Kurtz-Fernhout Software

Date: Mon, 30 Oct 2000 14:18:01 -0500

From:   Paul Fernhout
Reply-To: unrev-II@egroups.com
Organization: Kurtz-Fernhout Software

To:     unrev-II@egroups.com

Subject:   Plan for HyperScope-OHS launch and development

Rod Welch wrote:


Thanks for important perspective on developing KM. Here is some feedback...



From that page:

  491648 -     KM is an operating system for people and organizations.  It's new;
  491649 -     and so hasn't been around for 50 years.  There is no school where
  491650 -     engineers can learn about KM.  Education for KM was proposed to
  491651 -     Stanford on 000929, but it is not a priority. ref SDS 21 RT6N
  491652 -     There are a lot of of seminars, books and magazines on KM, but
  491653 -     none explain how to produce KM; everyone shows up and asks what is
  491654 -     it, as reported on 991217. ref SDS 3 9030
  491655 -
  491656 -     As a result, On 000615 the DKR team reported 6 months of research
  491657 -     at SRI showed there is not enough knowledge to develop Knowledge
  491658 -     Management. ref SDS 15 6271,

And later:

  491713 -     The meeting at SRI on 001017 indicates work on OHS is pending
  491714 -     award of funds from sponsors. ref SDS 26 4877  Open Source
  491715 -     nominally thrives on volunteer work.  Over 10 months nothing has
  491716 -     been volunteered.  This doesn't mean people are lazy, indifferent,
  491717 -     or un-talented.  Doug has assembled an outstanding team, and SRI
  491718 -     is planning a strong core development team.  However, if there is
  491719 -     not enough knowledge, progress is difficult.  On 001017 Eric
  491720 -     notified the team he will volunteer source code soon, which is a
  491721 -     constructive sign, ref SDS 25 0001, but that conflicts with the
  491722 -     meeting at SRI where Eugene indicated the architecture will be
  491723 -     developed by a new team. ref SDS 26 TJ3G
  491724 -
  491725 -     On 001025 Doug Engelbart requested comments linked to the OHS
  491726 -     Launch Plan.  No one has done this because they are busy doing
  491727 -     what they want, reported on 001012, ref SDS 24 PT5M, and they
  491728 -     don't want to do KM, for reasons in the record on 001025,
  491729 -     ref SDS 30 XF4N  Software engineers want to work on interesting
  491730 -     technology; but, KM is not interesting, because there isn't enough
  491731 -     knowledge.
  491732 -
  491733 -     This record suggests open source is a good development method for
  491734 -     enhancing KM.  A core capability may be needed to guide such an
  491735 -     effort, as Eric noted on 000120. ref SDS 4 3002  Creating the core
  491736 -     capability requires doing things people don't want to do, which
  491737 -     conflicts with empowerment precepts of open source.

Jack Park wrote:

We continue to confuse the notion of *Open Source Software* with the notion of homebrew software. I do not believe (though I could be wrong) that Doug ever had any notion of homebrew software. His goal, near as I can tell, was and remains to create a body of software that is *Open Source*, and that has a radically different and specific meaning from the notions of free software that is hacked by thousands of individuals the world over, even though those projects do, indeed, result in open source products. It appears, at least to me, that a professionally managed project in which contributors receive compensation for software development is a greatly different project than one managed and built by volunteers.

Rod, Jack-

Interesting comments by both of you on the open source nature of the effort and related issues (including Jack's latest comment).

While I am definitely very much "pro open source" (my wife and I released a six person year effort under the GPL), I do think Rod has some good points about open source issues in general and as evidenced in this project. Knowledge Management (in terms of software implementations) is indeed a newer field than for example OS development, and it is harder to get people together to make such solutions because less people are aware of the issues (especially given all the "hot" proprietary work happening in dot-coms).

This UnRevII effort is in my opinion not very much an "open source" one to date. Even though it wants to be one, organizational issues and attitudes and habits formed from decades of doing things in a proprietary way have shaped key decisions and assumptions. There is little code contributed to any core -- on my side certainly in large part because "Permission to use" is not an open source license (but still governs this list). Several people here have done independent open source things, for example I've made the Pointrel system available under an open source license, but for liability and other reasons wouldn't ever consider "contributing" it to the UnRevII effort under the "permission to use" license. I understand there is another more open source license for the contributions made on the other list -- so that is an improvement, but it still has some ways to go (i.e. a distinction between (multiple) licenses for code, content, and collection) as well as not (yet) having much to put under it. Most of all, the UnRevII project is not open source because there is no body of work under an open source license. (There is some content produced like this list, but it is under a proprietary license.)

It is hard to explain everything that has been said or done on this list over the past year to violate the "spirit" of open source development (especially without sounding overly critical of people who are trying to do good stuff and make the world a better place). Perhaps it started with the lack of open sourcing Augment (i.e. there was no initial "gift of code") or the DARPA related development (although I'm very pleased to see the recent efforts like Augment->HTML).

Recently, for example, the suggestion Doug and SRI outline for picking a core team is not really open source (as I usually think of it) in spirit. To me, it indicates a collision of the corporate notion of software development with the notion of volunteerism (i.e. you don't pick a core open source team if much of the work is voluntary -- the core team in effect picks you, and then maybe you negotiate what that participation means or whether they can join the core.) Obviously, if it is a paid effort, it is typical for a manager to pick employees from candidates, so I understand Doug and SRI's reason for planning to do it that way.

Again, for example, bringing in an outside manager is not really open source in spirit as opposed to management by founders or participants. I'm not saying bringing in an experienced manager from outside the effort is not a good idea, or not required to get a grant, or won't work (or wouldn't be a smart move in any conventional software effort) -- I'm just saying that is a very different than open source as usual given the context of an active discussion list. But then again, open source means a lot of things to different people, and obviously Software Carpentry is a novel attempt to do open source in a environment involving some money changing hands in various ways.

Can the result of such efforts be "open source"? Of course -- just use such a license at the end and it technically qualifies. Lots of people get paid to develop open source software as part of their jobs. Can it succeed? Quite possibly. Perhaps it is the best way to go, given that Doug has a very specific idea in mind of what he wants to see implemented (a better Augment). I think the Augment approach has weaknesses (especially in the issue of the notion of documents vs. nodes, and in the notion of linking vs. searching) but I'd still like to see Doug and others make the attempt to build on that.

But, if you think deeply about the psychology of such announcements as the effect existing (many silent) participants on the mailing list, you will realize they would tend to make volunteers and potential volunteers less enthusiastic.

Both my wife and I have seen such things happen at some established non-profits we have worked at (like zoos and environmental groups), where there ends up being a huge distinction between staff and volunteers, with volunteers sometimes treated as cannon fodder to do scut work, and staff feeling entitled and prestigious. Volunteers have to be pretty committed to persist in the face of the difficulties and unpleasant environment the staff (or even other "more senior" volunteers) at such organizations often (unknowingly) create. Rather than foster a sense of inclusion into an ongoing public efforts, some announcements sometimes make the OHS/DKR effort sound more like a for-profit venture producing a product (with an open source license admittedly) with some expectation of leaching off of technically adept but otherwise socially stupid volunteer programmers. I understand that may not be the intent -- but it is to me the way things sometimes come across, based on the limitations email has in conveying tone.

Open source done with some money involved can be a very complex political animal. I guarantee the SRI effort will have internal conflicts as it tries to decide how much to open source and what not to (no matter what is said at the start), as usually government grants let the contractor keep all rights (except some the government has just for itself). And if not expressed directly, it will be expressed in what employees are assigned to an open core and what are assigned to a proprietary surrounding layer. I'm not singling out SRI specifically. Most non-profits can't do open source well for the same reason -- they need continued funding, and thus a desire to charge for information or for an effort to become "self supporting through sales" (as many grant request-for-proposals require) gets in the way of the desire to be open with information. What non-profit manager wants to think about letting loyal and hardworking staff go at the end of a contract? This is also one consequence of granting agencies still not "getting" open source, given the academic culture of copyrights and competition most reviewers are steeped in. There is little acceptance or understanding yet in granting agencies of the concept of "getting over a hurdle" by seed funding a specific open source effort done by committed people and then seeing what emerges and grows from that seed.

I had been reluctant at the start of the year to get involved with Bootstrap because there was no open source track record (i.e. no actual code or content put out there). I still feel this reservation was (and is) warranted. Nonetheless, I continue to participate on this list anyway because I see the value in communicating with the others on this list -- not because I expect SRI (another organizations without a significant open source track record to my knowledge) to swoop in and implement something open source after they get $5M. I've seen large organizations that can consume $5M without even blinking (sometimes just by putting the first footstep in the wrong direction and refusing to acknowledge that until miles later). I've even done some of that myself. :-) Obviously, Augment is definitely in the right direction (my comments on weaknesses not withstanding) and so it's likely good will come of the money, if only an up-to-date agument that runs under Windows or Linux. I wish SRI well, but somehow their effort feels to me very disconnected from this forum -- perhaps because I haven't noticed SRI people as key continuing on-line discussion participants. (Yes they do post, the issue is how much and with what effect.) To an extent, the same thing happened earlier with the local UnRev meetings at Stanford vs. using this list.

The biggest thing that would change the open source situation regarding UnRevII in my opinion of this is the release of significant preexisting code and content by the main participants under one or more specific open source licenses. That action might also attract more developers.

Still, one should not underestimate the value of UnRevII as a good forum, even if no open source software is produced directly from it. UnRevII has succeeded at bringing together people interested in using KM to help solve some global problems, and it has served to collect pointers to much of the state of the art in that field. UnRevII has succeeded at that, even as some open source developers have come by, scratched their heads, and gone on to other projects. In that sense, Bootstrap and Stanford have played a key facilatory role, and I thank them for that.

As I have said before, perhaps UnRevII would have been even more successful if the original mission had been explicitely a more conservative one not involving any direct software development or related IP assignment, as "a forum to discuss global issues KM could help with, discover the state of the art in KM, and hopefully foster or inspire independent efforts to make open source / nonproprietary KM solutions building on the Augment legacy." If the charter was that, I for one would get more out of the list and use it slightly differently (in the sense of discussing more technology details without fear of potential "permission to use" liability, and propose and comment on short term small collaborations of a few people to code open source things on specific issues, again without fear from "permission to use").

As for me, I'm starting another contract tomorrow (Java & XML ) so I'll be a little less vocal on the list for a while.


-Paul Fernhout
Kurtz-Fernhout Software

Developers of custom software and educational simulations
Creators of the GPL Garden with Insight(TM) garden simulator