Logseq - Free Note-Taking That Doesn't Suck

Logseq is different from most note apps because it thinks in chunks (blocks) instead of documents. Don't expect to "get it" immediately - block-based thinking takes weeks to click. You'll probably create a mess the first month while you figure out your system, which is exactly what happened to countless users on Reddit's PKM communities.

Everything stays on your computer. No cloud accounts, no data harvesting, no subscription fees. Your notes are just Markdown files you own. This follows the local-first software philosophy - both the biggest selling point and the biggest pain in the ass. You control everything, which means you're responsible for everything including backups and sync.

The Block Thing Everyone Talks About

Instead of writing documents, you write in bullet points. Each bullet is a "block" that can be linked to from anywhere else. Think of it like having a conversation where you can reference any sentence you've ever said before. This concept was pioneered by Roam Research and implemented in various forms by tools like RemNote and Obsidian's block references.

The linking system is the killer feature. Your notes start connecting themselves as you add more content through automatic backlinks. It's pretty cool when you rediscover old ideas through random links, a concept called serendipitous discovery in the personal knowledge management community. But fair warning - the linking gets addictive once you get it, which means you'll spend way too much time connecting everything to everything else.

Block references break when you rename pages (you'll get "Block not found" everywhere), search is inconsistent and sometimes misses obvious matches, and the mobile apps feel like afterthoughts. But when it works, it works really well.

Privacy Means "Figure It Out Yourself"

Everything runs locally on your machine. No tracking bullshit, no phone-home connections, no analytics. Your data is yours. The downside? No customer support either. It's community help or figure it out yourself.

This is the trade-off: complete data sovereignty in exchange for DIY problem-solving. If you're the type who likes troubleshooting config files at 2am, you'll love this. If you expect hand-holding, look elsewhere.

Works completely offline, which is great until you realize sync is a nightmare. File conflicts happen regularly if you edit on multiple devices. Git works well if you're technical (check out this Git sync guide), but cloud services like Dropbox cause headaches.

More Than Just Notes (But Not Great at Any One Thing)

Logseq does note-taking, task management, PDF annotation, and whiteboards all in one app. It's convenient when it works, but each feature is decent rather than amazing.

PDF annotation is fantastic when it works, but complex PDFs will make you want to throw your laptop. The whiteboard feature is neat for quick sketches but don't expect Figma-level performance. Task management is basic compared to dedicated tools like Todoist or TickTick, and the plugin ecosystem is small - expect half the plugins to break when Logseq updates.

Current stable version is 0.10.9 as of late 2024. The team pushes regular updates that usually fix more things than they break, but always backup before updating.

But understanding what Logseq is doesn't tell you what it's like to actually use it day-to-day. That's where the rubber meets the road.

Logseq gets slow with large vaults. If you're planning to store 10,000+ notes, expect some lag - this is a known performance issue discussed extensively on their GitHub issues. The search indexing can take forever after the initial import (45+ minutes on my 2019 MacBook with 3000 notes), and performance degrades noticeably around 2000+ pages according to community benchmarks.

The Block Linking Thing That's Actually Useful

You link pages with [[Page Name]] and specific blocks with ((Block ID)). The block ID is some ugly UUID that Logseq generates automatically - you'll never remember these, so you mostly just search and click. This syntax is documented here but the UX could be better.

The real magic happens when you embed blocks using the `{{embed}}` syntax. Change the source block and it updates everywhere. This is powerful but easy to overdo - too many embeds create a tangled mess that's impossible to follow, as warned in Andy Matuschak's notes on transclusion.

Bi-directional linking means when you link to something, it automatically knows it was linked. The graph view shows these connections, which looks cool in demos but becomes useless noise once you have more than 100 notes - a common complaint in knowledge management communities.

PDF Annotation (When It Doesn't Crash)

PDF support in Logseq 0.10.x works great on some PDFs, crashes on others. Complex PDFs with lots of images or weird layouts will cause problems. Academic papers usually work fine, but don't try to annotate your tax documents.

Highlights and annotations become blocks you can reference elsewhere, which is genuinely useful for research. But the PDF viewer is basic - no advanced features like form filling or digital signatures. For serious PDF work, you'll still need a dedicated app.

EPUB support exists but barely. It can open simple EPUB files but anything with complex formatting looks like garbage. Better than nothing, but don't count on it for serious reading - stick with dedicated readers like Calibre for that.

Tasks: Basic But Functional

Any block can become a task by typing TODO or DOING or DONE. The system automatically collects these in queries, which is handy for seeing all your tasks across projects.

The daily journal is where most people live. It creates a new page for each day automatically. This becomes your inbox - dump everything here, then organize it later, following the Getting Things Done capture methodology. Daily journals get cluttered fast without discipline, but they're great for capturing random thoughts.

Task queries break regularly when you rename pages or restructure things. You'll spend time fixing broken queries after any major reorganization.

Whiteboards: Pretty But Slow

The Tldraw integration lets you draw diagrams and mind maps. It's neat for quick sketches but laggy with large diagrams. Don't expect Figma-level performance - it's more like a digital napkin.

You can embed notes directly into whiteboards, which sounds cooler than it is. The text is tiny and hard to read. The feature exists but most people ignore it after the novelty wears off.

Flashcards: If You're Into That

Logseq has built-in spaced repetition using #card tags. It works fine for memorizing facts but the algorithm is basic. Anki is still better for serious study, with more sophisticated spaced repetition algorithms.

The cards integrate with your notes, so you can turn any block into a flashcard. This is convenient but leads to inconsistent card quality - some blocks make terrible flashcards.

Plugins: Small Ecosystem, Frequent Breakage

The plugin ecosystem has maybe 100 useful plugins. Half of them break with each Logseq update, and some developers abandon their plugins when they get bored.

Popular plugins include calendar views, habit tracking, and themes. The plugin system uses web tech (JavaScript/CSS), so any web developer can contribute. But the API changes frequently, causing compatibility headaches for plugin developers.

So how does all this reality stack up against the competition? Let's be brutally honest about where Logseq wins and where it gets crushed.