I kept a license current for Curio for the equation editor. It’s only since the file-backed and synced figures features that I’ve started using it again for story planning.
It is not the right tool for large datasets.
The weakness with option-drag to create an automatic link in Curio from Devonthink is that it uses the file path, not the Devonthink hyperlink.
There probably isn’t an alternative to that if you want to edit-in-situ from the external app.
Being able to edit Devonthink documents from Curio is pretty slick. I have to admit it’s a fragile thing, though.
Would like to learn how to integrate DT with Curio. Do, don’ts, best practices. Is there an article or forum topic you can recommend? Thanks.
Would like to use a Kanban style todo / project board, too.  @Luminary99_0 big thank you for your solution and clear explanation
For simple tasks I use the Reminders app with separate lists, but it’s missing the requested overview. For projects I use Jira. On both cases I use Hook to generate deep links to documents and related information in DT.
See here - note I believe you need to turn on Spotlight for this to work.
Investigate file-backed text figures in Curio. You can put a file-backed text figure in a kanban stack. Initially, that may look messy but you can right click and tell Curio to display the file as an icon instead of the full text.
If you option-drag a DT document into Curio and tell Curio to create a file-backed text figure, you can edit it in either Curio or DT. Curio will use DT’s copy.
There is one way I’ve found to break the setup.
Imagine you sync your Curio project to another machine. If you open your Curio project before you import the DT database, Curio will see missing files, flag them all as broken links, and ask you to manually fix them individually.
If the files were in an indexed DT folder, you can navigate to them and re-establish the links. If the files were normal DT documents, Curio won’t let you navigate into the database package directory to find them.
It’s pretty cool to put a corkboard front end on Devonthink. Limitations apply and opening Curio before importing the DT database will break Curio’s links.
Nothing, of course, breaks Devonthink, at least in my experience.
Can Curio reference the DT item by x-devonthink link or only by file path?
It doesn’t seem it does…
I can copy and paste x-links, but ALT-drag result in a file-path reference.
Also a nice feature, but in that case I prefer to duplicate an item to the Curio repo or use a x-link to the source (in this case DT). For that I use Hook.
Curio can reference DT documents either way, link or file path - however, option-dragging a document from Devonthink to Curio creates a link that is apparently by file path. You can add a Devonthink x-link as a reference on any object in Curio. When you click such a link, DT will open and present the object you’re linking to.
The advantage of the file-backed text is that you can have what amounts to a bunch of index cards on a corkboard. Any or all of the index cards can be text documents in DT, and editing them in Curio updates them in DT.
The disadvantage is if you don’t have the referenced files available when you open Curio the links will be marked as broken.
One thing that helps is that the file path to a document in DT doesn’t seem to change for the life of the document.
DT’s mantra that tags are groups sort of works both ways in this case. If you move a document to a different group in DT, its physical file path doesn’t change, it’s just designated in DT as belonging to a different group.
Smart folks, those Devonthink developers.
Thanks - I think if you rebuild a database the filepath may change and there are other situations when it can happen too. So the item links are preferred.
You may be right. I’ve actually done that, rebuild a database after linking files to Curio. So far, no broken links. That’s not exactly rigorous, of course.
I’ve also found that if I move a Devonthink database to a different subdirectory Curio has (so far) not had trouble with file links. I don’t know how it locates files that have moved. In my dabbling it hasn’t had trouble.
If the file completely doesn’t exist, that’s when Curio reports the link broken.
Or, maybe I’ve just been lucky.
That’s interesting - do you know if the links by filepath work if you sync your DT3 database to another computer and load your Curio project on the second computer?
Yes, I’ve done that. The pitfall comes when opening Curio before the source files are created in the Devonthink import.
If the files Curio needs exist somewhere, it finds them, at least so far. I keep spotlight indexing turned off on DT databases, so I don’t think Curio is finding the files that way. Or, maybe Curio is registering linked files in Spotlight from the Curio side of the equation.
Other than that potentially rough edge, I’m happy with using Curio as a front end for DT.
This is the closest to what I’m used to in a Kanban board. I’ve used them both via GitLab Issue Boards and also as Aha! Workflow Boards. I particularly like the GitLab implementation, as it seems to “flow” the smoothest.
In both instances, the key is that the visual location in the Kanban board is based on a defined piece of metadata for the item you are displaying on the board; my usual use case was to have a process step label for each item. The item would be displayed on the board based on that label.
Another important attribute of both implementations was that it was a two-way street between the metadata and the visual display of the item. If the metadata got changed (i.e. you labeled an item as “done”), it would show up in the Kanban board column associated with the “done” label. The other side of this was that if you dragged an item between columns, the relevant metadata for the idem would be updated to reflect the move. This was extremely powerful in doing Kanban reviews during videoconferences; when somebody reported a change in an item, you visually dragged it during the meeting, and everyone saw the change - very powerful indication of progress. Since the metadata was updated, anyone could also look at the board or item outside the meeting and see the current status.
Conversely, I would not find a simple “corkboard” Kanban board very useful. You’d only end up keeping the metadata and visual appearance in sync manually (and who’s got time for that…), and it would inevitably become wrong very quickly.
Finally, updates to metadata can be automated using the many tools DT gives you for automated data handling. I could see this becoming very powerful very quickly.
i’d love to get that visual two-way street with DT as the backing store. Do any of the approaches mentioned in this thread (e.g., Curio shelves) provide it?
Yes, I hear you - I am not a Kanban user.
However, as you mentioned, things in DT3 can be automated. So for instance you could create a tag for each item and change them in the item itself. You would then use a smart rule to move the item to the group with the corresponding tag. So that automates part of the process.
You could create more sophisticated rules to place them in subfolders etc. Like you intimated: DT3 is very deep and flexible.