Definitely, and I appreciate your (and Edward's) responsiveness and openness. I don't grasp everything in your message (yet), I only installed leo on one computer yesterday and was a bit surprised to see it (pip actually) download PyQt5 just after I had manually installed PyQt6, and ... not installing leo6.6 but an older version. I expected Qt would be used for some visualization, but I have no experience with it either. I'm still out on the priority to cover my "information bowls" in leo (between my Python logfile mill, long-term food reserve management, Belgium addresses with fibers, home climate control, etc.) Most imply some future Python code, but I actually hope that leo would help me see more clearly where I should put my efforts first, by evaluating the "RoI" of everything I put in.
On Monday, March 21, 2022 at 3:27:10 PM UTC+1 [email protected] wrote: > I've used and liked mind maps very a long time, and I'm very picky about > their appearance. I started with Tony Buzan's book, before there were > programs to make mind maps. Mind Maps were among other things an excellent > memory aid for me. I used to take meeting notes as mind maps in real time > (by hand), and I found time after time that I could pick up one of them say > five years later, and in five or ten minutes I could reconstruct the whole > meeting. I also taught several graduate level classes using mind maps as > my lecture notes. It worked wonderfully well. > > I dislike almost all the mind mapping programs out there that I've tried, > including the Brain, mostly because their appearance doesn't suit me, or > the way you interact with them works poorly, etc., etc. I used to use the > old Mind Manager product (I'm talking 15+ years ago) and liked that pretty > well. But now the successor is way too complex and garbaged up, and pretty > expensive to boot. So I don't use it. > > The one thing a mind map doesn't do is to show incoming edges - that is, > "predecessor" nodes. That's not a bad thing, but it prevents one from > moving freely around a large knowledge base. It's not the intended use of > mind maps, so it's not defect in the basic idea. If you want to make > effective use of a knowledge base it's a weakness, one that The Brain has > tackled. But because of the emphasis on "thoughts", the Brain is weak on > the relationships, which is where the real meat of a knowledge base is or > should be. And I don't like the way their displays shift around. > > The zettelkasten approach is as you say about notes, but that's a > superficial view. It's much more about the connections, and also, if we go > along with the original, the "notes" are intended to be carefully thought > out summaries of various thought fragments, not simply reminders dashed off > in a hurry. > > Leo's basic structure of a node for everything has proved to be quite > wonderful. It is limited a bit because the nodes are text only, but that's > for the most part livable. A node has a single incoming edge (not counting > clones - clones introduce a sort of hybrid model, and I don't want to get > into it here). So if we want to have a node with multiple incoming edges, > and outgoing edges that link to nodes that aren't "child" nodes, we need to > carry more information, and we need a way to visualize and navigate the > system. > > There are many ways to carry more link information. For the zettelkasten > system I worked out some time ago (there are one or two threads on that > from a year or more ago) I use gnx identifiers (Leo's internal node ids) > with notation in each node's body to define a link. For my browser > bookmark manager - still in progress - I use Leo's UNL paths to construct > internal links each time a display is built. One could also use UAs (User > Attributes) though I haven't felt the need to so far. > > It's the visualization and navigation of any system that is based on > graphs that's the hard part. It's hard both from a design/UI point of view > and from an implementation point of view. The really nice part of basing > these systems on Leo is how easy it is to reorganize, restructure, and edit > the information. I have found that to be a real strong point. The > disadvantage is that there is no intrinsic graphical subsystem/API for > displaying the information (preferably with links). > > I'm inclined to think that using Graphviz may be good for generating some > of these displays. Since Graphviz is a stand-alone console-based C > program, it may turn out that a server is needed to interact with it > effectively. Leo already has the skeleton of such a server, and the > leointeg project (that lets you work on Leo outlines in Visual Studio Code) > is a proof of principle. > > Whew, that was longwinded! If you want to pursue this subject further, we > should probably start a new thread. You can see that it's a topic I'm very > interested in. > On Monday, March 21, 2022 at 4:31:38 AM UTC-4 [email protected] wrote: > >> It doesn't surprise me. That's indeed the strength of the Brain (to me), >> it is basically a mindmap but with the *current thought* as the center, >> it's not hierarchical. In that sense it is different from an outlook. >> OTOH, in zettelkasten I get the impression that the focus is on the >> taking of notes in an outline, and just adding #tags and links between the >> notes. I don't directly see there anything that leo couldn't do better. Can >> you clarify what you have in mind here ? >> >> On Monday, March 21, 2022 at 6:01:46 AM UTC+1 [email protected] wrote: >> >>> I fooled around with a sample Brain project on the Brain's web site. I >>> was able to import the "thoughts" and links. The project file contains a >>> number of .json files, although they are not actually valid json. But it's >>> possible to work with them anyway. So yes, it's not hard to do an import. >>> The bigger question would be what kind of interface would work well with >>> them and how it would fit into Leo's node system. Contrary to what I >>> speculated above, the thoughts and their links do not look much like >>> bookmark collections. Basically, each "thought" node has both incoming and >>> outgoing links, potentially of any number. It is possible that they would >>> fit into the zettelkasten paradigm, which is basically one way to >>> synthesize multiple incoming and outgoing links. But whether that would be >>> a good approach or not, the interface and display are what need to be >>> invented. There would be no point in trying to reproduce what The Brain >>> already has! >>> >>> On Friday, March 18, 2022 at 2:12:18 PM UTC-4 [email protected] wrote: >>> >>>> Actually, just the browser bookmark scripts and optionally the mind >>>> mapping visualization, would probably do it. >>>> >>>> On Friday, March 18, 2022 at 2:07:27 PM UTC-4 [email protected] wrote: >>>> >>>>> Having downloaded one of the example brains, and looked at a few of >>>>> the online examples, I'm coming to think that a combination of the >>>>> browser >>>>> bookmark manager scripts I'm working on together, perhaps, with the >>>>> zettelkasten-style organizing scripts I offered in a different thread, >>>>> would provide similar capabilities. Add some enhancements to the mind >>>>> mapping script I provided a while ago, and we might be able to have a >>>>> really good alternative. >>>>> >>>>> On Friday, March 18, 2022 at 11:28:31 AM UTC-4 [email protected] >>>>> wrote: >>>>> >>>>>> I've tried TheBrain maybe three times over these many years. It >>>>>> always seems so promising, and I always abandon it with frustration. >>>>>> >>>>>> On Friday, March 18, 2022 at 11:10:22 AM UTC-4 [email protected] >>>>>> wrote: >>>>>> >>>>>>> ... Now, maybe leo could also import TheBrain files ? ;-) >>>>>>> >>>>>> >>>>>> Now there is an interesting idea. A Brain file is a .brz (not >>>>>> actually .zip, but same idea). It has a lot of individual json files >>>>>> that >>>>>> obviously have to work together, along with some icon files. Much of the >>>>>> data is metadata, which might or might not be of interest. The big >>>>>> thing >>>>>> to settle would be how present it all, and one big question there is >>>>>> whether there can be cycles, and if so, how to handle them.. >>>>>> >>>>> -- You received this message because you are subscribed to the Google Groups "leo-editor" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/leo-editor/36d496a3-ce2f-4772-9562-957404768580n%40googlegroups.com.
