Mr. Rod Welch
rodwelch@pacbell.net
The Welch Company
440 Davis Court #1602
San Francisco, CA 94111 2496
..
Subject:
Agenda Meeting Mike Poremba
Dear Rod,
I will check the conversion op on 500 line lists and let you know.
..
Re Mike's agenda, I took a crack at converting the outline to get it into
SDS. It is indeed not a trivial undertaking. This is one reason that I want
to add programs or features to help with just this sort of problem. See my
record at
Conversation on status of SDS project Rod, I'd like to hear how things are
progressing regarding SDS. As you know, I'm interested in what you've come
up with, and would love to see it advanced beyond your own personal use.
..
glj -- Question: What are the *root* causes of KM failure? My take is that
the problem is that the culture failed to change, regardless of the degree
of suitability of the technology. This is an issue that we face with SDS as
well - no matter how good a technology may be, it is only as useful as
people will use it. We must be *very* careful in our analysis of KM
failures. If our understanding of why KM has failed is incorrect, we may
well address the wrong problems. It is easy and tempting to believe that KM
has failed simply because it is not SDS, but is that an accurate assessment?
More discussion later.
Brief report on the status of my own web application for collaborative
writing To catch you up on what I've accomplished in the past half-year,
and some key lessons that I've learned along the way.
glj -- My experience to date indicates that there is substantial value in
many of the ideas and most of the facilities within SDS, but that it is
hampered by a highly idiosyncratic interface. This is why I have tried,
elsewhere, to separate what is essential about SDS and what is dependent on
its implementation.
Re-implement SDS as a Windows or Java application Targeted towards
individual end-users.
*Not* Java. While there is some advantage to multiple platform support, Java
is still too slow in every implementation I have seen to support a good
editor. If I can find a good editor to use as a base, then perhaps, but I
expect that even the free ones will be burdened by licensing issues such as
GPL that could cause problems.
..
We have discussed incremental changes by adding programs written in other
languages.
..
A parallel, and hopefully first, step is trying to get a version of Medit
that gets beyond the 640k memory problem and can fully support SDS. I don't
have an assessment as to how achievable this is.
.. http://www.welchco.com/sd/08/00101/02/02/12/12/151344.HTM#1696
Integrate key functionality into more sophisticated tools For example,
Knownspace Helium
..
glj -- as you note, we need further analysis of other approaches to
understand the extent to which SDS features can be supported. There is an
issue that adding SDS features to other products may improve those products,
but it might not benefit SDS all that much.
..
glj -- If the focus is on services rather than product, and that can be
profitable, then using other products as a base for SDS is not as much of an
issue.
Work plan What are your next steps?
There are so many ideas in SDS that are worthy of re-implementing
elsewhere Ease-of-use will be important.
Ease of use may not be the issue Mike has in mind
..
glj -- We have been here before. Ease of use is not of paramount importance,
but neither is it of negligible concern. Given that we have an analysis of
the features that are essential to SDS, then we can investigate how to make
those features as easy to use as possible. Developing a technology that
people are likely to be willing to adopt is a trade between what the program
does or allows, and how easy it is to do it. If a necessary function is
missing, the program is weak. If a function is far harder to use than is
warranted by the benefit that it confers, then the function might as well
not exist.
.. http://www.welchco.com/sd/08/00101/02/02/12/12/151344.HTM#7M4Y
Porting SDS to another environment is, essentially, a rewrite There's
no such thing as an application "port" that involves changing
programming environments. These are called application "rewrites".
A rewrite should progress in small steps Modern software development
usually involves continuous incremental improvement through a series of
many low-risk goals, rather than large ambitious projects.
..
glj -- As you note, start with getting beyond the 640k problem, then look at
modularization of functions to support incremental rewriting. Your point
that experience with SDS helps with what (and how) to rewrite, is well
taken.
.. http://www.welchco.com/sd/08/00101/02/02/12/12/151344.HTM#7M7R
No one person can do it themselves You're talking about a multi-person
project.
Recouping on investment is challenging Covering the costs of creating
such software applications is challenging. Profiting is even more
difficult.
..
glj -- You note that "Launching disruptive technology is often slow and
challenging for developers," which is true. That is partly why consulting
and services help. It is possible for an expert to make use of a tool that
is not ready for mass adoption, as you have demonstrated. Clearly immediate
(or even near term) profit is not our main motivation or we would be doing
something else.
.. http://www.welchco.com/sd/08/00101/02/02/12/12/151344.HTM#7M92
Few people have such a practical vision of information management
"Exploiting knowledge management requires having a solution available when
demand rises to the level of intolerance for bumbling, loss, conflict, crisis
and calamity under the status quo using information management." Here is where
it is important to be "at the right place at the right time." Organizations can
make changes (sometimes) when the alternatives are sufficiently dire. Being in
a position to help them make changes which are in fact improvements is the
goal. Even those who have failed at KM have made big bucks from those who were
convinced that the proposed solution would correct the unacceptable problems
that they faced. The major difficulty comes down to the fact that, sooner or
later, somebody has to think, and most people and organizations will resist
that until the bitter end and organizational failure.
..
Chris Dent points out at
When people react poorly to the concept of co-evolution, I think what they are
sensing is that co-evolution makes serious demands on the people involved. It
says, "If you want success in your endeavor (whatever it might be), you might
need to adapt."
.. http://www.welchco.com/sd/08/00101/02/02/12/12/151344.HTM#TG92
Review Mike's progress in developing a web-based collaborative writing
tool
Since we last met, I've made good progress on my web site for
collaborative writing.
.. Accomplishments to date
Web-based outline editor Works in IE/Windows
glj -- You ask "How will Mike's Outline feature fare in relation to 90% of
Microsoft features under Windows and IE that users evidently ignore." This
is an excellent question. While I like outlines, and there is merit to an
online presence, I am not convinced that editing online is the most
productive approach. Experience with blogs and Wiki may shed some light on
this. I am convinced for now that what is needed is a local application to
allow features, speed, and operation independent of the internet coupled
with synchronization mechanisms for local and online content. Companies may
have problems with content online as well, even with authentication
mechanisms.
.. http://www.welchco.com/sd/08/00101/02/02/12/12/151344.HTM#PW9M
Backend database for storing outlines Relational database for storing
managed content.
Secure spaces for collaborating on outlines Backend support for secure
shared workspaces where users can collaborate on writings.
..
glj -- I consider this essential to any collaboration tool. What is missing
is the ability to establish outlines that are shared only with a specified
group. Currently outlines are either private or public, and there needs to
be a group mechanism. SDS need to have this as well. You asked "How is
outlined content integrated with other content?" The short answer is that
there is no other content at the moment. Outlines can have links associated
with them which allows parts of an outline to be in more than one place, and
also allows reference to external material. Mike's system supports outlines
as its primary content at the moment, and there is scant provision (yet) for
any other sort of content.
.. http://www.welchco.com/sd/08/00101/02/02/12/12/151344.HTM#PW54
Meet with Eric Armstrong for his perspective Eric has offered to serve
as a sounding board for ideas on how I should direct my focus to exploit
the technology I've developed.
Gain user base Goal is to get to 25 users who have logged in 25 times.
Product can then be called successful, and worthy of following through
on a business plan. Until then, it's all talk.
..
glj -- You ask "Are there examples of work product that improve
collaboration, as discussed in NWO?"
http://www.welchco.com/03/00050/01/09/03/02/03/0309.HTM#4077 ..
I doubt it at the moment. There are too few users for any collaboration to have
happened as yet, and group sharing and / or a searchable list of public
outlines are needed in order to support collaboration in any real sense at all.
Enhance existing editor to work in other browsers IE/Mac and Mozilla
support.
glj -- I like the idea of threaded discussions, particularly if there are
tools to support refactoring such a discussion.
.. That is all I have on the agenda.
..
*** Process of Responding to this Email ***
Let's talk for a moment about the best way for me to have gone about
responding to this email.
The agenda outline exists in at least the following forms:
..
Mike's original email
The web SDS record that you reference in your letter. This is what I
used, since it has your comments and was the record you referenced.
..
I used yours to generate a new record of my own to which I could add comments.
That seemed more than a bit strange, but it might have been the best approach.
..
What are your thoughts on how I might best have done this?