Simple stupid query eaten by logseq

I have to say that logseqs feature set looked enticing half a year ago when I started to use it. It still looks enticing, if only because I am no nearer to actually use these capabilities:

For fun (and to learn), I am trying this:

  • Sprinkle three-line-blocks like “[[expense]]”, “category:: foo”, “amount:: 20” throughout my journals.
  • Create a query that collects all these blocks and presents the categories and amounts in tabular form.
  • Have a sum of the amounts underneath the table.
  • And as a stretch goal, group the table rows by category and produce sums by group.

So we start with {{query (and [[expense]])). Fine. Now it is time to add two lines to the query block that I cribbed from somewhere else: “query-table:: true” and “query-properties:: [:category :amount]”.

Next: as a child of the query, add “{{function (sum :amount)}}”.

At that point, just as I hit the TAB key to demote the new block after the query block, Logseq eats the two properties I just entered. I was going to attach a short (35 sec) video, but .mkv and such are not authorized.

Much worse, though, is that Logseq is actively user-hostile. Pretty much nothing is “discoverable”, there is no printing, no decent export to PDF (yes, I tried the plug-ins), no intruduction for people like me who look for a tool and not for a hobby.

I may have to go back to Onenote. Much less powerful, but what there is works and is easy to use – it helps me working instead of becoming this huge time-suck.

Cheers,
Felix.

1 Like

Welcome to Logseq and the community @felixkasza! Sorry to see you’re running into issues.

You mention some issues that make me think the graph is stored in a location where another tool also accesses it. Are you perhaps using iCloud, Dropbox, Google Drive, etc.? That could cause Logseq to “eat” the notes (though in reality it’s the sync tool that’s overwriting the notes on disk).

As for queries themselves: this is a feature that’s more aimed at advanced users. You say that you just want to use a note-taking tool without it becoming a time-suck. However, I don’t understand why you’d want to immediately write queries in that case, instead of using the Linked References as a starting point.

Logseq has a Getting Started guide built-in. Click ? in the bottom right in Logseq and then Handbook for the onboarding manual.

If you have more questions, feel free to post them here, send me a DM, or email at support@logseq.com.

Thank you for taking the time to reply, I appreciate it!

Regarding the disappearing properties, I hasten to stress that (a) my notes are not shared, synced or anything of the sort and that (b) what disppears is the text of two properties, “query-table:: true” and “query-properties:: [:category :amount]”. I type them in, start a new block, and Shazam! the properties I just added in are gone.

The properties are still in the 2025_03_15.md file backing this journal-day, but making things typed by the user (me!) and belonging to the user (me!) vanish is something I consider very rude. This is an aspect of what I call “user-hostile” behaviour.

Getting back to the big picture. Yes, I read the Getting Started guide. I am so old that I read manuals and in fact expect them to exist; today’s haphazard piles of fragmented web pages that rely on each other in circular dependencies are not my thing at all. Me, I like a readable, narrative-style teaching section supported by a reference section.

But the Getting Started guide, I must say, is well-written, and what topics it explains, it explains very well. However, simple (“live”) queries at the very least are an absolute necessity for using logseq and truly have to be covered. If I need only an outliner with nothing beyond that, why, Word has an outline mode, and it can print and save to PDF.

Finally, while I termed myself a newbie, my earliest journal file is 2024_08_02.md, I have made repeated attempts upon querying, I have written in LISP long before there was Clojure or logseq, etc. I am only a newbie with respect to querying, and I may be condemned to remain in this state forever.

So allow me to sum up what I see as the major areas that could use some love:

  1. Print, print, print. And save to PDF. If I need one piece of paper with a shopping list on it, I am not going to drag a laptop into the local QFC and balance it on the cart handle while I shop.
  2. Print some more, namely query results. Is it really that hard to just send the rendered results to a printer or to a PDF? Printing is core functionality, not something to be handed of to a plugin that tries and fails at the task.
  3. An additional chapter to the Getting Started guide: Queries for Noobs.

As I am old as dirt, I may have extra time on my hands in the near-to-middle future. When I do, I might try my hand at something like a Queries chapter, at least for what logseq calls “live queries”. For instance, how one collects all the blocks with “[[some topic]]” into a list, extracts specific properties, outputs them as columns in a table, and aggregates specific colums, perhaps grouped by the values of some key columns.

But no Datalog. I would take the easy way out and just publish a list of mental-health hotlines.

1 Like