https://www.dendron.so/ logo
Join Discord
Powered by
Need help. Dendron prompted me to update my confi...
# questions
t
Need help. Dendron prompted me to update my config today so I allowed it and now am receiving the following error on startup and my tree view is blank and lookup doesn't work and I can't get enough info from the error or the logs to discern the culprit {"ctx":"ReloadIndex.execute","error":{"name":"DendronError","status":"unknown","severity":"fatal","payload":"{}"},"msg":"unable to initialize engine"}
k
it looks like its an issue with the config. in dendron.yml, can you replace the following "workspace": { "vaults": [{fsPath: "vault"}], }
s
I suggested the same, but they had the same error again.
k
ah, good to know. in that case, I would go also with your suggestion of using the backup 😅
s
They also said they couldn't find the backup 😅
I wonder if the upgrade crashed midway, that could explain a broken config & missing backup?
t
what's interesting is that that config file actually doesn't show any changes in 2 weeks. I tried reverting to the version in the repo but I am still getting the error
I've also tried removing and re-installing dendron with no change. I can initialize a new workspace with dendron and everything works without issue, so it seems the problem is in this vault
h
So if I'm understanding correctly, the config that you currently have worked prior to the upgrade? can you try reverting to 
0.86.1
and let us know if that solves the issue? This will help us pinpoint what went wrong and when the error was introduced.
@User You've also mentioned (or @User mentioned that you couldn't) that you couldn't find the backup file. Can you confirm that there is indeed no backup under 
{workspace root}/.backup/config
?
t
@User I can confirm there is no backup. I rolled to 0.86.1 with no change. I also tried adding the vault to a new and initialized workspace that I confirmed was working properly. Once I've added the vault in question, the new workspace stops working as well. I'm suspecting that one of my notes or something else in the vault is causing the issue and it was simply the restart and reload in the upgrade that caught it, not the actual upgrade causing the problem. I'm going to start going through the most recently changed files tonight and see if I can pinpoint an issue.
h
I can assist you with narrowing down the issue if you want. To start off, can you share the current state of 
dendron.yml
? If it contains sensitive information, you can DM me with the sensitive parts replaced. I'll try to recreate a minimal example of this issue with that.
t
@User nothing sensitive. I haven't made any direct changes so it should be pretty cookie cutter
h
That does look pretty normal. I'm wondering why the config update didn't create a backup. In this workspace, is Dendron currently not able to activate?
t
dendron appears to be active, I can use all the normal commands. It just seems that the index cannot load so anything related to it is non-functional
h
Can you see 
Dendron: Open Backup
command in the command palette, and if so does it list anything?
t
no
can't see it
h
Which version of Dendron are you currently on?
t
0.87.0 Rolled back last night to the previous version with no change so I just brought it back up to current
h
That is a bit concerning. After you brought it back up to current, did you reload vscode?
t
I did
h
Open Backup
command was introduced with 
0.87.0
so this should be present in the command palette. Otherwise during plugin activation something went wrong while Dendron was registering commands. This could be a clue to our issue here.
t
re-loaded again and It is there
h
cool. does it show you anything? or does it tell you there are no backups?
t
it found one backup from 3/26
version: 4 useFMTitle: true useNoteTitleForLink: true mermaid: true useKatex: true dev: enablePreviewV2: true site: copyAssets: true siteHierarchies: - root siteRootDir: docs usePrettyRefs: true title: Dendron description: Personal knowledge space siteLastModified: true gh_edit_branch: main commands: lookup: note: selectionMode: extract confirmVaultOnCreate: false leaveTrace: false bubbleUpCreateNew: true fuzzThreshold: 0.2 randomNote: {} insertNote: initialValue: templates insertNoteLink: aliasMode: none enableMultiSelect: false insertNoteIndex: enableMarker: false workspace: vaults: [] journal: dailyDomain: daily name: journal dateFormat: y.MM.dd addBehavior: childOfDomain scratch: name: scratch dateFormat: y.MM.dd.HHmmss addBehavior: asOwnDomain task: name: '' dateFormat: '' addBehavior: childOfCurrent statusSymbols: '': ' ' wip: w done: x assigned: a moved: m blocked: b delegated: l dropped: d pending: 'y' prioritySymbols: H: high M: medium L: low todoIntegration: false createTaskSelectionType: selection2link graph: zoomSpeed: 1 enableAutoCreateOnDefinition: false enableXVaultWikiLink: false enableRemoteVaultInit: true enableUserTags: true enableHashTags: true workspaceVaultSyncMode: noCommit enableAutoFoldFrontmatter: false enableEditorDecorations: true maxPreviewsCached: 10 maxNoteLength: 204800 preview: enableFMTitle: true enableNoteTitleForLink: true enableMermaid: true enablePrettyRefs: true enableKatex: true automaticallyShowPreview: false
h
Awesome. let me take a look.
So the error you mentioned at the beginning was caused because before the config update 
vaults
under 
workspace
was empty. This is usually not empty if a workspace was initialized and was never touched manually. Can you confirm that Dendron works as intended if you use the backed up 
dendron.yml
? with 
vaults
set to an empty list, I suspect that it will still have problems. But I'm assuming your workspace was functional before the upgrade so I'm curious.
t
that is what I just did and I can confirm that it did not fix the error. That version of the config file has been running for several weeks without issue as well
h
does it fix the error if you add in the vault? should be like this:
Copy code
...
workspace:
    vaults:
        -
            fsPath: vault
...
k
this could likely be it. you can find the bad note via the following process: Go into your workspace settings. Set [dendron.logLevel](https://wiki.dendron.so/notes/19a0ea9d-7292-4a68-bc6f-ffd462a54bc5.html) to 
DEBUG
. Restart Dendron. All logs should now be printed in debug mode. When you initialize your workspace again, run 
Open Logs
. The last note that Dendron loads - that is either the problematic note or it is right before the problematic note (the problematic note should be the next one as by alphabetical ordering)
@User I don't think this is a config version or related to upgrade but rather a malformatted note
t
"dendron.logLevel": "debug",
or 
"dendron.logLevel": "DEBUG",
?
k
ah, should be 'debug'
you might want to check the server logs which has more info -> https://wiki.dendron.so/notes/401c5889-20ae-4b3a-8468-269def4b4865.html#server-logs
h
It looks like Dendron isn't detecting any vault during initialization.
t
that's the impression I'm getting as well. Just not sure how to correct that. The vault is specified in the config file
Copy code
workspace:
    vaults:
        -
            fsPath: vault
k
i wonder if its a white space issue. can you c/p your entire 
dendron.yml
again with the updated vaults property?
t
I just managed to get a clean workspace to add the vault from my last commit and everything seems functional. My instinct at this point is just to move to the new workspace rather than chase this ghost. I can keep this broken repo in place for a bit if there is something you want to dig into from the dev side though
h
Glad you managed to get it working again. Is the repo public?
t
Actually, in an attempt to clean things up I tried creating another workspace, clean init, add remote, and now that is throwing undefined schema errors when reloading the index. I think there is a deeply buried problem that occured somewhere between my initial local vault and some errors pulling that repo to a new machine as it was syncing to two local directories on the new machine and that was creating conflict (the backup config was in the old directory for instance, not the new one I thought it was working in). At this point I think my best bet is to pull the indv. md and schema files into a new vault and start clean
The repo is not public, but I'd be willing to share it with you if you'd like
h
That would be great. post here or DM me a link any time 🙂
t
I'll send a link in the morning.
h
Take your time. Thanks for being patient with us.
28 Views
PreviousNext
Powered by