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: October 15, 2000 08:16 PM Sunday; Rod Welch

Eugene Kim comments about Grant Bowman's report on DKR architecture.

1...Summary/Objective
2...Vision Topology of OHS will be Email Repository
3...OHS Topology, Vision Set, Implementation will Come Later
4...Email OHS Objective with Lens on the Data
5...OHS Design for Addressable Email; Local Storage, View Control
.....2...Using Email to Deliver Intelligence Rather than Information
6...Open Source Requires Initial Design to Guide the Work
7...Implementation Requires Taking Steps to Get Started
8...Open Source Engineers Move Forward Even in the Wrong Direction
9...Compiler for Augment Expedites Updating Application
10...Augment Complexity Presents Challenge for Updating Application


..............
Click here to comment!

CONTACTS 
0201 - Eugene Eric Kim                                                                                                                                                    O-00000756 0101
020101 - Mr. Eugene Eric Kim                                                                                                                                              O-00000756 0101

SUBJECTS
Design OHS Architecture, Grant Bowman
Meeting Doug Engelbart on OHS Scope
Links Like Rod Welch System Are Not Effective for OHS
Local Storage with Serveletts for View Manipulation, Eric Armstrong
Topology Vision by Doug for for OHS being Formulated, Impelmentation
OHS Development Vector, Doug, 000419
Email Enhancement Tools, Hackable, 000428
OHS Architecture Email with Links, Eric Armstrong
Meeting with Grant Bowman Date Unknown on OHS Vision, Topology
Scope OHS Project Email System being Formulated, Impelmentation Later
Meeting Doug Engelbart Grant Bowman
Email System Proposed by Eugene, Lee (Strategy #2 on 000504??)
Email Conventional Best Com Method Available, 000502
Email XML Seed Team Use What is Available, 000302
Augment Small Capabilities
Email OHS/DKR Starter Technologies Develops Experience to Create KM
Augment NLS Update Doug's Program from the 60s Developed at SRI and
View Control Latency Objection Eric Raises
Spec Review OHS Completed Report to be Issued 001030

4821 -
4821 -    ..
4822 - Summary/Objective
4823 -
482301 - Follow up ref SDS 68 0000, ref SDS 67 0000.
482302 -
482303 - Received ref DRT 1 0001 from Eugene Kim responding to the letter from
482304 - Grant Bowman, ref DRP 5 0001, received on 001012, ref SDS 68 0001,
482305 - explaining discussions with Doug Engelbart on architecture for the OHS
482306 - and DKR system.
482307 -
482309 -  ..
482310 - Vision Topology of OHS will be Email Repository
482311 - OHS Topology, Vision Set, Implementation will Come Later
482312 - Email OHS Objective with Lens on the Data
482313 -
482314 - Follow up ref SDS 68 K49J.
482315 -
482316 - Eugene's letter today does not amend Grant's understanding that the
482317 - OHS system being developed is email with addressability.
482318 -
482320 -  ..
482321 - OHS Design for Addressable Email; Local Storage, View Control
482322 -
482323 - Follow up ref SDS 68 RI9J.
482324 -
482325 - I called Eugene Kim the next evening on 001016 and he advised that
482326 - Eric's OHS specifications, originally called the CDS in the record on
482327 - 000505, ref SDS 39 4864, were formally reviewed by the project some
482328 - time ago.  Earlier on 000725 Eugene mentioned at dinner following a
482329 - meeting at SRI that a review had occurred. ref SDS 48 YK4O
482331 -  ..
482332 - On 000614 Eric issued version 0.9. ref SDS 43 0782  The next day, on
482333 - 000615 Eugene proposed changing CDS to OHS. ref SDS 44 5880
482335 -  ..
482336 - Eugene plans to issue review comments by the end of the month.
482337 -
482338 -     [On 001017 Eric advises production is underway. ref SDS 69 0001
482339 -
482340 -     [On 010205 Eugene reports this review did not occur, ref SDS 70
482341 -     BM4F, and thus no report was issued.
482343 -  ..
482344 - Eugene understands Eric has two concerns that could potentially be
482345 - categorized as a "local storage" problem.
482346 -
482347 -     1.  View control
482348 -
482349 -         Eric wants to avoid the roundtrip latency involved in the
482350 -         current architecture. ref DRT 1 SQ7F
482351 -
482352 -            Need example to illustrate this problem.  Off hand it is
482353 -            not experienced in in SDS, but not sure without more
482354 -            explanation, and maybe an example.
482356 -          ..
482357 -         Eugene agree with this concern, but thinks that the current
482358 -         architecture is an important transitional step to start
482359 -         getting users, and that it can easily be modified to support
482360 -         local handling of views.
482361 -
482362 -
482363 -
482364 -
482365 -
482366 -
4824 -

SUBJECTS
OHS/DKR Architecture, 000824
Click Links Email Objectionable
Links Hard to Keep Orientation, Need Map
Links in Email Objection Eric Raises

6206 -
620701 -          ..
620702 -     2.  Using Email to Deliver Intelligence Rather than Information
620703 -
620704 -         Eric doesn't like the way Rod's system embeds links, not text.
620705 -         ref DRT 1 2R7M
620706 -
620707 -         Eugene agrees with Eric, even though his email system is set
620708 -         up to open a web browser with an e-mail link with one click.
620709 -         Eugene feels this is a very minor issue that can be solved in
620710 -         two ways:  include the text as well as the link for legacy
620711 -         clients, and build an "include content of link" feature into
620712 -         the OHS client at some point.
620713 -
620714 -            This aligns with Eric's letter on 000824 he does not like
620715 -            links in email and he does not like project archives.
620716 -            ref SDS 53 FJ5H
620717 -
620718 -
620719 -
620720 -
6208 -

SUBJECTS
Open Source Engineers Empowered to Do What They Want, If Useful Syner
Termite Production Uses Invisible Hand Rather than Directed Design
Productivity IT Declined Last 10 Years Info Highway
Management IT Poor
Open Souce Requires Initial Design

8707 -
870801 -  ..
870802 - Open Source Requires Initial Design to Guide the Work
870803 - Implementation Requires Taking Steps to Get Started
870804 - Open Source Engineers Move Forward Even in the Wrong Direction
870805 -
870806 - Follow up ref SDS 68 B4P2.
870807 -
870808 - Eugene falls in the middle of two competing methods for getting
870809 - started on the OHS project to implement Augment using Open Source
870810 - methods, ref DRT 1 ZU4H, even if the direction is incorrect, which
870811 - Grant discussed in his letter on 001012. ref SDS 68 B4P2
870813 -  ..
870814 - Eugene explains Doug's concern about moving in the wrong direction;
870815 - and notes that Grant is more concerned about not moving at all.
870817 -  ..
870818 - Eugene is equally averse to lack of action and action in the wrong
870819 - direction for the sake of action.  Successful open source development
870820 - depends heavily on strong initial design.  It's certainly feasible,
870821 - even highly recommended, to scrap initial code attempts entirely and
870822 - rewrite from scratch, but if you start off with a bad design to begin
870823 - with, an open source-style methodology is not going to help you in any
870824 - way whatsoever. ref DRT 1 PW4L
870825 -
870826 -    On 000426 Jack Park explained need for architecture to guide source
870827 -    code work. ref SDS 37 0304
870829 -     ..
870830 -    This seems to address Paul Fernhout's proposal on 000831 for a
870831 -    termite production method. ref SDS 55 0001
870833 -     ..
870834 -    [On 001017 Eric Armstrong submits notice of creating source code,
870835 -    but does not indicated the basis of the architecture. He is
870836 -    concurrently developing explanations of code so the architecture
870837 -    makes sense. ref SDS 69 0001
870838 -
870839 -
870840 -
870841 -
870842 -
870843 -
8709 -

SUBJECTS
Compiler Augment Run on Current Technology, Bowman, 001012
Augment Record Port to Current Technology, Bowman, 001012
Augment Complex to Update for Creating OHS Functions, 001012, Bowman
Augment Move Record to Modern Technology Compiler, Bowman, 001012
Augment NLS Update Doug's Program from the 60s Developed at SRI and
Augment Compiler Needed to Update Doug's Program from the 60s Develop
Compiler for Augment to Create Modern OHS

A509 -
A51001 -  ..
A51002 - Compiler for Augment Expedites Updating Application
A51003 - Augment Complexity Presents Challenge for Updating Application
A51004 -
A51005 - Follow up ref SDS 68 D4K2.
A51006 -
A51007 - Eugene reports that Augment source code is available in Augment, and
A51008 - Doug could potentially start dumping it once he gets going with
A51009 - a2h.pl.  Eugene thinks the idea of reimplementing L-10 compilers,
A51010 - which he attributes to Grant, would be a lot of fun, and would love to
A51011 - see someone do it.  Eugene would like to give it a go, if only for
A51012 - historical curiosity. What better way to study a language than to
A51013 - write a compiler for it? ref DRT 1 LJ5O
A51015 -  ..
A51016 - Eugene does not address Grant's report that...
A51017 -
A51018 -     1.  Implementing Augment is not an overnight project.
A51019 -
A51020 -     2.  Doug is concerned that the complexity of Augment restrains
A51021 -         progress on developing an update, and enlisting a cadre of
A51022 -         users in order to extend the knowledge developed from the
A51023 -         original work in the 60s.
A51025 -          ..
A51026 -     3.  Moving existing data and exporting it from Augment in a way
A51027 -         that links would be retained. If this is to evolve data should
A51028 -         be portable en mass. Unmoveable knowledge has less value.
A51029 -
A51030 -
A51031 -
A51032 -
A51033 -
A51034 -
A51035 -
A511 -
Distribution. . . . See "CONTACTS"