Ask HN: Do you have a personal wiki?
I have learned that taking notes is more of a habit than tools themselves. But tools help developing that habit.
I can not assign folders to notes. Many times a note belongs to multiple folders. Wikis are not hierarchical and make more sense for notes but they are complex. They don't store data in plain text (or Markdown).
Obsidian is adding features that most editors don't support which is kind of a vendor lock in. Typora is a great Markdown editor but that's all it is good for. Logseq is weird and doesn't work properly for writing plain text.
Do you put your notes, thoughts, code snippets in folders or use a wiki based solution? If so does it support Markdown (emphasis on Markdown because it works great with code) and store data in plain text files?
33 comments
[ 3.9 ms ] story [ 86.3 ms ] threadI almost like Markdown and almost like Restructured Text. What kills the deal for me is that every dual-mode editor is slow. (e.g. like driving nails with a hammer which has a 4 to 6 second delay between the strike and the recoil)
I have a hobby of printing cards on an inkjet printer and I've written a "book" that gives step my step instructions for the whole process from image selection to installation.
I have a bunch of other things like that too.
My cards have a front and back side and often there is a QR code on the back which points to either an existing web site (a Type 1 card) or to a site managed by my system (Type 2 card.)
It's the thin edge of a very long wedge: it's a new take on the idea of "content management" because the "digital twin" of a card could be a photo shell like this
https://GEN5.INFO/$/X$Q-01:$ZG*XHWPM6/
or a blog post or a small web application. Alternately a digital object has a "physical twin" I can share with people outside social media.
I also take notes on my todo items (C-c C-z) when I start a timer for it.
Search is not a big concern, but the fuzzy search (SPC s p) doom-emacs provides is good enough for my use cases.
1. Roam by Roam Research
2. Zettelkasten there is a good book by Sönke Ahrens that covers the method in detail.
3. Org-Mode, a part of emacs. It's a text format is similar to Markdown plus a set of tools to operate on them.
4. Org-Roam, an emacs/org-mode version of Roam ([1] above)
5. Several Zettelkasten plug-ins for Org-mode.
Last year, I went down the emacs rabbit hole, because like you I wanted a better note taking system. I've been learning org-mode and emacs, but haven't yet circled back around to note taking and I can't vouch much for of the above suggestions.
Today, most of my notes are fleeting and need only to be stored on paper (and later thrown away). Everything else finds its way into an org file or flat text file.
I can say, that org-mode works even better than markdown with code. To the point that I am converting most of my linux config into one master org document, which gets parsed out into the various individual system config files.
For example, I seem to always run into issues with copy/cut not grabbing the selected text or paste putting it into the wrong spot. Typing with slashes always brings up the menu. The block structure makes managing indentions/bullets/lists a pain when I think I'm starting a new paragraph, but I'm not.
Love the software and the wiki structure, but I just can't use it for notes.
https://bear.app
https://obsidian.md
Shadow knowledge definition:
Say you are writing a GUI application and you choose QT with C++. Now that QT has good documentation but definitely not everything is there, so there is a lot of trials and errors and even more Stackoverfoo. A lot of "shadow knowledge" is in understanding the documentation in the right way, knowing which phrase to Google and applying everything you learned in coding.
However in most of cases, those shadow knowledge becomes obsolete:
Scenario 1 - This is a one shot project, and you won't come back for a long time. So these "shadow knowledge" won't make much sense when you do come back, say, after 6 months. You have to record the "meta shadow knowledge" of these "shadow knowledge". For example, why did I bookmark this stackoverflow post? Why did I write these 3 lines of code? Essentially, you have to record every thought process and hopefully the future you understnad most of them.
Scenario 2 - You continue to make QT applications, one after another, and eventually these shadow knowledge becomes second instinct and you never need to look at the notes again.
At work I have started writing down a public troubleshooting document where I write down anything which I spent a lot of time and then turned to be something simple (wrong or missing configuration most of the times).
These notes help me and sometimes other team members too. But this is in confluence only, which doesn't enforce hierarchy. I don't have a single system of my own. My notes are spread in gists, Google Keep, Dynalist and Markdown files. I don't live in console and Linux isn't my primary OS and Org mode heavily relies on Emacs.
What would be really handy would be a way to scrawl annotations on web pages so that, when [as the posters above say] I revisit a web page in future, when looking up something I've forgotten, I could see my "This didn't work!" or "Try 3rd option down" or even just some highlighted or circled text on the page, to save me wasting time. I've tried a few browser plugins which seek to provide this kind of facility but they were never very reliable. To be fair, the ever-changing nature of the web means such things are always trying to label a moving target.
Perhaps a plugin that allows you to annotate a web page, archives that version of the page and then, if you visit that URL in future, lets you know 'You already have an annotated version of this URL saved. Do you want to view it?' is the answer.
I know, this is time consuming but currently I'm using Obsidian, Foam (VS Code plugin) and Logseq all at the same time to jot down notes thoughts etc. and see which one fits my needs.
I only wish it was encrypted at rest. Otherwise, great app, with an active community.
Quick, light texts are in Notion, but eventually all important ones are moved to DT.
It is an interesting merge between manual an automatic process. I keep a Joplin database, with a PARA structure. Then I wrote a script that export all the notes from the Resource section. I then take that folder and put it into my static site builder. Et voila!
Currently I have a bookstack[0] instance on my VPS which I'm trying to cultivate as more well structured long form knowledge .
Also giving Joplin a try to see if an offline solution might be better.
[0]: https://www.bookstackapp.com/
*I have a plugin called Markdown Paste[0] to use a keystroke combo to save an image from the clipboard into a designated subfolder and insert the corresponding Markdown underneath my cursor. But it's nothing that can't be done by manually saving an image into a folder and typing the corresponding Markdown by hand
[0]: https://marketplace.visualstudio.com/items?itemName=telesoho...
Link:https://notes.rishikeshs.com
hash tags for organizing
web based simple interface
sqlite for quick querying
generate static html for sharing
> Obsidian is adding features that most editors don't support which is kind of a vendor lock in
I'm assuming you mean block level referencing? Or Frontmatter at the beginning of the Markdown? I haven't found this to be an issue - but worst case, you can choose not to use it.