I’m not sure if DT’s database schema would support this feature or not, but:
When I use the See Also/Classify pane, DT suggests places I could move my document to based on its contents. This is great, and constitutes 95% of how I organize my document in a very large database.
With the new tagging support, what I’d really like to see rather than manually tagging something is a “Suggested Tags” section in the See Also pane. This would be determined by doing whatever DT currently does to find similar documents, and then instead of presented me with the folders those documents live in, present me with all the explicit tags those documents have. Odds are, DT can automatically figure out most of my tags for me.
In fact, since Groups are implicitly tags, I’m sure there’s a clever way of unifying the whole Classify mechanism.
If the tag is assigned by its group location…
If you move a document from group to another, the original tag is removed as the document is no longer in the group where it inherited the original tag. If you want a document to appear in a different group yet retain the original tag, you will need to replicate the document instead of moving it.
If the tag is assigned a tag contained in the Tags group…
The tag will move with the document from one group to another.
In other words, if you have:
GroupA
DocA
GroupB
Moving the document from GroupA to GroupB will remove the GroupA tag and assign the GroupB tag (assuming that both groups have not been excluded from tagging).
If you have:
GroupA
DocA
GroupB
Replicating the document from GroupA to GroupB will retain the GroupA tag and also assign the GroupB tag (again assuming that both groups have not been excluded from tagging).
Finally, if you have the following:
Tags
TagA
DocA
GroupA
DocA
GroupB
Moving the document from GroupA to GroupB will remove the GroupA tag, assign the GroupB tag, and TagA will be retained.
greg_jones, thanks for the examples. My understanding is improved but not yet quite clear.
Your last example,
confuses me a bit. Let me rephrase it without duplication of A & B.
if you have the following:
Tags
TagMammals
Doc”An article about mouse reproduction”
Groups
GroupCarnivores
Doc”An article about mouse reproduction”
GroupRodents
Seeing that I have misfiled the article under “Carnivores”, I move it to the group “Rodents” and it loses the LOCATION TAG “Carnivores” and acquires “Rodents “ instead but retains the NON-LOCATION TAG “Mammals”. Good.
But I still do not understand how tags are groups in one way, and not in another way.
Here are a couple of questions:
At the moment I have nothing in my Tags folder. I guess this is because I had 1500 tags (mostly words from titles of items in an RSS feed that I didn’t know I had), and so I deleted all the tags. [ Is there a way to add back all the Groups as Tags, all at once? ] If I select the database and View> As Tags, I see a list of every Group and Sub-group in my database, arranged alphabetically. None of these are Tags per se, and none of them would continue to be attached to an item if I moved it from one Group to another. In addition, if I select an item and view its Tags in the Info window, I have now way of knowing whether the “Mammals” that appears is a Tag proper or just a Group. Since one is linked to the item until I take it off, and the other is only connected with the item as long as it is located in the Group of that name, it matters quite a lot.
Nothing has been added to the empty Tags folder even though I have created new folders and subfolders with items in them. Theoretically it seems that now every time I add an item to a group, that group should show up as a new tag in the Tags folder. Why isn’t that happening?
I don’t see the “Tags bar” (sounds like a standalone item, taskbar or palette) mentioned in the Help, unless it is the little pane in the Info window. That does show the Groups an item is in, and enables me to add other Tags which do then appear in the Tags folder.
The Help says, about tagging,
I find this misleading. Tags are, apparently, not just “visualized as ‘labels’ attached to the item”, they function as labels and not locations (see greg_jones post with examples).
If they were identical, tags and groups, presumably I could not have duplicates: a Tag “Mammals” as well as a Group “Mammals”. If they functioned the same, removing an article from the Group “Mammals” would remove it from the Tag “Mammals”.
I apologize for going on. Some of my confusion is mechanical, how to do things. But some is fundamental, in the conceptual framework. It seems to me misleading that 2 different things are being presented as being the same.
Using your terminology, tags that are assigned by virtue of the group that contains them (LOCATION TAG “Carnivores” in the above example) will not also be duplicated in the special tags group ‘Tags’. The Tags folder only contains tags (NON-LOCATION TAG “Mammals”) that are explicitly created by the user and/or by importing (Open Meta tags, RSS feed categories, etc.).
Here is a picture of the Tags Bar, which appears below the Navigation Bar and above the View/Edit pane. It is activated by clicking on the far-right icon in the Navigation Bar. Note that the icon DOES NOT display in the picture-it was clipped off when posted to the forum, but it is to the right of the far-right icon that does appear below.
In the above picture, I am assigning tags to a document that resides in the group ‘Case Study Methods’, (LOCATION TAG) and has already been assigned the (NON-LOCATION TAG) tags ‘Harvard’ and ‘Staff Study’. I am in the process of assigning a third tag by typing ‘kno’ and DT has proposed the two tags that begin with ‘kno’-specifically ‘Knowles’ and ‘Knowledge Management’. It is worth noting that ‘Knowles’ is a tag in the Tags group (NON-LOCATION TAG) while ‘Knowledge Management’ is a group (LOCATION TAG). If I continue to type ‘Knowledge Worker’, then a new (NON-LOCATION TAG) group will be created in the Tags group, and this document will be assigned ‘Knowledge Worker’ as a tag.
They are identical, technically, and the fact that they can have the same name supports this. Leaving the concept of tags aside for the moment, I can have as many groups as I want in my database named ‘Misc. Notes’ because all the groups (folders) function the same. I can also create a tag named ‘Misc. Notes’, or multiple tags named ‘Misc. Notes’ if I want, because they all are technically identical. Now let me add that this is where I believe the otherwise excellent DT method of tagging shows a weakness. If I have as example 100 groups in my database that each contain the sub-group ‘Misc. Notes’, then in the Tag View (Command-6) I will see 100 groups named ‘Misc. Notes’. Also, if I want to assign the (LOCATION) tag ‘Misc. Notes’ using the Tag Bar, I have no way of knowing to which of the 100 groups named ‘Misc. Notes’ it will be replicated. I find this highly confusing and I hope that this can be done differently (see below).
No need to apologize at all-you’ll only understand the possibilities, and limitations, of tagging when you fully understand how they work. I would suggest that you turn on the Tags Bar and also look at your data using the Tags View (Command-6) The Tags View currently lists all tags differentiated by gray (LOCATION) tags and blue (NON-LOCATION) tags. Also, keep in mind that tagging is not yet finalized in pb8-the documentation mentions specifically that the Tags View is still incomplete.
Thanks, this information has helped considerably. I have been experimenting a bit with how it works.
Is it possible to use Search to find all articles/items that are tagged with a certain combination of tags (NON-LOCATION, or BLUE tags)? Of course that would be extremely useful. But my simple trial was unsuccessful. I tagged an item with 2 (blue) tags that were, to avoid confusion, distinct from any existing Group names, then did a Search in both “All” and “Metadata” for “Tag 1 AND Tag 2”. The result was “No items found”.
Hi,
maybe I’m missing something but regarding the behavior of tags when moving I think there’s another ‘strange’ thing:
I generally tag entries while they’re still in the inbox and after I move to the proper location. Well if this second part is made manually there’s no problem but if I use the “move” button in the classify window, once moved the tags (real tags not group tags) are deleted and the only visible tag is that of the recently-moved-to group. According to my understanding of tag principles in DT this behavior doesn’t make sense, what do you think? It seems rather buggy…
Has this feature been implemented? It would be a great help for my workflow.
I am a DT newbie. I am interested in using DT with gmail-style “flat hierarchies” : inbox and archive are my only top-level groups. All categorization happens via (hierarchical) tagging. The only other groups I make have to do with keeping different parts of one logical document together: for example, a PDF paper with it’s bibliography information, or all the parts of a multi-part article (or multi-volume book.)
Also, I read somewhere that tags do not affect classification. If this is true then my organization system basically defeats the purpose of using DT.
