DTP: Tags deletion seems not working as expected

I suspect that the removal of the thags in both the Tags group and the Tags view does not work as detailed in the DT Help.

Indeed, I noticed that I have hundreds of tags (in both the Tags group and the Tags view) wich do not refer to any document. In the tag view which consists of blue colored and grey colored tags, only the grey ones (the usual groups) are referring to one or more items, the blue ones (the specific document keywords) are empty.

To confirm, I made the following test.
1° I select a document > In the info panel I add a specific tag named “Testtag” which does not exist already > Testtag is properly added to both the Tags group and the Tags view (entry is blue colored).
2° I remove the tag “Testtag” from the info panel of the document > at this time there are no longer documents tagged with “Testtag” > but Testtag keeps being listed in both the Tags group and the Tags view although there are no related items in the database.

I guess that this deficiency is the reason why I have so many orphan tags in my database.

2 questions:
1° could somebody tell me whether this is a well known bug, whether there is a fix… or tell me if I am understanding/doing something wrong?
2° is there a proper way to re generate automatically the Tags group and the Tags view to have them consistent with the actual database content?

Thank you for your help :slight_smile:

Hi Christophe,

I don’t think you are doing anything wrong, the help has this to say:

So I would agree with you that it isn’t working as the documentation says it should. I also have lots of empty tag groups.

You can manually delete these fairly easily (there are probably better ways to do this), double click the top-level Tags group to open it in a new window, set that window to list view and use the Size column to sort them so that all the empty tags group together. Now you can select all the empty tags and delete them.

I don’t know the answer for your second question, I think that once you have deleted the empty tags that your database should be up to date.

Regards Alan

Although “empty” tags do not vanish, contrary to the documentation, I prefer it that way. As Alan noted above, it’s a simple matter to locate and delete tags that don’t have database items associated with them, and to remove them.

When I’m working on a project I can create in the Tags group a set of tags that can be useful while working on the project, such as a tag for marking useful references for that project, “workflow” tags that help me plan the stages of the project and that can be applied to my notes and drafts in progress, and so on.

I can create such tags in the Tags group. At the beginning, they have no content, but they exist for future use. As I work on the project, a particular tag may be applied to several items, and so have members, then at a later stage have no members, then at still another stage have members. I would find it distressing if my “workflow” tags were to disappear each time they become “empty” as I don’t want to have to recreate them each time I wish to use them again. Example: When I’m reviewing draft segments of a project, I might want to apply a tag such as “Needs rewrite” to a segment. That and similar “workflow” tags help me keep an eye on what needs to be done next. After I’ve rewritten the tag-designated segment(s) of the project, that tag may have no members. But I want to retain it for future use; I don’t want it to disappear from the Tags group.

My attitude towards tagging in general is rather churlish. I think it’s a waste of time and effort to tag everything as I enter content. Indeed, I love DEVONthink because I don’t need to use tags or keywords to find everything.

But I’m excited about the addition of tagging to DEVONthink because it adds a flexible and powerful set of tools to the user’s arsenal for categorizing and ‘marking’ items within a database. Now, in version 2.0.3 the full Search window adds Tags as searchable elements in a query.

My vote is to keep tags in the Tags group even if the last item so tagged is untagged. Just as I find value in retaining a group created for a purpose, even if it presently has no members, I find value in retaining a tag created for a purpose even if it presently has no members.

Pruning a database to identify no longer desired Tags is easy enough. Indeed, I often use a tag for a specific purpose, then later remove it, whether or not it has members.

Thank you so much Alan & Bill.

I know for certain now that DT does not behave as the documentation says.

Bill’s opinion is interesting. Indeed, having available a permanent list of “in common use” and chronic tags, is a stength for methodical people and it could be a real frustration to see such materials vanishing in the air when not named.

I am also used to use pre defined tags which structure and discipline (some of) my activities but I also often rely on “on the fly” tags which depend on the circunstances and those are (should be), in my opinion, short-lived.

Anyway, I know how to rid my Database of pollution and … should I be polled, I think I endly would vote to preserve the orphan tags.

Again thank you guys for your lights which helped me in removing some doubts :slight_smile:

One other option would be to have a preference that lets you decide whether to delete orphaned tags?

Whilst I can see the benefit in Bill’s way of working, it won’t necessarily suit everyone. Personally I’d prefer that the tags behaved as the documentation suggests and cleaned up when orphaned.

A preference would give me the choice of which way I wanted to work and means that I wouldn’t have to manually purge tags.

I’m of the same opinion as Bill because I use tags in more or less the same way he’s described and would find it very distressing if my tags that were temporarily at 0 vanished.

What I would find very useful and has been suggested before in the past is a way to sort Tags from least to most. That option doesn’t exist in Sort by, the closest option is word count which does not always give a accurate sorting order especially if there are some PDFs which are not OCRd and for whatever reasons either can’t be, or I don’t want that copy OCRd maybe because it has comments written in hand across the original which don’t show up or preclude accurate OCR. This is also true of image tags and various other criteria where “word count” doesn’t work as you’d want while sorting tags.

Ideally you should be able to sort tags from least to most, so that all of the tags which are at zero at the moment could be visually examined and deleted if desired, whereas right now the zero tag groups or groups with only 1 or 2 entires are mixed in all over the place.

Or having, on one hand, a kind of preferred list of permanent tags which would be maintained strictly manually and preserved… and, on the other hand, the chance to create “one shot” and passing tags which would be automatically deleted when orphaned.

That would be almost perfect :smiley: . May be the Devon team will regard all these suggestions as being appropriate :bulb: