I’m looking to use DT as my reference library in a way that complements my use of Zotero as a citation manager. After looking into this, I’ve settled on the current workflow for journal articles and the like (PDFs):
- Add references to Zotero - metadata and full text PDF
- Use Zotfile to rename and move them to a local subdirectory of my Dropbox folder
- Index that Dropbox folder in DT
I’m wondering if there are any special precautions that I need to take in handling those indexed items in DT.
I’ve read the Taking Control of DEVONthink book and the Importing & Indexing section of the DT docs (p 51-54) and played with these features enough to get a feel for things. Based on that I see no reason to avoid moving indexed items into other groups, tagging, replicating, labeling, etc. them, but want to be aware of any practical gotchas that I might be overlooking.
The extents of DT’s “tight integration” with the filesystem seem a little fuzzy, depending mostly (as I understand it) on whether the item was indexed as a file, in an individual folder, or in a folder hierarchy. This makes me a little uncomfortable.
I’ve searched this group and what I’ve read seems to support that. Just fishing for gotchas, I guess…
I appreciate any advice, recommendations, best practices, etc. you are willing to offer, or links to the same.
I use a similar system, but I use Paperpile, which saves the pdfs in Google Drive instead of Zotero and Dropbox. Everything works very well. I believe your system is good enough. You can do most of your academic work in DT and use Zotero only for references.
This makes me a little uncomfortable.
While it shouldn’t make you uncomfortable, you should approach indexing with caution and thoughtfulness.
Also, you should be aware you may have to use File > Update Indexed Items with the indexed group(s) as (1) the Dropbox sync may not initiate a filesystem event, and (2) DEVONthink tries to avoid creating any conflicting situation, i.e., getting in Dropbox’s way.
I absolutely agree re: caution and thoughtfulness. Hence the reading and my initial post. What makes me “uncomfortable” is some language in the related documentation that is not 100% concrete. For example,
If you rename in the Finder, the change should be automatically reflected in the database. DEVONthink tries to keep the group in the database and the folder in the Finder in the same state.
I would have appreciated more confident terms here, or at least some discussion of the limitations / concerns to expect.
Also, I would have benefited to more details and examples in your description of how DT’s behavior with indexed files will vary for individual files, folders, or hierarchies.
I understand that there is some complexity here, that the underlying system and dependencies may not be consistent, and that documentation can never answer all questions. Just trying to get this workflow mostly right from the start, or at least limit future setbacks.
With regards to the Dropbox sync concerns, if I move the DT records out of the indexed group (to file it by topic rather than source) and the references are subsequently broken, do I need to use File > Update Indexed Items on all the scattered references? Or can this be performed on at the database level? Or should I create replicants of the indexed items for filing, so that I can Update the original index folders in that event?
Thank you for your presence and responsiveness in this forum, and the dedication that it demonstrates to this tool.
As discussed in the Help > Documentation > In & Out > Importing & Indexing section, you should index locations that are fairly static, e.g., not moving things outside the indexed parent folder.
You could remove the indexed parent from the database and empty the Trash, removing only the database references.
You could instead index the locations you’re moving items to, e.g., the topic folders you mentioned. I would personally suggest this action.
However, you should read the aformentioned section to understand the filesystem ramifications.