This is a first cut at a Compendium FAQ. The questions and answers were gathered from the compendiuminstitute yahoogroup from 2003-2006. Your suggestions for improvement and contributions of further questions and answers most welcome!

Al Selvin
January 2006

Tags:requirements and design  

Tags:requirements and design  

As promised, we've posted a web export of the original functional
specification for Compendium. It was maintained from the software's
early development in 1998 or so until late 2000.

We'll probably make some better container for this eventually, and/or
include more of the maps that are in the database we exported it
from*. But for now, if you're interested in taking a look, it's at:

http://compendium.open.ac.uk/ReqSpec.htm

A community project that we've oft talked about would be to bring
this spec up to date. There are hundreds of newer features that are
not included in it, and some aspects of the design have changed since
2000 but are not reflected in this document. Still, though, it can
give people a good understanding of what we were trying to achieve in
the early development, and much of what needs to be preserved as the
tool expands. Astute readers will also find some requirements that
have still not been completely addressed. If anyone is interested in
helping to update this, please let us know at kmi-compendium@open.ac.uk.

Tags:requirements and design  

Tags:requirements and design  

Tags:navigation  

As a user I often make an effort to structure my database / maps so that I
impose my own logic of up and down. For example, I will create a designated
Home Page / map that helps me navigate back to a core schema, like a table
of contents. One reason I do this is that I use Exports in Outline html a
lot for browsing by a large audience-- and there must be SOME form of
logical back/forward imposed or the casual viewer is quickly lost. Also I
think this means that as an artful practitioner, its good to learn to take
care of the structures you create, make them elegant, and become more aware
of when you are creating confusing recursions and divergences because of
transcluded nodes.

(CP)

Tags:argumentation  

> Michelle forwarded everyone a web chat question from T. Tihamer regarding
> how Compendium would handle issues in logic, specifically arguments of the
> form modus ponens and argumentation with objections and rebuttals. I would
> also like to see someone knowledgeable in Compendium address at length the
> issue of Compendium and logic. The Reasonable software of Tim van Gelder is
> one software designed specifically to handle the logic of
> argumentation--modus ponens and objections and rebuttals, all of which is
> essential in order to evaluate an argument rather than map issue areas
> primarily. Is Compendium intended to address these areas as well but in a
> more dialogical fashion than Reasonable software? Or is Compendium more
> oriented to defining issue areas and logging points made rather than
> evaluating the argumentation that arises through dialog? Any insight from
> strong Compendium users who know Reasonable or similar software would be
> greatly appreciated.
>

Tags:argumentation  

Compendium, as a tool and methodology, is not meant to enshrine or put
forward a preferred or paradigm mode of representing argumentation, or
even necessarily to center around argumentation per se. Rather,
inspired by but taking off from its forebears like QuestMap,
Compendium is meant to provide a wide-ranging toolkit that allows many
types of representations (including, but not limited to, multiple ways
of showing arguments) to be interwoven together.

Individual users may prefer particular methods and can constrain their
use of the tool to particular schemas, but a quick look at examples
like those on our Showcase (soon to be expanded at
http://compendium.open.ac.uk/community/showcase.htm) will reveal the
diversity of representation types that Compendium can support.

That being said, there are many argumentation experts on the list that
may have strong opinions about the best ways to use Compendium's
affordances for argumentation (in addition to others that don't use
Compendium for argumentation at all).

(AMS)

Tags:argumentation  

I can indicate a decision to go ahead with an idea by switching the
icon from the "light bulb" to the "gavel" but there is no icon that I
can use to show that an idea has been rejected. In fact, I've never
seen any map which shows rejected ideas in any way. It is possible
that a team has rejected a couple ideas out of the group without
making their final decision on what ideas to accept. How are those
rejected ideas shown on the map?

(TL)

Tags:nodes  argumentation  tags  

Usually what I do is indicate the chosen idea with a Decision or Note
node. You could also assign a Tag of "rejected" to the rejected ideas, or
change the node type of the rejected idea(s) to a Reference node and supply
your own icon to the node using the Image field.

(AMS)
============

Sometimes a decision is to reject something ... Like a decision to reject a rejection
node type. How would you represent that in a compendium IBIS-type map about
new features?

New Features? (Q) <---- Rejection Node Type (I) <---- Decision? (Q) <----
Rejected (I)

Of course you can also create a code "rejected" and add it as a tag.

(MXS)

Here's what we teach in the Dialog Mapping Workshop.

An important aspect of the argumentation process is marking Ideas that
the group doesn't want to actively pursue, but which for group memory
purposes need to be kept in the map. Al and Maarten suggested a
couple of ways to do that. In Dialog Mapping we call such Ideas
"retired", and indicate it by simply putting parens ()'s or brackets [
]'s around the label. This is a simple notation that suggests
"removed from active consideration for the time being", and shows up
in the map (as opposed to a Tag). Probably the ultimate would be to
have an Idea icon with a line though it, and to have Compendium switch
to this icon automatically whenever an Idea was tagged as "retired".

By the way, it also happens that Arguments get retired. Arguments
represent claims which either support or object to adoption of an
Idea. Sometimes people what to argue against the claim, to show that
it's not valid. This can be done two ways: simply, with Arguments (a
new Argument objecting to the first Argument) or more robustly with a
challenging question such as "Really?" followed by a case for why the
claim isn't true. With either notation, if the person who stated the
original Argument retracts it (e.g. "Oh, yeah, I guess you're right.")
then the Argument is either deleted or kept in the map but "retired"
by putting parens around the label. This indicates that it no longer
has any rhetorical force. (Again, an Icon/Tag linkage would make this
easier and clearer.)

(JC)

Tags:database  nodes  maps  

Tags:database  nodes  maps  

Compendium at present allows for multiple users on a given database,
and you can have a database on a server that anyone can log into, but
yes, we are currently missing features that highlight what's changed
since a given user logged in. So far, Compendium does not do this
simply because the vast majority of users are using it in single-user
mode (QuestMap, the 1990s product on which Compendium is based, did
support this in fact -- we just haven't had the developer time so far
to replicate it).

However, you can do a Search (Tools / Search, Ctrl-F, or just click the
Search toolbar button) and search for nodes created/modified
before/after a date. So anything modified after the meeting date would
be of interest. You can inspect the search results directly, or place
them into a map or list to work with them as normal.

(SBS)

Tags:nodes  maps  

If the keeper identifies items that should be discussed the next time the group convenes, how can these be noted on the map?

Tags:nodes  maps  transclusions  export  navigation  

One way would be for the keeper to copy and paste them to a new map
(remembering that the nodes 'know' that they're now in two views,
which can be navigated from the Views right-click menu)

Another way (if the nodes of concern are all viewable on screen
together) would be to shift-select the nodes so they go yellow, and
then screen grab that as a 'photo' of the nodes of importance to be
returned to (by hitting your Print Screen function key). This can then
be pasted into a document or email as part of the agenda.

Tags:import  graphics and icons  reference nodes  

Tags:nodes  import  export  graphics and icons  reference nodes  sharing  

In PowerPoint I was looking at my slides in the slide view. As an
experiment I dragged an image of one slide over to the compendium map
and Wah-LA it became a reference node with the image in view. I knew
about using this function in other contexts but this was an unexpected
treat.

(NP)

-------------------------

Something I recently stumbled into is all the hypertext functionality
of PPT. I am doing slide shows now in PPT that are ALL hypertext-- no
slides appear in linear sequence; all slides are reached through links
internal to the show. For example, EVERY slide contains a navigation
button back to the visual model, and one to a table of contents. PLUS
out of this mode I link to Compendium html exports, so I can do fancy
visual modeling with drill downs in PPT, in which some of the drill
downs take the viewer into Compendium exports. Then that whole PPT
show can be given to the viewer to browse privately at a desktop, and
since it now looks just like a web site, they can browse the visual
model without getting that usual frustration inherent in a long deck
of linearly sequenced PPT slides.

(CP)

------------------------------

I had a wonderful experience with the PowerPoint drag-n-drop
functionality just a few days ago, facilitating a meeting at a
consulting company.

The meeting consisted of a series of PowerPoint presentations with
discussion during and after the slides. (Is there any other kind?
:-) ) Here's what I did:

1. Drag-n-drop slides into map. In my experience, you have to be
viewing the slides in PowerPoint's "Slide Sorter View", and you need
to have arranged PowerPoint and Compendium so that you can drag from
PP to C in one motion. (If you have trouble reading the thumbnails of
the slides in "Slide Sorter View" you can zoom them larger.)

2. Once a slide is in the map, relabel it with the title of the slide.
(Obviously these first 2 steps are better done before the meeting
starts, although that's not what happened in last week's meeting.)

3. Here's the way-cool part: In Compendium, turn on "Image Rollover"
(under the Map menu heading). Now you don't need PowerPoint any more
-- just by moving the cursor over the slide's node you get a nearly
full size image of the slide that everyone can read.

4. Capture discussion of the slide in the map. Simply create a
Question like "Issues/comments from this slide?" and link it to the
slide's node. You can smoothly move between viewing the slide and
capturing the discussion about it ... all in Compendium. And since
the map contains both the slides and the discussion, it is a coherent
record of the whole meeting.

Finally, note that the slide images created by the drag-n-drop
operation are saved in Compendium's "Linked Files" folder. (The new
vastly-improved "Export HTML Views" function coming in the next
release handles these embedded images well -- when you export a map
with embedded PowerPoint slides to the web the slide nodes are hot
links to separate pages that show the slide image.)

(JC)

-----------------------------

Tinkering with this PPT to Compendium method that Jeff describes (post
of Apr 7, Jeff Conklin as below): If I go into ppt 'file' menu, 'page
setup', choose custom size, and size to 20" by 15" (preserving the
original slide ratio), then when I drag the thumbnail into C, and do a
rollover of the image-- then it seems to enlarge nicely to the limit
of the C. frame. It might thus be best to create PPT with that page
size. One ppt I re-sized seemed fine; another thus had some text
sizing problems.

(CP)

----------------------------

OK, you're right ... it's not really "nearly full size". The rollover
image of the slide is just under 1/2 the height and 1/2 the width of
the screen, so, yeah, about 20% of the area of the screen. Perhaps I
was just lucky that these slides used a large font and no animation --
the group had no trouble reading the rollover image of the slide.

(JC)

----------------------------

Tags:sharing  

I know of three ways to do this:

1. Have two screens and two projectors. This way content (PP) get's
in own display, and process (C) gets its own shared display as well.
The only downside I've found to this arrangement is that it's possible
(though rare) for the two screens to be in a tug-of-war for the
group's attention. This is my favorite -- I always try to have two
screens and projectors ... AND the room tables set up in a way that
everyone can see both with ease (a flattened U-shape works pretty
well).

2. If there's only one screen and projector, you have to manage the
sharing of this critical resource. I've done it 3 ways:

a. One computer. I have both PP and C on my computer, and I Alt-Tab
between them. This gives me lots of control, but has a big downside:
I can't be capturing discussion or organizing the map while the group
is looking at the slide.

b. Two computers plus a video switch. This can work really well.
The C driver can be capturing brief comments while the group is
looking at a PP slide (or spreadsheet or Visio diagram, etc.). Then
if the group goes into serious discussion mode you press the button on
the switch that puts the C computer up on the screen and map the
discussion.

c. One computer configured to use 2 "side-by-side" monitors. I've
only done this once and it worked fairly well. I think there's a way
to do it with Macintosh machines as well, but on a Wintel machine, at
least with Windows XP, you can set up windows for two monitors:
number "1" is your notebook or computer screen, number "2" is the
display projector. Then you can put PP on monitor 2, C on monitor 1,
and you can type in C while the group is looking at PP. To put C on
the big screen, you drag it's window from monitor 1 to monitor 2, i.e.
from your notebook screen to the big screen. It's a very strange
sensation the first time you do it! I highly recommend (from personal
experience!) experimenting with this technique before doing it in a
client session!

Anyway, all this points to the magical dance of the group's attention
and ownership between each other, the facilitator, and the shared
display or displays. Has anyone else had other experience with this
dance?

(JC)

Tags:nodes  import  text  

Tags:nodes  import  text  

Here's the basic procedure for dragging and dropping Excel content at
present:

In an Excel spreadsheet, highlight (select) the area you want to bring into
Compendium. Notice the black border around the selection area.

'Grab' (click on) that border with your mouse pointer. Without releasing
the button, drag that selection onto a Compendium map, then release the
button.

'(If you have unchecked "Always drop text as Text without choices" in
Tools...Options...DnD Options) you should see a Drag and Drop Selection
dialog box. Choose Process Drop as Excel text then click Process Drop.

At present, there is a default mapping of row and column headers and data
cells. Compendium assumes the cell at the upper left of your selection is
the root node. All cells at the left column of your selection become the
next level of Question nodes. All cells at the top row of your selection
become the next level of Question nodes specializng the previous level.
Finally, the data cells become the Answers to those Questions.

So

A1 A2 A3
B1 B2 B3
C1 C2 C3

will become (approximately):

A2? -- B2
/
B1? ---- A3? -- B3
/
A1?
\
C1? -- A2? -- C2
\
A3? -- C3

If you play around with your spreadsheet you can get pretty consistent
results.

(AMS)

Tags:maps  graphics and icons  

Tags:maps  graphics and icons  

You can draw boxes and other graphics on maps by choosing
Tools...Activate Scribble Pad. This opens a toolbar that lets you
draw circles, boxes, and other shapes on maps. You can choose line
weights and colors for the shapes as well. The feature is in somewhat
of a primitive state at present, but it does work. For more info on
this, see the "Scribble Pad" item in Compendium's on-line Help
(under "Other Actions and Concepts").

(AMS)

Tags:maps  graphics and icons  

You can provide graphical backgrounds to maps that can provide
a lot of expressive possibilities. You do this by double-clicking on
any map background (the white space) and selecting an image for the
Background Image field. This image can be any .gif, .jpg, .png, etc. People
are starting to do some very exciting things with this. You can see
some examples (more to come!) at the Showcase
(http://compendium.open.ac.uk/community/showcase.htm).

(AMS)

The Compendium Showcase contains 'best practice' examples of many Compendium techniques.

Tags:maps  graphics and icons  

The ability to add background graphics to maps opens up lots of
possibilities. One is to add branding to your maps with the use of
watermarks. A watermark is a faint image of a logo or other graphic that
appears in the background.

I've created an (admittedly crude) example of one using the Compendium
Institute website logo (thanks to Raju Dave for making the watermark
version of the logo). If you don't see it in this email, you can go to the
Files area (http://groups.yahoo.com/group/compendiuminstitute/files) and
click on "watermark-map.jpg".

To do this, have an image file (.gif, .jpg, etc.) with your desired
watermark ready *. Create a map node, open it, then double-click on the
map's white space to open its Contents window. Click on the ./. button to
the right of the Background Image field. Navigate to the image file you
want to use, then click Open. Click OK to close the Contents window, and
voila! your watermark is in place.

(AMS)

Tags:clones  templates  

> Is there a way to perform a deep clone in Compendium. I'm trying to create
> templates with some nested maps, but the template idea with nodes is
> much less useful to unless I can clone a node deep, i.e. clone map,
> its contents and if it has maps, also its contents, etc.
>

Tags:clones  import  export  templates  

Doing an Export to XML with the right options selected (full depth),
then reimporting that export file, should accomplish what you're
after.

(AMS)

Tags:import  XML  

Tags:import  requirements and design  XML  

If we want to have a discussion about this, I can start.

1) First, one needs to determine how you want to map the mindmanager
entities to Compendium nodes and links.
2) What features do you want to extract (symbols, text blocks, free
floating text, relationships - arrows, etc.)?
3) Mindmanager has a macro language that can be used to traverse the
tree and extract the nodes and features.
There are a few add-ons already written for Mindmanager that might do
this or could be adapted to extract information to XML. Mindmanager
symbols can be placed on Nodes to help identify the meaning and use of
nodes.
4) The XSLT to transform the extracted info into a form that Compendium
will eat should not be too bad since MindManager is generally
hierarchical except for relationship(arrows).

If Praj has some specifications about how how the Mindmanager features
are used in the maps to be translated, that would be a good place to start.
a) Nodes - can the nodes be categorized and identified in Mindmanager
(questions, answers, notes, arguments pro and con, references,)
b) Text on Nodes?
c) Symbols on Nodes
d) Graphics?
e) Free floating text
f) arrow links
g) hyperlinks
www
other maps
files
h) colours on text and on lines
i) Boundaries

(RW)

You can choose any color at all for the link arrows. Go to Tools...Manage Link Groups. This lets you not only choose colors for the default link group, but create your own link types with any colors and labels you wish.

For example, to change the color of the basic default link type ("Related To" in the pre-defined Issue-Based Information System link group, go to Tools...Manage Link Groups, highlight the Issue-Based Information System item, then click Edit. Highlight the Related To item, then click Edit. Click the Choose button next to the (black) Link Colour. This brings up a Colour Chooser dialog that lets you choose colors in one of several ways, to any color on the spectrum.

For more info on this, see the "Link Groups" item in Compendium's on-line Help (under "Nodes").

(AMS)

Tags:database  

Tags:database  Java  import  export  XML  

Data sharing is part of the reason for using a database. Compendium allows sharing of its data on many levels (via SQL queries to the database, via a Java API, via the Jabber IM protocol, via XML import and export, etc.).

But understanding why we use a database rather than a file goes beyond this, to part of the intended purpose of the software. We call Compendium "The Power Tool for Knowledge Mapping." This is because we developed the tool and its underlying methods over years of experience working with groups with other tools. We wanted
to allow the maximum amount of power and flexibility to the tool's users to not only create graphical maps of ideas, but to enable deep re-use of material, wide-ranging customization and extension capabilities, ease of programmatic manipulation of the contents of a database, and many other features. Part of our motivation in working
on the tool has been to permit non-technical people to do things by manipulating graphical objects and their attributes that would otherwise require programmers to accomplish. You can take Compendium very far in that direction, if that is your desire and intent.

There are a number of our research papers and other writings that
explore these ideas. Explore the Library at
http://compendium.open.ac.uk/library/library.htm

(AMS)

Tags:database  

Tags:database  installation  

Using MS-Access/ODBC caused the same kind of problems we have seen with MySQL. In those days, many people tripped up over congfiguring the ODBC Control Panel in Windows, which had to be done for each database, not just at install time. We got just as many emails in support about that as we now do with "Administration database unable to be created" etc. with MySQL. And, of course, it pretty much limited us to one OS. It also didn't seem to work for sharing Compendium databases over networks, which we (with limitations) essentially got for "free" when we migrated to MySQL.

(AMS)

Tags:database  MySQL  

Tags:database  installation  MySQL  

For information, should others need to repeat this exercise, the exact
process followed was:

1/ Install MySQL 4.1 using default location & with Data file location
set to C:\MySQL (for ease of backing-up)
2/ Install MySQL Administrator using default options
3/ Install MySQL Query Browser & test database
4/ Install Compendium
5/ Edit file MySQL.properties for database connection - note URL
requires the localhost IP address (127.0.0.1) & not 'localhost', which
is OK for MySQLAdministrator tool access
6/ Start Compendium

In retrospect, my original installation failed at point (5). This
(finger trouble, admittedly) would not have occurred if the installer
included either a database connector configuration wizard (to edit
MySQL.properties) or simply displayed the readme.txt file. Clues
regarding the need to edit the MySQL.properties file are well hidden!!

(KC)
=====

You should have localhost defined in your host files in Windows.

c:\windows\system32\drivers\etc\lmhosts and
c:\windows\system32\drivers\etc\hosts
as
127.0.0.1 localhost

This will simplify lots of software installations and sooner or later
you will get caught again with this one so you should define it now so
that it is always there.

This definition is the same for Linux and Unix and Windows. I assume
that Mac OSX has the same thing as well.

(RW)

Tags:database  

Tags:database  

MySQL just needs to run when Compendium runs. Doing this requires removing winmysqladmin.exe from Startup, and manually launching it before launching Compendium. Possibly C:\mysql\bin\winmysqladmin.exe could be added to the compendium.bat file that launches Compendium.

I do the manual launch on one of my systems where I run Compendium less frequently.

(AMS)

Tags:database  MySQL  

Tags:database  MySQL  

In the section [mysqld] of my.ini, add

# Block external connections
bind-address=127.0.0.1

This restricts connections to applications running on the same host.

Restart MySQL to activate the change.

(RW)

Tags:database  security  MySQL  

Tags:database  security  MySQL  

It looks like you have not set up your Compendium to log into MySQL
correctly.

Are you sure that your compendium startup file has the same username
"root" and the root password that you used to set up MySQL?
You may not have set the password on the root account on MySQL.
Set it to something and then make the changes to the Compendium
configuration file and you will move on to the next problem or to
Nirvana if you are lucky.

It is very important to set the root password on MySQL since it is a
service that is listening on the network for connections and anyone can
get access to you database if ou leave the default password on MySQL. As
root they can do anything thta they want to your data. Not smart.
There are steps that you can take later to prevent this but right now
you need to get a new password and tell Compendium about it. Once it
works security can be added

(RW)

Tags:database  security  MySQL  

[If anyone has a similar problem in the future, here's how I solved
it. The problem was that mySQL only set up the account root@localhost,
whereas Compendium was trying to access root@localhost.localdomain. So
the problem was solved by the mySQL command:

GRANT ALL PRIVILEGES ON *.* to root@localhost.localdomain;

(JW)

--- RW wrote:
> It probably needs a username and URL otherwise how will it know
where to
> find the database (could be on another server) and how would it know
> what username to log in as.
>
> Try
> URL=localhost
> username=root
> password=
>
>
> JW wrote:
>
> >Yes. I'm running it off the root MySQL account. [Just in case you
> >meant the root Linux account, I tried that too, but without any
success]
> >
> >My MySQL.properties file is as supplied:
> >
> >#MySQL Connection Details
> >
> >#Mon Nov 03 12:14:19 GMT 2003
> >
> >url=
> >
> >password=
> >
> >username=
> >
> >
> >Given that my mySQL is setup without a password for root, as far as I
> >can see this *ought* to work.
> >
> >James
> >
> >--- In compendiuminstitute@yahoogroups.com, RW wrote:
> >
> >
> >>"I've tried running it as root, but it gives the same error message."
> >>
> >>You mean using mySQL's root account in MySQL.properties?
> >>
> >>
> >>Ron
> >>
> >>
> >>
> >>JW wrote:
> >>
> >>
> >>
> >>>I've tried running it as root, but it gives the same error message.
> >>>
> >>>(I also removed the password from my MySQL root account, and tried
> >>>with Compendium's default settings, but this still gave the error
> >>>message.)
> >>>
> >>>Any other suggestions?
> >>>
> >>>James
> >>>
> >>>--- In compendiuminstitute@yahoogroups.com, RW wrote:
> >>>
> >>>
> >>>
> >>>
> >>>>You might want to try running compendium as root. Not as secure as
> >>>>
> >>>>
> >what
> >
> >
> >>>>you are trying to do but if you do not have any other Mysql
> >>>>
> >>>>
> >>>>
> >>>>
> >>>pplications,
> >>>
> >>>
> >>>
> >>>
> >>>>it makes no difference since your "compendium" account will look
like
> >>>>root anyway.

Tags:database  security  MySQL  

Just to clarify the MySQL 'root' user situation.

MySQL has TWO root user accounts that are there by default (neither of
which have passwords) when you install MySQL.

1. 'root' user account for external host access (from outside your
computer over the net).
2. 'root' user account for localhost access (from inside your
computer only).

The first of these two accounts is the one which could potentially cause
a security issue.
When you install Compendium, it checks to see if you have already set a
password for the 'root' external host user account. If not, Compendium
sets one for you (which is an auto-generated non-dictionary password).

I have done this because I wanted to protect inexperienced MySQL users
from a security breach from the net.
You can still initally access MySQL with the 'root' localhost account
and no password, so you can always set your own passwords for either
root account at a later date.

Compendium, by default accesses MySQL with the 'root' localhost account
and no password.
If you give your 'root' localhost account a password, you must them tell
Compendium what that is via the MySQLAdministration.bat application
found in the Compendium home directory.

This application can also be used to give Compendium the details to
connect to a MySQL application running on another machine.

(MB)

Tags:database  

Tags:database  

These option work only within one running instance of Compendium.

You need to select the nodes in one database,
chose to 'Copy to another database',
then close that database,
open the other database
and select 'Paste from another '. which should then be available.
(MB)

---------------------------------

Another way of saying this is, you can't do this with two separate
instances of Compendium running. It can only be done within the same
Compendium session. If you want to have two separate Compendiums
running, you'd need to use Export as XML from one then Import from XML
in the other.
(AMS)

Tags:database  

> I use Compendium on my desktop machine at work and would like to keep
> it sync'ed with my laptop so I can work on projects at home. I know
> I can export/import and I suspect that is the only way to do what I
> want. Is there any other "easy" way?
>
> BTW, will the Derby based version have an easier way to sync? If not
> planned, can I add a feature request? :-)
>
(HT)

Tags:database  Derby  XML  

Besides export/import, there is Backup/Restore. That is what I use to
keep databases on different machines in sync, because that way you can
be sure you are getting everything (with export to XML you may choose
options that don't get everything).

The Derby version will also have backup and restore. We don't have any
current plans to do more with sync, at present (but the developer
community is welcome to explore!).

(AMS)

Tags:database  security  

Tags:database  security  

> When you create a project (database), you add your user details in the

> project creation dialog. By default, that user is an administrator,
> (unless, for some reason, you deselect the box to make that user an
> administrator?).
>
> Therefore, all you should have to do is use your normal login to edit
> the project name, as you should be an administrator.
>
---------------------
> Hi Compendium,
>
> Just a suggestion resulting from my new install of Compendium with the

> Derby database. I wanted to change my database name, but didn't know
> the admin username and password, which is required before being able
> to do this.
>
> I wasn't able to find it anywhere on your website. (Eventually I found

> it by googling.) I suggest that the install readme or the website FAQ
> mention what the default administrator's name and password are.
>


(MB)

Tags:database  

Q: Should I be worried about putting all of my stuff in one database?
Will that cause performance problems eventually? How do you decide
when to start a new database?

Tags:database  import  export  XML  

Databases can get pretty large without problems. One of my
database backup files is now almost 3mb and I've seen no performance
problems. The only performance slowdown I ever see these days is when
first opening a map with lots of large images, it can be a little
slow (though once open performance is fine). For me, a database
should contain a particular domain, however you define that. The main
reason to keep things separate is to avoid worrying about finding
unrelated stuff as the result of searches or the like. I guess a rule
of thumb would be that a database should contain the stuff that you
want to consider as a whole. If that is multiple projects, fine. If
you want to do something particular that should be kept unmixed with
other material, then that database can be narrower in scope. It's
also quite easy now to move chunks of data around in between
databases using the XML export and import capabilities; these chunks
can be as small as one node or as large as hundreds of nested maps.
You can (optionally) preserve the node IDs and transclusions when
doing this also.

(AMS)

Tags:database  

Q: Can I backup to a thumb drive just using the backup command? I'm
putting so much on it that I need to be sure it will be available
even if my machine crashes.

Tags:database  export  XML  

You should be able to, as long as the drive is mapped. Just select
it in the "Enter the file name to backup to" dialog. As long as you
can navigate to it, it should work. BTW I backup frequently, usually
on the fly just to my local drive and at the end of a session I copy
the backup to another machine located elsewhere. (I do this with any
important data, not just Compendium databases). I also sometimes do
quick XML exports of a map I'm in the middle of working on, if I
don't want to take the risk of losing even a few minutes worth of
work (right click on the map's white space, choose Export... Export
as XML.

(AMS)

Tags:nodes  labels  

Tags:nodes  labels  

This capability is on the enhancements list for future development. We intend that every node label can have its own formatting, but it hasn't been implemented yet (except in the data model).

(AMS)

Tags:nodes  transclusions  

Tags:nodes  transclusions  

You can easily do this. There are several ways to make identical but separate copies of existing nodes.

Note: Transclusions (the same node in different views) are created by copying a node in one view and pasting it into another (or by Searching for an existing node and inserting it into a view from the Search results, or by using the Auto Label Searching feature on the Map menu). They let you keep track of the same concept in multiple
contexts, and can play a variety of other powerful roles (in fact, the ease with which you can manipulate them is one of the unique features of Compendium). However, there is no need or requirement to use them when not desired.

(AMS)

Tags:clones  nodes  transclusions  

The simplest way to create a unique, non-transclusive copy of a node is to right-click on the node (or a selected group of nodes, using Shift-right click ) and choose Clone. This creates a node (s) with the same label, detail, properties, etc. as the original,
but it is completely separate and unique (it has a new node ID).

(AMS)

Tags:clones  nodes  transclusions  shortcuts  

You can also make Shortcut nodes from any node (same procedure as for Clone above but choose Shortcut instead of Clone). This is useful when you want to show multiple instances of the same idea on a single map (especially when the map is larger than one screen-ful), but indicate that the "child" instances point to an original.

Shortcuts are somewhat like Windows shortcuts; they point back to the original
but don't show changes if the original is changed.

(AMS)

Tags:nodes  

Tags:database  nodes  

A node that is in the database but not in any view is called a "stranded" node, and it's not that uncommon. People create nodes, realize their weren't the right type, and then delete them rather than changing their type, for example.

The way to find such nodes is to do a Search for "Deleted Nodes": on the right-most tab in the Search dialog box, select the "Context" of "All Views and Deleted Objects".
The stranded (and unstranded) nodes will be visible in the search window, and from there can be inserted into the current (map or list) view.

This is a powerful if somewhat kludgey way to recover from accidentally "deleting" nodes. If you can describe them somewhat uniquely, as in by Label or Creation Date, you can "undelete" them without having to wade through all the nodes in the database.

If you want to really remove objects from the database, you can purge them from the Trash Bin window.

In this sense all nodes (and links) in Compendium are transclusions from the database -- there is no sense of a "primary map" where a node lives. It's just visible in one or more views ... until it's stranded.

(JC)

=================================================

If you delete a map from the home view, and it has child maps/lists that appears in no other view that is traceable back to the home View, then they should go to the recycle bin.

Otherwise, if for some reason, there are dead views not in the recycle bin then all views are listed in the 'Find a Map/List' dialog window, ( on the Map Menu ). This should list all views whether they appear in the home view or another view, or nowhere. It just lists 'active' views, (not marked for deletion and therefore in the recycle bin).

(MB)

Tags:database  nodes  navigation  tags  

In the case where nodes are in views that aren't reachable via navigating:

What we mean by "stranded" is that an object is not reachable by any level of browsing from the Home Window. Stranded objects could be reachable from other stranded objects, but they're still cut off from normal user navigation and thus not viewable. (Note that you don't even need two maps ... you could Copy and Paste Map1 into itself!)

If I delete Map 1 from my Home Window is it considered 'stranded' even though it is transcluded in Map 2?

If the two maps are transcluded in each other, but neither is reachable from the Home Window, then, Yes, they are both stranded.

Map 1 is still in one view, although that view is not directly or indirectly accessible from the home window. The only way to find Map 1 is to search for it by name or tag.

You can always "unstrand" an object by searching with the "All Views and Deleted Objects" Context.

Compendium seems to be good about maintaining the Trash Bin as containing all and only stranded nodes. From there you could restore them to their original view or the current view, or purge them.

"The issue we have is that a bunch of nodes will have the same tags, but some are to remain in the database since they are valid and reachable while others are not and figuring out which ones to delete is tremendously labour intensive (given that during my clean up I ran into 2400 nodes that were not directly reachable and which I had to delete manually).

Can't you just Purge the Trash Bin to remove them from the database?

Note that with Compendium the word "delete" is pretty ambiguous. Technically it means to remove a node or nodes from the current view. Sometimes you want to strand a node ... then you have to go to every view containing that node and delete it. Finally, if you really want it gone from the database, you have to strand it and then go to the Trash Bin and Purge it. (QuestMap had a Purge command that allowed the user to remove unwanted nodes from the database directly, by right clicking over them ... but they could not be transcluded.)

(JC) (RvH)

Tags:nodes  

Tags:nodes  transclusions  

It may not be clear to everyone what the idea of a "parent" is, and how it relates to Compendium.

There really isn't the idea of one node being another node's parent in Compendium. Nodes themselves don't have parents or children -- that is, one node doesn't come from another node. For that matter, there really isn't an inherent "up" or "down" in Compendium in the same sense as, say, Windows Explorer or any other hierarchical system. This is something that makes Compendium different from most concept-mapping or mind-mapping applications. Even the Home Window is not really the "top" of a database, because maps and lists don't have to be navigable to from there (although it is a good practice to do so).

In Compendium, a node can appear on any number of views (maps or lists). This is called "transclusion," a hypertext term that means the same object is in more than one view. If you make a change to a node in one Compendium view, it changes everywhere else it appears.

Because of this capability, a node that is transcluded doesn't have a single "up" -- rather, we provide the Views capability that shows all the maps or lists that contain that node. In some sense all of them are "up". In the case of a Map or List node, they can even contain themselves -- i.e. you can copy a Map node, open the map, and paste it in. What would "up" be then?

Traditionally (actually taking off from QuestMap, Compendium's "parent" in many ways), we have made transclusions easy to create -- by Copying a node in one view and Pasting it into another, a transclusion is created. Perhaps it is too easy! We have found that for many people this is confusing, because it is not what happens in most other applications (in most applications copy/paste creates a brand new object that has nothing to do with the object it was copied from).

We are currently thinking about possibilities of changing the ways copy, paste, and transclusion work in Compendium to address these potential confusions.

(AMS)

Tags:nodes  navigation  stencils  

Is there a stencil-type menu of "frequently used map nodes" that I could drag and drop (not copy and paste) into maps?

Tags:nodes  navigation  stencils  

"I wish I had a stencil-type menu of "frequently used map nodes" --

You can use the Favorites menu for that purpose. Highlight a frequently used node and select Favorites > Add to Favorites.

To add a node from that menu to any map or list, just select it from the menu and it will be transcluded into the current view.

(AMS) (MB)

Tags:tags  

I think that they are one of the big advantages of Compendium (I can keep action items without really thinking about it), but I have the sense that I am under-using them.

Q: Is there a way to visually represent when nodes are 'tagged'? (like the * for description)?

(DT)

The "T" in the upper left corner of the node means that node has at least one tag. Make sure you have "Tags Indicator" checked under the View menu.

(AMS)

Tags:nodes  tags  

Tags are very powerful, and even more so when used in combination (e.g. "actionitem" "urgent" "management-issue" "jenny"). They let you create true multi-attribute data elements out of your nodes and views, and really help with searching and reporting. The Tags interface also lets you create groups of tags for different purposes.

(AMS)

I am wondering if anyone could describe the general process for
creating new skins for use in Compendium.

Tags:nodes  graphics and icons  

In the Compendium application folder there is a subfolder called 'Skins'. Inside this folder there are further subfolders, one for each skin (icon set) available in Compendium. Each skin type folder contains a set of gif images used for the node
type images (only).

To create a new skin:
1. Create a subfolder in the 'Skins' folder with the name of your skin (this is the name displayed on the Compendium menu).

2. Create a set of gifs for your new skin.
-The names and number of the gifs must match exactly the names and number of the the gifs in the other skin folders.

Compendium expects those names.
-Your new skin gifs need to match the size of the exisiting gifs.
Currently there are two sizes:-
standard = 32 pixels square
small = 16 pixels square

Once you have done this, when you open Compendium your new skin should be available on the 'Format/Icon Sets' menu.

(SBS)

Tags:nodes  tags  graphics and icons  

For example, if I use a node (and tag it) to represent a happy thought, I'd like to give it
a smiley face icon.

(DT)

Tags:nodes  shortcuts  graphics and icons  reference nodes  

There are at least two ways you can do this. One is by substituting your own icons for any of the node types in the Images folder for the skin (icon set) you're using. Notice that there are "shortcut" node icons -- you could easily substitute for any or all of those. So for example, make an Answer node, right-click to create a Shortcut of it, that could have the other icon.

The other way (which I prefer, and which other users have picked up on their own) is simply to have a folder of icons you like up on your desktop (for example, smiley faces, sad faces, etc.). Whenever you want to create, say, a "smiley" node, just drag the smiley icon off the folder and drop it on your map. That immediately creates a Reference node with the smiley as its icon (see the Image field inside the node).

(AMS)

Tags:nodes  maps  navigation  

Q: I find myself trying to keep to one screen, and reverting to
creating sub-maps a lot to make this happen. Is this ok form? Should
I try to get used to using maps that more than fill the screen?

Tags:nodes  maps  navigation  

This is partly a matter of style, personal preference, and context/intention (i.e. what the map(s) are for). In some contexts my maps can get very large, for example when I want to contain a whole analysis in one place. For more unstructured or conversational maps, it is probably a good idea to keep things down to a screenful or two, because otherwise things get hard to find.

In general, I'd say the more structure/formality you use, the larger maps can be without getting lost. Otherwise, using sub-maps is indeed good form. Going back later and "refactoring" larger maps into smaller maps is also a good idea. It's all about keeping things to manageable chunk- and grain-size, as well as deciding how you want to manage your information for later reporting, searching, linking, and publishing.

(AMS)

Tags:nodes  maps  lists  

I find them convenient for keeping the size of a map down, but I am not sure what I am giving up.

Tags:nodes  maps  lists  

Lists are essentially sortable maps (really "views") without links (and with less visual options, such as zooming). So they're useful when you want a container of items that you might want to sort and that you don't want the 'confusion' of links.

I use them primarily as "catalogs," collections of like items (often items that share
tags, like action items).

There is more about lists and catalogs in the training materials at http://compendium.open.ac.uk/training/training.htm.

(AMS)

Tags:nodes  maps  transclusions  

In 'proper' Compendium practice, you would have not created another map node with the same name inside a map, but rather copied and pasted the map node itself into the map. In other words, Map A can contain the map node "Map A" within itself even though they are actually the same object. This is what is meant by "transclusion" -- the same hypertext object in multiple containers (including, if desired, itself). This very powerful but sometimes tricky to understand concept is at the core of what makes Compendium different than other concept mapping approaches.


Al,
Actually, I wasn't clear about transclusions as I had not really used the feature. So, I must confess that I didn't fully understand what Michelle's comments implied on this subject. Now, I understand it somewhat better from the explanations in your mail. If I can think of some real-life mapping of concepts using transclusions, I think I will be able to appreciate better this feature. Are there some examples of usage of transclusions that I can look at to get an understanding?

As an user, I thought that creating a new map node with a name 'YFilter' is the same as copying and pasting an already existing map node with the same name, especially since I am creating the second one thru a double click on the first node. (Though I did see that the double-click brought up only an empty canvas for me and not a container with the same map node already embedded in it.).
I also have some doubts about the way I am using the product to produce argument maps. Let me explain.
If you look at my file, (in the map XTrie) I start out with an argument node at the top for which I have a Pro node and a con node. Often, the pro or con node in itself becomes a new argument for which I have to show further pros and cons. So, how do I specify that a pro node for an argument is also a new argument node as well? My diagram shows such an argument as just a pro node with some con nodes underneath it. Is this the way one is supposed to represent argument resolution?
Thanks for your help
Raghu

-----Original Message-----
Raghu,

Just wondering if the point MB makes about transclusions is clear to you. In 'proper' Compendium practice, you would have not created another map node with the same name inside a map, but rather copied and pasted the map node itself into the map. In other words, Map A can contain the map node "Map A" within itself even though they are actually the same object. This is what is meant by "transclusion" -- the same hypertext object in multiple containers (including, if desired, itself). This very powerful but sometimes tricky to understand concept is at the core of what makes
Compendium different than other concept mapping approaches.

I'd be happy to discuss further if the above is not clear or helpful.

(AMS)

Tags:Java  

Can compendium run as an applet?
If so, are there any supporting links or documents?
Can you point me to a place in the code?

Tags:Java  

Currently this isn't supported, but it did work this way at one time and it would not be difficult to refactor the application to run as an applet. Someone should try it :-)

(AMS)

Tags:export  

Help ... When I export to html outline, and I have links to Word docs
(reference nodes) thus embedded in the export, how can I get the Word
docs to open as a separate Word doc, in the Word environment, rather
than in the browser window?
(CP)

Tags:export  

In Windows, Open Windows Explorer, go to Tools->Folder Options menu.
Click on File Types Tab. Select "doc", from the registered file types list.

Click on Advanced button, uncheck "Browse in same window". Click OK.

(CB)

Tags:maps  reference nodes  

Tags:maps  reference nodes  

Select the Hide Icons item on the Map menu. There are other ways to avoid having icons display, such as using the Stencil with invisible icons that Simon Buckingham Shum provided at http://groups.yahoo.com/group/compendiuminstitute/message/166

(AMS)

Tags:stencils  

Tags:nodes  text  labels  

Tags:nodes  text  labels  

One of the quickest ways to convert a text list into nodes is to paste the text into the Detail of a node (i.e., copy the text from its source document, open a node in Compendium, and paste the text into the Detail field (the larger text field)). As long as the individual items are separated by blank linkes, clicking the "Convert Page text into Nodes" button in the window's toolbar (the colorful one that look like an arrow pointing to a map node, to the left of the Cut button) will convert the pasted text into nodes. You can even give the nodes types.

For more info, see the "Node Contents and Properties" item in Compendium's on-line Help (Help... Help Contents... Nodes... Node Contents and Properties, and scroll down to the "Text to Node Conversion Rules" section).

(AMS)

Tags:nodes  text  labels  

Given a block of text, with multiple sentences, each an item in a
list: What is some quick techniques for converting those into a
series of nodes in Compendium? Drag and Drop seems to work. Do I
have that right, "process drop as Excel text" brings the list into
Compendium as a series of items?

The Excel convertor works as shown in the screenshot at XXXXX. Drag the
top left corner of a selected region and it will convert row and
column headers into an Issue template...

Tags:text  

Tags:nodes  text  graphics and icons  stencils  

Concerning a "text box" node, you can create something close to what you
want by creating a empty bitmap icon which you attach to the node thus
creating a node with text but no visible icon above it. Using can use the
Stencil feature to allow you to create the node by selecting and dragging it
into the map. Of course, since it's invisible, you have to know where it is
in the stencil window. I've created a icon which is a thin horizontal line
that I use sometimes when I want the user to focus on the text.

Of course, since this approach uses nodes and nodes can't be manipulated
without removing the Scribble Pad, you would have to create the diagram
using the Scribble Pad, deactivate the scribble pad to place the text node
and then reactivate the Scribble Pad and hope you put the text in the
correct relative position to the diagram otherwise you need to deactivate,
drag the text, and reactivate again.

(SBS)

Tags:sharing  

I'm a newbie - sorry if this is covered in the docs somewhere, but I couldn't find it. I've got MySQL running on my computer, set up for TCP/IP connections. I'm trying to set it up so multiple users can work on the same project. We can all log in to the same database, but we can't see each other's work. We're not trying to do this simultaneously.

Tags:navigation  sharing  

Currently in Compendium each user has their own home view within a database project. So whatever nodes you put on your home view only appear on yours and no one else's.

If you want to share maps, then each user needs to add the relevant map/s to their home view, or a map they have in their home view. You can use the 'Find a Map/List' option on the Map menu to locate any map in the database, then add it to your home view or whatever map you currently have open.

(MB)

Tags:sharing  

Tags:templates  

Tags:templates  

There are many ways to do this. Templates in Compendium can be as 'deep' as you want -- i.e. not just some nodes and links on one map, but whole structures of maps within maps (if desired).

You can make templates by highlighting the node(s) and links you want (or a map containing them, or whatever), then choosing File.. Export.. XML file. Export them to the convenient "Templates" directory in your Compendium program directory. Import them the same way (File... Import .. XML file, or right click on a map white space and choose Import).

Or, you can do the above, and have the templates sitting as xml files in a directory, and just drag and drop those files onto a map. Or, you can have 'pallette' maps hanging around, and Clone the items from those maps (creates non-transcluded copies). Or, you can set up the type of thing you want in Excel or Word and drag and drop those structures in to use as templates. Also see the material on Stencils and the other drag and drop options in the Help, and the New Features section of the Help.
(AMS)

Tags:export  XML  templates  

You can export any structure in Compendium, of arbitrary depth, as an XML file and use it as a template. In other words, you can set up a map containing a set of nodes and links, including maps and lists of other nodes and links, including further maps etc., with all their metadata (tags, transclusions, node and link types), and export that whole structure as a single XML file. It can then be imported as a template as many times as you want. There are choices to make in how you do this, but you are on the right track to deeper/broader use of Compendium than just as a concept-mapping tool.

Once you have saved the structure as an XML file, you can (as you say) drag and drop that file in, or use File...Import...Import as XML.

Re Clones, they are a big help when you want to make what in most software would be "copies" of existing nodes, without having to recreate the labels, details, tags, etc. of those nodes by hand. They can be quite useful, but I really wouldn't use them for templates - -the XML mechanism is much better.

(AMS)

Tags:import  XML  templates  

Remember that you can drag+drop a Compendium XML file straight into the app, so if you make yourself a shortcut to open the folder where you keep your templates, then this is currently about the fastest way to have a 'Template Library'. Your template can of course be arbitrarily deep and complex, and can use custom Stencils (and yes, it
would be nice if the right Stencil(s) automatically opened whenever you brought in a template using them!)

On import, you can then choose your options of whether you bring nodes in as brand new nodes, or transclude them to any existing instances, or even overwrite existing nodes if there is a difference.

We started initial work on a real Template Library Manager user interface which could track uses of a template etc (ie. make Templates first class objects like nodes) but had to abandon due to a student leaving.

(SBS)

Tags:import  export  XML  tags  templates  

Templates in Compendium can serve many purposes and are a large subject, but here are the basics.

Compendium templates are predefined sets of ideas, maps, or sets of maps/lists) which can help organize a database. Templates can provide structure and repeatability. They can, for example, seed the modelling of a recognised, recurring situation, e.g. focusing attention on key issues, options or tradeoffs that are considered important.

Templates can be as simple as a single repeating question, or as complex as whole structures of linked nodes and maps. Use them with as much depth as needed for your tasks.

Examples of the use of templates include:

a) walking through an established procedure to ensure that all the right questions have been addressed

b) driving discussion in a very top down manner in order to elicit requirements that must then be fed to other tools

A template is distinguished not only by the particular nodes it contains, but by their layout which may have been carefully designed to support the analysis, and the metadata 'Tags' that have been assigned to the nodes to support their subsequent harvesting.

Templates can currently be reused in 2 ways:

1. Copy and paste (or Clone) direct from an existing map - in which case the nodes are actually 'transcluded', a hypertextual dimension of connection which complements connections shown graphically in maps:

i) transcluded nodes point to the same database object so any edit to one 'ripples through' the db to update all other instances immediately;

ii) transcluded nodes display the number, and on mouse rollover the names, of the views they are in (= the contexts in which the same idea has arisen) allowing you to jump direct between them and see at a glance how often they've been used.

2. Export as an XML file which you then import on demand. You then have the option to transclude the imported nodes to existing instances, setting up the kinds of 'deep connections' summarised above, or import them as brand new nodes which can be edited independently with no transclusion. This way lets you give your templates an easily remembered name, such as "TaskModel.xml" or "Project Proposal.xml".

Another advantage of constraining maps in this way means they can be exported to other tools for subsequent processing, because the content and structure is known. On the flip-side, you can break out of a template at any point if it turns out not to fit the situation and capture informal rationale and other discussion that is important ("this question is the wrong one for this problem - what we really need to know is..."; "we broke this guideline only as a temporary measure to make the deadline - this needs fixing for v2...").

For more info on Compendium templates, see the tutorial and hands-on exercsies at
http://compendium.open.ac.uk/training/training.htm and the video at http://compendium.open.ac.uk/training/videos/clips/modelling-templates.html
(requires Flash plug-in)

(SBS/AMS)

Tags:web  export  

> Has anyone had any success finding/generating URLs for the maps in the
> HTML-Views export?
>
> I am trying to post a set of HTML-Views export maps on a website, and then
> link to the submaps. It appears that all of the maps in such an export
> appear within a single HTML frame, so there's no URL for individual submaps
> (except perhaps for a scary javascript call that doesn't seem work).
>

(JC)

Tags:web  Java  export  

Personally I don't like using the frames. As long as you don't point people to the 'top' map, you don't get the frames and the below is not a problem.

Look in the directory that you exported to. All of the individual maps you exported will appear there, with the name of the map followed by some numbers then .html. The numbers are the unique identifier (since you could have more than one map with the same name in your export). Find the real top map from your export and give people that URL.

(AMS)