I've started using an under-appreciated feature in Devonthink that is proving to be extremely nice. It's enough to sidetrack me from Obsidian despite the normally undeniable pathology of CRIMPing.

You can set Devonthink groups to work as tags, either per group or per database, by un-excluding groups from tagging.

If you choose to do so per database, you can still set individual groups back to the default in which they are excluded from tagging.

If you un-exclude a group from tagging, that means it becomes a valid tag in addition to being a group (folder). You can create a group called Eye Gouging, un-exclude it from tagging, and then tag documents in the Larry, Moe, and Curly groups with Eye Gouging.

When you apply that tag, Devonthink creates a replicant of the tagged file in the Eye Gouging group.

You can then ignore how the replicant got created. It's just a replicant. Delete the source, the remaining replicants survive, all replicants of each other. Delete the group that formed the tag, any replicants are safe, they just lose the tag of the group you deleted.

There's one nice wrinkle. If you have a group called Stuff that is un-excluded from tagging and create files in it, each will automatically have the Stuff tag.

If you delete a file in Stuff, it's deleted (unless it lives on as a replicant somewhere).

If you just remove the Stuff tag, it disappears from the Stuff group but it isn't deleted. It either soldiers on in the form of other replicants, or it jumps to the database's local inbox for later re-filing.

There are some limitations.

You can add tags by path. For instance, if you have two traditional tags, Chap1 and Chap2, each could have a Reference subtag.

You can tag a document with Chap1/Reference or Chap2/Reference. The desired instance of Reference will be selected.

The tag search path doesn't work with group hierarchy. The best way to add a group tag that isn't unique is to replicate the document to the target group. The group picker will let you navigate the group tree to the specific group-as-tag you want to use .

Interestingly, tags in Devonthink are pretty much groups that are automatically un-excluded from tagging. In fact, if you have categories of tags that you use strictly as containers for subtags, you can exclude those top level tags from tagging. That will declutter a document's tag list when you use the subtags.

Anyway, thought I'd muse about tags in Devonthink. I should learn how to write Obsidian plugins. That would get my CRIMPing back on its feet.

... or you could just play with existing Obsidian plugins. I'm currently experimenting with Cardboard, a kanban plugin that allows you to use dates or tags as column identifiers (and set up as many "boards" as you like, using whichever convention you prefer and filtering on whichever tags, folders or documents you like). Potent stuff!

Amontillado wrote:

I've started using an under-appreciated feature in Devonthink that is
proving to be extremely nice. It's enough to sidetrack me from Obsidian
despite the normally undeniable pathology of CRIMPing.

You can set Devonthink groups to work as tags, either per group or per
database, by un-excluding groups from tagging.

If you choose to do so per database, you can still set individual groups
back to the default in which they are excluded from tagging.

If you un-exclude a group from tagging, that means it becomes a valid
tag in addition to being a group (folder). You can create a group called
Eye Gouging, un-exclude it from tagging, and then tag documents in the
Larry, Moe, and Curly groups with Eye Gouging.

When you apply that tag, Devonthink creates a replicant of the tagged
file in the Eye Gouging group.

You can then ignore how the replicant got created. It's just a
replicant. Delete the source, the remaining replicants survive, all
replicants of each other. Delete the group that formed the tag, any
replicants are safe, they just lose the tag of the group you deleted.

There's one nice wrinkle. If you have a group called Stuff that is
un-excluded from tagging and create files in it, each will automatically
have the Stuff tag.

If you delete a file in Stuff, it's deleted (unless it lives on as a
replicant somewhere).

If you just remove the Stuff tag, it disappears from the Stuff group but
it isn't deleted. It either soldiers on in the form of other replicants,
or it jumps to the database's local inbox for later re-filing.

There are some limitations.

You can add tags by path. For instance, if you have two traditional
tags, Chap1 and Chap2, each could have a Reference subtag.

You can tag a document with Chap1/Reference or Chap2/Reference. The
desired instance of Reference will be selected.

The tag search path doesn't work with group hierarchy. The best way to
add a group tag that isn't unique is to replicate the document to the
target group. The group picker will let you navigate the group tree to
the specific group-as-tag you want to use .

Interestingly, tags in Devonthink are pretty much groups that are
automatically un-excluded from tagging. In fact, if you have categories
of tags that you use strictly as containers for subtags, you can exclude
those top level tags from tagging. That will declutter a document's tag
list when you use the subtags.

Anyway, thought I'd muse about tags in Devonthink. I should learn how to
write Obsidian plugins. That would get my CRIMPing back on its feet.

Is there a plugin to allow arbitrary reordering of the notes in a folder? That's handy to be able to do.

Bartender?

https://forum.obsidian.md/t/file-explorer-custom-sort/1602/123

Cool!