Hello, I am hoping to get some feedback, answers and tips on my setup and wondering if anyone else has a similar setup:
All of my relevant data is stored on folders on my desktop which is synced with Google Drive. The intention is for all three locations -DT, Mac/desktop, and Google Drive - to be in sync. To start, I used DT (Pro) to index my desktop so that it initially formed 1 database, which would allow files to remain in sync with Google Drive. From here, I began to create a few separate databases to help DT self-organize based on specific folders/files (as I read that having separate databases containing relevant information is better compared to one large database containing a variety of information). I also started the process of tagging files and folders (with the “exclude from tagging” box unchecked because my folders were already organized).
My goal is to be able to effectively use and optimize the “See Also” feature to make connections within the various .pdfs, .doc, .ppt, .xml and .epubs within my files, specifically by reading the content within those documents.
Questions:
Is this the correct general approach? Should I be doing something different?
It is my understanding that databases are internal to DT. I maintained my original folders on my desktop however when I created new databases for specific folders and clicked the “update index items”, my desktop folders synced with the newly created DT folders, which I don’t mind, but want to understand what exactly syncs and what doesn’t.
Following the above, I now have separate databases for the few chosen folders on my desktop but I also have a “Desktop” database which contains all the folders on my desktop, including the newly created databases/folders. The newly created databases/folders sync to the Desktop database (and the Desktop itself). So, it’s fairly duplicative, although weirdly DT only identifies a few files as being duplicative from this setup. Should I be doing anything different?
Is there a way to remove files off the Duplicates list?
I’ve read that I don’t have to tag everything as DT will learn so is there a recommended tag amount e.g, 80% of all files?
All of my relevant data is stored on folders on my desktop
Why?
Question 2:
which is synced with Google Drive.
Why?
Is this the correct general approach? Should I be doing something different?
See above.
It is my understanding that databases are internal to DT. I maintained my original folders on my desktop however when I created new databases for specific folders and clicked the “update index items”, my desktop folders synced with the newly created DT folders, which I don’t mind, but want to understand what exactly syncs and what doesn’t.
Note, terminology is important. You don’t sync the database with folders in your filesystem. You can index and update them, but sync is a very specific and unrelated thing.
Following the above, I now have separate databases for the few chosen folders on my desktop but I also have a “Desktop” database which contains all the folders on my desktop, including the newly created databases/folders. The newly created databases/folders sync to the Desktop database (and the Desktop itself). So, it’s fairly duplicative, although weirdly DT only identifies a few files as being duplicative from this setup. Should I be doing anything different?
It is generally not a good idea to index the same location in multiple databases. This is often a cause for things to seem like they’re appearing or disappearing in a database, especially if you’ve forgotten you’ve indexed in more than one.
Is there a way to remove files off the Duplicates list?
No. You can enable Settings > General > General > Stricter recognition of duplicates but that will make no difference if you’ve literally indexed duplicates.
I’ve read that I don’t have to tag everything as DT will learn so is there a recommended tag amount e.g, 80% of all files?
Tagging, just like many things in DEVONthink is optional. There is no recommended amount.
Back up for a second and clarify why you’ve set things up the way you have, including why you’re redundantly indexing Finder folders. That way we can better understand and assist with your situation.
That doesn’t answer anything. Why are you syncing this data in the first place? And why is Google Drive your preferred solution?
Also, it is not clear if by “desktop” you’re referring to the computer itself, or your ~/Desktop folder.
You should not use your Desktop folder to store all your files.
You should not index your entire hard drive
It sounds like you’re new to DEVONthink and don’t really know what you’re doing. Take a step back. DEVONthink has a fantastic manual that you can access through the Help menu or download as PDF here. At least read the Getting Started chapter – and since you want to index, the section In & Out > Importing & Indexing.
DEVONthink is very flexible and can accomodate many different setups and workflows. It really depends on your goal(s), the nature of your work and what files you work with. Without further details it’s hard to say much.
I recommend you start out small and get to know the application.
Before you go crazy with tagging, read about indexing in the manual. Tagging imported files is different from tagging indexed files. Not all cloud solutions work with Finder tags, I’m not sure if Google Drive does.
If you need to use Drive that’s fine, but the setup seems extremely bizarre. Why are all the files and folders “on the desktop,” and not just files and folders IN Drive? Why index the desktop? If you want to index files in Drive, seems to me a better solution would be to just put the files there and index whatever folder(s) you want.
My advice would be to keep things as simple as possible, and what you describe sounds anything but simple. I’ve tried complex setups in the past and they are usually a recipe for chaos. Not only that, but they tend to be laborious to maintain. These days I just have one database and if I want to find something I use the search facility. I find that lots of folders can be more confusing than helpful at times and I’ve trimmed them down quite a bit recently.
Using the terms “Indexed” and “Google drive” in the same sentence when discussing DT can be pretty risky unless you fully understand the implications.
If you want to use DT for your data- and there are great reasons to do so - it is usually far safer and easier to put everything in DT and then use an approved method to sync data to a DT sync store. That sync store can be located on Dropbox, iCloud, CloudMe, a local WebDav server such as a Synology NAS Drive, or a web-baed WebDav server.
Then all of your computers/devices can in turn sync to that sync store.
Given the confusion, I guess most aren’t familiar with Google Drive’s capabilities - it is a cloud storage solution that also provides a downloadable application that allows you to sync your desktop (literally marketed as “sync your desktop with Google Drive”). However, “desktop” is just one example and files can be uploaded from anywhere, much like any other cloud storage solution. Regardless, I think the focal point here should be on how to properly index GD rather than what location is being indexed?
I suspect almost everyone on this Forum understands Google Drive’s capabilities well. Indeed I am a fan and even have a Google Workspace account for my business.
The point is that there are some specific use cases where indexing works really well with DT but rarely is indexing a good solution for your core critical documents. There are various scenarios where indexing can lead to loss of data and/or loss of metadata in DT -and you need to understand that well before indexing any critical data. Unless/until that point I suggest a solution that does not require indexing.
Tagging indexed files in DEVONthink propagate those tags to the original files in the Finder.
Regardless, I think the focal point here should be on how to properly index GD rather than what location is being indexed?
I will disagree with this assessment as you have not addressed my response.
So again… It is generally not a good idea to index the same location in multiple databases. This is often a cause for things to seem like they’re appearing or disappearing in a database, especially if you’ve forgotten you’ve indexed in more than one. â€
So why are you indexing the desktop in a database, then individually indexing folders from the desktop in other databases? This is not a wise idea and I indicated it is not the “correct general approach”.
PS: There is no correct way to index Google Drive. It’s not that specific.
†: And yes, I know how this can lead to confusion from personal experience.
Please provide more details. We understand you want to index files from Google Drive. But there is no single answer here. The best ways to do that depends on the context and your goals. Not the goal of “index Google drive”—your broader aims.
…and if a Finder tag on an indexed file is removed outside of DEVONthink, that is reflected in DEVONthink.
Finder tags are a type of extended metadata unique to the filesystem used in macOS and iOS. If a cloud storage solution is not compatible—many aren’t, since the server storing the data doesn’t run macOS—or if it syncs with a computer that doesn’t run macOS, you risk them disappearing.
To summarize (where “>” is a database and “>>” is a group/folder within the database), I first indexed everything via one large Desktop database e.g.:
I then created new databases for the groups because I was under the impression that DT’s AI, search and “See Also” functionality works best if databases are separated by information type (e.g, Projects, Research, Books) rather than having one large “Desktop” database containing all the information:
>Project
>Research
>Books
Because the newly created databases are the same as some of the groups in the Desktop database, I am at the point where I am deleting the groups in Desktop so that I am left with:
…while the actual folder organization on my desktop and by extension, Google Drive, seemingly remain intact and synced. Eventually, I plan to separate out all the groups so that the one large “Desktop” database is no longer needed.
If there are potential issues with this, I can delete those separate databases and go back to having the one large “Desktop” database.
I am at the point where I am deleting the groups in Desktop so that I am left with
Were these groups indexed individually or the entire Desktop folder was indexed into the Desktop database?
Deleted where?
There is no need to have one huge database with the Desktop contents.
But there is also no need to have one database per Finder folder.
Both are acceptable.
It’s just advisable to avoid indexing the same Finder folder in multiple databases.
PS: See Also & Classify can work across all open databases or just the current one.
