Evolving News: Turning Answers Into Stories
Turning Answers Into Stories (TAIS) introduction
turning answers into stories is a shorthand i use for an elaborate goal: using computer communications for collaborative development and discovery of details in context - in various "story" contexts where the details belong.
web interlinking and search indexing are some ways we currently organize the inundation of available information. what's lacking in those approaches is progressive relative locality - the ability to organize details in relation to one another across different, progressively expansive contexts. such progressively expansive context provides extendable orientation. (See TAIS Orienting for more along this line.)
to enable this, people need to be able to situate details according to the context where they feel the details belong - like bookmarking - and to navigate according to such arrangements. by capturing individuals arrangements, and consolidating the arrangements so people can benefit from one another's choices, people could collaboratively grow arrangements that suit their perspectives, and also discover and connect with the contexts where the details belong according to other people's perspectives, or according to their own perspectives in other situations.
by enabling people to see and use one another's contextual organizations, they can benefit from commonly useful arrangements as well as from insights gleaned from contrasting arrangements, collaboratively gaining and growing shared perspectives that are salient to many.
news ecologies - news stories - and actually, interconnected constellations of stories - that grow progressively, with easy, salient situation of new details as they arrive.
ongoing, incremental documentation of development activities and developing systems, using details in the various contexts where they belong.
(i've used my [pendref] EmacsAllout outliner for a mundane daily journal of my work activities for years, and have been eager for almost as long for something like what i describe above. this, and work on things like [pendref] ZWiki extensions, issue trackers, content management sytems, and so on, inform my appetites and choices here.)
progressively up-to-date, organized directories of business details of social and business organizations, including things like the offices, activities, policies, agreements, connections.
collaboratively extended topic trees - extensible layered classification system where classification topics and refinements are derived from usage. users have the option to add new choices in any collection, with the understanding that the most frequently used siblings - the most popular - are the most prominently offered as choices.
all these items could be seen as a kind of the "news ecology", and that in turn is a kind of view of the world wide web view with more provisions for extensive, orientable organization i describe in [pendref] TAISOrienting.
(click on the items with triangle bullets to collapse or expand their contents)
- collaborative organizing:
- individuals situating info of concern according to their own criteria
- facilities for indexing commonalities in arrangements provide the basis for sharing and growing collaborative arrangements (a la social bookmarking, collaborative filtering)
- info situated in multiple contexts, eg:
- those of each individual concerned with the info
- the arrangements of project details by project organizers
- activity logs within and across ongoing efforts like projects, daily activities, etc
- technical systems descriptions, from various concerns like operation, design, maintenance, marketing, constitute contexts which describe the systems
conventional outlines and !FAQs - "Frequently Answered Questions" documents - are rudimentary approaches to addressing this kind of issue, with serious scaling limitations. i'm aiming for something more, where:
- the same detail can be composed differently in different contexts.
- recipes for connecting and combining descriptions would themselves be developed and organized as descriptions, susceptible to all the capabilities of the system.
my representative application would be a development journal where you can describe and review project work in various ways.
in this journal, you could focus on the current day's work on one or another project, concentrating on the frontiers of particular tasks. you could see some detail with the context of other projects where it belongs - eg, a person involved, or an artifact, or a coordination date, or whatever. you could also look at the project from various cumulative views, including things like the current status of all the project's thrusts, or historical progressions of those thrusts, or anticipated efforts and actions - plus scheduling info for some of the anticipated elements - and so on.
as a single individual can approach each detail in a project in the light of diverse contexts, different individuals might be looking at and working with a project from different roles - those of a developer, an administrator, an end user, a marketer, and so on. each of these perspectives can be concerned with different (and multiple) aspects of some shared items, according to the context.
i am extending simple outline-like structures to represent shared items, including conveying their different contents in different contexts and also expressing their congruences - in identity and also structure - with other instances of the item in other contexts. (i am gradually extending my [pendref] EmacsAllout outliner as a prototype interface to a remotely accessed data structure which would provide these "spherical" organization features.)
i'll include mechanisms for referring to items for the description of item structure, providing an extensible structuring mechanism for representing context-sensitive organization of items that is, itself, context sensitive. on top of that i'll develop a project journal as sketched above.
ultimately, the usefulness of information is in its salient connection with other information. in order to scale that usefulness, "organizing" needs to be incrementally refinable and federate-able, so one person's contributions are useful not only to themselves, but also to others with related concerns. i am trying to head towards a way to express information's organization with adequately refinable and federate-able details, for collaborative connection of those details into their many stories.
i have to stress that the goal is not to develop a system capable of automatically organizing information. rather, my aim is to harness the overlap of the respective individual concerns of the people using the system.
the aim is a flexible framework that enables people to make choices about situating the information they encounter and contribute according to their own needs, informing but not precluding the choices of others - or their own choices in other contexts. the system would facilitate recognition and consolidation of related choices so that the overall organization as well as the cumulative content benefits from everyone's insights, as motivated by their own incentives.
this approach would complement and extend indexed search, providing scalable contexts to navigate rather than disjoint neighborhoods. (see [pendref] TAISOrienting for elaboration.)
often, details in one description of something belong in other descriptions, of that thing and of others. aspects of the details often vary depending on the context of concern. for example:
- view of a project detail today can have some refinements compared to yesterday's view, or that of last month.
- person as business contact can have different contact information than person as club member - eg, office vs home phone, boss or colleague vs spouse as backup contact, etc. the same details (phone number, backup contact) have different values in different contexts.
- some of the relevant details for use of a ticketing system are different for the end user, operators, various kinds of maintainers/repairman, developers, and marketers.
also, organization of knowledge about the system - the metadata - is an aspect of the descriptions the system implements, just like any other facts the system is used to convey. the description representation needs to be able to represent organization as well as other details.
- rich - to convey substantial descriptions of real things with their many and diverse facets
- flexible - to evolve along with and adapt to continually emerging understanding of the description's subjects
- simple and comprehensible - to enable personal navigation with a continuing sense of orientation, of relative location within coherent contexts (see [pendref] TAISOrienting)
- scalably coherent - so that diverse descriptions can be consolidated to form more comprehensive descriptions, and, diverse individuals can benefit from and collaborate with one another across this arrangement
see [pendref] TAISPrinciples for primary motivating issues and precedents, and [pendref] TAISOrienting for a central organizational premise.
|[sites]||individual web sites, and sections within site, provide one dimension of locality, and opportunities for structuring within it (hierarchical and topical containment, next/previous, etc). those opportunities can be used for the kinds of structuring that i describe, but without a comprehensive, extensible model and some technical infrastructure, they cannot coherently embody various contextual organizations. lacking that, web site hierarchies instead manifest the quandary of hierarchical obscurity as they grow: buried-too-deep in an overly nested arrangement or one-among-too-many in an overly-flat one. resorting to navigation by search indexing sacrifices any opportunity to develop a sense of relative locality.|
|[work]||the goals underlying Turning Answers Into Stories have driven many of my interests and efforts in software development, including my work on [pendref], [pendref] , [pendref] , [pendref] (driving ideas about [pendref] OrganizationObjects), and more|