created: 2024-03-24 05:56 tags: - daily-notes

Sunday, March 24, 2024

<< Timestamps/2024/03-March/2024-03-23-Saturday|Yesterday | Timestamps/2024/03-March/2024-03-25-Monday|Tomorrow >>

🎯 Goal

• To learn how to use Obsidian Dataview to access data from each header (#) and subheader (##)

accomplished:: false

🌟 Results

• Learned how to flatten lists in an Obsidian file via Obsidian Dataview
• Might be able to do more custom pulls with Obsidian Dataviewjs

🌱 Next Time

• Figure out a structure that would allow me to subdivide my notes (i.e treat each individual section like its own notes folder for that particular subject for that particular day)

📝 Notes

Ok so I quickly got distracted with the idea of Floaty Notes, and I think I've come across a forum discussion that talks about how I could implement this into my DIMMiN App (specifically somewhere in the BigBrain App). I'll have to come back to this idea because I think it will be useful for actually publishing / establishing my notes. For now I need to focus on header-level metadata access.

As I talked about in 2024-03-18-Monday, I don't want to create a new file / note for every single different concept I come across. In fact, if it were easily possible I would want to be able to write my notes in a format that had many points of metadata that I could integrate about a specific section. The example I gave last time was

Project A

• Status:: Check
• Summary:: Found out the meaning of life or something
• tags:: #ProjectATest

However, I'd also like to be able to collect similar information about project B

Project B

• Status:: Pending
• Summary:: Lost the meaning of life or something oops
• tags:: #ProjectBTest

TABLE status, summary
FROM #ProjectATest 
LIMIT 5

The above (which also changes its representation in the previous note, 2024-03-18-Monday) doesn't quite work the way I want. I want only this file to be referenced, and I want project A and Project B to be referenced as their own rows in the table from this specific note. I do like the fact that the different metadata is separated into its own columns correctly though so that's a start.

Maybe I could use a Dataview Object. Looking into the documentation on the dataview object type taught me how to filter dataview queries to the current file using the where file = this.file filter:

TABLE status, summary
WHERE file = this.file

Now I just need to separate each piece of metadata into its own row by subheader. I found this post that talked about how to implement something like this in Obsidian Dataviewjs. This wouldn't work for me, and when I asked ChatGPT about the problem it said that was because dataview doesn't natively allow access to the "meta" subheader field referenced in the solution. It did however suggest that I could use either the header syntax or a list syntax:

• Project: Project C
• Status:: Check
• Summary:: Earned the meaning of life or something

• Tags:: #ProjectCTest

• Project: Project D

• Status:: Pending
• Summary:: Pasta
• Tags:: #ProjectDTest

Which might mesh better with my current notetaking style. Unfortunately this also returned no results. What the..? For some reason the suggestions given by ChatGPT don't work well with a lot of obsidian dataview procedures.

I'm still trying to find a way to associate data at the bullet point / header level. An example of the format I'm trying is in the Issues file:

- issue: Fix DIMMiN header
    - desc:: "Found out that the 'Blog' section of the header is always highlighted because it uses the `nav__item  nav__item--active` list class. This could be updated so that whichever URL the user is currently on, that is the header that is bolded." 
    - tags:: #dimmin-app 
    - status:: #open 
    - priority:: #priority-low 
    - src:: [2023-11-18](/blog/vault/dimmin-notes/2023-11-18/)
    - resolved:: none
- issue: Use [PostgreSQL](/blog/vault/dimmin-notes/postgresql/) in [Local Version](/blog/vault/dimmin-notes/local-version/) instead of [SQLlite](/blog/vault/dimmin-notes/sqllite/)
    - desc:: "Should be using a PostgreSQL version locally instead of sqlite to better match between my developer environment and the production environment"
    - tags:: #dimmin-app 
    - status:: #closed 
    - priority:: #priority-low 
    - src:: [2023-06-20](/blog/vault/dimmin-notes/2023-06-20/)
    - resolved:: [2023-11-20](/blog/vault/dimmin-notes/2023-11-20/)

I want to be able to access each issue almost as its own file, then access the metadata around that issue (such as desc, tags, status, etc.). While this is well-established at the individual bullet level, I don't know if it's as well established at the nested bullet level I'm looking for here. Another option might be to use the easily traversable JSON format, but I don't know if that will actually be easier here.

Notes last touched today

List FROM "" WHERE file.mday = date("2024-03-24") SORT file.mtime asc

(Template referenced from Dann Berg, can be found here)