Hi Nathan, please see response in line...

Begin forwarded message:

From: "Nathan Bauer" <[EMAIL PROTECTED]>
Date: March 13, 2006 1:41:18 PM PST
To: "Mimi Yin" <[EMAIL PROTECTED]>
Subject: Re: chandler usage inquiry

Hi Mimi,

Thanks for getting back to me.  I would be happy if you forward any or
all of this discussion to the design list.  And thank you for further
describing Chandler's data structure.  It was very helpful to me.  The
possibilities of (a) setting up multiple hierarchical arragements and
(b) assigning some semantics to labels sounds perfect for how I'm
hoping to use Chandler.  Although some future citation system would be
great, what I really need right now is a way to organize the tons of
research notes I have on hand.  It's driving me crazy that I'm
currently unable to capture all of the data relationships that I know
are there.  It sounds like I'll soon be able to use Chandler to do
this.

I do have one further question.  You wrote:
What we won't have in the short-term are sophisticated ways to
navigate and large data sets (aka data visualizations of relationships).

I'm having some trouble parsing this sentence.  Did you mean that
there won't soon be _any_ methods of visualizing and navigating
through data structure, or did you mean that there won't soon be any
_sophisticated_ ways of doing so.  I am hoping you meant the latter. 
Even a fairly rudimentary way of seeing and moving among collections
via their semantic relationships would be very useful. 

In the short-term, the sidebar probably won't get much more complex than it is today (which is a flat list). We have explored designs for: 
1) various ways to "walk" the data structures like a tree: http://wiki.osafoundation.org/bin/view/Projects/BrowserDesign
2) UI for managing and defining attributes

However, we're unlikely to get to those features any time soon. For now, think of the sidebar more as a bookmarks bar, a place where you can plop your most frequently used collections, rather than a top-down view of how your data is organized.


In the current
version of Chandler, all I've managed to produce is a simple,
unstructured list of collections in the sidebar (kind of like the
labels in Picasa).  I assume that the sidebar will eventually have
some sort of mechanism for visualizing (and navigating among) these
collections based on their structural relationships.  From my
perspective, which is perhaps atypical, having something along these
lines would be highly desirable for 1.0.  It would immediately make
Chandler better than any of the notetaking / outlining software that
I'm currently aware of (like Keynote, for instance).  It would also
have some more calendar-centric value, I assume.

Anyway, I apologize if you've gone over all of this before.  Having
followed the design list for a while, I'm still failing to get what
sort of data structure navigation and visualization methods are
planned for 1.0.  It would be great if you could briefly describe
these.  Or, if it's already documented, perhaps you could just point
me in the right direction.

Thanks again.

Best,
Nathan


On 3/13/06, Mimi Yin <[EMAIL PROTECTED]> wrote:
Hi Nathan,

Thank you for your detailed email. I think you bring up some good
points, would you mind if I forward this to the Design list?

The short answer to your question is that this particular usage
scenario is not on the list of target use cases in Chandler 1.0.

The long answer is that what you're describing certainly falls within
the larger vision for Chandler and while we're not optimizing for it
in the short-term (ie. shipping out of the box with semantics
relevant to citations and research-related content), 1.0 will have
laid important ground-work for eventually supporting this type of use
case more explicitly in the future.

Including:
1. The ability to label items with your own attributes and eventually
to define new Kinds of content (ie. Citation Kind) with custom schema.
2. A taxonomy of content Kinds that takes into account the kind of
content you describe.

What we won't have in the short-term are sophisticated ways to
navigate and large data sets (aka data visualizations of relationships).

As for the question of hierarchy. We're not opposed to the use of
hierarchical displays of data. Instead, we're attempting to walk a
fine line between data structures and data displays.

So, there will certainly be a way for you to view and interact with
information in a hierarchical display. However, the data itself will
not be "stuck" in the hierarchy you view it  in. And, you will be
able to view the same data in a number of different hierarchies.

Let's take your example:
You could either arrange your research in the following hierarchy:
Philosopher: Plato > Area of philosophy: Metaphysics > Query: What is
Beauty? > Sub-query: What is the difference between Beauty and Goodness?

OR

You could rearrange to look at it like this:
Area of philosophy: Metaphysics > Query: What is Beauty? > Sub-query:
What is the difference between Beauty and Goodness? > Philosopher:
Plato > Area of philosophy: Metaphysics > Query: What is Beauty? >
Sub-query: What is the difference between Beauty and Goodness?

So there are a fewsignificant differences between Chandler's data
structure and traditional hierarchies:

1. The ability to assign semantics to "topical groupings". A "topic"
is not just <Plato> but, <Philosopher: Plato>.

2. The ability for "topics" to be arranged and re-arranged in an
unlimited number of organizations. e.g. The <Sub-query: What is the
different between Beauty and Goodness?> will show up as a child of
both <Query: What is Beauty?> and <Query: What is Goodness?>

3. The ability for a single topic to intersect multiple topics, yet
still maintain it's integrity as a single topic. e.g. The <Sub-query:
What is the different between Beauty and Goodness?> is presented as a
"child" of <Query: What is Beauty?> and <Query: What is Goodness?>,
but it is a topic in its own right.

I hope that answers at least some of your questions.

Mimi



On Mar 8, 2006, at 8:29 PM, Nathan Bauer wrote:

Dear Mimi,

I have been an interested spectator on the development of Chandler for
quite some time now.  I'm very excited about both the potential of the
application itself and the open process by which it is being
developed.  While I do look forward to all the calendar-specific
functionality on the immediate horizon, this is not my primary
interest in Chandler.  I'm an academic (in philosophy), and I'm very
intrigued by Chandler's potential to organize my research in new and
more complex ways.  I've compiled some examples of the sort of thing I
have in mind.  From my perspective, I'm interested in knowing (a) if
this is the sort of thing Chandler might eventually manage; and (b)
when I can expect this to be functional.  I also thought that, from
your perspective, it might be useful to have in mind some use cases
that are a bit different from those I've been reading about.

I was tempted to just submit my examples directly to the design list,
but I am worried that this will just be a distraction from the
calendar-centric focus of the next few releases.  As a compromise, I
thought I would send it to you to see what you think of it.  If you
think it useful, do with it what you will.  If not, feel free to
ignore it.

The examples follow.

Research-centric use cases for Chandler

My main hope for Chandler is to use it to manage my notes and
quotations.  Over the years, I've compiled a large collection of
philosophical quotations that I use in my research.  I also have a
large number of my own notes, ranging from brief comments on
quotations to longer free-standing passages.  All of these are
currently distributed in various files, making it difficult to find
the one I want (or even to remember which ones I have).  I would like
to use Chandler to organize these notes and quotations in various ways
via labels (or collections or whatever the appropriate term is).  Here
is some of what I have in mind.

(1) Manage quotation references
For quotations, it would be useful to label or tag them with
bibliographic information (author, book title, page reference, etc.).
Using labels for this information would allow me to avoid repeatedly
typing this information, and would also allow me to organize my
quotations by source (for instance, if I wanted to see all of my
quotations from a particular chapter of a book.  Eventually, I could
imagine someone developing a parcel that would turn Chandler into a
fairly full-featured bibliographic manager, but I could get by with
something much simpler.  I'm picturing a hierarchical arrangement of
labels (Author > Book > Chapter > etc.)

(2) Organize by topic
I would like to use labels to organize my notes and quotations in
multiple ways.  For instance, a single note or quotation might be
linked to various topics and, thus, indirectly to various other notes
and quotations.

(3) See these organizational relationships
(a) When I have a particular note open, I would like to be able to see
all of the topics that it belongs to
(b) When I'm looking at a topic (label), I'd like to see all of the
notes and quotations that fall under it.
(c) I'd like to be able to organize the topics (labels) themselves,
according to various, potentially overlapping hierarchies and to be
able to see these topic relationships.  One example might be something
like:
Philosopher > Area of philosophy > General issue > More specific issue

I understand that there is some resistance to this sort of
hierarchical method of classifying data.  I'm all for auto-generated
collections, where these are applicable, but I do think there remains
a place for more hierarchical organizations as well.  These
organizational relationships are, after all, themselves an important
kind of information.

I hope this gives you an idea of some of the ways in which I hope to
use Chandler.  None of this strikes me as being particularly
complicated, but I confess that I've only played around a little with
the current release and I don't yet understand Chandler's collections
model well enough to see if it will do the sorts of things I have in
mind.

Best,
Nathan

--




--

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Open Source Applications Foundation "Design" mailing list
http://lists.osafoundation.org/mailman/listinfo/design

Reply via email to