3D mindmapping with Topicscape - ideas exchange
Argey
pointers to information that will help us build an even better
Topicscape. Please keep them coming. One user, Tony, a musician and
electrical engineer, has been particularly prolific in a series of
email exchanges over the last week or so.
We have been having an conversation with Tony about suggestions, new
features and proposals. This has been full of ideas and I asked if I
could put much of it on this Group so that we can have reactions from
other Topicscape users and active Beta testers. He said that if I
remove personal stuff, that would be OK.
Rather than one long post, I'll clean it up, break it up by topic and
will put them up here over the next few days.
I hope you'll comment.
Thanks
Argey
Argey
the active/current topic so that you can see them before clicking on
the tunnel gate with separate zooming in/out just for this circling
loose associations [LAs].
LAs topics might be shown in the landscape view behind the history list
and search hits views. In my opinion it's important also to see the LAs
of the active topic visually on the landscape anywhere.
Argey: You can double-left click on a tunnel, go inside and see the
loose associations (LAs) in a small panel. The names of those topics
can be treated like all other topics (double-right click to make
current, double left-click to enter, drag to the pending tray, drop on
them, etc. You can also right-click on a topic and select View tunnel.
If LAs were circling the current topic, you wouldn't know which visible
topic they applied to would you? Similarly if they were on the left.
I feel that LAs are an important requirement of 3D mindmapping - I know
they are useful from long practical use, day to day. But I'm not too
satisfied with showing them as tunnels. The analogy is reasonable, but
and am working on other manifestations of the LA and this will lead to
more thoughts, thanks. There are issues of readability if you have too
many things flying around, (and we shall be introducing flagged topics,
one choice of which will be a flag image in top of the cone) it would
make the 'Scape confusing if it had other things in the same place.
Tony: Flags are a good idea. I hope you're also planning of
implementing "topic types" as available in Brain as "thought types".
For that also flag colors might be used. So each user could define for
himself the meaning of the colors. Brain even supports "link types". If
you're considering to implement an importer for Inspiration, too, link
types would be interesting here also. Or in general anywhere where
concept mapping principles are used "link types" would be great.
Argey: We have user-defined attributes planned (they are in the
database schema already to avoid conversion later, just not implemented
in the interface yet). This should be a more general solution.
We have association types (topic-maps terminology) working in our
Student Edition beta and our 1.5beta already but these are not in
public release yet. It uses relationship phrases. So Italy and Rome
might be connected by two phrases "has as its capital city" and "is the
capital of": giving factual statements "Italy has as its capital city
Rome" and "Rome is the capital of Italy". This allows semantic sense to
be derived from a Topicscape.
Argey
Argey
(This should do exactly the same what a mouse left-double click is
doing right now.)
Argey: Yes, why not . . . will do.
Tony: It should be possible to change topics in the landscape view
also via keyboard keystrokes. Especially when combined together with
ENTER for seeing the topic center. Then you could walk around in your
landscape without touching the mouse.
Argey: I like to walk around mindmaps with the keyboard, so I take
your point. I feel that a 3D environment is not so great to navigate
with a keyboard, though. The fact that we use the cursor-movement keys
(arrows) for flying means that we would probably have to use some sort
of A-S-W-Z diamond. We'll give it some thought.
Argey
. sea with boats (several boat models according to size), not green
gras but blue sea with jumping/diving dolphins as jump/dive tunnels!
. galaxy with stars and planets and ships
. on floor but with other motives like chess figures or soldier types
(generals)
. city looks (houses as in New York, Germany, Sweden, Japan, China,
Africa... or real sahara look with real looking pyramids or camels on
sahara sand.)
. animals in amazonas (for children)
. lego stones on lego field (for children)
. music persons (like musicians in a symphony orchestra. or fans of a
rock band in front of the main band on the stages.)
. military place with tents (CWC like colors) and comouflage filtering.
"Onplace filtering": Only the hits can be seen on their original
places, the rest get a camouflage object on them so that you can't see
them.
. or probably best would be a combination of pyramids (largest
objects), tents (middle sized objects) , camels (smaller objects) and
humans (smallest objects). a) on sand b) on grass >>> there might be
different skins with different pyramid, tent, camel, human forms!
(changing forms/roles relatively to situations/environments is very
natural to human behaviour, too.) huge marketing potential :) by
customizing figures to different end user types. Even there might be
always two versions of each object: with/without a tunnel look. (With
tunnel should look somehow more interesting/complex/important...)
Argey: Topicscape is trying to extend 3D beyond games and into the
serious work-place of mindmapping for information and project
management. It's a question of markets - I don't know who would be a
user of the above type of interface and what they would use it for.
Also on performance. Rendering complex 3D objects at the resolution we
have to use to allow readable text in a 3D environment could slow
things down a lot. We can have hundreds or even thousands of objects
rendered in the 3D environment.
We are experimenting with new skins at present.
Tony: As you wrote . . . the performance would decrease then it's
better to keep the textures simple.
Argey: It's not so much the textures as rendering the objects in 3D,
but there are also issues about the clarity of the interface - too much
visual confusion makes it difficult to read the text and see the
Topicscape's structure clearly.
Argey
definition. If Topicscape would allow the same but with unlimited own
"custom search definitions" that would establish a great link from your
topic collection to the outside world, meaning the internet.
You can apply your own search uri's directly from your topics
collection in the topicscape in the right moment. This would mean you
would be always just two clicks away from your beloved search command
on your actual active topic. (1.click: e.g. F4 for opening the list pop
up window, selecting via up/down keystrokes, 2.click: ENTER).
Argey wrote: You can more or less do it now. Do your search in a
browser for the first time, drag the favicon to a topic and select
Create a shortcut to a web page. You will not need to open the
occurrence directly (though you can), just click the URL hyperlink
which is automatically placed in the Source field in the Details Panel
when you saved it.
Additionally, you can mark that as a Favorite occurrence to find it
quickly. Or you could put all the favorite searches in one topic and
make that a favorite. Not quite two clicks but I think "two clicks"
ignores "selecting via up/down keystrokes".
I do see this as a valid addition to a mindmapping / information
management tool, but as it is doable now, there are other features I'd
like to add first. We will probably add Web searches to the Favorites
menu later and provide a keyboard shortcut to open the list as you
suggest.
Argey
Argey
a command line batch process. e.g. if you have a text file as an
occurence and you would like to apply a specific code on this text
file, one shortcut to send this file immediately as a command line to a
python/perl program would be great.
Freely definable dos commands on the topic name/occurence file would be
very useful in my opinion for speedy working. Sure the old way (doing
by hand) is still possible which is much slower. As an example think of
a personalbrain xml file which you want to convert into a different
format,
- you simply press your shortcut (e.g. F5),
- choose from the pulldown menu your previously defined command and
- press ENTER.
The result would be your xml file in it's new form. You could even
write your program so that it'll be opened anywhere else immediately,
e.g. in the browser or in Topicscape itself (if that would be
possible).
Argey wrote: I can see the value of this, but I feel it takes it away
from our target market, which is people who are familiar with
mindmapping and like mindmapping as a way of organizing information.
It would move Topicscape towards being an operating environment rather
than just a place to manage ideas, reference material and the like. It
would also further complicate the interface.
We probably will add something like this eventually (and default it to
invisible so that a user has to switch it on) . However, there are
many improvements which I think are needed first.
Argey
Argey
details view [Topic Center] (without going to Landscape view) e.g. by
using shortcuts.
Argey wrote: Tony suggested F6, F7, F8 but these have other uses
already. We will look into this and allow the same keyboard shortcuts
in the 'Scape as well as the Topic Center.
Argey
Argey
v1.1 support would be great also. (The problem with the original PB
Exporter is that it can't mostly convert bigger brains correctly to
xml, actually it's just crashing in the middle of the conversion
process... so the dependency of PB Exporter is not a good situation.)
Argey wrote: After hearing about this from Tony, further investigation
showed that 1.1 exports are produced by a fan product, not an official
Brain one and it can easily produce invalid XML (just include HTML in a
note). Using PB Exporter, we have converted Brains of up to 5,294
nodes, and have tested all inclusions like link types, thought types,
thought graphics, wallpaper, notes that include HTML and images, and
the full 127-character ASCII extension set including German special
characters (the latter are present in Tony's Brains he tells us). We
have been unable to make PB Exporter fail. Not that we doubt Tony has
a problem, but that we do not yet know what that problem is.
We have suggested that Tony try to break down one of his Brains that
fails to export, to find out what factor makes it fail. We can well
understand that our customers want to get their Brain-type mindmaps
into 3D but we are reluctant to advise our customers to move away from
the official export product.
Argey
Argey
specification here:
<http://www.ms.lt/ms/projects/formatkinds/000811specification.html>
<http://www.ms.lt/importexport.html> The Mindset CSV format is also
closely related to the web application Translucid which also supports
this Mindset CSV format. In the Mindset CSV format you can describe
any part of a Brain which is not possible in the structured text files
so nicely.
Another idea would be also to allow importing from specific wiki
formats directly into Topicscape. My favourite wiki implementation is
Oddmuse: www.oddmuse.org
Argey wrote: Those are useful references, thank you. There are many
mindmapping applications whose output we'd like to import, and will in
time. (BTW we are not trying to describe any part of a Brain in our
structured files!)
We think that Dave Winer's OPML is a useful de facto standard in this
aea which is why we adopted that early (it's in v.1.1beta), though it's
an outlining structure more than for mindmapping.
Argey
Argey
Landscape view directly via a shortcut without entering the topic
center. E.g. Alt-Enter or via mouse by Ctrl-double right click.
Argey wrote: Our next Beta (working internally) provides various ways
of seeing what's in a topic without opening it, and then of opening
occurrences directly. It also has a function to open an occurrence
automatically if you do a search and there's only one hit.
Tony wrote: Ok, that sounds nice, let's wait for the next beta then.
Argey
window, as available in Brain. If you could add anything similar to
Topicscape (e.g. as standard first occurence of any topic) that would
be a bomb.
Argey wrote: You have the Details panel for this - the Description. At
present that's only 256 characters, but it will be 2K in Student
Edition and the next version of the Personal Edition (already working).
You can also add a fileless occurrence (right click inside Topic Center
anywhere except on an occurrence or topic and select "Creat a new
occurrence". You can use it's decription, source, author and authority
for notes, including adding working URLs.
Finally, you can grab a portion of any Word or Excel document, and
other types, use paste Special in the Topic Center and Topicscape will
create a wrapper for you automatically.
As a counter to the original point, I would re-phrase this and say
"What's missing in the Brain is the ability to store as many files as
you like under a thought." I oten save scanned magazine articles and
mul-page web articles. It always used to annoy me that I had to make a
separate thought for every page, when it was really all one continuum
that happened to be broken up when I acquired it.
Tony wrote: Can't the description optionally be replaced by the first
occurence which would be loaded in that region? Right-click jumping
into different topics is what I'm missing here.
Argey wrote: I haven't yet grasped the advantage of this. It seems to
duplicate material and function, and adds files. Right-click jumping
is independent of Notes, anyway. See next item on Wikis.
Argey
existing as a topic name, then you can just right click this word and
jump to this thought even if it's not any parent, child or loose
association of the current active topic. This is missing in Topicscape.
The advantage of that is if you would import texts from other wikis
where many WikiWords were used you can still profit from those words.
Argey wrote: This is a good idea, but I can see some practical
difficulties - multiple topics with the same name being the biggest.
I rarely have one-word topics, but I do quite often have topics with
the same names multiple times in a Topicscape. So do others, I know.
There is an example in our user stories at
http://www.topicscape.com/user-story/mindmaps-for-research-projects.html
This 3D mindmap has an area where some legislation was analyzed,
country by country (see the 4th picture down). Under each country's
topic there is a set of topics names where research results were placed
for that country. The same set of names appears under every country.
Topicscape could not resolve these ambiguities. In a Wiki, there is an
ongoing effort to avoid such repetitions, but a Wiki has a different
structure and a different purpose from a Topicscape or a mindmap.
I think this is do-able, and later we will look at providing a
Wikki-style [[word]] marking and offering the user a choice of the link
to go to, or we may make words so marked to be targets of a topic
search if clicked on.
Argey
new topicscape [here Tony is referring to importing folders or
importing a structured text file] with an already available topic,
these equal topic names should be "somehow" autoconnected. One idea
might be that the autoconnected topics are like planets circling around
the current topic in the air/on top. So even would know immediately
that this circling topic came originally from another TopicScape
source.
Argey wrote: Like the proposal for wiki-style links, I'm concerned that
matching names do not necessarily represent matching topics. I'm
generally against making topic connections automagically for the user.
In 1.5Beta we have a merge function. If you drop a folder named
Mindmapping on a topic called Mindmapping, Topicscape will ask if you
want to merge or treat it as normal (make a new associated topic). If
you decide to merge, it will make all files in the folder into
occurrences of the target topic, and make all subfolders into children
of that topic. But even in this case, the user makes the decision.
I will think about your proposal, but right now cannot see a safe or
friendly way of doing it.
Argey
Argey
different Topicscapes to exact topic positions! [snipped] At least it
should be possible to drag'n'drop a topic into your desktop which
should lead you to exactly this position if you activate it later.
[snipped] Then you could have 1000 topicscapes which are connected from
various topic-connection points which would lead us to a "TopicNet".
Argey wrote: I have snipped Tony's extensive description of how he
thinks it should work from the above, because this is already working
in-house in 1.5Beta, and exactly as he describes. You'll get it soon!
Argey
Argey
you should get a question: "Do you just want to add the LINK to the new
topicscape or do you want to important all it's contents?"
Only adding the link has the disadvantage that you can't get back to
your previous topicscape again.
Argey wrote: Importing all the contents could be drastic if done by
mistake and in my experience the need for this is rare. Mostly, people
would make the Topicscape into an occurrence that just opens the other
one.
I prefer the present (1.1Beta) approach of exporting the whole
Topicscape to folders, going to the other Topicscape and importing the
top level folder as a floating topic. When dropped it rebuilds the 3D
mindmap as it was before. It's more deliberate and would never be done
by accident. We shall document this process.
And on the "disadvantage" that you mention, it is really easy to get
back to the previous Topicscape again - just go to the File menu and
click on item 2 in the previous-Topicscapes list
Argey
Argey
much faster when adding stuff into Brain. One reason for this is that
you can already define by dragging'n'dropping to a specific area around
the actual topic what kind of topic it should get (parent > drag to
top, child > drag below, jump > drag to it's left area). In Topicscape
you've always to click additionally the right option what kind of new
topic you're adding. A similar logic as available in Brain would make
things faster.
Argey wrote: When I make new topics in real-life use of Topicscape, I
find that 9 times out of 10 it is a child, and that is the default, so
no selection is necessary, so I don't agree that "you've always to
click additionally the right option". Note that the Hint function,
where Topicscape will analyze your new topic's proposed name against
existing ones in that Topicscape (and not just make a dumb 'start of
phrase' comparison). This involves an extra press of the Enter key,
admittedly, but we believe it adds real value in avoiding unwanted
topic duplication.
I'm glad you wrote "had the feeling" because I think your not comparing
like with like:- In the Brain, when you drop an item to make a new
thought, you can only connect it to the central thought. In Topicscape
you can drop it on any topic that is visible to make a new topic or
structure of topics. So in the Brain, you often have to navigate
first, sometimes you cannot see the thought you want, so there is
additional preparatory work before you get to that moment of dropping.
We will consider how an improvement might be made, though.
Argey
Tom Lieber
> Tony wrote: It would be nice if you could send a specific occurence to
> a command line batch process. e.g. if you have a text file as an
> occurence and you would like to apply a specific code on this text
> file, one shortcut to send this file immediately as a command line to a
> python/perl program would be great.
> Argey wrote: I can see the value of this, but I feel it takes it away
> from our target market, which is people who are familiar with
> mindmapping and like mindmapping as a way of organizing information.
> It would move Topicscape towards being an operating environment rather
> than just a place to manage ideas, reference material and the like. It
> would also further complicate the interface.
I agree here, especially as I've yet to have this case come up during
my everyday activities in Windows. Even in Linux, I think the only
commands I regularly run on whole files are grep and latex.
Sincerely,
Tom Lieber
http://AllTom.com/
http://GadgetLife.org/
Tom Lieber
> Argey: It's not so much the textures as rendering the objects in 3D,
> but there are also issues about the clarity of the interface - too much
> visual confusion makes it difficult to read the text and see the
> Topicscape's structure clearly.
The ocean theme sounds cool, but I do think that it would slow things
down too much on systems that aren't mine. >:)
Tom Lieber
> present that's only 256 characters, but it will be 2K in Student
> Edition and the next version of the Personal Edition (already working).
With the 2K limit, this seems to fit the need for something between
the tiny note of before, and a "full-blown" occurrance. Having a
click-shortcut that opens external programs (opening the first
occurrance) does not seem like a good idea, given how often accidental
clicks take place.
Argey
Personal Edition and are looking at metalic ones, so that users will
have a choice.
We can try an ocean as long as it doesn't have 3D dolphins and the like
to render.
I had an unexpected comment from a professional cartoonist. He thought
the present theme looked military - the browns and greens. I see what
he meant once he explained that about camouflage colors and rows of
tents! I'd always thought of the cones and pyramids as stylized
mountains, with vegetation up to a tree line, then rock/soil, then on
the highest, snow. That was the original thinking.
Argey
Argey
click:
Left-double click should change the topic. Right double-click should
enter the topic center. In my case I'm using much more often the
right-double click (changing topics) than entering the topic centers.
And left-double click is easier than right-double click.
Argey wrote: I feel this is just a matter of taste. If we changed it,
there would be an awful lot of annoyed users, some of whom have been
using Topicscape for 3D mindmapping for over a year.
Maybe one day we'll have a fully customizable interface, but that would
make the help system harder to maintain and use because the Help might
say one thing while the interface did another after a user change.
Argey
Argey
keep the previous view setting. (e.g. if I viewed my topics from
exactly the top view, after double-right-clicking the new view should
be also in the same angle view, meaning also from the top!)
Argey wrote: It does - I think Backspace is what you were missing.
This is mentioned in many places, but it is rather oblique and needs to
be made clearer. (Where it's covered: the shortcut summary, the
Quickstart document and the first page of the Help - "Can't I skip all
this and just get started?" several other places in Help). We'll also
make sure this is in the new demo feature that we're building into
Topicscape. Thanks for letting us see that the documentation is not
clear anough.
Tony wrote: Another idea would be to be able to fly back your whole
history in a looped way as smoothly as possible without having to press
all the time Backspace, e.g. if you're learning in a specific topic and
would like to go through the topic which you looked at so far. Some
more detail thoughts might be added to this basic idea. So there could
be a REPLAY button which would playback everything optionally
· in reverse order
· in exact the same order as you did from the beginning
Argey wrote: I think leaving the user in full control by holding the
Backspace key down is preferrable to giving control to a robot. Being
able to see your path through a 3D mindmap is really necessary
sometimes.
I have often wanted to replay forward in the same order after
backspacing though, and just have not got down to specifying it.
Thenks to Tony for that suggestion. I added it to the action list when
I received this from him, and it's working already. It will appear in
our next Beta release. We used Shift+Backspace as the control.
Argey
Argey
as occurences which are also displayed (in a given resolution) in the
landscape view on the object! I would use this feature just for some
topics, not many and that would improve the quality of the topicscape
in my opinion. e.g. the picture occurence(s) might have an additional
checkbox on the right side where you can select "show this picture on
the object in landscape view".
As an example, in a company the pictures of the workers / managers /
responsible persons could be used here. Or pictures of products...
Argey wrote: Yes, we have recently been working on a preview mode, and
that is working in 1.5Beta. It can preview a limited range of file
types, including text, PowerPoint, common graphics ones, live web pages
and MHT files. That is in the form of a large thumbnail (256pl x
256pl). This can be minimized for performance. But it is in the Topic
Center, not the 'Scape, so it is not quite what you are asking for.
We are also thinking about other forms of graphic display in the
'Scape, but don't expect to see this soon. The overriding
consideration is clarity of the scene. We need to think of a way of
adding the images without making the text harder to find and read.
Otherwise the great advantage that Topicscape has over all other ways
of visualizing information might be reduced rather than enhanced.
Everything added to the 'Scape may hide something behind it. Done
well, it would certainly be an advantage, I agree, and we'll keep
thinking about the best approach..
Argey
Argey
views (star view).
1. x-axis: horizontal view (input/output)
2. y-axis: vertical view (hiearchy with parents and children as
available right now in topicscape)
3. z-axis: timeline view where you can order your topics according to a
timeline, you can always follow later what came before timely and what
will come later timely
See here: <http://www.one-world-is.com/rer/owis/dem/slides/img033.gif>
With this model you can model/visualize (I would say) anything in the
world.
Argey wrote: I've done a lot of abstract data modelling as well as
business process modelling, and I find Roebuck's ideas very
interesting, though for enterprises, I prefer the Zachman framework. I
feel that Roebuck is in a very different universe to Topicscape. I see
Topicscape as an information landscape for visualization, not a
modelling tool.
Topicscape's History List is there to introduce the time dimension to
3D mindmapping, and though it defaults to a limit of 70 history items,
a user can make this much larger. When we get to the Enterprise
edition, we shall introduce version control as well to cover the
history of individual occurrences.
Argey
steveK
> Tony: The tunneling topics might be circled on top like planets around
> the active/current topic so that you can see them before clicking on
> the tunnel gate with separate zooming in/out just for this circling
Here is an interesting article review that IMHO is relevant to TS. It
actually is realated to several topics in this thread. For example the
feature of assigning colors to cones....
How Did Our Ancestors' Minds Really Work?
Main Category: Neurology / Neuroscience News
Article Date: 10 Sep 2006 - 5:00am (PDT)
http://www.medicalnewstoday.com/medicalnews.php?newsid=51357
How did our evolutionary ancestors make sense of their world? What
strategies did they use, for example, to find food? Fossils do not
preserve thoughts, so we have so far been unable to glean any insights
into the cognitive structure of our ancestors. However, in a study
recently published in Current Biology (September 5, 2006), researchers
at the Max Planck Institute for Psycholinguistics and their colleagues
at the Max Planck Institute for Evolutionary Anthropology were able to
find answers to these questions using an alternative research method:
comparative psychological research. In this way, they discovered that
some of the strategies shaped by evolution are evidently masked very
early on by the cognitive development process unique to humans.
Being able to remember and relocate particular places where there is
food is an asset to any species. There are two basic strategies for
remembering the location of something: either remembering the features
of the item (it was a tree, a stone, etc.), or knowing the spatial
placement (left, right, middle, etc.). All animal species tested so far
- from goldfish, pigeons and rats though to humans - seem to employ
both strategies. However, if the type of recall task is designed so
that the two strategies are in opposition, then some species (e.g.
fish, rats and dogs) have a preference for locational strategies, while
others (e.g. toads, chickens and children) favor those which use
distinctive features.
Until now, no studies had systematically investigated these preferences
along the phylogenetic tree. Recently, however, Daniel Haun and his
colleagues have carried out the first research of its kind into the
cognitive preferences of a whole biological family, the hominids. They
compared the five species of great apes - orangutans, gorillas,
bonobos, chimpanzees and humans - to establish which cognitive
strategies they prefer in order to uncover hidden characteristics. The
researchers worked on the assumption that if all five species share
particular preferences, these are very probably a part of the
evolutionary legacy of our most recent common ancestors, who died out
some 15 million years ago.
At the Wolfgang Koehler Primate Research Center at the Leipzig Zoo, the
researchers hid coveted items using two different strategies (see
Fig.2): In the place condition, the item remained in the same place it
was hidden in previously, but under a different object (e.g. a stone);
in the feature condition the object remained the same, but the place
changed. It was established that all four great ape species and
one-year-old children actually use the location as a way of finding
something hidden, even if it is hidden under a completely different
object. This outcome suggests that this preference has been part of our
cognitive structure for 15 million years.
The researchers then investigated three-year-old children and
discovered a difference: Unlike younger children, they considered the
object under which the item was hidden to be the most reliable
indication of its whereabouts, even if the location had changed
completely. The scientists have sufficient evidence to conclude that
1-year-old children and great apes do not lack the capability to
develop a feature-based strategy, but simply prefer to use a
place-based strategy. Evidently, humans reassess these preferences as
their cognitive development continues.
"The unique human cognitive development seems to mask some of our
evolved strategies even before we reach the age of three," says Daniel
Haun. "In future experiments, we therefore want to find out which areas
of cognitive development in humans, for example language acquisition,
are responsible for this restructuring of cognitive preferences." The
new methodical approach and the results it yields pave the way for the
systematic study of the cognitive structures of our evolutionary
ancestors and thus ultimately to an improved understanding of the
origins of human thinking.
###
Original work:
Daniel B. M. Haun, Josep Call, Gabriele Janzen, and Stephen C. Levinson
Evolutionary Psychology of Spatial Representations in the Hominidae
Current Biology 16, 1-5, September 5, 2006
steveK
Argey wrote:
> Tony: Texture ideas / motifs.
or probably best would be a combination of pyramids (largest
> objects), tents (middle sized objects) , camels (smaller objects) and
> humans (smallest objects). a) on sand b) on grass >>> there might be
> different skins with different pyramid, tent, camel, human forms!
> (changing forms/roles relatively to situations/environments is very
> natural to human behaviour, too.)
This has occured to me too ... My though was an ancient Rome motif.
:o)
I would disagree with the above idea of using pyramids vs tents vs
camels, etc though. I -do- like the idea for the motif, I just think
that having these extra components would move the visual representation
further AWAY from the abstract and more toward a literal land scape.
as such, the compenents of the landscpe should functionally mirror
thier real-world counterparts. FOr example, in the real world a person
may live in a tent but not even have a camel. Better to utilize camels
in some other way. Maybe they could walk from one to another related
topic. THey would leave camel tracks in the sand and this would show
the relationship lines ... Well, maybe not <g> But it -would- mirror
the functional relevance of a camel--to transport...
On a related topic that is related to visual components of the 'Scape,
have you guys checked out the new Google SketchUp? It's a free app
that lets you generate 3D objects then give them colors, textures, etc.
Apparently if you have Google Earth installed (need highspeed) you can
recreate your own house (or other building) in SketchUp then IMPORT it
into Google Earth -- pretty cool! Googles stuff seems to be fairly
stable (technically and comercially). As such, would it be possible to
have an "import SketchUp object" feature? I'm not sure about the
processor/memory requirements though ... might not be possible.
Last quick though for a motif... 3D renditions of MS Office gizmos?
THere's like a 1000 different bitmap icons you can use in windows.
Maybe 3D versions of these (already familiar) icons... Just a
though...
Next big idea: Quantum leaping 'Scape elements through wormholes... ;-)
Argey
And yes, it relates to the new color-of-topics features that will soon
be introduced.
Argey
Argey
above (it shows up as Sun, Sep 10 2006 1:11 am on my pc, not sure if
everyone would see that time though). Anyway it starts "Tony: Texture
ideas / motifs."
I'm really concerned about confusion in the field of view, obscuring
the topic names with unnecessary detail and avoiding slower performance
because of rendering complex objects. We knew about Google SketchUp,
but hadn't looked at it yet. I've looked at it now and it's very
interesting. We have to find a way of producing models in a form that
we can import into Topicscape.
I think eventually, we should let the landscape be a community effort,
where we provide the information to make and import 3D objects, a
number of simple and I hope elegant skins, and then users can decide
for themselves what they want to see, and whether self-made and more
complex models provide acceptable performance on their own PC.
But first, we have a lot of new functionality to roll out, and that's
the priority now.
Argey
PS Oh, and the wormholes? Already working . . . but we have trouble
with them snapping shut. Lost three staffers that way already, *and*
the black hole is hard to anchor to a topic.
steveK
> with them snapping shut. Lost three staffers that way already, *and*
> the black hole is hard to anchor to a topic.
BLackholes?! I was wondering why my face was smushed up against the
computer screen! ;-)
-steve