I am unhappy with how blogs are in the pursuit of unique insights yet hide them deep within blog posts. I feel the organization should be first finding the idea, then looking for arguments. So, I made Logic Map.
Visit here: www.logicmap.com. Click around to jump between ideas. Press W/A/S/D on desktop to fly, shift to fly faster. When outside an idea, click on an idea to immerse in it.
I felt that it needed to be rendered in 3D for two reasons:
-
Force directed graphs in higher dimensions are a great way to organize related ideas. This provides a recommendation engine as well as a useful experience when flying around
-
Seeking truth is a harsh endeavor – providing a beautiful place for it to take places inspires greatness.
Since the bar for performance are text based blogs, I had to be very stingy. The whole thing is 6 draw calls, mainly relying on extended Physical Shaders and instancing. Built using R3F. Press cmd+i to open the inspector view.
Happy to hear any questions / suggestions on the 3D side!
Thanks for sharing.
Presenting concepts maps in 3D solves one of the inherent problems of 2D mapping – crossing of lines. In 3D this almost never happens.
Here are some comments:
-
I find it hard to make use of the lines between idea blobs – i.e. the lines a nice 3D effect, but are not practically useful, because they are too many, too dashed, and it is hard to follow them visually. My suggestion would be to show thick lines (3D cylinders) between the central blob and its direct connection; thin lines for the next level of connections and the current lines for all other distant connections
-
having a map of idea blobs without labels seams strange, this it is hard to pick a related idea and go to it; random clicking on blobs does not provide the level of exploration that I would have expected from such map (yes, I can see related ideas in the text panel, but I cannot see where they are in the 3D scene). My suggestion is to add a label to the main blob and smaller labels to its directly connected blobs; I’m not sure whether a third level of labels is good, as it may clutter the screen. Maybe making this option toggable?!?
-
it might be good to give some context as why two ideas are related. For example, code (a set of instructions) has a nearby idea medium (a uniform way to express ideas). Why and how are these two ideas close?!?
-
it would be nice if users can use this tool with own data, i.e. somehow do define concepts and relations in a text form and import it
-
would it be nice to have the following functionality: the user picks two unrelated idea, the tool find paths from the first idea to the second idea via some intermediate ideas … and shows these routes in 3D
