> After some mental gymnastics weighing if I should continue with Obsidian, I found solace when asking myself "Can I see myself using this in 20 years?". I couldn't. The thought of cyclically migrating notes from one PKMS to another every 5 years, as I had done from Evernote to Notion to Obsidian, made me feel tired.
In point of fact this is actually an argument IN FAVOR of Obsidian. While the editor might be proprietary - the notes themselves are just standard markdown. If somehow all the copies of Obsidian magically disappeared off the earth tomorrow, I could easily switch over to Emacs org mode, VS Code, or literally anything else.
> Obsidian was a great tool for me personally for a long time. But I felt frustrated when I wanted to access my notes on my phone while on-the-go and saw that I had to pay for this feature.
Again, a little bit odd considering that the author is technically savvy enough to write an entire PKMS but didn't seem to consider that you can just check your markdown notes into a git repository and sync with the native android/iOS Obsidian app on a mobile device. All my notes sync up to Gitea hosted on my VPS and it works relatively seamlessly.
I'm glad the author had fun. Personally, I'm very happy with Obsidian and the plugin architecture has made it easy for me to extend it where necessary.
I had a very similar thought process about 15 years ago, and went on a quest to write my own notes system - after trying out a lot of ideas and giving up, I washed up in emacs and gave org-mode a try. It's actually good enough, and I can grep through my notes easiy, and sync them with git.
Currently vetting a way to sync my database files with my markdown files on my laptop, so it functions similar to Obsidian. I enjoy Vim too much to work constrained to Directus' markdown editor!
I saw that you didn’t want to use a 3rd party provider, but why not stick a git repo on your VPS (which you are trusting with your data today) and use that to coordinate syncs between your client devices?
I expect my PKMS to evolve and wouldn't rule out a self-hosted Git server if I find it's a better option long term.
I don't think you really get it. Git is distributed. There's no need for "a git server". You already have a machine on which you host the SQL database, you can just use that as yet another git remote.
I'd recommend you to look into SyncThing Fork or a similar tool if you never want your notes to leave your own server.
I wrote about ways to sync Obsidian here: https://bryanhogan.com/blog/how-to-sync-obsidian
100% this. The reason I started using Obsidian in the first place is that it's built on the exact directory structure and file formats that I was already using to manage my writing and notes, and if Obsidian goes away for some reason, that won't change.
Still far better than a proprietary format.
I’m all for doing projects like this as an intellectual exercise. It’s just that the motivation behind doing so in the article is a bit more “huh?”
Funny enough I had downloaded a-shell and experimented with it and going git based. But ultimately didn't want my notes stored through Github. If that way works for you, cool!
My concern with this approach would be I've read through Directus' codebase and can understand it. With a self-hosted Git server like this I'd be worried if shit hit the fan and corrupted my Git files or stopped being maintained I'd be a duck out of water
And if you really just want a simple hosting system, https://tangled.sh is really easy to set up. It uses atproto (network underlying bluesky) as their identity provider and for tracking issues, PRs, comments, etc. Their "knot server" is basically just a little self-hosted go node that manages git repos. The project is fairly small atm and it's pretty much all in go so it's not too hard to skim through if you want to see how it works under the hood (or if you are afraid of needing to be able to keep it maintained long term).
You should have the same concern with anything you're hosting yourself, and you should have 3-2-1 backups to mitigate that concern. Gitea just uses regular Git repositories under the hood last i checked, and Git is an extremely mature system; I'd expect 20+ year old repositories to work fine as long as the data are kept physically intact.
git or syncthing
(This is at least 1+ years old info, might have changed)
It will always cost more if you consider your own time for maintenance long term. Obsidian is one of the most consumer friendly business for note taking out of there, they are not VC so the Evernote comparison is unwarranted IMO.
They do also publish the “verify the encryption steps” for this.
Of course, depending on your threat model this could be insufficient, but then you probably wouldn’t trust obsidian in the first place.
Couldn’t avoid the computation panic of 2038 but it got by
As we’ve seen before, it takes one VC investment to change a source available license into something not so friendly and forks are never guaranteed.
Even if Obsidian vanished tomorrow and the application became unmaintainable, I'd still have all my notes in a text based format.
It's a non-trivial amount of money to a lot of people (myself included). I spend way more than that on Free Software, but I'm not throwing money to a proprietary program if I can choose.
Where is the limit ?
While $100/year maybe doesn't sound like much, it's hardly the only subscription service you have, and they all add up, from your mail provider, office suite, cloud storage, streaming services, phone bills, internet service, etc.
Personally I find $100/year to edit notes on my phone to be a bit much, but then again, I just use iOS Notes.
I am so fed up with everything turning into subscriptions, that I've just completely stopped buying things that are subscription based.
I understand developers need to make a living, but simply throwing a subscription on top of it won't convince me to buy your product. You convince me by making a compelling product, and by continuously updating it, adding new features, which will convince me to buy another version.
I track the ones I have so I can compare the cost, looking at either daily, monthly, or yearly cost. Sorting by price, I can look them over to judge if one of them seems unusually expensive for what it is, and regularly review to see if there are and I’m not using and need to be cancelled.
My most expensive is the could backup for my NAS. $8/month is about what I pay for Proton, which offers a lot more than just note syncing. So $8 for notes does seem like a lot. Looking at Obsidian’s pricing page[0], the $8/month is for publishing… hosting a website with your Obsidian data. Just syncing is only $4, and there are many free ways to do it. That part of the article felt like the author was trying to justify writing their own tool due to cost. That doesn’t feel justified, and they were stretching… but the good thing is there doesn’t need to be any financial justification at all. Just make your own tool for the sake of making your own tool. That’s good enough.
The biggest life hack I can recommend for a self hoster is to set up a VPN on your local network and then just never expose your services on the public internet unless you're specifically trying to serve people outside your own household.
Before I did this I was constantly worried about the security implications of each app I thought about installing or creating. Now it's not even worth setting up auth on a lot of simple services I build because if someone is able to hit their endpoints I'm already in deep trouble for many other reasons.
The security through obscurity (non-standard port, no root) are both kinda silly but why not.
That said, with awesome services like TailScale, it's pretty hard to get locked out of your network. TailScale is so so good at "just working".
I think these are decent controls when layered with others. The effectiveness differs depending on your threat models, of course, but at the very least it helps reduce the noise seen from most automated scans reducing the effort involved in monitoring your assets.
Ever since ssh almost got backdoor-ed, the only thing "exposed" on my servers is Wireguard, which is UDP based and therefore harder to know if it's running. SSH also goes over wireguard.
* LAN components are on a UPS, helps keep continuity between power blips and breaker flips
* Dynamic DNS, cron runs a script 4x per day to ensure a DNS name points to my IP, even if issued a new one by the ISP
* Rebooting everything occasionally to ensure the network and services come back up on their own and I didn't make a mistake with some config that loads at boot, etc.
Besides, I don't bother with auth for simple services, not stuff that actually hosts data. If someone unauthorized is inside my network they're not going to be interested in using my TTS/STT service or in finding out the last barcode I scanned or in using my tiny consumer GPU to generate tokens on an LLM—there are way worse things they could be doing at that point than fiddling with the many tiny services I have set up.
Also: I couldn't set up so many silly, inconsequential services if I didn't have a VPN. With my setup, every new idea I have can be a quick service on my network accessible by me anywhere in the world. If I had to expose each of these things to the internet I wouldn't bother running them at all lest they have an exploit that ends up being an entrypoint into my network.
I self-host a lot of services, and without Wireguard (or equivalent), remote access just wouldn't be realistic.
I use Tailscale, purely out of laziness and a willingness to trust them today, but I'd move to head scale if either of those caveats changed.
OpenVPN does seem to always work everywhere (presumably because outside contractors and support personnel use OpenVPN when onsite so it's a squeaky wheel that matters) so I've moved to that instead. Beyond that I can't figure out what the hell is the problem and the way IT works, they have no reason to fix it. I did get them to somewhat work on it by reporting Google VPN as randomly failing, but they just fixed Google VPN and nothing more than that. So anyway wireguard is great until you encounter bullshit corporate firewalls.
This is like the only piece of the puzzle for me.
The downsides are that I need to be connected to the VPN at home to use the domain and I currently don't have SSL set up on the domains, so browsers complain when I connect to them. The second problem I could fix, but I'm not sure if there's a solution for the first.
With modern tools like Wireguard, you can even set it up relatively easy, either as Wireguard alone or as Tailscale (or ZeroTier, though that's not Wireguard).
Wireguard (and Tailscale) allows you to setup the tunnel so that only local (RFC1918 ie) traffic is routed over it, meaning it won't eat up your battery like when just routing all traffic over it.
I have Wireguard setup like that. It enables on any Wi-Fi network that isn't mine, as well as cellular, and the battery impact is less than 2% over a day.
When I'm thinking about a hypothetical situation when I need to save the world by hacking into a hypothetical villain, my best hope will be him using your approach to security.
It certainly helps when your attack surface consists of numerous web apps of unknown quality.
Drive-by RCEs (e.g. log4j) then suddenly become much less of a headache.
For now my threat model consists of script kiddies and abusive corporations. Self-hosting gets me away from the corporations and keeping my stuff off of the public internet keeps me away from script kiddies.
notes()
{
if [ ! -z "$1" ]; then
mkdir -m 00750 -p /Users/User/iCloud/Documents/notes
Now=$(date '+%B %d %Y %H:%M')
echo -en "\n$Now\t$@\n" >> /Users/User/iCloud/Documents/notes/notes.txt
else
echo "${Now}"
cat /Users/User/iCloud/Documents/notes/notes.txt 2>/dev/null
fi
}
function nod { mg +-1 "/home/user/notes/$(date "+%Y-%m-%d").txt"; }
* 'nod' stands for "notes of (the) day" and was quick to type.* 'mg' is micro emacs, my first shell-based editor thanks to OpenBSD. The '+-1' syntax means "open at end of file" so I could easily append.
If you watch the animated gif, he is still using a third party service to store that graph.
I also think people to tend to like Markdown mostly because it’s plain text. The added benefits of that preview view is minimal. Like my gut feeling Markdown is popular 90% because of it’s in an accepted way to do plaintext and only 10% for the added formatting.
> Markdown is popular 90% because of it’s in an accepted way to do plaintext and only 10% for the added formatting.
For me Markdown allows me to write and format text at the speed of thought. Added bonus is that it's readable with "less xyz.md" or anything which can render text.
lotr-recipes
lotr-recipes/manflesh.md
lotr-recipes/media/manflesh_1.png
lotr-recipes/media/manflesh_2.png
Also makes it trivial to run a note through a static site generator and publish online.Yes, the file can grow large with many images, but it's a single file containing everything... even scripting!
TiddlyWiki is great until you want to add a structure to your Wiki. I was using it like mad, then I found out that linking pages took more time then writing notes, and I pulled the trigger and moved to Obsidian.
Markdown is great because you can easily add structure while typing compared to other format which have a more extensive markup format. I prefer org-mode because what Markdown can do, but also more extensive capabilities if you need so, but there's not a lot of editors for it especially on mobile.
In relative terms you may be right... but subjectively, having grown accustomed to Obsidian's live view in editor mode, I'd have a hard time giving it up.
Errm, no? Obsidian sync is optional. I pay for it to support them, but my main vaults are all synced by iCloud, which was auto set-up by Obisidan during initial setup on my iPhone.
On the Android side, any service which can sync files can work, I assume.
Note: Yes, I use Obsidian on my phone without sync, all the time, and it syncs.
My Office vault lives in a separate cloud service, and it works?
There are multitudes of pros and cons regarding choosing an iPhone. The restrictions of the Obsidian app is only a single one of those. Choosing an Android phone has drawbacks of its own.
Or does the iOS version of Obsidian do things differently?
However, if you don't store your vault in iCloud, it creates an Obsidian folder inside the area which can be accessed by Files app (as I just checked), which means, any application having files integration can access and sync that folder.
Even if you store your vault in iCloud, it's still accessible by any app which offer files integration [0].
But their solution is to depend on directus, which can lead to the exact same issues. To my eyes, they just added an extra step...
A gif would help clarify what your tool does. I've used an automated flow with Github Actions and Charm's VHS (https://github.com/charmbracelet/vhs) in my repo here to demo my CLI tool I built a while back (https://github.com/Amber-Williams/yall/blob/main/demo.gif). Might be of interest : )
I mean go nuts and roll your own if you want, but really, what’s not to like?
People in a similar position might be interested in Joplin, which is indeed FOSS, and has lots of sync options. I personally use SyncThing, which keeps things free, but you can also use a number of other free cloud providers. You can choose to encrypt your notes to protect your privacy.
I'm using Syncthing [0] to sync my vault between devices. On my main PC, Syncthing runs constantly in the background. Say, if I made a change, and want to send those changes to my phone, I open the application on my phone and let it fetch the changes. It's not perfectly smooth, like Obsidian's own integration, but I prefer this instead of setting a Git repository. Also, the files don't stay in a remote server.
And the notes are all just markdown files. If the obsidian software were to disappear you have all your notes. It's fine someone wanted to spend a load of time writing their own software but none of the reasons presented in this piece make sense.
I looked at my account, and I am charged $10 but it seems they automatically moved me to a "Plus" plan that has more storage. So no complaints from me really. Either that or the $4 plan is new. [1]
The $4 only comes with 1GB of storage. I would recommend the $10 for 50GB if you use images in your notes.
Why would anyone ever want to use more than one vault? I just use different folders. The only reason I can think of would be if you are using Obsidian for work where you aren't allowed to use unapproved services.
There are also several Obsidian community plugins for sync, I use Remotely Save via WebDAV.
I myself currently use Google Drive with DriveSync on Android to sync my notes, which works great. Other cloud providers also work well.
I wrote a comparison of different tools to sync here: https://bryanhogan.com/blog/how-to-sync-obsidian
It's $4 actually, for the normal plan that works perfectly well for most use cases. It's also end to end encrypted, which is great. And it's not just about syncing for me, it's about a backup solution for the notes.
> I started to have concerns about the longevity of the plugins and app itself. Some of you may remember when Evernote aggressively limited free users to 50 notes, many users migrated their notes elsewhere. I was one of those users.
The great thing about Obsidian (in comparison to Evernote), is that everything is just a plain text markdown file on disk. You can open those files in any app. If Obsidian goes away someday, all your notes can continue to be edited in any plain text editor. Sometimes I open notes in VS Code, because there are certain things I just prefer writing there.
https://web.archive.org/web/20240104200401/https://obsidian....
The profound difference lies in ontological fungibility – Emacs isn’t software you use, but cognitive clay that becomes an extension of your mind’s operating system. Where any specialized app is inevitably doomed to constrain you to some kind of constructed imagination of what note-taking/knowledge work should be, Emacs+Org erases the distinction between a tool and thought through radical philosophical pillars.
1. The Medium is the Message Paradox
Emacs rejects the app paradigm's fundamental axiom. Instead of being a "notes app" or "writing app", it's a meta-medium where:
- Your notes can spontaneously become a calendar event → spreadsheet formula → email draft → code compiler
- The act of writing is programming your environment (Org markup becomes executable functions)
- Tools aren't discrete entities but fluid expressions of your current mental state (e.g., I can run a shell command piping it to grep and then pipe the results into a text buffer)
1. Agency Through Textual Primordial Soup
By rooting everything in plain text + programmable buffers, you're working with the substrate of computation itself. Unlike database-driven apps that entomb your ideas in rigid schemas:
- Every thought remains perpetually protean – a TODO item can morph into a API documentation generator through markup alone
- You manipulate knowledge at the level of semantics (headings, tags, properties) rather than fighting GUI metaphors
- The friction between "taking notes" and "building systems" disappears – your journal entries are the configuration files of your life (I manage all my dotfiles — for Linux, Mac, home and work machines via Org-mode)
1. Compounding Selfhood
Specialized apps optimize for atomic efficiency; Emacs thrives on continuous identity investment. Each macro you write, each Org capture template, each minor mode becomes:
- A cognitive microhabitat that evolves with your thinking patterns
- Permanent infrastructure that pays compound interest (my 2010 Org config still works, while Evernote of 2010 is abandonware)
- A mirror of your epistemology – the keybindings/hierarchies are your neural pathways externalized
This creates an irreducible satisfaction: you're not just using tools but cultivating a personal universe where every interaction leaves permanent fertile ground for future growth. The specialized app user lives in rented apartments; the Emacs devotee walks through an ever-expanding mansion whose rooms rearrange themselves to their thoughts.
Do you think what you're talking about is hard to demo?
"- The act of writing is programming your environment (Org markup becomes executable functions)
- Tools aren't discrete entities but fluid expressions of your current mental state (e.g., I can run a shell command piping it to grep and then pipe the results into a text buffer)"
I haven't seen an impressive demo of this kind of stuff tbh.
I'm a regular dweller of https://www.meetup.com/emacsatx. We meet every first Wednesday of the month - if timezone permits it, come talk to us if you have specific questions. I will promise you though to make an effort to produce some demos and publish them.
The readme at https://github.com/kunalb/termdex/tree/main/markdown_files is probably the best bit.
Either way, like many others, I use SyncThing to sync my vault, and routinely edit it with vim, so Obsidian is just one comfortable shell that can (relatively easily) be replaced.
I use cherrytree currently, by the way.
For context, this is the quote that is "absurd":
> You can’t really know where you are going until you know where you have been
I feel like that'd be fine in a lot of different contexts.
Obsidian has end to end encryption and is $4 a month. I totally relate to it being fun to build your own tools but acting like it's a practical use of time... idk
You can do this with SOPS and age encryption and it's amazing, but can't view/edit notes outside a terminal or on mobile very easily that I've found.
Looking for a new solution like this, or maybe obscure configuration for an existing notes app that can support this workflow.
All of the "end-to-end" solutions seems like they just store your encrypted keys somewhere with the application files, sync them around to different machines, etc, and decrypt key with a password. But web frontends can be compromised and the master password intercepted, so I'd like to require a Yubikey touch for each document decrypt, which would make exfiltrating multiple documents more difficult.
Where I find Trilium to shine though, is when, after linking notes to one other for a little while, you realise "Well, I have a bunch of them that relate to `People`, others to `Products`, and, oh, a bunch of `Customers` as well, wouldn't it be nice if those were sharing the same properties?" (like People:{"Lives in", "Date of birth", "Partner of"…}, Customer:{"Address", "Contracts":-Multiple-, …}).
When you reach that point you can use Inheritance (from the hierarchy) and/or Composition (from Template notes) so that all your "People"-like or "Customer"-like notes share the same properties, and you can then easily manage them as data, giving the same organisational and queryable power of a RDBMS without having to commit on a data model from the get go (it evolves with you as you refine the inherited or templated attributes).
I think any sufficiently large collection of notes eventually reaches a point where it self-organises around a set of "Reference notes" more often linked to, and this is where Trilium saves you a ton of time instead of giving you more house-keeping work (good luck maintaining those "Reference notes" in sync with each-other in a system like Logseq or Obsidian, been there, done that).
I'll spin up a Trillium instance soon and see if it's still not for me (but I uh won't approach it with that mindset).
One way to go with your experiment would be to create an inbox¹ where all the new notes go (or use a day note² if you want calendar support) and a side notes hierarchy called "Collections" with subfolders like "Persons", "Companies", … each having their own Template³. That way, when you are in a new note and need to create a reference to a note that doesn't yet exist, it will pull from the available templates and create the new note from it. The template note under "Collections" will retain the backlink to all instances. Just like that you got yourself something as capable as Tana's supertags⁴.
¹: https://triliumnext.github.io/Docs/Wiki/attributes.html ²: https://triliumnext.github.io/Docs/Wiki/day-notes.html ³: https://triliumnext.github.io/Docs/Wiki/template.html ⁴: https://tana.inc/supertags
In my mind and experience, Trilium has a very unique and extensible model that lends itself to "growing with your PKMS": notes is the atom of information, attributes can be used to manage notes as structured and relational data, templates and inheritance provide structure and consistency at scale.
Trilium may not look like much on the surface, but it is incredibly capable while being approachable. Give it a serious try.
¹: https://github.com/TriliumNext/Notes/
²: you can use Trilium local-first/only, or cloud-only, or hybrid. It has its own sync protocol, you just point your instance to a server to sync with, and now you have a master-master replication. All my notes are available offline so I can keep working in-flight, notes shared with others are available via web whether I'm online or not, and I can edit my notes on the web where I don't need offline persistence. All of that is built-in/native to Trilium.
I've noticed that trillium has hierarchical notes; is there a view to look at an item higher on the tree and have it also have the contents of all its children?
You are right that the "atom" of content is the block in an outliner and the Note in Trilium. If you can tolerate⁰ the coarser-granularity, you can make Trilium behave pretty closely to an outliner: notes can be embedded within notes, either manually, or via the "Book" note-type¹ (that essentially renders a tree as embedded notes), hoisting² should be a familiar concept then.
⁰: when researching the topic, I immediately fell in love with outliners, thinking I would never go back to a note-based approach like Joplin which I was using then, but here I am, promoting a note-based solution. Metadata/tags at block level is not something I could get the hang of (I know how to manage collections of notes at scale, but not collections of blocks). ¹: https://triliumnext.github.io/Docs/Wiki/book-note.html ²: https://triliumnext.github.io/Docs/Wiki/note-hoisting.html
It's not perfect, but if I really want better search functionality, I'll just use the SQLite database that stores the notes. I've never needed to roll up my sleeves for that. I get around the limitations.
It's not perfect, but crafting one's own Personal Knowledge Management System sounds like a 5 year journey for 10 to 20 hours per week at least.
Rolling your own solution is especially limiting in the context of the sheer amount of integrations the popular ones (like Notion for example) support.
You're basically saying you will quickly build something better than the X hundred engineers at PKMS company Y quickly and it will continue to be better than what X hundred engineers will iterate upon.
I think that time is just better spent learning and picking the subset of features that, for example, Notion offers that really improves your learning rate.
That said I've played around with its API a few years ago and with page elements being block elements you need to loop through n amount of requests to get the content, it didn't make sense for my use case.
One of the things I dislike about moden software is the constant bloat and churn, because there are so many customers and so many different incentives for software companies to keep pushing features ad infinitum. In contrast, home-grown software like this has one customer and they know exactly what they want. It doesn't matter that a theoretical home-grown app doesn't integrate with the 10 social networks the user doesn't use, because it integrates perfectly with the one they do use.
This person isn't rebuilding the entirety of Obsidian, they're rebuilding the subset of parts they actually use and get value from, which is a much smaller project. By intelligently narrowing your scope like this, making stuff yourself is totally viable. Reframe "limiting" as "targeted".
After using Obsidian for a couple of years I realized use a very limited set of features - editing markdown in a directory structure.
All I needed was really the directory structure view to the left and content to the right.
Uh oh. I wouldn't use those. Of course they come and go - they're made by companies.
> Could you see yourself using your note-taking app you use today in 30 years?
Yes of course. Otherwise I wouldn't be using it.
> Do you ever have concerns around the privacy of your notes?
Not really.
> Are you spending more time setting up your notes system rather than managing your notes?
No.
> What does an effective and timeless PKMS even look like?
I use VimWiki[0]. There's a possibility it will go away, but I doubt it. There's a possibility both vim and neovim will go away, but I doubt it.
It stores everything as Markdown files. Should Markdown ever go away, it's all still very readable plain text files. I use UTF-8. Perhaps that'll go away at some point?
I version everything with git, I doubt git will fully go away, but I'm ok migrating to a different VCS if need be.
I bet the longevity of my setup is way better than the longevity of a backend written in TypeScript, backed by a SQL database, running in Docker, based on a CMS I've never heard of (Directus).
Yes, with caveats.
I actually publish my wiki on the web with about fourty lines of bash to transform the Markdown into a static HTML website[0]. So I can access it through the web browser. When people ask me for recipes or whatever, I can just give them a url.
I host one of my git remotes on GitHub (an extra backup, a service which is usually up and gives me a way to sync my notes should all my other devices be offline). I understand and admire that you didn't want to do that. Probably it's possible to install git on a phone and use a markdown editor? I don't particularly trust my phone tbh. Certainly not enough to put my git signing key on it!
I haven't found use for plugins yet since I'm really just searching, updating tasks and archiving. But if I do need extra functionality, Emacs is the most versatile editor out there, and org-mode is native to it.
> But if it's so obvious, why aren't other developers rolling out their own PKMS? Perhaps I'm the first to discover this or perhaps developers aren't writing about their custom PKMS.
Well, because of Standards, of course: https://xkcd.com/927/
I don't mean to shit on the OP's work here, but from what I can tell, the app they built is a multi-platform markdown editor and renderer that has an auth stack. Oh, and it's self hostable.
If I hop on https://awesome-selfhosted.net/ , head down to the note taking section: https://awesome-selfhosted.net/tags/note-taking--editors.htm... , I can see at least 7 that support this feature. Oh, also this category: https://awesome-selfhosted.net/tags/knowledge-management-too... has many more.
So while I think it's fun to do personal projects, I kinda feel like, if you had time to do this, it would probably have been better both for you and just like the world in general if you instead just created a PR with whatever feature you wanted on one of these more fleshed out projects. Bonus: you get a bunch more stuff, for free, since many other people are working on the same project. Bonus bonus: You can put a project with a shitload of github stars and users on your portfolio/resume/whatever and point to your PR.
Anyway as for PKMS thoughts, I've been using org mode since 2016. I've tried Obsidian and Logseq for completedness but in both cases ended up back in org mode for various reasons.
In PKMS, everyone goes on about knowledge graphs, linking etc, but I've realized lately I've never found that useful - I do use org-roam and link notes, but when I want to find links to, say, "machine learning," I'm just as likely to simply do a full-text file search for the term, which leads to the same results. As for the visual knowledge graphs, I've never seen them useful for anything other than showing off at coworking meetups.
What I've come back to is, what I really need my PKMS to do that I haven't really configured org mode to do yet for me is, in situations that happen to me CONSTANTLY when I'm out and about, I need my PKMS very quickly to answer questions for me like, "who was that guy I read recently that said something about modern capitalism causing us all to be alienated," or, "I vaguely remember reading about how social media categorizes us into advertising groups, what was that again?", or, "What was that city in Italy we went to with that crazy good ice cream? Actually on that note what islands did we go to on that trip?" I'm frequently in conversations with people where I want to share information with them, but maybe because I have ADHD brain or just am uniquely deficient and remembering very specific bits of info, I can't recall stuff (a great example of this, and I had to google to write this part: I ALWAYS forget Quentin Tarantino's name despite really liking his movies). Anyway, I tried using an org-roam org-to-html deploy tool to create a searchable, private website of my knowledge graph, and that's... fine I guess. I need to get it automated somehow, but even then I'm sure it won't be great. Of course I'm thinking of some kind of deployed solution that queries an LLM that can search my entire note repo, but that's a project and a half I don't have time to do.
So for now my plan is to just keep plugging away at org mode and org-to-html to see if I can get a really good flow there.
Would probably work with other similar options too.
LLMs are the missing piece that everyone has been desperately need to have the knowledge base come to life, instead of as a glorified key word search engine
It's quite easy to sync notes to your mobile device using a free method, or using a cloud service you might already be paying for [4].
The great thing about Obsidian is that the notes itself are just markdown files, so you can use them in any other program. This protects you as a user in case Obsidian enters a enshittification phase. A good alternative is haptic [0], it is very similar to Obsidian but can also be used in the browser. Or LogSeq [1], SilverBullet[2] and just Visual Studio Code also work well. For just editing a single file MarkText[3] is also good.
[0]: https://github.com/chroxify/haptic
[1]: https://logseq.com/
How much does your VPS costs vs Obsidian subscription? I wonder. Is it like 1 5$/month micro machine and you just pray that it will survive for 10 years without data loss?
williamsss•5h ago
busymom0•4h ago
vunderba•4h ago
williamsss•4h ago