A whiteboard for the main concepts in Logseq

mentaloid · June 9, 2023, 3:30pm

New user here. I have been working on a graph about Logseq itself. Of course Logseq can be used (and abused) in many ways, which are not always compatible with each other, so my work is primarily about my own use. Therefore, this is a disclaimer that the content of my work is subjective, not necessarily agreeing with some official description of Logseq.

The default Graph view is pretty much unusable, even for a graph of limited connections:

Arranging the nodes is difficult but possible, while precision is impossible and the space is used inefficiently:

In the end I have created a whiteboard to position the nodes as intended:

This is not perfect, but does the job well. Given some conventions, it could even be automated. The arrows are a nice improvement, they clarify that this diagram has two logical dimensions. Their abstract meaning is that each node:

is about and can be understood in terms of the nodes on its right
contains and can be analyzed down to the nodes below it

Therefore, this diagram defines and analyzes the top-left node, which is Logseq. With that meaning in mind, I think that the result has the explanatory value of several pages of descriptive text. At the very least it can tie all that text together, especially for new users, who are typically overwhelmed and without orientation.

Nevertheless, with something as rich as Logseq, a single diagram cannot hold every concept, so here are some more concepts, as seen from the perspective of the knowledge column:

phrase level: references, tags, other links
note level: TODOs, queries, other code
concept level: outlines, journals, files, other modules
graph level: whiteboards, folders, other packages
- Technically, every item at this level is also a file, but conceptually it contains other files.
between graph level and knowledge level: namespaces and other domains

Thoughts are welcome.

mentaloid · June 11, 2023, 8:30pm

To give an example of how this diagram can help, consider the long-standing confusion between pages and blocks (an open conversation for years now). As a new user, reading the many threads on the matter didn’t provide me any confidence with how to use them. Here are just a few of those threads:

Picking between page and block in a given occasion can feel obvious, and both options may work at the moment, but this doesn’t exclude regretting the choice down the way:

Not everyone has the time or expertise to make a smooth conversion when they change their mind.
Even worse can be the feeling of unrest when both options look justifiable.

To gain the needed confidence, should shift the focus from the tangible to the conceptual. This is what the structure of the diagram clears up:

A page is a node in a graph (Logseq graph points down to Logseq page).
A page represents a concept within the managed knowledge (Logseq page points right to concept, which is under knowledge).
- A concept is a conceived idea (concept points right to idea), meaningful enough to be given a unique title (not just an id).
  - as a soft rule, something that could be an article in a wiki
- A concept is made up of a list of notes (concept points down to note).
  - The notes of a concept provide an outline for it.
  - A concept is the subject of all its notes.
A page is made up of a tree of blocks (Logseq page points down to Logseq block).
- A block is a node in the tree of a page.
- A block represents a note or a bullet point within a note (Logseq block points right to note).
  - A note associates some concepts (which it may reference with [[]]) (note points right to association).
  - A note belongs to the page of the concept which has the role of its subject.
    - A note in a journal is not in its final position.

mthibode · June 14, 2023, 10:24pm

Here are some parallel thoughts based on Plato’s Divided Line.

mentaloid · June 15, 2023, 9:54am

Good observation, Platos’ line certainly relates to all this. More specifically, it corresponds to the horizontal dimension. It contains all the concepts except of the first column, i.e. Logseq, which is a tool that operates on the other columns.

However, your implementation has some issues:

Connections is used twice.
Areas, Sources, Connections, and References should be in singular.
Most importantly, your rows and columns don’t “move” according to the arrows.
- As a result, the positions of the concepts is highly questionable.

Let me remind the meaning of the two dimensions:

Right arrows express object of operation.
Down arrows express contained objects.

This is a suggested improvement:

Plato’s line:	highest	high	low	lowest
Logseq	Knowledge	Understanding	Belief	Opinion
Graph	Completeness	Comprehensiveness	Coherence	Conjecture
Page	Concept	Truth	Thought	Value
Block	Note	Reference	Connection	Assignment
Markup	Commentary	Area	Emphasis	Source

You may continue from here.

mthibode · June 15, 2023, 11:17am

Yes, you’re right. I didn’t take into account and correctly interpret or adopt the logic of the arrows. I couldn’t quite grasp your idea with a quick read, but I was attracted to the symmetry of your diagram, and its ability to explain how to think about Logseq. I only gave a little thought to ensuring that all of the elements that Logseq offers e.g. graph, page, etc. are accounted for.

In the diagram I was making I was re-imagining how I am trying to use the app. A lot of people describe it as a ‘second mind’ or a way to implement zettelcasten. If Logseq has a teleological purpose, because it’s open source and modifiable to anyone’s purpose, then what should guide its functionality and parts?

For my own part, I am trying to document the opinions annd snippets that I come across and want to further develop into understanding. So far it’s been a good replacement for MS OneNote and a better bicycle for the mind than Apple notes or Miro. I’ll revisit your ideas and arrows again and take onboard your suggested edits. I was also thinking of including the proportions in the divided line, but that can wait.

mentaloid · June 20, 2023, 9:43am

Concerning your edits, it is now apparent that arrows are used differently in each diagram:

In my diagram, arrows represent references (although of specific meaning).
In your diagram, arrows represent movement.
Plato’s single line implies a direction from lowest to highest.

Therefore, even though the same or similar concepts are used, each diagram follows its own conventions:

My diagram tries to improve on the conventions of Logseq’s Graph view.
- It incorporates Plato’s line as a dimension.
- It attempts to describe Logseq’s main concepts in relation to each-other and some other ones.
Your diagram follows conventions like those of a workflow.
- It incorporates Plato’s line as a stairway.
- It attempts to describe Logseq’s main concepts in relation to their usage.
Plato’s line is neither a graph of concepts nor a workflow of steps, but an arrangement of categories.
- It attempts to describe not Logseq, but reality.

Therefore, the diagrams are complimentary rather than directly comparable.

As about the way you use Logseq (i.e. your workflow), it has to be your own, and continuously adapted, guided by your evolving needs. There is no single workflow that can remain effective across all time or all users. Without knowing specific needs, commenting on a specific workflow can be arbitrary and misleading (thought I can still notice the repetition of Connection).

Back to your incorporation of Plato’s line, a lot depends on the definitions of the concepts, and that makes it a philosophical matter:

Is this the correct order? Some examples:
- Can there be any Knowledge without Understanding?
- Is Dialectics about Understanding or about Truth?
- Can a Truth lack Coherence (like some Beliefs)?
Should there be some loop in the flow? Some examples:
- Is understanding a black-or-white matter, or is it the gradual result of recursions?
- Is Logseq for producing understanding, for documenting it, or for both?
Are some concepts overlapping? Are some concepts missing?
- Could it be that some concepts are not nodes (i.e. distinct steps) but arrows (ways to move from one node to another)?
- Is the goal Understanding or Wisdom?

Some of those questions are very hard, and yet their answers can make a big difference. The symmetry of a limited approximation can be more helpful than the fight for precise and complete answers (like the earlier philosophers tend to be more helpful than the later ones). I suspect that this is the real value in Plato’s line as well. Not to oversimplify reality, but to help us understand it in smaller sections of parallel meanings.