Database setup ok?

If possible, I would like your opinion on this setup. Would it lead to local url’s being overwritten?

DTP + Sente

  • Two databases in DTP: database1 with all full text PDF’s, database2 with notes, quotes, inter alia on the full text articles in database1. That full text database1 can thus be excluded in order not to distract the AI system too much, but if I want to search it, it would still be possible.
  • The notes in database2 contain wikilinks to the PDF’s in database1
  • The notes also contain a link to the reference in Sente
  • The reference in Sente refers to the PDF in DTP through a local url

I got inspired by the Paperless Academic Workflow article but I think there’s something unlogical in my reasoning.

Then one extra question: what is wrong with putting all your documents in the Finder and letting both Sente and DTP index from there? This was suggested in a post by mbywater I read somewhere (see below) but was never answered. What’s the disadvantage with that approach?

I use DevonThink Pro and Sente. There's a good trick for using Sente to manage notes on books. Sente will link an indivdual reference to, among other things, RTF files. Keep a "stationery pad" RTF document on your desktop. Each time you want to take notes on a new reference, drop that document onto the reference in Sente.

If you've set Sente to manage attachments, it'll creat a new copy of the RTF file, name it with the book title (or author/date, or whatever you choose) and store it in its attachments folder (~/documents/Sente by default).

Now all you do is ask DevonThink to index that folder. Lo: your notes are accessible in Sente, indexed and searchable in DT, editable in whatever you choose (I use Bean for that sort of stuff) and everything is always in sync. Sweet.

If you're in this for a longish haul, I'd spring for DT Pro Office and a scanner. That's what I'm about to do, simply for this good workflow:

(1) Find book in library

(2) Photocopy relevant pages

(3) Scan photocopied pages into DevonThink

You now have text/pdf files you can do what you want with. Very handy.

Are these file URLs or x-devonthink-item:// URLs?

The main disadvantage is that indexed files are not yet editable. In case of PDF documents that might be no problem, e.g. you could use Preview/Skim for this task.