Dear DTP community
I am using DTP since years with different databases for work and private life. I am sorting new documents into groups of the databases, sometimes using replicants if they would fit into different groups. For new documents I use the auto-classify feature, which is quite helpful. But I do not use tags at all. I know some (or many?) of you use tags extensively. I have never used them in the finder, and do not use them in DTP. What is your workflow and use of tags? which benefit do they offer in practice? Would be really interested to learn from you.
There are certainly very sophisticated methods of using tags. I only need them for two simple things that help narrow down the search.
Letās say tag:energy and tag:Europe
Documents on these topics are spread across many different groups. So I canāt just select one group. But with both tags I can find all documents, wherever they are in the database.
Or if I search for a specific āreportā. If I only searched for this word, I would get too many results with all kinds of reports. So I search for the word āreportā with the tag:energy.
As I said, this is trivial. But thatās enough for me. Iām sure others use tags much better.
2 Likes
54+ years of daily diary entries safely ensconced in DEVONthink, grouped by year and date within year and most containing links to other entriesā¦
Significant dates tagged, Humour tagged, Holidays tagged (sub-tagged by destination): Iām sure you get the general picture.
The result: life recalled at the click of a button (so to speak). One example only and, Iām sure, peculiar to me. 
Stephen
6 Likes
My records are organized with tags
A single database, and minimal groups
Instead of nesting, my tag naming reflects hierarchy
For example; Budget, Budget-Housing, Budget-HousingRent, Budget-HousingUtilities
My top level tags are Type-ā¦
which identifies sub-tags
For example Type-Receipt records are also tagged with Vendor-ā¦, and Budget-ā¦
I use an Applescript to assist with processing records , including tag assignment
Benefit of tags;
- Multiple assignment to a record
without dealing with the group/replicant stuff
Search using multiple tag criteria - Itās easier to assign a tag than searching for a group
- Folders/groups are for separating things
Tags are for collecting them together
(Jorge Arango and āknowledge gardeningā)
3 Likes
Iāve been digging into tags vs. groups and I find there is only a very small difference between the two. Groups are simply tags that behave a little differently than other tags. But not much differently
That said, I use a lot of groups and havenāt used tags at all ā until now. I want to keep a books to-be-read list in DT, and I think it would be a good idea to have one document per book (probably a blank document, maybe with a quick sentence why I thought the book would be good to read), and a tag for the author and another tag for whether the book is fiction or nonfiction.
One way I have started to think of it is that tags, groups and documents are all objects that live in databases, and they are more similar than they are different.
I expect Bluefrog will be along momentarily to correct me. 
(And to be clear I will be grateful for the correction.)
1 Like
The presentation of tags and groups are similar in DEVONthink. This makes it easy to navigate for many people who prefer working with them. However, groups and tags arenāt synonymous. Groups are general containers, made for organizing and segregating items. Tags are for specifics, e.g., adding context to an item. They can be used to create virtual groups, i.e., smart groups, assist in sorting, or aid in searching. As they are arbitrary values, you can use them in any fashion youād like with very little limitation. And they also can be applied to any item in your databases, unlike the smaller scope of the Properties inspector which is limited to PDF, RTF, and a few media types.
4 Likes
Folders/groups are useful for navigation. Tags are useful for filtering a search.
It eventually depends on your preferred way to retrieve information. Your method is the best method (likely not a static one, though). If you are generally satisfied with text search alone, like I am, then it might not be necessary to maintain an elaborate structure in the first place.
2 Likes
By disabling the option to exclude groups from tagging (see File > Database Properties) itās also easily possible to use groups like tags.
I tend to have a tag āmethodā per database. Let me explain. I have a cookbook database where the groups are recipe types (starter, soup, main course, fish course, pudding, salad etc.) I use tags for the ingredients so I can look at an alternative dimension like all recipes using avocado.
For my work database, groups represent individual projects (often broken down into sub groups). Tags are used to define the document type (proposal, plan, budget etc.). I also use tags for document status (WIP , done, waiting on someone). So if working on a new project I can quickly call up all proposals I have previously written to copy and paste good bits into the new project proposal. Similarly tags here give an alternative view or ādimensionā on my documents.
Instead of tags I could have used custom metadata, but I personally find tags are much easier to add. With DT you can get as complicated as you like
3 Likes
Iām always fascinated on how you use tags inside DT, something I still havenāt the courage toā¦
May I ask some genuine questions?
-
How big is your database?
If I had only one database, it would be hundreds of GBs: would it still be manageable? -
Do you use DevonThink on mobile too?
I highly rely on DTTG, and alas it doesnāt support smart folders: how could I work on a single project without a regular folder?
My database size
. 30,000 records
. 22GB
I rely on DTTG for my mobile access
My projects are identified by tags; Project:ā¦
which can be added to Favorites
WOW I was afraid of my 4K records in 5 GBs!
How do you find the projects on mobile?
I know you can search by tags, but can you search by a chain of tags?
I ask because DTTG doesnāt allow smart groups.
Every now and then I wonder if I should start using tags. I donāt at the moment because: Every group and every filename, while not synonymous with tags serves much the same purpose. If I search on a keyword, the result is ranked by what DT thinks is likely to be more important, the search can be modified in all the usual ways to limit the number of results, and tagging a file when it arrives in DT is just another step Iām too busy to take.
However, comments in this article (thanks all), suggest using a tag for a term that does not appear in the filename or document content, which would sort of function like a replicant. That does sound worth the extra step; all I have to do now is remember to use it!
One minor suggestion regarding search results is that I wish I could change the filename in the results without having to stop and open the file. Or maybe there is a way to do it that I havenāt yet discovered? Sometimes I see a ridiculously long filename that doesnāt reveal what it is until the very end because I didnāt stop to rename it in the first place.
As per my screenshot above,
I go to the sidebar taglist, and scroll-down to my project tags
Current projects are accessed via Favorites
Iām using groups and replicants for similar purposes. Like you, I have one group per project, but then I use sub-groups for types of documents within a project ā one for discussion of the project itself, another sub-group for notes, and another for drafts.
I use replicants to surface important documents from the sub-group to the top level. Iām currently working on three projects; I have a note listing deadlines and timelines for all those projects, and that note is replicated in the top level of each project folder and the discussion sub-folder for each project.
I also use sub-groups freely to join together related documents. For example, today I need to send research findings to the sponsors of a report, and Iāve grouped a Word document prepared by an analyst with my own draft of the report that will go to the sponsor. Just two documents in a group, and Iām OK with that.
Folks who see a distinction between folders and tags always say stuff like that and I donāt get it.
Theyāll also say that documents live inside documents while tags are attached to documents. And non-Devonthink users will say that a document can only be in one folder but a document can have many tags.
For me, both the folder/group and the tags are characteristics of the document.
In non-DevonThink systems, there is often a major distinction between the visual presentation of folders and tags, but thatās not so much the case with DT. Both groups and tags are displayed similarly and hierarchically in the left sidebar.
What are some of the significant practical differences between groups and tags in DevonThink? How are they implemented differently from a technology perspective? Iāve read the manual and nothing jumps at me. They seem interchangeable.
But in DevonThink, tags are presented in the sidebar in the same manner as groups.
Why is it necessary to exclude groups from tagging for that purpose?
Iām not arguing for the sake of arguing here (much as I love a good argument!). I feel like understanding the distinction between groups and tags might help me use DT more effectively and hopefully benefit others.
Thatās true. Now imagine you have 200 tags in the sidebarā¦
Itās usually not an issue with groups, which are nested at multiple levels. Thatās not always the case with tags.
3 Likes
It can be quite useful, although with its own caveats.
Say I have a document at Topics > Philosophy > Aristotle > my note.md . When I disable exclude groups from tagging, this document will inherit three tags: Aristotle, Philosophy, and Topics. The first two is desirable. The third one is redundant. By excluding group Topics from tagging, I can get rid of the redundant tag, however all new groups under Topics will also be excluded from tagging by default.
Now I can make a smart rule to ensure that every single descendant group of Topics is not excluded from tagging. That is extra complexity, though.
In non-DevonThink systems, there is often a major distinction between the visual presentation of folders and tags, but thatās not so much the case with DT.
If you look carefully, you would see DEVONthink supports both a token-based appearance as well as a hierarchical structure for tags. Other apps generally only support tokens.
Both groups and tags are displayed similarly and hierarchically in the left sidebar.
Correct. As I already stated, this makes navigation of tags easy for many people, especially when using nested tags. And this displays like the common taxonomic structure people are familiar with, even from the simpler form taught in basic science education.
What are some of the significant practical differences between groups and tags in DevonThink? How are they implemented differently from a technology perspective?
There arenāt significant technological differences, but there are differences. For example, you canāt search for a document by using a group-based search prefix, e.g., group:2024 or location:Big Electric Co.. Those do not exist. You can choose a location in the scope bar of a search but that is not the same thing. Locations/groups are generally seen as important for visual organization. Tags are usually not unless youāre browsing in a hierarchical structure like DEVONthink also offers (and again, why we support both methods).
Tags are about the properties of a thing, the context of it. They usually reinforce concepts in a document or introduce arbitrary concepts about it. Consider a MP3 of a song (and donāt be blind to the fact weāre talking about more than documents like PDFs). It may have attached metadata about the artist, album, release year, etc. but it likely has no information about the players. You could add a tag of bassist: Michael Manring to the file and use it for later searching, smart groups, etc. It doesnāt matter whether that MP3 is in a group for the album or sitting in the Global Inbox, that tag has added arbitrary but useful information to the document, information āwhat group itās inā canāt necessarily provide.
As a second example, take a PDF of a receipt. Yes, the OCRād version has the date and time, totals, restaurant, etc. Adding a tag for the restaurant, month and year, VISA, etc. reinforces data in the document. You can also include tags for a category not included in the text of the receipt, like entertainment or the business contact you were dining with. These all allow easier corralling of similar documents, e.g., at tax time or other purposes. I could search for tags 2024, business, Harvey's, to find receipts for business dinners I went to last year.
Part of the beauty of tags: Instead of having to create separate groups and replicate into them (which actually affect all replicants since youāre now adding parents to every replicated item), you leave the document in situ and can dynamically gather them in smart (virtual) groups as needed. To illustrate, all these PDFs are in separate groups but easily gathered with a smart group looking for the tag Januaryā¦
And this is not and either/or situation. DEVONthink doesnāt force a method on the user, it accommodates one, the other, or both.
2 Likes
I saw an interview with Jorge Arango in which he made the comment that (roughly summarised) folders (groups) are for separating things and tags are for collecting or gathering things (among other uses). I found that a useful rough distinction of their functions.
6 Likes
If you move or replicate a document to another group, the tags stay attached to the document. So one ācharacteristicā changes, the others donāt.