"taco, what are you doing?"

running a bash script that converts 800 json entries to markdown by piping the wallabag json archive that has html nestled in it to pandoc and then adding frontmatter to the markdown file by copying the json data.

"it's four o'clock in the morning, why on earth are you doing that?"

…because i've lost control of my life.

in related news i fucking snapped today

#obsidian #logseq #pkm

the linking continues...
i'm honestly considering migrating some static sites to an obsidian vault + https://quartz.jzhao.xyz...
Welcome to Quartz 4

Quartz is a fast, batteries-included static-site generator that transforms Markdown content into fully functional websites.

Quartz 4
i've manually tagged and added a summary to 203/986 clippings from wallabag. some of them failed to be retrieved due to people blocking scraping of their site due to ai crawlers so i'm using the obsidian web clipper to extract markdown from those pages. the nice thing about obsidian as well is that i can just paste rich text and it will translate it to markdown automagically.
i keep stumbling upon articles that i thought i had archived correctly that actually fell to an llm scraper trap that produced some random text with bait links. gotta reimport a lot of 'em... :(
very disheartening that a large amount of links i need to clean up from 2024-2025 are just gone, link rot is so fast these days

implemented a newer, simpler tagging system that goes by domain rather than specifics and now it's all coming together...

#obsidian #pkm

@chirpbirb Damn this looks awesome; way more organized than my notes 
@chamomile this has "only" taken me like 2 weeks of thorough examination of literature and scripting to auto tag stuff
@chirpbirb Auto-tagging sounds pretty useful. Got a link? My tags are pretty bare because I always forget to and so I mostly maintain associations by obsessively linking pages together.
@chamomile it's a whole custom python thing i've got over here, i didn't find anything specific out of the box and it involves using a local llm to read articles and parse tags. i feel like there's a way easier way to do this though
@chirpbirb Ohh I see, interesting. I'd be surprised if there wasn't already a community plugin for that sort of thing tbh, but that's probably be doing about the same thing.
@chamomile honestly if i'm bored enough i might make a plugin that just maps keywords to specific tags and writes that to the file somehow
@chirpbirb this sure looks cool but does it help?
@colinstu yes, imo. narrowing down the categories for tagging and moving to [[linking]] for everything else helps generate spheres of thought i can then reference later or make category wells out of.
@chirpbirb my fav is when they rot while still open as tabs (that I've left open for so long). I need to almost copy the url into a new tab and reload to make sure I don't lose anything off the old cached copy..
@chirpbirb Sucks so hard how they've ruined it for everyone.
@chirpbirb Speaking of Obsidian Web Clipper... https://defuddle.md/
Defuddle — Get the main content of any page as Markdown.

Get the main content of any page as clean, readable Markdown.

@chiroplut i saw that! made by the same guy who made the web clipper :D