Bootstrap Institute
6505 Kaiser Drive
Fremont, CA 94555


Memorandum

Date: 5 Apr 2000 19:36 PDT

From:   Doug Engelbart
DOUG@bootstrap.org
Bootstrap Institute

To:     Eric Armstrong
eric.armstrong@eng.sun.com

Subject:   OHS development candidate vector
For Discussion

INTRODUCTION

I keep returning to this "vector" again and again, and decided that it is important to me as possibly the best course we could take. So let me describe it; and then we can talk about it.

It would defer until later some basic, and ultimately critical, parts of the OHS Tool System Architecture -- e.g., integrated editing and multi-class UIS. But it would have high immediate utility, and provide unique usage experience to many "outside users," and especially to categories of web users we'll want to have become seriously involved in subsequent co-evolution.

BASIC DESCRIPTION

Heavy involvement of IBM's WBI (Web-Based Intermediary)

  1. Assume basic operational mode ...

  1. URLs targeted toward designated files would be intercepted by WBI.

    Likely have extended syntax to these "URLs" -- see below.

  2. WBI retrieves the designated file from the designated server

  3. WBI "transcodes" the designated file into the transcoded-file form which the user's browswer can access and display.

    Simple step -- if designated file is HTML, then the transcoding employs reproducible algorithm to affix "purple numbers" as target-tags-on, and links-to each paragraph ("label/link tags").

    Every time this file goes through this retrieval and transcodin process, it will have the same label/link tags affixed to the sme paragraphs, so that e.g. "#5B" on a user-transmitted URL would always position her browser window to the same "5B-labelled" paragraph.

    Added nicety -- INDIRECT LINKING: Special syntax on the initial UL indicates, "go to position X in File Y, pick up the link there, thn do the above transcoding on that URL's designated file."

    Serious "OTHER-FILE LINKING" -- where the designated file is not oe which the browser can position or display. E.g., "native file fors" of MS Word, Lotus Notes, ..., or especially software source-code, requirements, specifications, status reports, user guides, etc.

    The transcoding produces specialy structured and formatted HTML (likely later extended into XML), with label/link additions.

    Then, when files are transcoded into browser-readable form, the transcoder would provide optional, additional "deliverable" alterations to the content (and formatting/structuring) to create any special "views."

    Considerable "co-evolution" here, exploring optional ways to do the structuring, formatting and viewing.

  4. WBI sends the transcoded file to the user's browser to satisfy the location/view specs held in the (extended-syntax) URL.

2. Staged goals:

  1. First, to get addressable label/link tags installed in our own DKR-web email and docs.

  2. Then, general web documents -- so we can begin usefully to cite specific passages in anybody's web pages.

  3. Then begin providing useful view options in our DKR email and docs

    -- and on "out" to anybody's web pages..

  4. Then pick a target software-development environment: source-code language? Likely trial-user communities?

  5. In collaboration with trial-user community, establish and implement initial set of transcoding specs for project source code, requirements, specs, project status reports, relevant email dialog, user docs, ... toward a trial prototype software-development environment.

    NOTE:

    Of special potential, for a targeted user communities, would be a/the Open-Source OHS-Developer Community.

  6. During all of this, establish a healthy set of user representatives as active stakeholders in the co-evolution of human- and tool-system development.

    3. Somewhere in the above sequence, develop at least the framework for an OHS architecture which would enable the integrated editing and other whole-system operation.

    As funding and participatory interest emerge, the above sequence would be modified appropriately.

AND ... ***********

This is all open for discussion; plus, there are interesting partnering possibilities emerging with the Software Productiviey Consortium. And, ... what companies/agencies/societies might be interested collaborators?

OpenGroup? ACM?


Sincerely,

Doug

Douglas C. Engelbart, Ph.D.
doug@bootstrap.org


Copy to:

  1. Jon Bosak, Jon.Bosak@eng.sun.com,
  2. Adam Cheyer, acheyer@verticalnet.com
  3. Bill Daul, bill.daul@zoomon.com
  4. John Deneen, JJDeneen@ricochet.net
  5. Philip Gust gust@NouveauSystems.com
  6. Marcelo Hoffmann, hoffmann@sri.com
  7. Lee Iverson, leei@ai.sri.com
  8. Tanya Jones, tanya@foresight.org
  9. Ted Kaehler, tedk@wdi.disney.com
  10. Eugene Kim, eekim@eekim.com
  11. Ruth Lang, rlang@ai.sri.com
  12. Harvey Lehtman, hlehtman@iftf.org
  13. Paul Maglio, pmaglio@almaden.ibm.com
  14. Jack Park, jackpark@verticalnet.com
  15. Chris Peterson, peterson@foresight.org
  16. Andy Poggio, andy.poggio@eng.sun.com
  17. Jeff Rulifson, Jeff.Rulifson@eng.sun.com
  18. Neil Scott, ngscott@arch.stanford.edu
  19. Jim Spohrer, spohrer@almaden.ibm.com
  20. Warren Stringer, warren@muse.com
  21. Rod Welch, rowelch@ATTGLOBAL.NET
  22. Joe Williams, altintdev@webtv.net
  23. Shinya Yamada, shinya@indigo.co.jp
  24. Su-Ling Yee, yeemailbox@yahoo.com>
  25. Paul Fernhout, pdfernhout@kurtz-fernhout.com
  26. Frode Hegland, frode@liquidinformation.com
  27. Hilary LaMonte, HLaMonte@nsba.org
  28. Jeff Miller, jeffm@dynamite.com.au
  29. Henry vanEyken, vaneyken@sympatico.ca
  30. Jon Winters, winters@obscurasite.com>