What actually counts as backlog in Dynalist
This is honestly where everything goes sideways.
When you first set up Dynalist thinking “Cool, infinite bullet points, infinite nesting, freedom,” what you don’t account for is: the mess you will leave for yourself in about six weeks. Entire projects hidden under collapsed nodes. Random ideas from meetings floating in your inbox with zero context. And somewhere in there, a task like “Call Dave” from three months ago that was still marked as urgent but you have no idea who Dave is anymore 🙂
So when I say “backlog,” I don’t mean just old tasks — I’m talking about untagged, unlinked, non-hierarchical fragments. Stuff that didn’t flow anywhere. Pages marked with “📌” which now only raise the question: why was this pinned?
Here’s how I started sorting out what even counted as backlog:
– Anything in my Daily Notes that had a checkbox and never got checked.
– Bullets tagged with #later, which I used liberally — and never revisited.
– Archived meeting notes that included next steps but weren’t linked to any current project node.
Fun fact: Dynalist doesn’t have real project management features like task aging, so you gotta roll your own system for this. That also means you may be cleaning up based on vibes.
To help yourself survive this, set up a single document called “Backlog Review Holding Zone.” Drag everything questionable there. It sounds stupidly simple, but trust me, having a single point for chaos dramatically reduces the number of mental tabs open.
Filtering out real tasks versus notes and ideas
The real pain when organizing Dynalist isn’t deletion — it’s decisions. You’ll open a bullet that says “Start podcast?” and stare at it for a solid 20 seconds wondering if it was a joke. Was that something you jotted down late at night? A real commitment? Something someone else said in a meeting?? ¯\\_(ツ)_/¯
That’s why I built a two-pass system:
**Pass one: Categorization, no decisions.**
Use tags — the less brainpower, the better. I added bulk #tag commands for sorting things as #task, #idea, #note, or #junk. Here’s a quick comparison I sketched that lives in the top of my Dynalist doc now:
“`
#task → clear action, person responsible, possible due date
#idea → no specific action, could float freely
#note → captured info, no action expected
#junk → irrelevant, old, abandon it
“`
This system saved me from spirals. You don’t even have to decide what to do — yet. You just point and label like a video game.
**Pass two: Handle only #tasks**
Everything else can wait. Sort just the bullets with #task into:
– Project they relate to (if they still do)
– Whether they can be done this week (I use another tag like #thisweek)
– Whether they require info first (they get tagged #blocked)
Once you clear this batch, it gives you quick dopamine. Also, it exposes how many of your tasks were actually hallucinations from two months ago.
Making custom views with search operators
So this was the thing that literally saved my brain: Dynalist’s search is ridiculously powerful… and also super easy to forget about.
I had been scrolling forever, combing through each project tree manually, until I remembered you could create saved searches with custom operators.
Let me give you an example of one that changed everything:
“`
checkbox:true -is:completed tag:#later
“`
That one line shows me all unchecked tasks tagged #later. Basically, the guilt list. I titled it “Unfinished Promises” because: realism 😂
Oh, and here’s another favorite:
“`
parent:ProjectX checkbox:true -is:completed
“`
That fetches all incomplete tasks related to a specific project (assuming ProjectX is in a parent node named that). Which means I don’t need to even open the main project outline anymore unless I’m actively working on something.
Once you build a few of these saved searches, they turn into a dashboard. They live in your bookmarks pane, and you can right click → Open All in Tabs if you’re into chaos (or if you’re me, with 34 saved filters spread across different client work). Totally worth it.
What broke when I tried GTD in Dynalist
Oh my god okay. So, I thought I’d be clever and implement David Allen’s whole GTD flow inside Dynalist. You know the drill — Inbox, Next Actions, Projects, Waiting For, Someday Maybe, etc. So I built each one as a header in one doc, created templates, added checkbox logic.
Everything broke.
First, I underestimated how badly Dynalist handles cross-referencing between nodes. If you move something out of one structure to another (say, Inbox → Next Actions), the collapse state resets, and suddenly half your nodes expand when you don’t want them to. Super annoying.
Second, I didn’t set up consistent tagging. I relied *entirely* on parent headers, so the moment I tried to build a saved search across the system like:
“`
checkbox:true parent:NextActions
“`
…suddenly it wasn’t finding items that were deeply nested, or ones accidentally left in Inbox. The parent filter only checks *immediate* parent, not grandparents. Whoops.
Last straw — my “Waiting For” bullets triggered no reminders. Because Dynalist has no date-based notification system — just static dates. I forgot that a task could sit there forever saying “@due July 4th” and nothing would happen 😛
Fixing it meant:
– I had to add redundant tags like #next, #waiting, #someday, even if items lived in the right section. It was annoying but prevented stuff from becoming invisible.
– I stopped nesting more than two layers. Too easy to forget you even *have* a task buried six pages deep.
– Also, tried turning off automatic expanding of nodes when moving bullets — there’s a weird setting in the daily logic doc (?) that affects this.
TLDR: GTD works better in apps that are actually task managers. In Dynalist, it’s more like cosplaying GTD.
Using date tags to prune time sensitive tasks
My backlog had seasonal ghosts. Like, legit tasks that were about Black Friday promotions or tax forms or even a coffee shop popup location that shut down six months ago.
The key trick was scanning anything with a date in it — which sounds harder than it is. Dynalist lets you insert a date by typing `!` — so `!2023-11-15` becomes a proper date tag. Even better, you can search for these with actual operators:
“`
within: -7d
“`
This shows anything with a date tag on it from the last week *backward*. You can flip it to:
“`
outline:true before:today
“`
And that shows you all bullets that were scheduled *before today* — basically missed deadlines.
Once I cleared out tasks with missed dates, I felt lighter. Then I went full-on and created a calendar-based archive. Here’s how:
– Made a folder called “Task Archive by Month”
– Each file was named Jan2024, Feb2024, etc.
– Once tasks were done or irrelevant, they got dragged to the right month file (especially if the date still mattered contextually).
I love folders way more than one mega document. The only tradeoff is that search operators on folders are slower.
Bonus: you can also create a search that finds all tasks with dates “next week” to preview your schedule:
“`
within:7d
“`
Play around with it. Feels like hacking, kinda.
Quick keyboard cuts that saved me hours
This part is surprisingly overlooked. People use Dynalist like a clicky notepad when in fact, most of its power comes from keyboard nav. And once you learn just a few of these, you stop wasting time with mouse movements:
– `Ctrl + Shift + M` → Moves a bullet to another document. Crucial when you’re mid-review and want to dump something into a focused backlog file.
– `Ctrl + ,` → Opens the settings menu. I did not even know this until year two 🙁
– `Ctrl + Enter` → Toggles checkbox on the current bullet.
– `Alt + Shift + Left/Right` → Collapse or expand nodes.
Also highly recommend muscle memorying the sidebar nav:
– `Alt + Shift + Up/Down` → Moves whole bullets up or down. Way faster than dragging.
One thing that tripped me up: I kept accidentally using `Tab` thinking it’d indent — but in the new desktop version, it sometimes triggers other stuff depending on context. So I now only indent using `Ctrl + ]`.
If you’re cleaning backlogs regularly, this kind of speed matters. Otherwise you’ll sit there dragging bullets across collapsed trees like some kind of medieval monk 😅
The forgotten Dynalist setting that solved everything
So while knee-deep in reorganizing all my backlog junk, I was frustrated with how collapsing worked. Sometimes I’d open a project node and *everything* inside would expand, even things I specifically remember collapsing.
I thought I was going insane.
Turns out, buried in Dynalist’s Settings, under *Advanced*, there’s a setting called “Remember collapsible states.” And yes, mine was turned off.
I literally have no memory of turning it off. Maybe during an update, maybe during a moment of rage-unchecking a bunch of options. But this one tiny toggle was the reason why opening a tree caused every sub-level to spray open.
After switching it back on, everything snapped into place. Now, when I collapse a node, it *stays* collapsed. Even across sessions. Even across devices.
Truly felt like discovering that half your misery was just one unlabeled checkbox in an unloved part of the UI 💀