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 31, 2000 05:15 PM Thursday; Rod Welch

SRI team reviews Open Source managment for software development.

1...Summary/Objective
2...Fun Work More Productive than Diligent Adherence to Requirements
3...Permission to Use Causes Objection to Open Source Development
4...Solve Permission to Use, Apply to Content of Mailing List Email
5...List Charter Proposed Alternate to Permission to Use
6...OHS/DKR Project Has Shown Status of KM Worldwide
7...Termite Software Development Requires Focus Work to Make Connections
8...Extensible Architecture Requires Look up Table


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

CONTACTS 

SUBJECTS
Property Rights in Knowledge Organization
Proposal on License
Participation Rights of Contributors
Open Source Assignments Require Authority
Open Source Causes Disputes Delays Progress
Management Cannot Make Assignments
Arguments About License and Ownership
Open Source Termites Develop What They Want, Hope It Fits Design
Command Control Record Leadership
Productivity Ergonomics Imact Management
Productivity Impacted by Motivations of Employees and Contributors
Payment Higher Rewards Reflecting Big Needs Improves Productivity
Termite Design Open Source Rejects Management Interference

3015 -
3015 -    ..
3016 - Summary/Objective
3017 -
301701 - Follow up ref SDS 50 H56L, ref SDS 44 0001.
301702 -
301703 - Received ref DRT 1 0001 from Paul Fernhout explaining advantage of
301704 - open source software development based on a metaphore of termites
301705 - constructing individual "nests" that eventually combine to form an
301706 - effective nest for the entire group. ref DRT 1 0001
301708 -  ..
301709 - On 000424 Eric Armstrong reported risk of failure increases from going
301710 - in wrong direction on project design. ref SDS 18 P95I  Feel Good
301711 - management empowers creativity, increases risks, reviewed on 911123.
301712 - ref SDS 1 1331
301713 -
301714 -     [On 001012 Grant Bowman promotes Open Source work regardless of
301715 -     whether it aligns with objectives, in order to develop synergy
301716 -     that may produce something useful. ref SDS 57 B4P2
301718 -      ..
301719 -     [On 001015 Eugene Kim proposes initial design for open source
301720 -     development. ref SDS 58 B4P2
301722 -      ..
301723 -     [On 001031 Paul discusses open source development. ref SDS 60 TX7J
301725 -      ..
301726 -     [On 001101 Jack Park submits article on using good management
301727 -     practices for open source projects. ref SDS 61 MX6H
301729 -  ..
301730 - Paul observes the OHS/DKR project does not follow the termite method,
301731 - but instead imposes on contributors (termites) pre-architected design
301732 - requirements. ref DRT 1 IC8I  Paul cautions against reliance on
301733 - finance to acquire and direct the talent needed to develop Knowledge
301734 - Management, ref DRT 1 EZ8I, based on research at following...
301735 -
301737 -      ..
301738 -     http://www.gnu.org/philosophy/motivation.html
301739 -
301741 -      ..
301742 -     [On 001017 SRI seeking NIH support that requires project audits.
301743 -     ref SDS 59 0784
301744 -
301746 -  ..
301747 - Fun Work More Productive than Diligent Adherence to Requirements
301748 -
301749 - Paul expects the Termite production method for open source
301750 - development would be more fun, and argues that fun is an important
301751 - part of a successful effort. ref DRT 1 DL4H
301752 -
301753 -     [On 000926 concern that working intelligently takes too much
301754 -     diligence. ref SDS 52 004M
301756 -      ..
301757 -     [On 001004 explained that adding "intelligence" to engineering and
301758 -     management is also fun, plus it saves time and improves earnings.
301759 -     ref SDS 56 6N9G
301761 -      ..
301762 -     [On 001101 article on IT development says fun is not an excuse not
301763 -     to get the work done. ref SDS 61 OX6N
301765 -  ..
301766 - This reflects classic tension between management by empowerment, and
301767 - top-down command and control methods, reviewed on 931130, ref SDS 3
301768 - 9677, i.e., viewing work as fun or drudgery.  On 970710 empowerment
301769 - was reviewed at Intel to enhance creativity, but impairs management.
301770 - ref SDS 5 1431  On 970119 managers and engineers empowered to express
301771 - emotions in confrontational communication because it feels good.
301772 -
301773 -         [On 001012 Grant Bowman promotes Open Source work regardless
301774 -         of whether it aligns with objectives, in order to develop
301775 -         synergy that may produce something useful. ref SDS 57 B4P2
301777 -          ..
301778 -         [On 001004 proposed that Eric develop schedule based on
301779 -         Termite methods. ref SDS 56 YP6I
301781 -      ..
301782 -     On 000125 Dick Karpinski related experience that shows software
301783 -     projects fail without strong management that sets an objective and
301784 -     identifies small scale work packages, which combine to provide
301785 -     desired capability using a general plan. ref SDS 8 8960
301786 -
301787 -        [...below Eric Armstrong comments on need for focused work to
301788 -        connect the product of "termite" software development.
301789 -        ref SDS 0 UW4G
301791 -      ..
301792 -     Brooklyn Bridge, BART, English Channel Tunnel, Landing on the
301793 -     moon, all seemed to get done with finance.
301795 -      ..
301796 -     The people who did the work were motivated by big needs reflected
301797 -     by payment of large sums to motivate productive, focused attention
301798 -     on getting specific things done that people would not otherwise
301799 -     perform absent incentive of payment.
301801 -      ..
301802 -     Termites are wired to build the same thing today that they built
301803 -     yesterday, last month, and 10,000 years ago.  That may be why a
301804 -     lot of individual termite structures can combine to form a useful
301805 -     composite environment.  However, if the result of connecting
301806 -     different termite structures, is not acceptable to one of the
301807 -     termites, who might want indoor plumbing, or cable TV, there is
301808 -     not a lot of opportunity to change the design.
301810 -      ..
301811 -     Creating Knowledge Management is about improving the design, not
301812 -     assembling existing designs.
301813 - ..
301814 - Paul notes that to attract open source developers, a bottom-up
301815 - termite-like approach might work better, per above. ref SDS 0 0001  He
301816 - explains a "termite" empowerment management method is less risky for
301817 - an open source developer to build their own little pile they like and
301818 - which makes sense to them and to then offer it under some terms they
301819 - choose to the rest of the world (which accepts or rejects it as the
301820 - world sees fit), than to build some custom part they have less
301821 - interest in as part of a grand design that may or may not ever be
301822 - complete enough to use. ref DRT 1 U08K
301823 -
301824 -     On 000125, reviewed challenge of getting things done using open
301825 -     source methods. ref SDS 8 8960
301827 -      ..
301828 -     This method could be explained as laziness that avoids the demands
301829 -     of hard work to formulate a useful design, preferring to focus on
301830 -     narrow, short-term perspectives that pop into the mind at the
301831 -     moment, and hope the result can be made useful in a larger cause,
301832 -     by others who are willing to make the effort to think in broader
301833 -     terms.
301834 -     ..
301835 -     Software "termites" don't know what to build in order to
301836 -     create better knowledge tools, reported on 000615, ref SDS 37
301837 -     6271, confirming earlier reports on 000324. ref SDS 12 4877  Many
301838 -     tasks are counterintuitive, and can only be grasped once the
301839 -     entire capability is assembled and used for several months, up to
301840 -     a year. So, unlike termites, who build the same thing over and
301841 -     over, engineers for Knowledge Management need guidance, leadership
301842 -     and incentives to build something new and better.
301843 -
301844 -        [On 000921 proposed meeting to review pending action items
301845 -        needed to move project forward. ref SDS 51 UP7F
301847 -         ..
301848 -        [On 000927 Eric reports activity on project quiet. ref SDS 53
301849 -        0001
301851 -         ..
301852 -        [On 000929 Eugene reports team was delayed by not understanding
301853 -        deliverable, but contributors can begin production. ref SDS 54
301854 -        0001
301856 -         ..
301857 -        [On 000929 offered to meet with Doug to support DKR effort, and
301858 -        review action items. ref SDS 55 0001
301859 -
301861 -  ..
301862 - Permission to Use Causes Objection to Open Source Development
301863 -
301864 - Paul advises that the "Permission to Use" rule...
301865 -
301866 -   http://www.bootstrap.org/dkr/discussion/1087.html
301867 -
301868 - ...(specifically the indemnification clause as it relates to software
301869 - patents) may prevent "termite" contributions through self-organizing,
301870 - described above, ref SDS 0 0001, from happening at least for him.
301871 - ref DRT 1 W69M and ref DRT 1 VT5O
301873 -  ..
301874 - He feels "Permission to use" makes it risky for developers to talk
301875 - about their own related open source efforts on this list, lest that be
301876 - taken to mean they are contributing their code under "Permission to
301877 - use" (even though they might be happy to contribute it under an open
301878 - source license of their choosing which disclaims warranty, liability
301879 - etc.). Thus, the piles such developers may be building will not find a
301880 - discussion here -- and thus never be connected with other related
301881 - piles produced by other list members. ref DRT 1 085F
301883 -  ..
301884 - This supplemnts Paul's letter yesterday ref DRP 10 0001 expressing
301885 - appreciation for extra effort Eugene Kim is making to address open
301886 - source licensing issues, following up announcement on 000802 that the
301887 - SRI team for the DKR has adopted an open source license. ref SDS 44
301888 - 0001
301889 -
301890 -
301892 -  ..
301893 - Solve Permission to Use, Apply to Content of Mailing List Email
301894 -
301895 - Paul proposes "Permission to use" would be much more effective for him
301896 - if it
301897 -
301898 -    a.  related specifically to the content of the mailing list (or the
301899 -        finished classroom video stream naturally), ref DRT 1 UH5O,
301901 -         ..
301902 -    b.  had a clause promising preserving attribution and keeping
301903 -        content intact, and, ref DRT 1 0G6H,
301905 -         ..
301906 -    c.  said nothing about code contributions. ref DRT 1 BG6K
301908 -  ..
301909 - Paul feels these terms and conditions permit code contributions to be
301910 - negotiated and accepted into a "main release" after licensing
301911 - negotiation with the organizers of any particular larger effort.
301912 - ref DRT 1 WG6M
301914 -  ..
301915 - Paul maintains that creating an unlimited scope agreement is
301916 - self-defeating as developers just won't contribute (and possibly might
301917 - not hold up in court anyway as it is overly broad and not signed).
301918 - ref DRT 1 GF6K
301919 -
301920 -
301922 -  ..
301923 - List Charter Proposed Alternate to Permission to Use
301924 -
301925 - Paul argues that a list charter and peer pressure would prevent a
301926 - developer of proprietary software with no open source component from
301927 - hawking their wares on the list, and one would expect (demand?) that
301928 - any code posted to the list be put under at least one specific open
301929 - source license (or made public domain) by the poster at the time of
301930 - posting in exchange for others reviewing it. So, to an extent
301931 - "Permission to use" addresses many issues perhaps better handled by a
301932 - list charter.
301933 -
301934 -
301936 -  ..
301937 - OHS/DKR Project Has Shown Status of KM Worldwide
301938 -
301939 - Paul lists four separate activities occurring on Doug's project...
301940 -
301941 -    A.  Discussion of world problems (mostly faded)
301942 -
301943 -    B.  Discussion of other related development efforts (posting links,
301944 -        etc.)
301946 -         ..
301947 -    C.  Discussion of individual's own efforts in the open source
301948 -        software arena
301950 -         ..
301951 -    D.  Development of a branded "Open Hyperdocument System" by
301952 -        Bootstrap
301954 -  ..
301955 - B is continuing to happen, and for me has been the greatest value of
301956 - this list. ref DRT 1 6W4K
301958 -  ..
301959 - We have through termite-like processes already built quite a resource
301960 - in the unrevii list in the B area of seeing what is currently going on
301961 - around the globe in the direction Doug and others set off in back in
301962 - the 60s. ref DRT 1 GW6O
301963 -
301964 -    This may reflect
301965 -
301966 -      a.  Lee Iverson's report on 000324. ref SDS 12 4877
301968 -           ..
301969 -      b.  Marcello Hoffman's report on 000330. ref SDS 14 0713
301970 -
301971 -
301972 -
301974 -  ..
3020 -
3021 -
3022 - 2318 received letter from Eric Armstrong
302301 -  ..
302302 - Termite Software Development Requires Focus Work to Make Connections
302303 -
302304 - Received ref DRT 2 0001 from Eric Armstrong responding to Paul's
302305 - letter, reported above.
302306 -
302307 - Eric expects a lot of code would be needed to make use of little
302308 - pieces of software submitted by termite software contributors,
302309 - proposed by Paul, ref SDS 0 0001, and seems to indicate this prevents
302310 - the method from being used often.
302312 -  ..
302313 - This seems to support classic project management methods, for the
302314 - reasons set out above, ref SDS 0 BN4J, on 000125. ref SDS 8 8960
302315 -
302316 -
302317 -
302318 -
3024 -

SUBJECTS
Extensible Code Users Customize to Suit Preference
OHS CDS Requirements,
DKR Design Ideas
Design Problems

3706 -
370701 -  ..
370702 - Extensible Architecture Requires Look up Table
370703 -
370704 - Follow up ref SDS 21 0583, ref SDS 17 5033.
370705 -
370706 - Eric notes Pauls' proposal for "termite" software development that
370707 - entails focused effort to connect contributions from open source
370708 - developers, per above, ref SDS 0 UW4G, leads to the subject of an
370709 - "extensible architecture", ref DRT 2 RM7G, which he cited on 000504,
370710 - ref SDS 21 0583, and is part of CDS specs reviewed in detail on
370711 - 000505. ref SDS 22 T75N
370713 -  ..
370714 - He recommends every object needs to be generated by a factory, which
370715 - is driven by a lookup table telling which class to use to instantiate
370716 - the object. You can then subclass an existing class to reuse its
370717 - behaviors, and also insert that object into the system. ref DRT 2 NM7J
370719 -  ..
370720 - This seems to align with Eric's proposed specs on extensible reviewed
370721 - on 000505. ref DRP 1 0528
370722 -
370723 -
370724 -
370725 -
370726 -
370727 -
370728 -
370729 -
370730 -
370731 -
3708 -