THE WELCH COMPANY
440 Davis Court #1602
San Francisco, CA 94111-2496
415 781 5700
rodwelch@pacbell.net


S U M M A R Y


DIARY: August 24, 2000 02:23 PM Thursday; Rod Welch

DKR architecture based on SDS according to design meeting 000822.

1...Summary/Objective
2...SDS Model Uniform Display and Access Work Record
3...OHS/DKR Trying to Accomplish Access to Record of Work History
4...Email Translated into XML HTML Access Work Record Displayed in Browser
5...Design Changes Failed OHS/DKR Development
6...Changing Tools Design OHS/DKR Development Failed
7...Wheel of Intelligence Creating Knowledge Only Invented by SDS
8...Design OHS/DKR Failed Search Fast Easy Path Avoid Reinvent Wheel
9...Reply in Context to Email that is Part of Original Message
10...OHS/DKR Based on SDS Design
11...Context Management Knowledge Space Organizational Memory Not Fun
12...Many people report that links provide useful, timely knowledge to

ACTION ITEMS.................. Click here to comment!

1...Need clarification on corrections Eric is actually making.
2...Need explanation of how a reply can be delivered in "context"
3...Don't "threads" very quickly diverge from the actual subject

CONTACTS 
0201 - Bootstrap Institute
020101 - Mr. Douglas C. Engelbart, Ph.D.

SUBJECTS
Design OHS Architecture
Architecture Translating Text to XML then HTML
OHS/DKR Architecture, 000824
SDS Proof of Concept for DKR, 000426, Jack Park
Links Confusing Prefers Printed Material, 000504
SDS Design Basis for OHS Proposed by Doug Engelbart, Eric Armstrong
Architecture What OHS/DKR Trying to Accomplish Translate Email Text

2209 -
2209 -    ..
2210 - Summary/Objective
2211 -
221101 - Follow up ref SDS 41 0000, ref SDS 37 0000.
221102 -
221103 - Received ref DRT 1 0001 from Eric Armstrong reporting meeting with
221104 - Doug Engelbart on 000822 to develop architecture for OHS/DKR.
221106 -  ..
221107 - This meeting aligns with Eugene Kim's letter on 000815 reporting plans
221108 - for Doug to meet with team members on finalizing the architecture for
221109 - the OHS/DKR project. ref SDS 41 LJ4L
221110 -
221111 -     [On 000825 Brian Lincoln reports meeting with Doug on OHS/DKR
221112 -     learning curve. ref SDS 42 0001
221114 -      ..
221115 -     [On 000830 Grant Bowman reports meeting with Doug on 000829 about
221116 -     marketing. ref SDS 46 L52W
221118 -  ..
221119 - Eric submitted a second letter, ref DRT 2 0001, a few hours later
221120 - clarifying the notes in the first letter.  He says...
221121 -
221122 -     I believe I got some things wrong. I know I got some things wrong.
221123 -     But a lot of the notes about how the system is "really" going to
221124 -     work were not reflected in the discussion I had. ref DRT 2 6V7H
221125 -
221126 -        Need clarification on corrections Eric is actually making.
221128 -  ..
221129 - In the second letter Eric relates the system uses totally vanilla
221130 - email clients. There is no plan to do anything that works with email
221131 - messages. Instead, the messages will be stored in OHS-form, accessible
221132 - via the web. ref DRT 2 0001
221133 -
221134 -     [On 010124 Eric recalls using a vanilla browser and proxy server,
221135 -     but does not cite the record. ref SDS 62 UK5G
221136 -
221137 -
221139 -  ..
221140 - SDS Model Uniform Display and Access Work Record
221141 - OHS/DKR Trying to Accomplish Access to Record of Work History
221142 - Email Translated into XML HTML Access Work Record Displayed in Browser
221143 -
221144 - Eric reports Doug presented an architecture for OHS/DKR technology
221145 - that would translate email into XML then HTML for display in browsers,
221146 - and which would then provide access to the historical record of daily
221147 - work. ref DRT 1 QM6J
221149 -  ..
221150 - This seems to accomplish the purpose of Knowledge Management explained
221151 - in Doug's 1972 paper reviewed on 000327. ref SDS 17 3971
221152 -
221153 -     [On 000929 followed up. ref SDS 50 6S9H
221155 -  ..
221156 - Eric feels Doug's proposal is the Rod Welch system, ref DRT 1 2W6M,
221157 - which uses links to create an effective Knowledge Space, developed on
221158 - 960620, ref SDS 4 3516, and explained and illustrated in a letter to
221159 - Eric and the unrev team, ref DIP 1 0001, on 000122. ref SDS 11 0001
221160 -
221161 -        [On 000929 Doug indicated support for SDS. ref SDS 50 6S9H
221163 -         ..
221164 -        [On 001027 confirmed support for the SDS design. ref SDS 57
221165 -        XV6F
221167 -         ..
221168 -        [On 010329 Jack Park reports plans to independently develop an
221169 -        SDS program called something like "Structured Dialog System."
221170 -        ref SDS 63 XZ4K
221172 -         ..
221173 -        [On 010913 Eric comments favorably on SDS. ref SDS 66 1U3O
221175 -  ..
221176 - This supplements Eric's letter on 000812 raising concerns about
221177 - aspects of OHS design, ref SDS 40 UY9L, which he feels makes
221178 - maintenance a "headache." ref SDS 40 HS9F  On 000811 he objects to
221179 - links in email, cited again today, per below. ref SDS 0 FJ5H
221180 -
221181 -        [On 000928 Eugene Kim reports that the project objective is
221182 -        settled. ref SDS 49 X57F
221184 -  ..
221185 - On 000405 Doug submitted a letter to the DKR team proposing
221186 - addressability in email. ref SDS 18 2484
221188 -  ..
221189 - On 991222 Doug's objective to augment human intelligence, ref SDS 9
221190 - 3977, is accomplished by SDS; Doug asked how Knowledge Space works.
221191 - ref SDS 9 1140  A letter to Doug requested feedback on forming
221192 - alliances for pilot testing. ref SDS 9 6324
221193 -
221194 -        [On 000829 discussed with Pat Lincoln. ref SDS 43 1056
221196 -         ..
221197 -        [...discussed with Jack Park. ref SDS 44 XH4K
221199 -         ..
221200 -        [...comments from Morris. ref SDS 45 0001
221202 -      ..
221203 -     Eric's report seems to align with Eugene Kim's favorable comments
221204 -     on SDS to the SRI team on 000504, which was supported by Jack
221205 -     Park. ref SDS 26 0625
221207 -      ..
221208 -     Earlier on 000426 Jack indicated SDS provides a proof of concept
221209 -     for aspects of KM. ref SDS 22 3315
221211 -      ..
221212 -     On 000407 submitted a letter with copy to Doug suggesting that SRI
221213 -     use the SDS record on the web to demonstrate ability to perform KM
221214 -     development, that supports fund raising. ref SDS 19 0786
221216 -  ..
221217 - Eric says Doug's proposal to implement SDS system, per above,
221218 - ref SDS 0 PU5N, does not include subjects, which he is calling
221219 - catagories. ref DRT 1 NO4N  Eric's letter on 000601 submitting version
221220 - 0.7 of OHS design specs proposed catagories are a significant
221221 - requirement for retreiving information from the record, based on the
221222 - Traction program he investigated. ref SDS 30 5933
221223 -
221224 -     This seems to align with the record on 000623 showing lack of
221225 -     agreement on incorporating catagories into the OHS design.
221226 -     ref SDS 34 5040  Eric is not aware of SDS subject indexing using
221227 -     organic structures for context management of complex, set out in
221228 -     the record on 890523.
221229 -
221230 -            [On 010916 Eric begins to grasp SDS power to organize work
221231 -            more effectively than other methods. ref SDS 67 0001
221233 -      ..
221234 -     On 000718 proposal to NASA omitted catagories, ontology and
221235 -     subjects. ref SDS 35 7520
221237 -      ..
221238 -     On 000725 Pat Lincoln reported that ontology would be included in
221239 -     OHS design proposal to NASA under a separate submission.
221240 -     ref SDS 36 7426
221241 -
221242 -
221243 -
2213 -

SUBJECTS
Knowledge Management SDS Invented Wheel OHS/DKR Failed High Tech Tou

2803 -
280401 -  ..
280402 - Design Changes Failed OHS/DKR Development
280403 - Changing Tools Design OHS/DKR Development Failed
280404 - Wheel of Intelligence Creating Knowledge Only Invented by SDS
280405 - Design OHS/DKR Failed Search Fast Easy Path Avoid Reinvent Wheel
280406 -
280407 - Eric relates history showing various design tasks have been started,
280408 - but not completed by the team. ref DRT 1 I94F
280410 -  ..
280411 - Eric continues in his letter...
280412 -
280413 -     a.  At SRI we started out doing use cases, then stopped that and
280414 -         considered...
280416 -  ..
280417 - On 000324 use case analysis proposed during OHS/DKR launch meeting.
280418 - ref SDS 16 6052  On 000503 Eric listed use cases for good management
280419 - that email helps accomplish; did not include requirement for
280420 - intelligence that converts information into knowledge. ref SDS 24 F82E
280421 -
280422 -            [On 001214 Jack cites a lot of use case work has been done,
280423 -            and suggests alternate track. ref SDS 60 01RR
280425 -  ..
280426 - Eric continues in his letter...
280427 -
280428 -     b.  Interfaces, then stopped that and started...
280430 -          ..
280431 -     c.  Architecture. We have now stopped that and have started this
280432 -         new process, where the...
280434 -          ..
280435 -     d.  System architecure comes predefined, rather than being built
280436 -         up in response to the needs demonstrated by use cases and
280437 -         analysis.
280439 -  ..
280440 - Additionally, OHS/DKR team discussed extensivley...
280442 -          ..
280443 -     e.  XML editors for organizing the record, disucssed by Eric on
280444 -         000223, ref SDS 14 QQ96, following up prior work on 000129
280445 -         when Eric suggested XML. ref SDS 13 0957
280447 -          ..
280448 -     f.  IBIS proposed in Eric's OHS/DKR specification on 000614 to
280449 -         organize the work. ref SDS 31 4939  IBIS-style discussion
280450 -         supported by categories, proposed on 000622. ref SDS 33 6204
280452 -          ..
280453 -     g.  3-layer architecture proposed by Jack Park on 000623.
280454 -         ref SDS 34 EK3W
280456 -          ..
280457 -     h.  Wiki technology for collaboration being developed for OHS/DKR
280458 -         on 000622. ref SDS 33 9000  On 000727 team discouraged about
280459 -         slow progress using Wiki. ref SDS 37 9000
280461 -  ..
280462 - As a result, Eric does not "trust" the design nor the process.
280463 - ref DRT 1 MF4M
280464 -
280465 -            [On 001130 IBM engineers unable to develop Knowledge
280466 -            Management solution by converting Lotus Notes, because
280467 -            nobody could agree on scope, tools, design, anything...
280468 -            ref SDS 59 F26K
280470 -             ..
280471 -            [On 080124 Morris switches development tools and design
280472 -            scheme again for SDS Windows application, similar to
280473 -            experience on OHS/DKR. ref SDS 76 FW4O
280475 -  ..
280476 - If Eric is correct, that Doug's proposed design is similar to SDS,
280477 - then he does not "trust" the SDS design.  This aligns with his letter
280478 - on 000811 objecting to email with links into Knowledge Space, and
280479 - restated today. ref SDS 0 PU5N
280481 -  ..
280482 - OHS/DKR team started on 000324 to apply knowledge of existing tools
280483 - that could be assembled so that people would not have to "reinvent the
280484 - wheel." ref SDS 16 JY4F  Over the ensuing months of effort, recounted
280485 - by Eric today, the team searched frantically for tools, but soon
280486 - recognized they did not know what tools to use.
280488 -  ..
280489 - On 000601 Marcelo Hoffman cautioned during a meeting at SRI to avoid
280490 - "high tech tourism" seaching the Internet for solutions that avoid
280491 - having develop capabilities required for Knowledge Management.
280492 - ref SDS 29 1007
280494 -  ..
280495 - People began asking "what are we trying to accomplish (see later case
280496 - study on 020530, ref SDS 71 2B3I) because nobody knows what the
280497 - "wheel" of intelligence does to convert information into knowledge,
280498 - noted by Eric on 000503. ref SDS 24 5033  A month later on 000615 the
280499 - entire team agreed they did not have enough knowledge to create
280500 - technology for Knowledge Management. ref SDS 32 6271
280502 -  ..
280503 - The SDS design is set out in POIMS explaining a technology based on
280504 - the architecture of human thought. ref OF 1 0561
280506 -  ..
280507 - Eric says the design does not provide a gui interface, but instead
280508 - puts Augment's view-control commands into a central position, which
280509 - requires using a complicated mnemonic language. which power users
280510 - eventually come to like, but normal users hate it, and can't be
280511 - bothered to use it. If the system depends on using it, the system is
280512 - doomed. ref DRT 1 WY3L
280514 -  ..
280515 - On 000622 Eric reported that Augment information he reviewed showed
280516 - the type of user interface that was presented on 000822. ref SDS 33
280517 - 9626
280519 -      ..
280520 -     [On 000825 Brian Lincoln reports Doug's concern for easing
280521 -     learning curve, and comments personal preference for direct
280522 -     functions to enhance productivity by saving time. ref SDS 42 PX3M
280524 -      ..
280525 -     [On 001010 Sheldon reports progress developing a Magic Lense
280526 -     application to support Augent efficient control and modifier hand
280527 -     and a keying hand. ref SDS 52 G8R3
280528 -
280529 -
280530 -
280531 -
2806 -

SUBJECTS
Email Reply in Context, Eric Armstrong 000824

3003 -
300401 -  ..
300402 - Reply in Context to Email that is Part of Original Message
300403 -
300404 - Eric wants...
300405 -
300406 -     ...an email system that delivers a reply "in context" so that it
300407 -     appears as part of the original message. ref DRT 1 YF4N
300408 -
300409 -         Need explanation of how a reply can be delivered in "context"
300410 -         that is different from placing the letter in the chronology of
300411 -         the information stream, as his letter today is incorporated
300412 -         into the SDS record.  Below, he objects to using historical
300413 -         archives. ref SDS 0 FJ5H  What then does "context" mean?  Need
300414 -         an example.
300415 -
300416 -            [On 001019 Gil Regev also requested context. ref SDS 56
300417 -            TF5H
300419 -             ..
300420 -            [On 011102 Traction supports Eric's objective for reply in
300421 -            context as part of original email. ref SDS 69 9I3U
300423 -  ..
300424 - Eric's letter continues...
300425 -
300426 -     Register subject threads of interest - always seeing threads that
300427 -     are really new, but not being bothered with additional messages to
300428 -     old threads that I've already chosen to ignore. ref DRT 1 PB5H
300429 -
300430 -         Don't "threads" very quickly diverge from the actual subject
300431 -         that describes the content of a letter, as evidenced from the
300432 -         record on the DKR project.
300434 -          ..
300435 -         The character of email as a stream of conscious reacation,
300436 -         causes people to raise ancillary issues under a common subject
300437 -         description, with the result that after a few responses there
300438 -         is no correlation to the original subject ("thread").  This
300439 -         pretty much means we have to open the letter and read a bit to
300440 -         see if it is material to our objectives, interests and
300441 -         commitments, and if not, then delete it.  If so, it needs to
300442 -         be incorporated into the management environment.  If we don't
300443 -         do it, and cannot get someone who does, then the complexity
300444 -         and risks of our life are increased exponentially.
300446 -      ..
300447 -     Search email content and display information in context, to
300448 -     improve current systems that he indicates are "awful." ref DRT 1
300449 -     AM6O
300450 -
300451 -         Organizing information by subject, Eric uses "catagories,"
300452 -         enables finding information in context, however, this is a
300453 -         fairly challenging undertaking.
300455 -          ..
300456 -         The solution requires more than a better email program,
300457 -         because very soon everything becomes a jumble, which leads to
300458 -         a need for an organic subject structure called out in POIMS.
300459 -         ref OF 1 1110
300460 -
300461 -
300462 -
300463 -
3005 -

SUBJECTS
Links Like Rod Welch System Are Not Effective for OHS
Design Problems
Links Rod Welch System Not Useful for Knowledge Repository
Archives Eric Not a Fan Does Not Want Access to Record
Links Recommends Ought to be a Law Against Links in Email
OHS/DKR Architecture, 000824
Click Links Email Objectionable

5209 -
521001 -  ..
521002 - OHS/DKR Based on SDS Design
521003 - Context Management Knowledge Space Organizational Memory Not Fun
521004 -
521005 - Eric objects to links for accessing the record that aids understanding
521006 - context; he indicates that...
521007 -
521008 -     Doug's current proposal based on the SDS design will clutter the
521009 -     inbox with link-containing messages, and feels that reconstructing
521010 -     an argument from a series of messages like that will be next to
521011 -     impossible.  It will require consulting the archive. ref DRT 1
521012 -     8E5M
521014 -  ..
521015 - Lack of interest in organizational memory aligns with Eric's objection
521016 - on 000811 reviewing a letter on Peter Drucker's work provided through
521017 - a link in an email. ref SDS 39 0001  On 000517 he expressly objected
521018 - to Knowledge Space, ref SDS 28 5184, reflecting resistance to a new
521019 - way of working, cited previously on 960612 by Lynn Conway. ref SDS 3
521020 - 1368  Similarly, on 990712 Morris objected to links in email.
521021 - ref SDS 6 5837
521023 -            ..
521024 -           [On 001015 Eugene Kim explained understandings for Eric's
521025 -           concerns about Rod's system. ref SDS 53 QG6J
521027 -            ..
521028 -           [On 001018 Eric reports linking presents design problem for
521029 -           OHS architecture. ref SDS 55 M4W7
521031 -            ..
521032 -           [On 010913 Eric comments favorably on SDS archive links.
521033 -           ref SDS 66 1U3O
521035 -            ..
521036 -           [On 020820 Murray Altheim reported people fail to use
521037 -           explicit links, ref SDS 72 UV6G, because accuracy for
521038 -           communication takes too much effort without SDS, ref SDS 73
521039 -           VN6H; experts who promoted purple numbers all seem to have
521040 -           given up adding granular addressability for working
521041 -           intelligently. ref SDS 72 0K64
521043 -  ..
521044 - If using the record is too difficult, i.e., next to impossible, why
521045 - does the SDS record on the Internet use the record, while others,
521046 - using conventional IT methods, rarely align their work with project
521047 - history, including Eric, except for his project in Texas reported on
521048 - 000517?
521050 -  ..
521051 - Maybe it is not as difficult as it seems to align the record, if
521052 - effective tools are available for managing the record.  This is why
521053 - experience is essential to enable people to allay false fears.  For
521054 - example on 000125 Eric worried that adding connections to communiation
521055 - would "boggle the mind." ref SDS 12 3867
521056 -
521057 -           [On 000926 SDS technology reduces diligence required to
521058 -           perform knowledge work. ref SDS 48 6Q6J
521060 -            ..
521061 -           [On 000920 asked who are customers for KM? ref SDS 47 2V5F
521063 -            ..
521064 -           [On 001004 letter to DKR team notes people don't really want
521065 -           to use KM, they want to talk about it, with the result that
521066 -           there are no customers. ref SDS 51 0A3H
521068 -            ..
521069 -           [On 020924 Jack Park reports that Eric changed his mind and
521070 -           has indicated Communication Metrics implemented by SDS is
521071 -           quite powerful. ref SDS 74 RN40
521073 -  ..
521074 - Eric continues...
521075 -
521076 -     Unlike many users, apparently, I am not a big fan of archives. In
521077 -     fact, I hate them. I have my own archives -- copies of the
521078 -     messages I care about. I search them when I need to. So a system
521079 -     in which the archive is the most (and possibly only) useful part
521080 -     of the system holds little interest for me. ref DRT 1 RM6I
521081 -
521082 -            [On 010122 BrowseUp makes creating links to the archives
521083 -            faster and easier, provides tools to manage complexity.
521084 -            ref SDS 61 IN9O
521086 -             ..
521087 -            [On 010912 Eric supports "maleable archives." ref SDS 65
521088 -            0001
521090 -             ..
521091 -            [On 010916 Eric seems to clarify the report today that he
521092 -            does not like archives because it is hard to find anything
521093 -            using conventional technology. ref SDS 68 KA6H
521095 -             ..
521096 -            [On 010916 Eric reported SDS has amazing memory that
521097 -            obviously works for finding everything. ref SDS 67 0001
521099 -             ..
521100 -            [On 020530 Eric says people will use SDS to create
521101 -            connected record if given opportunity. ref SDS 71 RK5M
521103 -  ..
521104 - On 000421 Doug Engelbart explained vision for Guide to Running Nics
521105 - that project history should be maintained. ref SDS 23 2867
521107 -  ..
521108 - On 000227 Eric expected connecting the record would improve
521109 - productivity, save time and encourage widespread use of knowledge,
521110 - contrary to the objection today. ref SDS 15 1248  But, earlier on
521111 - 000125 he worried that a connected record would "boggle the mind."
521112 - ref SDS 12 3975
521114 -  ..
521115 - This aligned with recommendations from a professional event on 940611
521116 - explaining that a good record aids creativity by providing a resource
521117 - that can be assembled in alternate ways to identify useful patterns
521118 - for different contexts. ref SDS 1 5EPD
521119 -
521120 -            [On 001108 Eric reports on creativity aids that make sense.
521121 -            ref SDS 58 0001
521123 -  ..
521124 - On 000426 Eric cited advantages of a project record connected to
521125 - relevant history. ref SDS 21 0786  On 000423, he called on the team to
521126 - support capability for discussion and deliberation that records
521127 - decisions and their rationales so that knowledge gained in the process
521128 - can be applied to future projects. ref SDS 20 5933
521130 -  ..
521131 - On 000505 project requirements prepared by Eric provide a connected
521132 - environment that "boggles the mind." ref SDS 27 4524
521133 -
521134 -           [On 001004 letter to DKR team notes people don't want their
521135 -           mind boggled. ref SDS 51 HP7M
521136 -
521137 -           [On 001017 Eric developed source code for OHS. ref SDS 54
521138 -           IO6J
521140 -            ..
521141 -           [On 070126, people give up on all 3 elements of knowledge
521142 -           management, reported ref SDS 75 QH8L, capturing the record,
521143 -           adding subjects to organize the record, ref SDS 75 WV5K, and
521144 -           adding connections that yield the power of knowledge.
521145 -           ref SDS 75 YV4F
521146 -
521148 -  ..
521149 - Many people report that links provide useful, timely knowledge to
521150 - expedite the work, see research on 990712. ref SDS 6 X3G6
521151 -
521152 -            [On 010907 many people have written unsolicited approval of
521153 -            SDS records on the Internet. ref SDS 64 KX3L
521154 -
521155 -            [On 020329 links used effectively to expedite understanding
521156 -            of POIMS. ref SDS 70 Z35I
521158 -  ..
521159 - This issue seems to divide along personality, context and scale.  Some
521160 - people don't want the responsiblity of knowledge under any
521161 - circumstances, per point 4, above.  Context impacts willingness to
521162 - click on a link.  Where knowledge is provided that solves an immediate
521163 - problem, people seem pleased to have links.  If, however, people are
521164 - doing something for which there is no immediate need or benefit, links
521165 - seem burdensome.  A lot of links seem confusing, per Doug's
521166 - explanation of the scale problem reviewed on 991222. ref SDS 9 1596
521168 -  ..
521169 - The larger issue that arises from objecting to using history is that
521170 - people feel uncomfortable analysing daily information to understand
521171 - correlations, implications and nuance that augments intelligence.
521172 - Desire to avoid responsibility for knowledge, per point 4 above,
521173 - ref SDS 0 PY3L, aligns with Eric's preference on 000503 to avoid
521174 - "knowledge" because it opens issues of cognitive science that make
521175 - many people uncomfortable. ref SDS 24 5033
521176 -
521177 -
521178 -
521179 -
521180 -
5212 -
Distribution. . . . See "CONTACTS"