Cliff Lynch: The Invisible Influencer in Information Infrastructure

Cliff Lynch: The Invisible Influencer in Information Infrastructure

Herbert Van de Sompel and Michael L. Nelson

Note: In the Summer of 2025, a shorter version of this blog post will be published as an essay in portal: Libraries and the Academy, in a Festschrift for Cliff Lynch, who retired from CNI in April 2025.

Abstract

The UPS Prototype was a proof-of-concept web portal built in preparation of the Universal Preprint Service Meeting held in October 1999 in Santa Fe, New Mexico. The portal provided search functionality for a set of metadata records that had been aggregated from a range of repositories that hosted preprints, working papers, and technical reports. Every search result was overlaid with a dynamically generated SFX-menu that provided a selection of value-adding links for the described scholarly work. The meeting outcome eventually led to the Open Archives Initiative and its Protocol for Metadata Harvesting (OAI-PMH), which remains widely used in scholarly communication, cultural heritage, and beyond. The SFX-menu approach became standardized as the NISO OpenURL Framework for Context-Sensitive Services (NISO OpenURL) and compliant linking servers remain operational in academic and research libraries worldwide. Both OAI-PMH and NISO OpenURL as well as associated systems and services, have been so widely deployed that they can safely be considered an integral part of the scholarly information infrastructure. The authors, who were deeply involved in devising the UPS Prototype and played core roles in the OAI-PMH and NISO OpenURL specification efforts, take the reader behind the scenes of the development of these technologies and reveal Clifford Lynch as the Invisible Influencer in the establishment of scholarly Information Infrastructure. 

Introduction

The book “Music with Roots in the Aether” bundles interviews and essays about seven contemporary American composers, including Philip Glass, Alvin Lucier, and Pauline Oliveros [1]. It was published in 2000 and is based on a 1975-1976 project by Robert Ashley – himself a contemporary American composer – for which he created 14 one-hour videotapes focusing on the creative genius of the selected composers [2]. Ashley was asked to write a foreword for the book and spends his first paragraphs emphasizing the challenge involved in reflecting on a project he did 25 years earlier:

But the Foreword turned out to be hard, even for me. I couldn’t remember who I was when the project was conceived. I couldn't remember any of the energies of the ideas that went into the project. Purposely I have not been good at remembering old ideas. I burn bridges. It keeps the path clear. [1]

Ashley’s sentiment resonates with us, because, for this 2025 essay, we impulsively chose to illustrate Clifford Lynch’s impact on the development of infrastructure for research and education by means of the Universal Preprint Service (UPS) project that we jointly initiated and executed in 1999. Some memories have remained strong, others have faded and become uncertain, and, undoubtedly a lot has just evaporated into the fabric of time. Fortunately, there are external memories that can serve as fall-backs when ours fail. Many aspects of the project and its context were documented in research papers. These papers reference documents with details about underlying discussions that are long gone from the organizational websites on which they were published, but fortunately were saved for posterity by the indispensable Internet Archive. The petites-histoires featuring people involved in our effort have not been publicly documented because the turn of the century was still social-media free. But personal touches often aptly illustrate the spirit of a project and the zeitgeist of the era in which it took place. Therefore, we feel compelled to also include a few anecdotes that remain glittering sharp in our memory. All in all, despite the fog of time, we are quite confident that the story we tell is an accurate reflection of events that were crucial to the eventual broad adoption of metadata harvesting using the Open Archives protocol (OAI-PMH) and open linking using OpenURL, and, especially, of the crucial role Cliff played in making that happen. 

Two PhD Candidates, in for a Surprise

The middle to late 1990s were exciting times for those into computers, networks, and information. Times that seemed to hold an unlimited potential, rather abruptly brought about by the combination of the HTTP/HTML Web, the mainstreaming of the Internet, affordable personal computing, and increased digitization capabilities. Like many others, we were excited about how these technologies could bring about a better world and consequently devoured Wired, a magazine that abounds with “techno-utopianism and hippie-idealism” [3]. We had jobs that presented challenges in which this powerful combination of technologies could be leveraged to imagine and implement innovative solutions. 

Herbert became systems librarian at the Ghent University Library in 1981, after completing an administrative automation project there to obtain a degree in Informatics. He didn’t exactly hit the ground running as he was trying to figure out what automation in academic libraries was all about. Most libraries were focusing their efforts on the catalogue but, given his science education, that didn’t seem to tick all the boxes.  Eventually, it was the science librarian who turned on the light by putting the automation challenge in terms of the “consultation chain”: first searching secondary sources to find journal articles, then searching catalogues to determine where the journals were, and then obtaining the articles. And so it was that, as soon as CD-ROMs became available, Herbert started providing public access to Abstract & Indexing (A&I) databases, initially on stand-alone PCs, later on PCs in Local Area Networks (LAN), and eventually on PCs across the university’s Wide Area Network. He initiated an effort to create a Belgian Union Catalogue on CD-ROM and hooked it up to the network too (Figure 1). Access dramatically improved but constraints remained: consultation was restricted to Windows PCs, the LANs had to run the Banyan Vines operating system, and networking a large collection of CD-ROMs published by a variety of vendors was a dark art. It all amounted to access being restricted to dedicated library PCs operated in departmental libraries, which was better than what most other European academic libraries had to offer but not good enough for Herbert. That is why he experienced the interoperability fabric introduced by the Web as the chains coming off regarding ways to deliver scholarly information to researchers and students. That enthusiasm resulted in the 1997 release of the Library’s Executive Lounge, a menu-driven environment that provided web-based access to all information that had previously only been available on library PCs, with the addition of some electronic journal collections for good measure. But something was still missing: the Web had links and the Executive Lounge didn’t. Herbert put it as follows:

When using a library solution, the expectations of a net-traveler are inspired by his hyperlinked Web-experiences. To such a user, it is not comprehensible that secondary sources, catalogues and primary sources, that are logically related, are not functionally linked. [4]

Figure 1 - September 1989, Ghent University Library: Herbert showing off a CD-ROM

The frustration expressed in this quote led to a collaboration with SilverPlatter and Ex Libris to implement dynamic links from journal article descriptions in A&I databases to journal holding information in the library catalogue. And it also provided fertile ground for PhD research on how to empower libraries to create links across their electronic collections by means of an open linking framework. 

Michael began his professional career at the NASA Langley Research Center (LaRC) in 1991, originally working in the Analysis and Computation Division of the supercomputer center.  Early experiences with Usenet and anonymous FTP began to divert his attention from supercomputing and cluster computing (now known as cloud computing) to information networks and libraries. In 1993, he set up an anonymous FTP server, the Langley Technical Report Server (LTRS), for technical memorandums and technical papers published by LaRC.  It effectively brought the culture to share and access technical reports via FTP, which already existed in computer science, to NASA.  Later, in 1993, he added a web interface to LTRS, providing a much-needed boost in usability. Browsing functionality improved, abstracts were indexed, and became searchable using the Z39.50-based Wide Area Information Server (WAIS), which was pretty much the only free search software at the time (for example, MySQL was not released until 1995). Around the same time, the Center for AeroSpace Information (CASI) brought their own WAIS server online; it provided abstracts for all publicly available, NASA authored reports and articles. Other centers and projects were inspired by this activity and wanted to set up their own "report server".  It became clear that a website - the term "digital library" was not yet widely adopted - allowing simultaneous WAIS search of all the NASA and NASA-affiliated report servers was needed.  A bit of Perl hacking later, by Michael and his colleagues, and the NASA Technical Report Server (NTRS) was released in 1994 (Figure 2).  

Figure 2: Summer 1999, NASA Langley Research Center: Michael among the desktop machines running LTRS and NTRS

The development of LTRS and NTRS assumed a 1:1 relationship from a metadata record to the URL of the associated full text document.  But with the progression from ".ps.Z" to ".pdf" files, the usefulness of that assumption started to break down. It became unworkable by 1998, when Michael created a separate digital library for the scanned documents of the National Advisory Committee for Aeronautics (NACA), the 1915–1958 predecessor of NASA.  Obviously, none of these documents were born digital, and a single NACA report presented on the web was composed of TIFF images, large and thumbnail JPEGs, and a PDF of the entire report.  Based on the experience of managing and presenting these collections of files as a single web object, Michael's dissertation [5] evolved in the direction of creating buckets, the smart web objects in the Smart Objects, Dumb Archives (SODA) model [6]. The basic premises of SODA were that individual reports are more important than the repositories that hold them, and that it should be possible for multiple digital libraries to simultaneously make them discoverable. This 1997 insight is now commonplace, but went against the conventional wisdom of the time. It precedes, yet aligns with, the perspective of the W3C Architecture of the World Wide Web that individual resources are more important than the web servers that host them [7]. As a matter of fact, the Architecture of the World Wide Web only mentions resources, not web servers.

As Herbert and Michael embarked on their respective PhD explorations on different sides of the Atlantic, they didn’t realize they were about to meet, to collaborate on the UPS project, and to present their results at a meeting that would be moderated brilliantly by Cliff Lynch, a man they both admired but had never met in person. 

The UPS Prototype

By early 1999, Herbert’s ideas to give libraries a say regarding links across their electronic collections had taken shape [8]. He had also conducted an experiment illustrating the components of the open linking framework he envisioned. A linking server operated by the library would feature a knowledge base detailing its collection as well as a rule engine that would dynamically decide which links to provide for which type of collection item. A user interested in links for a specific item would click the associated Special Effects link (SFX) that was targeted at the linking server and contained keys that allowed the server to collect sufficient metadata about the item to evaluate the rules and return item-specific links [9]. But inserting SFX links required control of the systems that provided access to the collection and, as such, the experiment only used sources operated locally by the Ghent University Library. Demonstrating the general feasibility of the approach required an experiment without such constraints. 

When Rick Luce, director of the Los Alamos National Laboratory Research Library, visited the Ghent Library to check out the linking approach, it became clear that his groundbreaking Library Without Walls project [10] would provide the ideal setting: its collection combined locally and remotely controlled sources, including locally operated full-text, and it maintained close relationships with various parties in the scholarly information industry. So, Herbert packed up in February 1999 for a six-month stint in Los Alamos and successfully conducted an elaborate experiment that demonstrated the feasibility of the approach with sources under both local and remote control, including full-text collections from Wiley and the American Physical Society, and involved linking servers at Los Alamos and Ghent [11]. 

Figure 3 - Summer 1999, Donna Bergmark’s home: Rick Luce, Herbert, and Paul Ginsparg celebrating the call to action

But Los Alamos was also where the famous physics preprint server - then known as xxx.lanl.gov (now known as arXiv) - ran under Paul Ginsparg’s desk [12]. Having witnessed many years of fierce discussions at Ghent University about subscriptions to journals and their ever-increasing price tag, Herbert very much understood the appeal of the new communication paradigm it entailed and had brought his video camera along to Los Alamos, hoping he might get a chance to interview the much-revered Ginsparg. He shouldn’t have bothered. It turned out that Rick and Paul were already exploring whether the Library Without Walls, which ran a mirror of the preprint server, could become its institutional host, and Herbert started taking part in those conversations. 

One brainstorm led to another and by the time Herbert got ready to return to Ghent, the trio published a call to action (Figure 3) for “the further promotion of author self-archived solutions” in which they announced a meeting with 25 invited experts to be held in Santa Fe, NM, in October 1999, to kick things off [13]. The stated goals were “to reach an agreement regarding an approach to build a promotional prototype multidisciplinary digital library service for the main existing e-print archives” and “to create a forum that will continue to address the interoperability of self-archiving solutions, as a means to promote their global adoption” [14]. To this day, Herbert vividly remembers the thrilling moment when he pushed the Send button on his Toshiba laptop to distribute the final version of the call to action to various listservs, while sitting in his tiny apartment in Santa Fe that was sparsely outfitted with rented furniture (Figure 4). 

Figure 4: July 27 1999, Calle Mejia: Herbert sends out the invitation for the Santa Fe Meeting

Over time, Herbert had come to understand and embrace the “seeing is believing” power of prototypes. He had decided that a concrete strawman to illustrate services across e-print repositories would be needed to fuel discussions; but he would need collaborators to pull that off. When reaching out to various e-print repositories to obtain metadata dumps, Thomas Krichel, a major force behind the Research Papers in Economics (RePEc) [15] effort, enthusiastically came on board. Rick Luce identified just the other person who was needed. Via the New Mexico Library Alliance, he knew Michael’s supervisor Mike Little; together, they engineered a meeting in Washington, DC, anticipating that their Young Turks would resonate. During a four-day meeting in April 1999 it became clear they very much did, although their initial meeting didn’t get off to a good start due to Michael getting hopefully lost driving around Dupont Circle, in those pre-GPS days. They drew up technical plans for a prototype and even managed to get a meeting with Deanna Marcum at the Council on Library Information and Resources (CLIR) and Donald (Don) Waters at the associated Digital Library Federation (DLF), securing support and funding for the meeting and the prototype. 

Figure 5 – October 1999: UPS Prototype - A bucket for a preprint shows ReDIF metadata and two SFX links

Together, Herbert, Michael, and Thomas, started working on the UPS Prototype to be presented at the very outset of the planned Santa Fe Meeting. And, although the prototype was intended “not to make statements about the architectural directions that UPS should take, but rather to facilitate discussions,” [11] its design did entail some significant technical choices. Metadata would be collected from various-print repositories using static dumps; metadata would be normalized to the ReDIF format [16] used in the RePEc initiative; the SODA model would be used to manage/present individual e-prints as buckets; search across the aggregated metadata would be realized using the NCSTRL+ extension of Dienst that supported buckets; each e-print-specific bucket would provide SFX linking capabilities (Figure 5). In order to realize this all in a six-month period, the prototype trio brought more help on board. And they met twice in person: once in Ghent, where Thomas showed up totally drenched on Herbert’s doorstep after having biked through heavy rain from Ostend; once in Los Alamos, prior to the Santa Fe Meeting, when Thomas arrived hours late having biked (Figure 6) through heavy snow from Albuquerque and spent the night in a drainage pipe (colleagues arriving late was a recurring theme for Herbert).  Michael mostly remembers it being bitterly cold, since his earlier visit to Albuquerque in the Summer of that same year had not taught him to pack a sweater for Santa Fe in Fall. And, despite hiccups that plague every project, the well-documented UPS Prototype [17] was finished on time, ready to be presented to the meeting participants. 

The Santa Fe Meeting for ...?

In order to further optimize the chances of success for the meeting, the collaboration of Cliff Lynch and Don Waters as moderators had been secured and turned out to be fundamentally important. In the Acknowledgments section of his PhD thesis, Herbert put Cliff’s impact on the direction of the meeting and on his own thinking as follows:

When starting to work on this thesis, I went back reading several of his early papers and could not feel other than intimidated by the far forward-looking vision expressed therein. At several occasions, I heard Cliff address large audiences, discussing complicated digital library matters with an amazing clarity. Cliff's work has always been a great inspiration to me. I met Cliff for the first time in person at the Open Archives meeting in Santa Fe, for which he had enthusiastically accepted my invitation to serve as a moderator. His involvement was crucial to the successful conclusion of the meeting. [18]

Figure 6 - October 21 1999, Fort Marcy: Thomas Kirchel’s bike made it to the Santa Fe Meeting

The meeting started off in a very concrete manner, with the presentation of the UPS Prototype, some exposes on repository interoperability, and reflections about institutional versus discipline-oriented archive initiatives. But, as the first day progressed, the discussions got increasingly distracted by back-and-forth arguments about the necessity (or not) of peer-review. The Stevan Harnad “self-archiving” camp (archiving the peer-reviewed version of a contribution on a personal or institutional server) insisted it is essential to keep scholarly communication trustworthy, whereas the Paul Ginsparg “preprint” camp (publishing unreviewed contributions on a discipline-oriented or institutional server) stated that knowledgeable readers can assess quality without external review and that novice readers should wait until a peer-reviewed version becomes available. Michael also remembers Paul saying something to the effect that the meeting would be a lot more productive if everyone just learned how to program in Perl and then do something instead of just talking about it. The peer-review tension had already been present prior to the meeting and is even reflected in the evolution of the title of its announcement: an unpublished version dated April 1999 was entitled “Call for participation aimed at the further promotion of the preprint concept”, the version published in July 1999 was entitled “Call for your participation in the UPS initiative aimed at the further promotion of author self-archived solutions”, whereas post-meeting the title was modified to become “The Open Archives initiative aimed at the further promotion of author self-archived solutions.” The choice of the term “archives” didn’t go down well with professional archivists [19], but it did neutralize the disagreement regarding peer-review. By the end of the first day, when participants mingled at the Santa Fe Institute (Figure 7), Herbert was frustrated despite a successful demonstration of the prototype. His bad mood must have been tangible because Ed Fox, whom Herbert had met for the first time at the meeting, volunteered one of his patented neck massages. 

Figure 7 – October 21 1999, Santa Fe Institute: Herbert and Michael at the end of the first meeting day

That night, sleep would not come and Herbert, jetlagged and sleep-deprived, had incessant incoherent thoughts on how to get the meeting back on track. Prior to the start of the second day, he vented his frustration about the lack of progress to Cliff, who was about to start moderating the first session. Cliff was nice enough to let him ramble on a bit, and, in a manner that exemplified one of Cliff’s many unparalleled capabilities, he went on to open the meeting by providing two discussion topics regarding interoperability that he somehow had been able to synthesize from the first day’s discussions, which most had experienced as enjoyable yet lacking in any sense of concrete direction. One was whether archive functions, such as data collection and maintenance, should be decoupled from user functions, such as search. The other was about the choice between distributed searching across repositories and harvesting from them to build cross-repository search engines. This is what the meeting report has to say about the outcome of discussion regarding the first topic:

Although archive initiatives can implement their own end-user services, it is essential that the archives remain "open" in order to allow others to equally create such services. This concept was formalized in the distinction between providers of data (the archive initiatives) and implementers of data services (the initiatives that want to create end-user services for archive initiatives). [20]

The outcome of discussions of the second topic in favor of a harvesting solution is somewhat remarkable because distributed search using WAIS/Z39.50 was quite in vogue in libraries and digital libraries in those days. Cliff himself had a significant track record in Z39.50 and its standardization [21, 22],, but he had also identified harvesting approaches as a topic for further research [23]. Motivated by complexity and scalability concerns, he gently nudged discussions in favor of harvesting. In a paper in which he clarifies the complementary nature of Z39.50 and OAI-PMH, Cliff credits the meeting participants for the decision that was considered controversial by some in the community:

The Santa Fe group wanted a very simple, low-barrier-to-entry interface, and to shift implementation complexity and operational processing load away from the repositories and to the developers of federated search services, repository redistribution services, and the like. They also wanted to minimize the interdependency between the quality of applications services as viewed by the user and the behavior of repositories that supplied data to the applications services. [24]

By the end of the meeting, there was a general sense that the UPS Prototype had been helpful to illustrate the potential of cross-repository services and, hence, to emphasize the need for cross-repository interoperability. A paper that provides a rich summary of the Santa Fe Meeting describes it as follows:

There was general agreement among the participants at the meeting that the Prototype was an extremely useful demonstration of potential.  There was also agreement, however, that trying to reach consensus on the full functionality of the prototype was "aiming too high" and that a more modest first step was in order. [25]

Towards OAI-PMH and OpenURL

By turning the focus of the meeting on these two topics, Cliff fundamentally changed its course. By thoughtfully guiding the discussions towards these concrete outcomes, he set the stage for work on what would become the Open Archives Initiative Protocol for Metadata Harvesting [26], of which both Herbert and Michael became editors. Undoubtedly, he had plenty of technical skills that would have allowed him to make significant contributions to the actual specification effort. But in a manner that characterizes Cliff, he silently took a step back and let the community decide its direction while expressing continued support for the work, on many occasions, and at venues around the world. There can be no doubt that his endorsement played a crucial role in the global adoption of OAI-PMH, which has been an integral part of the scholarly and cultural heritage infrastructure for over two decades.

The focus on interoperability to realize just a single aspect demonstrated by the prototype - cross-repository discovery - also meant that discussions about its other technical ingredients, including SFX linking, would have to be postponed. But both Cliff and Don were very much aware of the problem it addressed and about the nature of the proposed solution. They were both part of the NISO Reference Linking Working Group [27] that investigated how to tackle the so-called “appropriate copy problem”, which, simplifying the charge to the Group, can be summarized as follows: “how to resolve a reference link to a paper in such a manner that it ends up at one of potentially many distributed copies of that paper to which a user, covered by an institutional subscription, has access?” 

The Working Group resulted from a meeting in February 1999 [28], in which various models for a link localization solution had been explored [29, 30]., Don Waters invited Herbert to present his linking work at a second meeting in June 1999 [31]. And, although the meeting report has nice things to say about SFX linking [32], including its ability to address link localization challenges beyond the appropriate copy problem, he remembers profusely apologizing to Don about a presentation not done well. Still, after the demonstration at the Santa Fe Meeting, Cliff extended an invitation for a presentation at the Spring 2000 meeting of the Coalition for Networked Information [33]. The room was packed with representatives from libraries, the scholarly publishing industry, and library system vendors, and the talk became a veritable breakthrough moment for SFX linking. But significant tasks remained, including standardizing the SFX link syntax and demonstrating the ability of the approach to integrate with the emerging DOI-based reference linking approach pursued by journal publishers and instantiated by CrossRef [34]. 

The standardization’s history is well documented [35]; it started in December 2000 when the original SFX URL specification [36] - by then renamed OpenURL - was submitted to NISO and concluded five years later with the release of The OpenURL Framework for Context-Sensitive Services [37]. The DOI integration was explored by means of a limited prototype [38] that was demonstrated and discussed at the July 2000 NISO/DLF/CrossRef meeting [39]. As the meeting seemed to reach a consensus in favor of the proposed model with an institutional localization component powered by OpenURL - essentially the SFX open linking approach - a question was brought forward as to whether the model with a centralized localization component that had been identified in the first meeting of the Working Group should also be further discussed. At that point, Cliff decidedly stepped in stating “No. We have a solution!” In doing so, he paved the way for the endorsement of the OpenURL linking framework by the Working Group, the rigorous testing of its feasibility in an extended prototype [40], and its eventual acceptance in the US scholarly communication community and beyond. Afterwards, Cliff continued to express support for the approach at numerous venues and gave it his strongest possible endorsement by becoming a member of Herbert’s PhD jury.

Thank you, Cliff

By means of the UPS Prototype effort, this essay has illustrated Cliff’s fundamental impact on the direction infrastructure for research, education, and cultural heritage has taken in the past decades. Two technologies, OpenURL that was used in the Prototype and OAI-PMH that resulted from the Prototype, became an integral part of that infrastructure. Hopefully, the essay has adequately shown that Cliff had a significant part in making that happen, not as an author of specifications, a writer of code, or a builder of tools. But rather as an identifier of problems to come and as a perceptive influencer, gently nudging forward the solutions he believed in and strongly supporting the community efforts that realized them. We have witnessed the same impact in other efforts we have been involved in since the UPS Prototype and can safely assume that others have experienced it in their projects aimed at improving the status-quo of scholarly information infrastructure. 

We do want to emphasize that, as we dreamt up the outlines of the UPS Prototype, we were early career researchers with a visible, yet modest track record. Cliff (CNI), along with Paul Ginsparg (LANL), Rick Luce (LANL), Deanna Marcum (CLIR), and Don Waters (DLF) strongly and publicly endorsed our effort, shone the spotlight on us, and in doing so had a major impact on our career trajectories. We vividly remember receiving that support and the experience has led us to similarly support the young researchers we have mentored since. 

Figure 8 – December 12 2017, Washington, DC: Cliff and Herbert at the Fall 2017 CNI Membership Meeting

As we were selected to write a contribution for this Festschrift, on behalf of all infrastructure plumbers, we want to profoundly thank Cliff. Scholarly infrastructure would not have progressed the way it did without him. We don’t envy the person who will step into his shoes once he has retired. The work ahead is enormous, with needs for new infrastructure and existing infrastructure crumbling. Indeed, OAI-PMH is being supplanted due to its reliance on XML, a technology that has become arcane in a JSON world. And the OpenURL Framework is under attack by the centralized Get Full Text Research [41] effort, launched by the major commercial publishers, that mutes the capabilities of libraries to influence the nature of links across their electronic collections. While 25 years of OAI-PMH and OpenURL do not put those technologies in the same IT infrastructure league of - say - UNIX, it is a substantial period considering that the lifetime of many digital library phenomena can typically be measured in terms of months or years, not decades. Cliff’s influence is directly traceable in the global penetration and longevity of these two technologies that go all the way back to the 1999 UPS Prototype. 

References

[1] Robert Ashley, “Music with roots in the aether: interviews with and essays about seven American composers” 1st ed. Köln : MusikTexte, 2000. 244 p. https://archive.org/details/ashley-robert-music-with-roots-in-the-aether-2000

[2]  Robert Ashley, David Behrman, Philip Glass, Alvin Lucier, Gordon Mumma, Pauline Oliveros, Terry Riley. “Music with roots in the aether”. June 1, 1977-June 18, 1977. The Kitchen, New York, New York. https://thekitchen.org/on-file/music-with-roots-in-the-aether/

[3] Anna Wiener, “On Reading Issues of Wired from 1993 to 1995,” The New Yorker, June 7, 2016, https://www.newyorker.com/culture/cultural-comment/on-reading-issues-of-wired-from-1993-to-1995

[4]  Herbert Van de Sompel, Patrick Hochstenbach, and Tobias De Pessemier, “The hybrid information environment and our Intranet solution to access it“, Ghent University Academic Bibliography, 1997, accessed on January 27, 2025, https://hdl.handle.net/1854/LU-1056689

[5]  Michael L. Nelson, "Buckets: smart objects for digital libraries,", PhD dissertation, Old Dominion University Digital Commons, 2000, accessed on January 28, 2025, https://doi.org/10.25777/gbh6-7d07

[6]  Michael L. Nelson et al. “SODA: Smart Objects, Dumb Archives,” Lecture Notes in Computer Science 1696, (1999): 453-464, accessed on January 27, 2025, https://doi.org/10.1007/3-540-48155-9_28

[7]  Ian Jacobs and Norman Walsh, “Architecture of the World Wide Web, Volume One,” W3C Recommendation, 15 December, 2004, accessed on January 27, 2025, https://www.w3.org/TR/webarch/

[8]  Herbert Van de Sompel and Patrick Hochstenbach. “Reference Linking in a Hybrid Library Environment. Part 1: Frameworks for Linking,” D-Lib Magazine 5, no 4 (1999), accessed on January 27, 2025, https://doi.org/10.1045/april99-van_de_sompel-pt1

[9]  Herbert Van de Sompel and Patrick Hochstenbach. “Reference Linking in a Hybrid Library Environment. Part 2: SFX, a Generic Linking Solution,” D-Lib Magazine 5, no 4 (1999), accessed on January 27, 2025, https://doi.org/10.1045/april99-van_de_sompel-pt2

[10] “Library Without Walls Welcome Page,” 1999, archived at the Wayback Machine, April 28, 1999, 

http://web.archive.org/web/19990428111624/http://lib-www.lanl.gov/lww/welcome.html 

[11]  Herbert Van de Sompel and Patrick Hochstenbach. “Reference Linking in a Hybrid Library Environment. Part 3: Generalizing the SFX Solution in the SFX@Ghent & SFX@LANL experiment,” D-Lib Magazine 5 no 10, accessed on January 27, 2025, https://doi.org/10.1045/october99-van_de_sompel

 

[12]  Sheon Han. “Inside arXiv—the Most Transformative Platform in All of Science,” WIRED, March 27, 2025, accessed on April 2, 2025, https://www.wired.com/story/inside-arxiv-most-transformative-code-science/  

[13]  Paul Ginsparg, Rick Luce, and Herbert Van de Sompel. “List of participants at the Santa Fe meeting of the Open Archives initiative, October 21-22 1999,” 1999, archived at the Wayback Machine, August 18, 2001, http://web.archive.org/web/20010818161630/http://vole.lanl.gov/ups-participants.htm

[14]  Paul Ginsparg, Rick Luce, and Herbert Van de Sompel. “The Open Archives initiative aimed at the further promotion of author self-archived solutions,” July 27th, 1999, archived at the Wayback Machine, January 19, 2001, http://web.archive.org/web/20010119091100/http://vole.lanl.gov/ups-invitation-ori.htm

[15]  RePEc: “Research Papers in Economics,” accessed on January 26, 2025, http://repec.org/

[16]  Thomas Krichel. “ReDIF version 1.0,” accessed on January 27, 2025, https://openlib.org/acmes/root/docu/redif_1.html 

[17]  Herbert Van de Sompel et al. “The UPS Prototype,” D-Lib Magazine 6, no 2 (2000), accessed on January 27, 2025, https://doi.org/10.1045/february2000-vandesompel-ups

[18]  Herbert Van de Sompel.  “Dynamic and context-sensitive linking of scholarly information,“  Ghent University Academic Bibliography, 2000, accessed on January 27, 2025, https://hdl.handle.net/1854/LU-522209

[19]  Michael L. Nelson, "To the Editor: Response to Peter Hirtle's April 2001 editorial, OAI and OAIS: What's in a Name?," D-Lib Magazine 7 no 5, accessed on February 12, 2025, https://doi.org/10.1045/may2001-letters

[20]  Paul Ginsparg, Rick Luce, and Herbert Van de Sompel. “First meeting of the Open Archives initiative,” 1999, archived at the Wayback Machine, January 19, 2001, http://web.archive.org/web/20010119090800/http://vole.lanl.gov/ups1-press.htm

[21]  Clifford A. Lynch. “RFC1729: Using the Z39.50 Information Retrieval Protocol in the Internet Environment,” December, 1994, accessed on January 27, 2025, https://datatracker.ietf.org/doc/rfc1729/

[22]  Clifford A. Lynch. “The Z39.50 Information Retrieval Standard - Part I: A Strategic View of Its Past, Present and Future,” D-Lib Magazine 3, no 4 (1997), accessed on January 27, 2025, https://hdl.handle.net/cnri.dlib/april97-lynch

[23]  Thomas Baker and Clifford Lynch. “EU-NSF Working Group on Metadata,” D-Lib Magazine 4, no 3 (1998), accessed on January 27, 2025, https://www.dlib.org/dlib/march98/03clips.html#MULTILINGUAL

[24] Clifford A. Lynch, "Metadata Harvesting and the Open Archives Initiative," ARL: A

Bimonthly Report, no. 217 (August 2001): 1–9,

archived at the Wayback Machine, May 25, 2012,  https://web.archive.org/web/20120525000130/http://www.arl.org/resources/pubs/br/br217/br217mhp.shtm

[25]  Herbert Van de Sompel and Carl Lagoze. “The Santa Fe Convention of the Open Archives Initiative,” D-Lib Magazine 6, no 2 (2000), accessed on January 27, 2025, https://doi.org/10.1045/february2000-vandesompel-oai

[26]  Carl Lagoze, Herbert Van de Sompel, Michael L. Nelson, and Simeon Warner. "The Open Archives Initiative Protocol for Metadata Harvesting," June 14, 2002, accessed on January 28, 2025, https://www.openarchives.org/OAI/2.0/openarchivesprotocol.htm

[27]  National Information Standards Organization. “Reference Linking Working Group, Sponsored by NISO, DLF, NFAIS, SSP, and CNI,” February 1999, archived at the Wayback Machine, November 17, 1999, http://web.archive.org/web/19991117151813/http://www.niso.org/reflink.html

[28]  National Information Standards Organization. “First Workshop on linkage from citations to electronic journal literature,” February 1999, archived at the Wayback Machine, April 21, 2000, http://web.archive.org/web/20000421205230/http://www.niso.org/linkge.html

[29]  Priscilla Caplan. “A model for reference linking,” June 10, 1999, archived at the Wayback Machine, October 3, 1999, http://web.archive.org/web/19991003125020/http://www.lib.uchicago.edu/Annex/pcaplan/reflink.html

[30]  Priscilla Caplan and William Y. Arms. “Reference Linking for Journal Articles,” D-Lib Magazine 5, no 7/8 (1999), accessed on January 27, 2025, https://doi.org/10.1045/july99-caplan

[31] National Information Standards Organization. “Second Workshop on linkage from citations to electronic journal literature,” June 1999, archived at the Wayback Machine, July 7, 2000, 

http://web.archive.org/web/20000709204615/https://www.niso.org/linkge2.html

[32]  National Information Standards Organization. “Report of the Second Workshop on Linkage from Citations to Journal Literature,” 1999, archived at the Wayback Machine, June 5, 2000, http://web.archive.org/web/20000605232443/http://www.niso.org/linkrept.html

[33]  Herbert Van de Sompel. “The SFX framework for context-sensitive reference linking,” CNI 2000 Spring Task Force Meeting, March 27-28, 2000, accessed on January 27, 2025, https://www.cni.org/wp-content/uploads/2013/06/Sompel-SFX2000Stf.pdf

[34]  Helen Atkins et al. “Reference Linking with DOIs: A Case Study,” D-Lib Magazine 6, no 2 (2000), accessed on January 27, 2025, https://doi.org/10.1045/february2000-risher

[35]  Arthur Hendricks. “The Development of the NISO Committee AX's OpenURL Standard,” Information Technology & Libraries 22, no 3 (2003): 129-133, accessed on January 27, 2025, http://archives.pdx.edu/ds/psu/11519 and https://ital.corejournals.org/index.php/ital/issue/view/955/7

[36]  Herbert Van de Sompel, Patrick Hochstenbach, and Oren Beit-Arie. “OpenURL Syntax Description,” May 16, 2000, archived at the Wayback Machine, April 10, 2001   https://web.archive.org/web/20010410235911/http://www.sfxit.com/openurl/openurl.html

[37]  National Information Standards Organization. “NISO's OpenURL Now a National Standard,” May 2, 2005, accessed on January 27, 2025, https://www.niso.org/press-releases/2005/05/nisos-openurl-now-national-standard

[38]  Herbert Van de Sompel and Oren Beit-Arie. “Open Linking in the Scholarly Information Environment Using the OpenURL Framework,” D-Lib Magazine 7, no 3 (2001), accessed on January 27, 2025, https://doi.org/10.1045/march2001-vandesompel

[39] "Meeting Report of the NISO/DLF/CrossRef Workshop on Localization in Reference Linking,” July, 2000, archived at the Wayback Machine, December 6, 2000,  

http://web.archive.org/web/20001206170000/http://www.niso.org/CNRI-mtg.html

[40]  Oren Beit-Arie et al. “Linking to the Appropriate Copy,” D-Lib Magazine 7, no 9 (2001), accessed on January 27, 2025, https://doi.org/10.1045/september2001-caplan

[41]  “Get Full Text Research,” accessed on January 27, 2025, https://www.getfulltextresearch.com/