@bitbonk @logseq It may help to hide the journals (and I wish I could hide property pages too). But if you keep everything in the journal (as it seems to be encouraged), the graph view doesn't show many connections, as it only shows pages and not blocks. I'm still a bit confused about the point of this.
@eldelacajita @bitbonk @logseq
Yeah it's confusing but let's try to make sense of it:
In Logseq the so called pages act as tags (no matter if using [[wikilink]] or hashtag syntax). And these pages/tags are nodes of a graph created using references as connections.
So while for every tag you have a page, it doesn't mean you need to write something in it. You may want to treat most pages as tags that will form a cloud of nodes in the graph view.
Using the journal only is an extreme option, that way you will treat every single page as a tag. Instead you can use a small number of pages to gather content but still using a lot of references to empty pages so that your graph is an accurate representation of the knowledge in your notes.
My advice is to use the journal as an inbox for quick notes, then move them to a few pages and once a page become as long as the chapter of a book, move a portion of its content to another page.
@post @bitbonk The thing is, if you just use the journal pages and leave the other pages as empty "tags" (as it seems to be intended by design), the graph won't show any connections between those "tags", because the block-level links (references) aren't displayed.
So if I use #Logsec as intended the graph will be useless.
For it to show a proper network of concepts, all blocks should be in the pages, which would become containers and not only empty "tags".
But if you use the "everything into pages" approach to make the graph view show those links, the linked references section in each page becomes less powerful, as most of the related blocks will be inside the page and with no such filtering features.
And if I combine both approaches (some things in the journals, other things inside other pages) the resulting graph will just show half the picture, which can be even more misleading.
Or am I misunderstanding something?
Your question definitely makes sense, you are asking how to keep using a page full of content like a tag (without using queries).
If [[X]] has actual content but you want to filter it according to references to [[Y]], why don't you just open [[Y]] and filter its references with [[X]] and eventually other pages?
I.e. use an actual page-as-tag as the entry point of you filter.
In order to avoid such a (at least to me) counterintuitive workaround, I'm missing a simple thing in #Logseq: a "page filter" on the top right of the page, with the same behavior as the linked references filter at the bottom, but acting inside the page contents themselves.
https://discuss.logseq.com/t/filters-for-note-body-not-just-backlinks/1305/37
This alone would make putting content inside the pages a viable approach. And it's basically just copying or extending an existing feature and a known behavior.
@post @bitbonk
Thanks for the suggestion. That's a way of working around that limitation. Is it
Now, imagine I want to focus on a Project so I go to its [[Project]] page, and then want to see the #fleeting notes, the TODOs, and things related to both [[Project]] and [[topic]].
Sure, I could go to [[TODO]] and then filter the linked references by [[Project]], then go to #fleeting and do the same, then go to [[topic]], etc.
But that doesn't seem a very practical workflow.