Thank you all! Much to digest and I will reread the thread and try and use the various ideas here.
As of now, I have indexed my Documents folder - which is mostly documents.
I know this was advised against, but this is the way I could start using DT3.
Things that are working well:
The good thing that DT3 seems to do is if I move an item from Inbox into a indexed group (which is a Folder in Finder), it creates the item as a file in Finder structure. So my indexed structure gets replicated on Finder. I know I am potentially giving up on some capabilities of DT3 but this seems to make sense for now. I wonder if there are others who use DT3 by indexing their Documents folder
The linking items that can be accessed on iOS and Mac is very promising.
Things that I need to figure out:
Documents folder (which I have indexed) does have files and directories that I would prefer not to index in DT3. Based on search on this forum, it seems there is no way to have DT3 ignore certain files/directories or better still path patterns. If there is, would be great!
I have some directories where I keep receipts and invoices, and then another subfolder has quickbooks files. Would be useful to index the receipts, invoices etc in DT3 and not index Qbooks files. Or there are folders which has Python scripts, would like to index the python script (because would like to search them for future snippets), but not the virtual environment directories.
iCloud sync is not quite satisfactory. Need to figure out another way to do this sync.
I wouldnāt recommend to rely heavy on indexing. Indexing is a kind of trade-off between your file system and DT, and as such it has some limitations.
Where I use indexing:
Searchable external archives
Shared cloud folders
Files whose location is used by some system functions and you are not recommended to change it
If I am not going to use any of above, I import, not index.
Good practice for me is to index top level folders in Inbox (e.g. shared Dropbox folders). Their structure and contents may change for some reasons, but āUpdate Indexed Itemsā works fast and nice to update all changes. I keep this structure in Inbox, to have the same structure with the other participants. But if I want to see some files and folder within the DT group structure which is logically closer to me, I just replicate them to any comfortable location.
What is very handy is that you may have files and folders in this shared structure, which are not visible for others (participants of shared folder) except you. If you want to show it you use āMove to external folderā - it turns this file to indexed file, removes from DT storage and places it to the the location where you see it in folder structure. If you want it back, or close access - you use āMove into Databaseā, you still see this file as if it was there in the shared folder, but it is in DT storage now and other participants donāt see it any more.
And donāt forget, that if you decide to move your files out of DT it will restore all the folder structure as a corresponding group structure. Nothing to loose.
You may choose any other sync store, if iCloud is not good for you. I use iCloud and donāt have any problems.
May be I need to create more of a distinction between files that need to be accessed through the Finder / terminal interface (scripts etc), and documents that I can launch through DT3 interface.
And if I do this later thing, then I need to redo some simple Hazel rules and Hazel launched scripts that I currently have. And I need to develop new muscle memory for classification as currently I either move files or more likely use Launcher to move files to a suitable folder. I am guessing others have already done so and need to lean on what people may have shared in this forum.
Well then I will stick with iCloud. The only thing that may be causing it to not work for me is the number of files. I have not checked but I am guessing there are a lot of files from the scripting directory virtual environments and some non documents such has Quickbooks / Quicken files.
I guess so, though you can open any file from DT3, if it canāt open itself it uses external app (uses systemās default), like MS Word or Script Debugger.
Iām sure you may do more complex and handy things with DT, than with Hazel, as for what DT is for - documents stuff.
I have 4 databases with about 35 Gb and over 40 000 files. Didnāt have any problems yetā¦ iCloud has its own timetable for what and when to upload, yes, it may be not as strait forward as dropbox, but its much cheaper and is used by my ecosystem as a whole (photo, video, backups etc), so Iāve chosen this option.
When all of Documents is indexed, then you are essentially using DEVONthink in place of Finder. And so the next question is, āwhy botherā?
I have used DEVONthink for a very long time, and always by creating databases that are either focused on specific projects, or databases covering a broad topic area. But a database that is nothing but everything in Documents would, I think, just muddy the picture.
Iām late this this topic ā but the original question, āLooking for use cases where DEVONthink is evidently superior to using Finderā¦ā seems to me an unanswerable question. And, with respect, I would suggest that if anyone needs to be convinced to use DEVONthink ā or any other software for that matter ā especially after having experimented with it ā then reconsidering āwhy DEVONthinkā might be helpful.
In other words, if you need software ā if you need DEVONthink ā then a bit of time trialing it will make it obvious. If the āuse caseā is not obvious, then donāt use the software. Iāve seen lots of folks buy DEVONthink or Tinderbox or Scrivener, or other complex software, often because some blogger suggested it, and then struggle for weeks and months to figure out what to do next. That kind of stress isnāt necessary. The best thing to do if the tool is not obviously useful is to go with your gut instinct and skip the purchase, or get a refund.
Agreed. But scripts need to sit in a folder for navigation via command line and virtual env etc.
Need to better figure out what to simply import and what to index.
And need to think through and get inspired by how others are creating the structure for different databases.
But you find the sync slow? I think its pretty fast and Iām using iCloud.
Considering all the uploading and downloading DevonThink3 has to do in the background damn right itās fast.
After being quite skeptical about DTās benefits, Iāve been using it on a project which I had started with the files in the Finder, mostly searching with HoudahSpot, and Iām finding it much easier going with DT. Itās a small project - about 150 files, about 40 pieces of old legislation and 110 court cases involving them, but a lot of the files are big OCRed PDFs from the early 20th century, with hundreds of pages of cruft and a few pages of interest. The search UI and OCR help but mostly itās the search speed - which is basically instantaneous - with a small group of docs.
