HI, I’ve read some handbooks and stuffs, but I really am confsed what should I do in my case.
I have lots of PDFs and Lecture Videos, which I annotate and study in ‘MarginNote’ App. I don’t use DT’s annotating tool.
MN doesn’t support DT’s location to bring up a file. It does allow external storage to monitor the files, but DT’s database structure isn’t allowed.
However, handling hundreds of documents and files, finder and MN both sucks at managing those files (tagging, custom metadata, etc…), so I chose DT to handle those.
Importing makes a copy, and I couldn’t delete the original one (MN wouldn’t recognize file in the DT database. It only recognizes in iCloudDrive/MN folder, so I should leave the original file to use MN ), so it would certainly make double the storage. So, I haven’t found any reason to import.
The questions are…
If I index the files, and the indexed files get lost or something, if I ‘re-index’ it again, does it recover the metadata or anything that was before? If so, I want to know the mechanism behind it. Does it recognize by comparing MD5 data?
If indexed, does search, AI feature work properly?
1.- No, once you delete an indexed file, and DT removes from his internal database metadata, if you put it back, it is a “new” one: new UUID, new annotations, new all.
2.- Yes, it does. But files must be local, you cannot have any cloud-only file indexed in DT. And managing a lot of indexed files have its own handicaps and you must be very cautelous about them.
Not an answer to your questions, just an observation on MarginNote.
I looked at MarginNote a while back as part of my “quest” to stay abreast of the best tools.
It’s a great app and I can see the appeal, however because it uses a proprietary form of annotations, I have avoided it. I try and and stick with PDF tools that use “standard” annotation formats, which tools like Adobe, PDF Expert, DEVONthink etc all understand and use.
As an alternative to MarginNote, I sometimes use Zengobi Curio, when I want to “spread” a PDF onto a visual canvas, and do similar things with it that MarginNote does.
Sure! And I understand the downsides of MN. It really is a big bummer.
For me, MN really is a exceptional allowance, since I highly prefer local, export-friendly, trasnfer-friendly apps.
Though MN’s feature to quickly make the structure of the subjects and its crossover liks with the files, video annotation tools….and much more, was the thing that I couldn’t give up. (Though it has many downsides…kinda laggy, closed source, poor support of the team…)
I also searched for substitute a lot, and curio was also one of them with heptabase, but it wasn’t matching my use case, since MN didn’t just match my usecase, but it expanded my usecase as I used it for a quite long time.
Oh! Good to know.
Then, if it’s not missing by deleting, but because ‘accident’ (maybe indexed file/folder location change), does it re-recognize by MD5 or something?
I just tested it out, looks like it recovers the metadata and DEVONthink tags.
If DT ends aware of the changes, normally movements of same file or folder while DT is running in the background, yes, it “moves” de data in his internal stuff.
Another trick for DT reinforcement, even in indexed files (and a must for no-indexed), is to launch the file from inside DT. For example, I have PDF Perfect as default PDF viewer and set double click opens externally in DT. That way, DT “annotates” that that file could be modified externally and is aware of changes. I don’t know if this is a thing that DT does specifically for indexed files, but a file modified when has been opened that way is less prone to generate duplicates or be treated as a new file.
And as cgrunenberg says, if you export, on import the files are “the same”.