I mean Trilium is fantastic app, lots of potential but the developer is struggling on his own, maybe it’s because it’s younger than logseq or maybe because is open source compared to obsidian. I think it’s the best note-taking/knowledge-base/second-brain i know it virtually could link everything you posses toghter to create a gigantic wiki, so much potential. Plus it has its own self hostable syncing server and web app. Guys give it a look and tell me what you think
For whatever it’s worth, this is the first I’ve ever heard of it (I thought you were referring to the IM client at first). It doesn’t seem to be on any of the popular self-hosted software lists (like https://github.com/awesome-selfhosted/awesome-selfhosted).
That was my first thought as well. IM client that I hadn’t heard the name of in many years.
That’s the problem, an unknown open source software can’t grow as fast as it should. Thanks for checking it out.
Jesus Christ, another one written in Javascript. Can’t we have something like this written in Rust, C, or Zig?
I refuse to use these things. I’m not going to host a local server just an app
Why what’s wrong with it?
I used it for a while and right up to the point I actually set it up to give it a go, I’d never even heard of it before. Definitely feels like it flies under the radar.
I loved the idea behind it, but a few points ended up sticking out so I ended up dropping it at some point. Namely, I didn’t like the markdown editor much, plus it was very awkward to use on mobile (which, granted, is an issue with most competitors). I also don’t like how it’s dabatase based, vastly prefer using local markdown files. Plus, it’s more of a personal thing but I came to prefer graph based implementations better than hierarchical ones. Still, really impressive project and I very much appreciate that it’s open source and fully self hostable, not much in the way of competition there when it comes to second brain alternatives.
After exploring a few options, I’ve ended up on Logseq. Shame that one isn’t self hostable, not fully anyway, since last I checked you still need to open a local folder even if you do host it yourself.
Thanks, actually it made a big strides towards being less hierarchical, you can put everything in a folder and use internal links now, plus it has relationship links that are very useful some times. Regarding it using a database instead of a file system it has a bunch of pros but has its cons too, in my opinion there should be a feature were you can use both and if you use the file system in one of your notes, the note is less powerful so that you can use both the database and the the file system. Lastly the phone app is a bit of a sore point for me too, I just write stuff in markdown and then import them in as soon as I can. If you want to check it out again it would be amazing if not thank you anyways for you comment, appreciate it
Thanks, actually it made a big strides towards being less hierarchical, you can put everything in a folder and use internal links now, plus it has relationship links that are very useful some times.
Ah, good to know!
Regarding it using a database instead of a file system it has a bunch of pros but has its cons too, in my opinion there should be a feature were you can use both and if you use the file system in one of your notes, the note is less powerful so that you can use both the database and the the file system.
Yeah, I 100% understand and respect the reasoning behind doing it that way but it’s just not something I can get fully behind myself. As a developer myself, I wouldn’t want to support both code paths so I also get why it isn’t really a priority.
Lastly the phone app is a bit of a sore point for me too, I just write stuff in markdown and then import them in as soon as I can.
That’s fair, do you just use something like Obsidian or GitJournal? I need to look into those at some point.
That all said, I think I’m entrenched too deeply into my current solution to reconsider but I might spin up an instance to check it out. Thanks for the heads-up!
No problem at all. I actually use logseq on my phone just because it’s FOSS.
Could you talk me through your setup? I’m interested, thank you
I have a pretty basic org-roam setup I think. I keep my org files all in a directory called “org” that I sync with syncthing (previously I used Dropbox), and whenever I setup a new machine, I just grab that folder and put it at my user root (with Dropbox I would just symlink the folder from “~/Dropbox/org” to “~/org”).
Now no matter what machine I am on and where I make my changes I have them all up to date.
I generally have large nodes that contain all my knowledge, and I split them up as they get too big. E.g I used to have a single UnityEngine node, but over time I have split it up into many different nodes: EditorWindow, ScriptableObjects/UnitTesting/etc…
I have at least one node for each of my projects, and there is a “Tasklist” heading in each of those project nodes which contain all my TODOs, those project files are tagged with the name of the project, so that I can easily write an org-agenda search to grab all the TODOs from a single project into a single view without anything else I have stored in the file (which includes a project synopsis, architecture notes/UML diagrams, general notes, etcc…).
Since I am already in emacs when I am writing code, this keep it very simple for me to have this information as accessible as all my code files are. When I discover a new language feature or have to look something up, I just open up the node for that language, and put that new information in, linking to the source where i grabbed that snippet, or where the full MSDN documentation is stored if I need to go more in depth that my short description I write it. Copying down the information helps me internalize it, and I can easily just search through that file for information I have stored. This means that even if I don’t have internet access, I have access to all my previously looked up information I maybe have forgotten.
That is very interesting, thank you very much
Looks good but I’ll be sticking with vscodium+foam
Understandable, now I’m curious about what that is, could you elaborate?
Foam is essentially core Obsidian inside VSCode/VSCodium, no access to plugins etc but there are VSCode plugins you can use as a replacement i.e. GitDocs can replace the Obsidian Git Sync plugin
If you search for comparisons most are pretty outdated now, for example Foam does have a graph view but Obsidian will still pull ahead in terms of features.
The main advantage for me is having everything in one program especially on my work PC where I can’t install Obsidian but I am allowed VSCode.
Cool, makes a lot of sense, thanks for the explanation.
Has anyone used both Obsidian and Trillium? I use Obsidian, so that’s my reference point here.
- How does using Trillium feel compared to Obsidian? Rich text is nice, but I actually like how much Obsidian gets out of your way.
Edit: Hmm, so hierarchy trees means things need to be in one place on the tree, but it can be “cloned” so it can have multiple locations on the tree. That seems like a very complicated way to day “in obsidian, you can link any note to any other note.” Am I misunderstanding anything here?
- How’s the conversion process? I have a lot of notes lol, and I find the [[Link]] format very convenient. It looks like this only wants to use text, which is fine but it very slightly breaks my train of writing thought as I prepare the link properly.
I used to use both obsidian and logseq the rich text of trilium isn’t that rich it is practically markdown but made simpler with a menu, more like the obsidian mobile app. Practically trilium isn’t strictly hierarchy based, if you wanted you could just slap everything in one folder and link what you want. The clone feature comes handy form people like me that have to have everything neatly organized so if I want a note in 2 folders at the same time I can do that. Sadly the [[link]] format isn’t supported as of now but it has its own internal linking, just press ctrl+L and link away. I hope this helps, maybe it’s not the perfect option for you but I would love if you could spread the word, Thank you very much
@da_g Yeah, I actually basically do slap everything in one folder on Obsidian lol. I love that I don’t need to spend time organizing folders, but I still find things easily by taking “paths” through my vault!
Ah, no [[links]]? Well. I’m a couple thousand notes deep on that format hahah.
I don’t think this is for me, but I’ve given your post a boost and an upvote! Good luck!
I tried that method but it hurt my brain, logseq practically forces you to use that because of its lack of folders.
For what I know [[links]] are not supported sorry, maybe in the future. But still I don’t know very much lol.
No worries I appreciate immensely the help that you gave, thanks.
Ok, this looks slick. First thing that came to mind was the IM client, which I used for a long time. I’m looking to consolidate a bunch of information (docs, knowledge base DBs, bookmarks) in this beast.
Nice, welcome the very small family of trilium.
Looks pretty cool. I need to check it out
Thanks man
Looks cool! I’ll have a look soon and share the news with people if I think it’s worth it!
Thanks man, means a lot
I know about Trilium, but never had the incentive to try it out. Maybe I will spare some time now to have a look and investigate. The sheer number of features is astonishing. Thank you for reminding me about Trilium again.
No, thank you for trying it out, as a big supporter means a lot
Thanks for sharing, as a fan of obsidian, I’m always looking for similar software.
I was an obsidian user too, thanks for checking trilium out
I may be starting a gig soon where I’ll be tasked with creating a knowledge base, this might be just the ticket. Thanks for sharing!
No problem, hope this works out well for you, thanks for the trust.
I didn’t know about it, but I wouldn’t say it’s unknown considering it has 22k stars on GitHub
What’s the main difference or improvement compared to logseq?
Well first it is more powerful, just editing files and linking can be done I a few ways, all pretty Uluseful. The main thing is the cloning of files, they are the same file in different directories plus you can self host the web app and synchronization tool and much more, give it a quick read.
Looking at this today, here are a few reasons for why it may not be adopted on a wider scale.
Looking for their web clipper extensions
- I see no extension in the Firefox store. The source code is at GitHub. Most won’t bother.
- I see an extension for chrome. Yay! However, in the permissions section, for those privacy minded, it appears to request permission to “read and change all your data on all websites.” Had this been a Firefox extension this would mean it can see all of your passwords and anything you see on the Web. This is probably also true for chrome but I can’t state it as fact. Perhaps someone can explain why this isn’t a red flag.
If you want to sync between different devices, it’s great that it supports a sync server. However, how many users are going to self host a server? 1% or less of the population? For the remaining 99%… The website appears to be silent as to how to manually sync between two or more desktop computers manually. Maybe it can be done? I don’t eat to be the first to try
Look at the details of the upgrade path. It looks rough going. And issues at GitHub reflect this . Take a look.
A single SQL lite database file hiddenly stores everything. This is scary for those of us who have lost our data on other SQL lite platforms like the darktable photo editor. SQL lite databases are supposed to be so reliable they say… Until they are not.
Markdown does not appear to be it’s standard. That can be good or bad 😔
For me that might be bad because the more privacy friendly markdown web clippers might not work . I’ll have to try .That’s a lot of friction to overcome. Not to mention the extent of cross platform support compared to logseq. Still, I might quickly try it just because of it’s code notes and math notes. I could really use these. In fact I may only use it for those and use logseq for web and general knowledge stuff.
I hope someone can correct any misinterpretations I’ve made. Please feel free. I’m trying to like Trillium.
I may have been too harsh on Trilium in my previous note.
It turns out that logseq is going to release a database version using SQLlite as well. Therefore it’s not fair to dis Trilium for using SQLlite.
Also, logseq has many documented sync problems. Logseq, at the time of this writing is alpha software IMO. Logseq can’t be considered beta because data integrity issues exist.
The issues with extension privacy I mentioned is still an issue for me but I can give Trilium a try using other extensions that copy to markdown while being privacy friendly.
I plan to try Trilium and see if I can find a desktop only solution with sneakernet / manual syncing to be safe, at least at first.
