FreeMind certainly has potential and I may end up using it, but it's really hard to settle for the crutches when you already know about exoskeletons.
There is Google Browser (a Java application for navigating Google index by relationships between sites). http://www.touchgraph.com/TGGoogleBrowser.html
The interface there is not necessarily much better than that of FreeMind, but it has much more potential. Mind maps are not supposed to be hierarchical, but organically growing and intertwined a lot. In FreeMind I can add links, but they aren't really good if compared with something universal like GoogleBrowser.
I think FreeMind would become much more powerful (not to mention cool looking) if it had a mode more similar to GoogleBrowser.
It would be great if arbitrary links (not just parent/child) between any nodes were supported, if text size and colors would dynamically adjust according to the distance, relevance and other factors, and if the layout would be dynamic, with the graph rearranging itself on the fly.
Just want to add that the same technology was already appled for Wiki visualisation and some other cool stuff (visit the TouchGraph site for details, as well as this page http://www.jspwiki.org/PageInfo.jsp?page=TouchGraphWikiBrowser\).
I would recommend you read more about Mind Maps and Tony Buzan. The links supplied do not appear to reflect the design concepts intended by freemind.
"I would recommend you read more about Mind Maps and Tony Buzan. The links supplied do not appear to reflect the design concepts intended by freemind."
You may be right, but I think that the program would be infinitely more useful if crosslinks and interconnections were at least an option. So much information is NOT hierarchical, and Freemind ONLY allows hierarchical organization. This is truly something I can accomplish, albeit less elegantly, with MS Word. If Freemind offered a truly organic network of connections, it would have a real selling point, I think.
Indeed, raw information is not hierarchical. But information by itself is not useful. It is the abstract model build to interpret a pertinent part of it which is useful. Useful abstractions (without unconsistencies) are necessarily hierarchical (the underlying relationship leads to a tree and not to a directed cyclic graph which becomes very quickly messy).
Process control experts well know that only loops can give unstability (even if loops are very useful to model or to cause a suitable dynamical behaviour).
Then FreeMind is very useful to build true static hierarchical models to interpret a bunch of raw non hierarchical informations.
At least, it is my point of view.
PS: beg your pardon for my english which is not as quite clear as i whish.
To follow last post. The possibiloity to draw a DAG (direct acyclic graph) instead of a tree could be a great enhancement (suppress the need to duplicate some nodes). But it is not so easy to image a procedure to do so.
PS: these kinds of restrictions permit to associate th graph to an algebra.
You might like to check out
As far as I have seen in the two previous references, the presented graphs are DAG. It seems there are no cycles.
They are exactly the king of graphs whose freemind permits only to build the spanning tree and not the true DAG.
Nevertheless the graphical rendering is impressive in the two cases.