 610 is simply that the properties of such code points cannot be 611 completely known until actual characters are assigned to them. If, 612 for example, such a code point was permitted to be included in a 613 label to be looked up, and the code point was later to be assigned to 614 a character that required some set of contextual rules, un-updated 615 instances of IDNA-aware software might permit lookup of labels 616 containing the previously-unassigned characters while updated 617 versions of IDNA-aware software might restrict their use in lookup, 618 depending on the contextual rules. It should be clear that under no 619 circumstance should an UNASSIGNED character be permitted in a label 620 to be registered as part of a domain name. 622 3.2. Registration Policy 624 While these recommendations cannot and should not define registry 625 policies, registries should develop and apply additional restrictions 626 as needed to reduce confusion and other problems. For example, it is 627 generally believed that labels containing characters from more than 628 one script are a bad practice although there may be some important 629 exceptions to that principle. Some registries may choose to restrict 630 registrations to characters drawn from a very small number of 631 scripts. For many scripts, the use of variant techniques such as 632 those as described in RFC 3843 [RFC3743] and RFC 4290 [RFC4290], and 633 illustrated for Chinese by the tables described in RFC 4713 [RFC4713] 634 may be helpful in reducing problems that might be perceived by users. 636 In general, users will benefit if registries only permit characters 637 from scripts that are well-understood by the registry or its 638 advisers. If a registry decides to reduce opportunities for 639 confusion by constructing policies that disallow characters used in 640 historic writing systems or characters whose use is restricted to 641 specialized, highly technical contexts, some relevant information may 642 be found in Section 2.4 "Specific Character Adjustments", Table 4 643 "Candidate Characters for Exclusion from Identifiers" of 644 [Unicode-UAX31] and Section 3.1. "General Security Profile for 645 Identifiers" in [Unicode-Security]. 647 The requirement (in [IDNA2008-Protocol] [[anchor10: ?? Section 648 number]]) that registration procedures use only U-labels and/or 649 A-labels is intended to ensure that registrants are fully aware of 650 exactly what is being registered as well as encouraging use of those 651 canonical forms. That provision should not be interpreted as 652 requiring that registrant need to provide characters in a particular 653 code sequence. Registrant input conventions and management are part 654 of registrant-registrar interactions and relationships between 655 registries and registrars and are outside the scope of these 656 standards. 658 It is worth stressing that these principles of policy development and 659 application apply at all levels of the DNS, not only, e.g., TLD or 660 SLD registrations and that even a trivial, "anything is permitted 661 that is valid under the protocol" policy is helpful in that it helps 662 users and application developers know what to expect. 664 3.3. Layered Restrictions: Tables, Context, Registration, Applications 666 The character rules in IDNA2008 are based on the realization that 667 there is no single magic bullet for any of the security, 668 confusability, or other issues associated with IDNs. Instead, the 669 specifications define a variety of approaches. The character tables 670 are the first mechanism, protocol rules about how those characters 671 are applied or restricted in context are the second, and those two in 672 combination constitute the limits of what can be done in the 673 protocol. As discussed in the previous section (Section 3.2), 674 registries are expected to restrict what they permit to be 675 registered, devising and using rules that are designed to optimize 676 the balance between confusion and risk on the one hand and maximum 677 expressiveness in mnemonics on the other. 679 In addition, there is an important role for user agents in warning 680 against label forms that appear problematic given their knowledge of 681 local contexts and conventions. Of course, no approach based on 682 naming or identifiers alone can protect against all threats. 684 4. Issues that Constrain Possible Solutions 686 4.1. Display and Network Order 688 Domain names are always transmitted in network order (the order in 689 which the code points are sent in protocols), but may have a 690 different display order (the order in which the code points are 691 displayed on a screen or paper). When a domain name contains 692 characters that are normally written right to left, display order may 693 be affected although network order is not. It gets even more 694 complicated if left to right and right to left labels are adjacent to 695 each other within a domain name. The decision about the display 696 order is ultimately under the control of user agents --including Web 697 browsers, mail clients, hosted Web applications and many more -- 698 which may be highly localized. Should a domain name abc.def, in 699 which both labels are represented in scripts that are written right 700 to left, be displayed as fed.cba or cba.fed? Applications that are 701 in deployment today are already diverse, and one can find examples of 702 either choice. 704 The picture changes once again when an IDN appears in a 705 Internationalized Resource Identifier (IRI) [RFC3987]. An IRI or 706 Internationalized Email address contains elements other than the 707 domain name. For example, IRIs contain protocol identifiers and 708 field delimiter syntax such as "http://" or "mailto:" while email 709 addresses contain the "@" to separate local parts from domain names. 710 An IRI in network order begins with "http://" followed by domain 711 labels in network order, thus "http://abc.def". 713 User agents are not required to display and allow input of IRIs 714 directly but often do so. Implementors have to choose whether the 715 overall direction of these strings will always be left to right (or 716 right to left) for an IRI or email address. The natural order for a 717 user typing a domain name on a right to left system is fed.cba. 718 Should the R2L user agent reverse the entire domain name each time a 719 domain name is typed? Does this change if the user types "http://" 720 right before typing a domain name, thus implying that the user is 721 beginning at the beginning of the network order IRI? Experience in 722 the 1980s and 1990s with mixing systems in which domain name labels 723 were read in network order (left to right) and those in which those 724 labels were read right to left would predict a great deal of 725 confusion. 727 If each implementation of each application makes its own decisions on 728 these issues, users will develop heuristics that will sometimes fail 729 when switching applications. However, while some display order 730 conventions, voluntarily adopted, would be desirable to reduce 731 confusion, such suggestions are beyond the scope of these 732 specifications. 734 4.2. Entry and Display in Applications 736 Applications can accept and display domain names using any character 737 set or character coding system. The IDNA protocol does not 738 necessarily affect the interface between users and applications. An 739 IDNA-aware application can accept and display internationalized 740 domain names in two formats: the internationalized character set(s) 741 supported by the application (i.e., an appropriate local 742 representation of a U-label), and as an A-label. Applications may 743 allow the display of A-labels, but are encouraged to not do so except 744 as an interface for special purposes, possibly for debugging, or to 745 cope with display limitations. In general, they should allow, but 746 not encourage, user input of A-labels. A-labels are opaque, ugly, 747 and malicious variations on them are not easily detected by users. 748 Where possible, they should thus only be exposed when they are 749 absolutely needed. Because IDN labels can be rendered either as 750 A-labels or U-labels, the application may reasonably have an option 751 for the user to select the preferred method of display. Rendering 752 the U-label should normally be the default. 754 Domain names are often stored and transported in many places. For 755 example, they are part of documents such as mail messages and web 756 pages. They are transported in many parts of many protocols, such as 757 both the control commands of SMTP and associated message body parts, 758 and in the headers and the body content in HTTP. It is important to 759 remember that domain names appear both in domain name slots and in 760 the content that is passed over protocols. 762 In protocols and document formats that define how to handle 763 specification or negotiation of charsets, labels can be encoded in 764 any charset allowed by the protocol or document format. If a 765 protocol or document format only allows one charset, the labels must 766 be given in that charset. Of course, not all charsets can properly 767 represent all labels. If a U-label cannot be displayed in its 768 entirety, the only choice (without loss of information) may be to 769 display the A-label. 771 Where a protocol or document format allows IDNs, labels should be in 772 whatever character encoding and escape mechanism the protocol or 773 document format uses at that place. This provision is intended to 774 prevent situations in which, e.g., UTF-8 domain names appear embedded 775 in text that is otherwise in some other character coding. 777 All protocols that use domain name slots (See Section 2.3.1.6 in 778 [IDNA2008-Defs]) already have the capacity for handling domain names 779 in the ASCII charset. Thus, A-labels can inherently be handled by 780 those protocols. 782 These documents do not specify required mappings between one 783 character or code point and others. An extended discussion of 784 mapping issues occurs in Section 6 and specific recommendations 785 appear in [IDNA2008-Mapping]. In general, IDNA2008 prohibits 786 characters that would be mapped to others by normalization or other 787 rules. As examples, while mathematical characters based on Latin 788 ones are accepted as input to IDNA2003, they are prohibited in 789 IDNA2008. Similarly, upper-case characters, double-width characters, 790 and other variations are prohibited as IDNA input although mapping 791 them as needed in user interfaces is strongly encouraged. 793 Since the rules in [IDNA2008-Tables] have the effect that only 794 strings that are not transformed by NFKC are valid, if an application 795 chooses to perform NFKC normalization before lookup, that operation 796 is safe since this will never make the application unable to look up 797 any valid string. However, as discussed above, the application 798 cannot guarantee that any other application will perform that 799 mapping, so it should be used only with caution and for informed 800 users. 802 In many cases these prohibitions should have no effect on what the 803 user can type as input to the lookup process. It is perfectly 804 reasonable for systems that support user interfaces to perform some 805 character mapping that is appropriate to the local environment. This 806 would normally be done prior to actual invocation of IDNA. At least 807 conceptually, the mapping would be part of the Unicode conversions 808 discussed above and in [IDNA2008-Protocol]. However, those changes 809 will be local ones only -- local to environments in which users will 810 clearly understand that the character forms are equivalent. For use 811 in interchange among systems, it appears to be much more important 812 that U-labels and A-labels can be mapped back and forth without loss 813 of information. 815 One specific, and very important, instance of this strategy arises 816 with case-folding. In the ASCII-only DNS, names are looked up and 817 matched in a case-independent way, but no actual case-folding occurs. 818 Names can be placed in the DNS in either upper or lower case form (or 819 any mixture of them) and that form is preserved, returned in queries, 820 and so on. IDNA2003 approximated that behavior for non-ASCII strings 821 by performing case-folding at registration time (resulting in only 822 lower-case IDNs in the DNS) and when names were looked up. 824 As suggested earlier in this section, it appears to be desirable to 825 do as little character mapping as possible as long as Unicode works 826 correctly (e.g., NFC mapping to resolve different codings for the 827 same character is still necessary although the specifications require 828 that it be performed prior to invoking the protocol) in order to make 829 the mapping between A-labels and U-labels idempotent. Case-mapping 830 is not an exception to this principle. If only lower case characters 831 can be registered in the DNS (i.e., be present in a U-label), then 832 IDNA2008 should prohibit upper-case characters as input even though 833 user interfaces to applications should probably map those characters. 834 Some other considerations reinforce this conclusion. For example, in 835 ASCII case-mapping for individual characters, uppercase(character) 836 must be equal to uppercase(lowercase(character)). That may not be 837 true with IDNs. In some scripts that use case distinctions, there 838 are a few characters that do not have counterparts in one case or the 839 other. The relationship between upper case and lower case may even 840 be language-dependent, with different languages (or even the same 841 language in different areas) expecting different mappings. User 842 agents can meet the expectations of users who are accustomed to the 843 case-insensitive DNS environment by performing case folding prior to 844 IDNA processing, but the IDNA procedures themselves should neither 845 require such mapping nor expect them when they are not natural to the 846 localized environment. 848 4.3. Linguistic Expectations: Ligatures, Digraphs, and Alternate 849 Character Forms 851 Users have expectations about character matching or equivalence that 852 are based on their own languages and the orthography of those 853 languages. These expectations may not always be met in a global 854 system, especially if multiple languages are written using the same 855 script but using different conventions. Some examples: 857 o A Norwegian user might expect a label with the ae-ligature to be 858 treated as the same label as one using the Swedish spelling with 859 a-diaeresis even though applying that mapping to English would be 860 astonishing to users. 862 o A user in German might expect a label with an o-umlaut and a label 863 that had "oe" substituted, but was otherwise the same, treated as 864 equivalent even though that substitution would be a clear error in 865 Swedish. 867 o A Chinese user might expect automatic matching of Simplified and 868 Traditional Chinese characters, but applying that matching for 869 Korean or Japanese text would create considerable confusion. 871 o An English user might expect "theater" and "theatre" to match. 873 A number of languages use alphabetic scripts in which single phonemes 874 are written using two characters, termed a "digraph", for example, 875 the "ph" in "pharmacy" and "telephone". (Such characters can also 876 appear consecutively without forming a digraph, as in "tophat".) 877 Certain digraphs may be indicated typographically by setting the two 878 characters closer together than they would be if used consecutively 879 to represent different phonemes. Some digraphs are fully joined as 880 ligatures. For example, the word "encyclopaedia" is sometimes set 881 with a U+00E6 LATIN SMALL LIGATURE AE. When ligature and digraph 882 forms have the same interpretation across all languages that use a 883 given script, application of Unicode normalization generally resolves 884 the differences and causes them to match. When they have different 885 interpretations, matching must utilize other methods, presumably 886 chosen at the registry level, or users must be educated to understand 887 that matching will not occur. 889 The nature of the problem can be illustrated by many words in the 890 Norwegian language, where the "ae" ligature is the 27th letter of a 891 29-letter extended Latin alphabet. It is equivalent to the 28th 892 letter of the Swedish alphabet (also containing 29 letters), U+00E4 893 LATIN SMALL LETTER A WITH DIAERESIS, for which an "ae" cannot be 894 substituted according to current orthographic standards. That 895 character (U+00E4) is also part of the German alphabet where, unlike 896 in the Nordic languages, the two-character sequence "ae" is usually 897 treated as a fully acceptable alternate orthography for the "umlauted 898 a" character. The inverse is however not true, and those two 899 characters cannot necessarily be combined into an "umlauted a". This 900 also applies to another German character, the "umlauted o" (U+00F6 901 LATIN SMALL LETTER O WITH DIAERESIS) which, for example, cannot be 902 used for writing the name of the author "Goethe". It is also a 903 letter in the Swedish alphabet where, like the "a with diaeresis", it 904 cannot be correctly represented as "oe" and in the Norwegian 905 alphabet, where it is represented, not as "o with diaeresis", but as 906 "slashed o", U+00F8. 908 Some of the ligatures that have explicit code points in Unicode were 909 given special handling in IDNA2003 and now pose additional problems 910 in transition. See Section 7.2. 912 Additional cases with alphabets written right to left are described 913 in Section 4.5. 915 Matching and comparison algorithm selection often requires 916 information about the language being used, context, or both -- 917 information that is not available to IDNA or the DNS. Consequently, 918 these specifications make no attempt to treat combined characters in 919 any special way. A registry that is aware of the language context in 920 which labels are to be registered, and where that language sometimes 921 (or always) treats the two- character sequences as equivalent to the 922 combined form, should give serious consideration to applying a 923 "variant" model [RFC3743][RFC4290], or to prohibiting registration of 924 one of the forms entirely, to reduce the opportunities for user 925 confusion and fraud that would result from the related strings being 926 registered to different parties. 928 4.4. Case Mapping and Related Issues 930 In the DNS, ASCII letters are stored with their case preserved. 931 Matching during the query process is case-independent, but none of 932 the information that might be represented by choices of case has been 933 lost. That model has been accidentally helpful because, as people 934 have created DNS labels by catenating words (or parts of words) to 935 form labels, case has often been used to distinguish among components 936 and make the labels more memorable. 938 Since DNS servers do not get involved in parsing IDNs, they cannot do 939 case-independent matching. Thus, keeping the cases separate in 940 lookup or registration, and doing matching at the server, is not 941 feasible with IDNA or any similar approach. Case-matching must be 942 done, if desired, by IDN clients even though it wasn't done by ASCII- 943 only DNS clients. That situation was recognized in IDNA2003 and 944 nothing in these specifications fundamentally changes it or could do 945 so. In IDNA2003, all characters are case-folded and mapped by 946 clients in a standardized step. 948 Some characters do not have upper case forms. For example the 949 Unicode case folding operation maps Greek Final Form Sigma (U+03C2) 950 to the medial form (U+03C3) and maps Eszett (German Sharp S, U+00DF) 951 to "ss". Neither of these mappings is reversible because the upper 952 case of U+03C3 is the Upper Case Sigma (U+03A3) and "ss" is an ASCII 953 string. IDNA2008 permits, at the risk of some incompatibility, 954 slightly more flexibility in this area by avoiding case folding and 955 treating these characters as themselves. Approaches to handling one- 956 way mappings are discussed in Section 7.2. 958 Because IDNA2003 maps Final Sigma and Eszett to other characters, and 959 the reverse mapping is never possible, that in some sense means that 960 neither Final Sigma nor Eszett can be represented in a IDNA2003 IDN. 961 With IDNA2008, both characters can be used in an IDN and so the 962 A-label used for lookup for any U-label containing those characters, 963 is now different. See Section 7.1 for a discussion of what kinds of 964 changes might require the IDNA prefix to change; after extended 965 discussions, the WG came to consensus that the change for these 966 characters did not justify a prefix change. 968 4.5. Right to Left Text 970 In order to be sure that the directionality of right to left text is 971 unambiguous, IDNA2003 required that any label in which right to left 972 characters appear both starts and ends with them and that it not 973 include any characters with strong left to right properties (that 974 excludes other alphabetic characters but permits European digits). 975 Any other string that contains a right to left character and does not 976 meet those requirements is rejected. This is one of the few places 977 where the IDNA algorithms (both in IDNA2003 and in IDAN2008) examine 978 an entire label, not just individual characters. The algorithmic 979 model used in IDNA2003 rejects the label when the final character in 980 a right to left string requires a combining mark in order to be 981 correctly represented. 983 That prohibition is not acceptable for writing systems for languages 984 written with consonantal alphabets to which diacritical vocalic 985 systems are applied, and for languages with orthographies derived 986 from them where the combining marks may have different functionality. 987 In both cases the combining marks can be essential components of the 988 orthography. Examples of this are Yiddish, written with an extended 989 Hebrew script, and Dhivehi (the official language of Maldives) which 990 is written in the Thaana script (which is, in turn, derived from the 991 Arabic script). IDNA2008 removes the restriction on final combining 992 characters with a new set of rules for right to left scripts and 993 their characters. Those new rules are specified in [IDNA2008-Bidi]. 995 5. IDNs and the Robustness Principle 997 The "Robustness Principle" is often stated as "Be conservative about 998 what you send and liberal in what you accept" (See, e.g., Section 999 1.2.2 of the applications-layer Host Requirements specification 1000 [RFC1123]) This principle applies to IDNA. In applying the principle 1001 to registries as the source ("sender") of all registered and useful 1002 IDNs, registries are responsible for being conservative about what 1003 they register and put out in the Internet. For IDNs to work well, 1004 zone administrators (registries) must have and require sensible 1005 policies about what is registered -- conservative policies -- and 1006 implement and enforce them. 1008 Conversely, lookup applications are expected to reject labels that 1009 clearly violate global (protocol) rules (no one has ever seriously 1010 claimed that being liberal in what is accepted requires being 1011 stupid). However, once one gets past such global rules and deals 1012 with anything sensitive to script or locale, it is necessary to 1013 assume that garbage has not been placed into the DNS, i.e., one must 1014 be liberal about what one is willing to look up in the DNS rather 1015 than guessing about whether it should have been permitted to be 1016 registered. 1018 If a string cannot be successfully found in the DNS after the lookup 1019 processing described here, it makes no difference whether it simply 1020 wasn't registered or was prohibited by some rule at the registry. 1021 Application implementors should be aware that where DNS wildcards are 1022 used, the ability to successfully resolve a name does not guarantee 1023 that it was actually registered. 1025 6. Front-end and User Interface Processing for Lookup 1027 Domain names may be identified and processed in many contexts. They 1028 may be typed in by users either by themselves or embedded in an 1029 identifier such as email addresses, URIs, or IRIs. They may occur in 1030 running text or be processed by one system after being provided in 1031 another. Systems may try to normalize URLs to determine (or guess) 1032 whether a reference is valid or two references point to the same 1033 object without actually looking the objects up (comparison without 1034 lookup is necessary for URI types that are not intended to be 1035 resolved). Some of these goals may be more easily and reliably 1036 satisfied than others. While there are strong arguments for any 1037 domain name that is placed "on the wire" -- transmitted between 1038 systems -- to be in the zero-ambiguity forms of A-labels, it is 1039 inevitable that programs that process domain names will encounter 1040 U-labels or variant forms. 1042 An application that implements the IDNA protocol [IDNA2008-Protocol] 1043 will always take any user input and convert it to a set of Unicode 1044 code points. That user input may be acquired by any of several 1045 different input methods, all with differing conversion processes to 1046 be taken into consideration (e.g., typed on a keyboard, written by 1047 hand onto some sort of digitizer, spoken into a microphone and 1048 interpreted by a speech-to-text engine, etc.). The process of taking 1049 any particular user input and mapping it into a Unicode code point 1050 may be a simple one: If a user strikes the "A" key on a US English 1051 keyboard, without any modifiers such as the "Shift" key held down, in 1052 order to draw a Latin small letter A ("a"), many (perhaps most) 1053 modern operating system input methods will produce to the calling 1054 application the code point U+0061, encoded in a single octet. 1056 Sometimes the process is somewhat more complicated: a user might 1057 strike a particular set of keys to represent a combining macron 1058 followed by striking the "A" key in order to draw a Latin small 1059 letter A with a macron above it. Depending on the operating system, 1060 the input method chosen by the user, and even the parameters with 1061 which the application communicates with the input method, the result 1062 might be the code point U+0101 (encoded as two octets in UTF-8 or 1063 UTF-16, four octets in UTF-32, etc.), the code point U+0061 followed 1064 by the code point U+0304 (again, encoded in three or more octets, 1065 depending upon the encoding used) or even the code point U+FF41 1066 followed by the code point U+0304 (and encoded in some form). And 1067 these examples leave aside the issue of operating systems and input 1068 methods that do not use Unicode code points for their character set. 1070 In every case, applications (with the help of the operating systems 1071 on which they run and the input methods used) need to perform a 1072 mapping from user input into Unicode code points. 1074 The original version of the IDNA protocol [RFC3490] used a model 1075 whereby input was taken from the user, mapped (via whatever input 1076 method mechanisms were used) to a set of Unicode code points, and 1077 then further mapped to a set of Unicode code points using the 1078 Nameprep profile specified in [RFC3491]. In this procedure, there 1079 are two separate mapping steps: First, a mapping done by the input 1080 method (which might be controlled by the operating system, the 1081 application, or some combination) and then a second mapping performed 1082 by the Nameprep portion of the IDNA protocol. The mapping done in 1083 Nameprep includes a particular mapping table to re-map some 1084 characters to other characters, a particular normalization, and a set 1085 of prohibited characters. 1087 Note that the result of the two step mapping process means that the 1088 mapping chosen by the operating system or application in the first 1089 step might differ significantly from the mapping supplied by the 1090 Nameprep profile in the second step. This has advantages and 1091 disadvantages. Of course, the second mapping regularizes what gets 1092 looked up in the DNS, making for better interoperability between 1093 implementations which use the Nameprep mapping. However, the 1094 application or operating system may choose mappings in their input 1095 methods, which when passed through the second (Nameprep) mapping 1096 result in characters that are "surprising" to the end user. 1098 The other important feature of the original version of the IDNA 1099 protocol is that, with very few exceptions, it assumes that any set 1100 of Unicode code points provided to the Nameprep mapping can be mapped 1101 into a string of Unicode code points that are "sensible", even if 1102 that means mapping some code points to nothing (that is, removing the 1103 code points from the string). This allowed maximum flexibility in 1104 input strings. 1106 The present version of IDNA differs significantly in approach from 1107 the original version. First and foremost, it does not provide 1108 explicit mapping instructions. Instead, it assumes that the 1109 application (perhaps via an operating system input method) will do 1110 whatever mapping it requires to convert input into Unicode code 1111 points. This has the advantage of giving flexibility to the 1112 application to choose a mapping that is suitable for its user given 1113 specific user requirements, and avoids the two-step mapping of the 1114 original protocol. Instead of a mapping, the current version of IDNA 1115 provides a set of categories that can be used to specify the valid 1116 code points allowed in a domain name. 1118 In principle, an application ought to take user input of a domain 1119 name and convert it to the set of Unicode code points that represent 1120 the domain name the user intends. As a practical matter, of course, 1121 determining user intent is a tricky business, so an application needs 1122 to choose a reasonable mapping from user input. That may differ 1123 based on the particular circumstances of a user, depending on locale, 1124 language, type of input method, etc. It is up to the application to 1125 make a reasonable choice. 1127 7. Migration from IDNA2003 and Unicode Version Synchronization 1129 7.1. Design Criteria 1131 As mentioned above and in RFC 4690, two key goals of the IDNA2008 1132 design are 1134 o to enable applications to be agnostic about whether they are being 1135 run in environments supporting any Unicode version from 3.2 1136 onward, 1138 o to permit incrementally adding new characters, character groups, 1139 scripts, and other character collections as they are incorporated 1140 into Unicode, doing so without disruption and, in the long term, 1141 without "heavy" processes (an IETF consensus process is required 1142 by the IDNA2008 specifications and is expected to be required and 1143 used until significant experience accumulates with IDNA operations 1144 and new versions of Unicode). 1146 7.1.1. Summary and Discussion of IDNA Validity Criteria 1148 The general criteria for a label to be considered valid under IDNA 1149 are (the actual rules are rigorously defined in the "Protocol" and 1150 "Tables" documents): 1152 o The characters are "letters", marks needed to form letters, 1153 numerals, or other code points used to write words in some 1154 language. Symbols, drawing characters, and various notational 1155 characters are intended to be permanently excluded. There is no 1156 evidence that they are important enough to Internet operations or 1157 internationalization to justify expansion of domain names beyond 1158 the general principle of "letters, digits, and hyphen". 1159 (Additional discussion and rationale for the symbol decision 1160 appears in Section 7.6). 1162 o Other than in very exceptional cases, e.g., where they are needed 1163 to write substantially any word of a given language, punctuation 1164 characters are excluded. The fact that a word exists is not proof 1165 that it should be usable in a DNS label and DNS labels are not 1166 expected to be usable for multiple-word phrases (although they are 1167 certainly not prohibited if the conventions and orthography of a 1168 particular language cause that to be possible). 1170 o Characters that are unassigned (have no character assignment at 1171 all) in the version of Unicode being used by the registry or 1172 application are not permitted, even on lookup. The issues 1173 involved in this decision are discussed in Section 7.7. 1175 o Any character that is mapped to another character by a current 1176 version of NFKC is prohibited as input to IDNA (for either 1177 registration or lookup). With a few exceptions, this principle 1178 excludes any character mapped to another by Nameprep [RFC3491]. 1180 The principles above drive the design of rules that are specified 1181 exactly in [IDNA2008-Tables]. Those rules identify the characters 1182 that are valid under IDNA. The rules themselves are normative, and 1183 the tables are derived from them, rather than vice versa. 1185 7.1.2. Labels in Registration 1187 Any label registered in a DNS zone must be validated -- i.e., the 1188 criteria for that label must be met -- in order for applications to 1189 work as intended. This principle is not new. For example, since the 1190 DNS was first deployed, zone administrators have been expected to 1191 verify that names meet "hostname" requirements [RFC0952] where those 1192 requirements are imposed by the expected applications. Other 1193 applications contexts, such as the later addition of special service 1194 location formats [RFC2782] imposed new requirements on zone 1195 administrators. For zones that will contain IDNs, support for 1196 Unicode version-independence requires restrictions on all strings 1197 placed in the zone. In particular, for such zones: 1199 o Any label that appears to be an A-label, i.e., any label that 1200 starts in "xn--", must be valid under IDNA, i.e., they must be 1201 valid A-labels, as discussed in Section 2 above. 1203 o The Unicode tables (i.e., tables of code points, character 1204 classes, and properties) and IDNA tables (i.e., tables of 1205 contextual rules such as those that appear in the Tables 1206 document), must be consistent on the systems performing or 1207 validating labels to be registered. Note that this does not 1208 require that tables reflect the latest version of Unicode, only 1209 that all tables used on a given system are consistent with each 1210 other. 1212 Under this model, registry tables will need to be updated (both the 1213 Unicode-associated tables and the tables of permitted IDN characters) 1214 to enable a new script or other set of new characters. The registry 1215 will not be affected by newer versions of Unicode, or newly- 1216 authorized characters, until and unless it wishes to support them. 1217 The zone administrator is responsible for verifying validity for IDNA 1218 as well as its local policies -- a more extensive set of checks than 1219 are required for looking up the labels. Systems looking up or 1220 resolving DNS labels, especially IDN DNS labels, must be able to 1221 assume that applicable registration rules were followed for names 1222 entered into the DNS. 1224 7.1.3. Labels in Lookup 1226 Anyone looking up a label in a DNS zone is required to 1228 o Maintain IDNA and Unicode tables that are consistent with regard 1229 to versions, i.e., unless the application actually executes the 1230 classification rules in [IDNA2008-Tables], its IDNA tables must be 1231 derived from the version of Unicode that is supported more 1232 generally on the system. As with registration, the tables need 1233 not reflect the latest version of Unicode but they must be 1234 consistent. 1236 o Validate the characters in labels to be looked up only to the 1237 extent of determining that the U-label does not contain 1238 "DISALLOWED" code points or code points that are unassigned in its 1239 version of Unicode. 1241 o Validate the label itself for conformance with a small number of 1242 whole-label rules. In particular, it must verify that 1244 * there are no leading combining marks, 1246 * the "bidi" conditions are met if right to left characters 1247 appear, 1249 * any required contextual rules are available, and 1251 * any contextual rules that are associated with Joiner Controls 1252 (and "CONTEXTJ" characters more generally) are tested. 1254 o Do not reject labels based on other contextual rules about 1255 characters, including mixed-script label prohibitions. Such rules 1256 may be used to influence presentation decisions in the user 1257 interface, but not to avoid looking up domain names. 1259 To further clarify the rules about handling characters that require 1260 contextual rules, note that one can have a context-required character 1261 (i.e., one that requires a rule), but no rule. In that case, the 1262 character is treated the same way DISALLOWED characters are treated, 1263 until and unless a rule is supplied. That state is more or less 1264 equivalent to "the idea of permitting this character is accepted in 1265 principle, but it won't be permitted in practice until consensus is 1266 reached on a safe way to use it". 1268 The ability to add a rule more or less exempts these characters from 1269 the prohibition against reclassifying characters from DISALLOWED to 1270 PVALID. 1272 And, obviously, "no rule" is different from "have a rule, but the 1273 test either succeeds or fails". 1275 Lookup applications that follow these rules, rather than having their 1276 own criteria for rejecting lookup attempts, are not sensitive to 1277 version incompatibilities with the particular zone registry 1278 associated with the domain name except for labels containing 1279 characters recently added to Unicode. 1281 An application or client that processes names according to this 1282 protocol and then resolves them in the DNS will be able to locate any 1283 name that is registered, as long as those registrations are valid 1284 under IDNA and its version of the IDNA tables is sufficiently up-to- 1285 date to interpret all of the characters in the label. Messages to 1286 users should distinguish between "label contains an unallocated code 1287 point" and other types of lookup failures. A failure on the basis of 1288 an old version of Unicode may lead the user to a desire to upgrade to 1289 a newer version, but will have no other ill effects (this is 1290 consistent with behavior in the transition to the DNS when some hosts 1291 could not yet handle some forms of names or record types). 1293 7.2. Changes in Character Interpretations 1295 In those scripts that make case distinctions, there are a few 1296 characters for which an obvious and unique upper case character has 1297 not historically been available to match a lower case one or vice 1298 versa. For those characters, the mappings used in constructing the 1299 Stringprep tables for IDNA2003, performed using the Unicode CaseFold 1300 operation (See Section 5.8 of the Unicode Standard [Unicode51]), 1301 generate different characters or sets of characters. Those 1302 operations are not reversible and lose even more information than 1303 traditional upper case or lower case transformations, but are more 1304 useful than those transformations for comparison purposes. Two 1305 notable characters of this type are the German character Eszett 1306 (Sharp S, U+00DF) and the Greek Final Form Sigma (U+03C2). The 1307 former is case-folded to the ASCII string "ss", the latter to a 1308 medial (Lower Case) Sigma (U+03C3). 1310 The decision to eliminate mandatory and standardized mappings, 1311 including case folding, from the IDNA2008 protocol in order to make 1312 A-labels and U-labels idempotent made these characters problematic. 1313 If they were to be disallowed, important words and mnemonics could 1314 not be written in orthographically reasonable ways. If they were to 1315 be permitted as distinct characters, there would be no information 1316 loss and registries would have more flexibility, but IDNA2003 and 1317 IDNA2008 lookups might result in different A-labels. 1319 With the understanding that there would be incompatibility either way 1320 but a judgment that the incompatibility was not significant enough to 1321 justify a prefix change, the WG concluded that Eszett and Final Form 1322 Sigma should be treated as distinct and Protocol-Valid characters. 1324 Registries, especially those maintaining zones for third parties, 1325 must decide how to introduce a new service in a way that does not 1326 create confusion or significantly weaken or invalidate existing 1327 identifiers. This is not a new problem; registries were faced with 1328 similar issues when IDNs were introduced and when other new forms of 1329 strings have been permitted as labels. 1331 There are several approaches to problems of this type. Without any 1332 preference or claim to completeness, some of these, all of which have 1333 been used by registries in the past for similar transitions, are: 1335 o Do not permit use of the newly-available character at the registry 1336 level. This might cause lookup failures if a domain name were to 1337 be written with the expectation of the IDNA2003 mapping behavior, 1338 but would eliminate any possibility of false matches. 1340 o Hold a "sunrise"-like arrangement in which holders of labels 1341 containing "ss" in the Eszett case or Lower Case Sigma are given 1342 priority (and perhaps other benefits) for registering the 1343 corresponding string containing Eszett or Final Sigma 1344 respectively. 1346 o Adopt some sort of "variant" approach in which registrants obtain 1347 labels with both character forms. 1349 o Adopt a different form of "variant" approach in which registration 1350 of additional names is either not permitted at all or permitted 1351 only by the registrant who already has one of the names. 1353 7.3. Character Mapping 1355 As discussed at length in Section 6, IDNA2003, via Nameprep (see 1356 Section 7.5), mapped many characters into related ones. Those 1357 mappings no longer exist as requirements in IDNA2008. These 1358 specifications strongly prefer that only A-labels or U-labels be used 1359 in protocol contexts and as much as practical more generally. 1360 IDNA2008 does anticipate situations in which some mapping at the time 1361 of user input into lookup applications is appropriate and desirable. 1362 The issues are discussed in Section 6 and specific recommendations 1363 are made in [IDNA2008-Mapping]. 1365 7.4. The Question of Prefix Changes 1367 The conditions that would require a change in the IDNA ACE prefix 1368 ("xn--" for the version of IDNA specified in [RFC3490]) have been a 1369 great concern to the community. A prefix change would clearly be 1370 necessary if the algorithms were modified in a manner that would 1371 create serious ambiguities during subsequent transition in 1372 registrations. This section summarizes our conclusions about the 1373 conditions under which changes in prefix would be necessary and the 1374 implications of such a change. 1376 7.4.1. Conditions Requiring a Prefix Change 1378 An IDN prefix change is needed if a given string would be looked up 1379 or otherwise interpreted differently depending on the version of the 1380 protocol or tables being used. An IDNA upgrade would require a 1381 prefix change if, and only if, one of the following four conditions 1382 were met: 1384 1. The conversion of an A-label to Unicode (i.e., a U-label) yields 1385 one string under IDNA2003 (RFC3490) and a different string under 1386 IDNA2008. 1388 2. In a significant number of cases, an input string that is valid 1389 under IDNA2003 and also valid under IDNA2008 yields two different 1390 A-labels with the different versions. This condition is believed 1391 to be essentially equivalent to the one above except for a very 1392 small number of edge cases which may not justify a prefix change 1393 (See Section 7.2). 1395 Note that if the input string is valid under one version and not 1396 valid under the other, this condition does not apply. See the 1397 first item in Section 7.4.2, below. 1399 3. A fundamental change is made to the semantics of the string that 1400 is inserted in the DNS, e.g., if a decision were made to try to 1401 include language or script information in the encoding in 1402 addition to the string itself. 1404 4. A sufficiently large number of characters is added to Unicode so 1405 that the Punycode mechanism for block offsets can no longer 1406 reference the higher-numbered planes and blocks. This condition 1407 is unlikely even in the long term and certain not to arise in the 1408 next several years. 1410 7.4.2. Conditions Not Requiring a Prefix Change 1412 As a result of the principles described above, none of the following 1413 changes require a new prefix: 1415 1. Prohibition of some characters as input to IDNA. This may make 1416 names that are now registered inaccessible, but does not change 1417 those names. 1419 2. Adjustments in IDNA tables or actions, including normalization 1420 definitions, that affect characters that were already invalid 1421 under IDNA2003. 1423 3. Changes in the style of the IDNA definition that does not alter 1424 the actions performed by IDNA. 1426 7.4.3. Implications of Prefix Changes 1428 While it might be possible to make a prefix change, the costs of such 1429 a change are considerable. Registries could not convert all IDNA2003 1430 ("xn--") registrations to a new form at the same time and synchronize 1431 that change with applications supporting lookup. Unless all existing 1432 registrations were simply to be declared invalid (and perhaps even 1433 then) systems that needed to support both labels with old prefixes 1434 and labels with new ones would first process a putative label under 1435 the IDNA2008 rules and try to look it up and then, if it were not 1436 found, would process the label under IDNA2003 rules and look it up 1437 again. That process could significantly slow down all processing 1438 that involved IDNs in the DNS especially since a fully-qualified name 1439 might contain a mixture of labels that were registered with the old 1440 and new prefixes. That would make DNS caching very difficult. In 1441 addition, looking up the same input string as two separate A-labels 1442 creates some potential for confusion and attacks, since the labels 1443 could map to different targets and then resolve to different entries 1444 in the DNS. 1446 Consequently, a prefix change is to be avoided if at all possible, 1447 even if it means accepting some IDNA2003 decisions about character 1448 distinctions as irreversible and/or giving special treatment to edge 1449 cases. 1451 7.5. Stringprep Changes and Compatibility 1453 The Nameprep [RFC3491] specification, a key part of IDNA2003, is a 1454 profile of Stringprep [RFC3454]. While Nameprep is a Stringprep 1455 profile specific to IDNA, Stringprep is used by a number of other 1456 protocols. Were Stringprep to be modified by IDNA2008, those changes 1457 to improve the handling of IDNs could cause problems for non-DNS 1458 uses, most notably if they affected identification and authentication 1459 protocols. Several elements of IDNA2008 give interpretations to 1460 strings prohibited under IDNA2003 or prohibit strings that IDNA2003 1461 permitted. Those elements include the proposed new inclusion tables 1462 [IDNA2008-Tables], the reduction in the number of characters 1463 permitted as input for registration or lookup (Section 3), and even 1464 the proposed changes in handling of right to left strings 1465 [IDNA2008-Bidi]. IDNA2008 does not use Nameprep or Stringprep at 1466 all, so there are no side-effect changes to other protocols. 1468 It is particularly important to keep IDNA processing separate from 1469 processing for various security protocols because some of the 1470 constraints that are necessary for smooth and comprehensible use of 1471 IDNs may be unwanted or undesirable in other contexts. For example, 1472 the criteria for good passwords or passphrases are very different 1473 from those for desirable IDNs: passwords should be hard to guess, 1474 while domain names should normally be easily memorable. Similarly, 1475 internationalized SCSI identifiers and other protocol components are 1476 likely to have different requirements than IDNs. 1478 7.6. The Symbol Question 1480 One of the major differences between this specification and the 1481 original version of IDNA is that the original version permitted non- 1482 letter symbols of various sorts, including punctuation and line- 1483 drawing symbols, in the protocol. They were always discouraged in 1484 practice. In particular, both the "IESG Statement" about IDNA and 1485 all versions of the ICANN Guidelines specify that only language 1486 characters be used in labels. This specification disallows symbols 1487 entirely. There are several reasons for this, which include: 1489 1. As discussed elsewhere, the original IDNA specification assumed 1490 that as many Unicode characters as possible should be permitted, 1491 directly or via mapping to other characters, in IDNs. This 1492 specification operates on an inclusion model, extrapolating from 1493 the original "hostname" rules (LDH, see [IDNA2008-Defs]) -- which 1494 have served the Internet very well -- to a Unicode base rather 1495 than an ASCII base. 1497 2. Symbol names are more problematic than letters because there may 1498 be no general agreement on whether a particular glyph matches a 1499 symbol; there are no uniform conventions for naming; variations 1500 such as outline, solid, and shaded forms may or may not exist; 1501 and so on. As just one example, consider a "heart" symbol as it 1502 might appear in a logo that might be read as "I love...". While 1503 the user might read such a logo as "I love..." or "I heart...", 1504 considerable knowledge of the coding distinctions made in Unicode 1505 is needed to know that there more than one "heart" character 1506 (e.g., U+2665, U+2661, and U+2765) and how to describe it. These 1507 issues are of particular importance if strings are expected to be 1508 understood or transcribed by the listener after being read out 1509 loud. 1511 3. Design of a screen reader used by blind Internet users who must 1512 listen to renderings of IDN domain names and possibly reproduce 1513 them on the keyboard becomes considerably more complicated when 1514 the names of characters are not obvious and intuitive to anyone 1515 familiar with the language in question. 1517 4. As a simplified example of this, assume one wanted to use a 1518 "heart" or "star" symbol in a label. This is problematic because 1519 those names are ambiguous in the Unicode system of naming (the 1520 actual Unicode names require far more qualification). A user or 1521 would-be registrant has no way to know -- absent careful study of 1522 the code tables -- whether it is ambiguous (e.g., where there are 1523 multiple "heart" characters) or not. Conversely, the user seeing 1524 the hypothetical label doesn't know whether to read it -- try to 1525 transmit it to a colleague by voice -- as "heart", as "love", as 1526 "black heart", or as any of the other examples below. 1528 5. The actual situation is even worse than this. There is no 1529 possible way for a normal, casual, user to tell the difference 1530 between the hearts of U+2665 and U+2765 and the stars of U+2606 1531 and U+2729 or the without somehow knowing to look for a 1532 distinction. We have a white heart (U+2661) and few black 1533 hearts. Consequently, describing a label as containing a heart 1534 is hopelessly ambiguous: we can only know that it contains one of 1535 several characters that look like hearts or have "heart" in their 1536 names. In cities where "Square" is a popular part of a location 1537 name, one might well want to use a square symbol in a label as 1538 well and there are far more squares of various flavors in Unicode 1539 than there are hearts or stars. 1541 The consequence of these ambiguities is that symbols are a very poor 1542 basis for reliable communication. Consistent with this conclusion, 1543 the Unicode standard recommends that strings used in identifiers not 1544 contain symbols or punctuation [Unicode-UAX31]. Of course, these 1545 difficulties with symbols do not arise with actual pictographic 1546 languages and scripts which would be treated like any other language 1547 characters; the two should not be confused. 1549 7.7. Migration Between Unicode Versions: Unassigned Code Points 1551 In IDNA2003, labels containing unassigned code points are looked up 1552 on the assumption that, if they appear in labels and can be mapped 1553 and then resolved, the relevant standards must have changed and the 1554 registry has properly allocated only assigned values. 1556 In the protocol described in these documents, strings containing 1557 unassigned code points must not be either looked up or registered. 1558 In summary, the status of an unassigned character with regard to the 1559 DISALLOWED, PROTOCOL-VALID, and CONTEXTUAL RULE REQUIRED categories 1560 cannot be evaluated until a character is actually assigned and known. 1561 There are several reasons for this, with the most important ones 1562 being: 1564 o Tests involving the context of characters (e.g., some characters 1565 being permitted only adjacent to others of specific types) and 1566 integrity tests on complete labels are needed. Unassigned code 1567 points cannot be permitted because one cannot determine whether 1568 particular code points will require contextual rules (and what 1569 those rules should be) before characters are assigned to them and 1570 the properties of those characters fully understood. 1572 o It cannot be known in advance, and with sufficient reliability, 1573 whether a newly-assigned code point will be associated with a 1574 character that would be disallowed by the rules in 1575 [IDNA2008-Tables] (such as a compatibility character). In 1576 IDNA2003, since there is no direct dependency on NFKC (many of the 1577 entries in Stringprep's tables are based on NFKC, but IDNA2003 1578 depends only on Stringprep), allocation of a compatibility 1579 character might produce some odd situations, but it would not be a 1580 problem. In IDNA2008, where compatibility characters are 1581 DISALLOWED unless character-specific exceptions are made, 1582 permitting strings containing unassigned characters to be looked 1583 up would violate the principle that characters in DISALLOWED are 1584 not looked up. 1586 o The Unicode Standard specifies that an unassigned code point 1587 normalizes (and, where relevant, case folds) to itself. If the 1588 code point is later assigned to a character, and particularly if 1589 the newly-assigned code point has a combining class that 1590 determines its placement relative to other combining characters, 1591 it could normalize to some other code point or sequence. 1593 It is possible to argue that the issues above are not important and 1594 that, as a consequence, it is better to retain the principle of 1595 looking up labels even if they contain unassigned characters because 1596 all of the important scripts and characters have been coded as of 1597 Unicode 5.1 and hence unassigned code points will be assigned only to 1598 obscure characters or archaic scripts. Unfortunately, that does not 1599 appear to be a safe assumption for at least two reasons. First, much 1600 the same claim of completeness has been made for earlier versions of 1601 Unicode. The reality is that a script that is obscure to much of the 1602 world may still be very important to those who use it. Cultural and 1603 linguistic preservation principles make it inappropriate to declare 1604 the script of no importance in IDNs. Second, we already have 1605 counterexamples in, e.g., the relationships associated with new Han 1606 characters being added (whether in the BMP or in Unicode Plane 2). 1608 Independent of the technical transition issues identified above, it 1609 can be observed that any addition of characters to an existing script 1610 to make it easier to use or to better accommodate particular 1611 languages may lead to transition issues. Such changes may change the 1612 preferred form for writing a particular string, changes that may be 1613 reflected, e.g., in keyboard transition modules that would 1614 necessarily be different from those for earlier versions of Unicode 1615 where the newer characters may not exist. This creates an inherent 1616 transition problem because attempts to access labels may use either 1617 the old or the new conventions, requiring registry action whether the 1618 older conventions were used in labels or not. The need to consider 1619 transition mechanisms is inherent to evolution of Unicode to better 1620 accommodate writing systems and is independent of how IDNs are 1621 represented in the DNS or how transitions among versions of those 1622 mechanisms occur. The requirement for transitions of this type is 1623 illustrated by the addition of Malayalam Chillu in Unicode 5.1.0. 1625 7.8. Other Compatibility Issues 1627 The 2003 IDNA model includes several odd artifacts of the context in 1628 which it was developed. Many, if not all, of these are potential 1629 avenues for exploits, especially if the registration process permits 1630 "source" names (names that have not been processed through IDNA and 1631 Nameprep) to be registered. As one example, since the character 1632 Eszett, used in German, is mapped by IDNA2003 into the sequence "ss" 1633 rather than being retained as itself or prohibited, a string 1634 containing that character but that is otherwise in ASCII is not 1635 really an IDN (in the U-label sense defined above) at all. After 1636 Nameprep maps the Eszett out, the result is an ASCII string and so 1637 does not get an xn-- prefix, but the string that can be displayed to 1638 a user appears to be an IDN. The newer version of the protocol 1639 eliminates this artifact. A character is either permitted as itself 1640 or it is prohibited; special cases that make sense only in a 1641 particular linguistic or cultural context can be dealt with as 1642 localization matters where appropriate. 1644 8. Name Server Considerations 1646 8.1. Processing Non-ASCII Strings 1648 Existing DNS servers do not know the IDNA rules for handling non- 1649 ASCII forms of IDNs, and therefore need to be shielded from them. 1650 All existing channels through which names can enter a DNS server 1651 database (for example, master files (as described in RFC 1034) and 1652 DNS update messages [RFC2136]) are IDN-unaware because they predate 1653 IDNA. Other sections of this document provide the needed shielding 1654 by ensuring that internationalized domain names entering DNS server 1655 databases through such channels have already been converted to their 1656 equivalent ASCII A-label forms. 1658 Because of the distinction made between the algorithms for 1659 Registration and Lookup in [IDNA2008-Protocol] (a domain name 1660 containing only ASCII codepoints can not be converted to an A-label), 1661 there can not be more than one A-label form for any given U-label. 1663 As specified in RFC 2181 [RFC2181], the DNS protocol explicitly 1664 allows domain labels to contain octets beyond the ASCII range 1665 (0000..007F), and this document does not change that. However, 1666 although the interpretation of octets 0080..00FF is well-defined in 1667 the DNS, many application protocols support only ASCII labels and 1668 there is no defined interpretation of these non-ASCII octets as 1669 characters and, in particular, no interpretation of case-independent 1670 matching for them (see, e.g., [RFC4343]). If labels containing these 1671 octets are returned to applications, unpredictable behavior could 1672 result. The A-label form, which cannot contain those characters, is 1673 the only standard representation for internationalized labels in the 1674 DNS protocol. 1676 8.2. DNSSEC Authentication of IDN Domain Names 1678 DNS Security (DNSSEC) [RFC2535] is a method for supplying 1679 cryptographic verification information along with DNS messages. 1680 Public Key Cryptography is used in conjunction with digital 1681 signatures to provide a means for a requester of domain information 1682 to authenticate the source of the data. This ensures that it can be 1683 traced back to a trusted source, either directly or via a chain of 1684 trust linking the source of the information to the top of the DNS 1685 hierarchy. 1687 IDNA specifies that all internationalized domain names served by DNS 1688 servers that cannot be represented directly in ASCII MUST use the 1689 A-label form. Conversion to A-labels MUST be performed prior to a 1690 zone being signed by the private key for that zone. Because of this 1691 ordering, it is important to recognize that DNSSEC authenticates a 1692 domain name containing A-labels or conventional LDH-labels, not 1693 U-labels. In the presence of DNSSEC, no form of a zone file or query 1694 response that contains a U-label may be signed or the signature 1695 validated. 1697 One consequence of this for sites deploying IDNA in the presence of 1698 DNSSEC is that any special purpose proxies or forwarders used to 1699 transform user input into IDNs must be earlier in the lookup flow 1700 than DNSSEC authenticating nameservers for DNSSEC to work. 1702 8.3. Root and other DNS Server Considerations 1704 IDNs in A-label form will generally be somewhat longer than current 1705 domain names, so the bandwidth needed by the root servers is likely 1706 to go up by a small amount. Also, queries and responses for IDNs 1707 will probably be somewhat longer than typical queries historically, 1708 so EDNS0 [RFC2671] support may be more important (otherwise, queries 1709 and responses may be forced to go to TCP instead of UDP). 1711 9. Internationalization Considerations 1713 DNS labels and fully-qualified domain names provide mnemonics that 1714 assist in identifying and referring to resources on the Internet. 1715 IDNs expand the range of those mnemonics to include those based on 1716 languages and character sets other than Western European and Roman- 1717 derived ones. But domain "names" are not, in general, words in any 1718 language. The recommendations of the IETF policy on character sets 1719 and languages, (BCP 18 [RFC2277]) are applicable to situations in 1720 which language identification is used to provide language-specific 1721 contexts. The DNS is, by contrast, global and international and 1722 ultimately has nothing to do with languages. Adding languages (or 1723 similar context) to IDNs generally, or to DNS matching in particular, 1724 would imply context dependent matching in DNS, which would be a very 1725 significant change to the DNS protocol itself. It would also imply 1726 that users would need to identify the language associated with a 1727 particular label in order to look that label up. That knowledge is 1728 generally not available because many labels are not words in any 1729 language and some may be words in more than one. 1731 10. IANA Considerations 1733 This section gives an overview of IANA registries required for IDNA. 1734 The actual definitions of, and specifications for, the first two, 1735 which must be newly-created for IDNA2008, appear in 1736 [IDNA2008-Tables]. This document describes the registries but does 1737 not specify any IANA actions. 1739 10.1. IDNA Character Registry 1741 The distinction among the major categories "UNASSIGNED", 1742 "DISALLOWED", "PROTOCOL-VALID", and "CONTEXTUAL RULE REQUIRED" is 1743 made by special categories and rules that are integral elements of 1744 [IDNA2008-Tables]. While not normative, an IANA registry of 1745 characters and scripts and their categories, updated for each new 1746 version of Unicode and the characters it contains, will be convenient 1747 for programming and validation purposes. The details of this 1748 registry are specified in [IDNA2008-Tables]. 1750 10.2. IDNA Context Registry 1752 IANA will create and maintain a list of approved contextual rules for 1753 characters that are defined in the IDNA Character Registry list as 1754 requiring a Contextual Rule (i.e., the types of rule described in 1755 Section 3.1.2). The details for those rules appear in 1756 [IDNA2008-Tables]. 1758 10.3. IANA Repository of IDN Practices of TLDs 1760 This registry, historically described as the "IANA Language Character 1761 Set Registry" or "IANA Script Registry" (both somewhat misleading 1762 terms) is maintained by IANA at the request of ICANN. It is used to 1763 provide a central documentation repository of the IDN policies used 1764 by top level domain (TLD) registries who volunteer to contribute to 1765 it and is used in conjunction with ICANN Guidelines for IDN use. 1767 It is not an IETF-managed registry and, while the protocol changes 1768 specified here may call for some revisions to the tables, these 1769 specifications have no direct effect on that registry and no IANA 1770 action is required as a result. 1772 11. Security Considerations 1774 11.1. General Security Issues with IDNA 1776 This document is purely explanatory and informational and 1777 consequently introduces no new security issues. It would, of course, 1778 be a poor idea for someone to try to implement from it; such an 1779 attempt would almost certainly lead to interoperability problems and 1780 might lead to security ones. A discussion of security issues with 1781 IDNA, including some relevant history, appears in [IDNA2008-Defs]. 1783 12. Acknowledgments 1785 The editor and contributors would like to express their thanks to 1786 those who contributed significant early (pre-WG) review comments, 1787 sometimes accompanied by text, Paul Hoffman, Simon Josefsson, and Sam 1788 Weiler. In addition, some specific ideas were incorporated from 1789 suggestions, text, or comments about sections that were unclear 1790 supplied by Vint Cerf, Frank Ellerman, Michael Everson, Asmus 1791 Freytag, Erik van der Poel, Michel Suignard, and Ken Whistler. 1792 Thanks are also due to Vint Cerf, Lisa Dusseault, Debbie Garside, and 1793 Jefsey Morfin for conversations that led to considerable improvements 1794 in the content of this document. 1796 A meeting was held on 30 January 2008 to attempt to reconcile 1797 differences in perspective and terminology about this set of 1798 specifications between the design team and members of the Unicode 1799 Technical Consortium. The discussions at and subsequent to that 1800 meeting were very helpful in focusing the issues and in refining the 1801 specifications. The active participants at that meeting were (in 1802 alphabetic order as usual) Harald Alvestrand, Vint Cerf, Tina Dam, 1803 Mark Davis, Lisa Dusseault, Patrik Faltstrom (by telephone), Cary 1804 Karp, John Klensin, Warren Kumari, Lisa Moore, Erik van der Poel, 1805 Michel Suignard, and Ken Whistler. We express our thanks to Google 1806 for support of that meeting and to the participants for their 1807 contributions. 1809 Useful comments and text on the WG versions of the draft were 1810 received from many participants in the IETF "IDNABIS" WG and a number 1811 of document changes resulted from mailing list discussions made by 1812 that group. Marcos Sanz provided specific analysis and suggestions 1813 that were exceptionally helpful in refining the text, as did Vint 1814 Cerf, Martin Duerst, Andrew Sullivan, and Ken Whistler. Lisa 1815 Dusseault provided extensive editorial suggestions during the spring 1816 of 2009, most of which were incorporated. 1818 13. Contributors 1820 While the listed editor held the pen, the core of this document and 1821 the initial WG version represents the joint work and conclusions of 1822 an ad hoc design team consisting of the editor and, in alphabetic 1823 order, Harald Alvestrand, Tina Dam, Patrik Faltstrom, and Cary Karp. 1824 Considerable material describing mapping principles has been 1825 incorporated from a draft of [IDNA2008-Mapping] by Pete Resnick and 1826 Paul Hoffman. In addition, there were many specific contributions 1827 and helpful comments from those listed in the Acknowledgments section 1828 and others who have contributed to the development and use of the 1829 IDNA protocols. 1831 14. References 1833 14.1. Normative References 1835 [ASCII] American National Standards Institute (formerly United 1836 States of America Standards Institute), "USA Code for 1837 Information Interchange", ANSI X3.4-1968, 1968. 1839 ANSI X3.4-1968 has been replaced by newer versions with 1840 slight modifications, but the 1968 version remains 1841 definitive for the Internet. 1843 [IDNA2008-Bidi] 1844 Alvestrand, H. and C. Karp, "An updated IDNA criterion for 1845 right to left scripts", August 2009, . 1848 [IDNA2008-Defs] 1849 Klensin, J., "Internationalized Domain Names for 1850 Applications (IDNA): Definitions and Document Framework", 1851 August 2009, . 1854 [IDNA2008-Protocol] 1855 Klensin, J., "Internationalized Domain Names in 1856 Applications (IDNA): Protocol", August 2009, . 1859 [IDNA2008-Tables] 1860 Faltstrom, P., "The Unicode Code Points and IDNA", 1861 August 2009, . 1864 A version of this document is available in HTML format at 1865 http://stupid.domain.name/idnabis/ 1866 draft-ietf-idnabis-tables-06.html 1868 [RFC3490] Faltstrom, P., Hoffman, P., and A. Costello, 1869 "Internationalizing Domain Names in Applications (IDNA)", 1870 RFC 3490, March 2003. 1872 [RFC3492] Costello, A., "Punycode: A Bootstring encoding of Unicode 1873 for Internationalized Domain Names in Applications 1874 (IDNA)", RFC 3492, March 2003. 1876 [Unicode-UAX15] 1877 The Unicode Consortium, "Unicode Standard Annex #15: 1878 Unicode Normalization Forms", March 2008, 1879 . 1881 [Unicode51] 1882 The Unicode Consortium, "The Unicode Standard, Version 1883 5.1.0", 2008. 1885 defined by: The Unicode Standard, Version 5.0, Boston, MA, 1886 Addison-Wesley, 2007, ISBN 0-321-48091-0, as amended by 1887 Unicode 5.1.0 1888 (http://www.unicode.org/versions/Unicode5.1.0/). 1890 14.2. Informative References 1892 [BIG5] Institute for Information Industry of Taiwan, "Computer 1893 Chinese Glyph and Character Code Mapping Table, Technical 1894 Report C-26", 1984. 1896 There are several forms and variations and a closely- 1897 related standard, CNS 11643. See the discussion in 1898 Chapter 3 of Lunde, K., CJKV Information Processing, 1899 O'Reilly & Associates, 1999 1901 [GB18030] "Chinese National Standard GB 18030-2000: Information 1902 Technology -- Chinese ideograms coded character set for 1903 information interchange -- Extension for the basic set.", 1904 2000. 1906 [IDNA2008-Mapping] 1907 Resnick, P., "Mapping Characters in IDNA", August 2009, . 1911 [RFC0810] Feinler, E., Harrenstien, K., Su, Z., and V. White, "DoD 1912 Internet host table specification", RFC 810, March 1982. 1914 [RFC0952] Harrenstien, K., Stahl, M., and E. Feinler, "DoD Internet 1915 host table specification", RFC 952, October 1985. 1917 [RFC1034] Mockapetris, P., "Domain names - concepts and facilities", 1918 STD 13, RFC 1034, November 1987. 1920 [RFC1035] Mockapetris, P., "Domain names - implementation and 1921 specification", STD 13, RFC 1035, November 1987. 1923 [RFC1123] Braden, R., "Requirements for Internet Hosts - Application 1924 and Support", STD 3, RFC 1123, October 1989. 1926 [RFC2136] Vixie, P., Thomson, S., Rekhter, Y., and J. Bound, 1927 "Dynamic Updates in the Domain Name System (DNS UPDATE)", 1928 RFC 2136, April 1997. 1930 [RFC2181] Elz, R. and R. Bush, "Clarifications to the DNS 1931 Specification", RFC 2181, July 1997. 1933 [RFC2277] Alvestrand, H., "IETF Policy on Character Sets and 1934 Languages", BCP 18, RFC 2277, January 1998. 1936 [RFC2535] Eastlake, D., "Domain Name System Security Extensions", 1937 RFC 2535, March 1999. 1939 [RFC2671] Vixie, P., "Extension Mechanisms for DNS (EDNS0)", 1940 RFC 2671, August 1999. 1942 [RFC2673] Crawford, M., "Binary Labels in the Domain Name System", 1943 RFC 2673, August 1999. 1945 [RFC2782] Gulbrandsen, A., Vixie, P., and L. Esibov, "A DNS RR for 1946 specifying the location of services (DNS SRV)", RFC 2782, 1947 February 2000. 1949 [RFC3454] Hoffman, P. and M. Blanchet, "Preparation of 1950 Internationalized Strings ("stringprep")", RFC 3454, 1951 December 2002. 1953 [RFC3491] Hoffman, P. and M. Blanchet, "Nameprep: A Stringprep 1954 Profile for Internationalized Domain Names (IDN)", 1955 RFC 3491, March 2003. 1957 [RFC3743] Konishi, K., Huang, K., Qian, H., and Y. Ko, "Joint 1958 Engineering Team (JET) Guidelines for Internationalized 1959 Domain Names (IDN) Registration and Administration for 1960 Chinese, Japanese, and Korean", RFC 3743, April 2004. 1962 [RFC3987] Duerst, M. and M. Suignard, "Internationalized Resource 1963 Identifiers (IRIs)", RFC 3987, January 2005. 1965 [RFC4290] Klensin, J., "Suggested Practices for Registration of 1966 Internationalized Domain Names (IDN)", RFC 4290, 1967 December 2005. 1969 [RFC4343] Eastlake, D., "Domain Name System (DNS) Case Insensitivity 1970 Clarification", RFC 4343, January 2006. 1972 [RFC4690] Klensin, J., Faltstrom, P., Karp, C., and IAB, "Review and 1973 Recommendations for Internationalized Domain Names 1974 (IDNs)", RFC 4690, September 2006. 1976 [RFC4713] Lee, X., Mao, W., Chen, E., Hsu, N., and J. Klensin, 1977 "Registration and Administration Recommendations for 1978 Chinese Domain Names", RFC 4713, October 2006. 1980 [Unicode-Security] 1981 The Unicode Consortium, "Unicode Technical Standard #39: 1982 Unicode Security Mechanisms", August 2008, 1983 . 1985 [Unicode-UAX31] 1986 The Unicode Consortium, "Unicode Standard Annex #31: 1987 Unicode Identifier and Pattern Syntax", March 2008, 1988 . 1990 [Unicode-UTR36] 1991 The Unicode Consortium, "Unicode Technical Report #36: 1992 Unicode Security Considerations", July 2008, 1993 . 1995 Appendix A. Change Log 1997 [[ RFC Editor: Please remove this appendix. ]] 1999 A.1. Changes between Version -00 and Version -01 of 2000 draft-ietf-idnabis-rationale 2002 o Clarified the U-label definition to note that U-labels must 2003 contain at least one non-ASCII character. Also clarified the 2004 relationship among label types. 2006 o Rewrote the discussion of Labels in Registration (Section 7.1.2) 2007 and related text about IDNA-validity (in the "Defs" document as of 2008 -04 of this one) to narrow its focus and remove more general 2009 restrictions. Added a temporary note in line to explain the 2010 situation. 2012 o Changed the "IDNA uses Unicode" statement to focus on 2013 compatibility with IDNA2003 and avoid more general or 2014 controversial assertions. 2016 o Added a discussion of examples to Section 7.1 2018 o Made a number of other small editorial changes and corrections 2019 suggested by Mark Davis. 2021 o Added several more discussion anchors and notes and expanded or 2022 updated some existing ones. 2024 A.2. Version -02 2026 o Trimmed change log, removing information about pre-WG drafts. 2028 o Adjusted discussion of Contextual Rules to match the new location 2029 of the tables and some conceptual material. 2031 o Rewrote the material on preprocessing somewhat. 2033 o Moved the material about relationships with IDNA2003 to be part of 2034 a single section on transitions. 2036 o Removed several placeholders and made editorial changes in 2037 accordance with decisions made at IETF 72 in Dublin and not 2038 disputed on the mailing list. 2040 A.3. Version -03 2042 This special update to the Rationale document is intended to try to 2043 get the discussion of what is normative or not under control. While 2044 the IETF does not normally annotate individual sections of documents 2045 with whether they are normative or not, concerns that we don't know 2046 which is which, claims that some material is normative that would be 2047 problematic if so classified, etc., argue that we should at least be 2048 able to have a clear discussion on the subject. 2050 Two annotations have been applied to sections that might reasonably 2051 be considered normative. One annotation is based on the list of 2052 sections in Mark Davis's note of 29 September (http:// 2053 www.alvestrand.no/pipermail/idna-update/2008-September/002667.html). 2054 The other is based on an elaboration of John Klensin's response on 7 2055 October (http://www.alvestrand.no/pipermail/idna-update/2008-October/ 2056 002691.html). These should just be considered two suggestions to 2057 illuminate and, one hopes, advance the Working Group's discussions. 2059 Some additional editorial changes have been made, but they are 2060 basically trivial. In the editor's judgment, it is not possible to 2061 make significantly more progress with this document until the matter 2062 of document organization is settled. 2064 A.4. Version -04 2066 o Definitional and other normative material moved to new document 2067 (draft-ietf-idnabis-defs). Version -03 annotations removed. 2069 o Material on differences between IDNA2003 and IDNA2008 moved to an 2070 appendix in Protocol. 2072 o Material left over from the origins of this document as a 2073 preliminary proposal has been removed or rewritten. 2075 o Changes made to reflect consensus call results, including removing 2076 several placeholder notes for discussion. 2078 o Added more material, including discussion of historic scripts, to 2079 Section 3.2 on registration policies. 2081 o Added a new section (Section 7.2) to contain specific discussion 2082 of handling of characters that are interpreted differently in 2083 input to IDNA2003 and 2008. 2085 o Some material, including this section/appendix, rearranged. 2087 A.5. Version -05 2089 o Many small editorial changes, including changes to eliminate the 2090 last vestiges of what appeared to be 2119 language (upper-case 2091 MUST, SHOULD, or MAY) and small adjustments to terminology. 2093 A.6. Version -06 2095 o Removed Security Considerations material and pointed to Defs, 2096 where it now appears as of version 05. 2098 o Started changing uses of "IDNA2008" in running text to "in these 2099 specifications" or the equivalent. These documents are titled 2100 simply "IDNA"; once they are standardized, "the current version" 2101 may be a more appropriate reference than one containing a year. 2102 As discussed on the mailing list, we can and should discuss how to 2103 refer to these documents at an appropriate time (e.g., when we 2104 know when we will be finished) but, in the interim, it seems 2105 appropriate to simply start getting rid of the version-specific 2106 terminology where it can naturally be removed. 2108 o Additional discussion of mappings, etc., especially for case- 2109 sensitivity. 2111 o Clarified relationship to base DNS specifications. 2113 o Consolidated discussion of lookup of unassigned characters. 2115 o More editorial fine-tuning. 2117 A.7. Version -07 2119 o Revised terminology by adding terms: NR-LDH-label, Invalid-A-label 2120 (or False-A-label), R-LDH-label, valid IDNA-label in 2121 Section 1.3.2. 2123 o Moved the "name server considerations" material to this document 2124 from Protocol because it is non-normative and not part of the 2125 protocol itself. 2127 o To improve clarity, redid discussion of the reasons why looking up 2128 unassigned code points is prohibited. 2130 o Editorial and other non-substantive corrections to reflect earlier 2131 errors as well as new definitions and terminology. 2133 A.8. Version -08 2135 o Slight revision to "contextual" discussion (Section 3.1.2) and 2136 moving it to a separate subsection, rather than under "PVALID", 2137 for better parallelism with Tables. Also reflected Mark's 2138 comments about the limitations of the approach. 2140 o Added placeholder notes as reminders of where references to the 2141 other documents need Section numbers. More of these will be added 2142 as needed (feel free to identify relevant places), but the actual 2143 section numbers will not be inserted until the documents are 2144 completely stable, i.e., on their way to the RFC Editor. 2146 A.9. Version -09 2148 o Small editorial changes to clarify transition possibilities. 2150 o Small clarification to the description of DNS "exact match". 2152 o Added discussion of adding characters to an existing script to the 2153 discussion of unassigned code point transitions in Section 7.7. 2155 o Tightened up the discussion of non-ASCII string processing 2156 (Section 8.1) slightly. 2158 o Removed some placeholders and comments that have been around long 2159 enough to be considered acceptable or that no longer seem 2160 necessary for other reasons. 2162 A.10. Version -10 2164 o Extensive editorial improvements, mostly due to suggestions from 2165 Lisa Dusseault. 2167 o Changes required for the new "mapping" approach and document have, 2168 in general, not been incorporated despite several suggestions. 2169 The editor intends to wait until the mapping model is stable, or 2170 at least until -11 of this document, before trying to incorporate 2171 those suggestions. 2173 A.11. Version -11 2175 o Several placeholders for additional material or editing have been 2176 removed since no comments have been received. 2178 o Updated references. 2180 o Corrected an apparent patching error in Section 1.6 and another 2181 one in Section 4.3. 2183 o Adjusted several sections that had not properly reflected removal 2184 of the material that is now in the Definitions document and 2185 removed an unnecessary one. 2187 o New material added to Section 3.2 about registration policy issues 2188 to reflect discussions on the mailing list. 2190 o Incorporated mapping material from the former "Architectural 2191 Principles" of version -01 of the Mapping draft into Section 6 and 2192 removed most of the prior mapping material and explanations. 2194 o Eliminated the former Section 7.3 ("More Flexibility in User 2195 Agents"), moving its material into Section 4.2. The replacement 2196 section is basically a placeholder to retain the mapping issues as 2197 one of the migration topics. Note that this item and the previous 2198 one involve considerable text, so people should check things 2199 carefully. 2201 o Corrected several typographical and editorial errors that don't 2202 fall into any of the above categories. 2204 A.12. Version -12 2206 o Got rid of the term "IDNA-valid". It no longer appears in 2207 Definitions and we didn't really need the extra term. Where the 2208 concept was needed, the text now says "valid under IDNA" or 2209 equivalent. 2211 o Adjusted Acknowledgments to remove Mark Davis's name, per his 2212 request and advice from IETF Trust Counsel. 2214 o Incorporated other changes from WG Last Call. 2216 o Small typographical and editorial corrections. 2218 Author's Address 2220 John C Klensin 2221 1770 Massachusetts Ave, Ste 322 2222 Cambridge, MA 02140 2223 USA 2225 Phone: +1 617 245 1457 2226 Email: john+ietf@jck.com