Hello there! 
I have used LogSeq for quite some time now, but I’m not 100% convinced that should be my go-to app yet. So I thought I should turn to the community, and therefore here’s a little presentation of my relationship with the app.
I have two kinds of usecase I want to cover with the app:
- Organizational/Report pages: take notes on the fly during a meeting/conference/class…
- Encyclopedia/Knowledge pages: document things, structure notes about things I learn and I want to remember
- And then, find back any of this very quickly when I need it
What I like about LogSeq:
- Privacy: I own my data, and can sync it between devices (I use Syncthing, so far so good)
- Extensibility and openness: there are plugins and I can develop things if I want to
- Outlining: hiearchical/tree nested-bullet-point-style note-taking
- Relationships: the graph/network/mindmap logic intending to make easy to link pages together
What is dislike about LogSeq:
- Performance issues, it’s the biggest caveat in my eyes (see Performance and battery issues - #2 by tayflo)
- Overall the app feels slow and not blazing-fast reactive (right now my graph is around 400 pages, a few are 1000+ lines long, some are of a few hundred lines, most are below that but I would need longer pages and a lot more of them)
- Long pages are barely usable (big problem; I write a lot, I need a lot of details)
- Therefore, switching between pages is problematic
- It slows down the mental process. I would like my tool to be swift so it does not hinder my toughts
- As a result, I don’t use LogSeq yet as much as I would like; for instance for quick-and-easy notes, I use something lighter and which feel more reliable (simple markdown files)
- Search is inefficient because of how are handled long pages and collapsed blocks (see Performance and battery issues - #2 by tayflo and Searching in the editor does not find text inside collapsed block · Issue #6913 · logseq/logseq · GitHub)
Kind of things that I would like in the long run (through new functionalities, plugins, tweaks or such):
- Be able to quickly have an history overview at graph-level, like: what pages I have created/edited in anti-chronological order (I have activated git, but the view is per-page, right? I probably could do a little better using the journal and references to it)
- Be table to connect to other apps my LogSeq data (such as Zotero, Grist, maybe someday a CardDAV/CalDAV server or such)
- I don’t want to import the data (irc, that’s what the Zotero plugin do), because it carries sync issues along (or at least the sync needs to be both-ways, because I want things to be coherent; coherence is more easily reached with a single source of truth)
- I rather want to be able to easily create relationships, and have some info from that (for instance, I put the ID of a Zotero document, and it’s linked; then, some kind of plugin could display infos fetched from Zotero, and/or allow me to click to open it in Zotero; the other way around, it would be nice to have LogSeq notes embedded in Zotero)
- To put in another way: I don’t want LogSeq to be my everything-app. I want it to be a tool fabulently competent at taking note, that I can connect to specialized app that are also excellent in their domain (and I intend to use Grist as my “structured data” app, even if there could be some overlap with LogSeq)
- Create a page directly from a template, with also auto-name in the templating (I feel like it would cover some of my usecases and speed up my workflow, especially on mobile)
- Maybe sticky parent-bullet-points, to keep track of where I’m at (there is a plugin for headers only, which also is not very reliable)
- Maybe smart headers: deduce the header hierarchy level from the bullet-points hierarchy
I also feel like I’m not using LogSeq correctly and I’m asking myself questions:
- What should be used as a page name?
- Given the “unlinked reference” feature, I guess it should be the name of the object
- But for notes that are event-related, what should it be? It could almost be a random string, with important metadata in properties
- I tend to use slashes in names to organize things (“Category/Subtopic/Subject”)
- But I feel like it’s not the way, I goes against the network logic
- I do that because it’s quicker than to set properties or reference (create a new page > enter the name with slashes > done)
- How to use properties efficently?
- What should go to a page? What’s the page perimeter, when to use block/sub-bullet-point, when to make a new page, when to use embed, what’s the workflow?
- Are there ways to easily split a page? That is, transform a block into a page, and embed it automatically in the original page
- How does an efficient logseq knowledge base look like?
Does the community has some kind of resources, guidelines, examples of workflows? About the “proper/efficient usage of LogSeq”.
There are a lot of topics on the forum but I don’t quite know where I should start, and I’m not sure if/how/where some of my subjects of interest are adressed (I know there is a DB-version in ongoing development which is intended to speed up performance).
So, I’m happy to join the community, feel free to react and share your thoughts 