Am new to DT and have been reading about Import vs Index in their help manual. Am yet to decide which approach to use. I am curious to see what other users prefer. Do you guys prefer Importing or Indexing and any specific reasons for choosing one over the other?
I import. Mainly because I use the same DT databases on multiple computers (syncing them) and they have somewhat different file structures.
I use indexing for some specific things beyond support but the vast majority of my data is imported.
On a modern computer there is so much room that I don’t find any point in indexing really. Including a bunch, maybe 50 sound files (MP3?) I have only about 6 GB of data. All my stuff though is text and pdf really.
I have looked at it, and maybe tried it some time ago but off hand I don’t think I have anything indexed at all.
I never index, just import.
Ages ago I indexed, until I decided a clean install with a new MacOS. That’s when I ran into problems getting the data files back using Time Machine. (Long story). I reinstalled the original configuration and then imported the indexed files. Once Time Machine had backed up the revised DT databases, I was able to do another clean install followed by restoring data files which included the DT databases with their imported data.
Now I have one database that indexes files associated with a few external apps like Obsidian. All other databases are imported.
I used to have a lot Indexed but I often got into trouble moving things in finder (you get broken links). I then stood back and considered what did I actually need inside and outside of DTP. The answer was surprisingly little outside. I now have most things imported and just two groups indexed (e.g. recipies with obsidian as the preview interface is simple for actual cooking).
Take time to look at each group of data (e.g. folder in finder) and ask “do I really access this frequently within another app?” If the answer is no, import it. Don’t forget even from within DTP you can open a file in another app. And also if you have imported and find it would be better indexed, it is easy to export and index.
It will take a bit of experimenting to find the best combination.
Import; I prefer having my data contained in the database package
and the auto-sync between my devices
There’s a few indexed folders for files that must be in the OS file system
I want my backups to include my files complete with folders and tags so I run a weekly export
edit; I tried full index; with no issues. However I’m concerned about a disconnect between the files and the DT metadata
The key considerations:
Do you want to synchronise documents between devices while preserving document links (e.g. links back to document in Annotation files)? If so, Importing is the only way to preserve these without manual intervention (e.g. searching and replacing a document title manually in x-devonthink links).
Do you want to make documents available to the Finder or third-party filesystem apps (such as Default Folder X or Hazel) as opposed to working on them in place in their in-database location? If so, you’ll want to go with Indexing.
Do you want to have the same editable file available in multiple databases? If so, this is a job for Indexing.
A handful of tips I’ve found useful in striking my own balance:
With very large databases that you need to work with on different MacOS devices, Indexing may paradoxically be more efficient than Importing. With ChronoSync + ChronoAgent, syncing a large folder is blisteringly fast, and then you can simply run the Data => Update indexed items of all databases script to update the indexed database(s) on the other device. (But document links will be broken.)
DTTG loves Imported files. If you’re syncing between desktop and iOS, this is usually the way to go.
A lot of users, myself included, prefer DT’s file management to the Finder, at least for the kinds of document we work with in DT databases – but not everyone may be ready to dispense with the Finder altogether, particularly for things like rearranging documents within an existing file hierarchy (where dragging documents between a bunch of open Finder windows is hard to beat as a UI). If that’s the case, stick with Indexing for now.
Because I have a main desktop Mac but also need to sync to a laptop and an iPad, I tend to end up using Indexing for large databases and Importing for smaller ones, the latter comprising mainly active, heavily-synced working documents with live document links. I’ve got used to manually search-and-replacing document addresses in links when I move a finished-with working document and its associated Annotations and other satellite files out of my Imported syncing database into my Indexed database, but that’s not a workflow that will appeal to everyone, and DEVONtech are right to recommend Importing as the default unless you have one of the many perfectly good reasons not to.
I import everything with the exception of two specific folders that are maintained via another app and I just require to be searchable in DT.
I mostly import for the same reasons as others have listed here. DT is my file management system (it’s more than that!) - it’s easier to just have everything inside my databases where I want it. The two folders I index are carefully named and I know they’re indexed. Their content might be looked at in DT, but it’s never edited in DT as another app is responsible for them and I’m careful.
Which other app, if I may ask?
You may. The other app is Notenik, which is a free markdown app. I love Notenik, and use it for two specific tasks that it is excellent for (it has other well-developed use cases, which I don’t need). It manages my collection of quotes, and it manages my collection of poetry. Each “collection” on Notenik is a folder. The files are Markdown, and I use Notenik on iCloud, so DT is basically indexing two iCloud folders of markdown files.
Notenik is Mac only, so I am basically just using DT as a funnel to bring the files to my search on DTTG. If there was a Notenik app for iOS I wouldn’t bother using DT as the go-between, but the developer has been clear that a mobile app is not needed (and I agree it isn’t). I only ever create new files in these collections when I’m on my Mac, in Notenik.
My quotes/poetry collection is stored/organized in Devonthink
What are the advantages for you of using Notenik?
Kudos for the folder structure, allowing for indexing by Devonthink
Well it is personal taste. There were two points for me:
The interface for writing an entry (= 1 markdown file) can be customised by collection. I thus have an interface set up with the “default” fields I want for each of my collections, e.g. an author field, tag field, title field, quote field, etc. All of this writes to the Markdown file.
You can sort by different criteria. I had previously tried to manage my quotes in Obsidian, but sometimes I need to sort by author, sometimes by title, sometimes by tags, etc. Obsidian struggled with this. Notenik can sort by different fields.
That’s an interesting approach to document/Markdown composition.
I used to import everything, until I realised that, by indexing my Bookends attachments folder, I could keep that folder synced across two machines (via a DTP Sync store). Syncing that folder this way has advantages for integration with Bookends (compared with using Synology Drive to sync) in that Wikilinks in DTP (from importing references from Bookends) can link to the corresponding Bookends attachments.
As a much less expert user than many others on this forum, I will suggest that although linking can be very useful, it is more important to use it carefully.
Index. I store almost all files in Dropbox, I have 3TB storage. I used to import but often I want to access the files on different devices so having them all in one place is easier for me.
I don’t use Bookends - I had a go but the Bookends developer told me it would struggle at 10,000 files so for pdfs I use Finder and my own ordering system. I do use Scrivener for some short writing, for example, my journal (in which incidentally I have info from BCE so as to position my place in the history of the world ) but mostly for long writing Mellel.
I used to import, but now I index everything. I prefer to have my files sitting in the Mac file structure, accessible and backed up from there. I love working on the indexed files from within DT and using its capabilities, so I get the best of both worlds. However, indexing does need some experience with DT in order to handle the occasional missing file links and ensure everything is in sync.
In the “best of both worlds” category; the indexed files are sync’d and can be accessed with my iPad
I know there’s overhead with indexing for the Devonthink app but it’s handled so well
I’m wondering if there’s additional negatives