Screenshot of the group/tree structure in your DT/DN?

Newbie, my second post. I’ve enjoyed very much reading the threads in DT/DN Usage scenarios.
For the moment I’m doing an ultimate “cleaning & tidying” in my various folders in Finder/Documents prior to importing the most of it into DTPro. I’m looking forward to being able to check fast the contents of zillions of files laying around and later on - delete the unwanted from Finder/Sync with DT and so on. Writing and bookmarking/archiving is another scenario plus whatever idea comes later.

One thing that has always been the most difficult to me was the" tree/folder structure" in Finder - and will be my problem later on in DT. Grouping projects in their own folders and then putting all related files in them is easy, but the structure bellow that level, the hierarchy of subfolders is the main “problem”. I’m quoting the word problem because not all my files are related to projects, some stuff is just plain archive that I still don’t know how to group.

I’m wondering if it’s possible for you to send a left pane tree structure screenshot, the contents in the right pane are not needed, that could also be quite private so… I think that this would be of great help to me.

I have been to the DT Support pages and downloaded and checked out all example databases but still - living people/you users at this forum, that’s what’s even more interesting :slight_smile:

I’d like to provide a screenshot, but I can’t :open_mouth: I have 1,323 groups, so it’d take a while, even for just a second level :wink:

My advice, though, is to look at each topic and see how it’s categorized on Wikipedia. Not that it’s perfect, but it might give you good ideas.

For instance, Mythology > Greek Mythology > Gods and Goddesses.

Note that you might have:

Mythology > Greek Mythology > Gods and Goddesses
Mythology > Greek Mythology > Creation Myths
Mythology > Norse Mythology > Gods and Goddesses
Mythology > Norse Mythology > Creation Myths

… which means you could also have

Mythology > Creation Myths > Greek Mythology
Mythology > Creation Myths > Norse Mythology

and so on. Which I suspect might be your problem, because I’ve wrangled with similar issues myself. I have many things that should be categorized in multiple categories, since I’m working on a rather complex series of books:

Characters > Major Characters > Jonny McCharacter > Events > 12:34PM – Jonny enters the bar.rtf
should also be
Volumes > Volume One > Chapter 53 > Events 12:34PM – Jonny enters the bar.rtf

Personally, I dislike DEVONthink Pro’s smart groups. I recognize that they are extremely powerful, but at the same time I don’t like them since they do not update in real-time, since I don’t like replicants, etc etc etc.

So what I’m doing right now is just putting each document in one (1) group, wherever it seems most appropriate, and waiting (eagerly) for real Smart Groups or something similar in DTP2.

So, uh, I don’t guess this was very helpful. I suppose my advice to you is to just do it however you see fit, and don’t worry about having large large numbers of groups, because DEVONthink is very navigable. And if a particular group tree gets too deep, just drag it apart and send its parts to the root directory. It’s where it should be anyway, if you’re spending that much time on it :wink:

Hope that helps somewhat, and if you have any other questions, please ask.

Why don’t you like replicants? You can look at it as a form of tagging but then in a spatial and not a linear way. It can be used to solve problems where you want to file the same item in multiple groups but always referring to the same item.

I don’t like it because sometimes I change a whole lot of…

I think it’s about time I write one of them “Usage Scenario” posts or something, because this would get off-topic :slight_smile:

Thanks for the replies, good idea this with Wikipedia. Keep 'em coming, I’m really grateful :slight_smile:

I was intrigued by the suggestion about using the Wikipedia structure for organizing and went to Wikipedia and found their outline for categories here:

They break it down into different levels of complexity. However, as to be expected from an opensource project, not all categories are outlined yet :slight_smile:


R. Joe - Founder

Yeah :slight_smile: They have a heck of a problem to deal with – categorizing has always frustrated me immensely. Almost everything should be in at least five or six categories, it seems, and every category should be in five or six other categories. I didn’t really go into the problem that much above, except to mention that it exists.

I tried to write a “How I Use DEVONthink” article, to clarify my issues with DT’s smart groups and replicants, but I then realized it’d be hideously uninteresting and it might result in me driven from the forums for spreading ennui. So I gave up.

But I’ll try to make a long story short (and helpful) – smart grouping as it currently exists is difficult to easily control. I’ll list my reasons and give a couple examples, and hopefully it won’t run too long. I don’t want anyone to get the impression that I’m badmouthing the developers or anything – DEVONthink is clearly the best, which is why I’m using it :slight_smile: I like the existing system, but I would prefer it to exist as an option for advanced users, on top of a Journler-esque (ie, nestable) smart folder system.


  1. Insufficient power in existing Applescripts. As it is, we are offered only a few conditions – match by label, match by comment, match by content. Anything more complicated requires custom scripting and/or a personal markup language of sorts, which means you’re quickly spending more time organizing than using. Standard boolean operators are thus completely inaccessible to most people.

  2. Groups do not have adequate real-estate for properly specifying conditions. I have to specify conditions in the name of the group (hideous and difficult to properly organize), in the comment of the group (tiny text area which is resizeable but still uncomfortable), or in another file inside the group (!). Again, any sophisticated classification system requires considerable work on the part of the user to a) devise the system and b) make it workable.

  3. Lag. This is a minor issue, but some of my smart groups currently depend on the contents of other smart groups (see complaint #1). As a result, a smart group applescript needs to trigger another smart group’s applescript (which sometimes must trigger that of another). I prefer the “real-time updating” of the iTunes/Journler/etc set, especially since I am usually exploiting Automator and various other programs at the same time. The lag becomes annoying when switching quickly between groups, and I have no desire to have thirty-seven variations of the same smart group applescript that I’d have to change whenever I decided to revise my scheme (or, more likely, found a bug in my script).

  4. Difficulty of testing. As a newbie to Applescript, I’m constantly screwing things up. I generally have to rewrite a line, save it, update DT’s script menu, and re-trigger a group in order to get it to show me whether I’ve screwed up somewhere. This is because AppleScript, while powerful and versatile, is largely opaque to me at this point, even though I am looking for functionality that is already available in other programs.

As for replicants, I’m not sure I can express why I dislike them. For whatever reason, I feel less confident in them than I do in knowing that I have one file, with one checksum, with one name. Again, I can’t really justify this belief. It’s just a personal preference.