idnits 2.17.1 draft-kunze-ark-18.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 'Intended status' indicated for this document; assuming Proposed Standard 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 1687 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 (April 5, 2013) is 4039 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: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Missing Reference: 'Qualifier' is mentioned on line 486, but not defined -- Possible downref: Non-RFC (?) normative reference: ref. 'ANVL' -- Possible downref: Non-RFC (?) normative reference: ref. 'ARK' -- Possible downref: Non-RFC (?) normative reference: ref. 'DCKernel' -- Possible downref: Non-RFC (?) normative reference: ref. 'DOI' -- Possible downref: Non-RFC (?) normative reference: ref. 'ERC' -- Possible downref: Non-RFC (?) normative reference: ref. 'Handle' -- Possible downref: Non-RFC (?) normative reference: ref. 'Kernel' -- Possible downref: Non-RFC (?) normative reference: ref. 'N2T' -- Possible downref: Non-RFC (?) normative reference: ref. 'NLMPerm' -- Possible downref: Non-RFC (?) normative reference: ref. 'NOID' -- Possible downref: Non-RFC (?) normative reference: ref. 'PURL' ** Obsolete normative reference: RFC 2141 (Obsoleted by RFC 8141) ** Downref: Normative reference to an Informational RFC: RFC 2288 ** 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) ** Downref: Normative reference to an Informational RFC: RFC 5013 -- Possible downref: Non-RFC (?) normative reference: ref. 'THUMP' Summary: 10 errors (**), 0 flaws (~~), 5 warnings (==), 15 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 Expires: October 7, 2013 R. Rodgers 5 University of California San 6 Francisco 7 April 5, 2013 9 The ARK Identifier Scheme 10 http://www.ietf.org/internet-drafts/draft-kunze-ark-18.txt 12 Abstract 14 The ARK (Archival Resource Key) naming scheme is designed to 15 facilitate the high-quality and persistent identification of 16 information objects. A founding principle of the ARK is that 17 persistence is purely a matter of service and is neither inherent in 18 an object nor conferred on it by a particular naming syntax. The 19 best that an identifier can do is to lead users to the services that 20 support robust reference. The term ARK itself refers both to the 21 scheme and to any single identifier that conforms to it. An ARK has 22 five components: 24 [http://NMAH/]ark:/NAAN/Name[Qualifier] 26 an optional and mutable Name Mapping Authority Hostport (usually a 27 hostname), the "ark:" label, the Name Assigning Authority Number 28 (NAAN), the assigned Name, and an optional and possibly mutable 29 Qualifier supported by the NMA. The NAAN and Name together form the 30 immutable persistent identifier for the object independent of the URL 31 hostname. An ARK is a special kind of URL that connects users to 32 three things: the named object, its metadata, and the provider's 33 promise about its persistence. When entered into the location field 34 of a Web browser, the ARK leads the user to the named object. That 35 same ARK, inflected by appending a single question mark (`?'), 36 returns a brief metadata record that is both human- and machine- 37 readable. When the ARK is inflected by appending dual question marks 38 (`??'), the returned metadata contains a commitment statement from 39 the current provider. Tools exist for minting, binding, and 40 resolving ARKs. 42 Status of this Memo 44 This Internet-Draft is submitted in full conformance with the 45 provisions of BCP 78 and BCP 79. 47 Internet-Drafts are working documents of the Internet Engineering 48 Task Force (IETF). Note that other groups may also distribute 49 working documents as Internet-Drafts. The list of current Internet- 50 Drafts is at http://datatracker.ietf.org/drafts/current/. 52 Internet-Drafts are draft documents valid for a maximum of six months 53 and may be updated, replaced, or obsoleted by other documents at any 54 time. It is inappropriate to use Internet-Drafts as reference 55 material or to cite them other than as "work in progress." 57 This Internet-Draft will expire on October 7, 2013. 59 Copyright Notice 61 Copyright (c) 2013 IETF Trust and the persons identified as the 62 document authors. All rights reserved. 64 This document is subject to BCP 78 and the IETF Trust's Legal 65 Provisions Relating to IETF Documents 66 (http://trustee.ietf.org/license-info) in effect on the date of 67 publication of this document. Please review these documents 68 carefully, as they describe your rights and restrictions with respect 69 to this document. Code Components extracted from this document must 70 include Simplified BSD License text as described in Section 4.e of 71 the Trust Legal Provisions and are provided without warranty as 72 described in the Simplified BSD License. 74 Table of Contents 76 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 77 1.1. Reasons to Use ARKs . . . . . . . . . . . . . . . . . . . 5 78 1.2. Three Requirements of ARKs . . . . . . . . . . . . . . . . 6 79 1.3. Organizing Support for ARKs: Our Stuff vs. Their Stuff . 7 80 1.4. Definition of Identifier . . . . . . . . . . . . . . . . . 8 81 2. ARK Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . 10 82 2.1. The Name Mapping Authority Hostport (NMAH) . . . . . . . . 10 83 2.2. The ARK Label Part (ark:/) . . . . . . . . . . . . . . . . 11 84 2.3. The Name Assigning Authority Number (NAAN) . . . . . . . . 12 85 2.4. The Name Part . . . . . . . . . . . . . . . . . . . . . . 13 86 2.5. The Qualifier Part . . . . . . . . . . . . . . . . . . . . 13 87 2.5.1. ARKs that Reveal Object Hierarchy . . . . . . . . . . 15 88 2.5.2. ARKs that Reveal Object Variants . . . . . . . . . . . 16 89 2.6. Character Repertoires . . . . . . . . . . . . . . . . . . 17 90 2.7. Normalization and Lexical Equivalence . . . . . . . . . . 18 91 3. Naming Considerations . . . . . . . . . . . . . . . . . . . . 20 92 3.1. ARKS Embedded in Language . . . . . . . . . . . . . . . . 20 93 3.2. Objects Should Wear Their Identifiers . . . . . . . . . . 20 94 3.3. Names are Political, not Technological . . . . . . . . . . 21 95 3.4. Choosing a Hostname or NMA . . . . . . . . . . . . . . . . 21 96 3.5. Assigners of ARKs . . . . . . . . . . . . . . . . . . . . 23 97 3.6. NAAN Namespace Management . . . . . . . . . . . . . . . . 23 98 3.7. Sub-Object Naming . . . . . . . . . . . . . . . . . . . . 25 99 4. Finding a Name Mapping Authority . . . . . . . . . . . . . . . 26 100 4.1. Looking Up NMAHs in a Globally Accessible File . . . . . . 27 101 5. Generic ARK Service Definition . . . . . . . . . . . . . . . . 29 102 5.1. Generic ARK Access Service (access, location) . . . . . . 29 103 5.1.1. Generic Policy Service (permanence, naming, etc.) . . 29 104 5.1.2. Generic Description Service . . . . . . . . . . . . . 31 105 5.2. Overview of The HTTP URL Mapping Protocol (THUMP) . . . . 31 106 5.3. The Electronic Resource Citation (ERC) . . . . . . . . . . 34 107 5.4. Advice to Web Clients . . . . . . . . . . . . . . . . . . 36 108 5.5. Security Considerations . . . . . . . . . . . . . . . . . 37 109 6. References . . . . . . . . . . . . . . . . . . . . . . . . . . 38 110 Appendix A. ARK Maintenance Agency . . . . . . . . . . . . . . . 40 111 Appendix B. Looking up NMAHs Distributed via DNS . . . . . . . . 41 112 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 44 114 1. Introduction 116 This document describes a scheme for the high-quality naming of 117 information resources. The scheme, called the Archival Resource Key 118 (ARK), is well suited to long-term access and identification of any 119 information resources that accommodate reasonably regular electronic 120 description. This includes digital documents, databases, software, 121 and websites, as well as physical objects (books, bones, statues, 122 etc.) and intangible objects (chemicals, diseases, vocabulary terms, 123 performances). Hereafter the term "object" refers to an information 124 resource. The term ARK itself refers both to the scheme and to any 125 single identifier that conforms to it. A reasonably concise and 126 accessible overview and rationale for the scheme is available at 127 [ARK]. 129 Schemes for persistent identification of network-accessible objects 130 are not new. In the early 1990's, the design of the Uniform Resource 131 Name [RFC2141] responded to the observed failure rate of URLs by 132 articulating an indirect, non-hostname-based naming scheme and the 133 need for responsible name management. Meanwhile, promoters of the 134 Digital Object Identifier [DOI] succeeded in building a community of 135 providers around a mature software system [Handle] that supports name 136 management. The Persistent Uniform Resource Locator [PURL] was 137 another scheme that had the advantage of working with unmodified web 138 browsers. ARKs represent an approach that attempts to build on the 139 strengths and to avoid the weaknesses of these schemes. 141 A founding principle of the ARK is that persistence is purely a 142 matter of service. Persistence is neither inherent in an object nor 143 conferred on it by a particular naming syntax. Nor is the technique 144 of name indirection -- upon which URNs, Handles, DOIs, and PURLs are 145 founded -- of central importance. Name indirection is an ancient and 146 well-understood practice; new mechanisms for it keep appearing and 147 distracting practitioner attention, with the Domain Name System (DNS) 148 [RFC1034] being a particularly dazzling and elegant example. What is 149 often forgotten is that maintenance of an indirection table is an 150 unavoidable cost to the organization providing persistence, and that 151 cost is equivalent across naming schemes. That indirection has 152 always been a native part of the web while being so lightly utilized 153 for the persistence of web-based objects indicates how unsuited most 154 organizations will probably be to the task of table maintenance and 155 to the much more fundamental challenge of keeping the objects 156 themselves viable. 158 Persistence is achieved through a provider's successful stewardship 159 of objects and their identifiers. The highest level of persistence 160 will be reinforced by a provider's robust contingency, redundancy, 161 and succession strategies. It is further safeguarded to the extent 162 that a provider's mission is shielded from funding and political 163 instabilities. These are by far the major challenges confronting 164 persistence providers, and no identifier scheme has any direct impact 165 on them. In fact, some schemes may actually be liabilities for 166 persistence because they create short- and long-term dependencies for 167 every object access on complex, special-purpose infrastructures, 168 parts of which are proprietary and all of which increase the carry- 169 forward burden for the preservation community. It is for this reason 170 that the ARK scheme relies only on educated name assignment and light 171 use of general-purpose infrastructures that are maintained mostly by 172 the internet community at large (the DNS, web servers, and web 173 browsers). 175 1.1. Reasons to Use ARKs 177 If no persistent identifier scheme contributes directly to 178 persistence, why not just use URLs? A particular URL may be as 179 durable an identifier as it is possible to have, but nothing 180 distinguishes it from an ordinary URL to the recipient who is 181 wondering if it is suitable for long-term reference. An ARK embedded 182 in a URL provides some of the necessary conditions for credible 183 persistence, inviting access to not one, but to three things: to the 184 object, to its metadata, and to a nuanced statement of commitment 185 from the provider in question (the NMA, described below) regarding 186 the object. Existence of the two extra services can be probed 187 automatically by appending `?' and `??' to the ARK. 189 The form of the ARK also supports the natural separation of naming 190 authorities into the original name assigning authority and the 191 diverse multiple name mapping (or servicing) authorities that in 192 succession and in parallel will take over custodial responsibilities 193 from the original assigner (assuming the assigner ever held that 194 responsibility) for the large majority of a long-term object's 195 archival lifetime. The name mapping authority, indicated by the 196 hostname part of the URL that contains the ARK, serves to launch the 197 ARK into cyberspace. Should it ever fail (and there is no reason why 198 a well-chosen hostname for a 100-year-old cultural memory institution 199 shouldn't last as long as the DNS), that host name is considered 200 disposeable and replaceable. Again, the form of the ARK helps 201 because it defines exactly how to recover the core immutable object 202 identity, and simple algorithms (one based on the URN model) or even 203 by-hand internet query can be used for for locating another mapping 204 authority. 206 There are tools to assist in generating ARKs and other identifiers, 207 such as [NOID] and "uuidgen", both of which rely for uniqueness on 208 human-maintained registries. This document also contains some 209 guidelines and considerations for managing namespaces and choosing 210 hostnames with persistence in mind. 212 1.2. Three Requirements of ARKs 214 The first requirement of an ARK is to give users a link from an 215 object to a promise of stewardship for it. That promise is a multi- 216 faceted covenant that binds the word of an identified service 217 provider to a specific set of responsibilities. It is critical for 218 the promise to come from a current provider and almost irrelevant, 219 over a long period of time, what the original assigner's intentions 220 were. No one can tell if successful stewardship will take place 221 because no one can predict the future. Reasonable conjecture, 222 however, may be based on past performance. There must be a way to 223 tie a promise of persistence to a provider's demonstrated or 224 perceived ability -- its reputation -- in that arena. Provider 225 reputations would then rise and fall as promises are observed 226 variously to be kept and broken. This is perhaps the best way we 227 have for gauging the strength of any persistence promise. 229 The second requirement of an ARK is to give users a link from an 230 object to a description of it. The problem with a naked identifier 231 is that without a description real identification is incomplete. 232 Identifiers common today are relatively opaque, though some contain 233 ad hoc clues reflecting assertions that were briefly true, such as 234 where in a filesystem hierarchy an object lived during a short stay. 235 Possession of both an identifier and an object is some improvement, 236 but positive identification may still be uncertain since the object 237 itself might not include a matching identifier or might not carry 238 evidence obvious enough to reveal its identity without significant 239 research. In either case, what is called for is a record bearing 240 witness to the identifier's association with the object, as supported 241 by a recorded set of object characteristics. This descriptive record 242 is partly an identification "receipt" with which users and archivists 243 can verify an object's identity after brief inspection and a 244 plausible match with recorded characteristics such as title and size. 246 The final requirement of an ARK is to give users a link to the object 247 itself (or to a copy) if at all possible. Persistent access is the 248 central duty of an ARK. Persistent identification plays a vital 249 supporting role but, strictly speaking, it can be construed as no 250 more than a record attesting to the original assignment of a never- 251 reassigned identifier. Object access may not be feasible for various 252 reasons, such as a transient service outage, a catastrophic loss, a 253 licensing agreement that keeps an archive "dark" for a period of 254 years, or when an object's own lack of tangible existence confuses 255 normal concepts of access (e.g., a vocabulary term might be 256 "accessed" through its definition). In such cases the ARK's 257 identification role assumes a much higher profile. But attempts to 258 simplify the persistence problem by decoupling access from 259 identification and concentrating exclusively on the latter are of 260 questionable utility. A perfect system for assigning forever unique 261 identifiers might be created, but if it did so without reducing 262 access failure rates, no one would be interested. The central issue 263 -- which may be summed up as the "HTTP 404 Not Found" problem -- 264 would not have been addressed. 266 1.3. Organizing Support for ARKs: Our Stuff vs. Their Stuff 268 An organization and the user community it serves can often be seen to 269 struggle with two different areas of persistent identification: the 270 Our Stuff problem and the Their Stuff problem. In the Our Stuff 271 problem, we in the organization want our own objects to acquire 272 persistent names. Since we possess or control these objects, our 273 organization tackles the Our Stuff problem directly. Whether or not 274 the objects are named by ARKs, our organization is the responsible 275 party, so it can plan for, maintain, and make commitments about the 276 objects. 278 In the Their Stuff problem, we in the organization want others' 279 objects to acquire persistent names. These are objects that we do 280 not own or control, but some of which are critically important to us. 281 But because they are beyond our influence as far as support is 282 concerned, creating and maintaining persistent identifiers for Their 283 Stuff is not especially purposeful or feasible for us to engage in. 284 There is little that we can do about someone else's stuff except 285 encourage their uptake or adoption of persistence services. 287 Co-location of persistent access and identification services is 288 natural. Any organization that undertakes ongoing support of true 289 persistent identification (which includes description) is well-served 290 if it controls, owns, or otherwise has clear internal access to the 291 identified objects, and this gives it an advantage if it wishes also 292 to support persistent access to outsiders. Conversely, persistent 293 access to outsiders requires orderly internal collection management 294 procedures that include monitoring, acquisition, verification, and 295 change control over objects, which in turn requires object 296 identifiers persistent enough to support auditable record keeping 297 practices. 299 Although, organizing ARK services under one roof thus tends to make 300 sense, object hosting can successfully be separated from name 301 mapping. An example is when a name mapping authority centrally 302 provides uniform resolution services via a protocol gateway on behalf 303 of organizations that host objects behind a variety of access 304 protocols. It is also reasonable to build value-added description 305 services that rely on the underlying services of a set of mapping 306 authorities. 308 Supporting ARKs is not for every organization. By requiring 309 specific, revealed commitments to preservation, to object access, and 310 to description, the bar for providing ARK services is higher than for 311 some other identifier schemes. On the other hand, it would be hard 312 to grant credence to a persistence promise from an organization that 313 could not muster the minimum ARK services. Not that there isn't a 314 business model for an ARK-like, description-only service built on top 315 of another organization's full complement of ARK services. For 316 example, there might be competition at the description level for 317 abstracting and indexing a body of scientific literature archived in 318 a combination of open and fee-based repositories. The description- 319 only service would have no direct commitment to the objects, but 320 would act as an intermediary, forwarding commitment statements from 321 object hosting services to requestors. 323 1.4. Definition of Identifier 325 An identifier is not a string of character data -- an identifier is 326 an association between a string of data and an object. This 327 abstraction is necessary because without it a string is just data. 328 It's nonsense to talk about a string's breaking, or about its being 329 strong, maintained, and authentic. But as a representative of an 330 association, a string can do, metaphorically, the things that we 331 expect of it. 333 Without regard to whether an object is physical, digital, or 334 conceptual, to identify it is to claim an association between it and 335 a representative string, such as "Jane" or "ISBN 0596000278". What 336 gives a claim credibility is a set of verifiable assertions, or 337 metadata, about the object, such as age, height, title, or number of 338 pages. In other words, the association is made manifest by a record 339 (e.g., a cataloging or other metadata record) that vouches for it. 341 In the complete absence of any testimony (metadata) regarding an 342 association, a would-be identifier string is a meaningless sequence 343 of characters. To keep an externally visible but otherwise internal 344 string from being perceived as an identifier by outsiders, for 345 example, it suffices for an organization not to disclose the nature 346 of its association. For our immediate purpose, actual existence of 347 an association record is more important than its authenticity or 348 verifiability, which are outside the scope of this specification. 350 It is a gift to the identification process if an object carries its 351 own name as an inseparable part of itself, such as an identifier 352 imprinted on the first page of a document or embedded in a data 353 structure element of a digital document header. In cases where the 354 object is large, unwieldy, or unavailable (such as when licensing 355 restrictions are in effect), a metadata record that includes the 356 identifier string will usually suffice. That record becomes a 357 conveniently manipulable object surrogate, acting as both an 358 association "receipt" and "declaration". 360 Note that our definition of identifier extends the one in use for 361 Uniform Resource Identifiers [RFC3986]. The present document still 362 sometimes (ab)uses the terms "ARK" and "identifier" as shorthand for 363 the string part of an identifier, but the context should make the 364 meaning clear. 366 2. ARK Anatomy 368 An ARK is represented by a sequence of characters (a string) that 369 contains the label, "ark:", optionally preceded by the beginning part 370 of a URL. Here is a diagrammed example. 372 http://example.org/ark:/12025/654xz321/s3/f8.05v.tiff 373 \________________/ \__/ \___/ \______/ \____________/ 374 (replaceable) | | | Qualifier 375 | ARK Label | | (NMA-supported) 376 | | | 377 Name Mapping Authority | Name (NAA-assigned) 378 Hostport (NMAH) | 379 Name Assigning Authority Number (NAAN) 381 The ARK syntax can be summarized, 383 [http://NMAH/]ark:/NAAN/Name[Qualifier] 385 where the NMAH and Qualifier parts are in brackets to indicate that 386 they are optional. 388 2.1. The Name Mapping Authority Hostport (NMAH) 390 Before the "ark:" label may appear an optional Name Mapping Authority 391 Hostport (NMAH) that is a temporary address where ARK service 392 requests may be sent. It consists of "http://" (or any service 393 specification valid for a URL) followed by an Internet hostname or 394 hostport combination having the same format and semantics as the 395 hostport part of a URL. The most important thing about the NMAH is 396 that it is "identity inert" from the point of view of object 397 identification. In other words, ARKs that differ only in the 398 optional NMAH part identify the same object. Thus, for example, the 399 following three ARKs are synonyms for just one information object: 401 http://loc.gov/ark:/12025/654xz321 402 http://rutgers.edu/ark:/12025/654xz321 403 ark:/12025/654xz321 405 Strictly speaking, in the realm of digital objects, these ARKs may 406 lead over time to somewhat different or diverging instances of the 407 originally named object. In an ideal world, divergence of persistent 408 objects is not desirable, but it is widely believed that digital 409 preservation efforts will inevitably lead to alterations in some 410 original objects (e.g, a format migration in order to preserve the 411 ability to display a document). If any of those objects are held 412 redundantly in more than one organization (a common preservation 413 strategy), chances are small that all holding organizations will 414 perform the same precise transformations and all maintain the same 415 object metadata. More significant divergence would be expected when 416 the holding organizations serve different audiences or compete with 417 each other. 419 The NMAH part makes an ARK into an actionable URL. As with many 420 internet parameters, it is helpful to approach the NMAH being liberal 421 in what you accept and conservative in what you propose. From the 422 recipient's point of view, the NMAH part should be treated as 423 temporary, disposable, and replaceable. From the NMA's point of 424 view, it should be chosen with the greatest concern for longevity. A 425 carefully chosen NMAH should be at least as permanent as the 426 providing organization's own hostname. In the case of a national or 427 university library, for example, there is no reason why the NMAH 428 should not be considerably more permanent than soft-funded proxy 429 hostnames such as hdl.handle.net, dx.doi.org, and purl.org. In 430 general and over time, however, it is not unexpected for an NMAH 431 eventually to stop working and require replacement with the NMAH of a 432 currently active service provider. 434 This replacement relies on a mapping authority "resolver" discovery 435 process, of which two alternate methods are outlined in a later 436 section. The ARK, URN, Handle, and DOI schemes all use a resolver 437 discovery model that sooner or later requires matching the original 438 assigning authority with a current provider servicing that 439 authority's named objects; once found, the resolver at that provider 440 performs what amounts to a redirect to a place where the object is 441 currently held. All the schemes rely on the ongoing functionality of 442 currently mainstream technologies such as the Domain Name System 443 [RFC1034] and web browsers. The Handle and DOI schemes in addition 444 require that the Handle protocol layer and global server grid be 445 available at all times. 447 The practice of prepending "http://" and an NMAH to an ARK is a way 448 of creating an actionable identifier by a method that is itself 449 temporary. Assuming that infrastructure supporting [RFC2616] 450 information retrieval will no longer be available one day, ARKs will 451 then have to be converted into new kinds of actionable identifiers. 452 By that time, if ARKs see widespread use, web browsers would 453 presumably evolve to perform this (currently simple) transformation 454 automatically. 456 2.2. The ARK Label Part (ark:/) 458 The label part distinguishes an ARK from an ordinary identifier. In 459 a URL found in the wild, the string, "ark:/", indicates that the URL 460 stands a reasonable chance of being an ARK. If the context warrants, 461 verification that it actually is an ARK can be done by testing it for 462 existence of the three ARK services. 464 Since nothing about an identifier syntax directly affects 465 persistence, the "ark:" label (like "urn:", "doi:", and "hdl:") 466 cannot tell you whether the identifier is persistent or whether the 467 object is available. It does tell you that the original Name 468 Assigning Authority (NAA) had some sort of hopes for it, but it 469 doesn't tell you whether that NAA is still in existence, or whether a 470 decade ago it ceased to have any responsibility for providing 471 persistence, or whether it ever had any responsibility beyond naming. 473 Only a current provider can say for certain what sort of commitment 474 it intends, and the ARK label suggests that you can query the NMAH 475 directly to find out exactly what kind of persistence is promised. 476 Even if what is promised is impersistence (i.e., a short-term 477 identifier), saying so is valuable information to the recipient. 478 Thus an ARK is a high-functioning identifier in the sense that it 479 provides access to the object, the metadata, and a commitment 480 statement, even if the commitment is explicitly very weak. 482 2.3. The Name Assigning Authority Number (NAAN) 484 Recalling that the general form of the ARK is, 486 [http://NMAH/]ark:/NAAN/Name[Qualifier] 488 the part of the ARK directly following the "ark:" is the Name 489 Assigning Authority Number (NAAN) enclosed in `/' (slash) characters. 490 This part is always required, as it identifies the organization that 491 originally assigned the Name of the object. It is used to discover a 492 currently valid NMAH and to provide top-level partitioning of the 493 space of all ARKs. NAANs are registered in a manner similar to URN 494 Namespaces, but they are pure numbers consisting of 5 digits or 9 495 digits. Thus, the first 100,000 registered NAAs fit compactly into 496 the 5 digits, and if growth warrants, the next billion fit into the 9 497 digit form. In either case the fixed odd numbers of digits helps 498 reduce the chances of finding a NAAN out of context and confusing it 499 with nearby quantities such as 4-digit dates. 501 The NAAN designates a top-level ARK namespace. Once registered for a 502 namespace, a NAAN is never re-registered. It is possible, however, 503 for there to be a succession of organizations that manage of an ARK 504 namespace. 506 2.4. The Name Part 508 The part of the ARK just after the NAAN is the Name assigned by the 509 NAA, and it is also required. Semantic opaqueness in the Name part 510 is strongly encouraged in order to reduce an ARK's vulnerability to 511 era- and language-specific change. Identifier strings containing 512 linguistic fragments can create support difficulties down the road. 513 No matter how appropriate or even meaningless they are today, such 514 fragments may one day create confusion, give offense, or infringe on 515 a trademark as the semantic environment around us and our communities 516 evolves. 518 Names that look more or less like numbers avoid common problems that 519 defeat persistence and international acceptance. The use of digits 520 is highly recommended. Mixing in non-vowel alphabetic characters a 521 couple at a time is a relatively safe and easy way to achieve a 522 denser namespace (more possible names for a given length of the name 523 string). Such names have a chance of aging and traveling well. 524 Tools exists that mint, bind, and resolve opaque identifiers, with or 525 without check characters [NOID]. More on naming considerations is 526 given in a subsequent section. 528 2.5. The Qualifier Part 530 The part of the ARK following the NAA-assigned Name is an optional 531 Qualifier. It is a string that extends the base ARK in order to 532 create a kind of service entry point into the object named by the 533 NAA. At the discretion of the providing NMA, such a service entry 534 point permits an ARK to support access to individual hierarchical 535 components and subcomponents of an object, and to variants (versions, 536 languages, formats) of components. A Qualifier may be invented by 537 the NAA or by any NMA servicing the object. 539 In form, the Qualifier is a ComponentPath, or a VariantPath, or a 540 ComponentPath followed by a VariantPath. A VariantPath is introduced 541 and subdivided by the reserved character `.', and a ComponentPath is 542 introduced and subdivided by the reserved character `/'. In this 543 example, 545 http://example.org/ark:/12025/654xz321/s3/f8.05v.tiff 547 the string "/s3/f8" is a ComponentPath and the string ".05v.tiff" is 548 a VariantPath. The ARK Qualifier is a formalization of some 549 currently mainstream URL syntax conventions. This formalization 550 specifically reserves meanings that permit recipients to make strong 551 inferences about logical sub-object containment and equivalence based 552 only on the form of the received identifiers; there is great 553 efficiency in not having to inspect metadata records to discover such 554 relationships. NMAs are free not to disclose any of these 555 relationships merely by avoiding the reserved characters above. 556 Hierarchical components and variants are discussed further in the 557 next two sections. 559 The Qualifier, if present, differs from the Name in several important 560 respects. First, a Qualifier may have been assigned either by the 561 NAA or later by the NMA. The assignment of a Qualifier by an NMA 562 effectively amounts to an act of publishing a service entry point 563 within the conceptual object originally named by the NAA. For our 564 purposes, an ARK extended with a Qualifier assigned by an NMA will be 565 called an NMA-qualified ARK. 567 Second, a Qualifier assignment on the part of an NMA is made in 568 fulfillment of its service obligations and may reflect changing 569 service expectations and technology requirements. NMA-qualified ARKs 570 could therefore be transient, even if the base, unqualified ARK is 571 persistent. For example, it would be reasonable for an NMA to 572 support access to an image object through an actionable ARK that is 573 considered persistent even if the experience of that access changes 574 as linking, labeling, and presentation conventions evolve and as 575 format and security standards are updated. For an image "thumbnail", 576 that NMA could also support an NMA-qualified ARK that is considered 577 impersistent because the thumbnail will be replaced with higher 578 resolution images as network bandwidth and CPU speeds increase. At 579 the same time, for an originally scanned, high-resolution master, the 580 NMA could publish an NMA-qualfied ARK that is itself considered 581 persistent. Of course, the NMA must be able to return its separate 582 commitments to unqualified, NAA-assigned ARKs, to NMA-qualified ARKs, 583 and to any NAA-qualified ARKs that it supports. 585 A third difference between a Qualifier and a Name concerns the 586 semantic opaqueness constraint. When an NMA-qualified ARK is to be 587 used as a transient service entry point into a persistent object, the 588 priority given to semantic opaqueness observed by the NAA in the Name 589 part may be relaxed by the NMA in the Qualifier part. If service 590 priorities in the Qualifier take precedence over persistence, short- 591 term usability considerations may recommend somewhat semantically 592 laden Qualifier strings. 594 Finally, not only is the set of Qualifiers supported by an NMA 595 mutable, but different NMAs may support different Qualifier sets for 596 the same NAA-identified object. In this regard the NMAs act 597 independently of each other and of the NAA. 599 The next two sections describe how ARK syntax may be used to declare, 600 or to avoid declaring, certain kinds of relatedness among qualified 601 ARKs. 603 2.5.1. ARKs that Reveal Object Hierarchy 605 An NAA or NMA may choose to reveal the presence of a hierarchical 606 relationship between objects using the `/' (slash) character after 607 the Name part of an ARK. Some authorities will choose not to 608 disclose this information, while others will go ahead and disclose so 609 that manipulators of large sets of ARKs can infer object 610 relationships by simple identifier inspection; for example, this 611 makes it possible for a system to present a collapsed view of a large 612 search result set. 614 If the ARK contains an internal slash after the NAAN, the piece to 615 its left indicates a containing object. For example, publishing an 616 ARK of the form, 618 ark:/12025/654/xz/321 620 is equivalent to publishing three ARKs, 622 ark:/12025/654/xz/321 623 ark:/12025/654/xz 624 ark:/12025/654 626 together with a declaration that the first object is contained in the 627 second object, and that the second object is contained in the third. 629 Revealing the presence of hierarchy is completely up to the assigner 630 (NMA or NAA). It is hard enough to commit to one object's name, let 631 alone to three objects' names and to a specific, ongoing relatedness 632 among them. Thus, regardless of whether hierarchy was present 633 initially, the assigner, by not using slashes, reveals no shared 634 inferences about hierarchical or other inter-relatedness in the 635 following ARKs: 637 ark:/12025/654_xz_321 638 ark:/12025/654_xz 639 ark:/12025/654xz321 640 ark:/12025/654xz 641 ark:/12025/654 643 Note that slashes around the ARK's NAAN (/12025/ in these examples) 644 are not part of the ARK's Name and therefore do not indicate the 645 existence of some sort of NAAN super object containing all objects in 646 its namespace. A slash must have at least one non-structural 647 character (one that is neither a slash nor a period) on both sides in 648 order for it to separate recognizable structural components. So 649 initial or final slashes may be removed, and double slashes may be 650 converted into single slashes. 652 2.5.2. ARKs that Reveal Object Variants 654 An NAA or NMA may choose to reveal the possible presence of variant 655 objects or object components using the `.' (period) character after 656 the Name part of an ARK. Some authorities will choose not to 657 disclose this information, while others will go ahead and disclose so 658 that manipulators of large sets of ARKs can infer object 659 relationships by simple identifier inspection; for example, this 660 makes it possible for a system to present a collapsed view of a large 661 search result set. 663 If the ARK contains an internal period after Name, the piece to its 664 left is a base name and the piece to its right, and up to the end of 665 the ARK or to the next period is a suffix. A Name may have more than 666 one suffix, for example, 668 ark:/12025/654.24 669 ark:/12025/xz4/654.24 670 ark:/12025/654.20v.78g.f55 672 ark:/12025/654.24 673 ark:/12025/xz4/654.24 674 ark:/12025/654.20v.78g.f55 676 There are two main rules. First, if two ARKs share the same base 677 name but have different suffixes, the corresponding objects were 678 considered variants of each other (different formats, languages, 679 versions, etc.) by the assigner (NMA or NAA). Thus, the following 680 ARKs are variants of each other: 682 ark:/12025/654.20v.78g.f55 683 ark:/12025/654.321xz 684 ark:/12025/654.44 686 Second, publishing an ARK with a suffix implies the existence of at 687 least one variant identified by the ARK without its suffix. The ARK 688 otherwise permits no further assumptions about what variants might 689 exist. So publishing the ARK, 691 ark:/12025/654.20v.78g.f55 693 is equivalent to publishing the four ARKs, 695 ark:/12025/654.20v.78g.f55 696 ark:/12025/654.20v.78g 697 ark:/12025/654.20v 698 ark:/12025/654 700 Revealing the possibility of variants is completely up to the 701 assigner. It is hard enough to commit to one object's name, let 702 alone to multiple variants' names and to a specific, ongoing 703 relatedness among them. The assigner is the sole arbiter of what 704 constitutes a variant within its namespace, and whether to reveal 705 that kind of relatedness by using periods within its names. 707 A period must have at least one non-structural character (one that is 708 neither a slash nor a period) on both sides in order for it to 709 separate recognizable structural components. So initial or final 710 periods may be removed, and adjacent periods may be converted into a 711 single period. Multiple suffixes should be arranged in sorted order 712 (pure ASCII collating sequence) at the end of an ARK. 714 2.6. Character Repertoires 716 The Name and Qualifier parts are strings of visible ASCII characters 717 and should be less than 128 bytes in length. The length restriction 718 keeps the ARK short enough to append ordinary ARK request strings 719 without running into transport restrictions (e.g., within HTTP GET 720 requests). Characters may be letters, digits, or any of these six 721 characters: 723 = # * + @ _ $ 725 The following characters may also be used, but their meanings are 726 reserved: 728 % - . / 730 The characters `/' and `.' are ignored if either appears as the last 731 character of an ARK. If used internally, they allow a name assigner 732 to reveal object hierarchy and object variants as previously 733 described. 735 Hyphens are considered to be insignificant and are always ignored in 736 ARKs. A `-' (hyphen) may appear in an ARK for readability, or it may 737 have crept in during the formatting and wrapping of text, but it must 738 be ignored in lexical comparisons. As in a telephone number, hyphens 739 have no meaning in an ARK. It is always safe for an NMA that 740 receives an ARK to remove any hyphens found in it. As a result, like 741 the NMAH, hyphens are "identity inert" in comparing ARKs for 742 equivalence. For example, the following ARKs are equivalent for 743 purposes of comparison and ARK service access: 745 ark:/12025/65-4-xz-321 746 http://sneezy.dopey.com/ark:/12025/654--xz32-1 747 ark:/12025/654xz321 749 The `%' character is reserved for %-encoding all other octets that 750 would appear in the ARK string, in the same manner as for URIs 751 [RFC3986]. A %-encoded octet consists of a `%' followed by two hex 752 digits; for example, "%7d" stands in for `}'. Lower case hex digits 753 are preferred to reduce the chances of false acronym recognition; 754 thus it is better to use "%acT" instead of "%ACT". The character `%' 755 itself must be represented using "%25". As with URNs, %-encoding 756 permits ARKs to support legacy namespaces (e.g., ISBN, ISSN, SICI) 757 that have less restricted character repertoires [RFC2288]. 759 2.7. Normalization and Lexical Equivalence 761 To determine if two or more ARKs identify the same object, the ARKs 762 are compared for lexical equivalence after first being normalized. 763 Since ARK strings may appear in various forms (e.g., having different 764 NMAHs), normalizing them minimizes the chances that comparing two ARK 765 strings for equality will fail unless they actually identify 766 different objects. In a specified-host ARK (one having an NMAH), the 767 NMAH never participates in such comparisons. 769 Normalization of an ARK for the purpose of octet-by-octet equality 770 comparison with another ARK consists of four steps. First, any upper 771 case letters in the "ark:" label and the two characters following a 772 `%' are converted to lower case. The case of all other letters in 773 the ARK string must be preserved. Second, any NMAH part is removed 774 (everything from an initial "http://" up to the next slash) and all 775 hyphens are removed. 777 Third, structural characters (slash and period) are normalized. 778 Initial and final occurrences are removed, and two structural 779 characters in a row (e.g., // or ./) are replaced by the first 780 character, iterating until each occurrence has at least one non- 781 structural character on either side. Finally, if there are any 782 components with a period on the left and a slash on the right, either 783 the component and the preceding period must be moved to the end of 784 the Name part or the ARK must be thrown out as malformed. 786 The fourth and final step is to arrange the suffixes in ASCII 787 collating sequence (that is, to sort them) and to remove duplicate 788 suffixes, if any. It is also permissible to throw out ARKs for which 789 the suffixes are not sorted. 791 The resulting ARK string is now normalized. Comparisons between 792 normalized ARKs are case-sensitive, meaning that upper case letters 793 are considered different from their lower case counterparts. 795 To keep ARK string variation to a minimum, no reserved ARK characters 796 should be %-encoded unless it is deliberately to conceal their 797 reserved meanings. No non-reserved ARK characters should ever be 798 %-encoded. Finally, no %-encoded character should ever appear in an 799 ARK in its decoded form. 801 3. Naming Considerations 803 The most important threats faced by persistence providers include 804 such things as funding loss, natural disaster, political and social 805 upheaval, processing faults, and errors in human oversight. There is 806 nothing that an identifer scheme can do about such things. Still, a 807 few observed identifier failures and inconveniences can be traced 808 back to naming practices that we now know to be less than optimal for 809 persistence. 811 3.1. ARKS Embedded in Language 813 The ARK has different goals from the URI, so it has different 814 character set requirements. Because linguistic constructs imperil 815 persistence, for ARKs non-ASCII character support is unimportant. 816 ARKs and URIs share goals of transcribability and transportability 817 within web documents, so characters are required to be visible, non- 818 conflicting with HTML/XML syntax, and not subject to tampering during 819 transmission across common transport gateways. Add the goal of 820 making an undelimited ARK recognizable in running prose, as in ark:/ 821 12025/=@_22*$, and certain punctuation characters (e.g., comma, 822 period) end up being excluded from the ARK lest the end of a phrase 823 or sentence be mistaken for part of the ARK. 825 This consideration has more direct effect on ARK usability in a 826 natural language context than it has on ARK persistence. The same is 827 true of the rule preventing hyphens from having lexical significance. 828 It is fine to publish ARKs with hyphens in them (e.g., such as the 829 output of UUID/GUID generators), but the uniform treatment of hyphens 830 as insignificant reduces the possibility of users transcribing 831 identifiers that will have been broken through unpredictable 832 hyphenation by word processors. Any measure that reduces user 833 irritation with an identifier will increase its chances of survival. 835 3.2. Objects Should Wear Their Identifiers 837 A valuable technique for provision of persistent objects is to try to 838 arrange for the complete identifier to appear on, with, or near its 839 retrieved object. An object encountered at a moment in time when its 840 discovery context has long since disappeared could then easily be 841 traced back to its metadata, to alternate versions, to updates, etc. 842 This has seen reasonable success, for example, in book publishing and 843 software distribution. An identifier string only has meaning when 844 its association is known, and this a very sure, simple, and low-tech 845 method of reminding everyone exactly what that association is. 847 3.3. Names are Political, not Technological 849 If persistence is the goal, a deliberate local strategy for 850 systematic name assignment is crucial. Names must be chosen with 851 great care. Poorly chosen and managed names will devastate any 852 persistence strategy, and they do not discriminate by identifier 853 scheme. Whether a mistakenly re-assigned name is a URN, DOI, PURL, 854 URL, or ARK, the damage -- failed access and confusion -- is not 855 mitigated more in one scheme than in another. Conversely, in-house 856 efforts to manage names responsibly will go much further towards 857 safeguarding persistence than any choice of naming scheme or name 858 resolution technology. 860 Branding (e.g., at the corporate or departmental level) is important 861 for funding and visibility, but substrings representing brands and 862 organizational names should be given a wide berth except when 863 absolutely necessary in the hostname (the identity-inert) part of the 864 ARK. These substrings are not only unstable because organizations 865 change frequently, but they are also dangerous because successor 866 organizations often have political or legal reasons to actively 867 suppress predecessor names and brands. Any measure that reduces the 868 chances of future political or legal pressure on an identifier will 869 decrease the chances that our descendants will be obliged to 870 deliberately break it. 872 3.4. Choosing a Hostname or NMA 874 Hostnames appearing in any identifier meant to be persistent must be 875 chosen with extra care. The tendency in hostname selection has 876 traditionally been to choose a token with recognizable attributes, 877 such as a corporate brand, but that tendency wreaks havoc with 878 persistence that is supposed to outlive brands, corporations, subject 879 classifications, and natural language semantics (e.g., what did the 880 three letters "gay" mean in 1958, 1978, and 1998?). Today's 881 recognized and correct attributes are tomorrow's stale or incorrect 882 attributes. In making hostnames (any names, actually) long-term 883 persistent, it helps to eliminate recognizable attributes to the 884 extent possible. This affects selection of any name based on URLs, 885 including PURLs and the explicitly disposable NMAHs. 887 There is no excuse for a provider that manages its internal names 888 impeccably not to exercise the same care in choosing what could be an 889 exceptionally durable hostname, especially if it would form the 890 prefix for all the provider's URL-based external names. Registering 891 an opaque hostname in the ".org" or ".net" domain would not be a bad 892 start. Another way is to publish your ARKs with an organizational 893 domain name that will be mapped by DNS to an appropriate NMA host. 894 This makes for shorter names with less branding vulnerability. 896 It is a mistake to think that hostnames are inherently unstable. If 897 you require brand visibility, that may be a fact of life. But things 898 are easier if yours is the brand of long-lived cultural memory 899 institution such as a national or university library or archive. 900 Well-chosen hostnames from organizations that are sheltered from the 901 direct effects of a volatile marketplace can easily provide longer- 902 lived global resolvers than the domain names explicitly or implicitly 903 used as starting points for global resolution by indirection-based 904 persistent identifier schemes. For example, it is hard to imagine 905 circumstances under which the Library of Congress' domain name would 906 disappear sooner than, say, "handle.net". 908 For smaller libraries, archives, and preservation organizations, 909 there is a natural concern about whether they will be able to keep 910 their web servers and domain names in the face of uncertain funding. 911 One option is to form or join a consortium [N2T] of like-minded 912 organizations with the purpose of providing mutual preservation 913 support. The first goal of such a consortium would be to perpetually 914 rent a hostname on which to establish a web server that simply 915 redirects incoming member organization requests to the appropriate 916 member server; using ARKs, for example, a 150-member consortium could 917 run a very small server (24x7) that contained nothing more than 150 918 rewrite rules in its configuration file. Even more helpful would be 919 additional consortial support for a member organization that was 920 unable to continue providing services and needed to find a successor 921 archival organization. This would be a low-cost, low-tech way to 922 publish ARKs (or URLs) under highly persistent hostnames. 924 There are no obvious reasons why the organizations registering DNS 925 names, URN Namespaces, and DOI publisher IDs should have among them 926 one that is intrinsically more fallible than the next. Moreover, it 927 is a misconception that the demise of DNS and of HTTP need adversely 928 affect the persistence of URLs. At such a time, certainly URLs from 929 the present day might not then be actionable by our present-day 930 mechanisms, but resolution systems for future non-actionable URLs are 931 no harder to imagine than resolution systems for present-day non- 932 actionable URNs and DOIs. There is no more stable a namespace than 933 one that is dead and frozen, and that would then characterize the 934 space of names bearing the "http://" prefix. It is useful to 935 remember that just because hostnames have been carelessly chosen in 936 their brief history does not mean that they are unsuitable in NMAHs 937 (and URLs) intended for use in situations demanding the highest level 938 of persistence available in the Internet environment. A well-planned 939 name assignment strategy is everything. 941 3.5. Assigners of ARKs 943 A Name Assigning Authority (NAA) is an organization that creates (or 944 delegates creation of) long-term associations between identifiers and 945 information objects. Examples of NAAs include national libraries, 946 national archives, and publishers. An NAA may arrange with an 947 external organization for identifier assignment. The US Library of 948 Congress, for example, allows OCLC (the Online Computer Library 949 Center, a major world cataloger of books) to create associations 950 between Library of Congress call numbers (LCCNs) and the books that 951 OCLC processes. A cataloging record is generated that testifies to 952 each association, and the identifier is included by the publisher, 953 for example, in the front matter of a book. 955 An NAA does not so much create an identifier as create an 956 association. The NAA first draws an unused identifier string from 957 its namespace, which is the set of all identifiers under its control. 958 It then records the assignment of the identifier to an information 959 object having sundry witnessed characteristics, such as a particular 960 author and modification date. A namespace is usually reserved for an 961 NAA by agreement with recognized community organizations (such as 962 IANA and ISO) that all names containing a particular string be under 963 its control. In the ARK an NAA is represented by the Name Assigning 964 Authority Number (NAAN). 966 The ARK namespace reserved for an NAA is the set of names bearing its 967 particular NAAN. For example, all strings beginning with "ark:/ 968 12025/" are under control of the NAA registered under 12025, which 969 might be the National Library of Finland. Because each NAA has a 970 different NAAN, names from one namespace cannot conflict with those 971 from another. Each NAA is free to assign names from its namespace 972 (or delegate assignment) according to its own policies. These 973 policies must be documented in a manner similar to the declarations 974 required for URN Namespace registration [RFC2611]. 976 To register for a NAAN, please read about the mapping authority 977 discovery file in the next section and send email to ark@cdlib.org. 979 3.6. NAAN Namespace Management 981 Every NAA must have a namespace management strategy. A time-honored 982 technique is to hierarchically partition a namespace into 983 subnamespaces using prefixes that guarantee non-collision of names in 984 different partition. This practice is strongly encouraged for all 985 NAAs, especially when subnamespace management will be delegated to 986 other departments, units, or projects within an organization. For 987 example, with a NAAN that is assigned to a university and managed by 988 its main library, care should be taken to reserve semantically opaque 989 prefixes that will set aside large parts of the unused namespace for 990 future assignments. Prefix-based partition management is an 991 important responsibility of the NAA. 993 This sort of delegation by prefix is well-used in the formation of 994 DNS names and ISBN identifiers. An important difference is that in 995 the former, the hierarchy is deliberately exposed and in the latter 996 it is hidden. Rather than using lexical boundary markers such as the 997 period (`.') found in domain names, the ISBN uses a publisher prefix 998 but doesn't disclose where the prefix ends and the publisher's 999 assigned name begins. This practice of non-disclosure, borrowed from 1000 the ISBN and ISSN schemes, is encouraged in assigning ARKs, because 1001 it reduces the visibility of an assertion that is probably not 1002 important now and may become a vulnerability later. 1004 Reasonable prefixes for assigned names usually consist of consonants 1005 and digits and are 1-5 characters in length. For example, the 1006 constant prefix "x9t" might be delegated to a book digitization 1007 project that creates identifiers such as 1009 http://444.berkeley.edu/ark:/28722/x9t38rk45c 1011 If longevity is the goal, it is important to keep the prefixes free 1012 of recognizable semantics; for example, using an acronym representing 1013 a project or a department is discouraged. At the same time, you may 1014 wish to set aside a subnamespace for testing purposes under a prefix 1015 such as "fk..." that can serve as a visual clue and reminder to 1016 maintenance staff that this "fake" identifier was never published. 1018 There are other measures one can take to avoid user confusion, 1019 transcription errors, and the appearance of accidental semantics when 1020 creating identifiers. If you are generating identifiers 1021 automatically, pure numeric identifiers are likeley to be 1022 semantically opaque enough, but it's probably useful to avoid leading 1023 zeroes because some users mistakenly treat them as optional, thinking 1024 (arithmetically) that they don't contribute to the "value" of the 1025 identifier. 1027 If you need lots of identifiers and you don't want them to get too 1028 long, you can mix digits with consonants (but avoid vowels since they 1029 might accidentally spell words) to get more identifiers without 1030 increasing the string length. In this case you may not want more 1031 than a two letters in a row because it reduces the chance of 1032 generating acronyms. Generator tools such as [NOID] provide support 1033 for these sorts of identifiers, and can also add a computed check 1034 character as a guarantee against the most common transcription 1035 errors. 1037 3.7. Sub-Object Naming 1039 As mentioned previously, semantically opaque identifiers are very 1040 useful for long-term naming of abstract objects, however, it may be 1041 appropriate to extend these names with less opaque extensions that 1042 reference contemporary service entry points (sub-objects) in support 1043 of the object. Sub-object extensions beginning with a digit or 1044 underscore (`_') are reserved for the possibilty of developing a 1045 future registry of canonical service points (e.g., numeric references 1046 to versions, formats, languages, etc). 1048 4. Finding a Name Mapping Authority 1050 In order to derive an actionable identifier (these days, a URL) from 1051 an ARK, a hostport (hostname or hostname plus port combination) for a 1052 working Name Mapping Authority (NMA) must be found. An NMA is a 1053 service that is able to respond to the three basic ARK service 1054 requests. Relying on registration and client-side discovery, NMAs 1055 make known which NAAs' identifiers they are willing to service. 1057 Upon encountering an ARK, a user (or client software) looks inside it 1058 for the optional NMAH part (the hostport of the NMA's ARK service). 1059 If it contains an NMAH that is working, this NMAH discovery step may 1060 be skipped; the NMAH effectively uses the beginning of an ARK to 1061 cache the results of a prior mapping authority discovery process. If 1062 a new NMAH needs to found, the client looks inside the ARK again for 1063 the NAAN (Name Assigning Authority Number). Querying a global 1064 database, it then uses the NAAN to look up all current NMAHs that 1065 service ARKs issued by the identified NAA. 1067 The global database is key, and ideally the lookup would be automatic 1068 and transparent to the user. For this, the most promising method is 1069 probably the Name-to-Thing (N2T) Resolver [N2T] at n2t.info. It is a 1070 proposed low-cost, highly reliable, consortially maintained NMAH that 1071 simply exists to support actionable HTTP-based URLs for as long as 1072 HTTP is used. One of its big advantages over the other two methods 1073 and the URN, Handle, DOI, and PURL methods, is that N2T addresses the 1074 namespace splitting problem. When objects maintained by one NMA are 1075 inherited by more than one successor NMA, until now one of those 1076 successors would be required to maintain forwarding tables on behalf 1077 of the other successors. 1079 There are two other ways to discover an NMAH, one of them described 1080 in a subsection below. Another way, described in an appendix, is 1081 based on a simplification of the URN resolver discovery method, 1082 itself very similar in principle to the resolver discovery method 1083 used by Handles and DOIs. None of these methods does more than what 1084 can be done with a very small, consortially maintained web server 1085 such as [N2T]. 1087 In the interests of long-term persistence, however, ARK mechanisms 1088 are first defined in high-level, protocol-independent terms so that 1089 mechanisms may evolve and be replaced over time without compromising 1090 fundamental service objectives. Either or both specific methods 1091 given here may eventually be supplanted by better methods since, by 1092 design, the ARK scheme does not depend on a particular method, but 1093 only on having some method to locate an active NMAH. 1095 At the time of issuance, at least one NMAH for an ARK should be 1096 prepared to service it. That NMA may or may not be administered by 1097 the Name Assigning Authority (NAA) that created it. Consider the 1098 following hypothetical example of providing long-term access to a 1099 cancer research journal. The publisher wishes to turn a profit and 1100 the National Library of Medicine wishes to preserve the scholarly 1101 record. An agreement might be struck whereby the publisher would act 1102 as the NAA and the national library would archive the journal issue 1103 when it appears, but without providing direct access for the first 1104 six months. During the first six months of peak commercial 1105 viability, the publisher would retain exclusive delivery rights and 1106 would charge access fees. Again, by agreement, both the library and 1107 the publisher would act as NMAs, but during that initial period the 1108 library would redirect requests for issues less than six months old 1109 to the publisher. At the end of the waiting period, the library 1110 would then begin servicing requests for issues older than six months 1111 by tapping directly into its own archives. Meanwhile, the publisher 1112 might routinely redirect incoming requests for older issues to the 1113 library. Long-term access is thereby preserved, and so is the 1114 commercial incentive to publish content. 1116 Although it will be common for an NAA also to run an NMA service, it 1117 is never a requirement. Over time NAAs and NMAs will come and go. 1118 One NMA will succeed another, and there might be many NMAs serving 1119 the same ARKs simultaneously (e.g., as mirrors or as competitors). 1120 There might also be asymmetric but coordinated NMAs as in the 1121 library-publisher example above. 1123 4.1. Looking Up NMAHs in a Globally Accessible File 1125 This subsection describes a way to look up NMAHs using a simple name 1126 authority table represented as a plain text file. For efficient 1127 access the file may be stored in a local filesystem, but it needs to 1128 be reloaded periodically to incorporate updates. It is not expected 1129 that the size of the file or frequency of update should impose an 1130 undue maintenance or searching burden any time soon, for even 1131 primitive linear search of a file with ten-thousand NAAs is a 1132 subsecond operation on modern server machines. The proposed file 1133 strategy is similar to the /etc/hosts file strategy that supported 1134 Internet host address lookup for a period of years before the advent 1135 of DNS. 1137 The name authority table file is updated on an ongoing basis and is 1138 available for copying over the internet from the California Digital 1139 Library at http://www.cdlib.org/inside/diglib/ark/natab and from a 1140 number of mirror sites. The file contains comment lines (lines that 1141 begin with `#') explaining the format and giving the file's 1142 modification time, reloading address, and NAA registration 1143 instructions. There is even a Perl script that processes the file 1144 embedded in the file's comments. The currently registered Name 1145 Assigning Authorities are: 1147 12025 National Library of Medicine 1148 12026 Library of Congress 1149 12027 National Agriculture Library 1150 13030 California Digital Library 1151 13038 World Intellectual Property Organization 1152 20775 University of California San Diego 1153 29114 University of California San Francisco 1154 28722 University of California Berkeley 1155 21198 University of California Los Angeles 1156 15230 Rutgers University 1157 13960 Internet Archive 1158 64269 Digital Curation Centre 1159 62624 New York University 1160 67531 University of North Texas 1161 27927 Ithaka Electronic-Archiving Initiative 1162 12148 Bibliotheque nationale de France 1163 / National Library of France 1164 78319 Google 1165 88435 Princeton University 1166 78428 University of Washington 1167 89901 Archives of the Region of Vaestra Goetaland 1168 and City of Gothenburg, Sweden 1169 80444 Northwest Digital Archives 1170 25593 Emory University 1171 25031 University of Kansas 1172 17101 Centre for Ecology & Hydrology, UK 1173 65323 University of Calgary 1174 61001 University of Chicago 1175 52327 Bibliotheque et Archives Nationales du Quebec 1176 / National Libary and Archives of Quebec 1177 39331 National Szechenyi Library / National Library of Hungary 1178 26677 Library and Archives Canada / Bibliotheque et Archives Canada 1180 5. Generic ARK Service Definition 1182 An ARK request's output is delivered information; examples include 1183 the object itself, a policy declaration (e.g., a promise of support), 1184 a descriptive metadata record, or an error message. The experience 1185 of object delivery is expected to be an evolving mix of information 1186 that reflects changing service expectations and technology 1187 requirements; contemporary examples include such things as an object 1188 summary and component links formatted for human consumption. ARK 1189 services must be couched in high-level, protocol-independent terms if 1190 persistence is to outlive today's networking infrastructural 1191 assumptions. The high-level ARK service definitions listed below are 1192 followed in the next section by a concrete method (one of many 1193 possible methods) for delivering these services with today's 1194 technology. 1196 5.1. Generic ARK Access Service (access, location) 1198 Returns (a copy of) the object or a redirect to the same, although a 1199 sensible object proxy may be substituted. Examples of sensible 1200 substitutes include, 1202 o a table of contents instead of a large complex document, 1204 o a home page instead of an entire web site hierarchy, 1206 o a rights clearance challenge before accessing protected data, 1208 o directions for access to an offline object (e.g., a book), 1210 o a description of an intangible object (a disease, an event), or 1212 o an applet acting as "player" for a large multimedia object. 1214 May also return a discriminated list of alternate object locators. 1215 If access is denied, returns an explanation of the object's current 1216 (perhaps permanent) inaccessibility. 1218 5.1.1. Generic Policy Service (permanence, naming, etc.) 1220 Returns declarations of policy and support commitments for given 1221 ARKs. Declarations are returned in either a structured metadata 1222 format or a human readable text format; sometimes one format may 1223 serve both purposes. Policy subareas may be addressed in separate 1224 requests, but the following areas should should be covered: object 1225 permanence, object naming, object fragment addressing, and 1226 operational service support. 1228 The permanence declaration for an object is a rating defined with 1229 respect to an identified permanence provider (guarantor), which will 1230 be the NMA. It may include the following aspects. 1232 (a) "object availability" -- whether and how access to the object 1233 is supported (e.g., online 24x7, or offline only), 1235 (b) "identifier validity" -- under what conditions the identifier 1236 will be or has been re-assigned, 1238 (c) "content invariance" -- under what conditions the content of 1239 the object is subject to change, and 1241 (d) "change history" -- access to corrections, migrations, and 1242 revisions, whether through links to the changed objects themselves 1243 or through a document summarizing the change history 1245 One approach to a permanence rating framework, conceived 1246 independently from ARKs, is given in [NLMPerm]. Under ongoing 1247 development and limited deployment at the US National Library of 1248 Medicine, it identifies the following "permanence levels": 1250 Not Guaranteed: No commitment has been made to retain this 1251 resource. It could become unavailable at any time. Its 1252 identifier could be changed. 1254 Permanent: Dynamic Content: A commitment has been made to keep 1255 this resource permanently available. Its identifier will always 1256 provide access to the resource. Its content could be revised or 1257 replaced. 1259 Permanent: Stable Content: A commitment has been made to keep this 1260 resource permanently available. Its identifier will always 1261 provide access to the resource. Its content is subject only to 1262 minor corrections or additions. 1264 Permanent: Unchanging Content: A commitment has been made to keep 1265 this resource permanently available. Its identifier will always 1266 provide access to the resource. Its content will not change. 1268 Naming policy for an object includes an historical description of the 1269 NAA's (and its successor NAA's) policies regarding differentiation of 1270 objects. Since it the NMA who responds to requests for policy 1271 statements, it is useful for the NMA to be able to produce or 1272 summarize these historical NAA documents. Naming policy may include 1273 the following aspects. 1275 (i) "similarity" -- (or "unity") the limit, defined by the NAA, to 1276 the level of dissimilarity beyond which two similar objects 1277 warrant separate identifiers but before which they share one 1278 single identifier, and 1280 (ii) "granularity" -- the limit, defined by the NAA, to the level 1281 of object subdivision beyond which sub-objects do not warrant 1282 separately assigned identifiers but before which sub-objects are 1283 assigned separate identifiers. 1285 Subnaming policy for an object describes the qualifiers that the NMA, 1286 in fulfilling its ongoing and evolving service obligations, allows as 1287 extensions to an NAA-assigned ARK. To the conceptual object that the 1288 NAA named with an ARK, the NMA may add component access points and 1289 derivatives (e.g., format migrations in aid of preservation) in order 1290 to provide both basic and value-added services. 1292 Addressing policy for an object includes a description of how, during 1293 access, object components (e.g., paragraphs, sections) or views 1294 (e.g., image conversions) may or may not be "addressed", in other 1295 words, how the NMA permits arguments or parameters to modify the 1296 object delivered as the result of an ARK request. If supported, 1297 these sorts of operations would provide things like byte-ranged 1298 fragment delivery and open-ended format conversions, or any set of 1299 possible transformations that would be too numerous to list or to 1300 identify with separately assigned ARKs. 1302 Operational service support policy includes a description of general 1303 operational aspects of the NMA service, such as after-hours staffing 1304 and trouble reporting procedures. 1306 5.1.2. Generic Description Service 1308 Returns a description of the object. Descriptions are returned in 1309 either a structured metadata format or a human readable text format; 1310 sometimes one format may serve both purposes. A description must at 1311 a minimum answer the who, what, when, and where questions concerning 1312 an expression of the object. Standalone descriptions should be 1313 accompanied by the modification date and source of the description 1314 itself. May also return discriminated lists of ARKs that are related 1315 to the given ARK. 1317 5.2. Overview of The HTTP URL Mapping Protocol (THUMP) 1319 The HTTP URL Mapping Protocol (THUMP) is a way of taking a key (any 1320 identifier) and asking such questions as, what information does this 1321 identify and how permanent is it? [THUMP] is in fact one specific 1322 method under development for delivering ARK services. The protocol 1323 runs over HTTP to exploit the web browser's current pre-eminence as 1324 user interface to the Internet. THUMP is designed so that a person 1325 can enter ARK requests directly into the location field of current 1326 browser interfaces. Because it runs over HTTP, THUMP can be 1327 simulated and tested via keyboard-based interactions [RFC0854]. 1329 The asker (a person or client program) starts with an identifier, 1330 such as an ARK or a URL. The identifier reveals to the asker (or 1331 allows the asker to infer) the Internet host name and port number of 1332 a server system that responds to questions. Here, this is just the 1333 NMAH that is obtained by inspection and possibly lookup based on the 1334 ARK's NAAN. The asker then sets up an HTTP session with the server 1335 system, sends a question via a THUMP request (contained within an 1336 HTTP request), receives an answer via a THUMP response (contained 1337 within an HTTP response), and closes the session. That concludes the 1338 connected portion of the protocol. 1340 A THUMP request is a string of characters beginning with a `?' 1341 (question mark) that is appended to the identifier string. The 1342 resulting string is sent as an argument to HTTP's GET command. 1343 Request strings too long for GET may be sent using HTTP's POST 1344 command. The three most common requests correspond to three 1345 degenerate special cases that keep the user's learning and typing 1346 burden low. First, a simple key with no request at all is the same 1347 as an ordinary access request. Thus a plain ARK entered into a 1348 browser's location field behaves much like a plain URL, and returns 1349 access to the primary identified object, for instance, an HTML 1350 document. 1352 The second special case is a minimal ARK description request string 1353 consisting of just "?". For example, entering the string, 1355 ark.nlm.nih.gov/12025/psbbantu? 1357 into the browser's location field directly precipitates a request for 1358 a metadata record describing the object identified by ark:/12025/ 1359 psbbantu. The browser, unaware of THUMP, prepares and sends an HTTP 1360 GET request in the same manner as for a URL. THUMP is designed so 1361 that the response (indicated by the returned HTTP content type) is 1362 normally displayed, whether the output is structured for machine 1363 processing (text/plain) or formatted for human consumption (text/ 1364 html). 1366 In the following example THUMP session, each line has been annotated 1367 to include a line number and whether it was the client or server that 1368 sent it. Without going into much depth, the session has four pieces 1369 separated from each other by blank lines: the client's piece (lines 1370 1-3), the server's HTTP/THUMP response headers (4-7), and the body of 1371 the server's response (8-13). The first and last lines (1 and 13) 1372 correspond to the client's steps to start the TCP session and the 1373 server's steps to end it, respectively. 1375 1 C: [opens session] 1376 C: GET http://ark.nlm.nih.gov/ark:/12025/psbbantu? HTTP/1.1 1377 C: 1378 S: HTTP/1.1 200 OK 1379 5 S: Content-Type: text/plain 1380 S: THUMP-Status: 0.6 200 OK 1381 S: 1382 S: erc: 1383 S: who: Lederberg, Joshua 1384 10 S: what: Studies of Human Families for Genetic Linkage 1385 S: when: 1974 1386 S: where: http://profiles.nlm.nih.gov/BB/A/N/T/U/_/bbantu.pdf 1387 S: [closes session] 1389 The first two server response lines (4-5) above are typical of HTTP. 1390 The next line (6) is peculiar to THUMP, and indicates the THUMP 1391 version and a normal return status. 1393 The balance of the response consists of a single metadata record 1394 (8-12) that comprises the ARK description service response. The 1395 returned record is in the format of an Electronic Resource Citation 1396 [ERC], which is discussed in overview in the next section. For now, 1397 note that it contains four elements that answer the top priority 1398 questions regarding an expression of the object: who played a major 1399 role in expressing it, what the expression was called, when is was 1400 created, and where the expression may be found. This quartet of 1401 elements comes up again and again in ERCs. 1403 The third degenerate special case of an ARK request (and no other 1404 cases will be described in this document) is the string "??", 1405 corresponding to a minimal permanence policy request. It can be seen 1406 in use appended to an ARK (on line 2) in the example session that 1407 follows. 1409 1 C: [opens session] 1410 C: GET http://ark.nlm.nih.gov/ark:/12025/psbbantu?? HTTP/1.1 1411 C: 1412 S: HTTP/1.1 200 OK 1413 5 S: Content-Type: text/plain 1414 S: THUMP-Status: 0.6 200 OK 1415 S: 1416 S: erc: 1417 S: who: Lederberg, Joshua 1418 10 S: what: Studies of Human Families for Genetic Linkage 1419 S: when: 1974 1420 S: where: http://profiles.nlm.nih.gov/BB/A/N/T/U/_/bbantu.pdf 1421 S: erc-support: 1422 S: who: USNLM 1423 15 S: what: Permanent, Unchanging Content 1424 S: when: 20010421 1425 S: where: http://ark.nlm.nih.gov/yy22948 1426 S: [closes session] 1428 Each segment in an ERC tells a different story relating to the 1429 object, so although the same four questions (elements) appear in 1430 each, the answers depend on the segment's story type. While the 1431 first segment tells the story of an expression of the object, the 1432 second segment tells the story of the support commitment made to it: 1433 who made the commitment, what the nature of the commitment was, when 1434 it was made, and where a fuller explanation of the commitment may be 1435 found. 1437 5.3. The Electronic Resource Citation (ERC) 1439 An Electronic Resource Citation (or ERC, pronounced e-r-c) [ERC] is a 1440 kind of object description that uses Dublin Core Kernel metadata 1441 elements [DCKernel]. The ERC with Kernel elements provides a simple, 1442 compact, and printable record for holding data associated with an 1443 information resource. As originally designed [Kernel], Kernel 1444 metadata balances the needs for expressive power, very simple machine 1445 processing, and direct human manipulation. 1447 The previous section shows two limited examples of what is fully 1448 described elsewhere [ERC]. The rest of this short section provides 1449 some of the background and rationale for this record format. 1451 A founding principle of Kernel metadata is that direct human contact 1452 with metadata will be a necessary and sufficient condition for the 1453 near term rapid development of metadata standards, systems, and 1454 services. Thus the machine-processable Kernel elements must only 1455 minimally strain people's ability to read, understand, change, and 1456 transmit ERCs without their relying on intermediation with 1457 specialized software tools. The basic ERC needs to be succinct, 1458 transparent, and trivially parseable by software. 1460 In the current Internet, it is natural seriously to consider using 1461 XML as an exchange format because of predictions that it will obviate 1462 many ad hoc formats and programs, and unify much of the world's 1463 information under one reliable data structuring discipline that is 1464 easy to generate, verify, parse, and render. It appears, however, 1465 that XML is still only catching on after years of standards work and 1466 implementation experience. The reasons for it are unclear, but for 1467 now very simple XML interpretation is still out of reach. Another 1468 important caution is that XML structures are hard on the eyeballs, 1469 taking up an amount of display (and page) space that significantly 1470 exceeds that of traditional formats. Until these conflicts with ERC 1471 principle are resolved, XML is not a first choice for representing 1472 ERCs. Borrowing instead from the data structuring format that 1473 underlies the successful spread of email and web services, the first 1474 ERC format uses [ANVL], which is based on email and HTTP headers 1475 [RFC2822]. There is a naturalness to ANVL's label-colon-value format 1476 (seen in the previous section) that barely needs explanation to a 1477 person beginning to enter ERC metadata. 1479 Besides simplicity of ERC system implementation and data entry 1480 mechanics, ERC semantics (what the record and its constituent parts 1481 mean) must also be easy to explain. ERC semantics are based on a 1482 reformulation and extension of the Dublin Core [RFC5013] hypothesis, 1483 which suggests that the fifteen Dublin Core metadata elements have a 1484 key role to play in cross-domain resource description. The ERC 1485 design recognizes that the Dublin Core's primary contribution is the 1486 international, interdisciplinary consensus that identified fifteen 1487 semantic buckets (element categories), regardless of how they are 1488 labeled. The ERC then adds a definition for a record and some 1489 minimal compliance rules. In pursuing the limits of simplicity, the 1490 ERC design combines and relabels some Dublin Core buckets to isolate 1491 a tiny kernel (subset) of four elements for basic cross-domain 1492 resource description. 1494 For the cross-domain kernel, the ERC uses the four basic elements -- 1495 who, what, when, and where -- to pretend that every object in the 1496 universe can have a uniform minimal description. Each has a name or 1497 other identifier, a location, some responsible person or party, and a 1498 date. It doesn't matter what type of object it is, or whether one 1499 plans to read it, interact with it, smoke it, wear it, or navigate 1500 it. Of course, this approach is flawed because uniformity of 1501 description for some object types requires more semantic contortion 1502 and sacrifice than for others. That is why at the beginning of this 1503 document, the ARK was said to be suited to objects that accommodate 1504 reasonably regular electronic description. 1506 While insisting on uniformity at the most basic level provides 1507 powerful cross-domain leverage, the semantic sacrifice is great for 1508 many applications. So the ERC also permits a semantically rich and 1509 nuanced description to co-exist in a record along with a basic 1510 description. In that way both sophisticated and naive recipients of 1511 the record can extract the level of meaning from it that best suits 1512 their needs and abilities. Key to unlocking the richer description 1513 is a controlled vocabulary of ERC record types (not explained in this 1514 document) that permit knowledgeable recipients to apply defined sets 1515 of additional assumptions to the record. 1517 5.4. Advice to Web Clients 1519 ARKs are envisaged to appear wherever durable object references are 1520 planned. Library cataloging records, literature citations, and 1521 bibliographies are important examples. In many of these places URLs 1522 (Uniform Resource Locators) are currently used, and inside some of 1523 those URLs are embedded URNs, Handles, and DOIs. Unfortunately, 1524 there's no suggestion of a way to probe for extra services that would 1525 build confidence in those identifiers; in other words, there's no way 1526 to tell whether any of those identifiers is any better managed than 1527 the average URL. 1529 ARKs are also envisaged to appear in hypertext links (where they are 1530 not normally shown to users) and in rendered text (displayed or 1531 printed). A normal HTML link for which the URL is not displayed 1532 looks like this. 1534 Click Here 1536 A URL with an embedded ARK invites access (via `?' and `??') to extra 1537 services: 1539 Click Here 1541 Using the [N2T] resolver to provide identifier-scheme-agnostic 1542 protection against hostname instability, this ARK could be published 1543 as: 1545 Click Here 1547 An NAA will typically make known the associations it creates by 1548 publishing them in catalogs, actively advertizing them, or simply 1549 leaving them on web sites for visitors (e.g., users, indexing 1550 spiders) to stumble across in browsing. 1552 5.5. Security Considerations 1554 The ARK naming scheme poses no direct risk to computers and networks. 1555 Implementors of ARK services need to be aware of security issues when 1556 querying networks and filesystems for Name Mapping Authority 1557 services, and the concomitant risks from spoofing and obtaining 1558 incorrect information. These risks are no greater for ARK mapping 1559 authority discovery than for other kinds of service discovery. For 1560 example, recipients of ARKs with a specified hostport (NMAH) should 1561 treat it like a URL and be aware that the identified ARK service may 1562 no longer be operational. 1564 Apart from mapping authority discovery, ARK clients and servers 1565 subject themselves to all the risks that accompany normal operation 1566 of the protocols underlying mapping services (e.g., HTTP, Z39.50). 1567 As specializations of such protocols, an ARK service may limit 1568 exposure to the usual risks. Indeed, ARK services may enhance a kind 1569 of security by helping users identify long-term reliable references 1570 to information objects. 1572 6. References 1574 [ANVL] Kunze, J. and B. Kahle, "A Name-Value Language", 2008, 1575 . 1577 [ARK] Kunze, J., "Towards Electronic Persistence Using ARK 1578 Identifiers", IWAW/ECDL Annual Workshop Proceedings 3rd, 1579 August 2003, 1580 . 1582 [DCKernel] 1583 DCMI, "Kernel Metadata Working Group", 2001-2008, 1584 . 1586 [DOI] IDF, "The Digital Object Identifier (DOI) System", 1587 February 2001, . 1589 [ERC] Kunze, J. and A. Turner, "Kernel Metadata and Electronic 1590 Resource Citations", October 2007, 1591 . 1593 [Handle] Lannom, L., "Handle System Overview", ICSTI Forum No. 30, 1594 April 1999, . 1596 [Kernel] Kunze, J., "A Metadata Kernel for Electronic Permanence", 1597 Journal of Digital Information Vol 2, Issue 2, ISSN 1368- 1598 7506, January 2002, 1599 . 1601 [N2T] CDL, "Name-to-Thing Resolver", August 2006, 1602 . 1604 [NLMPerm] Byrnes, M., "Defining NLM's Commitment to the Permanence 1605 of Electronic Information", ARL 212:8-9, October 2000, 1606 . 1608 [NOID] Kunze, J., "Nice Opaque Identifiers", February 2005, 1609 . 1611 [PURL] Shafer, K., "Introduction to Persistent Uniform Resource 1612 Locators", 1996, . 1614 [RFC0854] Postel, J. and J. Reynolds, "Telnet Protocol 1615 Specification", STD 8, RFC 854, May 1983. 1617 [RFC1034] Mockapetris, P., "Domain names - concepts and facilities", 1618 STD 13, RFC 1034, November 1987. 1620 [RFC2141] Moats, R., "URN Syntax", RFC 2141, May 1997. 1622 [RFC2288] Lynch, C., Preston, C., and R. Jr, "Using Existing 1623 Bibliographic Identifiers as Uniform Resource Names", 1624 RFC 2288, February 1998. 1626 [RFC2611] Daigle, L., van Gulik, D., Iannella, R., and P. Faltstrom, 1627 "URN Namespace Definition Mechanisms", BCP 33, RFC 2611, 1628 June 1999. 1630 [RFC2616] Fielding, R., Gettys, J., Mogul, J., Frystyk, H., 1631 Masinter, L., Leach, P., and T. Berners-Lee, "Hypertext 1632 Transfer Protocol -- HTTP/1.1", RFC 2616, June 1999. 1634 [RFC2822] Resnick, P., "Internet Message Format", RFC 2822, 1635 April 2001. 1637 [RFC2915] Mealling, M. and R. Daniel, "The Naming Authority Pointer 1638 (NAPTR) DNS Resource Record", RFC 2915, September 2000. 1640 [RFC3986] Berners-Lee, T., Fielding, R., and L. Masinter, "Uniform 1641 Resource Identifier (URI): Generic Syntax", STD 66, 1642 RFC 3986, January 2005. 1644 [RFC5013] Kunze, J. and T. Baker, "The Dublin Core Metadata Element 1645 Set", RFC 5013, August 2007. 1647 [THUMP] Gamiel, K. and J. Kunze, "The HTTP URL Mapping Protocol", 1648 August 2007, 1649 . 1651 Appendix A. ARK Maintenance Agency 1653 Production settings in which ARKs are used include the University of 1654 California, the National Library of France, the Internet Archive, and 1655 Portico, with maintenance based at the California Digital Library 1656 (CDL), housed at the University of California Office of the 1657 President. 1659 http://ark.cdlib.org/ 1661 Appendix B. Looking up NMAHs Distributed via DNS 1663 This subsection introduces an older method for looking up NMAHs that 1664 is based on the method for discovering URN resolvers described in 1665 [RFC2915]. It relies on querying the DNS system already installed in 1666 the background infrastructure of most networked computers. A query 1667 is submitted to DNS asking for a list of resolvers that match a given 1668 NAAN. DNS distributes the query to the particular DNS servers that 1669 can best provide the answer, unless the answer can be found more 1670 quickly in a local DNS cache as a side-effect of a recent query. 1671 Responses come back inside Name Authority Pointer (NAPTR) records. 1672 The normal result is one or more candidate NMAHs. 1674 In its full generality the [RFC2915] algorithm ambitiously 1675 accommodates a complex set of preferences, orderings, protocols, 1676 mapping services, regular expression rewriting rules, and DNS record 1677 types. This subsection proposes a drastic simplification of it for 1678 the special case of ARK mapping authority discovery. The simplified 1679 algorithm is called Maptr. It uses only one DNS record type (NAPTR) 1680 and restricts most of its field values to constants. The following 1681 hypothetical excerpt from a DNS data file for the NAAN known as 12026 1682 shows three example NAPTR records ready to use with the Maptr 1683 algorithm. 1685 12026.ark.arpa. 1686 ;; US Library of Congress 1687 ;; order pref flags service regexp replacement 1688 IN NAPTR 0 0 "h" "ark" "USLC" lhc.nlm.nih.gov:8080 1689 IN NAPTR 0 0 "h" "ark" "USLC" foobar.zaf.org 1690 IN NAPTR 0 0 "h" "ark" "USLC" sneezy.dopey.com 1692 All the fields are held constant for Maptr except for the "flags", 1693 "regexp", and "replacement" fields. The "service" field contains the 1694 constant value "ark" so that NAPTR records participating in the Maptr 1695 algorithm will not be confused with other NAPTR records. The "order" 1696 and "pref" fields are held to 0 (zero) and otherwise ignored for now; 1697 the algorithm may evolve to use these fields for ranking decisions 1698 when usage patterns and local administrative needs are better 1699 understood. 1701 When a Maptr query returns a record with a flags field of "h" (for 1702 hostport, a Maptr extension to the NAPTR flags), the replacement 1703 field contains the NMAH (hostport) of an ARK service provider. When 1704 a query returns a record with a flags field of "" (the empty string), 1705 the client needs to submit a new query containing the domain name 1706 found in the replacement field. This second sort of record exploits 1707 the distributed nature of DNS by redirecting the query to another 1708 domain name. It looks like this. 1710 12345.ark.arpa. 1711 ;; Digital Library Consortium 1712 ;; order pref flags service regexp replacement 1713 IN NAPTR 0 0 "" "ark" "" dlc.spct.org. 1715 Here is the Maptr algorithm for ARK mapping authority discovery. In 1716 it replace with the NAAN from the ARK for which an NMAH is 1717 sought. 1719 1. Initialize the DNS query: type=NAPTR, query=.ark.arpa. 1721 2. Submit the query to DNS and retrieve (NAPTR) records, discarding 1722 any record that does not have "ark" for the service field. 1724 3. All remaining records with a flags fields of "h" contain 1725 candidate NMAHs in their replacement fields. Set them aside, if 1726 any. 1728 4. Any record with an empty flags field ("") has a replacement field 1729 containing a new domain name to which a subsequent query should 1730 be redirected. For each such record, set query= 1731 then go to step (2). When all such records have been recursively 1732 exhausted, go to step (5). 1734 5. All redirected queries have been resolved and a set of candidate 1735 NMAHs has been accumulated from steps (3). If there are zero 1736 NMAHs, exit -- no mapping authority was found. If there is one 1737 or more NMAH, choose one using any criteria you wish, then exit. 1739 A Perl script that implements this algorithm is included here. 1741 #!/depot/bin/perl 1743 use Net::DNS; # include simple DNS package 1744 my $qtype = "NAPTR"; # initialize query type 1745 my $naa = shift; # get NAAN script argument 1746 my $mad = new Net::DNS::Resolver; # mapping authority discovery 1748 &maptr("$naa.ark.arpa"); # call maptr - that's it 1750 sub maptr { # recursive maptr algorithm 1751 my $dname = shift; # domain name as argument 1752 my ($rr, $order, $pref, $flags, $service, $regexp, 1753 $replacement); 1754 my $query = $mad->query($dname, $qtype); 1755 return # non-productive query 1756 if (! $query || ! $query->answer); 1757 foreach $rr ($query->answer) { 1758 next # skip records of wrong type 1759 if ($rr->type ne $qtype); 1760 ($order, $pref, $flags, $service, $regexp, 1761 $replacement) = split(/\s/, $rr->rdatastr); 1762 if ($flags eq "") { 1763 &maptr($replacement); # recurse 1764 } elsif ($flags eq "h") { 1765 print "$replacement\n"; # candidate NMAH 1766 } 1767 } 1768 } 1770 The global database thus distributed via DNS and the Maptr algorithm 1771 can easily be seen to mirror the contents of the Name Authority Table 1772 file described in the previous section. 1774 Authors' Addresses 1776 John A. Kunze 1777 California Digital Library 1778 415 20th St, 4th Floor 1779 Oakland, CA 94612 1780 USA 1782 Email: jak@ucop.edu 1784 R. P. C. Rodgers 1785 University of California San Francisco 1786 Box 0134, 185 Berry, China Basin, Lobby 6 290 1787 San Francisco, CA 94143-0134 1788 USA 1790 Email: rodgers@arborvitae.com