Here's a contribution from John Rothermel from 1994. John conducted a
6-month IR&D project at then ESL titled "OHS Technology Evaluation
Project" with the Bootstrap Institute. This summarizes how he
differentiates OHS/Augment from other systems.
Date: Sat, 21 May 94 15:55:02 -0700
From: John G. Rothermel <jgr>
Inspired by our collective conversation yesterday, I've tried to gather
my thoughts on some of the differentiators that set Augment apart from
anything else that's out there either commercially or in academia.
There are a couple of levels at which to do this; a marketing level, and
a functionality level. I'm definitely more comfortable with the
latter(!), but was forced to give the former some thought a couple of
months ago as part of ESL's New Ventures process.
So, at the marketing level, I think the major differentiator is that
Augment is an "end-to-end solution"; it's involved with (at least can
be, or should be) all aspects of one's daily (softcopy information)
tasks. This is the opposite of "point solution". It's an environment
(not just an application or tool) that supports collaborative knowledge
work (where knowledge work is basically anything that could be done on a
computer, and so is very broad).
The analogy that one of our marketing people came up with (which seemed
to help him make sense of it all) was an "information factory". I think
we could really run with this; the concepts of ingestion (getting
appropriate raw material inventory), warehousing of info (in the
Journal), reuse, groups working on parts of a whole, roles, work-process
examples, etc. all seem to have useful analogies.
I usually describe Augment as a collaborative information management
environment which provides powerful capabilities for individuals
(integrating and managing personal information bases), teams,
organizations, and the whole enchilada (enterprise integration). So the
significant thing to me is that Augment's capabilities *scale*. I.e.
the underlying capabilities support any size group / information-space,
without discontinuities (exceptions to be fixed :-).
Since no software system with such broad aims is going to hit the nail
on the head for every customer, it needs to be extensible, both for
tuning to a particular team's / organization's peculiar tasks, and in
support of process evolution for the long haul. I think user-level
extensibility is significant (vs having to rely on programming staff for
changes / experimentation).
Another thing that I often mention as a differentiator is that Augment
came top-down whole-cloth to address the needs of collaborative
knowledge work. And this shows up with how well all the capabilities in
the environment play together when going end-to-end across a series of
related tasks. All the tools play together, and all of the basic
linking, structuring, and addressing functionality is available
uniformly. All of which makes for a system of high conceptual integrity
(a holy grail for me).
All of this can sound like marketing fluff, so I usually give examples
like the Conference subsystem - which allows the other tools to operate
in a normal fashion (by the Shower) while in a screen sharing session,
i.e. any tool's data is available for normal manipulation with its
native tool while the screen sharing tool is in use by the group. (This
gets back to the interoperability of information among any of the 4 CSCW
modes (same/different time/place), another point I like to bring up.)
So much for marketing.
At the functionality level, I like to compare systems along the
following categories: document architecture, filing, linking, tagging,
collaboration support, etc. all of which have more detailed
functionality called out within them. (I can't find my references for
all of the categories right now, so I'll save that for another day.)
But some of the basic functionality differentiators for Augment are:
- a more sophisticated document-centered paradigm (moreso than OLE or
OpenDoc). The decoupling of (1) the tool, (2) the data, and (3) the
window allows a much cleaner partitioning of functionality, and is
potentially more powerful than anything that's out there. (E.g.
filtering and structure display is provided by the window, so not
every tool has to worry this.) In general, allowing any tool to work
on any *part* of a document to which its operations could apply (even
if created by another tool) is unique and optimal.
- The content analyzer's filtering-out / filtering-in capabilities
are also very powerful as is, and have incredible potential (but it
needs an easier to use filter-specifying tool, vs writing in the
pattern matching language?).
- The sequence generator is also incredibly powerful, unique, and
full of potential. Have available a sequence generator oriented
towards set operations, and a hypertext web oriented "trail
- Hideable names or labels (bookmarks) in documents are a great
feature, especially since you can specify them directly as link
destinations (and the links resolve even if the names are hidden).
- The window-based document visitation/return stack is unique. It's
more general and powerful than conventional hypertext link-return
stacks. (But it should be kept as an editable, reusable, linkable,
shareable file for more general "trail" or "tour" capabilities.)
- The frozen statement capability is unique. It should get
integrated with upgrades such as per-window command and status panes.
I.e. have a frozen-statement pane laying around near the top of the
window that's zero-height (hidden) to start out with.
- human readable links are unique, but their power doesn't show up
'till you talk end-to-end system usage scenarios (when working
computerless, e.g. reading hardcopy on a plane, you can call from an
AirPhone or from the airport to have some referenced item FAXed to
you; or working on a portable computer on a plane and being able to
see a human-readable link reference for a document that that you
already know about, have loaded on your portable, or have hardcopy
of). (Maybe much of Augment marketing-level differentiation should
be done with broad-scope mini scenarios - accentuating the end-to-end
strength vs just function-by-function comparisons?)
- INCLUDE links are unique and potentially incredibly powerful
(allowing "virtual documents"). (They need a little clean-up though,
so that when turned on, they take on the structure addresses of the
*referencing* document, and allow navigation to be contained within
the referencing document (i.e. so you don't all of a sudden pop up
into the referenced file, as happens now.)
- Being able to "file" more fine-grained chunks of information in
editable "locator files" is more powerful than traditional
"directory" or "folder" filing capabilities. Again, needs some user
support for doing traditional folder-related operations easily.
- Able to modify information that's not on the screen.
- Able to script anything you can do directly from the UI.
- The screen sharing tool has near optimal functionality. I don't
know if it's still unique though.
- Roles (the Act command) integrated with digital signatures and the
work-flow-management subsystem (Suspense - which in turn is
integrated with Reminder (and e-mail? and the Journal?)) This is
dynamite stuff, calling for more end-to-end mini scenarios. (And
some animated multimedia presentation, like with Macromedia Director!
- The tagging of each statement with the modifier's name and
timestamp is unique. If extended to a full blown user-editable and
searchable/filterable property list (attribute-value list), it'd be
- The Journal is nicely integrated with the rest of the system
(e-mail, archiving, etc.). There are more and more document library
management tools appearing out there, but not integrated with other
system functionality or work processes to the extent that the Journal
- The Reach subsystem is significant for interoperability and dealing
with legacy systems. I'm sure it's unique.
Actually, most of this stuff is discussed in more detail in
<ROTHERMEL, FINALREPORT, :x>.
This archive was generated by hypermail 2b29 : Thu Jan 06 2000 - 11:01:17 PST