idnits 2.17.00 (12 Aug 2021) /tmp/idnits14059/draft-leiba-cotton-iana-5226bis-01.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- ** The document seems to lack an IANA Considerations section. (See Section 2.2 of https://www.ietf.org/id-info/checklist for how to handle the case when there are no actions for IANA.) (A line matching the expected section header was found, but with an unexpected indentation: ' their IANA considerations. Instead, the intention is to permit' ) Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (October 2, 2012) is 3517 days in the past. Is this intentional? Checking references for intended status: Best Current Practice ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Missing Reference: 'BCP26' is mentioned on line 929, but not defined == Missing Reference: 'RFC9876' is mentioned on line 1169, but not defined -- Obsolete informational reference (is this intentional?): RFC 4005 (Obsoleted by RFC 7155) -- Obsolete informational reference (is this intentional?): RFC 4288 (Obsoleted by RFC 6838) -- Obsolete informational reference (is this intentional?): RFC 4395 (Obsoleted by RFC 7595) -- Obsolete informational reference (is this intentional?): RFC 5246 (Obsoleted by RFC 8446) -- Obsolete informational reference (is this intentional?): RFC 6195 (Obsoleted by RFC 6895) Summary: 1 error (**), 0 flaws (~~), 3 warnings (==), 6 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group M. Cotton 3 Internet-Draft Internet Assigned Numbers 4 Obsoletes: 5226 (if approved) Authority (IANA) 5 Intended status: BCP B. Leiba 6 Expires: April 5, 2013 Huawei Technologies 7 T. Narten 8 IBM Corporation 9 October 2, 2012 11 Guidelines for Writing an IANA Considerations Section in RFCs 12 draft-leiba-cotton-iana-5226bis-01 14 Abstract 16 Many protocols make use of identifiers consisting of constants and 17 other well-known values. Even after a protocol has been defined and 18 deployment has begun, new values may need to be assigned (such as for 19 a new option type in DHCP, or a new encryption or authentication 20 transform for IPsec). To ensure that such quantities have consistent 21 values and interpretations across all implementations, their 22 assignment must be administered by a central authority. For IETF 23 protocols, that role is provided by the Internet Assigned Numbers 24 Authority (IANA). 26 In order for IANA to manage a given namespace prudently, it needs 27 guidelines describing the conditions under which new values can be 28 assigned or when modifications to existing values can be made. If 29 IANA is expected to play a role in the management of a namespace, 30 IANA must be given clear and concise instructions describing that 31 role. This document discusses issues that should be considered in 32 formulating a policy for assigning values to a namespace and provides 33 guidelines for authors on the specific text that must be included in 34 documents that place demands on IANA. 36 This document obsoletes RFC 5226. 38 Status of this Memo 40 This Internet-Draft is submitted in full conformance with the 41 provisions of BCP 78 and BCP 79. 43 Internet-Drafts are working documents of the Internet Engineering 44 Task Force (IETF). Note that other groups may also distribute 45 working documents as Internet-Drafts. The list of current Internet- 46 Drafts is at http://datatracker.ietf.org/drafts/current/. 48 Internet-Drafts are draft documents valid for a maximum of six months 49 and may be updated, replaced, or obsoleted by other documents at any 50 time. It is inappropriate to use Internet-Drafts as reference 51 material or to cite them other than as "work in progress." 53 This Internet-Draft will expire on April 5, 2013. 55 Copyright Notice 57 Copyright (c) 2012 IETF Trust and the persons identified as the 58 document authors. All rights reserved. 60 This document is subject to BCP 78 and the IETF Trust's Legal 61 Provisions Relating to IETF Documents 62 (http://trustee.ietf.org/license-info) in effect on the date of 63 publication of this document. Please review these documents 64 carefully, as they describe your rights and restrictions with respect 65 to this document. Code Components extracted from this document must 66 include Simplified BSD License text as described in Section 4.e of 67 the Trust Legal Provisions and are provided without warranty as 68 described in the Simplified BSD License. 70 Table of Contents 72 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 5 73 1.1. Terminology Used In This Document . . . . . . . . . . . . 5 74 2. Why Management of a Namespace May Be Necessary . . . . . . . . 6 75 3. Designated Experts . . . . . . . . . . . . . . . . . . . . . . 7 76 3.1. The Motivation for Designated Experts . . . . . . . . . . 7 77 3.2. The Role of the Designated Expert . . . . . . . . . . . . 8 78 3.3. Designated Expert Reviews . . . . . . . . . . . . . . . . 9 79 3.4. Expert Reviews and the Document Lifecycle . . . . . . . . 10 80 4. Creating a Registry . . . . . . . . . . . . . . . . . . . . . 11 81 4.1. Well-Known IANA Policy Definitions . . . . . . . . . . . . 11 82 4.1.1. Policy: Private Use . . . . . . . . . . . . . . . . . 12 83 4.1.2. Policy: Experimental Use . . . . . . . . . . . . . . . 13 84 4.1.3. Policy: Hierarchical Allocation . . . . . . . . . . . 13 85 4.1.4. Policy: First Come First Served . . . . . . . . . . . 13 86 4.1.5. Policy: Expert Review . . . . . . . . . . . . . . . . 13 87 4.1.6. Policy: Specification Required . . . . . . . . . . . . 14 88 4.1.7. Policy: RFC Required . . . . . . . . . . . . . . . . . 14 89 4.1.8. Policy: IETF Review . . . . . . . . . . . . . . . . . 15 90 4.1.9. Policy: Standards Action . . . . . . . . . . . . . . . 15 91 4.1.10. Policy: IESG Approval . . . . . . . . . . . . . . . . 15 92 4.2. Best Practice for Selecting an Appropriate Policy . . . . 16 93 4.3. Using Multiple Policies in Combination . . . . . . . . . . 19 94 4.4. What to Put in Documents That Create a Registry . . . . . 19 95 4.5. Updating IANA Guidelines for Existing Registries . . . . . 22 96 5. Registering New Values in an Existing Registry . . . . . . . . 23 97 5.1. What to Put in Documents When Registering Values . . . . . 23 98 5.2. Updating Registrations . . . . . . . . . . . . . . . . . . 24 99 5.3. Overriding Registration Procedures . . . . . . . . . . . . 25 100 6. Documentation References in IANA Registries . . . . . . . . . 26 101 7. What to Do in "bis" Documents . . . . . . . . . . . . . . . . 26 102 8. Miscellaneous Issues . . . . . . . . . . . . . . . . . . . . . 27 103 8.1. When There Are No IANA Actions . . . . . . . . . . . . . . 27 104 8.2. Namespaces Lacking Documented Guidance . . . . . . . . . . 28 105 8.3. After-the-Fact Registrations . . . . . . . . . . . . . . . 28 106 8.4. Reclaiming Assigned Values . . . . . . . . . . . . . . . . 29 107 8.5. Contact Person vs Assignee or Owner . . . . . . . . . . . 29 108 8.6. Closing or Obsoleting a Registry . . . . . . . . . . . . . 30 109 8.7. BCP 78/79 Issues in Registries . . . . . . . . . . . . . . 30 110 9. Appeals . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 111 10. Mailing Lists . . . . . . . . . . . . . . . . . . . . . . . . 30 112 11. Security Considerations . . . . . . . . . . . . . . . . . . . 30 113 12. Changes Relative to Earlier Editions of BCP 26 . . . . . . . . 31 114 12.1. 2012: Changes in This Document Relative to RFC 5226 . . . 31 115 12.2. 2008: Changes in RFC 5226 Relative to RFC 2434 . . . . . . 32 116 13. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 33 117 13.1. Acknowledgments for This Document (2012) . . . . . . . . . 33 118 13.2. Acknowledgments from the second edition (2008) . . . . . . 33 119 13.3. Acknowledgments from the first edition (1998) . . . . . . 33 120 14. References . . . . . . . . . . . . . . . . . . . . . . . . . . 33 121 14.1. Normative References . . . . . . . . . . . . . . . . . . . 33 122 14.2. Informative References . . . . . . . . . . . . . . . . . . 33 123 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 36 125 1. Introduction 127 Many protocols make use of fields that contain constants and other 128 well-known values (such as the Protocol field in the IP header 129 [RFC0791] and MIME media types [RFC4288]). Even after a protocol has 130 been defined and deployment has begun, new values may need to be 131 assigned (such as a new option type in DHCP [RFC2132] or a new 132 encryption or authentication transform for IPsec [RFC4301]). To 133 ensure that such fields have consistent values and interpretations in 134 different implementations, their assignment must be administered by a 135 central authority. For IETF protocols, that role is provided by the 136 Internet Assigned Numbers Authority (IANA) [RFC2860]. 138 In this document, we call the set of possible values for such a field 139 a "namespace"; its actual value may be a text string, a number, or 140 another kind of value. The binding or association of a specific 141 value with a particular purpose within a namespace is called an 142 assigned number (or assigned value, or sometimes a "code point", 143 "protocol constant", or "protocol parameter"). Each assignment of a 144 value in a namespace is called a registration. 146 In order for IANA to manage a given namespace prudently, it needs 147 guidelines describing the conditions under which new values should be 148 assigned or when (and how) modifications to existing values can be 149 made. This document provides guidelines to authors on what sort of 150 text should be added to their documents in order to provide IANA 151 clear guidelines, and it reviews issues that should be considered in 152 formulating an appropriate policy for assigning numbers to name 153 spaces. 155 Not all namespaces require centralized administration. In some 156 cases, it is possible to delegate a namespace in such a way that 157 further assignments can be made independently and with no further 158 (central) coordination. In the Domain Name System, for example, IANA 159 only deals with assignments at the higher levels, while subdomains 160 are administered by the organization to which the space has been 161 delegated. As another example, Object Identifiers (OIDs) as defined 162 by the ITU are also delegated [RFC3232]; IANA manages the subtree 163 rooted at "iso.org.dod.internet" (1.3.6.1) . When a namespace is 164 delegated, the scope of IANA is limited to the parts of the namespace 165 where IANA has authority. 167 1.1. Terminology Used In This Document 169 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 170 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 171 document are to be interpreted as described in RFC 2119 [RFC2119]. 172 For this document, "the specification" as used by RFC 2119 refers to 173 the processing of protocol documents within the IETF standards 174 process. 176 2. Why Management of a Namespace May Be Necessary 178 One issue to consider in managing a namespace is its size. If the 179 space is small and limited in size, assignments must be made 180 carefully to prevent exhaustion of the space. If the space is 181 essentially unlimited, on the other hand, potential exhaustion will 182 probably not be a practical concern at all. Even when the space is 183 essentially unlimited, however, it is usually desirable to have at 184 least a minimal review prior to assignment in order to: 186 o prevent the hoarding of or unnecessary wasting of values. For 187 example, if the space consists of text strings, it may be 188 desirable to prevent entities from obtaining large sets of strings 189 that correspond to desirable names (existing company names, for 190 example). 192 o provide a sanity check that the request actually makes sense and 193 is necessary. Experience has shown that some level of minimal 194 review from a subject matter expert is useful to prevent 195 assignments in cases where the request is malformed or not 196 actually needed (for example, an existing assignment for an 197 essentially equivalent service already exists). 199 A second consideration is whether it makes sense to delegate the 200 namespace in some manner. This route should be pursued when 201 appropriate, as it lessens the burden on IANA for dealing with 202 assignments. 204 A third, and perhaps most important, consideration concerns potential 205 impact on the interoperability of unreviewed extensions. Proposed 206 protocol extensions generally benefit from community review; indeed, 207 review is often essential to avoid future interoperability problems 208 [RFC6709]. 210 When the namespace is essentially unlimited and there are no 211 potential interoperability issues, assigned numbers can safely be 212 given out to anyone without any subjective review. In such cases, 213 IANA can make assignments directly, provided that IANA is given 214 specific instructions on what types of requests it should grant, and 215 what information must be provided as part of a well-formed request 216 for an assigned number. 218 3. Designated Experts 220 3.1. The Motivation for Designated Experts 222 It should be noted that IANA does not create or define assignment 223 policy itself; rather, it carries out policies that have been defined 224 by others and published in RFCs. IANA must be given a set of 225 guidelines that allow it to make allocation decisions with minimal 226 subjectivity and without requiring any technical expertise with 227 respect to the protocols that make use of a registry. 229 In many cases, some review of prospective allocations is appropriate, 230 and the question becomes who should perform the review and what is 231 the purpose of the review. One might think that an IETF working 232 group familiar with the namespace at hand should be consulted. In 233 practice, however, working groups eventually disband, so they cannot 234 be considered a permanent evaluator. It is also possible for 235 namespaces to be created through individual submission documents, for 236 which no working group is ever formed. 238 One way to ensure community review of prospective assignments is to 239 have the requester submit a document for publication as an RFC. Such 240 an action helps ensure that the specification is publicly and 241 permanently available, and it allows some review of the specification 242 prior to publication and assignment of the requested code points. 243 This is the preferred way of ensuring review, and is particularly 244 important if any potential interoperability issues can arise. For 245 example, some assignments are not just assignments, but also involve 246 an element of protocol specification. A new option may define fields 247 that need to be parsed and acted on, which (if specified poorly) may 248 not fit cleanly with the architecture of other options or the base 249 protocols on which they are built. 251 In some cases, however, the burden of publishing an RFC in order to 252 get an assignment is excessive. However, it is generally still 253 useful (and sometimes necessary) to discuss proposed additions on a 254 mailing list dedicated to the purpose (such as the 255 media-types@iana.org for media types) or on a more general mailing 256 list (such as that of a current or former IETF working group). Such 257 a mailing list provides a way for new registrations to be publicly 258 reviewed prior to getting assigned, or gives advice to persons 259 wanting help in understanding what a proper registration should 260 contain. 262 While discussion on a mailing list can provide valuable technical 263 feedback, opinions may vary and discussions may continue for some 264 time without clear resolution. In addition, IANA cannot participate 265 in all of these mailing lists and cannot determine if or when such 266 discussions reach consensus. Therefore, IANA relies on a "designated 267 expert" for advice regarding the specific question of whether an 268 assignment should be made. The designated expert is an individual 269 who is responsible for carrying out an appropriate evaluation and 270 returning a recommendation to IANA. 272 It should be noted that a key motivation for having designated 273 experts is for the IETF to provide IANA with a subject matter expert 274 to whom the evaluation process can be delegated. IANA forwards 275 requests for an assignment to the expert for evaluation, and the 276 expert (after performing the evaluation) informs IANA as to whether 277 or not to make the assignment or registration. 279 It will often be useful to use a designated expert only some of the 280 time, as a supplement to other processes. For more discussion of 281 that topic, see Section 4.3. 283 3.2. The Role of the Designated Expert 285 The designated expert is responsible for initiating and coordinating 286 the appropriate review of an assignment request. The review may be 287 wide or narrow, depending on the situation and the judgment of the 288 designated expert. This may involve consultation with a set of 289 technology experts, discussion on a public mailing list, consultation 290 with a working group (or its mailing list if the working group has 291 disbanded), etc. Ideally, the designated expert follows specific 292 review criteria as documented with the protocol that creates or uses 293 the namespace. See the IANA Considerations sections of [RFC3748] and 294 [RFC3575] for examples that have been done for specific namespaces. 296 Designated experts are expected to be able to defend their decisions 297 to the IETF community, and the evaluation process is not intended to 298 be secretive or bestow unquestioned power on the expert. Experts are 299 expected to apply applicable documented review or vetting procedures, 300 or in the absence of documented criteria, follow generally accepted 301 norms such as those in Section 3.3. 303 Section 5.2 discusses disputes and appeals in more detail. 305 Designated experts are appointed by the IESG (normally upon 306 recommendation by the relevant Area Director). They are typically 307 named at the time a document creating or updating a namespace is 308 approved by the IESG, but as experts originally appointed may later 309 become unavailable, the IESG will appoint replacements if necessary. 311 For some registries, it has proven useful to have multiple designated 312 experts. Sometimes those experts work together in evaluating a 313 request, while in other cases additional experts serve as backups. 315 In cases of disagreement among those experts, it is the 316 responsibility of those experts to make a single clear recommendation 317 to IANA. It is not appropriate for IANA to resolve disputes among 318 experts. In extreme situations, such as deadlock, the IESG may need 319 to step in to resolve the problem. 321 In registries where a pool of experts evaluates requests, the pool 322 should have a single chair responsible for defining how requests are 323 to be assigned to and reviewed by experts. In some cases, the expert 324 pool may consist of a primary and backups, with the backups involved 325 only when the primary expert is unavailable. In other cases, IANA 326 might assign requests to individual members in sequential or 327 approximate random order. In the event that IANA finds itself having 328 received conflicting advice from its experts, it is the 329 responsibility of the pool's chair to resolve the issue and provide 330 IANA with clear instructions. 332 Since the designated experts are appointed by the IESG, they may be 333 removed by the IESG. 335 3.3. Designated Expert Reviews 337 In the years since RFC 2434 was published and has been put to use, 338 experience has led to the following observations: 340 o A designated expert must respond in a timely fashion, normally 341 within a week for simple requests to a few weeks for more complex 342 ones. Unreasonable delays can cause significant problems for 343 those needing assignments, such as when products need code points 344 to ship. This is not to say that all reviews can be completed 345 under a firm deadline, but they must be started, and the requester 346 and IANA should have some transparency into the process if an 347 answer cannot be given quickly. 349 o If a designated expert does not respond to IANA's requests within 350 a reasonable period of time, either with a response or with a 351 reasonable explanation for the delay (some requests may be 352 particularly complex), and if this is a recurring event, IANA must 353 raise the issue with the IESG. Because of the problems caused by 354 delayed evaluations and assignments, the IESG should take 355 appropriate actions to ensure that the expert understands and 356 accepts his or her responsibilities, or appoint a new expert. 358 o The designated expert is not required to personally bear the 359 burden of evaluating and deciding all requests, but acts as a 360 shepherd for the request, enlisting the help of others as 361 appropriate. In the case that a request is denied, and rejecting 362 the request is likely to be controversial, the expert should have 363 the support of other subject matter experts. That is, the expert 364 must be able to defend a decision to the community as a whole. 366 When a designated expert is used, the documentation should give clear 367 guidance to the designated expert, laying out criteria for performing 368 an evaluation and reasons for rejecting a request. In the case where 369 there are no specific documented criteria, the presumption should be 370 that a code point should be granted unless there is a compelling 371 reason to the contrary. Possible reasons to deny a request include 372 these: 374 o Scarcity of code points, where the finite remaining code points 375 should be prudently managed, or when a request for a large number 376 of code points is made, when a single code point is the norm. 378 o Documentation is not of sufficient clarity to evaluate or ensure 379 interoperability. 381 o The code point is needed for a protocol extension, but the 382 extension is not consistent with the documented (or generally 383 understood) architecture of the base protocol being extended, and 384 would be harmful to the protocol if widely deployed. It is not 385 the intent that "inconsistencies" refer to minor differences "of a 386 personal preference nature". Instead, they refer to significant 387 differences such as inconsistencies with the underlying security 388 model, implying a change to the semantics of an existing message 389 type or operation, requiring unwarranted changes in deployed 390 systems (compared with alternate ways of achieving a similar 391 result), etc. 393 o The extension would cause problems with existing deployed systems. 395 o The extension would conflict with one under active development by 396 the IETF, and having both would harm rather than foster 397 interoperability. 399 3.4. Expert Reviews and the Document Lifecycle 401 Review by the designated expert is necessarily done at a particular 402 point in time, and represents review of a particular version of the 403 document. Deciding when the review should take place is a question 404 of good judgment. And while re-reviews might be done when it's 405 acknowledged that the documentation of the registered item has 406 changed substantially, making sure that re-review happens requires 407 attention and care. 409 It is possible, through carelessness, accident, inattentiveness, or 410 even willful disregard, that changes might be made after the 411 designated expert's review and approval that would, if the document 412 were re-reviewed, cause the expert not to approve the registration. 413 It is up to the IESG, with the token held by the responsible Area 414 Director, to be alert to such situations and to recognize that such 415 changes need to be checked. 417 4. Creating a Registry 419 Creating a registry involves describing the namespaces to be created, 420 an initial set of assignments (if appropriate), and guidelines on how 421 future assignments are to be made. 423 Once a registry has been created, IANA records assignments that have 424 been made. The following labels describe the status of an individual 425 (or range) of assignments: 427 Private Use: Private use only (not assigned), as described in 428 Section 4.1.1. 430 Experimental: Available for general experimental use as described 431 in [RFC3692]. IANA does not record specific assignments for 432 any particular use. 434 Unassigned: Not currently assigned, and available for assignment 435 via documented procedures. While it's generally clear that 436 any values that are not registered are unassigned and 437 available for assignment, it is sometimes useful to 438 explicitly specify that situation. Note that this is 439 distinctly different from "Reserved". 441 Reserved: Not assigned and not available for assignment. 442 Reserved values are held for special uses, such as to extend 443 the namespace when it becomes exhausted. Note that this is 444 distinctly different from "Unassigned". 446 4.1. Well-Known IANA Policy Definitions 448 The following are some defined policies, most of which are in use 449 today. These cover a range of typical policies that have been used 450 to describe the procedure for assigning new values in a namespace. 451 It is not strictly required that documents use these terms; the 452 actual requirement is that the instructions to IANA be clear and 453 unambiguous. However, use of these terms is strongly RECOMMENDED, 454 because their meanings are widely understood. The terms are fully 455 explained in the following subsections. 457 1. Private Use 458 2. Experimental Use 459 3. Hierarchical Allocation 460 4. First Come First Served 461 5. Expert Review 462 6. Specification Required 463 7. RFC Required 464 8. IETF Review 465 9. Standards Action 466 10. IESG Approval 468 It should be noted that it often makes sense to partition a namespace 469 into multiple categories, with assignments within each category 470 handled differently. Many protocols now partition namespaces into 471 two or more parts, with one range reserved for Private or 472 Experimental Use while other ranges are reserved for globally unique 473 assignments assigned following some review process. Dividing a 474 namespace into ranges makes it possible to have different policies in 475 place for different ranges and different use cases. 477 Similarly, it will often be useful to specify multiple policies in 478 parallel, with each policy being used under different circumstances. 479 For more discussion of that topic, see Section 4.3. 481 Examples: 482 LDAP [RFC4520] 483 TLS ClientCertificateType Identifiers [RFC5246] (as detailed in 484 the subsections below) 485 Pseudowire Edge to Edge Emulation (PWE3) [RFC4446] 487 4.1.1. Policy: Private Use 489 For private or local use only, with the type and purpose defined by 490 the local site. No attempt is made to prevent multiple sites from 491 using the same value in different (and incompatible) ways. There is 492 no need for IANA to review such assignments (since IANA does not 493 record them) and assignments are not generally useful for broad 494 interoperability. It is the responsibility of the sites making use 495 of the Private Use range to ensure that no conflicts occur (within 496 the intended scope of use). 498 Examples: 499 Site-specific options in DHCP [RFC2939] 500 Fibre Channel Port Type Registry [RFC4044] 501 TLS ClientCertificateType Identifiers 224-255 [RFC5246] 503 4.1.2. Policy: Experimental Use 505 Similar to private or local use only, with the purpose being to 506 facilitate experimentation. See [RFC3692] for details. 508 Example: 509 Experimental Values in IPv4, IPv6, ICMPv4, ICMPv6, UDP, and TCP 510 Headers [RFC4727] 512 4.1.3. Policy: Hierarchical Allocation 514 Delegated managers can assign values provided they have been given 515 control over that part of the namespace. IANA controls the higher 516 levels of the namespace according to one of the other policies. 518 Examples: 519 DNS names 520 Object Identifiers 521 IP addresses 523 4.1.4. Policy: First Come First Served 525 Assignments are made to anyone on a first come, first served basis. 526 There is no substantive review of the request, other than to ensure 527 that it is well-formed and doesn't duplicate an existing assignment. 528 However, requests must include a minimal amount of clerical 529 information, such as a point of contact (including an email address) 530 and a brief description of how the value will be used. Additional 531 information specific to the type of value requested may also need to 532 be provided, as defined by the namespace. For numbers, the exact 533 value is generally assigned by IANA; with names, specific text 534 strings can usually be requested. 536 Examples: 537 SASL mechanism names [RFC4422] 538 LDAP Protocol Mechanisms and LDAP Syntax [RFC4520] 540 4.1.5. Policy: Expert Review 542 (Sometimes also called "Designated Expert" in earlier editions of 543 this document.) Review and approval by a designated expert is 544 required. The required documentation and review criteria for use by 545 the designated expert should be provided when defining the registry. 546 For example, see Sections 6 and 7.2 in [RFC3748]. 548 It is particularly important, when using a designated expert, to give 549 clear guidance to the expert, laying out criteria for performing an 550 evaluation and reasons for rejecting a request. When specifying a 551 policy that involves a designated expert, the IANA Considerations 552 SHOULD contain such guidance. It is also a good idea to include, 553 when possible, a sense of whether many registrations are expected 554 over time, or if the registry is expected to be updated infrequently 555 or in exceptional circumstances only. 557 Examples: 558 EAP Method Types [RFC3748] 559 HTTP Digest AKA algorithm versions [RFC4169] 560 URI schemes [RFC4395] 561 GEOPRIV Location Types [RFC4589] 563 4.1.6. Policy: Specification Required 565 Review and approval by a Designated Expert is required, (as in 566 Section 4.1.5) and the values and their meanings must be documented 567 in a permanent and readily available public specification, in 568 sufficient detail so that interoperability between independent 569 implementations is possible. The Designated Expert will review the 570 public specification and evaluate whether it is sufficiently clear to 571 allow interoperable implementations. The intention behind "permanent 572 and readily available" is that a document can reasonably be expected 573 to be findable and retrievable long after IANA assignment of the 574 requested value. Publication of an RFC is an ideal means of 575 achieving this requirement, but Specification Required is intended to 576 also cover the case of a document published outside of the RFC path. 577 For RFC publication, the normal RFC review process is expected to 578 provide the necessary review for interoperability, though the 579 designated expert may be a particularly well-qualified person to 580 perform such a review. 582 When specifying this policy, just use the term "Specification 583 Required". Some specifications have chosen to refer to it as "Expert 584 Review with Specification Required", and that only causes confusion. 586 Examples: 587 Diffserv-aware TE Bandwidth Constraints Model Identifiers 588 [RFC4124] 589 TLS ClientCertificateType Identifiers 64-223 [RFC5246] 590 ROHC Profile Identifiers [RFC5795] 592 4.1.7. Policy: RFC Required 594 RFC publication suffices, as an IETF submission or in any other 595 stream (currently an RFC Editor Independent Submission [RFC5742] or 596 an RFC in the IRTF or IAB Stream). Unless otherwise specified, any 597 type of RFC is sufficient (currently Standards Track, BCP, 598 Informational, Experimental, Historic). 600 4.1.8. Policy: IETF Review 602 (Formerly called "IETF Consensus" in the first edition of this 603 document.) New values are assigned only through RFCs in the IETF 604 Stream -- those that have been shepherded through the IESG as AD- 605 Sponsored or IETF working group Documents [RFC2026] [RFC5378]. The 606 intention is that the document and proposed assignment will be 607 reviewed by the IESG and appropriate IETF working groups (or experts, 608 if suitable working groups no longer exist) to ensure that the 609 proposed assignment will not negatively impact interoperability or 610 otherwise extend IETF protocols in an inappropriate or damaging 611 manner. 613 To ensure adequate community review, such documents are shepherded 614 through the IESG as AD-sponsored or working group documents with an 615 IETF Last Call. 617 Examples: 618 IPSECKEY Algorithm Types [RFC4025] 619 Accounting-Auth-Method AVP values in DIAMETER [RFC4005] 620 TLS Extension Types [RFC5246] 622 4.1.9. Policy: Standards Action 624 Values are assigned only for Standards Track RFCs approved by the 625 IESG. 627 Examples: 628 BGP message types [RFC4271] 629 Mobile Node Identifier option types [RFC4283] 630 TLS ClientCertificateType Identifiers 0-63 [RFC5246] 631 DCCP Packet Types [RFC4340] 633 4.1.10. Policy: IESG Approval 635 New assignments may be approved by the IESG. Although there is no 636 requirement that the request be documented in an RFC, the IESG has 637 discretion to request documents or other supporting materials on a 638 case-by-case basis. 640 IESG Approval is not intended to be used often or as a "common case"; 641 indeed, it has seldom been used in practice during the period RFC 642 2434 was in effect. Rather, it is intended to be available in 643 conjunction with other policies as a fall-back mechanism in the case 644 where one of the other allowable approval mechanisms cannot be 645 employed in a timely fashion or for some other compelling reason. 646 IESG Approval is not intended to circumvent the public review 647 processes implied by other policies that could have been employed for 648 a particular assignment. IESG Approval would be appropriate, 649 however, in cases where expediency is desired and there is strong 650 consensus (such as from a working group) for making the assignment. 652 The following guidelines are suggested for any evaluation under IESG 653 Approval: 655 o The IESG can (and should) reject a request if another path for 656 registration is available that is more appropriate and there is no 657 compelling reason not to use that path. 659 o Before approving a request, the community should be consulted, via 660 a "call for comments" that provides as much information as is 661 reasonably possible about the request. 663 Examples: 664 IPv4 Multicast address assignments [RFC5771] 665 IPv4 IGMP Type and Code values [RFC3228] 666 Mobile IPv6 Mobility Header Type and Option values [RFC6275] 668 4.2. Best Practice for Selecting an Appropriate Policy 670 The definitions above from "First Come First Served" to "Standards 671 Action" specify a range of policies in increasing order of 672 strictness: 674 4. First Come First Served: 675 No review, minimal documentation. 677 5. Expert Review: 678 Expert review, sufficient documentation for review. 680 6. Specification Required: 681 Expert review, significant, stable public documentation. 683 7. RFC Required: 684 Any RFC publication, IETF or a non-IETF Stream. 686 8. IETF Review: 687 RFC publication, IETF Stream only, but need not be Standards 688 Track. 690 9. Standards Action: 691 RFC publication, IETF Stream, Standards Track only. 693 In considering which of those policies to apply, it's important to 694 get the right balance of review and ease of registration. In many 695 cases, those needing to register items will not be IETF participants; 696 requests often come from other standards organizations, from 697 organizations not directly involved in standards, from ad-hoc 698 community work (from an open-source project, for example), and so on. 699 We must not make registration policies and procedures unnecessarily 700 difficult to navigate, unnecessarily costly (in terms of time and 701 other resources), nor unnecessarily subject to denial. 703 While it is sometimes necessary to restrict what gets registered (for 704 limited resources such as bits in a byte or numbers within a 705 relatively small range, or for items for which unsupported values can 706 be damaging to protocol operation), in many cases having items 707 registered is more important than putting restrictions on the 708 registration. A pattern of denial through overly strict review 709 criteria, or because of excessive cost in time and effort to get 710 through the process, discourages people from even attempting to 711 register their items. And failure to have in-use items registered 712 adversely affects the protocols in use on the Internet. 714 In particular, because policies 7 through 9 require involvement of 715 working groups, directorates, and/or communities of former working- 716 group participants to be actively involved and to support the effort, 717 requests frequently run into concerns that "it's not worth doing a 718 Standards-Track RFC for something this trivial," when, in fact, that 719 requirement was created by the working group in the first place, with 720 its selection of a Standards Action policy for the registry. Indeed, 721 publishing any RFC is costly, and a Standards Track RFC is especially 722 so, requiring a great deal of community time for review and 723 discussion, IETF-wide last call, involvement of the entire IESG as 724 well as concentrated time and review from the sponsoring AD, review 725 and action by IANA, and RFC-Editor processing. 727 Working groups and other document developers should use care in 728 selecting appropriate registration policies when their documents 729 create registries. They should select the least strict policy that 730 suits a registry's needs, and look for specific justification for 731 policies stricter than Specification Required. Examples of 732 situations that might merit RFC Required, IETF Review, or Standards 733 Action include the following. 735 o Registries of limited resources, such as bits in a byte (or in two 736 bytes, or four), or numbers in a limited range. In these cases, 737 allowing registrations that haven't been carefully reviewed and 738 agreed by community consensus could too quickly deplete the 739 allowable values. 741 o Registries for which thorough community review is necessary to 742 avoid extending or modifying the protocol in ways that could be 743 damaging. One example is in defining new command codes, as 744 opposed to options that use existing command codes: the former 745 might require a strict policy, where a more relaxed policy could 746 be adequate for the latter. Another example is in defining things 747 that change the semantics of existing operations. 749 There will be other cases, as well, of course; much assessment and 750 judgment is needed. And it will sometimes be the case that using 751 multiple policies in combination is appropriate (see Section 4.3). 752 It's not the intent here to put limits on the applicability of 753 particular registration policies, but to recommend laxity, rather 754 than strictness, in general, and to encourage document developers to 755 think carefully about each registry before deciding on policies. 757 The description in Section 4.1.10 of "IESG Approval" suggests that 758 the IESG "can (and should) reject a request if another path for 759 registration is available that is more appropriate and there is no 760 compelling reason not to use that path." The IESG should give 761 similar consideration to any registration policy more stringent than 762 Specification Required, asking for justification and ensuring that 763 more relaxed policies have been considered, and the strict policy is 764 the right one. This is a situation that will -- and should -- 765 involve a substantive discussion between the IESG and the working 766 group, chairs, document editors, and/or document shepherd. The 767 important point, again, is not to relax the registration policy just 768 to get the document through quickly, but to carefully choose the 769 right policy for each registry. 771 Accordingly, document developers need to anticipate this and document 772 their considerations for selecting the specified policy. Ideally, 773 they should include that in the document. At the least, it should be 774 included in the shepherd writeup for the document, and in any case 775 the document shepherd should ensure that the selected policies have 776 been justified before sending the document to the IESG. 778 When specifications are revised, registration policies should be 779 reviewed in light of experience since the policies were set. It is 780 also possible to produce a small document at any time, which 781 "updates" the original specification and changes registration 782 policies. In either case, a policy can be relaxed or made more 783 strict, as appropriate to the actual situation. 785 Once again, it cannot be stressed enough that this must not be a 786 mechanical process, but one to which the document developers apply 787 thought, consideration, assessment, and judgment in choosing the 788 right policy for each registry. 790 The recommendations in this section apply whether the well-defined 791 policy names defined herein are used, or whether the document 792 contains other policy definitions. The point, again, is not to limit 793 registration policies, but to ensure that the policies selected are 794 appropriate, and that proper consideration has been given to the 795 level of strictness required by them. 797 4.3. Using Multiple Policies in Combination 799 It is often desirable to allow registrations through the normal IETF 800 process, and to also provide a mechanism for registration outside the 801 process. Thus, a particular registry might want to use a policy such 802 as "RFC Required" or "IETF Review" sometimes, with a designated 803 expert checking a "Specification Required" policy at other times. 804 Such combinations are frequently appropriate, and are encouraged. 806 The alternative to using a combination requires either that all 807 requests come through RFCs or that requests in RFCs go through review 808 by the designated expert, despite the review and consensus that RFCs 809 represent. 811 For example, if it is felt that IETF consensus will provide good 812 review for a particular registry, but we expect frequent 813 registrations from other SDOs and we do not want those other 814 organizations always to be required to go through the IETF RFC 815 process, we might put the following in the IANA Considerations 816 section when we create the registry: 818 IANA is asked to create the registry "Fruit Access Flags" as a 819 sub-registry of "Fruit Parameters". New registrations will be 820 permitted through either the IETF Review policy or the 821 Specification Required policy [BCP26]. 823 Such combinations will commonly use one of {Standards Action, IETF 824 Review, RFC Required} in combination with one of {Specification 825 Required, Expert Review}. 827 4.4. What to Put in Documents That Create a Registry 829 The previous sections presented some issues that should be considered 830 in formulating a policy for assigning values in namespaces. It is 831 the working group and/or document author's job to formulate an 832 appropriate policy and specify it in the appropriate document. In 833 almost all cases, having an explicit "IANA Considerations" section is 834 appropriate. The following and later sections define what is needed 835 for the different types of IANA actions. 837 Documents that create a new namespace (or modify the definition of an 838 existing space) and that expect IANA to play a role in maintaining 839 that space (serving as a repository for registered values) MUST 840 provide clear instructions on details of the namespace. In 841 particular, instructions MUST include: 843 1. The name of the registry (or sub-registry) being created and/or 844 maintained. 845 The name will appear on the IANA web page and will be referred to 846 in future documents that need to allocate a value from the new 847 space. The full name (and abbreviation, if appropriate) should 848 be provided. It is highly desirable that the chosen name not be 849 easily confusable with the name of another registry. When 850 creating a sub-registry, the registry that it is a part of must 851 be clearly identified using its exact name (look it up, to be 852 sure). Providing a URL to precisely identify the registry is 853 helpful. Such URLs will be removed from the RFC prior to final 854 publication, but help to ensure that IANA will understand exactly 855 what is being requested. For example, a document could contain 856 something like this: 858 [TO BE REMOVED: This registration should be made in the Foobar 859 Operational Parameters registry, located at 860 http://www.iana.org/assignments/foobar-registry] 862 2. What information must be provided as part of a request in order 863 to assign a new value. This information may include the need to 864 document relevant security considerations, if any. 866 3. The review process that will apply to all future requests for a 867 value from the namespace. 869 Note: When a designated expert is used, documents MUST NOT name 870 the designated expert in the document itself; instead, any 871 suggested names should be relayed to the appropriate Area 872 Director at the time the document is sent to the IESG for 873 approval. This is usually done in the document shepherd writeup. 875 If the request should also be reviewed on a specific public 876 mailing list (such as the media-types@iana.org for media types), 877 that mailing address should be specified. Note, however, that 878 when mailing lists are specified, the requirement for a 879 designated expert MUST also be specified (see Section 3). 881 If IANA is expected to make assignments without requiring an 882 outside review, sufficient guidance MUST be provided so that the 883 requests can be evaluated with minimal subjectivity. 885 4. The size, format, and syntax of registry entries. When creating 886 a new name/number space, authors must describe any technical 887 requirements on registry (and sub-registry) values (valid ranges 888 for integers, length limitations on strings, etc.) as well as the 889 exact format in which registry values should be displayed. For 890 number assignments, one should specify whether values are to be 891 recorded in decimal, hexadecimal, or some other format. For 892 strings, the encoding format should be specified (ASCII, UTF8, 893 etc.). Authors should also clearly specify what fields to record 894 in the registry. 896 5. Initial assignments and reservations. Clear instructions should 897 be provided to identify any initial assignments or registrations. 898 In addition, any ranges that are to be reserved for "Private 899 Use", "Reserved", "Unassigned", etc. should be clearly indicated. 901 When specifying the process for making future assignments, it is 902 quite acceptable to pick one (or more) of the example policies listed 903 in Section 4.1 and refer to it by name. Indeed, this is the 904 preferred mechanism in those cases where the sample policies provide 905 the desired level of review. It is also acceptable to cite one of 906 the above policies and include additional guidelines for what kind of 907 considerations should be taken into account by the review process. 908 For example, RADIUS [RFC3575] specifies the use of a Designated 909 Expert, but includes specific additional criteria the Designated 910 Expert should follow. 912 For example, a document could say something like this: 914 --------------------------------------------------------------- 915 This document defines a new DHCP option, entitled "FooBar" (see 916 Section y), assigned a value of TBD1 from the DHCP Option space 917 [to be removed upon publication: 918 http://www.iana.org/assignments/bootp-dhcp-parameters] 919 [RFC2132] [RFC2939]: 920 Data 921 Tag Name Length Meaning 922 ---- ---- ------ ------- 923 TBD1 FooBar N FooBar server 925 The FooBar option also defines an 8-bit FooType field, for which 926 IANA is to create and maintain a new sub-registry entitled 927 "FooType values" under the FooBar option. Initial values for the 928 DHCP FooBar FooType registry are given below; future assignments 929 are to be made through Expert Review [BCP26]. 930 Assignments consist of a DHCP FooBar FooType name and its 931 associated value. 933 Value DHCP FooBar FooType Name Definition 934 ---- ------------------------ ---------- 935 0 Reserved 936 1 Frobnitz See Section y.1 937 2 NitzFrob See Section y.2 938 3-254 Unassigned 939 255 Reserved 940 --------------------------------------------------------------- 942 For examples of documents that provide detailed guidance to IANA on 943 the issue of assigning numbers, consult [RFC6195], [RFC3575], 944 [RFC3968], and [RFC4520]. 946 4.5. Updating IANA Guidelines for Existing Registries 948 Updating the registration process for an already existing (previously 949 created) namespace (whether created explicitly or implicitly) follows 950 a process similar to that used when creating a new namespace. That 951 is, a document is produced that makes reference to the existing 952 namespace and then provides detailed guidelines for handling 953 assignments in each individual namespace. Such documents are 954 normally processed as Best Current Practices (BCPs) [RFC2026]. 956 Example documents that updated the guidelines for managing (then) 957 pre-existing registries include: [RFC6195], [RFC3228], and [RFC3575]. 959 5. Registering New Values in an Existing Registry 961 5.1. What to Put in Documents When Registering Values 963 Often, documents request an assignment from an already existing 964 namespace (one created by a previously published document). In such 965 cases: 967 o Documents should clearly identify the namespace in which each 968 value is to be registered. If the registration goes into a sub- 969 registry, the author should clearly describe where the assignment 970 or registration should go. It is helpful to use the *exact* 971 namespace name as listed on the IANA web page (please look it up, 972 and don't guess), and cite the RFC where the namespace is defined. 974 Note 1: There is no need to mention what the assignment policy for 975 new assignments is, as that should be clear from the references. 977 Note 2: When referring to an existing registry, providing a URL to 978 precisely identify the registry is helpful. Such URLs, however, 979 should usually be removed from the RFC prior to final publication, 980 since IANA URLs are not guaranteed to be stable in the future. In 981 cases where it is important to include a URL in the document, IANA 982 should concur on its inclusion. 984 For example, a document could contain something like this: 986 [TO BE REMOVED: This registration should be made in the Foobar 987 Operational Parameters registry, located at 988 http://www.iana.org/assignments/foobar-registry] 990 o Each value requested should be given a unique reference. When the 991 value is numeric, use the notation: TBD1, TBD2, etc. Throughout 992 the document where an actual IANA-assigned value should be filled 993 in, use the "TBDx" notation. This helps ensure that the final RFC 994 has the correct assigned values inserted in all of the relevant 995 places where the value is expected to appear in the final 996 document. For values that are text strings, a specific name can 997 be suggested. IANA will normally assign the name, unless it 998 conflicts with a name already in use. 1000 o Normally, the values to be used are chosen by IANA and documents 1001 should specify values of "TBD". However, in some cases, a value 1002 may have been used for testing or in early implementations. In 1003 such cases, it is acceptable to include text suggesting what 1004 specific value should be used (together with the reason for the 1005 choice). For example, one might include the text "the value XXX 1006 is suggested as it is used in implementations". However, it 1007 should be noted that suggested values are just that; IANA will 1008 attempt to assign them, but may find that impossible, if the 1009 proposed number has already been assigned for some other use. For 1010 some registries, IANA has a long-standing policy prohibiting 1011 assignment of names or codes on a vanity or organization-name 1012 basis. For example, codes are always assigned sequentially unless 1013 there is a strong reason for making an exception. Nothing in this 1014 document is intended to change those policies or prevent their 1015 future application. 1017 o The IANA Considerations section should summarize all of the IANA 1018 actions, with pointers to the relevant sections elsewhere in the 1019 document as appropriate. When multiple values are requested, it 1020 is generally helpful to include a summary table. It is also 1021 helpful for this table to be in the same format as it appears or 1022 will appear on the IANA web site. For example: 1024 Value Description Reference 1025 -------- ------------------- --------- 1026 TBD1 Foobar [[this RFC]] 1028 Note: In cases where authors feel that including the full table is 1029 too verbose or repetitive, authors should still include the table 1030 in the draft, but may include a note asking that the table be 1031 removed prior to publication of the final RFC. 1033 As an example, the following text could be used to request assignment 1034 of a DHCPv6 option number: 1036 IANA has assigned an option code value of TBD1 to the DNS 1037 Recursive Name Server option and an option code value of TBD2 to 1038 the Domain Search List option from the DHCP option code space 1039 defined in Section 24.3 of RFC 3315. 1041 5.2. Updating Registrations 1043 Registrations are a request to assign a new value, including the 1044 related information needed to evaluate and document the request. 1045 Even after a number has been assigned, some types of registrations 1046 contain additional information that may need to be updated over time. 1047 For example, MIME media types, character sets, and language tags, 1048 etc. typically include more information than just the registered 1049 value itself. Example information can include point-of-contact 1050 information, security issues, pointers to updates, literature 1051 references, etc. In such cases, the document defining the namespace 1052 must clearly state who is responsible for maintaining and updating a 1053 registration. In different cases, it may be appropriate to specify 1054 one or more of the following: 1056 o Let the author update the registration, subject to the same 1057 constraints and review as with new registrations. 1059 o Allow some mechanism to attach comments to the registration, for 1060 cases where others have significant objections to claims in a 1061 registration, but the author does not agree to change the 1062 registration. 1064 o Designate the IESG, a designated expert, or another entity as 1065 having the right to change the registrant associated with a 1066 registration and any requirements or conditions on doing so. This 1067 is mainly to get around the problem when a registrant cannot be 1068 reached in order to make necessary updates. 1070 5.3. Overriding Registration Procedures 1072 Since RFC 2434 was published, experience has shown that the 1073 documented IANA considerations for individual protocols do not always 1074 adequately cover the reality after the protocol is deployed. For 1075 example, many older routing protocols do not have documented, 1076 detailed IANA considerations. In addition, documented IANA 1077 considerations are sometimes found to be too stringent to allow even 1078 working group documents (for which there is strong consensus) to 1079 obtain code points from IANA in advance of actual RFC publication. 1080 In other cases, the documented procedures are unclear or neglected to 1081 cover all the cases. In order to allow assignments in individual 1082 cases where there is strong IETF consensus that an allocation should 1083 go forward, but the documented procedures do not support such an 1084 assignment, the IESG is granted authority to approve assignments in 1085 such cases. The intention is not to overrule properly documented 1086 procedures, or to obviate the need for protocols to properly document 1087 their IANA considerations. Instead, the intention is to permit 1088 assignments in individual cases where it is obvious that the 1089 assignment should just be made, but updating the IANA process just to 1090 assign a particular code point is viewed as too heavy a burden. 1092 In general, the IETF would like to see deficient IANA registration 1093 procedures for a namespace revised through the IETF standards 1094 process, but not at the cost of unreasonable delay for needed 1095 assignments. If the IESG has had to take the action in this section, 1096 it is a strong indicator that the IANA registration procedures should 1097 be updated, possibly in parallel with ongoing protocol work. 1099 6. Documentation References in IANA Registries 1101 Usually, registries and registry entries include references to 1102 documentation (RFCs or other documents). The purpose of these 1103 references is to provide pointers for implementors to find details 1104 necessary for implementation, NOT to simply note what document 1105 created the registry or entry. Therefore: 1107 o If a document registers an item that is defined and explained 1108 elsewhere, the registered reference should be to that document, 1109 and not to the document that is merely performing the 1110 registration. 1112 o If the registered item is defined and explained in the current 1113 document, it is important to include sufficient information to 1114 enable implementors to understand the item and to create a proper 1115 implementation. 1117 o If the registered item is explained primarily in a specific 1118 section of the reference document, it is useful to include a 1119 section reference. For example, "[RFC9876], Section 3.2", rather 1120 than just "[RFC9876]". 1122 o For documentation of a new registry, the reference should provide 1123 information about the registry itself, not just a pointer to the 1124 creation of it. Useful information includes the purpose of the 1125 registry, a rationale for its creation, documentation of the 1126 process and policy for new registrations, guidelines for new 1127 registrants or designated experts, and other such related 1128 information. 1130 7. What to Do in "bis" Documents 1132 We often produce a new edition of an RFC, which obsoletes the 1133 previous edition (we sometimes call these "bis" documents, such as 1134 when RFC 9876 is updated by draft-ietf-foo-rfc9876bis). When the 1135 original document created registries and/or registered entries, there 1136 is a question of how to handle the IANA Considerations section in the 1137 "bis" document. 1139 If the registrations specify the original document as a reference, 1140 those registrations should be updated to point to the current (not 1141 obsolete) documentation for those items. Usually, that will mean 1142 changing the reference to be the "bis" document. 1144 For example, suppose RFC 9876 registered the "BANANA" flag in the 1145 "Fruit Access Flags" registry, and the documentation for that flag is 1146 in Section 3.2. The current registry might look, in part, like this: 1148 Name Description Reference 1149 -------- ------------------- --------- 1150 BANANA Flag for bananas [RFC9876], Section 3.2 1152 If draft-ietf-foo-rfc9876bis obsoletes RFC 9876 and, because of some 1153 rearrangement, now documents the flag in Section 4.1.2, the IANA 1154 Considerations of the bis document might contain text such as this: 1156 IANA is asked to change the registration information for the 1157 BANANA flag in the "Fruit Access Flags" registry to the following: 1159 Name Description Reference 1160 -------- ------------------- --------- 1161 BANANA Flag for bananas [[this RFC]], Section 4.2.1 1163 In many cases, if there are a number of registered references to the 1164 original RFC and the document organization has not changed the 1165 registered section numbering much, it may simply be reasonable to do 1166 this: 1168 Because this document obsoletes RFC 9876, IANA is asked to change 1169 all registration information that references [RFC9876] to instead 1170 reference [[this RFC]]. 1172 If information for registered items has been or is being moved to 1173 other documents, then, of course, the registration information should 1174 be changed to point to those other documents. In no case is it 1175 reasonable to leave documentation pointers to the obsoleted document 1176 for any registries or registered items that are still in current use. 1178 8. Miscellaneous Issues 1180 8.1. When There Are No IANA Actions 1182 Before an Internet-Draft can be published as an RFC, IANA needs to 1183 know what actions (if any) it needs to perform. Experience has shown 1184 that it is not always immediately obvious whether a document has no 1185 IANA actions, without reviewing the document in some detail. In 1186 order to make it clear to IANA that it has no actions to perform (and 1187 that the author has consciously made such a determination), such 1188 documents should include an IANA Considerations section that states: 1190 This document has no IANA actions. 1192 This statement, or an equivalent, must only be inserted after the 1193 working group or individual submitter has carefully verified it to be 1194 true. Using such wording as a matter of "boilerplate" or without 1195 careful consideration can lead to incomplete or incorrect IANA 1196 actions being performed. 1198 If a specification makes use of values from a namespace that is not 1199 managed by IANA, it may be useful to note this fact, with wording 1200 such as this: 1202 The values of the Foobar parameter are assigned by the Barfoo 1203 registry on behalf of the Rabfoo Forum. Therefore, this document 1204 has no IANA actions. 1206 In some cases, the absence of IANA-assigned values may be considered 1207 valuable information for future readers; in other cases, it may be 1208 considered of no value once the document has been approved, and may 1209 be removed before archival publication. This choice should be made 1210 clear in the draft, for example, by including a sentence such as 1212 [RFC Editor: please remove this section prior to publication.] 1214 or 1216 [RFC Editor: please do not remove this section.] 1218 8.2. Namespaces Lacking Documented Guidance 1220 For all existing RFCs that either explicitly or implicitly rely on 1221 IANA to evaluate assignments without specifying a precise evaluation 1222 policy, IANA (in consultation with the IESG) will continue to decide 1223 what policy is appropriate. Changes to existing policies can always 1224 be initiated through the normal IETF consensus process. 1226 All future RFCs that either explicitly or implicitly rely on IANA to 1227 register or otherwise manage namespace assignments MUST provide 1228 guidelines for managing the namespace. 1230 8.3. After-the-Fact Registrations 1232 Occasionally, IANA becomes aware that an unassigned value from a 1233 managed namespace is in use on the Internet or that an assigned value 1234 is being used for a different purpose than originally registered. 1235 IANA will not condone such misuse; procedures of the type described 1236 in this document MUST be applied to such cases. In the absence of 1237 specifications to the contrary, values may only be reassigned for a 1238 different purpose with the consent of the original assignee (when 1239 possible) and with due consideration of the impact of such a 1240 reassignment. In cases of likely controversy, consultation with the 1241 IESG is advised. 1243 8.4. Reclaiming Assigned Values 1245 Reclaiming previously assigned values for reuse is tricky, because 1246 doing so can lead to interoperability problems with deployed systems 1247 still using the assigned values. Moreover, it can be extremely 1248 difficult to determine the extent of deployment of systems making use 1249 of a particular value. However, in cases where the namespace is 1250 running out of unassigned values and additional ones are needed, it 1251 may be desirable to attempt to reclaim unused values. When 1252 reclaiming unused values, the following (at a minimum) should be 1253 considered: 1255 o Attempts should be made to contact the original party to which a 1256 value is assigned, to determine if the value was ever used, and if 1257 so, the extent of deployment. (In some cases, products were never 1258 shipped or have long ceased being used. In other cases, it may be 1259 known that a value was never actually used at all.) 1261 o Reassignments should not normally be made without the concurrence 1262 of the original requester. Reclamation under such conditions 1263 should only take place where there is strong evidence that a value 1264 is not widely used, and the need to reclaim the value outweighs 1265 the cost of a hostile reclamation. In any case, IESG Approval is 1266 needed in this case. 1268 o It may be appropriate to write up the proposed action and solicit 1269 comments from relevant user communities. In some cases, it may be 1270 appropriate to write an RFC that goes through a formal IETF 1271 process (including IETF Last Call) as was done when DHCP reclaimed 1272 some of its "Private Use" options [RFC3942]. 1274 8.5. Contact Person vs Assignee or Owner 1276 Many registries include designation of a technical or administrative 1277 contact associated with each entry. Often, this is recorded as 1278 contact information for an individual. It is unclear, though, what 1279 role the individual has with respect to the registration: is this 1280 item registered on behalf of the individual, the company the 1281 individual worked for, or perhaps another organization the individual 1282 was acting for? 1284 This matters because some time later, when the individual has changed 1285 jobs or roles, and perhaps can no longer be contacted, someone might 1286 want to update the registration. IANA has no way to know what 1287 company, organization, or individual should be allowed to take the 1288 registration over. For registrations rooted in RFCs, the stream 1289 owner (such as the IESG or the IAB) can make an overriding decision. 1290 But in other cases, there is no recourse. 1292 Registries can include, in addition to a "Contact" field, an 1293 "Assignee" or "Owner" field that can be used to address this 1294 situation, giving IANA clear guidance as to the actual owner of the 1295 registration. Alternatively, organizations can put an organizational 1296 role into the "Contact" field in order to make their ownership clear. 1298 8.6. Closing or Obsoleting a Registry 1300 [[anchor2: This section needs to be resolved before publication.]] 1302 8.7. BCP 78/79 Issues in Registries 1304 [[anchor3: This section needs to be resolved before publication.]] 1306 9. Appeals 1308 Appeals of registration decisions made by IANA can be made using the 1309 normal IETF appeals process as described in Section 6.5 of [RFC2026]. 1310 Specifically, appeals should be directed to the IESG, followed (if 1311 necessary) by an appeal to the IAB, etc. 1313 10. Mailing Lists 1315 All IETF mailing lists associated with evaluating or discussing 1316 assignment requests as described in this document are subject to 1317 whatever rules of conduct and methods of list management are 1318 currently defined by Best Current Practices or by IESG decision. 1320 11. Security Considerations 1322 Information that creates or updates a registration needs to be 1323 authenticated and authorized. IANA updates registries according to 1324 instructions in published RFCs and from the IESG. It also may accept 1325 clarifications from document authors, relevant working group chairs, 1326 Designated Experts, and mail list participants, too. 1328 Information concerning possible security vulnerabilities of a 1329 protocol may change over time. Likewise, security vulnerabilities 1330 related to how an assigned number is used may change as well. As new 1331 vulnerabilities are discovered, information about such 1332 vulnerabilities may need to be attached to existing registrations, so 1333 that users are not misled as to the true security issues surrounding 1334 the use of a registered number. 1336 An analysis of security issues is generally required for all 1337 protocols that make use of parameters (data types, operation codes, 1338 keywords, etc.) used in IETF protocols or registered by IANA. Such 1339 security considerations are usually included in the protocol document 1340 [RFC3552]. It is the responsibility of the IANA considerations 1341 associated with a particular registry to specify what (if any) 1342 security considerations must be provided when assigning new values, 1343 and the process for reviewing such claims. 1345 12. Changes Relative to Earlier Editions of BCP 26 1347 12.1. 2012: Changes in This Document Relative to RFC 5226 1349 Significant additions: 1351 o Added Section 3.4, Expert Reviews and the Document Lifecycle 1353 o Moved well-known policies into a separate section for each, 1354 subsections of Section 4.1. 1356 o Added Section 4.2, Best Practice for Selecting an Appropriate 1357 Policy. 1359 o Added Section 4.3, Using Multiple Policies in Combination. 1361 o Added Section 6, Documentation References in IANA Registries 1363 o Added Section 7, What to Do in "bis" Documents 1365 o Added Section 8.5, Contact Person vs Assignee or Owner 1367 o Added Section 8.6, Closing or Obsoleting a Registry 1369 o Added Section 8.7, BCP 78/79 Issues in Registries 1371 Clarifications and such: 1373 o Made clarifications about identification of IANA registries and 1374 use of URLs for them. 1376 o Clarified the distinction between "Unassigned" and "Reserved". 1378 o Made some clarifications in "Expert Review" about instructions to 1379 the designated expert. 1381 o Made some clarifications in "Specification Required" about how to 1382 declare this policy. 1384 o Assorted minor clarifications and editorial changes throughout. 1386 12.2. 2008: Changes in RFC 5226 Relative to RFC 2434 1388 Changes include: 1390 o Major reordering of text to expand descriptions and to better 1391 group topics such as "updating registries" vs. "creating new 1392 registries", in order to make it easier for authors to find the 1393 text most applicable to their needs. 1395 o Numerous editorial changes to improve readability. 1397 o Changed the term "IETF Consensus" to "IETF Review" and added more 1398 clarifications. History has shown that people see the words "IETF 1399 Consensus" (without consulting the actual definition) and are 1400 quick to make incorrect assumptions about what the term means in 1401 the context of IANA Considerations. 1403 o Added "RFC Required" to list of defined policies. 1405 o Much more explicit directions and examples of "what to put in 1406 RFCs". 1408 o "Specification Required" now implies use of a Designated Expert to 1409 evaluate specs for sufficient clarity. 1411 o Significantly changed the wording in the Designated Experts 1412 section. Main purpose is to make clear that Expert Reviewers are 1413 accountable to the community, and to provide some guidance for 1414 review criteria in the default case. 1416 o Changed wording to remove any special appeals path. The normal 1417 RFC 2026 appeals path is used. 1419 o Added a section about reclaiming unused value. 1421 o Added a section on after-the-fact registrations. 1423 o Added a section indicating that mailing lists used to evaluate 1424 possible assignments (such as by a Designated Expert) are subject 1425 to normal IETF rules. 1427 13. Acknowledgments 1429 13.1. Acknowledgments for This Document (2012) 1431 Thomas Narten and Harald Tveit Alvestrand edited the two earlier 1432 editions of this document (RFCs 2434 and 5226), and Thomas continues 1433 his role in this third edition. Most of the text from RFC 5226 1434 remains in this edition. 1436 13.2. Acknowledgments from the second edition (2008) 1438 The original acknowledgments section in RFC 5226 was: 1440 This document has benefited from specific feedback from Jari Arkko, 1441 Marcelo Bagnulo Braun, Brian Carpenter, Michelle Cotton, Spencer 1442 Dawkins, Barbara Denny, Miguel Garcia, Paul Hoffman, Russ Housley, 1443 John Klensin, Allison Mankin, Blake Ramsdell, Mark Townsley, Magnus 1444 Westerlund, and Bert Wijnen. 1446 13.3. Acknowledgments from the first edition (1998) 1448 The original acknowledgments section in RFC 2434 was: 1450 Jon Postel and Joyce Reynolds provided a detailed explanation on what 1451 IANA needs in order to manage assignments efficiently, and patiently 1452 provided comments on multiple versions of this document. Brian 1453 Carpenter provided helpful comments on earlier versions of the 1454 document. One paragraph in the Security Considerations section was 1455 borrowed from [RFC4288]. 1457 14. References 1459 14.1. Normative References 1461 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1462 Requirement Levels", BCP 14, RFC 2119, March 1997. 1464 14.2. Informative References 1466 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, 1467 September 1981. 1469 [RFC2026] Bradner, S., "The Internet Standards Process -- Revision 1470 3", BCP 9, RFC 2026, October 1996. 1472 [RFC2132] Alexander, S. and R. Droms, "DHCP Options and BOOTP Vendor 1473 Extensions", RFC 2132, March 1997. 1475 [RFC2860] Carpenter, B., Baker, F., and M. Roberts, "Memorandum of 1476 Understanding Concerning the Technical Work of the 1477 Internet Assigned Numbers Authority", RFC 2860, June 2000. 1479 [RFC2939] Droms, R., "Procedures and IANA Guidelines for Definition 1480 of New DHCP Options and Message Types", BCP 43, RFC 2939, 1481 September 2000. 1483 [RFC3228] Fenner, B., "IANA Considerations for IPv4 Internet Group 1484 Management Protocol (IGMP)", BCP 57, RFC 3228, 1485 February 2002. 1487 [RFC3232] Reynolds, J., "Assigned Numbers: RFC 1700 is Replaced by 1488 an On-line Database", RFC 3232, January 2002. 1490 [RFC3552] Rescorla, E. and B. Korver, "Guidelines for Writing RFC 1491 Text on Security Considerations", BCP 72, RFC 3552, 1492 July 2003. 1494 [RFC3575] Aboba, B., "IANA Considerations for RADIUS (Remote 1495 Authentication Dial In User Service)", RFC 3575, 1496 July 2003. 1498 [RFC3692] Narten, T., "Assigning Experimental and Testing Numbers 1499 Considered Useful", BCP 82, RFC 3692, January 2004. 1501 [RFC3748] Aboba, B., Blunk, L., Vollbrecht, J., Carlson, J., and H. 1502 Levkowetz, "Extensible Authentication Protocol (EAP)", 1503 RFC 3748, June 2004. 1505 [RFC3942] Volz, B., "Reclassifying Dynamic Host Configuration 1506 Protocol version 4 (DHCPv4) Options", RFC 3942, 1507 November 2004. 1509 [RFC3968] Camarillo, G., "The Internet Assigned Number Authority 1510 (IANA) Header Field Parameter Registry for the Session 1511 Initiation Protocol (SIP)", BCP 98, RFC 3968, 1512 December 2004. 1514 [RFC4005] Calhoun, P., Zorn, G., Spence, D., and D. Mitton, 1515 "Diameter Network Access Server Application", RFC 4005, 1516 August 2005. 1518 [RFC4025] Richardson, M., "A Method for Storing IPsec Keying 1519 Material in DNS", RFC 4025, March 2005. 1521 [RFC4044] McCloghrie, K., "Fibre Channel Management MIB", RFC 4044, 1522 May 2005. 1524 [RFC4124] Le Faucheur, F., "Protocol Extensions for Support of 1525 Diffserv-aware MPLS Traffic Engineering", RFC 4124, 1526 June 2005. 1528 [RFC4169] Torvinen, V., Arkko, J., and M. Naslund, "Hypertext 1529 Transfer Protocol (HTTP) Digest Authentication Using 1530 Authentication and Key Agreement (AKA) Version-2", 1531 RFC 4169, November 2005. 1533 [RFC4271] Rekhter, Y., Li, T., and S. Hares, "A Border Gateway 1534 Protocol 4 (BGP-4)", RFC 4271, January 2006. 1536 [RFC4283] Patel, A., Leung, K., Khalil, M., Akhtar, H., and K. 1537 Chowdhury, "Mobile Node Identifier Option for Mobile IPv6 1538 (MIPv6)", RFC 4283, November 2005. 1540 [RFC4288] Freed, N. and J. Klensin, "Media Type Specifications and 1541 Registration Procedures", BCP 13, RFC 4288, December 2005. 1543 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 1544 Internet Protocol", RFC 4301, December 2005. 1546 [RFC4340] Kohler, E., Handley, M., and S. Floyd, "Datagram 1547 Congestion Control Protocol (DCCP)", RFC 4340, March 2006. 1549 [RFC4395] Hansen, T., Hardie, T., and L. Masinter, "Guidelines and 1550 Registration Procedures for New URI Schemes", BCP 35, 1551 RFC 4395, February 2006. 1553 [RFC4422] Melnikov, A. and K. Zeilenga, "Simple Authentication and 1554 Security Layer (SASL)", RFC 4422, June 2006. 1556 [RFC4446] Martini, L., "IANA Allocations for Pseudowire Edge to Edge 1557 Emulation (PWE3)", BCP 116, RFC 4446, April 2006. 1559 [RFC4520] Zeilenga, K., "Internet Assigned Numbers Authority (IANA) 1560 Considerations for the Lightweight Directory Access 1561 Protocol (LDAP)", BCP 64, RFC 4520, June 2006. 1563 [RFC4589] Schulzrinne, H. and H. Tschofenig, "Location Types 1564 Registry", RFC 4589, July 2006. 1566 [RFC4727] Fenner, B., "Experimental Values In IPv4, IPv6, ICMPv4, 1567 ICMPv6, UDP, and TCP Headers", RFC 4727, November 2006. 1569 [RFC5246] Dierks, T. and E. Rescorla, "The Transport Layer Security 1570 (TLS) Protocol Version 1.2", RFC 5246, August 2008. 1572 [RFC5378] Bradner, S. and J. Contreras, "Rights Contributors Provide 1573 to the IETF Trust", BCP 78, RFC 5378, November 2008. 1575 [RFC5742] Alvestrand, H. and R. Housley, "IESG Procedures for 1576 Handling of Independent and IRTF Stream Submissions", 1577 BCP 92, RFC 5742, December 2009. 1579 [RFC5771] Cotton, M., Vegoda, L., and D. Meyer, "IANA Guidelines for 1580 IPv4 Multicast Address Assignments", BCP 51, RFC 5771, 1581 March 2010. 1583 [RFC5795] Sandlund, K., Pelletier, G., and L-E. Jonsson, "The RObust 1584 Header Compression (ROHC) Framework", RFC 5795, 1585 March 2010. 1587 [RFC6195] Eastlake, D., "Domain Name System (DNS) IANA 1588 Considerations", BCP 42, RFC 6195, March 2011. 1590 [RFC6275] Perkins, C., Johnson, D., and J. Arkko, "Mobility Support 1591 in IPv6", RFC 6275, July 2011. 1593 [RFC6709] Carpenter, B., Aboba, B., and S. Cheshire, "Design 1594 Considerations for Protocol Extensions", RFC 6709, 1595 September 2012. 1597 Authors' Addresses 1599 Michelle Cotton 1600 Internet Assigned Numbers Authority (IANA) 1601 4676 Admiralty Way, Suite 330 1602 Marina del Rey, CA 90292 1603 USA 1605 Phone: +1 310 301 5812 1606 Email: michelle.cotton@icann.org 1608 Barry Leiba 1609 Huawei Technologies 1611 Phone: +1 646 827 0648 1612 Email: barryleiba@computer.org 1613 URI: http://internetmessagingtechnology.org/ 1614 Thomas Narten 1615 IBM Corporation 1616 3039 Cornwallis Ave., PO Box 12195 - BRQA/502 1617 Research Triangle Park, NC 27709-2195 1618 USA 1620 Phone: +1 919 254 7798 1621 Email: narten@us.ibm.com