A folder of markdown docs in your favourite text editor, ftw!
+ It’s all within the editor you already know really well. Uses your existing tools.
+ Many editors have really good support for markdown built in. Treat H1’s like notes and along with modern fuzzy search for files/symbols you can easily get to any note and jump around.
+ If you want smarter [[liking]] there’s some good plugins out there to bring this to your editor.
+ Simple, future proof and no lock-in.
I’m currently enjoying markdown-oxide, an LSP for markdown docs. Captures all your notes as symbols so you can fuzzy search and “find references”, etc. supports #tags, too.
There's definitely a lot to recommend w/ your approach, which actually overlaps a bit w/ my reasons for using Obsidian.
My obd vault is "a folder of markdown docs" (which retains the "future-proof, no lock-in" benefits you cited). But the excellent WYSIWYG UX (open files in Edit mode, w/ "Live Preview" enabled) is something I haven't seen replicated in any VSCode or Cursor extension/plugin. I also prefer a dedicated tool for note-taking and "PKM" (Personal Knowledge Management"), as a peer of my IDE(s) for coding. I get to use the best tool for the job, w/ no compromises. Switching to a different IDE for a given project (eg IntelliJ for wrangling Kotlin) doesn't disrupt my workflow, and having clear context boundaries (note-taking vs coding) is a personal preference.
YMMV, different strokes,.... I know some emacs org-mode fans out there will extoll the benefits of using "one tool to rule them all", which does sound compelling... (shrug).
I love that there are such a variety of quality tools and approaches -- and, I'm very happy with mine.
I used a similar setup for a while: Obsidian for taking notes in markdown, and vscode for coding.
Eventually I moved to using vscode for both. My gigantic notes.md file is always open in tab 1, so I can go to it immediately with Ctrl + 1.
Finding notes in a single file is easier for me than finding them in a bazillion tiny files. And there's less friction whenever I need to make a note (no need to create and name a new file).
This! I have all my notes in markdown files in folders. I self host Silverbullet on top which is a nice web UI for managing notes, tagging them etc. All tags are stored in the individual markdown files using frontmatter. I also push my notes to forgejo (gitea) which means I have versioning. Works well for me!
I currently use flatnotes[1] as a Frontend with a background cronjob to add, commit and push every night.
Works pretty well so far with some acceptable nitpicks. It was a quick and dirty solution to overcome logseq couldn't run in a Server that time... Never needed anything else until now :-)
I have fallen in love with SVG+text Mermaid diagrams embedded in markdown code blocks, and LLMs are really great at generating them (by reading design documents and source code) and editing them with natural language commands. And they're not inaccessible unsearchable dead-end images you have to laboriously regenerate and check into the repo whenever you want to change something! ;)
Mermaid: Diagramming and charting tool:
JavaScript based diagramming and charting tool that renders Markdown-inspired text definitions to create and modify diagrams dynamically.
Include diagrams in your Markdown files with Mermaid:
A picture tells a thousand words. Now you can quickly create and edit diagrams in markdown using words with Mermaid support in your Markdown files.
I still feel like Roam Research did it best. Conor was truly a pioneer in the digital note taking field. I still miss the times when all of it was new, it was so exciting and everything felt revolutionary. A lot of Roam users are still using it even though the pricing hasn't changed a bit and neither has the design, but I don't think there's anything better even now. Zettelkasten is severely underrated and misunderstood. Notion and Evernote are disgusting, and PARA is a nonsensical grift. If you need more context regarding that, I recommend reading Conor's Twitter threads and watching interviews and livestreams featuring him.
I was an early adopter and "true believer" (aka paying supporter) of Roam Research, and will always be grateful for the game-changing (for me) paradigm shift it helped bring about in the PKM / note-taking space. I've long since moved on to Obsidian (which has met or exceeded my needs), but credit where it's due, Roam was transformational.
I was sure Zettelkasten was going to blow my mind, but it didn't. It seems like a nice way to organize thoughts to write. I'm usually not writing, just trying to externalize my brain. And for that Zettelkasten doesn't seem to offer much over Linking Your Thinking. (TL;DR pervasive wikilinks.)
I bought “How To Write Smart Notes”, but it’s misnamed: it should've been “Why To…”. I hoped it would tell me how to use Zettelkasten, but by the end it seemed to be a long sales manual without a how-to guide.
I would be more inclined to use it if there was offline support and it wasn't so expensive. I used (and really really like) Logseq for awhile but have since moved to Obsidian. I'd switch if there was a strong reason to.
I'm not the parent commenter, but made a similar transition (years ago at this point). I have always loved outliner formats, but the obsidian ecosystem is quite strong, and because I already had thousands of notes that were just plain old markdown, it was a more natural home.
I wrote code to facilitate the migration. Nothing too crazy, but in general I wrote scripts that:
- Add lines between logseq's daily notes format and the rest of the content
- Moving daily notes to month-based subfolders
- Automatically adding frontmatter to files that didn't have any
- Removing indentation when unnecessary
- Covert everything to space-based indentation rather than tabs
Bugs and performance mostly. The elusive database refactor might bring me back but we'll see.
I definitely have a preference for outliner, flat zk style, but I'm able to get the majority of the benefit from Obsidian while having access to a stronger ecosystem of plugins and first class publishing and syncing support. Meanwhile Logseq seems to have lost a lot of steam.
Found some documentation & sample website that answers a few of my questions, but I’d love to know more about pipelines and how they work with your notes. https://memex.bubbletea.dev/faq
That's actually my personal instance, but I made a random pipeline public so you can view it [1]! Here's a screenshot of the editing view if you're interested [2]
How would you support someone who wanted to migrate to this and bring all their data with them? Just wondering if you build something like this hoping people start fresh, or do you build tunnels to help people migrate in and out. Kudos.
I made an export function that just basically dumps the entire schema out to JSON that can be processed with something like jq, but I should see if I can make some sort of bulk import function that would be easy to use
I also haven't really stress tested it, this was really more of a personal wiki that I wanted to create specifically with the hierarchical linking/backlinking that I didn't see any other notetaking applications having just the way I wanted. But since the application is relatively simple, I think Postgres should be able to handle the load.
It uses fulltext search built into Postgres for all the searching capabilities though, so hopefully it is good!
Why on Earth does the README start with a huge image with absolutely unreadable teeny tiny light gray text on a dark gray background?
When I press Cmd-+ to zoom into the page in so I can actually read it, exactly the opposite of what I want happens: the image SHRINKS even smaller as the surrounding text grows to push it into a black hole! So I reset by pressing Cmd-0 and then try un-zooming by pressing Cmd--, and the image STILL shrinks even smaller.
What is the point? Did a Vibe Coding Anti-Accessibility LLM do this, or was it actually intentional? That first impression makes me completely uninterested in looking at the rest of the project, rightfully afraid that it might be as purposefully badly designed and inaccessible and contemptuous of users at the first readme page.
Why would anyone bother to go to so much effort to make an image of inaccessible unlinkable extremely low contrast tiny unreadable text wasting huge amounts of horizontal and vertical white space with no usable links (so you have to type in the contact email address with the keyboard, below an ironic unclickable but button-like call to action "read more on how to use memEx": no thank you!), and then put that at the very top of the page, when it's only a few lines of all lower case pointlessly centered but otherwise unformatted text, and there's no reason at all and extra effort to do that? It would have been less effort to post a blurry cell phone photo of a Word document displayed on a laptop screen. Is this some kind of Andy Kauffmanesque trolling? Vannevar Bush would roll over in his grave.
As We May Think: “Consider a future device … in which an individual stores all his books, records, and communications, and which is mechanized so that it may be consulted with exceeding speed and flexibility. It is an enlarged intimate supplement to his memory.” By Vannevar Bush
Vannevar Bush believed the scientific record had become "inadequate for [its] purpose" due to outdated methods of access and review. He warned that "the summation of human experience is being expanded at a prodigious rate," yet we still relied on systems from "the days of square-rigged ships." He criticized indexing as "artificial" and noted that it forced information into "only one place," making retrieval cumbersome. Instead, he advocated for “selection by association,” enabling users to “snap instantly to the next [item] suggested by the association of thoughts.” His proposed memex device embodied this principle—designed to be "consulted with exceeding speed and flexibility," storing vast personal libraries with the ability to “build a trail” of connected ideas, permanently linking and instantly retrieving them.
I've been meaning to make a website for this like my other project [1], but I was lazy and just posted this to HN thinking it'd get like one upvote lol. It's really more of just my personal wiki that I wanted to create specifically with the hierarchical linking/backlinking that I didn't see any other notetaking applications having just the way I wanted.
Looks like interesting idea and concept... I struggled with docker and setting variables... it seems that it is meant to be run on a server somewhere. Also unce I realized it was Phoenix it was easier but still.
Was there any environment variables in particular that were difficult to set up? I'd like to make the installation instructions as easy to install as possible!
I felt like the most difficult/frustrating setup step would be signing up for a SMTP service like mailjet, mailgun, etc in order to get emails working, but I haven't had many people try to set it up before and talk to me about it :)
Oh I also almost forgot, if you'd like a reverse proxy that's easy to setup, I'd recommend Nginx Proxy Manager [1]
Did anyone try out docs[1], the open source self-hostable project from the French government?
1: https://docs.numerique.gouv.fr/
I have yet to find a solid obsidian competitor, plugins + git repo really do cover most things.
A folder of markdown docs in your favourite text editor, ftw!
+ It’s all within the editor you already know really well. Uses your existing tools.
+ Many editors have really good support for markdown built in. Treat H1’s like notes and along with modern fuzzy search for files/symbols you can easily get to any note and jump around.
+ If you want smarter [[liking]] there’s some good plugins out there to bring this to your editor.
+ Simple, future proof and no lock-in.
I’m currently enjoying markdown-oxide, an LSP for markdown docs. Captures all your notes as symbols so you can fuzzy search and “find references”, etc. supports #tags, too.
There's definitely a lot to recommend w/ your approach, which actually overlaps a bit w/ my reasons for using Obsidian.
My obd vault is "a folder of markdown docs" (which retains the "future-proof, no lock-in" benefits you cited). But the excellent WYSIWYG UX (open files in Edit mode, w/ "Live Preview" enabled) is something I haven't seen replicated in any VSCode or Cursor extension/plugin. I also prefer a dedicated tool for note-taking and "PKM" (Personal Knowledge Management"), as a peer of my IDE(s) for coding. I get to use the best tool for the job, w/ no compromises. Switching to a different IDE for a given project (eg IntelliJ for wrangling Kotlin) doesn't disrupt my workflow, and having clear context boundaries (note-taking vs coding) is a personal preference.
YMMV, different strokes,.... I know some emacs org-mode fans out there will extoll the benefits of using "one tool to rule them all", which does sound compelling... (shrug).
I love that there are such a variety of quality tools and approaches -- and, I'm very happy with mine.
I used a similar setup for a while: Obsidian for taking notes in markdown, and vscode for coding.
Eventually I moved to using vscode for both. My gigantic notes.md file is always open in tab 1, so I can go to it immediately with Ctrl + 1.
Finding notes in a single file is easier for me than finding them in a bazillion tiny files. And there's less friction whenever I need to make a note (no need to create and name a new file).
This! I have all my notes in markdown files in folders. I self host Silverbullet on top which is a nice web UI for managing notes, tagging them etc. All tags are stored in the individual markdown files using frontmatter. I also push my notes to forgejo (gitea) which means I have versioning. Works well for me!
Silverbullet looks pretty cool; its landing page describes a feature set and workflows virtually identical to my use of Obsidian. Thanks for sharing!
I currently use flatnotes[1] as a Frontend with a background cronjob to add, commit and push every night.
Works pretty well so far with some acceptable nitpicks. It was a quick and dirty solution to overcome logseq couldn't run in a Server that time... Never needed anything else until now :-)
1: https://github.com/Dullage/flatnotes
I have fallen in love with SVG+text Mermaid diagrams embedded in markdown code blocks, and LLMs are really great at generating them (by reading design documents and source code) and editing them with natural language commands. And they're not inaccessible unsearchable dead-end images you have to laboriously regenerate and check into the repo whenever you want to change something! ;)
Mermaid: Diagramming and charting tool: JavaScript based diagramming and charting tool that renders Markdown-inspired text definitions to create and modify diagrams dynamically.
https://mermaid.js.org/
Include diagrams in your Markdown files with Mermaid: A picture tells a thousand words. Now you can quickly create and edit diagrams in markdown using words with Mermaid support in your Markdown files.
https://github.blog/developer-skills/github/include-diagrams...
Creating diagrams: Create diagrams to convey information through charts and graphs:
https://docs.github.com/en/get-started/writing-on-github/wor...
The Official Guide to Mermaid.js:
https://github.com/PacktPublishing/The-Official-Guide-to-Mer...
Mermaid looks amazing. Did not know gh had support just via
Thanks for mentioning this.Would be interested to know how this compares to logseq etc
I still feel like Roam Research did it best. Conor was truly a pioneer in the digital note taking field. I still miss the times when all of it was new, it was so exciting and everything felt revolutionary. A lot of Roam users are still using it even though the pricing hasn't changed a bit and neither has the design, but I don't think there's anything better even now. Zettelkasten is severely underrated and misunderstood. Notion and Evernote are disgusting, and PARA is a nonsensical grift. If you need more context regarding that, I recommend reading Conor's Twitter threads and watching interviews and livestreams featuring him.
What ended up happening with that weird Roam cult stuff? Did Conor finally go fully off the deep end or something?
Tried to figure it out a few months ago and it seems like they just disappeared.
I was an early adopter and "true believer" (aka paying supporter) of Roam Research, and will always be grateful for the game-changing (for me) paradigm shift it helped bring about in the PKM / note-taking space. I've long since moved on to Obsidian (which has met or exceeded my needs), but credit where it's due, Roam was transformational.
Denote, the org-mode package, takes the cake for me.
(From Protesilaos, whose introduction to emacs was also on the front page at the time of writing this comment)
I was sure Zettelkasten was going to blow my mind, but it didn't. It seems like a nice way to organize thoughts to write. I'm usually not writing, just trying to externalize my brain. And for that Zettelkasten doesn't seem to offer much over Linking Your Thinking. (TL;DR pervasive wikilinks.)
I bought “How To Write Smart Notes”, but it’s misnamed: it should've been “Why To…”. I hoped it would tell me how to use Zettelkasten, but by the end it seemed to be a long sales manual without a how-to guide.
I would be more inclined to use it if there was offline support and it wasn't so expensive. I used (and really really like) Logseq for awhile but have since moved to Obsidian. I'd switch if there was a strong reason to.
Why and how did you switch from Logseq to Obsidian?
I'm not the parent commenter, but made a similar transition (years ago at this point). I have always loved outliner formats, but the obsidian ecosystem is quite strong, and because I already had thousands of notes that were just plain old markdown, it was a more natural home.
I wrote code to facilitate the migration. Nothing too crazy, but in general I wrote scripts that:
Bugs and performance mostly. The elusive database refactor might bring me back but we'll see.
I definitely have a preference for outliner, flat zk style, but I'm able to get the majority of the benefit from Obsidian while having access to a stronger ecosystem of plugins and first class publishing and syncing support. Meanwhile Logseq seems to have lost a lot of steam.
It's all cargo cult anyway.
Was excited to click on it to begin with, but nearly jumped out of my seat when I saw that it was Elixir and LiveView.
My other app [1] is also Phoenix and Liveview, but it probably isn't interesting to you unless you have the same hobby as I do XD
[1] https://cannery.app
Found some documentation & sample website that answers a few of my questions, but I’d love to know more about pipelines and how they work with your notes. https://memex.bubbletea.dev/faq
That's actually my personal instance, but I made a random pipeline public so you can view it [1]! Here's a screenshot of the editing view if you're interested [2]
[1] https://memex.bubbletea.dev/pipeline/evaluate-life
[2] https://misskey.bubbletea.dev/files/73e9088a-a696-45de-9f17-...
How would you support someone who wanted to migrate to this and bring all their data with them? Just wondering if you build something like this hoping people start fresh, or do you build tunnels to help people migrate in and out. Kudos.
I made an export function that just basically dumps the entire schema out to JSON that can be processed with something like jq, but I should see if I can make some sort of bulk import function that would be easy to use
How does it compare to org-roam? Seems like colab and web-serving is the main diff?
This looks potentially interesting.
How does it scale? Would 200.000 notes be a problem? Does it handle pdf, epub or enex?
and most importantly
Search, what does it have and how good is it? (Is it better than evernote)?
Unfortunately, it only supports text currently :(
I also haven't really stress tested it, this was really more of a personal wiki that I wanted to create specifically with the hierarchical linking/backlinking that I didn't see any other notetaking applications having just the way I wanted. But since the application is relatively simple, I think Postgres should be able to handle the load.
It uses fulltext search built into Postgres for all the searching capabilities though, so hopefully it is good!
I assume the name is based on Vannevar Bush's Memex¹?
¹https://en.wikipedia.org/wiki/Memex
The ex suffix could also come from Elixir (its a common pattern in many libraries)
This and the previous comment are both correct :)
I'm guessing it's not my mate DC's old software company¹.
¹https://en.wikipedia.org/wiki/Memex_Technology_Limited
Why on Earth does the README start with a huge image with absolutely unreadable teeny tiny light gray text on a dark gray background?
When I press Cmd-+ to zoom into the page in so I can actually read it, exactly the opposite of what I want happens: the image SHRINKS even smaller as the surrounding text grows to push it into a black hole! So I reset by pressing Cmd-0 and then try un-zooming by pressing Cmd--, and the image STILL shrinks even smaller.
What is the point? Did a Vibe Coding Anti-Accessibility LLM do this, or was it actually intentional? That first impression makes me completely uninterested in looking at the rest of the project, rightfully afraid that it might be as purposefully badly designed and inaccessible and contemptuous of users at the first readme page.
Why would anyone bother to go to so much effort to make an image of inaccessible unlinkable extremely low contrast tiny unreadable text wasting huge amounts of horizontal and vertical white space with no usable links (so you have to type in the contact email address with the keyboard, below an ironic unclickable but button-like call to action "read more on how to use memEx": no thank you!), and then put that at the very top of the page, when it's only a few lines of all lower case pointlessly centered but otherwise unformatted text, and there's no reason at all and extra effort to do that? It would have been less effort to post a blurry cell phone photo of a Word document displayed on a laptop screen. Is this some kind of Andy Kauffmanesque trolling? Vannevar Bush would roll over in his grave.
As We May Think: “Consider a future device … in which an individual stores all his books, records, and communications, and which is mechanized so that it may be consulted with exceeding speed and flexibility. It is an enlarged intimate supplement to his memory.” By Vannevar Bush
https://www.theatlantic.com/magazine/archive/1945/07/as-we-m...
Vannevar Bush believed the scientific record had become "inadequate for [its] purpose" due to outdated methods of access and review. He warned that "the summation of human experience is being expanded at a prodigious rate," yet we still relied on systems from "the days of square-rigged ships." He criticized indexing as "artificial" and noted that it forced information into "only one place," making retrieval cumbersome. Instead, he advocated for “selection by association,” enabling users to “snap instantly to the next [item] suggested by the association of thoughts.” His proposed memex device embodied this principle—designed to be "consulted with exceeding speed and flexibility," storing vast personal libraries with the ability to “build a trail” of connected ideas, permanently linking and instantly retrieving them.
I've been meaning to make a website for this like my other project [1], but I was lazy and just posted this to HN thinking it'd get like one upvote lol. It's really more of just my personal wiki that I wanted to create specifically with the hierarchical linking/backlinking that I didn't see any other notetaking applications having just the way I wanted.
[1] https://cannery.app
Looks like interesting idea and concept... I struggled with docker and setting variables... it seems that it is meant to be run on a server somewhere. Also unce I realized it was Phoenix it was easier but still.
It is not trivial to run locally at least for me.
Was there any environment variables in particular that were difficult to set up? I'd like to make the installation instructions as easy to install as possible!
I felt like the most difficult/frustrating setup step would be signing up for a SMTP service like mailjet, mailgun, etc in order to get emails working, but I haven't had many people try to set it up before and talk to me about it :)
Oh I also almost forgot, if you'd like a reverse proxy that's easy to setup, I'd recommend Nginx Proxy Manager [1]
[1] https://nginxproxymanager.com/
As we may think.
zettELkasten
I see what you did here ;)
I initially thought it was related to https://memex.garden/ but it seems not.
As khaledh points out (https://news.ycombinator.com/item?id=43668156), more likely to Vannevar Bush's Memex (https://en.wikipedia.org/wiki/Memex).