All my vault checking runs on Dataview queries I mostly copy pasted from around the internet. Orphan notes, dead links, that sort of thing. They work, kind of. But I barely understand half of them, so when one breaks I’m lost. Tried moving a couple to Bases and it felt cleaner for like five minutes, then I hit the stuff that just doesn’t translate. Block level things mainly. Now I’m wondering if I’m trading one headache for another.
Anyone done the full switch for this maintenance kind of use? Did stuff break? Was it worth it?
Hi, again
. Can’t speak to the headache as I don’t do vault maintenance; still, sorry it’s been a bother for you. As for comparisons, nothing would break per say. You’ve already encountered that Dataview can do stuff Base can’t, some with tradeoffs. Would explicitly discussing the limitations help you decide?
In most situations, Bases
- is more efficient,
- is more user friendly to make,
- and lets you edit properties from the base;
- but it can only ever return one existing file per row/card/line,
- and it cannot read regular content in notes.
Dataview falls behind in all of those expect the last two:
- it can display results in any number of pleasant formats and groupings,
- and it is excellent at reading some in-note things—less excellent at reading others, but usually possible.
For orphans[1], Bases takes the cake.
For zombies[2] and really all ghosts[3], you’re at a mid-ground. Both can do it. Dataview can display one zombie per row along with the number of sightings and locations, making 'em easy to corral and pick off. Bases can only list every note with a zombie alongside that note’s list of zombies; counts are challenging to make (and a bit useless when made), and groups are awkward when any note has more than one zomb/ghost.
Personally, I would take Bases ghost hunting with me and just put up with the awkward display. But it’s your call.
I’d be very interested to see a list, to give me some ideas.
I’m only starting out with Obsidian, but think I can migrate from both OneNote and Todo using Bases, the query statement and the Tasks community plugin.
The documentation is all excellent and (if you aren’t dead set against it) a good AI chatbot can help you with specific queries.
This is exactly the comparison I was hoping someone would write. Thank you. The one file per row limit is the part that worries me most. Most of my audit stuff wants one row per problem, not per file. A note with three dead links showing up once with a blob of text after it is workable but ugly. Your zombie counting example nails it: doable in Bases, but awkward enough that you feel it every time. I think you have talked me into the split you describe. Bases for the orphan sweeps where it wins cleanly, dataview where I need per item rows and counts. Running both side by side feels wrong somehow like admitting defeat but it might just be the honest answer
Choosing the right tool for the right task is never defeat!
welcome! good timing doing this early by the way. way easier than untangling a 5 year vault later, ask me how i know lol.
short list id start with, roughly in this order:
- orphan notes. nothing linking in or out. these pile up so quietly.
- dead links pointing at notes that dont exist. renames cause most of these.
- notes with no tags if youre gonna use tags. catches the ones that slip through.
- notes missing dates or titles in frontmatter if you use those downstream.
honestly just the orphan check alone is worth it in month one. rest is gravy. are you doing these in bases or dataview? happy to compare notes, im mid switch myself.
Haha fair, I’ll take that. Right tool for the right task it is. Bases for sweeps, Dataview for the detailed stuff. Defeat officially not admitted.