Blogory Fancytree Client Design Notes.
How is Blogory using FancyTree? My server is a tree of Python objects, each at their own URL. This is called Object Publishing Currently I download the HTML for the given node on the tree, and in the background download the json for the whole tree. It only takes a second or two two download 2000 objects. When they are downloaded, I switch from static HTML to fancytree. The user never notices.
The first interesting change I have made is in search. If you type multiple words in the search box, it checks every fancytree node object to see if both of those words match the text. I like this approach better than the experimental fancytree filter. I could share this. My problem is that it is a lot of work to release it, and no perceived benefit. If you are interested talk to me, maybe you can motivate me. Maybe no one cares.
The harder thing to do is to integrate search with downloading the whole tree. As each node is downloaded, I have to run the search against it to see if it should get displayed and the tree counts updated and the display updated.
Thanks for letting me write about this here. No one else would understand or care. Writing clears my thinking.
Right now the server and client only represent trees. Actually knowledge is a graph. Nodes can be in two different locations. I hope to implement this in the future, but have not done so yet.
I am struggling with the issue of which part of the software to release. I have made a lot of changes, it would be a lot of work to release this softwre, so I am posting here to start a conversation. If you are interested in what I have done, let me know.