it's also possible to toggle any subset of the modifiers by defining your own custom keyboard shortcut in vscode: https://dendron.so/notes/ad270a7d-2aed-4273-8319-eb6536e38b29.html#modifiers

I think I fixed it. The issue probably was that I have 2 vaults configured one for my main (work-related sensitive local-only) notes and another one coming from git journal to be able to add new notes while on the go and have access to a subset of the main notes. Before multi-vault was a thing I had copied some notes from the main vault into the gitjournal one. This has led to notes in multiple vaults with the same id which apparently confuses Dendron. The isolated notes were exactly those. I thought to fix it via

Dendron:Doctor

after deleting the id value but for some reason that didn't work out. So I deleted them in the main vaults. Thanks to the new multi-vault capabilities I still can find all my notes easily. I should have checked all this before bothering you, apologies!

@User - what is your mobile note-taking workflow?

Question: is anyone using the link snippet without removing the https://part from it?

hey guys does anybody know of a good plugin for vscode that will notify you when a task is due?

I would like to know an extension like this as well

yeah so far I haven't found one

they all have to do with tasks for coding etc

🙂

I'm sure kevin could do that with all his free time he has lol j/k

all good. and you just reminded me that i need to update the good doctor now that multi-vault is a thing to check for duplicates across all vaults 😅

what link snippet are you referring to?

@User submitted a feature request to that effect https://github.com/dendronhq/dendron/issues/282 🙂 free time is a little in short supply though to start working on it 😓

@User what was the command to create a page out of a todo you showed me?

highlight todo -> some command -> new page with link to it inserted

oh CMD + SHIFT + S

There something I don't quite understand about hierarchies in dendron. I don't understand why we want hierarchies to be a tree rather then a graph. It seems to me that since any tree is also a graph we could get all of the same benefits. I read the pages mentioning "hierarchies" in "kevinslin.com" and to me it seems that most if not all of them are relevant to graph hierarchies. That does not mean there aren't some beautiful novelties in the arcticles regarding "hierarchies", there is a lot of insight there: 1. We could and should have nodes that are more "high level" or "higher in the tree" and some that are "nodes". 2. Unlike roam other hypertext tools, there should be some "parent child" hierarchies (i.e - not all hyper-links are born equal) But couldn't we still achieve this with some (directional) graph rather then a tree?

I do know that tree search is faster then graph search. But (from my understanding) dendron supports search that is not part of the tree (as it should, since you might not know which note you are looking for), and search can be made very efficient anyhow

Hope this is the right place to ask such questions 🙂 If not - I'm sorry for bothering you guys 🙂

great question and yes, its absolutely the right place to ask such a question. from a "maximizing degrees of freedom" point of view, you are correct -> a graph/directed graph is more flexible than a tree. one of the early design decisions when it came to working on dendron was whether to build it on a foundation of trees vs graphs. i ultimately settled on trees after trying both and realizing that the constraints of trees were a feature and not a bug. this is because while a topic (eg. tomatos) in theory can fit under multiple parents (eg. cooking, fruits, my garden, etc), as a human trying to find information, the act of choosing one thing helps me find it again. without choosing one canonical parent, everything remains an option which hinders both adding and retrieving data: - when adding a note, i'm never sure if i linked to all the possible relations for the new note - when i'm retrieving a note, i'm no longer sure where to start looking i compare the limitations with trees to the limitations of not being able to use

goto

in most modern programming languages: theoretically, goto has more freedom but it becomes impossible for most humans to reason about which is why we have programming paradigms which are more restrictive but let us work with larger quantities of code.

finally, there are many ways to still represent relationships that are not parent-child in dendron. backlinks and note references are the most immediate examples. in the future, we will also add an alias option that lets you make a single note the child of multiple parents. even though we're now wandering into directed graph territory, in this scenario, the note still has its canonical hierarchy, which is the principle that makes dendron works for humans

it seems that a graph would flatten hierarchies out making them not really hierarchies at all. This is an interesting design challenge. Thanks for sharing Kevin!

Thanks for the response @kevins8 . Ok, so I completly understand how constraints can be a feature rather then a bug - as a designer its our (your?) job to "lead" the user toward good practice. (btw, this is something I realsied just 2-3 months ago) However, if I'm hearing you correctly, the main advantage of tree rather then a grath is the ability to search (maybe retreving is a better word). But isn't that a "solved" problem? With the rise of ML and NLP search as become stronger then ever before in finding relevent things from weak undirected searches. In dentron you even allow such unhiearched searches.

That is also a great point! But i guess now that I want to ask is: is it possible to have a graph structure when some nodes are "higher" then others? There are such structure. This can even be done automaticaly by looking at hyper-reference in notes (this was done for wikipedia: the automatically detected datahierarchies in wikipedia notes by looking at the hyperlnk garph. I can find the article latter)

It all comes down to what happes in practice, could be inresting to try and get some data to compare retreval.

Btw - It could be that searching for hierarchies is the part of the process that makes you undersand and synthesize the matirial. But I wonder if its stronger the connecting to graph nodes :)

I would rephrase that - to be, the main advantage of using is a tree is because it forces users to choose. just like the act of writing can help you remember material better, the act of choosing a canonical hierarchy for your notes makes it easier to reference. one of the main benefits with dendron is that you can manage a large number of notes without searching because of hierarchies and lookup. my thesis is that searching is the dominant paradigm in most knowledge management tools because said tools don't give users an easy way to reference their notes which is why people resort to search - not because it is better but because there is no alternative. as for ML and NLP search, I see these things like voice recognition - until they are +98% effective, these solutions are immensely frustrating when they fail. search works well for general knowledge on the internet via google but i find it quite terrible for the personal knowledge (eg. this is why no one can find anything in gdrive/gdocs).

i'm quite sure that such structures exist but it all comes down to tradeoffs in the end. take the great serialization wars.

json

ultimately beat out

xml

as the standard data format for most things because of its simplicity and ease of use. what we're aiming for with dendron's hierarchies is something like json - just enough structure to be useful but not too much to become overbearing

Cant you just think of the file as a node that just has a long name anyway? The files are bidirectional linked anyway. I really like the tree structure, by molding the structure yourself, refactoring, subtracting adding you make these the connections stronger in your brain. But I’m by no means an expert in graph/category theory or neuroscientist 😉