I’ve seen that this question has been asked before over the years but not finally addressed. I find it a bit cumbersome (and unnecessarily so) that indexation is done with absolute paths.This means that the indexed folders must stay put or else the indexation and tagging etc of them in the DT database is lost.
Why not add ability of relative paths for the indexation?
sorry for being unclear: I would want the indexed item to have path relative to the database. So as long as I keep DB and item ‘together’ I can move them around.
Ah, I see. (I think. 
)
This actually works now. If the database and the indexed files are at the same relative location in the file system from the root of the current user on each machine you use, then the indexed files will be available on each machine where you open that database. A simple example:
~/Documents/Workfiles/Databases/myDatabase.dtBase2
~/Documents/Workfiles/Indexed Files/…
In this example, if your indexing occurs at or below /Workfiles/…, then the indexed documents will be available on each machine that’s set up this way and has instances of the database and files. (The point of this example is that “Workfiles” and its children can be copied among machines and the indexed structure will stay intact.)
Same is true for indexing material that is located inside Dropbox, Box, Google Drive, Creative Cloud … any cloud … if the parent for those folders is at the root of the user on each machine.
If you look at Tools > Show Info you’ll see that DEVONthink displays the relative path:
Aha!
Sounds good! My problem though is that the root location of DB and items is not the same on each machine. I’ll make sure to get it right from the beginning in the future but right now I have a 2800 item DB and the absolute path issue destroys the tags I’ve added to quite a number of the indexed files :-/
(Unless you know of some ingenious way of saving the tags?)
Thanks for the help!
If you feel like investing the time, it’s possible to move things into the database (thus preserving the tags) then export them to a new location (still preserving the tags) and then index that new location. Clunky, but workable.
Greg Jones has written about this at length in the forum – just don’t happen to have the citation at hand right now, but you might want to search for his articles on indexing.
I would definitely prefer export/re-import that to re-tagging 1700 articles! I tried to to find Greg Jones’ posts about this but no luck so far. If you find a link/reference that can help me locate it that would be by much appreciated!