Journals are nested, why and how to fix?

Hello, I just set up a git repository on the desktop and noticed the file path is
/home/rapi5/Documents/logseq-repo/journals/journals

This doesn’t seem right as these are the files in /home/rapi5/Documents/logseq-repo/journals

 rapi5  raspberrypi  ~  Documents  logseq-repo  journals  main  $  ls
2024_03_18.md    2024_08_19.md  2024_10_21.md  2024_11_11.md  2024_12_10.md
2024_03_22.md    2024_08_24.md  2024_10_22.md  2024_11_12.md  2024_12_11.md
2024_03_23.md    2024_09_09.md  2024_10_24.md  2024_11_13.md  2024_12_12.md
2024_03_29.md    2024_09_10.md  2024_10_27.md  2024_11_14.md  2024_12_13.md
2024_04_03.md    2024_09_12.md  2024_10_28.md  2024_11_15.md  2024_12_14.md
2024_04_15.md    2024_10_03.md  2024_10_29.md  2024_11_16.md  2024_12_17.md
2024_04_21.md    2024_10_04.md  2024_10_30.md  2024_11_18.md  2024_12_18.md
2024_04_22.md    2024_10_07.md  2024_10_31.md  2024_11_21.md  2024_12_19.md
2024_04_22.md-2  2024_10_08.md  2024_11_01.md  2024_11_22.md  2024_12_20.md
2024_05_08.md    2024_10_09.md  2024_11_03.md  2024_11_23.md  2024_12_21.md
2024_06_08.md    2024_10_10.md  2024_11_04.md  2024_12_04.md  2024_12_22.md
2024_07_01.md    2024_10_13.md  2024_11_05.md  2024_12_05.md  2024_12_23.md
2024_07_03.md    2024_10_15.md  2024_11_06.md  2024_12_06.md  assets
2024_07_03.md-2  2024_10_16.md  2024_11_07.md  2024_12_07.md  journals
2024_08_08.md    2024_10_18.md  2024_11_09.md  2024_12_08.md  logseq
2024_08_15.md    2024_10_19.md  2024_11_10.md  2024_12_09.md  pages

and these are the files in /home/rapi5/Documents/logseq-repo/journals/journals

rapi5  raspberrypi  ~  Documents  logseq-repo  journals  journals  main  $  ls
2024_12_24.md  2025_01_06.md  2025_01_25.md  2025_02_07.md  2025_02_22.md
2024_12_26.md  2025_01_07.md  2025_01_26.md  2025_02_08.md  2025_02_23.md
2024_12_27.md  2025_01_08.md  2025_01_27.md  2025_02_10.md  2025_02_24.md
2024_12_28.md  2025_01_09.md  2025_01_28.md  2025_02_11.md  2025_02_25.md
2024_12_29.md  2025_01_10.md  2025_01_30.md  2025_02_12.md  2025_02_27.md
2024_12_30.md  2025_01_13.md  2025_01_31.md  2025_02_13.md  2025_03_02.md
2024_12_31.md  2025_01_15.md  2025_02_01.md  2025_02_14.md  2025_03_03.md
2025_01_01.md  2025_01_18.md  2025_02_02.md  2025_02_15.md  2025_03_05.md
2025_01_02.md  2025_01_21.md  2025_02_03.md  2025_02_17.md  2025_03_06.md
2025_01_03.md  2025_01_22.md  2025_02_04.md  2025_02_18.md  2025_03_07.md
2025_01_04.md  2025_01_23.md  2025_02_05.md  2025_02_19.md  2025_03_08.md
2025_01_05.md  2025_01_24.md  2025_02_06.md  2025_02_20.md

Just synced the Desktop Logseq 0.10.9 to the Github repository and Logseq prompts; these errors;

My Logseq data is now on the Desktop, however I am confused what needs cleaning up.

Also, is it possible to create a new graph which stays local and private without syncing?

As the error says, it looks like you’ve somehow nested a graph into another graph. That’s possible when you tried to create another graph but did so within an existing graph folder (in this case it seems like you’ve created it inside the journals folder?).

As you use GitHub, it could also be an improperly setup repo which caused this nesting.

The best solution now is to manually untangle it, by moving the nested folders out of the journals folder (it should not contain any subfolders).

Yes, Logseq is a local app and all you do by creating a graph is selecting a folder on your device’s disk where to store your notes. It’s best to store your graph in a local folder. Unless you want share a graph between devices or want automated backups, there’s no need to have a syncing solution in place.

However, as you use GitHub I assume you have a need for it?

1 Like

Thank you, I’ll try that manual corrections.
Well, I use a public graph which syncs from desktop to github to phone.
I would like to also have a private local graph.
Having the public and private graphs I can switch around in would be ideal.

You can. Just create a new graph and point it to a local folder (that’s not synced with any tool like iCloud, Dropbox, Google Drive, etc.). Now you have a local-only graph.

You can switch between graphs by clicking your graph name in the top left sidebar. The name of a graph is the name of the local folder you selected/created when creating your graph.

1 Like

All my Logseq app data is deleted from 20250326 - 20250409. I tried to merge stuff, but it’s a mess.
Need a pro to walk me through, syncing and clean up gives me a headache.

Now that Logseq app is wiped of all data after 20250326, I typed “test” in 20250409. A sync completed, but
https://github.com//logseq-repo/tree/main/journals/journals
and
https://github.com//logseq-repo/tree/main/journals shows no “test” from 20250409, just synced data up to 20250326.

Just checked the phone → Files app → Browse → Locations → On My iPhone → Logseq → SyncLogseqViaWorkingcopyWithGit → journals → journals → 2025_04_09 → today’s new data is there.

@Ken sorry to say, but it seems like you don’t feel comfortable with file systems nor with Git. No shame in that, but if you’re having so many problems then Logseq might not be the tool for you.

Why not look into a browser-based tool like Roam or Tana? That will take any worry about messing with local files and git merge conflicts. Just my two cents.

Thanks, I guess I’m searching for FOSS which must be why I didn’t choose Joplin, Obsidian, Roam or Tana.
I think they cost a fortune also, and I’m not an enterprise, just a normal individual person.

Joplin in FOSS. You only need to pay if you want sync and don’t want to self-host the (also FOSS) sync server.

Although I’m very comfortable with git, for logseq I found icloud was by far the easiest way to get sync working. If you’re not on iOS, I’m sure dropbox etc would work just as well.

1 Like

A word of warning here: file-sync tools like iCloud and Dropbox tend to cause data corruption when used with Logseq. And you still need to be comfortable with the file system to avoid nesting graphs (if you have multiple graphs, of course).

The only way iCloud could potentially work is if you mark the graph folder as “Keep Downloaded” on all of your devices. Maybe the same could work for Dropbox, though my experience is that it will still make folders online-only at some point after which the Logseq graph can get corrupted easily.

1 Like

I’ve been using icloud for months without an issue. But yes, you have to mark the folder as Keep Downloaded on all devices.

1 Like

I assume the GitHub, iCloud, Google and Dropbox cloud sync suggestions are a paid service?

I managed to set up a sync from iPhone iOS 18.1.1 to WorkingCopy app to GitHub to Desktop. However due to the confusion as per the post title, I am struggling to clean up the nested mess which I’m unsure how it was created.
Additionally, WorkingCopy app was good, however there were often bugs, potentially caused by iOS Automations that don’t complete before the screen lock in 30 seconds. I then find cleaning up git commits, syncs and the other tech jargon confusing.

Now that I’ve mistakenly deleted all my phone’s Logseq data from 20250336-20250409, I need to decide on my options.
1: paid sync, not an option
2: use another PKM, must be FOSS and maybe I’ll try Joplin although Joplin seems like more of a GUI app…which could be great.
3: set up Logseq phone and desktop again from scratch; unsure but I guess I’ll move all the .md files from before 20250326 manually to cloud storage.
4: set up Logseq phone and desktop again from scratch and don’t use sync, just manually move data from .md files when needed? Not an optimal solution though.
5: unsure of other solutions?

I’m a long-time Joplin user and relatively recent logseq convert. Joplin is excellent at what it does, which is provide I nice UI (desktop and app) for managing and editing markdown files organised in a folder hierarchy. It’s sync solution is quite solid, too.

The thing I’m loving about logseq though is how fluid it is to use - I just throw everything on the day’s journal page and then re-construct the blocks into something useful later. It has been a godsend for the way my brain works.

All the cloud sync services generally have a free tier of something like 5gb, so you don’t have to pay unless you start using it for other things (or have a crazily huge logseq folder).

1 Like

Yes, IMHO Logseq has potential, probably needing some user prompts to increase the uptake.

As an individual, I would be happy with a <15GB free option, similar to Google Drive, which is where I can fit all my individual stuff. 5 GB is too small (although perhaps text .md files are smaller?).
Enterprise would need >15GB storage and should rightly pay for the service.

Logseq sync costs $4 per month or $36 per year.
Apple Cloud is free <5GB.
Bitbucket is free <1GB.
GitHub is free <15GB.
Google Drive is free <15GB.
Working Copy app is free if registering when a student.

Might need a new post, but I’m clearing all Logseq data from phone iOS 18.1.1, Working Copy app, phone Files On My Phone, GitHub and local Desktop.
However, the local Desktop Logseq still shows some old data even though I removed directories for logseq. I’m confused where Logseq Desktop was storing the data? I did just run a Logseq Desktop → Settings → Advanced → Clear cache → and now the Logseq Desktop shows the new welcome page.

Do you know about the Refresh and Re-Index options?

Refresh scans the disk for changed files, and Re-index completely rebuilds the in-memory logseq graph from what’s on disk. Both are very useful when you’re making changes to files outside of logseq (including when you have syncing happening in the background).

This might be why I haven’t had any data corruption using iCloud - I regularly re-index, especially if I notice something seems to be missing.

1 Like

Thanks, I’m unfamiliar with the Refresh and Re-Index options, as I must tried ages ago and lost data. I just Cleared the Cache and that solved the making a totally new Logseq system.
I’ll have to research and invest time into the Refresh and Re-Index another time.