Transitioning links from Obsidian to the Archive

2»

• @micahredding said:
I'm not looking for security.

I'd love to have a system capable of maintaining integrity despite the ever changing nature of its parts but I'm under no illusion that I can have that.

Just like you are aware that [[ID]] links don't mean editing everything else about linked file is "safe", I do not expect auto-updating [[Title]] links to give me impunity. I'm prepared to see certain things break, I consider it a natural part of the process and I have my strategies for fixing what's broken.

I can see your point about "burned in" titles preserving more context for why the connection was made in the first place but it does not mean self-updating links are completely bare of context or that it can't be supplemented.

• First of all, I make conscious effort to sort the mess out at the moment of its creation. Usually it's not hard to tell that a potential edit of either the filename or body of the note may affect notes linking to it. I try my best to inspect backlinks, break things up and create new notes (even if empty placeholders) at the time when I have the clearest view of both the past and the future of the note.
• I do store the old title as an alias in the note's meta if the change was important enough. I mostly do it for the ease of searching but it's same "burned in" title just in one centralized place instead of being scattered across multiple notes. I may even attach a more deliberate comment on the change to the bottom of the note if the edit really turns things upside down.
• Links that are most likely to "break" usually involve claims and ideas and would have explicit commentary on why they are being connected. Just because links have self-descriptive titles doesn't mean they exist in a vacuum.
• A note containing the link and a note linked to, even if edited, still provide context
• If everything else fails, there's git. Old link title in the linked from note is there, old content of the note linked to is there. Every change to every file is tracked and the context at any point is recoverable.

We all store different things in our ZKs and types of notes differ naturally from person to person. I find that a huge portion of my notes are rather risk-free when it comes to editing. I'm very attentive with notes that carry such risks and handle them "by hand". For everything else automatic cleanup is a blessing. I enjoy having order where possible.

Also self-updating is not the only or even the primary reason I'm using title based links. But it's a huge help in ensuring those links work.

• @Will, I've been a lurker on this forum for several years now, and have seen quite a few of your screenshots and videos. I find your notes layout very simple and pleasing and have definitely been influenced by it somewhat, even if I can't borrow much directly due to adapting a slightly different workflow.

So I of course understood that you did not suggest using ID-only filenames within the body of a note. It's the file lists that would be the biggest problem were one to go that route. Even if one particular app learns to go around this problem by displaying something else in the sidebar, you would not want to depend on that app's availability. One day you may need to access your notes through something else and a wall of digits won't be pleasant to deal with.

@Will said:
I'm not sure I'd ever use this though as this is not how I form relationships between ideas or 'links'. I have to search established notes and read them to determine if they are candidates for connection. I can't remember more than the last 3-5 notes by their filenames to be able to choose them from a list. This requires thinking, reading, sophisticated search strategies, and "a lot of toggling views, hopping between the notes and copy & pasting." I wouldn't have it any other way. I find this the most enjoyable part of zettelkasting.

I my experience are 2 primary scenarios for linking:

• registering a connection that I know is there and
• exploring possible connections

There are a lot of connections that I find in my head rather than in my vault. Some of them are obvious like a broader topic or a related person for example. Other times I type something and it makes me think of something from a completely different area. I'm in my composing mode, I need to quickly establish a link to that very specific thing and move on. Light bulbs go off in my head all the time, so it's a very important part of my workflow. Link auto-completion allows me to fly through such insertions without leaving the flow.

Yes, of course it's possible to achieve completion for filenames that contain IDs but

• It's still the title part of the filename that makes it possible. It has to be there and it has to be written in a particular way to ensure maximum findability. It's the part that matters
• The ID part acts as a hindrance in this case. It requires the output to be processed in a very specific way, so it demands a custom solution. KM is a powerful tool that you seem to have a great command of but it won't be much of a help on a Windows system or iOS, while completion of a filename "as is" is something available "out of the box" across multiple applications and platforms.

For a second scenario, where I actively explore my vault for possible connections, I do resort to a more complex routine of course. It still doesn't require a lot of toggling views though, as Obsidian lets you run search queries right from the body of a note through a code block – an approach I prefer personally, because it keeps me in the same view and provides more space for results than a narrow sidebar does. And with some help from plugins, it is possible to run complex searches that extract parts of the note's metadata or body and present results as a sortable table for example (screenshot from github)

I'm not a big fan of relying on an app-specific functionality, so I don't keep such embeds long term with an exception of few notes I have marked as disposable. But I enjoy having this powerful database-like view in my toolbox. Having a simple list display as a Kanban board is a nice little tool as well.

None of this is directly enabled by naming files in a particular way, obviously. This functionality is provided by the particular application and plugins I'm using. But filenames that contain nothing but the names of the entities I expect them to represent are what keeps the output readable and the worflow simple and enjoyable to use. So they do enable it in a way.