That’s also my use case. I also do not recommend this setup 
1 Like
The one caveat I can think of is that the fact that I have continuous backups in BackBlaze and weekly backup to two external drives. I’m not an expert on these mattes but am I right in assuming that given my backup regimen that even if DT corrupted a file/files indexed from iCloud I’d have good backups given that there are multiple versions of the backups across both Time Machine and Backblaze?
DEVONthink not likely to corrupt files. Onedrive with their changes to keep files only on server and delete from local drive (by default) will likely lead to problems as DEVONthink will not be able to find the deleted files. can change how files sync in Onedrive but many people are oblivious. Onedrive does all the syncing—not DEVONthink.
Obsidian now includes unsupported file types in the file tree view, and you can link to them as if they were markdown or other supported file types. That is definitely true for Microsoft Word documents—I have dozens in my Obsidian vault.
1 Like
Well, that sounds like good news. I don’t use Onedrive except for the occasional Word doc or Excel spreadsheet and never for research. I may be fine to trying again to index my research folder with DT.
where are the files to be indexed? local drive? networked drive? folder that is synced by other software eg dropbox, google drive, synology drive? and is there a specific reason you do not import? (indexing adds complexity but as long as you take the advice in the “DEVONthink Handbook” will be ok.
As I noted above, my files are stored in a research folder in iCloud.
The vast majority of the files are PDFs or markdown text files.
The reasons I’ve historically indexed:
- To avoid having files in two places (if I import AND keep the files in iCloud). I prefer not to have the files residing only in a proprietary application even though I know that I can convert and export from DT if/ as needed. I’ll add that while DT’s iOS app is serviceable, it is not great. When on my iPad I’d rather use Obsidian, iA Writer or other application to access the files.
- So that the files are accessible using any compatible program on any of my devices, e.g, Obsidian, iA Writer, etc.
I realize I can open a native app from within DT to open a file residing in DT but I’d rather do so by using the application directly rather than going “through” DT to do so.
That said, I’m always open to hearing if my process needs improvement or if there is a compelling reason why I should import rather than index my files.
Happy I’m not the only person using Devonthink like this!
I rarely interact with the DT UI, and I don’t use it for search.
I use it as a background utility which automatically creates PDFs from RSS feeds and files them; OCRs scans and files them etc.
DT’s background automation functionality is so powerful it remains really useful to me, even though I rarely interact it ‘directly’.
These are pretty valid and I find myself in the same situation with DT. I checked EagleFiler today and it seems like a “lite” DT which could work for most of my use cases but I have not found the capability to index folders external to its database, it gulps everything into its database --which is a regular filesystem structure, unlike DT’s heavily indexed internal folder format.
Yes, each library = one folder (and subfolders) - there’s no concept of external folders. It might be possible to use symbolic links, but I’ve never tried. The developer is very responsive on the forums.
You can, of course, have multiple libraries.
I store my files on iCloud, which seems to cause DT all sorts of issues; no problems so far with EF, possibly because it uses a different method to maintain its indexes.
I would not put it that way. They are not “gulped into a database.”
When you “import” them, your files are copied or moved (your choice) to the root folder that you selected when you set up the EagleFiler library. But the files themselves remain unchanged and keep their original names and file types. They can be manipulated and edited with external apps.You can use the EagleFiler command “Reveal in Finder” at any time to see where they are stored. There are no package contents to reveal. No UID-type filenames are assigned.
The only database involved is the one that stores the structure of the index that EagleFiler creates for you that gives its excellent search capability. And this index is completely separate from your files and could be rebuilt if necessary at any time from those files. EagleFiler has a simple and straightforward app architecture that I find to be very useful.