Kindle clippings

Kumiyo Nakakoji (University of Tokyo) calls this representational talkback. The writer seeks a way to represent the problem space, sketches notes that describe an initial approach, and then reviews the notes as if approaching the problem anew. If the representation is good, the problem now seems simpler. If the representation fails to make sense, that incoherence often suggests a new and better approach to the problem. Nakakoji’s studies of tools for enhancing this internal dialog, Amplified Representational Talkback, influenced the design of Tinderbox. (loc: 200-204)

Conceptual Clustering In choosing a camera, we progressed from readily-accessible but excessively concrete information – today’s camera models – to improve our understanding of camera selection and, indeed, of the digital camera industry. We used notes to focus our attention and to discover patterns and exceptions. (loc: 204-6)

Bags, Bunches, and Sequences Tinderbox containers can play three distinct roles in organizing your information. (loc: 376)

Adornments express loose and experimental relationships, while containers express more concrete and lasting ideas of subordination or containment. (loc: 400-401)

One of the earliest and still most sophisticated approaches to the link was Randy Trigg’s doctoral dissertation, which explored a hypertext system for reviewing scientific papers for publication [24]. (loc: 1114-16)

At the same time, the complexity of writing (which the hypertext research community had initially underestimated) was gradually becoming clear. In particular, both critical theory [12] and careful examination of actual hypertexts [3] showed that it is often impossible to assign a single function or meaning to each unit of a text. The same link carries a different meaning in different contexts; worse, that meaning may well depend on what the reader has already seen, and what the reader already knows. A link might, in the first stages of a long project, represent an eager insight. Later, its more knowledgeable creator might look at the same link as a cautionary flag, an example of early naïveté. Still later, the same link might be seen as ironic commentary, or satire, or as a proposal for a daring and unexpected return to a long-abandoned position. (loc: 1141-47)

link types. not worth it. not capturing emrgent understanding. people dont use it. slows down writimg. link to tag discussion w hewitt (loc: 1141)

links can fulfill many different functions (loc: 1147)

link between goto in programming and complex link structures (loc: 1159)

Reacting to van Dam’s warning, a number of workers adopted a vision of links that was essentially prescriptive: links should express the connection between modules of information, with each link performing a declared and defined role. Large documents, they suggested, could be mastered by hierarchical decomposition, using links to express structure. Links might also play a carefully limited role in drawing connections across the structure, providing glossaries or media services. The result was link minimalism, a hypertext movement which, ironically, sought to use links as little as possible. This impulse was expressed in hypertext engineering [7], Hypertext Design Methodology [6], and similar efforts to simplify and tame the potential complexity of the link network. My own early hypertext work, in retrospect, fits into this line of development. The chief concern of “The Bookmark and The Compass” [2] was the fear that readers would find hypertexts disorienting and confusing; it proposed a host of tools – bookmarks, breadcrumbs, thumb tabs, margin notes – to remind readers where they were. The lasting contribution of link minimalism, in the end, was its attempt to draw a distinction between structural links, which were intended to express the articulation of the hypertext into sections and subsections, and “applicative” links which were intended to express connections relating to the subject matter or “application” of the document. (loc: 1160-70)

structural links and applicative links (loc: 1170)

If the reception of Moulthrop’s paper showed how large the chasm was, his hypertext novel Victory Garden [20] made a crucial contribution to bridging it. Joyce’s afternoon, though widely admired, was not widely understood. Moulthrop’s episodic novel about the first Gulf War and the ongoing controversy of the canon then embroiling world campuses was a convincing demonstration of hypertext’s power – and of its complexity. Moulthrop’s links are filled with puns, tricks, feints, and deceptions. No one can believe these links can be clearly labeled, or that they epitomize “free and knowing navigation” [21], and yet these links are also clearly planned with care and intelligence. For a time, link minimalists tried to argue informally that such effects would prove useful only in the arts, but David Kolb, a philosopher, published Socrates in the Labyrinth [10] to preempt this line of argument by showing that the philosophical underpinnings of argumentation were at least as complex as the artistic effects the minimalists sought to dismiss. (loc: 1184-91)

Trigg’s link types declare what a link is. A second line of development asks links to carry an operational, rather than a formal, meaning. Links, in this view, say nothing about what their endpoints are or what their relationship may be; they offer an opportunity to go from one place to another. (loc: 1193-95)

More recently, frustrated attempts to build large sites and indexes based on a grand taxonomy – a single, consensual, topical organization – led to enthusiasm for informal folksonomies and ad hoc tag schemes. But tag clouds and folksonomies necessarily abandon the hope of an ideal or optimal navigation scheme in which everything finds its proper place. Informal organization demands informal (and multiple) navigation. Indeed, while minimalists tried to eliminate as many confusing links as they could and navigationalists strove to make links universally clear and intuitive, some hypertext writers were deliberately seeking to disorient readers, and found the work surprisingly difficult. (loc: 1208-13)

Folding Space. In a spatial hypertext, proximity takes on special, semantic meaning. Adjacent notes, or notes that lie on the same adornment or cluster together, are connected. Distant notes are separated. At times, however, we may need to “fold” space, to assert a connection between notes that cannot be clustered together. Here, the link might not represent a one-to-one relationship; we might link a representative note in one cluster to some representative note in another cluster to remind us that these widely separated clusters are, in fact, related. (loc: 1241-45)

Finally, it is vital to consider the meaning of the link, the manner in which readers extract information from the link and its trajectory. When we follow a navigation link from here to there, we naturally inquire why the link was made as it was. What did the link’s label suggest we would find at its destination? How does this explanation refract our encounter with the destination text? The link situates the meaning of its source and destination while carrying its own interstitial meaning in the chasm between one note and another. The connection between links and the cinematic cut (and Eisenstein’s theory of montage) is intimate and requires careful thought [17, 18]. (loc: 1270-75)

external and internal links… markdown and html (loc: 1285)

Internal links encourage malleable texts – editing is no more difficult than conventional text editing – but internal links imply that the owner of the text controls all outbound links. If you want to let readers annotate, you either have to give them indiscriminate permission to edit the page [14] or you need to support external links. If you want to support external links, you must either avoid editing the page or you need an elaborate notification service to update all the link tables whenever the page changes. (loc: 1286-90)

External link advocates are often document-centric, treating the nodes as fixed and unchanging atoms. (Indeed, it is an explicit requirement of Nelson’s Xanadu that texts never change; new versions may be created but a text, once published, remains forever unaltered.) (loc: 1292-94)

We may also distinguish a third mode, which Miner calls social and which (because “social software” means something else) I will here call intimate or nobitic. Here, the notional audience lies beyond the author but does not extend to the world at large. Letters intended for publication are nobitic. So, too, are stories notionally told among friends or written for our family, Heart of Darkness or Winnie the Pooh. We have always recognized that journals and diaries adopt any of these modes. Thoreau’s Journal is notionally a diary, but clearly it was composed and intended to influence the general public and the electorate. Da Vinci and Pepys, on the other hand, wrote for themselves: Da Vinci wrote in code, and Pepys in an obscure shorthand to which, in the more sensitive amatory passages, he added a further layer of pidgin French and Italian. Other diaries seem fundamentally nobitic: Anne Frank invents the diary as a sort of character she can address because she wants someone outside the family to whom she can complain. Understanding the distinction between private, nobitic, and public weblogs is crucial to appreciating the craft of these shared notes, and for thinking clearly about what notes should be shared and how those notes can best be expressed. (loc: 1822-31)

Henry Darger never seems to have mentioned his immense, unpublishable fantasies of the Vivian Girls to anyone. (loc: 1858-59)

Three fundamentally different and incompatible metaphors shape both the design of hypertext systems like Tinderbox and the way these designs are usually discussed and marketed. (loc: 2083-84)

info farming manufacturing gardening (loc: 2084)

info farming mining gardening (loc: 2084)

Information Mining views pertinent information as a valuable resource to be efficiently extracted and refined. Successful mining tools are those which can most quickly extract valuable nuggets of precious fact from large repositories of base data. The classic measures of successful mining are recall, precision, and cost. Information Manufacturing views the acquisition, refinement, assembly and maintenance of information as a continuous enterprise, employing teams of specialists to implement acquisition procedures and management strategies. In contrast to mining, with its individualistic emphasis on seizing information and opportunity, the rhetoric of the information factory values continuity of process over individual entrepreneurship. Successful information factory tools are those which create stocks of corporate information through inexpensive, interchangeable labor. The classic measure of successful factory systems is usability: ease of initial training, productivity, and accuracy of work product. 4 Portions of this chapter originally appeared, in somewhat different form, in “Enactment in Information Farming”, Proceedings of Hypertext ’93, ACM, Baltimore. © 1993 by the Association for Computing Machinery. Information farming (or information gardening) views the cultivation of information as a continuing and frequently collaborative activity performed by people working to achieve changing individual and common goals. Where the mine and factory serve the organization, the information farm is a computational space where colleagues and employees may work together on shared tasks and also pursue individual goals. The focus is neither on extraction (as in the mine) nor on stockpiling (as in the information factory), but on continuous cultivation and community. The vision of information farming is integral to the romantic school of hypertext, and the different goals of the information farm and information mine have been a fertile source of misunderstanding between the hypertext and information-retrieval communities. The design goals of farm tools differ from the goals we usually apply to mining tools and factory systems. Because farming is a continuous activity, measures of discrete transactions (e.g. recall rates) are less central to farming than to mining. Moreover, traditional human- computer interface studies concentrate on easily measured, everyday phenomena – the repetitive activities of the factory rather than the disparate activities of the information farm. The success of farming systems lies at least as much in the extent to which they can convey insight or explain extraordinarily difficult concepts as in their facility for expressing routine matters. Anecdotes and reading logs – records of personal experience over an extended course of thoughtful use and introspection – are a more characteristic approach to studying actual information farms and farmers than statistical surveys of sample subjects. (loc: 2084-2104)

when to use different documents in tb? (loc: 2245)

the balnce betw text in notes and titles (loc: 2267)

1. Allen, D., Getting Things Done: The Art Of Stress-Free Productivity. 2002, New York: Penguin. 2. Bernstein, M., The Bookmark and the Compass: Orientation Tools for Hypertext Users. SIGOIS Journal, 1988. 9(1988): p. 34-45. 3. Bernstein, M., M. Joyce, and D.B. Levine. Contours of Constructive Hypertext. in European Conference on Hypermedia Technology. 1992. Milano: Asssociation for Computing Machinery. (loc: 2285-88)

13. Landow, G.P., editor, The Victorian Web. 1987-present. (loc: 2295-96)

22. Shipman, F. and C.C. Marshall. Formality Considered Harmful: Experience, emerging themes, and directions on the use of formal representations in interactive systems. in CSCW. 1999. 23. Taylor, P.G., “Hypertext-based art education: Implications for liberatory learning in high school.” Doctoral Dissertation. 1999: Penn State University. (loc: 2301-3)