Outlining with unsorted DT groups

I am about to embark on a MAJOR, four-year non-fiction book project, and my plan is to put my DT notes into the order they need to be in for the book (by replicating the relevant notes into chapter and sub-chapter groups) and to use that as my working outline – by leaving the database unsorted, with notes and groups arranged as I need for the outline.
My question is, is this SAFE (given good backups etc) or is there anything that makes the ORDER I put my notes in less stable than the actual content of those notes. Imagine the disaster if I get all my thousands of notes sorted as required for the book, only to lose the order I’ve sorted them into.
Any advice/thoughts VERY much appreciated – including on other ways to achieve my outlining goals.

Although I’ve never had as many items as you in a group that has its document sort order set as View > Sort > Unsorted, I frequently do that when, for example, I create a group intended to display the individual sections or chapters of a writing project. That allows me to manually drag the documents into the order I want to establish.

I’ve had no problems with the stability of that order. I usually add at least one additional Sort order column to the view window in which I’m working, such as Date Modified and/or Word Count, that allow me to see which document(s) I was most recently working on, or which still have little or no word count. I can then return to my unsorted order by invoking View > Sort > Unsorted.

I suppose I could do something dumb to mess up my unsorted order. One approach to correcting that would be to use Tools > Restore Backup and return to a previous state of the database in which my unsorted order was still as intended.

With your many documents that you have sorted manually in the Unsorted setting, I would likely want to document that order for future reference. That can be done by selecting all the documents in a group, then copying the list to the clipboard. The clipboard content can then be pasted as a list of document names into a plain text document. Alternatively (if you have the space, as this could create a big file) pasting the clipboard contents into a rich text document will actually copy the documents (including their contents) into the rich text document, in your Unsorted order.

To make this even easier and more useful, imho, have a look at Make an Index Document (RTF) for a Selection of Documents where korm has provided a great script that creates an index which preserves the order and can be used to go to individual items.

These comments are from a more basic perspective than the very useful advice from Bill, above. I have to confess I would find it a bit of a nightmare to try to maintain order for files in a completely unsorted folder (though it appears that Korm’s script, posted by Allsop, would do the job). It seems counterintuitive to me, but each to their own, of course. I would wonder about the feasibility of maintaining disorderly order ( :smiley: ) for a four year long, non-fiction project, where your research and conclusions are likely to substantially evolve over the project’s lifetime. This will surely entail frequent shifting of notes and ideas around?

My second question would be: do you need to replicate the notes/files? If I were working in this scenario, I would have a miscellaneous folder(s) into which I would place notes/files when I am unsure of what chapter they will eventually go into. Tools > Show Groups and Tags provides a quick and easy way to move files within your database, when you do decide on their final placement. This will work whether you import files directly to Devonthink, or index them. I would also structure the database as simply as possible: with the database sorted by ‘kind’ (so that the Inbox and Tags are top), parent folders are sorted alphabetically. I would simply name parent folders Chapter 1; Chapter 2, etc., to have them sorted in the order you want. Subfolders nested within these Chapter folders could similarly be labelled Section 1; Section 2, etc., to achieve the same result within the parent folder. After that, Bill and Allsop have better solutions.

I respect the other suggestions, but I wouldn’t bet four years work on a hierarchy of documents that I manually structured in DEVONthink. Personally, I would use Scrivener for structuring my notes – which plays very well with DEVONthink for link-backs to research in my databases. But even before that, the first thing I would do is experiment with this workflow – whether in DEVONthink, or in some other tools – on a small scale for a month or so. I’ve never found it easy to maintain sort order in DEVONthink and with my fat fingers it is too easy to wreck the order by inadvertently changing View > Sort. Just my two cents. Test for yourself how it goes.

Thanks, all – It sounds pretty much as I thought: That my system might work, but have certain risks. (The reason I’m replicating, Kinsey, is so that I can have the original versions of the notes remain in the groups they started out in – for instance, separate groups for the notes pertaining to each source I’m reading – and ALSO have them live within groups that represent each chapter in my book, and within each chapter in the unsorted – or me-sorted – order I need the notes to be in, as the topics they address come up in the chapter.) Korm, I was hoping to AVOID having to master yet another program, such as Scrivener, but if you are right that Scrivener “plays very well with DEVONthink for link-backs to research in my databases” (not even sure what that means!) maybe I will try that. I agree that the risk of simply choosing the wrong sort command, and losing all my careful ordering, may be notable.
Sigh. Merging note-taking and outlining is far more complicated than it should be…

@Blake: in that case, then Korm’s suggestion regarding Scrivener is very sound. I work on academic book projects (and conference papers, etc) using Scrivener and DTPO in conjunction. Devonthink is used for what it’s best at; storing and sorting my research notes in a reasonably sensible manner. Scrivener is used to write these notes up. It’s a reasonably-priced, very powerful word processing tool that’s nonetheless easy to get started with on a basic level. No, I don’t work for them!
That would be my preference, rather than replicating every single note in another database. It seems to me that would be very time consuming. There are lots of threads here and on the Literature and Latte forum about using DTPO in conjunction with Scrivener. DTPO is excellent at what it does, but I would never consider using it as a tool to outline a major writing project.

Well, @Kinsey, I guess I’ll look into Scrivener, then. What I’m looking for is something equivalent to a pile of cue cards, inscribed with relevant notes from my sources, arranged into the order in which those notes will be relevant in writing. DT seems the obvious place to STORE those virtual cue cards, so now I have to find a virtual corkboard where they can be pinned up in the correct order for writing. I wonder how much of the content of each DT cuecard I will be able to “see” from inside Scrivener, as I put them in order for writing…
Guess I’ll find out as I explore.

:smiley: Then, as Korm points out, you’re looking for Scrivener. Good illustration of the cork board feature here: davidhewson.com/2012/09/11/quick … corkboard/