idnits 2.17.1 draft-kunze-ark-21.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.) ** The document seems to lack separate sections for Informative/Normative References. All references will be assumed normative when checking for downward references. ** The abstract seems to contain references ([Qualifier]), which it shouldn't. Please replace those with straight textual mentions of the documents in question. == There are 2 instances of lines with non-RFC2606-compliant FQDNs in the document. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year == Line 1762 has weird spacing: '... regexp repla...' -- The document seems to lack a disclaimer for pre-RFC5378 work, but may have content which was first submitted before 10 November 2008. If you have contacted all the original authors and they are all willing to grant the BCP78 rights to the IETF Trust, then this is fine, and you can ignore this comment. If not, you may need to add the pre-RFC5378 disclaimer. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (May 29, 2019) is 1793 days in the past. Is this intentional? -- Found something which looks like a code comment -- if you have code sections in the document, please surround them with '' and '' lines. Checking references for intended status: Informational ---------------------------------------------------------------------------- == Missing Reference: 'Qualifier' is mentioned on line 513, but not defined ** Obsolete normative reference: RFC 2141 (Obsoleted by RFC 8141) ** Obsolete normative reference: RFC 2611 (Obsoleted by RFC 3406) ** Obsolete normative reference: RFC 2616 (Obsoleted by RFC 7230, RFC 7231, RFC 7232, RFC 7233, RFC 7234, RFC 7235) ** Obsolete normative reference: RFC 2822 (Obsoleted by RFC 5322) ** Obsolete normative reference: RFC 2915 (Obsoleted by RFC 3401, RFC 3402, RFC 3403, RFC 3404) Summary: 8 errors (**), 0 flaws (~~), 4 warnings (==), 3 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group J. Kunze 3 Internet-Draft California Digital Library 4 Intended status: Informational E. Bermes 5 Expires: November 30, 2019 Bibliotheque nationale de France 6 May 29, 2019 8 The ARK Identifier Scheme 9 draft-kunze-ark-21 11 Abstract 13 The ARK (Archival Resource Key) naming scheme is designed to 14 facilitate the high-quality and persistent identification of 15 information objects. A founding principle of the ARK is that 16 persistence is purely a matter of service and is neither inherent in 17 an object nor conferred on it by a particular naming syntax. The 18 best that an identifier can do is to lead users to the services that 19 support robust reference. The term ARK itself refers both to the 20 scheme and to any single identifier that conforms to it. An ARK has 21 five components: 23 [http://NMAH/]ark:[/]NAAN/Name[Qualifier] 25 an optional and mutable Name Mapping Authority Hostport (usually a 26 hostname), the "ark:" label, the Name Assigning Authority Number 27 (NAAN), the assigned Name, and an optional and possibly mutable 28 Qualifier supported by the NMA. The NAAN and Name together form the 29 immutable persistent identifier for the object independent of the URL 30 hostname. An ARK is a special kind of URL that connects users to 31 three things: the named object, its metadata, and the provider's 32 promise about its persistence. When entered into the location field 33 of a Web browser, the ARK leads the user to the named object. That 34 same ARK, inflected by appending a single question mark (`?'), 35 returns a brief metadata record that is both human- and machine- 36 readable. When the ARK is inflected by appending dual question marks 37 (`??'), the returned metadata contains a commitment statement from 38 the current provider. Tools exist for minting, binding, and 39 resolving ARKs. 41 Status of This Memo 43 This Internet-Draft is submitted in full conformance with the 44 provisions of BCP 78 and BCP 79. 46 Internet-Drafts are working documents of the Internet Engineering 47 Task Force (IETF). Note that other groups may also distribute 48 working documents as Internet-Drafts. The list of current Internet- 49 Drafts is at https://datatracker.ietf.org/drafts/current/. 51 Internet-Drafts are draft documents valid for a maximum of six months 52 and may be updated, replaced, or obsoleted by other documents at any 53 time. It is inappropriate to use Internet-Drafts as reference 54 material or to cite them other than as "work in progress." 56 This Internet-Draft will expire on November 30, 2019. 58 Copyright Notice 60 Copyright (c) 2019 IETF Trust and the persons identified as the 61 document authors. All rights reserved. 63 This document is subject to BCP 78 and the IETF Trust's Legal 64 Provisions Relating to IETF Documents 65 (https://trustee.ietf.org/license-info) in effect on the date of 66 publication of this document. Please review these documents 67 carefully, as they describe your rights and restrictions with respect 68 to this document. 70 Table of Contents 72 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 73 1.1. Reasons to Use ARKs . . . . . . . . . . . . . . . . . . . 4 74 1.2. Three Requirements of ARKs . . . . . . . . . . . . . . . 5 75 1.3. Organizing Support for ARKs: Our Stuff vs. Their Stuff . 6 76 1.4. Definition of Identifier . . . . . . . . . . . . . . . . 7 77 2. ARK Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . 8 78 2.1. The Name Mapping Authority Hostport (NMAH) . . . . . . . 9 79 2.2. The ARK Label Part (ark:) . . . . . . . . . . . . . . . . 11 80 2.3. The Name Assigning Authority Number (NAAN) . . . . . . . 11 81 2.4. The Name Part . . . . . . . . . . . . . . . . . . . . . . 12 82 2.5. The Qualifier Part . . . . . . . . . . . . . . . . . . . 13 83 2.5.1. ARKs that Reveal Object Hierarchy . . . . . . . . . . 14 84 2.5.2. ARKs that Reveal Object Variants . . . . . . . . . . 15 85 2.6. Character Repertoires . . . . . . . . . . . . . . . . . . 16 86 2.7. Normalization and Lexical Equivalence . . . . . . . . . . 17 87 3. Naming Considerations . . . . . . . . . . . . . . . . . . . . 18 88 3.1. ARKS Embedded in Language . . . . . . . . . . . . . . . . 19 89 3.2. Objects Should Wear Their Identifiers . . . . . . . . . . 19 90 3.3. Names are Political, not Technological . . . . . . . . . 19 91 3.4. Choosing a Hostname or NMA . . . . . . . . . . . . . . . 20 92 3.5. Assigners of ARKs . . . . . . . . . . . . . . . . . . . . 21 93 3.6. NAAN Namespace Management . . . . . . . . . . . . . . . . 22 94 3.7. Sub-Object Naming . . . . . . . . . . . . . . . . . . . . 23 95 4. Finding a Name Mapping Authority . . . . . . . . . . . . . . 24 96 4.1. Looking Up NMAHs in a Globally Accessible File . . . . . 25 97 5. Generic ARK Service Definition . . . . . . . . . . . . . . . 26 98 5.1. Generic ARK Access Service (access, location) . . . . . . 27 99 5.1.1. Generic Policy Service (permanence, naming, etc.) . . 27 100 5.1.2. Generic Description Service . . . . . . . . . . . . . 29 101 5.2. Overview of The HTTP URL Mapping Protocol (THUMP) . . . . 29 102 5.3. The Electronic Resource Citation (ERC) . . . . . . . . . 32 103 5.4. Advice to Web Clients . . . . . . . . . . . . . . . . . . 34 104 5.5. Security Considerations . . . . . . . . . . . . . . . . . 35 105 6. References . . . . . . . . . . . . . . . . . . . . . . . . . 35 106 Appendix A. ARK Maintenance Agency: arks.org . . . . . . . . . . 37 107 Appendix B. Looking up NMAHs Distributed via DNS . . . . . . . . 38 108 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 40 110 1. Introduction 112 [ Note about this transitional draft. The ARKsInTheOpen.org 113 Technical Working Group (https://wiki.duraspace.org/display/ARKs/ 114 Technical+Working+Group) is in the process of revising the ARK spec 115 via a series of Internet-Drafts. No breaking changes from the 2008 116 spec are envisaged. Some minor changes are being deferred to later 117 in order to make it easier to review more important changes; some of 118 those small changes would result in "noisy diffs" since they are 119 global in scope, for example, converting all instances of http:// and 120 NMAH to https:// and NMA, respectively. ] 122 This document describes a scheme for the high-quality naming of 123 information resources. The scheme, called the Archival Resource Key 124 (ARK), is well suited to long-term access and identification of any 125 information resources that accommodate reasonably regular electronic 126 description. This includes digital documents, databases, software, 127 and websites, as well as physical objects (books, bones, statues, 128 etc.) and intangible objects (chemicals, diseases, vocabulary terms, 129 performances). Hereafter the term "object" refers to an information 130 resource. The term ARK itself refers both to the scheme and to any 131 single identifier that conforms to it. A reasonably concise and 132 accessible overview and rationale for the scheme is available at 133 [ARK]. 135 Schemes for persistent identification of network-accessible objects 136 are not new. In the early 1990's, the design of the Uniform Resource 137 Name [RFC2141] responded to the observed failure rate of URLs by 138 articulating an indirect, non-hostname-based naming scheme and the 139 need for responsible name management. Meanwhile, promoters of the 140 Digital Object Identifier [DOI] succeeded in building a community of 141 providers around a mature software system [Handle] that supports name 142 management. The Persistent Uniform Resource Locator [PURL] was 143 another scheme that had the advantage of working with unmodified web 144 browsers. ARKs represent an approach that attempts to build on the 145 strengths and to avoid the weaknesses of these schemes. 147 A founding principle of the ARK is that persistence is purely a 148 matter of service. Persistence is neither inherent in an object nor 149 conferred on it by a particular naming syntax. Nor is the technique 150 of name indirection -- upon which URNs, Handles, DOIs, and PURLs are 151 founded -- of central importance. Name indirection is an ancient and 152 well-understood practice; new mechanisms for it keep appearing and 153 distracting practitioner attention, with the Domain Name System (DNS) 154 [RFC1034] being a particularly dazzling and elegant example. What is 155 often forgotten is that maintenance of an indirection table is an 156 unavoidable cost to the organization providing persistence, and that 157 cost is equivalent across naming schemes. That indirection has 158 always been a native part of the web while being so lightly utilized 159 for the persistence of web-based objects indicates how unsuited most 160 organizations will probably be to the task of table maintenance and 161 to the much more fundamental challenge of keeping the objects 162 themselves viable. 164 Persistence is achieved through a provider's successful stewardship 165 of objects and their identifiers. The highest level of persistence 166 will be reinforced by a provider's robust contingency, redundancy, 167 and succession strategies. It is further safeguarded to the extent 168 that a provider's mission is shielded from funding and political 169 instabilities. These are by far the major challenges confronting 170 persistence providers, and no identifier scheme has any direct impact 171 on them. In fact, some schemes may actually be liabilities for 172 persistence because they create short- and long-term dependencies for 173 every object access on complex, special-purpose infrastructures, 174 parts of which are proprietary and all of which increase the carry- 175 forward burden for the preservation community. It is for this reason 176 that the ARK scheme relies only on educated name assignment and light 177 use of general-purpose infrastructures that are maintained mostly by 178 the internet community at large (the DNS, web servers, and web 179 browsers). 181 1.1. Reasons to Use ARKs 183 If no persistent identifier scheme contributes directly to 184 persistence, why not just use URLs? A particular URL may be as 185 durable an identifier as it is possible to have, but nothing 186 distinguishes it from an ordinary URL to the recipient who is 187 wondering if it is suitable for long-term reference. An ARK embedded 188 in a URL provides some of the necessary conditions for credible 189 persistence, inviting access to not one, but to three things: to the 190 object, to its metadata, and to a nuanced statement of commitment 191 from the provider in question (the NMA, described below) regarding 192 the object. Existence of the two extra services can be probed 193 automatically by appending `?' and `??' to the ARK. 195 The form of the ARK also supports the natural separation of naming 196 authorities into the original name assigning authority and the 197 diverse multiple name mapping (or servicing) authorities that in 198 succession and in parallel will take over custodial responsibilities 199 from the original assigner (assuming the assigner ever held that 200 responsibility) for the large majority of a long-term object's 201 archival lifetime. The name mapping authority, indicated by the 202 hostname part of the URL that contains the ARK, serves to launch the 203 ARK into cyberspace. Should it ever fail (and there is no reason why 204 a well-chosen hostname for a 100-year-old cultural memory institution 205 shouldn't last as long as the DNS), that host name is considered 206 disposeable and replaceable. Again, the form of the ARK helps 207 because it defines exactly how to recover the core immutable object 208 identity, and simple algorithms (one based on the URN model) or even 209 by-hand internet query can be used for for locating another mapping 210 authority. 212 There are tools to assist in generating ARKs and other identifiers, 213 such as [NOID] and "uuidgen", both of which rely for uniqueness on 214 human-maintained registries. This document also contains some 215 guidelines and considerations for managing namespaces and choosing 216 hostnames with persistence in mind. 218 1.2. Three Requirements of ARKs 220 The first requirement of an ARK is to give users a link from an 221 object to a promise of stewardship for it. That promise is a multi- 222 faceted covenant that binds the word of an identified service 223 provider to a specific set of responsibilities. It is critical for 224 the promise to come from a current provider and almost irrelevant, 225 over a long period of time, what the original assigner's intentions 226 were. No one can tell if successful stewardship will take place 227 because no one can predict the future. Reasonable conjecture, 228 however, may be based on past performance. There must be a way to 229 tie a promise of persistence to a provider's demonstrated or 230 perceived ability -- its reputation -- in that arena. Provider 231 reputations would then rise and fall as promises are observed 232 variously to be kept and broken. This is perhaps the best way we 233 have for gauging the strength of any persistence promise. 235 The second requirement of an ARK is to give users a link from an 236 object to a description of it. The problem with a naked identifier 237 is that without a description real identification is incomplete. 238 Identifiers common today are relatively opaque, though some contain 239 ad hoc clues reflecting assertions that were briefly true, such as 240 where in a filesystem hierarchy an object lived during a short stay. 241 Possession of both an identifier and an object is some improvement, 242 but positive identification may still be uncertain since the object 243 itself might not include a matching identifier or might not carry 244 evidence obvious enough to reveal its identity without significant 245 research. In either case, what is called for is a record bearing 246 witness to the identifier's association with the object, as supported 247 by a recorded set of object characteristics. This descriptive record 248 is partly an identification "receipt" with which users and archivists 249 can verify an object's identity after brief inspection and a 250 plausible match with recorded characteristics such as title and size. 252 The final requirement of an ARK is to give users a link to the object 253 itself (or to a copy) if at all possible. Persistent access is the 254 central duty of an ARK. Persistent identification plays a vital 255 supporting role but, strictly speaking, it can be construed as no 256 more than a record attesting to the original assignment of a never- 257 reassigned identifier. Object access may not be feasible for various 258 reasons, such as a transient service outage, a catastrophic loss, a 259 licensing agreement that keeps an archive "dark" for a period of 260 years, or when an object's own lack of tangible existence confuses 261 normal concepts of access (e.g., a vocabulary term might be 262 "accessed" through its definition). In such cases the ARK's 263 identification role assumes a much higher profile. But attempts to 264 simplify the persistence problem by decoupling access from 265 identification and concentrating exclusively on the latter are of 266 questionable utility. A perfect system for assigning forever unique 267 identifiers might be created, but if it did so without reducing 268 access failure rates, no one would be interested. The central issue 269 -- which may be summed up as the "HTTP 404 Not Found" problem -- 270 would not have been addressed. 272 1.3. Organizing Support for ARKs: Our Stuff vs. Their Stuff 274 An organization and the user community it serves can often be seen to 275 struggle with two different areas of persistent identification: the 276 Our Stuff problem and the Their Stuff problem. In the Our Stuff 277 problem, we in the organization want our own objects to acquire 278 persistent names. Since we possess or control these objects, our 279 organization tackles the Our Stuff problem directly. Whether or not 280 the objects are named by ARKs, our organization is the responsible 281 party, so it can plan for, maintain, and make commitments about the 282 objects. 284 In the Their Stuff problem, we in the organization want others' 285 objects to acquire persistent names. These are objects that we do 286 not own or control, but some of which are critically important to us. 287 But because they are beyond our influence as far as support is 288 concerned, creating and maintaining persistent identifiers for Their 289 Stuff is not especially purposeful or feasible for us to engage in. 290 There is little that we can do about someone else's stuff except 291 encourage their uptake or adoption of persistence services. 293 Co-location of persistent access and identification services is 294 natural. Any organization that undertakes ongoing support of true 295 persistent identification (which includes description) is well-served 296 if it controls, owns, or otherwise has clear internal access to the 297 identified objects, and this gives it an advantage if it wishes also 298 to support persistent access to outsiders. Conversely, persistent 299 access to outsiders requires orderly internal collection management 300 procedures that include monitoring, acquisition, verification, and 301 change control over objects, which in turn requires object 302 identifiers persistent enough to support auditable record keeping 303 practices. 305 Although, organizing ARK services under one roof thus tends to make 306 sense, object hosting can successfully be separated from name 307 mapping. An example is when a name mapping authority centrally 308 provides uniform resolution services via a protocol gateway on behalf 309 of organizations that host objects behind a variety of access 310 protocols. It is also reasonable to build value-added description 311 services that rely on the underlying services of a set of mapping 312 authorities. 314 Supporting ARKs is not for every organization. By requiring 315 specific, revealed commitments to preservation, to object access, and 316 to description, the bar for providing ARK services is higher than for 317 some other identifier schemes. On the other hand, it would be hard 318 to grant credence to a persistence promise from an organization that 319 could not muster the minimum ARK services. Not that there isn't a 320 business model for an ARK-like, description-only service built on top 321 of another organization's full complement of ARK services. For 322 example, there might be competition at the description level for 323 abstracting and indexing a body of scientific literature archived in 324 a combination of open and fee-based repositories. The description- 325 only service would have no direct commitment to the objects, but 326 would act as an intermediary, forwarding commitment statements from 327 object hosting services to requestors. 329 1.4. Definition of Identifier 331 An identifier is not a string of character data -- an identifier is 332 an association between a string of data and an object. This 333 abstraction is necessary because without it a string is just data. 334 It's nonsense to talk about a string's breaking, or about its being 335 strong, maintained, and authentic. But as a representative of an 336 association, a string can do, metaphorically, the things that we 337 expect of it. 339 Without regard to whether an object is physical, digital, or 340 conceptual, to identify it is to claim an association between it and 341 a representative string, such as "Jane" or "ISBN 0596000278". What 342 gives a claim credibility is a set of verifiable assertions, or 343 metadata, about the object, such as age, height, title, or number of 344 pages. In other words, the association is made manifest by a record 345 (e.g., a cataloging or other metadata record) that vouches for it. 347 In the complete absence of any testimony (metadata) regarding an 348 association, a would-be identifier string is a meaningless sequence 349 of characters. To keep an externally visible but otherwise internal 350 string from being perceived as an identifier by outsiders, for 351 example, it suffices for an organization not to disclose the nature 352 of its association. For our immediate purpose, actual existence of 353 an association record is more important than its authenticity or 354 verifiability, which are outside the scope of this specification. 356 It is a gift to the identification process if an object carries its 357 own name as an inseparable part of itself, such as an identifier 358 imprinted on the first page of a document or embedded in a data 359 structure element of a digital document header. In cases where the 360 object is large, unwieldy, or unavailable (such as when licensing 361 restrictions are in effect), a metadata record that includes the 362 identifier string will usually suffice. That record becomes a 363 conveniently manipulable object surrogate, acting as both an 364 association "receipt" and "declaration". 366 Note that our definition of identifier extends the one in use for 367 Uniform Resource Identifiers [RFC3986]. The present document still 368 sometimes (ab)uses the terms "ARK" and "identifier" as shorthand for 369 the string part of an identifier, but the context should make the 370 meaning clear. 372 2. ARK Anatomy 374 An ARK is represented by a sequence of characters (a string) that 375 contains the label, "ark:", optionally preceded by the beginning part 376 of a URL. Here is a diagrammed example. 378 ARK ANATOMY Core Immutable Identity 379 =========== _______________|_______________ 380 / \ 381 Resolver Service Base Object Name Qualifier 382 _________|_______ ________|_______ ______|______ 383 / \/ \/ \ 384 http://example.org/ark:12025/654xz321/s3/f8.05v.tiff 385 \_________/ \__/\___/ \______/\____/\_______/ 386 | | | | | | 387 | Label | | Sub-parts Variants 388 | | | 389 Name Mapping Authority | Assigned Name 390 Hostport (NMAH) | 391 Name Assigning Authority Number (NAAN) 393 The ARK syntax can be summarized, 395 [http://NMAH/]ark:[/]NAAN/Name[Qualifier] 397 where the NMAH, '/', and Qualifier parts are in brackets to indicate 398 that they are optional. The Base Object Name is the substring 399 comprising the "ark:" label, the NAAN and the assigned Name. The 400 Resolver Service is replaceable and makes the ARK actionable for a 401 period of time. Without the Resolver Service part, what remains is 402 the Core Immutable Identity (the "persistible") part of the ARK. 404 2.1. The Name Mapping Authority Hostport (NMAH) 406 Before the "ark:" label may appear an optional Name Mapping Authority 407 Hostport (NMAH) that is a temporary address where ARK service 408 requests may be sent. Preceded by a URI-type protocol designation 409 such as "https://", it specifies a Resolver Service. The NMAH itself 410 is an Internet hostname or hostport combination having the same 411 format and semantics as the hostport part of a URL. The most 412 important thing about the NMAH is that it is "identity inert" from 413 the point of view of object identification. In other words, ARKs 414 that differ only in the optional NMAH part identify the same object. 415 Thus, for example, the following three ARKs are synonyms for just one 416 information object: 418 http://loc.gov/ark:12025/654xz321 419 http://rutgers.edu/ark:12025/654xz321 420 ark:12025/654xz321 422 Strictly speaking, in the realm of digital objects, these ARKs may 423 lead over time to somewhat different or diverging instances of the 424 originally named object. In an ideal world, divergence of persistent 425 objects is not desirable, but it is widely believed that digital 426 preservation efforts will inevitably lead to alterations in some 427 original objects (e.g, a format migration in order to preserve the 428 ability to display a document). If any of those objects are held 429 redundantly in more than one organization (a common preservation 430 strategy), chances are small that all holding organizations will 431 perform the same precise transformations and all maintain the same 432 object metadata. More significant divergence would be expected when 433 the holding organizations serve different audiences or compete with 434 each other. 436 The NMAH part makes an ARK into an actionable URL. As with many 437 internet parameters, it is helpful to approach the NMAH being liberal 438 in what you accept and conservative in what you propose. From the 439 recipient's point of view, the NMAH part should be treated as 440 temporary, disposable, and replaceable. From the NMA's point of 441 view, it should be chosen with the greatest concern for longevity. A 442 carefully chosen NMAH should be at least as permanent as the 443 providing organization's own hostname. In the case of a national or 444 university library, for example, there is no reason why the NMAH 445 should not be considerably more permanent than soft-funded proxy 446 hostnames such as hdl.handle.net, dx.doi.org, and purl.org. In 447 general and over time, however, it is not unexpected for an NMAH 448 eventually to stop working and require replacement with the NMAH of a 449 currently active service provider. 451 This replacement relies on a mapping authority "resolver" discovery 452 process, of which two alternate methods are outlined in a later 453 section. The ARK, URN, Handle, and DOI schemes all use a resolver 454 discovery model that sooner or later requires matching the original 455 assigning authority with a current provider servicing that 456 authority's named objects; once found, the resolver at that provider 457 performs what amounts to a redirect to a place where the object is 458 currently held. All the schemes rely on the ongoing functionality of 459 currently mainstream technologies such as the Domain Name System 460 [RFC1034] and web browsers. The Handle and DOI schemes in addition 461 require that the Handle protocol layer and global server grid be 462 available at all times. 464 The practice of prepending "http://" and an NMAH to an ARK is a way 465 of creating an actionable identifier by a method that is itself 466 temporary. Assuming that infrastructure supporting [RFC2616] 467 information retrieval will no longer be available one day, ARKs will 468 then have to be converted into new kinds of actionable identifiers. 469 By that time, if ARKs see widespread use, web browsers would 470 presumably evolve to perform this (currently simple) transformation 471 automatically. 473 2.2. The ARK Label Part (ark:) 475 The label part distinguishes an ARK from an ordinary identifier. 476 There is a new form of the label, "ark:", and an old form, "ark:/", 477 both of which must be recognized in perpetuity. Implementations 478 should generate new ARKs in the new form (without the "/") and 479 resolvers must always treat received ARKs as equivalent if they 480 differ only regard to new form versus old form labels. Thus these 481 two ARKs are equivalent: 483 ark:/12025/654xz321 484 ark:12025/654xz321 486 In a URL found in the wild, the label indicates that the URL stands a 487 reasonable chance of being an ARK. If the context warrants, 488 verification that it actually is an ARK can be done by testing it for 489 existence of the three ARK services. 491 Since nothing about an identifier syntax directly affects 492 persistence, the "ark:" label (like "urn:", "doi:", and "hdl:") 493 cannot tell you whether the identifier is persistent or whether the 494 object is available. It does tell you that the original Name 495 Assigning Authority (NAA) had some sort of hopes for it, but it 496 doesn't tell you whether that NAA is still in existence, or whether a 497 decade ago it ceased to have any responsibility for providing 498 persistence, or whether it ever had any responsibility beyond naming. 500 Only a current provider can say for certain what sort of commitment 501 it intends, and the ARK label suggests that you can query the NMAH 502 directly to find out exactly what kind of persistence is promised. 503 Even if what is promised is impersistence (i.e., a short-term 504 identifier), saying so is valuable information to the recipient. 505 Thus an ARK is a high-functioning identifier in the sense that it 506 provides access to the object, the metadata, and a commitment 507 statement, even if the commitment is explicitly very weak. 509 2.3. The Name Assigning Authority Number (NAAN) 511 Recalling that the general form of the ARK is, 513 [http://NMAH/]ark:[/]NAAN/Name[Qualifier] 515 the part of the ARK directly following the "ark:" (or older "ark:/") 516 label is the Name Assigning Authority Number (NAAN), up to but not 517 including the next `/' (slash) character. This part is always 518 required, as it identifies the organization that originally assigned 519 the Name of the object. It is used to discover a currently valid 520 NMAH and to provide top-level partitioning of the space of all ARKs. 522 An organization may request a NAAN from the ARK Maintenance Agency 523 [ARKagency] (described in Appendix A) by filling out the form at 524 [NAANrequest]. NAANs are opaque strings of one or more characters 525 drawn from this set, 527 0123456789bcdfghjkmnpqrstvwxz 529 which consists of digits and consonants, minus the letter 'l'. 530 Restricting NAANs to this set serves two goals. It reduces the 531 chances that words -- past, present, and future -- will appear in 532 NAANs and carry unintended semantics. It also helps usability by not 533 mixing commonly confused characters ('0' and 'O', '1' and 'l') and by 534 being compatible with strong transcription error detection (eg, the 535 [NOID] check digit algorithm). Since 2001, every assigned NAAN has 536 consisted of exactly five digits, and no immediate change in that 537 practice is foreseen. 539 The NAAN designates a top-level ARK namespace. Once registered for a 540 namespace, a NAAN is never re-registered. It is possible, however, 541 for there to be a succession of organizations that manage an ARK 542 namespace. 544 2.4. The Name Part 546 The part of the ARK just after the NAAN is the Name assigned by the 547 NAA, and it is also required. Semantic opaqueness in the Name part 548 is strongly encouraged in order to reduce an ARK's vulnerability to 549 era- and language-specific change. Identifier strings containing 550 linguistic fragments can create support difficulties down the road. 551 No matter how appropriate or even meaningless they are today, such 552 fragments may one day create confusion, give offense, or infringe on 553 a trademark as the semantic environment around us and our communities 554 evolves. 556 Names that look more or less like numbers avoid common problems that 557 defeat persistence and international acceptance. The use of digits 558 is highly recommended. Mixing in non-vowel alphabetic characters a 559 couple at a time is a relatively safe and easy way to achieve a 560 denser namespace (more possible names for a given length of the name 561 string). Such names have a chance of aging and traveling well. 562 Tools exists that mint, bind, and resolve opaque identifiers, with or 563 without check characters [NOID]. More on naming considerations is 564 given in a subsequent section. 566 2.5. The Qualifier Part 568 The part of the ARK following the NAA-assigned Name is an optional 569 Qualifier. It is a string that extends the base ARK in order to 570 create a kind of service entry point into the object named by the 571 NAA. At the discretion of the providing NMA, such a service entry 572 point permits an ARK to support access to individual hierarchical 573 components and subcomponents of an object, and to variants (versions, 574 languages, formats) of components. A Qualifier may be invented by 575 the NAA or by any NMA servicing the object. 577 In form, the Qualifier is a ComponentPath, or a VariantPath, or a 578 ComponentPath followed by a VariantPath. A VariantPath is introduced 579 and subdivided by the reserved character `.', and a ComponentPath is 580 introduced and subdivided by the reserved character `/'. In this 581 example, 583 http://example.org/ark:12025/654xz321/s3/f8.05v.tiff 585 the string "/s3/f8" is a ComponentPath and the string ".05v.tiff" is 586 a VariantPath. The ARK Qualifier is a formalization of some 587 currently mainstream URL syntax conventions. This formalization 588 specifically reserves meanings that permit recipients to make strong 589 inferences about logical sub-object containment and equivalence based 590 only on the form of the received identifiers; there is great 591 efficiency in not having to inspect metadata records to discover such 592 relationships. NMAs are free not to disclose any of these 593 relationships merely by avoiding the reserved characters above. 594 Hierarchical components and variants are discussed further in the 595 next two sections. 597 The Qualifier, if present, differs from the Name in several important 598 respects. First, a Qualifier may have been assigned either by the 599 NAA or later by the NMA. The assignment of a Qualifier by an NMA 600 effectively amounts to an act of publishing a service entry point 601 within the conceptual object originally named by the NAA. For our 602 purposes, an ARK extended with a Qualifier assigned by an NMA will be 603 called an NMA-qualified ARK. 605 Second, a Qualifier assignment on the part of an NMA is made in 606 fulfillment of its service obligations and may reflect changing 607 service expectations and technology requirements. NMA-qualified ARKs 608 could therefore be transient, even if the base, unqualified ARK is 609 persistent. For example, it would be reasonable for an NMA to 610 support access to an image object through an actionable ARK that is 611 considered persistent even if the experience of that access changes 612 as linking, labeling, and presentation conventions evolve and as 613 format and security standards are updated. For an image "thumbnail", 614 that NMA could also support an NMA-qualified ARK that is considered 615 impersistent because the thumbnail will be replaced with higher 616 resolution images as network bandwidth and CPU speeds increase. At 617 the same time, for an originally scanned, high-resolution master, the 618 NMA could publish an NMA-qualfied ARK that is itself considered 619 persistent. Of course, the NMA must be able to return its separate 620 commitments to unqualified, NAA-assigned ARKs, to NMA-qualified ARKs, 621 and to any NAA-qualified ARKs that it supports. 623 A third difference between a Qualifier and a Name concerns the 624 semantic opaqueness constraint. When an NMA-qualified ARK is to be 625 used as a transient service entry point into a persistent object, the 626 priority given to semantic opaqueness observed by the NAA in the Name 627 part may be relaxed by the NMA in the Qualifier part. If service 628 priorities in the Qualifier take precedence over persistence, short- 629 term usability considerations may recommend somewhat semantically 630 laden Qualifier strings. 632 Finally, not only is the set of Qualifiers supported by an NMA 633 mutable, but different NMAs may support different Qualifier sets for 634 the same NAA-identified object. In this regard the NMAs act 635 independently of each other and of the NAA. 637 The next two sections describe how ARK syntax may be used to declare, 638 or to avoid declaring, certain kinds of relatedness among qualified 639 ARKs. 641 2.5.1. ARKs that Reveal Object Hierarchy 643 An NAA or NMA may choose to reveal the presence of a hierarchical 644 relationship between objects using the `/' (slash) character after 645 the Name part of an ARK. Some authorities will choose not to 646 disclose this information, while others will go ahead and disclose so 647 that manipulators of large sets of ARKs can infer object 648 relationships by simple identifier inspection; for example, this 649 makes it possible for a system to present a collapsed view of a large 650 search result set. 652 If the ARK contains an internal slash after the NAAN, the piece to 653 its left indicates a containing object. For example, publishing an 654 ARK of the form, 656 ark:12025/654/xz/321 658 is equivalent to publishing three ARKs, 659 ark:12025/654/xz/321 660 ark:12025/654/xz 661 ark:12025/654 663 together with a declaration that the first object is contained in the 664 second object, and that the second object is contained in the third. 666 Revealing the presence of hierarchy is completely up to the assigner 667 (NMA or NAA). It is hard enough to commit to one object's name, let 668 alone to three objects' names and to a specific, ongoing relatedness 669 among them. Thus, regardless of whether hierarchy was present 670 initially, the assigner, by not using slashes, reveals no shared 671 inferences about hierarchical or other inter-relatedness in the 672 following ARKs: 674 ark:12025/654_xz_321 675 ark:12025/654_xz 676 ark:12025/654xz321 677 ark:12025/654xz 678 ark:12025/654 680 Note that slashes around the ARK's NAAN (/12025/ in these examples) 681 are not part of the ARK's Name and therefore do not indicate the 682 existence of some sort of NAAN super object containing all objects in 683 its namespace. A slash must have at least one non-structural 684 character (one that is neither a slash nor a period) on both sides in 685 order for it to separate recognizable structural components. So 686 initial or final slashes may be removed, and double slashes may be 687 converted into single slashes. 689 2.5.2. ARKs that Reveal Object Variants 691 An NAA or NMA may choose to reveal the possible presence of variant 692 objects or object components using the `.' (period) character after 693 the Name part of an ARK. Some authorities will choose not to 694 disclose this information, while others will go ahead and disclose so 695 that manipulators of large sets of ARKs can infer object 696 relationships by simple identifier inspection; for example, this 697 makes it possible for a system to present a collapsed view of a large 698 search result set. 700 If the ARK contains an internal period after Name, the piece to its 701 left is a root name and the piece to its right, and up to the end of 702 the ARK or to the next period is a suffix. A Name may have more than 703 one suffix, for example, 704 ark:12025/654.24 705 ark:12025/xz4/654.24 706 ark:12025/654.20v.78g.f55 708 There are two main rules. First, if two ARKs share the same root 709 name but have different suffixes, the corresponding objects were 710 considered variants of each other (different formats, languages, 711 versions, etc.) by the assigner (NMA or NAA). Thus, the following 712 ARKs are variants of each other: 714 ark:12025/654.20v.78g.f55 715 ark:12025/654.321xz 716 ark:12025/654.44 718 Second, publishing an ARK with a suffix implies the existence of at 719 least one variant identified by the ARK without its suffix. The ARK 720 otherwise permits no further assumptions about what variants might 721 exist. So publishing the ARK, 723 ark:12025/654.20v.78g.f55 725 is equivalent to publishing the four ARKs, 727 ark:12025/654.20v.78g.f55 728 ark:12025/654.20v.78g 729 ark:12025/654.20v 730 ark:12025/654 732 Revealing the possibility of variants is completely up to the 733 assigner. It is hard enough to commit to one object's name, let 734 alone to multiple variants' names and to a specific, ongoing 735 relatedness among them. The assigner is the sole arbiter of what 736 constitutes a variant within its namespace, and whether to reveal 737 that kind of relatedness by using periods within its names. 739 A period must have at least one non-structural character (one that is 740 neither a slash nor a period) on both sides in order for it to 741 separate recognizable structural components. So initial or final 742 periods may be removed, and adjacent periods may be converted into a 743 single period. Multiple suffixes should be arranged in sorted order 744 (pure ASCII collating sequence) at the end of an ARK. 746 2.6. Character Repertoires 748 The Name and Qualifier parts are strings of visible ASCII characters. 749 For received ARKs, implementations must support a minimum length of 750 255 octets for the string composed of the Base ARK plus Qualifier. 751 Implementations generating strings exceeding this length should 752 understand that receiving implementations may not be able to index 753 such ARKs properly. Characters may be letters, digits, or any of 754 these seven characters: 756 = ~ * + @ _ $ 758 The following characters may also be used, but their meanings are 759 reserved: 761 % - . / 763 The characters `/' and `.' are ignored if either appears as the last 764 character of an ARK. If used internally, they allow a name assigner 765 to reveal object hierarchy and object variants as previously 766 described. 768 Hyphens are considered to be insignificant and are always ignored in 769 ARKs. A `-' (hyphen) may appear in an ARK for readability, or it may 770 have crept in during the formatting and wrapping of text, but it must 771 be ignored in lexical comparisons. As in a telephone number, hyphens 772 have no meaning in an ARK. It is always safe for an NMA that 773 receives an ARK to remove any hyphens found in it. As a result, like 774 the NMAH, hyphens are "identity inert" in comparing ARKs for 775 equivalence. For example, the following ARKs are equivalent for 776 purposes of comparison and ARK service access: 778 ark:12025/65-4-xz-321 779 http://sneezy.dopey.com/ark:12025/654--xz32-1 780 ark:12025/654xz321 782 The `%' character is reserved for %-encoding all other octets that 783 would appear in the ARK string, in the same manner as for URIs 784 [RFC3986]. A %-encoded octet consists of a `%' followed by two hex 785 digits; for example, "%7d" stands in for `}'. Lower case hex digits 786 are preferred to reduce the chances of false acronym recognition; 787 thus it is better to use "%acT" instead of "%ACT". The character `%' 788 itself must be represented using "%25". As with URNs, %-encoding 789 permits ARKs to support legacy namespaces (e.g., ISBN, ISSN, SICI) 790 that have less restricted character repertoires [RFC2288]. 792 2.7. Normalization and Lexical Equivalence 794 To determine if two or more ARKs identify the same object, the ARKs 795 are compared for lexical equivalence after first being normalized. 796 Since ARK strings may appear in various forms (e.g., having different 797 NMAHs), normalizing them minimizes the chances that comparing two ARK 798 strings for equality will fail unless they actually identify 799 different objects. In a specified-host ARK (one having an NMAH), the 800 NMAH never participates in such comparisons. Normalization described 801 here serves to define lexical equivalence but does not restrict how 802 implementors normalize ARKs locally for storage. 804 Normalization of a received ARK for the purpose of octet-by-octet 805 equality comparison with another ARK consists of several steps. 806 First, the NMAH part (eg, everything from an initial "http://" up to 807 the next slash), if present is removed. Second, any URI query string 808 is removed (everything from the first literal '?' to the end of the 809 string). Third, the first case-insensitive match on "ark:/" or 810 "ark:" is converted to "ark:" (replacing any upper case letters and 811 removing any terminal '/'). Fourth, in the string that remains, the 812 two characters following every occurrence of `%' are converted to 813 lower case. The case of all other letters in the ARK string must be 814 preserved. Fifth, all hyphens are removed. 816 Sixth, if normalization is being done as part of a resolution step, 817 and if the end of the remaining string matches a known inflection, 818 the inflection is noted and removed. Seventh, structural characters 819 (slash and period) are normalized: initial and final occurrences are 820 removed, and two structural characters in a row (e.g., // or ./) are 821 replaced by the first character, iterating until each occurrence has 822 at least one non-structural character on either side. Finally, if 823 there are any components with a period on the left and a slash on the 824 right, either the component and the preceding period must be moved to 825 the end of the Name part or the ARK must be thrown out as malformed. 827 The fourth and final step is to arrange the suffixes in ASCII 828 collating sequence (that is, to sort them) and to remove duplicate 829 suffixes, if any. It is also permissible to throw out ARKs for which 830 the suffixes are not sorted. 832 The resulting ARK string is now normalized. Comparisons between 833 normalized ARKs are case-sensitive, meaning that upper case letters 834 are considered different from their lower case counterparts. 836 To keep ARK string variation to a minimum, no reserved ARK characters 837 should be %-encoded unless it is deliberately to conceal their 838 reserved meanings. No non-reserved ARK characters should ever be 839 %-encoded. Finally, no %-encoded character should ever appear in an 840 ARK in its decoded form. 842 3. Naming Considerations 844 The most important threats faced by persistence providers include 845 such things as funding loss, natural disaster, political and social 846 upheaval, processing faults, and errors in human oversight. There is 847 nothing that an identifer scheme can do about such things. Still, a 848 few observed identifier failures and inconveniences can be traced 849 back to naming practices that we now know to be less than optimal for 850 persistence. 852 3.1. ARKS Embedded in Language 854 The ARK has different goals from the URI, so it has different 855 character set requirements. Because linguistic constructs imperil 856 persistence, for ARKs non-ASCII character support is unimportant. 857 ARKs and URIs share goals of transcribability and transportability 858 within web documents, so characters are required to be visible, non- 859 conflicting with HTML/XML syntax, and not subject to tampering during 860 transmission across common transport gateways. Add the goal of 861 making an undelimited ARK recognizable in running prose, as in 862 ark:12025/=@_22*$, and certain punctuation characters (e.g., comma, 863 period) end up being excluded from the ARK lest the end of a phrase 864 or sentence be mistaken for part of the ARK. 866 This consideration has more direct effect on ARK usability in a 867 natural language context than it has on ARK persistence. The same is 868 true of the rule preventing hyphens from having lexical significance. 869 It is fine to publish ARKs with hyphens in them (e.g., such as the 870 output of UUID/GUID generators), but the uniform treatment of hyphens 871 as insignificant reduces the possibility of users transcribing 872 identifiers that will have been broken through unpredictable 873 hyphenation by word processors. Any measure that reduces user 874 irritation with an identifier will increase its chances of survival. 876 3.2. Objects Should Wear Their Identifiers 878 A valuable technique for provision of persistent objects is to try to 879 arrange for the complete identifier to appear on, with, or near its 880 retrieved object. An object encountered at a moment in time when its 881 discovery context has long since disappeared could then easily be 882 traced back to its metadata, to alternate versions, to updates, etc. 883 This has seen reasonable success, for example, in book publishing and 884 software distribution. An identifier string only has meaning when 885 its association is known, and this a very sure, simple, and low-tech 886 method of reminding everyone exactly what that association is. 888 3.3. Names are Political, not Technological 890 If persistence is the goal, a deliberate local strategy for 891 systematic name assignment is crucial. Names must be chosen with 892 great care. Poorly chosen and managed names will devastate any 893 persistence strategy, and they do not discriminate by identifier 894 scheme. Whether a mistakenly re-assigned name is a URN, DOI, PURL, 895 URL, or ARK, the damage -- failed access and confusion -- is not 896 mitigated more in one scheme than in another. Conversely, in-house 897 efforts to manage names responsibly will go much further towards 898 safeguarding persistence than any choice of naming scheme or name 899 resolution technology. 901 Branding (e.g., at the corporate or departmental level) is important 902 for funding and visibility, but substrings representing brands and 903 organizational names should be given a wide berth except when 904 absolutely necessary in the hostname (the identity-inert) part of the 905 ARK. These substrings are not only unstable because organizations 906 change frequently, but they are also dangerous because successor 907 organizations often have political or legal reasons to actively 908 suppress predecessor names and brands. Any measure that reduces the 909 chances of future political or legal pressure on an identifier will 910 decrease the chances that our descendants will be obliged to 911 deliberately break it. 913 3.4. Choosing a Hostname or NMA 915 Hostnames appearing in any identifier meant to be persistent must be 916 chosen with extra care. The tendency in hostname selection has 917 traditionally been to choose a token with recognizable attributes, 918 such as a corporate brand, but that tendency wreaks havoc with 919 persistence that is supposed to outlive brands, corporations, subject 920 classifications, and natural language semantics (e.g., what did the 921 three letters "gay" mean in 1958, 1978, and 1998?). Today's 922 recognized and correct attributes are tomorrow's stale or incorrect 923 attributes. In making hostnames (any names, actually) long-term 924 persistent, it helps to eliminate recognizable attributes to the 925 extent possible. This affects selection of any name based on URLs, 926 including PURLs and the explicitly disposable NMAHs. 928 There is no excuse for a provider that manages its internal names 929 impeccably not to exercise the same care in choosing what could be an 930 exceptionally durable hostname, especially if it would form the 931 prefix for all the provider's URL-based external names. Registering 932 an opaque hostname in the ".org" or ".net" domain would not be a bad 933 start. Another way is to publish your ARKs with an organizational 934 domain name that will be mapped by DNS to an appropriate NMA host. 935 This makes for shorter names with less branding vulnerability. 937 It is a mistake to think that hostnames are inherently unstable. If 938 you require brand visibility, that may be a fact of life. But things 939 are easier if yours is the brand of long-lived cultural memory 940 institution such as a national or university library or archive. 941 Well-chosen hostnames from organizations that are sheltered from the 942 direct effects of a volatile marketplace can easily provide longer- 943 lived global resolvers than the domain names explicitly or implicitly 944 used as starting points for global resolution by indirection-based 945 persistent identifier schemes. For example, it is hard to imagine 946 circumstances under which the Library of Congress' domain name would 947 disappear sooner than, say, "handle.net". 949 For smaller libraries, archives, and preservation organizations, 950 there is a natural concern about whether they will be able to keep 951 their web servers and domain names in the face of uncertain funding. 952 One option is to form or join a consortium [N2T] of like-minded 953 organizations with the purpose of providing mutual preservation 954 support. The first goal of such a consortium would be to perpetually 955 rent a hostname on which to establish a web server that simply 956 redirects incoming member organization requests to the appropriate 957 member server; using ARKs, for example, a 150-member consortium could 958 run a very small server (24x7) that contained nothing more than 150 959 rewrite rules in its configuration file. Even more helpful would be 960 additional consortial support for a member organization that was 961 unable to continue providing services and needed to find a successor 962 archival organization. This would be a low-cost, low-tech way to 963 publish ARKs (or URLs) under highly persistent hostnames. 965 There are no obvious reasons why the organizations registering DNS 966 names, URN Namespaces, and DOI publisher IDs should have among them 967 one that is intrinsically more fallible than the next. Moreover, it 968 is a misconception that the demise of DNS and of HTTP need adversely 969 affect the persistence of URLs. At such a time, certainly URLs from 970 the present day might not then be actionable by our present-day 971 mechanisms, but resolution systems for future non-actionable URLs are 972 no harder to imagine than resolution systems for present-day non- 973 actionable URNs and DOIs. There is no more stable a namespace than 974 one that is dead and frozen, and that would then characterize the 975 space of names bearing the "http://" prefix. It is useful to 976 remember that just because hostnames have been carelessly chosen in 977 their brief history does not mean that they are unsuitable in NMAHs 978 (and URLs) intended for use in situations demanding the highest level 979 of persistence available in the Internet environment. A well-planned 980 name assignment strategy is everything. 982 3.5. Assigners of ARKs 984 A Name Assigning Authority (NAA) is an organization that creates (or 985 delegates creation of) long-term associations between identifiers and 986 information objects. Examples of NAAs include national libraries, 987 national archives, and publishers. An NAA may arrange with an 988 external organization for identifier assignment. The US Library of 989 Congress, for example, allows OCLC (the Online Computer Library 990 Center, a major world cataloger of books) to create associations 991 between Library of Congress call numbers (LCCNs) and the books that 992 OCLC processes. A cataloging record is generated that testifies to 993 each association, and the identifier is included by the publisher, 994 for example, in the front matter of a book. 996 An NAA does not so much create an identifier as create an 997 association. The NAA first draws an unused identifier string from 998 its namespace, which is the set of all identifiers under its control. 999 It then records the assignment of the identifier to an information 1000 object having sundry witnessed characteristics, such as a particular 1001 author and modification date. A namespace is usually reserved for an 1002 NAA by agreement with recognized community organizations (such as 1003 IANA and ISO) that all names containing a particular string be under 1004 its control. In the ARK an NAA is represented by the Name Assigning 1005 Authority Number (NAAN). 1007 The ARK namespace reserved for an NAA is the set of names bearing its 1008 particular NAAN. For example, all strings beginning with 1009 "ark:12025/" are under control of the NAA registered under 12025, 1010 which might be the National Library of Finland. Because each NAA has 1011 a different NAAN, names from one namespace cannot conflict with those 1012 from another. Each NAA is free to assign names from its namespace 1013 (or delegate assignment) according to its own policies. These 1014 policies must be documented in a manner similar to the declarations 1015 required for URN Namespace registration [RFC2611]. 1017 To register for a NAAN, please read about the mapping authority 1018 discovery file in the next section and send email to ark@cdlib.org. 1020 3.6. NAAN Namespace Management 1022 Every NAA must have a namespace management strategy. A time-honored 1023 technique is to hierarchically partition a namespace into 1024 subnamespaces using prefixes that guarantee non-collision of names in 1025 different partition. This practice is strongly encouraged for all 1026 NAAs, especially when subnamespace management will be delegated to 1027 other departments, units, or projects within an organization. For 1028 example, with a NAAN that is assigned to a university and managed by 1029 its main library, care should be taken to reserve semantically opaque 1030 prefixes that will set aside large parts of the unused namespace for 1031 future assignments. Prefix-based partition management is an 1032 important responsibility of the NAA. 1034 This sort of delegation by prefix is well-used in the formation of 1035 DNS names and ISBN identifiers. An important difference is that in 1036 the former, the hierarchy is deliberately exposed and in the latter 1037 it is hidden. Rather than using lexical boundary markers such as the 1038 period (`.') found in domain names, the ISBN uses a publisher prefix 1039 but doesn't disclose where the prefix ends and the publisher's 1040 assigned name begins. This practice of non-disclosure, borrowed from 1041 the ISBN and ISSN schemes, is encouraged in assigning ARKs, because 1042 it reduces the visibility of an assertion that is probably not 1043 important now and may become a vulnerability later. 1045 Reasonable prefixes for assigned names usually consist of consonants 1046 and digits and are 1-5 characters in length. For example, the 1047 constant prefix "x9t" might be delegated to a book digitization 1048 project that creates identifiers such as 1050 http://444.berkeley.edu/ark:28722/x9t38rk45c 1052 If longevity is the goal, it is important to keep the prefixes free 1053 of recognizable semantics; for example, using an acronym representing 1054 a project or a department is discouraged. At the same time, you may 1055 wish to set aside a subnamespace for testing purposes under a prefix 1056 such as "fk..." that can serve as a visual clue and reminder to 1057 maintenance staff that this "fake" identifier was never published. 1059 There are other measures one can take to avoid user confusion, 1060 transcription errors, and the appearance of accidental semantics when 1061 creating identifiers. If you are generating identifiers 1062 automatically, pure numeric identifiers are likeley to be 1063 semantically opaque enough, but it's probably useful to avoid leading 1064 zeroes because some users mistakenly treat them as optional, thinking 1065 (arithmetically) that they don't contribute to the "value" of the 1066 identifier. 1068 If you need lots of identifiers and you don't want them to get too 1069 long, you can mix digits with consonants (but avoid vowels since they 1070 might accidentally spell words) to get more identifiers without 1071 increasing the string length. In this case you may not want more 1072 than a two letters in a row because it reduces the chance of 1073 generating acronyms. Generator tools such as [NOID] provide support 1074 for these sorts of identifiers, and can also add a computed check 1075 character as a guarantee against the most common transcription 1076 errors. 1078 3.7. Sub-Object Naming 1080 As mentioned previously, semantically opaque identifiers are very 1081 useful for long-term naming of abstract objects, however, it may be 1082 appropriate to extend these names with less opaque extensions that 1083 reference contemporary service entry points (sub-objects) in support 1084 of the object. Sub-object extensions beginning with a digit or 1085 underscore (`_') are reserved for the possibilty of developing a 1086 future registry of canonical service points (e.g., numeric references 1087 to versions, formats, languages, etc). 1089 4. Finding a Name Mapping Authority 1091 In order to derive an actionable identifier (these days, a URL) from 1092 an ARK, a hostport (hostname or hostname plus port combination) for a 1093 working Name Mapping Authority (NMA) must be found. An NMA is a 1094 service that is able to respond to the three basic ARK service 1095 requests. Relying on registration and client-side discovery, NMAs 1096 make known which NAAs' identifiers they are willing to service. 1098 Upon encountering an ARK, a user (or client software) looks inside it 1099 for the optional NMAH part (the hostport of the NMA's ARK service). 1100 If it contains an NMAH that is working, this NMAH discovery step may 1101 be skipped; the NMAH effectively uses the beginning of an ARK to 1102 cache the results of a prior mapping authority discovery process. If 1103 a new NMAH needs to found, the client looks inside the ARK again for 1104 the NAAN (Name Assigning Authority Number). Querying a global 1105 database, it then uses the NAAN to look up all current NMAHs that 1106 service ARKs issued by the identified NAA. 1108 The global database is key, and ideally the lookup would be automatic 1109 and transparent to the user. For this, the most promising method is 1110 probably the Name-to-Thing (N2T) Resolver [N2T] at n2t.net. It is a 1111 proposed low-cost, highly reliable, consortially maintained NMAH that 1112 simply exists to support actionable HTTP-based URLs for as long as 1113 HTTP is used. One of its big advantages over the other two methods 1114 and the URN, Handle, DOI, and PURL methods, is that N2T addresses the 1115 namespace splitting problem. When objects maintained by one NMA are 1116 inherited by more than one successor NMA, until now one of those 1117 successors would be required to maintain forwarding tables on behalf 1118 of the other successors. 1120 There are two other ways to discover an NMAH, one of them described 1121 in a subsection below. Another way, described in an appendix, is 1122 based on a simplification of the URN resolver discovery method, 1123 itself very similar in principle to the resolver discovery method 1124 used by Handles and DOIs. None of these methods does more than what 1125 can be done with a very small, consortially maintained web server 1126 such as [N2T]. 1128 In the interests of long-term persistence, however, ARK mechanisms 1129 are first defined in high-level, protocol-independent terms so that 1130 mechanisms may evolve and be replaced over time without compromising 1131 fundamental service objectives. Either or both specific methods 1132 given here may eventually be supplanted by better methods since, by 1133 design, the ARK scheme does not depend on a particular method, but 1134 only on having some method to locate an active NMAH. 1136 At the time of issuance, at least one NMAH for an ARK should be 1137 prepared to service it. That NMA may or may not be administered by 1138 the Name Assigning Authority (NAA) that created it. Consider the 1139 following hypothetical example of providing long-term access to a 1140 cancer research journal. The publisher wishes to turn a profit and 1141 the National Library of Medicine wishes to preserve the scholarly 1142 record. An agreement might be struck whereby the publisher would act 1143 as the NAA and the national library would archive the journal issue 1144 when it appears, but without providing direct access for the first 1145 six months. During the first six months of peak commercial 1146 viability, the publisher would retain exclusive delivery rights and 1147 would charge access fees. Again, by agreement, both the library and 1148 the publisher would act as NMAs, but during that initial period the 1149 library would redirect requests for issues less than six months old 1150 to the publisher. At the end of the waiting period, the library 1151 would then begin servicing requests for issues older than six months 1152 by tapping directly into its own archives. Meanwhile, the publisher 1153 might routinely redirect incoming requests for older issues to the 1154 library. Long-term access is thereby preserved, and so is the 1155 commercial incentive to publish content. 1157 Although it will be common for an NAA also to run an NMA service, it 1158 is never a requirement. Over time NAAs and NMAs will come and go. 1159 One NMA will succeed another, and there might be many NMAs serving 1160 the same ARKs simultaneously (e.g., as mirrors or as competitors). 1161 There might also be asymmetric but coordinated NMAs as in the 1162 library-publisher example above. 1164 4.1. Looking Up NMAHs in a Globally Accessible File 1166 This subsection describes a way to look up NMAHs using a simple name 1167 authority table represented as a plain text file. For efficient 1168 access the file may be stored in a local filesystem, but it needs to 1169 be reloaded periodically to incorporate updates. It is not expected 1170 that the size of the file or frequency of update should impose an 1171 undue maintenance or searching burden any time soon, for even 1172 primitive linear search of a file with ten-thousand NAAs is a 1173 subsecond operation on modern server machines. The proposed file 1174 strategy is similar to the /etc/hosts file strategy that supported 1175 Internet host address lookup for a period of years before the advent 1176 of DNS. 1178 The name authority table file is updated on an ongoing basis and is 1179 available for copying over the internet from the California Digital 1180 Library at http://www.cdlib.org/inside/diglib/ark/natab and from a 1181 number of mirror sites. The file contains comment lines (lines that 1182 begin with `#') explaining the format and giving the file's 1183 modification time, reloading address, and NAA registration 1184 instructions. There is even a Perl script that processes the file 1185 embedded in the file's comments. The currently registered Name 1186 Assigning Authorities are: 1188 12025 National Library of Medicine 1189 12026 Library of Congress 1190 12027 National Agriculture Library 1191 13030 California Digital Library 1192 13038 World Intellectual Property Organization 1193 20775 University of California San Diego 1194 29114 University of California San Francisco 1195 28722 University of California Berkeley 1196 21198 University of California Los Angeles 1197 15230 Rutgers University 1198 13960 Internet Archive 1199 64269 Digital Curation Centre 1200 62624 New York University 1201 67531 University of North Texas 1202 27927 Ithaka Electronic-Archiving Initiative 1203 12148 Bibliotheque nationale de France 1204 / National Library of France 1205 78319 Google 1206 88435 Princeton University 1207 78428 University of Washington 1208 89901 Archives of the Region of Vaestra Goetaland 1209 and City of Gothenburg, Sweden 1210 80444 Northwest Digital Archives 1211 25593 Emory University 1212 25031 University of Kansas 1213 17101 Centre for Ecology & Hydrology, UK 1214 65323 University of Calgary 1215 61001 University of Chicago 1216 52327 Bibliotheque et Archives Nationales du Quebec 1217 / National Libary and Archives of Quebec 1218 39331 National Szechenyi Library / National Library of Hungary 1219 26677 Library and Archives Canada / Bibliotheque et Archives Canada 1221 5. Generic ARK Service Definition 1223 An ARK request's output is delivered information; examples include 1224 the object itself, a policy declaration (e.g., a promise of support), 1225 a descriptive metadata record, or an error message. The experience 1226 of object delivery is expected to be an evolving mix of information 1227 that reflects changing service expectations and technology 1228 requirements; contemporary examples include such things as an object 1229 summary and component links formatted for human consumption. ARK 1230 services must be couched in high-level, protocol-independent terms if 1231 persistence is to outlive today's networking infrastructural 1232 assumptions. The high-level ARK service definitions listed below are 1233 followed in the next section by a concrete method (one of many 1234 possible methods) for delivering these services with today's 1235 technology. 1237 5.1. Generic ARK Access Service (access, location) 1239 Returns (a copy of) the object or a redirect to the same, although a 1240 sensible object proxy may be substituted. Examples of sensible 1241 substitutes include, 1243 o a table of contents instead of a large complex document, 1245 o a home page instead of an entire web site hierarchy, 1247 o a rights clearance challenge before accessing protected data, 1249 o directions for access to an offline object (e.g., a book), 1251 o a description of an intangible object (a disease, an event), or 1253 o an applet acting as "player" for a large multimedia object. 1255 May also return a discriminated list of alternate object locators. 1256 If access is denied, returns an explanation of the object's current 1257 (perhaps permanent) inaccessibility. 1259 5.1.1. Generic Policy Service (permanence, naming, etc.) 1261 Returns declarations of policy and support commitments for given 1262 ARKs. Declarations are returned in either a structured metadata 1263 format or a human readable text format; sometimes one format may 1264 serve both purposes. Policy subareas may be addressed in separate 1265 requests, but the following areas should should be covered: object 1266 permanence, object naming, object fragment addressing, and 1267 operational service support. 1269 The permanence declaration for an object is a rating defined with 1270 respect to an identified permanence provider (guarantor), which will 1271 be the NMA. It may include the following aspects. 1273 (a) "object availability" -- whether and how access to the object 1274 is supported (e.g., online 24x7, or offline only), 1276 (b) "identifier validity" -- under what conditions the identifier 1277 will be or has been re-assigned, 1278 (c) "content invariance" -- under what conditions the content of 1279 the object is subject to change, and 1281 (d) "change history" -- access to corrections, migrations, and 1282 revisions, whether through links to the changed objects themselves 1283 or through a document summarizing the change history 1285 A recent approach to persistence statements, conceived independently 1286 from ARKs, can be found at [PStatements], with ongoing work available 1287 at Appendix A.. An older approach to a permanence rating framework 1288 is given in [NLMPerm], which identified the following "permanence 1289 levels": 1291 Not Guaranteed: No commitment has been made to retain this 1292 resource. It could become unavailable at any time. Its 1293 identifier could be changed. 1295 Permanent: Dynamic Content: A commitment has been made to keep 1296 this resource permanently available. Its identifier will always 1297 provide access to the resource. Its content could be revised or 1298 replaced. 1300 Permanent: Stable Content: A commitment has been made to keep this 1301 resource permanently available. Its identifier will always 1302 provide access to the resource. Its content is subject only to 1303 minor corrections or additions. 1305 Permanent: Unchanging Content: A commitment has been made to keep 1306 this resource permanently available. Its identifier will always 1307 provide access to the resource. Its content will not change. 1309 Naming policy for an object includes an historical description of the 1310 NAA's (and its successor NAA's) policies regarding differentiation of 1311 objects. Since it the NMA who responds to requests for policy 1312 statements, it is useful for the NMA to be able to produce or 1313 summarize these historical NAA documents. Naming policy may include 1314 the following aspects. 1316 (i) "similarity" -- (or "unity") the limit, defined by the NAA, to 1317 the level of dissimilarity beyond which two similar objects 1318 warrant separate identifiers but before which they share one 1319 single identifier, and 1321 (ii) "granularity" -- the limit, defined by the NAA, to the level 1322 of object subdivision beyond which sub-objects do not warrant 1323 separately assigned identifiers but before which sub-objects are 1324 assigned separate identifiers. 1326 Subnaming policy for an object describes the qualifiers that the NMA, 1327 in fulfilling its ongoing and evolving service obligations, allows as 1328 extensions to an NAA-assigned ARK. To the conceptual object that the 1329 NAA named with an ARK, the NMA may add component access points and 1330 derivatives (e.g., format migrations in aid of preservation) in order 1331 to provide both basic and value-added services. 1333 Addressing policy for an object includes a description of how, during 1334 access, object components (e.g., paragraphs, sections) or views 1335 (e.g., image conversions) may or may not be "addressed", in other 1336 words, how the NMA permits arguments or parameters to modify the 1337 object delivered as the result of an ARK request. If supported, 1338 these sorts of operations would provide things like byte-ranged 1339 fragment delivery and open-ended format conversions, or any set of 1340 possible transformations that would be too numerous to list or to 1341 identify with separately assigned ARKs. 1343 Operational service support policy includes a description of general 1344 operational aspects of the NMA service, such as after-hours staffing 1345 and trouble reporting procedures. 1347 5.1.2. Generic Description Service 1349 Returns a description of the object. Descriptions are returned in a 1350 structured metadata format, human readable text format, or in one 1351 format that serves both purposes (such as human-readable HTML with 1352 embedded machine-readable metadata). A description must at a minimum 1353 answer the who, what, when, and where questions concerning an 1354 expression of the object. Standalone descriptions should be 1355 accompanied by the modification date and source of the description 1356 itself. May also return discriminated lists of ARKs that are related 1357 to the given ARK. 1359 5.2. Overview of The HTTP URL Mapping Protocol (THUMP) 1361 The HTTP URL Mapping Protocol (THUMP) is a way of taking a key (any 1362 identifier) and asking such questions as, what information does this 1363 identify and how permanent is it? [THUMP] is in fact one specific 1364 method under development for delivering ARK services. The protocol 1365 runs over HTTP to exploit the web browser's current pre-eminence as 1366 user interface to the Internet. THUMP is designed so that a person 1367 can enter ARK requests directly into the location field of current 1368 browser interfaces. Because it runs over HTTP, THUMP can be 1369 simulated and tested via keyboard-based interactions [RFC0854]. 1371 The asker (a person or client program) starts with an identifier, 1372 such as an ARK or a URL. The identifier reveals to the asker (or 1373 allows the asker to infer) the Internet host name and port number of 1374 a server system that responds to questions. Here, this is just the 1375 NMAH that is obtained by inspection and possibly lookup based on the 1376 ARK's NAAN. The asker then sets up an HTTP session with the server 1377 system, sends a question via a THUMP request (contained within an 1378 HTTP request), receives an answer via a THUMP response (contained 1379 within an HTTP response), and closes the session. That concludes the 1380 connected portion of the protocol. 1382 A THUMP request is a string of characters beginning with a `?' 1383 (question mark) that is appended to the identifier string. The 1384 resulting string is sent as an argument to HTTP's GET command. 1385 Request strings too long for GET may be sent using HTTP's POST 1386 command. The three most common requests correspond to three 1387 degenerate special cases that keep the user's learning and typing 1388 burden low. First, a simple key with no request at all is the same 1389 as an ordinary access request. Thus a plain ARK entered into a 1390 browser's location field behaves much like a plain URL, and returns 1391 access to the primary identified object, for instance, an HTML 1392 document. 1394 The second special case is a minimal ARK description request string 1395 consisting of just "?". For example, entering the string, 1397 ark.nlm.nih.gov/12025/psbbantu? 1399 into the browser's location field directly precipitates a request for 1400 a metadata record describing the object identified by ark:12025/ 1401 psbbantu. The browser, unaware of THUMP, prepares and sends an HTTP 1402 GET request in the same manner as for a URL. THUMP is designed so 1403 that the response (indicated by the returned HTTP content type) is 1404 normally displayed, whether the output is structured for machine 1405 processing (text/plain) or formatted for human consumption (text/ 1406 html). 1408 In the following example THUMP session, each line has been annotated 1409 to include a line number and whether it was the client or server that 1410 sent it. Without going into much depth, the session has four pieces 1411 separated from each other by blank lines: the client's piece (lines 1412 1-3), the server's HTTP/THUMP response headers (4-7), and the body of 1413 the server's response (8-13). The first and last lines (1 and 13) 1414 correspond to the client's steps to start the TCP session and the 1415 server's steps to end it, respectively. 1417 1 C: [opens session] 1418 C: GET http://ark.nlm.nih.gov/ark:12025/psbbantu? HTTP/1.1 1419 C: 1420 S: HTTP/1.1 200 OK 1421 5 S: Content-Type: text/plain 1422 S: THUMP-Status: 0.6 200 OK 1423 S: 1424 S: erc: 1425 S: who: Lederberg, Joshua 1426 10 S: what: Studies of Human Families for Genetic Linkage 1427 S: when: 1974 1428 S: where: http://profiles.nlm.nih.gov/BB/A/N/T/U/_/bbantu.pdf 1429 S: [closes session] 1431 The first two server response lines (4-5) above are typical of HTTP. 1432 The next line (6) is peculiar to THUMP, and indicates the THUMP 1433 version and a normal return status. 1435 The balance of the response consists of a single metadata record 1436 (8-12) that comprises the ARK description service response. The 1437 returned record is in the format of an Electronic Resource Citation 1438 [ERC], which is discussed in overview in the next section. For now, 1439 note that it contains four elements that answer the top priority 1440 questions regarding an expression of the object: who played a major 1441 role in expressing it, what the expression was called, when is was 1442 created, and where the expression may be found. This quartet of 1443 elements comes up again and again in ERCs. 1445 The third degenerate special case of an ARK request (and no other 1446 cases will be described in this document) is the string "??", 1447 corresponding to a minimal permanence policy request. It can be seen 1448 in use appended to an ARK (on line 2) in the example session that 1449 follows. 1451 1 C: [opens session] 1452 C: GET http://ark.nlm.nih.gov/ark:12025/psbbantu?? HTTP/1.1 1453 C: 1454 S: HTTP/1.1 200 OK 1455 5 S: Content-Type: text/plain 1456 S: THUMP-Status: 0.6 200 OK 1457 S: 1458 S: erc: 1459 S: who: Lederberg, Joshua 1460 10 S: what: Studies of Human Families for Genetic Linkage 1461 S: when: 1974 1462 S: where: http://profiles.nlm.nih.gov/BB/A/N/T/U/_/bbantu.pdf 1463 S: erc-support: 1464 S: who: USNLM 1465 15 S: what: Permanent, Unchanging Content 1466 S: when: 20010421 1467 S: where: http://ark.nlm.nih.gov/yy22948 1468 S: [closes session] 1470 Each segment in an ERC tells a different story relating to the 1471 object, so although the same four questions (elements) appear in 1472 each, the answers depend on the segment's story type. While the 1473 first segment tells the story of an expression of the object, the 1474 second segment tells the story of the support commitment made to it: 1475 who made the commitment, what the nature of the commitment was, when 1476 it was made, and where a fuller explanation of the commitment may be 1477 found. 1479 5.3. The Electronic Resource Citation (ERC) 1481 An Electronic Resource Citation (or ERC, pronounced e-r-c) [ERC] is a 1482 kind of object description that uses Dublin Core Kernel metadata 1483 elements [DCKernel]. The ERC with Kernel elements provides a simple, 1484 compact, and printable record for holding data associated with an 1485 information resource. As originally designed [Kernel], Kernel 1486 metadata balances the needs for expressive power, very simple machine 1487 processing, and direct human manipulation. 1489 The previous section shows two limited examples of what is fully 1490 described elsewhere [ERC]. The rest of this short section provides 1491 some of the background and rationale for this record format. 1493 A founding principle of Kernel metadata is that direct human contact 1494 with metadata will be a necessary and sufficient condition for the 1495 near term rapid development of metadata standards, systems, and 1496 services. Thus the machine-processable Kernel elements must only 1497 minimally strain people's ability to read, understand, change, and 1498 transmit ERCs without their relying on intermediation with 1499 specialized software tools. The basic ERC needs to be succinct, 1500 transparent, and trivially parseable by software. 1502 In the current Internet, it is natural seriously to consider using 1503 XML as an exchange format because of predictions that it will obviate 1504 many ad hoc formats and programs, and unify much of the world's 1505 information under one reliable data structuring discipline that is 1506 easy to generate, verify, parse, and render. It appears, however, 1507 that XML is still only catching on after years of standards work and 1508 implementation experience. The reasons for it are unclear, but for 1509 now very simple XML interpretation is still out of reach. Another 1510 important caution is that XML structures are hard on the eyeballs, 1511 taking up an amount of display (and page) space that significantly 1512 exceeds that of traditional formats. Until these conflicts with ERC 1513 principle are resolved, XML is not a first choice for representing 1514 ERCs. Borrowing instead from the data structuring format that 1515 underlies the successful spread of email and web services, the first 1516 ERC format uses [ANVL], which is based on email and HTTP headers 1517 [RFC2822]. There is a naturalness to ANVL's label-colon-value format 1518 (seen in the previous section) that barely needs explanation to a 1519 person beginning to enter ERC metadata. 1521 Besides simplicity of ERC system implementation and data entry 1522 mechanics, ERC semantics (what the record and its constituent parts 1523 mean) must also be easy to explain. ERC semantics are based on a 1524 reformulation and extension of the Dublin Core [RFC5013] hypothesis, 1525 which suggests that the fifteen Dublin Core metadata elements have a 1526 key role to play in cross-domain resource description. The ERC 1527 design recognizes that the Dublin Core's primary contribution is the 1528 international, interdisciplinary consensus that identified fifteen 1529 semantic buckets (element categories), regardless of how they are 1530 labeled. The ERC then adds a definition for a record and some 1531 minimal compliance rules. In pursuing the limits of simplicity, the 1532 ERC design combines and relabels some Dublin Core buckets to isolate 1533 a tiny kernel (subset) of four elements for basic cross-domain 1534 resource description. 1536 For the cross-domain kernel, the ERC uses the four basic elements -- 1537 who, what, when, and where -- to pretend that every object in the 1538 universe can have a uniform minimal description. Each has a name or 1539 other identifier, a location, some responsible person or party, and a 1540 date. It doesn't matter what type of object it is, or whether one 1541 plans to read it, interact with it, smoke it, wear it, or navigate 1542 it. Of course, this approach is flawed because uniformity of 1543 description for some object types requires more semantic contortion 1544 and sacrifice than for others. That is why at the beginning of this 1545 document, the ARK was said to be suited to objects that accommodate 1546 reasonably regular electronic description. 1548 While insisting on uniformity at the most basic level provides 1549 powerful cross-domain leverage, the semantic sacrifice is great for 1550 many applications. So the ERC also permits a semantically rich and 1551 nuanced description to co-exist in a record along with a basic 1552 description. In that way both sophisticated and naive recipients of 1553 the record can extract the level of meaning from it that best suits 1554 their needs and abilities. Key to unlocking the richer description 1555 is a controlled vocabulary of ERC record types (not explained in this 1556 document) that permit knowledgeable recipients to apply defined sets 1557 of additional assumptions to the record. 1559 5.4. Advice to Web Clients 1561 ARKs are envisaged to appear wherever durable object references are 1562 planned. Library cataloging records, literature citations, and 1563 bibliographies are important examples. In many of these places URLs 1564 (Uniform Resource Locators) are currently used, and inside some of 1565 those URLs are embedded URNs, Handles, and DOIs. Unfortunately, 1566 there's no suggestion of a way to probe for extra services that would 1567 build confidence in those identifiers; in other words, there's no way 1568 to tell whether any of those identifiers is any better managed than 1569 the average URL. 1571 ARKs are also envisaged to appear in hypertext links (where they are 1572 not normally shown to users) and in rendered text (displayed or 1573 printed). A normal HTML link for which the URL is not displayed 1574 looks like this. 1576 Click Here 1578 A URL with an embedded ARK invites access (via `?' and `??') to extra 1579 services: 1581 Click Here 1583 Using the [N2T] resolver to provide identifier-scheme-agnostic 1584 protection against hostname instability, this ARK could be published 1585 as: 1587 Click Here 1589 An NAA will typically make known the associations it creates by 1590 publishing them in catalogs, actively advertizing them, or simply 1591 leaving them on web sites for visitors (e.g., users, indexing 1592 spiders) to stumble across in browsing. 1594 5.5. Security Considerations 1596 The ARK naming scheme poses no direct risk to computers and networks. 1597 Implementors of ARK services need to be aware of security issues when 1598 querying networks and filesystems for Name Mapping Authority 1599 services, and the concomitant risks from spoofing and obtaining 1600 incorrect information. These risks are no greater for ARK mapping 1601 authority discovery than for other kinds of service discovery. For 1602 example, recipients of ARKs with a specified hostport (NMAH) should 1603 treat it like a URL and be aware that the identified ARK service may 1604 no longer be operational. 1606 Apart from mapping authority discovery, ARK clients and servers 1607 subject themselves to all the risks that accompany normal operation 1608 of the protocols underlying mapping services (e.g., HTTP, Z39.50). 1609 As specializations of such protocols, an ARK service may limit 1610 exposure to the usual risks. Indeed, ARK services may enhance a kind 1611 of security by helping users identify long-term reliable references 1612 to information objects. 1614 6. References 1616 [ANVL] Kunze, J. and B. Kahle, "A Name-Value Language", 2008, 1617 . 1619 [ARK] Kunze, J., "Towards Electronic Persistence Using ARK 1620 Identifiers", IWAW/ECDL Annual Workshop Proceedings 3rd, 1621 August 2003, 1622 . 1624 [ARKagency] 1625 ARKs-in-the-Open, "ARK Maintenance Agency", 2019, 1626 . 1628 [DCKernel] 1629 Initiative, D. C. M., "Kernel Metadata Working Group", 1630 2001-2008, . 1632 [DOI] Foundation, I. D., "The Digital Object Identifier (DOI) 1633 System", February 2001, . 1635 [ERC] Kunze, J. and A. Turner, "Kernel Metadata and Electronic 1636 Resource Citations", October 2007, 1637 . 1639 [Handle] Lannom, L., "Handle System Overview", ICSTI Forum No. 30, 1640 April 1999, . 1642 [Kernel] Kunze, J., "A Metadata Kernel for Electronic Permanence", 1643 Journal of Digital Information Vol 2, Issue 2, 1644 ISSN 1368-7506, January 2002, 1645 . 1647 [N2T] Library, C. D., "Name-to-Thing Resolver", August 2006, 1648 . 1650 [NAANrequest] 1651 ARKs.org, "NAAN Request Form", 2018, 1652 . 1654 [NLMPerm] Byrnes, M., "Defining NLM's Commitment to the Permanence 1655 of Electronic Information", ARL 212:8-9, October 2000, 1656 . 1658 [NOID] Kunze, J., "Nice Opaque Identifiers", February 2005, 1659 . 1661 [PStatements] 1662 Kunze, J., "Persistence statements: describing digital 1663 stickiness", October 2016, 1664 . 1666 [PURL] Shafer, K., "Introduction to Persistent Uniform Resource 1667 Locators", 1996, . 1669 [RFC0854] Postel, J. and J. Reynolds, "Telnet Protocol 1670 Specification", STD 8, RFC 854, DOI 10.17487/RFC0854, May 1671 1983, . 1673 [RFC1034] Mockapetris, P., "Domain names - concepts and facilities", 1674 STD 13, RFC 1034, DOI 10.17487/RFC1034, November 1987, 1675 . 1677 [RFC2141] Moats, R., "URN Syntax", RFC 2141, DOI 10.17487/RFC2141, 1678 May 1997, . 1680 [RFC2288] Lynch, C., Preston, C., and R. Daniel, "Using Existing 1681 Bibliographic Identifiers as Uniform Resource Names", 1682 RFC 2288, DOI 10.17487/RFC2288, February 1998, 1683 . 1685 [RFC2611] Daigle, L., van Gulik, D., Iannella, R., and P. Faltstrom, 1686 "URN Namespace Definition Mechanisms", BCP 33, RFC 2611, 1687 DOI 10.17487/RFC2611, June 1999, 1688 . 1690 [RFC2616] Fielding, R., Gettys, J., Mogul, J., Frystyk, H., 1691 Masinter, L., Leach, P., and T. Berners-Lee, "Hypertext 1692 Transfer Protocol -- HTTP/1.1", RFC 2616, 1693 DOI 10.17487/RFC2616, June 1999, 1694 . 1696 [RFC2822] Resnick, P., Ed., "Internet Message Format", RFC 2822, 1697 DOI 10.17487/RFC2822, April 2001, 1698 . 1700 [RFC2915] Mealling, M. and R. Daniel, "The Naming Authority Pointer 1701 (NAPTR) DNS Resource Record", RFC 2915, 1702 DOI 10.17487/RFC2915, September 2000, 1703 . 1705 [RFC3986] Berners-Lee, T., Fielding, R., and L. Masinter, "Uniform 1706 Resource Identifier (URI): Generic Syntax", STD 66, 1707 RFC 3986, DOI 10.17487/RFC3986, January 2005, 1708 . 1710 [RFC5013] Kunze, J. and T. Baker, "The Dublin Core Metadata Element 1711 Set", RFC 5013, DOI 10.17487/RFC5013, August 2007, 1712 . 1714 [THUMP] Gamiel, K. and J. Kunze, "The HTTP URL Mapping Protocol", 1715 August 2007, 1716 . 1718 Appendix A. ARK Maintenance Agency: arks.org 1720 The ARK Maintenance Agency [ARKagency] at arks.org has several 1721 functions. 1723 o To manage the registry of organizations that will be assigning 1724 ARKs. Organizations can request or update a NAAN by filling out a 1725 form [NAANrequest]. 1727 o To be a clearinghouse for information about ARKs, such as best 1728 practices, introductory documentation, tutorials, community 1729 forums, etc. These supplemental resources help ARK implementor in 1730 high-level applications across different sectors and disciplines, 1731 and with a variety of metadata standards. 1733 o To be a locus of discussion about future versions of the ARK 1734 specification. 1736 Appendix B. Looking up NMAHs Distributed via DNS 1738 This subsection introduces an older method for looking up NMAHs that 1739 is based on the method for discovering URN resolvers described in 1740 [RFC2915]. It relies on querying the DNS system already installed in 1741 the background infrastructure of most networked computers. A query 1742 is submitted to DNS asking for a list of resolvers that match a given 1743 NAAN. DNS distributes the query to the particular DNS servers that 1744 can best provide the answer, unless the answer can be found more 1745 quickly in a local DNS cache as a side-effect of a recent query. 1746 Responses come back inside Name Authority Pointer (NAPTR) records. 1747 The normal result is one or more candidate NMAHs. 1749 In its full generality the [RFC2915] algorithm ambitiously 1750 accommodates a complex set of preferences, orderings, protocols, 1751 mapping services, regular expression rewriting rules, and DNS record 1752 types. This subsection proposes a drastic simplification of it for 1753 the special case of ARK mapping authority discovery. The simplified 1754 algorithm is called Maptr. It uses only one DNS record type (NAPTR) 1755 and restricts most of its field values to constants. The following 1756 hypothetical excerpt from a DNS data file for the NAAN known as 12026 1757 shows three example NAPTR records ready to use with the Maptr 1758 algorithm. 1760 12026.ark.arpa. 1761 ;; US Library of Congress 1762 ;; order pref flags service regexp replacement 1763 IN NAPTR 0 0 "h" "ark" "USLC" lhc.nlm.nih.gov:8080 1764 IN NAPTR 0 0 "h" "ark" "USLC" foobar.zaf.org 1765 IN NAPTR 0 0 "h" "ark" "USLC" sneezy.dopey.com 1767 All the fields are held constant for Maptr except for the "flags", 1768 "regexp", and "replacement" fields. The "service" field contains the 1769 constant value "ark" so that NAPTR records participating in the Maptr 1770 algorithm will not be confused with other NAPTR records. The "order" 1771 and "pref" fields are held to 0 (zero) and otherwise ignored for now; 1772 the algorithm may evolve to use these fields for ranking decisions 1773 when usage patterns and local administrative needs are better 1774 understood. 1776 When a Maptr query returns a record with a flags field of "h" (for 1777 hostport, a Maptr extension to the NAPTR flags), the replacement 1778 field contains the NMAH (hostport) of an ARK service provider. When 1779 a query returns a record with a flags field of "" (the empty string), 1780 the client needs to submit a new query containing the domain name 1781 found in the replacement field. This second sort of record exploits 1782 the distributed nature of DNS by redirecting the query to another 1783 domain name. It looks like this. 1785 12345.ark.arpa. 1786 ;; Digital Library Consortium 1787 ;; order pref flags service regexp replacement 1788 IN NAPTR 0 0 "" "ark" "" dlc.spct.org. 1790 Here is the Maptr algorithm for ARK mapping authority discovery. In 1791 it replace with the NAAN from the ARK for which an NMAH is 1792 sought. 1794 1. Initialize the DNS query: type=NAPTR, query=.ark.arpa. 1796 2. Submit the query to DNS and retrieve (NAPTR) records, discarding 1797 any record that does not have "ark" for the service field. 1799 3. All remaining records with a flags fields of "h" contain 1800 candidate NMAHs in their replacement fields. Set them aside, if 1801 any. 1803 4. Any record with an empty flags field ("") has a replacement field 1804 containing a new domain name to which a subsequent query should 1805 be redirected. For each such record, set query= 1806 then go to step (2). When all such records have been recursively 1807 exhausted, go to step (5). 1809 5. All redirected queries have been resolved and a set of candidate 1810 NMAHs has been accumulated from steps (3). If there are zero 1811 NMAHs, exit -- no mapping authority was found. If there is one 1812 or more NMAH, choose one using any criteria you wish, then exit. 1814 A Perl script that implements this algorithm is included here. 1816 #!/depot/bin/perl 1818 use Net::DNS; # include simple DNS package 1819 my $qtype = "NAPTR"; # initialize query type 1820 my $naa = shift; # get NAAN script argument 1821 my $mad = new Net::DNS::Resolver; # mapping authority discovery 1823 &maptr("$naa.ark.arpa"); # call maptr - that's it 1825 sub maptr { # recursive maptr algorithm 1826 my $dname = shift; # domain name as argument 1827 my ($rr, $order, $pref, $flags, $service, $regexp, 1828 $replacement); 1829 my $query = $mad->query($dname, $qtype); 1830 return # non-productive query 1831 if (! $query || ! $query->answer); 1832 foreach $rr ($query->answer) { 1833 next # skip records of wrong type 1834 if ($rr->type ne $qtype); 1835 ($order, $pref, $flags, $service, $regexp, 1836 $replacement) = split(/\s/, $rr->rdatastr); 1837 if ($flags eq "") { 1838 &maptr($replacement); # recurse 1839 } elsif ($flags eq "h") { 1840 print "$replacement\n"; # candidate NMAH 1841 } 1842 } 1843 } 1845 The global database thus distributed via DNS and the Maptr algorithm 1846 can easily be seen to mirror the contents of the Name Authority 1847 Table file described in the previous section. 1849 Authors' Addresses 1851 John A. Kunze 1852 California Digital Library 1853 415 20th St, 4th Floor 1854 Oakland, CA 94612 1855 USA 1857 Email: jak@ucop.edu 1858 Emmanuelle Bermes 1859 Bibliotheque nationale de France 1860 Quai Francois Mauriac 1861 Paris, Cedex 13 75706 1862 France 1864 Email: emmanuelle.bermes@bnf.fr