Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: Obsidian – A knowledge base that works on local Markdown files (obsidian.md)
1087 points by ericax 11 months ago | hide | past | favorite | 477 comments



Maker here, excited to be on Hacker News. Obsidian is going into public beta today! (We're also on Product Hunt, come check us out!)

We made Obsidian to be your long-term second brain and personal knowledge base. As you put in more notes and make more connections, the knowledge base gets more valuable, so we think it's important that you can 100% own your data and not rely on any cloud services.

We believe you second brain should work similarly to your own brain and connections are crucial in thinking. Obsidian supports [[internal links]] between your notes out of the box, and provide a powerful graph view and backlink pane to help you understand your knowledge.

We also noticed how personal note-taking and knowledge management is, so we built Obsidian to be very extensible from the start, and let you put together your own workflow with plugins like daily notes and page preview as building blocks.

This leads to our three fundamental values of Obsidian:

1. Local-first, Markdown plain text based; 2. Link as first-citizen. 3. As extensible as possible.

Obsidian is a powerful front-end for your knowledge, like an IDE for your notes.

Learn more about Obsidian's features: https://obsidian.md/features

Read the story of the project and the team: https://obsidian.md/about


Interesting stuff!

I currently use Andy Matuschak's [1] system, using his note-link-janitor script [2] to generate backlinks and Typora to edit. The only thing Obsidian adds is the graph view for me, but it seems that Obsidian generates backlinks using file name, not title. I prefer linking by title. Perhaps this can be an option? The editor also seems to be lacking a little... for instance I can't seem to render math. Hopefully some of my feedback will be useful to you.

Overall really cool idea, but probably not going to use for now. Will keep tabs, and wish you the best of luck!

[1] https://notes.andymatuschak.org/About_these_notes [2] https://github.com/andymatuschak/note-link-janitor


I discovered andy's notes in the past and has been trying to determined what he uses to publish those clean yet powerfull notes. The janitor is only one part. do you also publish your notes as HTML? How to you make use of the backlinks generated by janitor?


It's possible to get a similar system with TiddlyWiki and the Krystal theme plus a few plugins:

https://twitter.com/Learn_Awesome/status/1265574525342793730...


Intriguing... thanks for sharing!


I don't. I share my notes with some friends as a private GitHub repo. The backlinks I just use as click-throughs to help me navigate my own notes. I too admire his notes site. Making my own is too much effort for me right now, but it is something I really want...


I believe you can use this Gatsby theme to get the behavior you're looking for https://github.com/aravindballa/gatsby-theme-andy


Woah, looks neat! Thank you!


Linking by title is an anti-pattern. Titles change and titles are not unique. Link-rot should always be prevented. Best solution is to use a uuid and hide it from the user.


This is what Quiver does underneath the covers. Every note has an unique id, same as attachments (photos, files, pdfs, etc).

My only problem with Quiver is that it seems that development has stopped, so the chances of adding new bits (like link autocomplete, for example), are thin. Other than that it's a pretty useful tool.


In Zettlr, new files are created based on a timestamp by default, e.g. 20200528171636.md. You can add YAML frontmatter like this:

---

title: Something

---

and the title will be reflected in the file navigator, instead of 20200528171636.md.


That's interesting.

I'm imagining a version that runs as a daemon, watching the folder containing all the notes. It then looks for files that have been modified, and are not currently edited (.swp files for vim, for example), and runs an update.

I think I'd prefer something running in the browser, though that is of course not ideal for several reasons...


2. Link as first-citizen.

sits up in chair, attention captured

Where can I read more about this? My current personal wiki is powered by TiddlyWiki and while I don't necessarily love the performance, I do LOVE the link structure of TiddlyWiki (I can create a "table of contents" page a random tag, and then every page using that tag gets rendered on said page). I have similar plugins for VSCode to collect all of my todo comments all into one document, linking back to their respective files.

Curious if Obsidian has a similar feature beyond the mind-map view shown on the features page?


The second box on our feature page might help!

https://obsidian.md/features


If I got you right, then you can just open the said tag in Obsidian and see all references to it as backlinks on the side panel.


Backlinks, thank you. That was the word I was trying to think of and describe exactly this, missed it by THAT much, heh-drinks more coffee. Cheers, I'll definitely be taking this for a spin


Psst, if you end up wanting to do backlinks in context by your notes in Tiddlywiki like Roam does, as shown here... https://i.imgur.com/n9ef25L.png

I have how to do that here: https://lesser.occult.institute/an-opinionated-approach-to-t... :)


Great article on Tiddlywiki, thanks! I too hate the name tiddlers but love the software so I tolerate it.


Dang, my optical nerve retracted and disconnected from the retina, and is now sitting in some dark corner of some bar in the adrenal cortex drinking whisky in pure protest to your H1 styling.


¯\_(ツ)_/¯ kids these days can't handle a little 𝕵𝖗𝖆𝖐𝖙𝖚𝖗


Kids? I'm probably old enough to vehemently request that you kindly get off my greensward.

But let us not get bogged down with ageism, the real question is what kind of monster uses both cyan and magenta text-shadow diagonally offset against each other???? !!! .... ?

:-)


Any commenter with a greensward is a friend of mine!

The answer is Google, sort of: https://developers.google.com/fonts/docs/getting_started#ena...

The anaglyph effect is what I tweaked. It's not too unreadable with simpler fonts--I 100% acknowledge readability takes a massive hit with both the fraktur and the anaglyph. Ultimately, though, my goal with that blog is mostly to have fun, so the expressive aspect is important to me. I may play around later with doing an animated effect (a la https://codepen.io/anatravas/pen/mOyNWR ) so the glitchy vibe remains strong but there are more readable keyframes. (Would a proper designer use animated text like this? Certainly not. Am I a proper designer? Certainly certainly not.)


I'm all for artistic freedom. Just letting you know in a jestful way that the design (and I am certainly not a proper, nor any other kind, designer either) that the design detracted/distracted from the message for this reader, to the point where I just hit back.

Then I went back in again, and started poking around with the element inspector. Then we had this discussion, and I am still not sure what the page content was actually about, but I have a sense that I might have found it interesting if I had stuck with it.


Fair enough; that artistic aspect means that for me, the design is part of the message so it's worth keeping. I think I do a similar thing with unstyled websites a la https://motherfuckingwebsite.com/ ; the content has to look very interesting to me to be worth copying and pasting in CSS a la http://bettermotherfuckingwebsite.com/ to actually read the thing. It's interesting to think about the wide range of tolerances out there for different kinds of visual presentation.


Looks pretty great to me. I have a great book called Fraktur Mon Amour. Unfortunately it is out of print but if you can still get it somewhere I wholeheartedly recommend it.


Powell's will send me an email when it's back in stock. Thanks!


In all fairness, they're pretty hard to read.


Yup, acked in another comment. The "k" is especially out of the norm, so with the colors as well, it's Not Ideal


Do you know a single plugin to get the two columns layout of stroll?


Yup! The cool thing about Tiddlywiki is that because it's one massive blob that executes clientside, and because it's designed to be extensible, you can yank the plugin Stroll uses right out of Stroll. If you go to https://giffmex.org/stroll/stroll.html and click plugins on the sidebar, you can see a list of installed plugins; Stories is what he's using to get the second column, and it shows up with this link: https://giffmex.org/stroll/stroll.html#%24%3A%2Fplugins%2Fsq...

Click and drag that to the tab with your own TW open, and you're good to go!

I haven't used that plugin so I don't know if it's good or whatnot, but there you are.


Thank you!

I did not think of grabbing it out of stroll. Tried searching for plugins but never encountered the "Stories" plugin.


Done, cheers for that!


> 3. As extensible as possible.

The app would need to be open source for this to be possible.

It's still possible to make money with an open-source product. When you're targeting a developer audience, it might even be more profitable to be open-source.


proprietary software can totally be extensible, that's what plugin APIs are for. foss is cool and all, but i don't buy the implication that it's some sort of civic duty for developers to release under a foss license.


Is an open-source piece of software more extensible than closed-source? If so, then if the goal is to maximize extensibility, then open-source is required.


Open source software is readily and legally user modifiable. It is often (but not always) readily extensible - it can actually be quite difficult to extend in some cases. (There's plenty of difficult to work with spaghetti on GitHub.)

Source available proprietary software is readily inspectable, but often not legal to modify (at least not for free). It might still be extensible (ex plugins) though!

Closed source proprietary software is difficult to inspect, difficult to modify, and generally illegal to modify (for most use cases, in most jurisdictions). As above, it could easily be extensible though.


Hardly. Extending a piece of software requires generally extensive familiarization with the codebase, possibly ranging into millions of lines.

On the other hand, a well documented plugin or scripting system, which sits on top of the existing domain logic and is well documented and full of examples, generally is an excellent way to allow extending the base app.

The base app can be open or closed source. Without scripting or plugin system it's still a black box for most intents (as the time needed to study and change it would likely be too much).


A piece of software being open-source does not preclude it from also having a plugin or scripting system.


Sure! But the question was "is an open source software more extensible...". And for most situations where software is used the answer is no.


If being open-source improves extensibility by even 0.0000001%, then to make something as extensible as possible requires it to be open-source.

As someone who writes plugins, being able to see "behind the curtain" is very helpful, especially for taking advantage of un-documented features.


someone created something that they'd like to make a little money off of, while still taking very extensive steps to make it available, extensible, and prevent lock in. they chose a path to profit with all of these goals in mind. you're demanding an awful lot for something that "improves extensibility by even 0.0000001%". in the spectrum between "i want to extract money from the users of this product" and "i want this product to be used as the users see fit", i would say this product's goals lean towards the latter.

please don't say the wording is the problem. "as extensible as possible" is fine. "as possible" is a qualifier, it clearly means "as extensible as possible without undermining other goals". you're pretending that they've promised to make it extensible at the expense of _everything_ else, and that assumption in context in unfounded.

it's true that paths to monetization exist for open source software, but they usually aren't accessible to an individual developer who is building a small bootstrapped side-project to generate small amounts of passive income.


obviously the "as possible" is qualified.

setting that aside, though, the best way to make software (open- or closed-source) extensible is via a plugin API. without it, the only way for end users to extend open-source software is to fork and maintain the fork forever, or attempt to merge upstream. so one might argue that the quality of the plugin API (and its documentation) is the primary measure of a software's "extensibility".


Hah, is it not opensource? If it really is closed then it's useless, not worth discussing. Then it's funny they maket it as being better than clouds, that clouds shuts down, change owners or policies. With closed source software they will be able to do the same. Yes, markdown is plain text but as more and more additional features would be used, one would get more vendor locked-in to this software. And then what - write an opensource alternative with all the features from scratch again. Clouds at least give online sync...


Early user here. Obsidian is literally the best app in 2020 so far IMO.

- Blazingly fast

- Clean UI

- Free

- Sync with Dropbox, Github, iCloud...

- Great community

I have never looked back to Notion and Bear since I found it.

Btw, the Obs team is moving so fast.


> Free

Not really. The license says you are not allowed to take notes about work you get compensated for. So free only for 100% hobby projects.

Working in a start-up I don't have big 100% hobby projects that would require a lot of note keeping. YMMV


I'd be more concerned if you had a hard time getting reimbursed for this kind of software at your job.


What's the concern for? A lot of great jobs still have penny-pinching managers. I've worked somewhere notoriously Frugal that was good and paid well but had no problem denying requests like this.


It's sometimes not just about money. Any organisation that alows individuals and teams to freely pick and choose tools will end up with a huge pile of semi-forgotten systems and organisational knowledge spread over multiple wikis, sharepoint sites, trello boards, monday projects and word docs on shared drives and dropboxes. Seen it happen multiple times. Corparate use of "personal/hobbyist" free tiers of popular tools is not just a legal liability, it's also a security nightmare apart from the already mentioned infromation spread problems.


That's a super helpful counter-point, thanks!


I'll respond with a connected set of questions...

Is it worth the time to deny a (well intentioned) $150 request made by a software developer who costs a company (for example) $150K total (salary, benefits, and fixed costs)? $150 is 0.1% of $150K.

What are the benefits and costs? Some questions to ask include:

- How much time did the company spend in the process of denying this request? By the time you include emails, research, context switching, etc, I think it is fair to say this might consume 15 minutes of the engineer's time and 15 minutes of the denier's time. That would be 0.5 hours; assuming $80/hr, that is $40.

- What is the likely effect? Will it discourage an employee from trying a new tool? Might it encourage a culture of "build it here" rather than pay to use something that already exists? Might it encourage an employee to abuse the license? Might this increase the legal risk of the company? In the case of, say, online learning, it might encourage employees to browse dozens of crappy web resources rather than simply paying for a high quality learning resource. My point: being stingy has real effects on human behavior.

- Does the company pay the same level of detail to other areas that could easily save more money? Three examples: (1) Does a company need to spend $200/year/person on junk food or bottled water? (2) Does a company look carefully at ways to improve energy efficiency? (3) Does a company have a smart, regularly practiced, data-recovery plan in place? I could go on; I think you see my point: it is wiser to allocate effort and oversight in proportion to impact.

- To what degree does the company have issues with trust and accountability? What is the effect on morale are discouraged from trying and paying for useful software and tooling?

I think my overarching point is, again, illustrated by questions: What does a company truly value? Are they mindful and realistic about their costs and benefits?

P.S. This may be obvious to some (but not all): paying for software is not necessarily a bad thing. Open source has many advantages, but without ongoing contributions and/or a funding strategy, open source software is not necessarily a "safer" bet than closed-source software. A better litmus test is "can I export my data in a useful way if I decide to leave or switch?"


This seems like an argument for buying SDEs needed software. It doesn't seem like an argument against the idea that you can work somewhere where they refuse to buy things like this, and still have it be a good job.

If your job pays well, work is interesting, good commute, good benefits, good manager, good coworkers, important mission, or a net positive combination of these things, that seems way more important than whether or not they'll buy you arbitrary software.


> This seems like an argument for buying SDEs needed software.

It depends on what the company values. This is what I meant when I wrote: "What does a company truly value? Are they mindful and realistic about their costs and benefits?"

> ALittleLight: It doesn't seem like an argument against the idea that you can work somewhere where they refuse to buy things like this, and still have it be a good job.

Correct, I made no such argument. You are free to make that value judgment.

It depends where you sit. Maybe you want to dig into ways an organization can improve? If so, that gets into questions about organizational values as well as costs and benefits of various options.

> ALittleLight: If your job pays well, work is interesting, good commute, good benefits, good manager, good coworkers, important mission, or a net positive combination of these things, that seems way more important than whether or not they'll buy you arbitrary software.

Again, feel free to make such a value judgement.

However, I would not use the word 'arbitrary' here, since in employment situations, there will be some understanding around expenses, often set out in policies and conversations. Even in organizations that are more flexible with expenses, employees are expected to use good judgment for business expenses.


I'm curious: were you expecting that my comment offer "an argument against the idea that you can work somewhere where they refuse to buy things like this, and still have it be a good job."?

If so, why?


The thread, as I understand it, goes something like:

1. The company should buy you this.

2. They may not.

3. If they didn't, that would be a concern.

4. Why?

5. Argument that the company should buy extra software.

Point 5 is off topic. There might be a good job that wouldn't buy you extra software and that wouldn't really be a concern.

It's like if someone said "Your job is bad if they don't offer free lunch". I might say "My job doesn't offer it, but it's not a concern because I like my job for other reasons" and your reply would be advocating for the benefits of free lunch. Free lunch might be great, but the topic is whether it's a concern that the company doesn't offer it, not whether it's great or not.


Another response: your argument style reminds me a little bit of a straw man:

  > A straw man (or strawman) is a form of argument
  > and an informal fallacy based on giving the 
  > impression of refuting an opponent's argument, 
  > while actually refuting an argument that was not
  > presented by that opponent. - Wikipedia
... with the twist: instead of engaging with anything I said, you mischaracterize it and large parts of the preceding discussion. Then you use that as a way to dismiss what I wrote as "off topic".

I'm going to have some fun looking up what others call that kind of rhetoric.


Ah, this appears to be the core of it: according to your interpretation of the thread, my comments are off-topic.

After re-reading the thread, it is clear that your 'understanding' (as written above) of the thread is inaccurate.

I'll annotate the first five comments in the thread, with your 'understandings' and my responses:

  Original post:
  > allenleein: "Early user here. Obsidian is literally the best
  > app in 2020 so far IMO. -Blazingly fast -Clean UI -Free -Sync
  > [...] -Great community"

  Your 'understanding':
  > 1. The company should buy you this.
Your understanding is inaccurate: allenleein does not say a company should buy Obsidian.

  Follow-up comment:
  > usr1106: "[it is] Not really [free]. The license says you are
  > not allowed to take notes about work you get compensated for.
  > So free only for 100% hobby projects.
  >
  > Working in a start-up I don't have big 100% hobby projects
  > that would require a lot of note keeping. YMMV

  Your 'understanding':
  > 2. They may not [buy you this]
Your understanding is inaccurate: usr1106 says nothing about what an organization should or should not buy. A better summary would be: usr1106 disagrees about what 'free' means, explains the license, and does have hobby projects that would qualify as free.

  > codezero: I'd be more concerned if you had a hard time getting
  > reimbursed for this kind of software at your job.

  Your 'understanding':
  > 3. If they didn't, that would be a concern.
Fair enough.

  Follow-up comment:
  > libria: "What's the concern for? A lot of great jobs still 
  > have penny-pinching managers. I've worked somewhere notoriously
  > Frugal that was good and paid well but had no problem denying
  > requests like this.""

  Your 'understanding':
  > Why?
Your understanding is incomplete. In addition to asking why, libria also offers a framing and makes a value judgment about what constitutes a good job.

Some of your other comments in this thread refer to this framing and value judgment. That's fine, but other comments are in no way obligated to agree or buy-in to that framing.

  Follow-up comment:
  > xpe: I'll respond with a connected set of questions...
  > 
  > Is it worth the time to deny a (well intentioned) $150 request
  > made by a software developer who costs a company (for example)
  > $150K total (salary, benefits, and fixed costs)? $150 is 0.1%
  > of $150K.
  >
  > What are the benefits and costs? Some questions to ask include:
  >
  > [... time spent? likely effects? consistency in other areas? ...]
  >
  > I think my overarching point is, again, illustrated by questions: 
  > What does a company truly value? Are they mindful and realistic
  > about their costs and benefits? [... P.S. ... snip]
 
  Your 'understanding':
  5. Argument that the company should buy extra software.
Your understanding is inaccurate. I asked many questions that don't give one particular universal answer about purchasing; the questions, I hope, suggest an approach to finding an answer that works for you.

A charitable reading (see the HN Guidelines) of my comment would see that I was responding to this part of libria's comment: "What's the concern for?" To put it very simply, I would be concerned by a company that was not mindful and realistic about their costs and benefits. Why? It is simple: I value working for mindful and realistic companies.

In summary, your understanding is inaccurate.


Because it seems like you are talking to a dev, not a manager. Assume the dev is already onboard with the idea of being compensated; the original comment was "I'd be more concerned if you had a hard time getting reimbursed" i.e from the POV of the employee.


I don’t understand.

Is it fair to say you (Chris2048) have some expectation around how the thread evolved based on your assumptions of what is relevant?

I see the best conversations here as trying to understand each other.


I don't understand what you are asking.

I am not the same person, but I'm not sure that matters unless you know of OP specifically, you are asking: "how was my intention interpreted as X".

Im telling you the context of the thread implied it, otherwise your comments arent relevant to the POV of a dev.


Yes, I saw you aren't using the same account name. That's why I asked.

At the risk of saying something we already know, HN discussion isn't limited to one person's definition of what a "hacker" or "developer" is:

  > On-Topic: Anything that good hackers would find
  > interesting. That includes more than hacking and
  > startups. If you had to reduce it to a sentence,
  > the answer might be: anything that gratifies one's
  > intellectual curiosity.
I think it is easy to forget, so I just want to write this here... People here may work on building technology (hardware, software, biotech), running companies, leading teams, thinking about problems to solve, how to be effective, how to deal with stress and mental issues, and lots more. So, talking about organizational culture is on-topic.

You may prefer to discuss something from the point of view of a software developer. You may have desire to keep threads organized and on-topic.

Personally, I think your assessment of the "context of the thread" is both overly narrow and off target. But my goal is not to convince you my interpretation is correct...

...My goal is to show that your interpretation of the context of the thread is subjective. Again, subjective is fine; we don't need to agree. I want to emphasize that reasonable people can see it differently from you. I hope that you (and everyone on HN) can recognize this and think it through before they say a comment is "irrelevant".

So, forgive me for asking, but I can't help but wonder how you approached this thread. With curiosity? With a goal of understanding? With some other driving factors?

I would like to highlight a few important points from the HN guidelines that think we can all learn from (myself included) ...

  > Please don't post shallow dismissals, especially of
  > other people's work. A good critical comment teaches
  > us something.

  > Please respond to the strongest plausible 
  > interpretation of what someone says, not a weaker one
  > that's easier to criticize. Assume good faith.

  > Be kind. Don't be snarky. Have curious conversation;
  > don't cross-examine. Comments should get more 
  > thoughtful and substantive, not less, as a topic gets
  > more divisive.
On a personal note, I put considerable thought into my comment -- phrasing it with open-ended questions where I implied some arguments in play. I offered food for thought. I was hoping for substantive responses, not claims that the comment was off topic. In addition, I found some of the resulting comments to be unconstructive. I think we owe it to each other to help each other and make the most of our precious time here on the planet.

One motivation for reading this comment, frankly, is a plea for people to re-examine their communication tendencies and the resulting effects on (a) your ability to learn (e.g. a growth mindset); (b) this community; (c) all communities.


Can be a bit more specific? Your words imply accusations without explicitly making them, and it comes across as two-faced to me.

* why are you highlighting those HN guidelines?

* who are the "people" you want to "re-examine their communication tendencies", and what motivates this plea?

* why are you curious about how I "approached this thread"?

That aside,

I answered your question: replying to a thread in an ambiguous or open-ended manner will cause people to fill in the gaps (infer your meaning) from the context of the thread. If your meaning does not follow from the context (i.e. is a non-sequitur) it's likely you will be misunderstood; In this case that you were offering "an argument against the idea that you can work somewhere where they refuse to buy things like this, and still have it be a good job."


Responses to your bullet points:

Re: 1 & 2: I would like HN participants to consider the HN guidelines, because I often see what appears to be a lack of awareness. Following the guidelines (many of which are about self-reflection and tone) helps shape the community discussion constructively.

Re 3: I asked because I want to understand your motivations here.

Yes, I understand the general idea of non-sequiturs.

In summary, I think an accurate and charitable reading of my comment will realize that it was not a non sequitur nor off-topic.

Above your comment said:

> Im telling you the context of the thread implied it, otherwise your comments arent relevant to the POV of a dev.

My comments are relevant to the point of view of a developer. Moreover, HN discussion is about more than the POV of a developer.


Why did you to cite those particular guideline in this specific thread? Why ask about my motivations.

It seems like you are dodging the question(s) i.e If you have some generality about HN participants, why put it in this thread (and not others). Since you don't put it in your profile, or copy-paste it in every comment you make, it seems to me there's a reason.

> In summary, I think an accurate and charitable reading of my comment will realize that it was not a non sequitur nor off-topic.

Another back-handed response, as it implies my own comments (which made the opposite conclusion) is therefore either/or not accurate, or not charitable. If you believe this, then why not explicitly say so - and then defend that position? you say "In summary", but I can't see what part of this post you are summarising.

> My comments are relevant to the point of view of a developer. Moreover, HN discussion is about more than the POV of a developer.

They might relevant, if there is enough context to understand them. And we are not talking about what is relevant to "HN discussion" - we are talking about this thread in particular.


I don’t know you or anything about your life experiences. I asked about your motivations because I was curious.

My goal here is to use a calm, measured language. I was hoping this would help the conversation, but perhaps it upset you. You called my comments ‘back-handed’ and ‘two-faced’. I didn’t intend them that way.

You could have chosen different words. You may realize the words you chose were harsher than necessary. Even if you were correct in your assessment, which I don’t think you were, those choice of words will likely have a negative effect in a conversation. Especially online, particularly with someone you don’t know.

BTW, I am genuinely sorry if you think I’m trying to insult you in an obscure or sneaky way. I’m not. Doing that would be unkind.

Speaking of your claims that my comments were ‘two-faced’ or ‘back-handed’, there is another explanation. (Skip two paragraphs down for that)

If there’s one thing I could get across to you, it is: please open your mind to other explanations. Be charitable towards others. Don’t assume malice.

If you think you are already as charitable as you can be, then I don’t expect this advice to bother you. If you feel bothered by it, perhaps you should take a closer look at yourself. (I’m not claiming that I am perfect in this regard. It is a process.)

You might have reached the point in life when you realize and respect that people have different communication styles. Many people may not be as direct as you would like.

You say I ‘dodged’ your question. I hope you realize there are other ways to say the same thing with nicer connotations.

You also may realize you didn’t answer my questions, which I asked first. I don’t mind if you don’t want to answer.

I’ll try to phrase my thinking over the last few messages in a different way. My take is that many of your claims are overconfident, possibly because you aren’t actively asking yourself ‘how might other people see this’.

I think a big reason I’ve been replying is out of some (misplaced, perhaps) desire to help you. I think you would benefit by finding more ways to understand other people’s points of view.

I will admit, you seem capable of arguing just fine. So, I don’t see intelligence being a limiter. I would guess (with about 75% probability) that a lack of empathy is a limiter for you.

This is not meant to be harsh even though it may be direct. If true, you certainly aren’t alone and you definitely aren’t alone in a community of technical people. There’s plenty of rationality and technical knowledge but too little empathy.

Example in point: You did a nice job of criticizing my use of ‘in summary’. I’m both joking and not. My usage could be improved, but I think the intent was clear.

Based on what I’ve seen in your behavior, I predict you will reply. However, I don’t expect it to be much different in tone. Feel free to surprise me!

In any case, maybe you will check back in a few years and re-read this thread. Maybe you will see it with new eyes. Maybe it will be some value to you.

Just so you know, if you reply, I don’t expect to reply in timely manner (or ever). So, feel free to have the last word here.


I was asking ALittleLight.

Chris2048: are you the same person?


It's not like there's a realistic way to enforce that license, so it's essentially free.

Edit: typo


Very few licenses are easily enforced, but that does not entitle folks to violate them. Most desktop software can easily be found with the copy protection removed, but that doesn’t make it free - there’s no moral difference between warez, GPL violations, and violating a free-for-non-commercial-use license.


Curious if you use your markdown files on mobile and if so, what tools do you use? I’m still using bear because of their iPhone/iPad apps.


Not GP but Markor is fantastic for anyone running Android (https://f-droid.org/en/packages/net.gsantner.markor/).


You should try NotePlan https://noteplan.co/ - pretty decent search, tags, markdown-like syntax and calendar on both mac and iPhone apps.


How does it compare to Roam?


1. Obsidian makes use of a local folder of Markdown text files, rather than syncing to a cloud database like Roam.

2. Each page in Roam is an outline (as in outliner), whereas each file in Obsidian is a Markdown file.

3. #tag is not the same as [[tag]] in Obsidian.

4. Rather than a built-in feature, "daily notes" is available as a plugin in Obsidian.


Wow, based on your history of submissions you've really been obsessing over this problem for quite a while. Kudos on continuing to improve your products/positioning.


Hi Eric(?)

I downloaded Obsidian and it looks really nice. I use markdown already to make notes on my side project, but in my brain my thoughts are more diffuse as it's an MVP stage thing with so many ways to extend, so many things to think about. I think this would, at firs glance, help me alot.

One small thing I'd recommend would be when you first open it, could it create a default folder like (on Windows for example) Documents\Obsidian. And pre fill that as the folder (and I can change if I like). Then when I first use it there is less friction.

The other thing is to UX test that first screen where you are made to choose between new document and reading docs. I felt that made me think too much, and it might be the wording. I'd sort of prefer something like:

New User? Do you want to read help on getting started? Yes/No.

However I might not be typical, so watch over some shoulders as people first use the app and ask them what they are thinking.

Good luck!



Thank you so much for the tips, they make a lot of sense! Our onboarding surely needs some re-thinking.


I've been looking for something like this for... so long. It's like tiddlywiki but with first class & local markdown! Thank you!!!

Add a separate GUI for a workflow-y like list-editing (that saves as a markdown list format) and you've got a serious competitor to Roam as well as beating out more traditional competitors like Notion, Boostnote, etc.

e: ah, no inline LaTex.. I knew there was a catch!


> Add a separate GUI for a workflow-y like list-editing (that saves as a markdown list format) and you've got a serious competitor to Roam as well as beating out more traditional competitors like Notion, Boostnote, etc.

FYI, they're the ones behind Dynalist.


Can you clarify who "they" is? Are Dynalist and Obsidian made by the same folks?


AFAIK yes.


Dynalist doesn't have the first class local markdown support like this!


> e: ah, no inline LaTex.. I knew there was a catch!

Sadly, this makes it completely useless for me.


Agreed, although according to a reply a fix is in the works.


The catch shall be fixed soon! :)


This looks really cool! I can't wait to check it out.

I've been using various personal wikis for years, and this hits most of what I am looking for.

But, I'd like to offer some suggestions.

- It would be cool if you would consider some tweaks for how the text is rendered. It can still be Markdown compatible, but for example allow linebreaks to be interpreted as linebreaks. Text used for reference often has different needs than what is being exported to HTML.

- Git integration. I just use git for syncing my knowledge bases. It makes the most sense for my multiple devices, and it is very rare that I even have conflicts and can't automatically merge differences.

I actually have a script that lets me use Dropbox or another syncing provider to sync my working tree, separate from the repo itself, so that my history isn't polluted with excessive automated commits, but it is still tracked relative to where it was checked out, and resolved automatically. That way you can have the strengths of git without the drawbacks.

You may also want to check out git-annex/datalad. I combine it with my home grown Markdown wikis for embedding references to files in my wikis, keeping my git history just pure text. One of my goals is to bridge file management with text management.

I'd be happy to share any of this if you're interested. I've had plenty of free time recently to further develop it.


+1 for a git-annex integration! I'm hoping this may come to pass at least as a plug-in.

I'm pretty much ready to abandon org-mode for this, but the git/git-annex combo is harder to leave behind.

Would love to hear more about your setup!


This looks incredible, I’m going to try it out today.

One suggestion - would you consider adding videos/gifs of your product on the features page? I feel like that would demonstrate your product much better, especially the linking part.


Downloaded this a few hours ago, and ported all my notes about my novel over to it, and I have to say: this is a really great piece of software. The interface is intelligent, ergonomic and fast, and it does a lot of things automatically which I really like. It cuts down on the boring book keeping.

I was also able to move my plot outline into this, which was a godsend for me because previously it was in a word document, meaning I couldn't check things off. Now I have an interactive task list for it! And everything is all in one place!

One small suggestion: I would really like some place where it lists all the key commands (for instance, are there key commands for making headings? what about the key command for Replace, which I couldn't find?). Also, it would be nice if tags weren't necessarily represented inside the document from a viewing perspective. Maybe as a bar along the bottom, which is then copied into the markdown document transparently?

Thanks for making this!


Settings -> Hotkeys


Why isn't the snap package signed?

    $ snap install obsidian_0.6.4_amd64.snap 
    error: cannot find signatures with metadata for snap "obsidian_0.6.4_amd64.snap"


Apologies for our lack of experience on Linux, we'll look into it!


You'd need some snap store to be able to install it without using --dangerous flag. Snapcraft.io is Ubuntu's (and most popular one), though that might not work for you if you want the download to be available exclusively from your own website.

But then again, if it's on Snapcraft, you don't need to worry about distributing updates to your users.


Might I suggest distributing your app as an AppImage? snaps require the snap package manager (snapd) to be installed on your system. AppImages, on the other hand, run out of the box after a chmod +x.


Are you going to add some features for more structured data? I'm thinking about tables like in Notion where you can sort by any column. Working with Markdown tables is very clunky.


The closest thing in Markdown is probably front-matters. You're right in that Obsidian is not great for structured data right now, and for that purpose Notion or Airtable is much better.


If you need to edit tables then you could just use TableFlip, gives you a spreadsheet style interface to edit tables, automatically updates your markdown file without the need to copy paste https://tableflipapp.com


That might be a good candid it feature for a plugin when the API is released. It was mentioned before in the discord.


Not sure if this BoneAppleTea material or DamnYouAutoCorrect :-)


Have you any good resources for getting started with this graph-type note-taking? Currently I just used notable.md in a NextCloud folder and it works pretty slickly.


Not aware of many good resources since it's pretty new, but you can probably look into the Zettelkasten method.

The gist is to make your notes atomic and link them together whenever possible. [[ links provides auto-complete, and the backlink plugin tells you what notes links to this note, and what notes mention it but don't explicit link it.

The graph view is for discovery and navigation mostly. It lets you discover clusters, and identify "orphan" nodes that aren't connected anything. That might inspire you to think of connections and strength your knowledge network.

Hope that makes sense!



Unless you follow the semantic wiki model (https://en.wikipedia.org/wiki/Semantic_wiki) rather than the plain wiki model what the user is building is neither a knowledge base nor a second brain. Nearly all software in this area makes this mistake so it's not your fault. Links in human brains are typed (not untyped) and the regular wiki model (and web, and Obsidian) only support untyped links. If you are truly aiming to deliver software that allows users to create a personal knowledge base then please consider spending some time researching semantic web technology.


I just installed it, and it looks pretty slick, congratulations on the launch!

One piece of feedback, I couldn't discover in the app itself how to make references to other documents. I finally figured it out by looking at your web page and seeing the [[connections]] bit.


Had the same issue, found it on this HN post.


> Obsidian is a powerful front-end for your knowledge, like an IDE for your notes.

Could you describe Obsidian's pros'n'cons in comparison to VNote[0] + Viki[1]?

[0] https://github.com/tamlok/vnote

[1] https://github.com/tamlok/viki


this is awesome! i see the 'zettelkasten link fixer' in the markdown format importer ("Fixes [[UID]] links to full [[UID File Name]].") - any plans/thoughts on just supporting [[UID]]? this seems less brittle to me in that you won't break all of your backlinks if you change the (non-ID) part of the filename.


Obsidian will auto-update all backlinks when you rename it.

Essentially, we think wikilinks should identify the destination exactly and should not perform a search instead.

Thankfully it's all just plain text at the end of the day, and it's not hard to do some text processing to convert both ways.


+1 for this request, which would allow better interoperability with "The Archive" (from zettelkasten.de). I built my Zettelkasten links in "The Archive" using [[UID]]. Obsidian looks fantastic and I'd love to give it a try but my existing links do not work. As a side note, the other feature from "The Archive" I miss is the ability to create buttons for custom searches in the left sidebar.


Could you add the releases on Arch Linux's AUR? Downloading a binary from a website is not something Linux users often do. I couldn't find Obsidian on the snap store[0] either.

[0]: https://snapcraft.io/search?q=obsidian


Ive been using Trello Card's markdown notes and the connecting cards feature to approximate obsidian's main features for years now. Really love it.

Very interested in the graph and multiplexer features...

A Trello integration would be nice.


Awesome work guys! Judging by such an active and enthusiastic community, I could see this having a very sizable market share of knowledge bases like roam / bear / notion in the future. So far I haven't found a good filesystem based note taking app that fit my needs. Quiver came the closest but has not been active in development recently, so I'm switching to this and I'm excited where it goes.

You and Shida have some serious product & programming chops ~ UWaterloo represent!


can you do a quick and dirty screenshare demo of how you use it and post that to youtube? e.g. setting up some notes after a fake brainstorming session or micro LSD dose etc..


There are a few videos made by beta testers already, but it's a good idea to have some official ones. We've just been too busy shipping releases...

Some videos that might help: https://www.youtube.com/watch?v=cFYaWC_86W0 https://www.youtube.com/watch?v=KAkJMHg-dGw https://www.youtube.com/watch?v=rh2ysYig8Wo


Obsidian seems very cool. One thing I noticed on install as someone with a Windows device is that the install location cannot be specified. It would be great if your team supported this as it's pretty standard for many Windows apps and aids in system organization. I was able to move Obsidian and the obsidian-installer and things seem okay but I have no idea if this will break the updater at a later point in time.


Obsidian looks great! And I think it will be greater if I can write `[](another_markdown)` as internal links too, it will be easier to generate HTML files to host, using Hugo or something. If people save all markdown files in one folder then it's the same word count as [[another_markdown]].


I would just like to point out John Gruber disavowed the []() syntax on Twitter. He even said he would like to accept reverse brackets, since it's such a common problem.

I have been using personal wikis for about a decade, and 99% of the time I make an internal link, I want the caption to be the same as the page name.

I wish that the Markdown standard would just specify to render [[Page]] as [Page](Page).


To be fair, John Gruber has disavowed a lot of modern Markdown features and syntaces, so it's not exactly an authority statement anymore for today's Markdown. It's why CommonMark exists.


Yes we do support the standard format, navigation will work in Obsidian but unfortunately not link auto-completion.


What do you think about CMS usecases like blogging? Where you'd write locally and maybe define some notes or a folder for static site generation, maybe triggering rebuilds from within Obsidian?


We designed the "Publish" add-on service for that purpose (not yet available): https://obsidian.md/pricing

We still want to give you the option to export internal links to standard [Markdown](link) links so that static site generators or sites like GitBooks can understand them out of the box.


Good to know, thx.


Why not just use Jekyll or Hugo at that point?


Idea was to use that feature with Jekyll or Hugo.


I know I'm a day late, but I wanted to pass on that I am in love with this product, so cool! I have used it nonstop since I downloaded it. Can't wait to see how this evolves


How does it compare to Emacs Org Mode?


As several people have pointed out, both the "Personal" and "Catalyst" licenses are intended for personal use only, making this quite an expensive product should I want to use it for work as well. But in the full license text, there seems to be an even more problematic phrasing:

> The use of OBSIDIAN for the exercise of your own trade or profession (...) does not qualify as personal use.

I would interpret this to mean that I as a developer can not use this to exercise the "trade" of software development. That would in turn mean that I can not use this to make notes of stuff I learn on my own time, if it is related to software development.

I would imagine most people not caring about this kind of license limitation, but it would be interesting if it was intended this way, or if this is just me being bad at licenses.


If we're internet armchair lawyering then I'd point out that a carpenter building something is "exercising their trade or profession" but a carpenter reading a textbook or watching a how-to video is not.

Analogously, a salaried software developer learning something and taking notes on their own time would be in the clear but a freelance developer tracking their projects or clients would be in violation.

What's not clear is how it applies to a salaried developer learning something and taking notes on the job (ie is time provided for professional development part of practicing your trade or merely a perk?) or a freelance developer using this product at all (same issue as previous scenario, except now it's not even clear where your job ends and personal time begins).


What about a carpenter testing out new techniques in their garage and taking notes as they go - are they exercising their trade?


I'm pretty sure that doesn't qualify. To my mind that's training or study of a trade which is distinct from practicing (ie getting paid for) it.

Where it gets weird is cases where the delineation between practice and study is unclear. Researchers are professionally employed to study things. On the clock professional development is study, but it's paid for by your employer. Freelancers manage their own time making even the distinction between paid and unpaid work unclear in some cases - were you studying as part of a current contract, or to better prepare yourself to take on similar work in the future?


If you ain't getting paid, you ain't doing trade.


Commercial license is 50$ per year. Not even 5 bucks per month. I do not think this qualifies as "expensive", especially if you intend to use it for work related stuff


If you are going to use it as a knowledge base, you are probably looking at a few years of use.

Currently the closest match to this software I have installed is Quiver. Let’s compare costs over a period of 5 years:

Quiver: $9.99 Obsidian: $250.00

That may not be a lot of money to you personally, but it fits my mental model of “expensive” pretty well.


There is also Joplin (https://joplinapp.org/), which is open source and has multiple Apps available (including mobile). I don’t think it uses plain markdown files as the primary data store, but notes are written in it and they can be exported as Markdown files.


I got myself stuck in quiver which is a nice tool but has no movement in its project and has horrible exporting. Beware. I still need to move my .qlite notes into markdown.

My problem with Quiver wasn't just no updates, its a good tool how it is, but that plus no web/android/windows. Just figured it was time to finally hop of that sinking ship.

I.. wish someone had told me this 2 years ago when I started using quiver and it was already dead, before I put hundreds of notes in it that are now stuck in .qlite or .. pdf i think?


In Quiver, you can right-click a notebook (or All Notes) and export to markdown – as well as png, plain text, pdf, or png.


Thanks for the reminder. I think I tried this but had problems but I definitely need to try again, I totally forgot.


Does Quiver solve the same problems? Sorry it's a hard word to Google so not sure I found the app you are thinking of.

50/y is definitely a think about it and review yearly when budgeting type of expense as a personal expense, because all those recurring expenses add up! But as a business expense, if a $100k/y dev asks for a $50 tool I think their boss will say "why you waste my time asking, just get it and we'll pay for it".


I would say Quiver targets the same problem space. There’s some overlap and there are some healthy differences. I didn’t mention it to propose it as an alternative, but to illustrate that “expensive” is not just about absolute cost.

https://github.com/HappenApps/Quiver/wiki


I suspect OP is talking about this: https://github.com/HappenApps/Quiver/wiki (there's a link to an iOS app there)


My biggest problem is the license is pretty strict, so I don't really have room to try it and see if it works for me.

90% of my notes are work related, so I don't really generate enough "hobby" notes to stress the tool and see if I want to use it. I'm not going to drop $50 to try and it and see if I like it. I'm also not going to violate their license to see if I like it.

I wish there were a trial for the commercial version. 14 or 30 days would be really useful as an evaluation period. Without being able to try it before I buy, though, it ends up being a pass from me even though $50 isn't _that_ expensive.


wow $4 bucks a month fits your mental model of "expensive", you must live a very frugal life :-)


Not all people are living in an environment where they have an income that allows one to afford this. For some $4 is a lot, for some it is peanuts. Context matters.


If every piece of software I use day-to-day cost $4/mo, I'm not sure I would even use computers.


Do you create/release/maintain any free software? Or do you just feel entitled to have all the software that individual people do create for free?


I don't think that's relevant, I don't feel entitled to anything. My point was just that the $4's would add up. There are 2286 packages on my PC at the moment, which ones should I pay for? I can't even imagine earning that much money.

Moreover, are you suggesting that people should feel guilty for using free software?


The whole discussion was about how 'expensive' this particular software was, at $4/mo. $4 is the price of a cup of coffee. If the software makes you more productive than one cup of coffee per month would, then it seems like $4 is not expensive at all.

Of those 2286 packages on your PC, how many of them are written by individual software engineers who are trying to make a living by writing software? And how many of them contribute to your overall wellbeing? I'm guessing that 99% of those packages are used by programs that you yourself don't use, that just come packaged with your OS or some other product. The other 1%, yes, I think you should think about paying for.

We live in a world where a certain kind of extractive personality--Zuckerberg, Gates, Ellison--makes billions from their software, because they are willing to be an asshole-to-the-hilt with their anti-competitive, unethical, and often downright illegal behavior. The 'nice guys' who write and maintain your compiler (if gcc) or editor (if vim or emacs) or scripting language (virtually all of them) and any number of other tools that we use daily, will struggle to pay their hospital bills in retirement. And yet people balk at $4/mo for software that they might use for years. And even if they only used it for 1 month, that $4 is inconsequential compared to the other factors (like learning curve, time investment, etc).

I do think people should be spending some money on their software tools each year. Maybe 1% of their income? That's around $100/mo, which seems like an awful lot, but only in comparison to the current price they're paying of $0. But think of how much better your tools would be, if the developer who makes them could earn a living wage from them, without having to embrace the capitalistic ideals that make us hate computers and software in the firstplace.


While much of that is true, you again seem to be assuming everyone who uses notetaking software earns some kind of US/SF SWE salary from your $100/mo calculation.

I really dislike this cup of coffee that is commonplace. Buying coffee from Starbucks is really expensive! Coffee costs about 30c to make yourself. Some people buy coffee every day - over a year that can up to a month's rent in a world-class city.

Personally I have paid for a lot of software over the years. I think 95% of the time I have ended up getting almost no value out of those purchases, especially with iOS apps and PC games I never play. I wish I had not been so frivolous with those $4 coffee-cup Steam purcahses and instead spent all that money donating to open source projects which I actually do use every day. In the future I will be trying to do so.


By your calculations I should be spending less than $10 per month on software. And I'm a software developer, just not in USA. If I spent $4 just for Obsidian, I'd be left with $6 for cloud file storage, music and video streaming, and other subscriptions.


Oh, I'll jump in here.

Yes. I've built multiple pieces of free software, including open source drivers, data packages, libraries and even the odd game mod. Those projects have been used by individuals, corporations and other projects, both opensource and closed.

And no, I'm not entitled to free software, but I am entitled to where I spend my time, money and support. I do financially support open source projects that I use (and keep a spreadsheet in Notion to track that on a yearly basis). I also believe that if I'm going to integrate software into my day-to-day and depend on that software, it must be open source or have an open source available alternative (ala Android).

Open source isn't about making (or saving) a dollar. It's about software being more than just a means to profit.


I disagree. $50 for note taking app only is expensive compared to something like Google Docs which does much more functionality and includes cloud storage.

The product is applicable to anyone who needs a knowledge base so it’s $50/employee and that adds up quickly.

While the developer is free to charge whatever they like, I don’t like the trend of these products priced based on people thinking $50 isn’t that much and that spiraling into what should be a one time fee to every piece of software being $50/year forever.

This reminds me of the 4 hour workweek where there’s a plan to create recurring streams of revenue for little to no work.

I think the amount of time put into this product doesn’t warrant that price. Even though I think it’s a really neat product.


Yeah. Photoshop is only $10/mo. Many IntelliJ IDEs are $12.50/mo and you get to keep them if you stop paying. Both of those are core revenue drivers for many of the employees who use them. $4/mo is quite expensive for a markdown editor IMO.


I think it’s cool that the markdown still works when I stop paying.

Imagine if the author of grep charged $5/month for the utility.


I mean, if you're in a software job, sure. But in total for a year, it's almost 1 month worth of rent in the city that's close to me.


Rent is $50 a month?


whups! corrected


I don’t think the message really changed?


"But in total for a year,"

In other words, a year's worth of this (which is based off the idea that you would actually be using it. If you plan to use something you don't just stop using it after a month), is an entire month's worth of rent that you could pay instead.

Or maybe a different perspective, a month's worth of this is the same price as a month's worth of groceries.


The service is $50 per year. unless I’m also confused here.


When I read it it was 50$ per month, so many they have changed their pricing? Or I could have misread?

fake edit: yes, they reviewed their pricing since I read it. the "sync" tier used to be at the 50$ mark.


Hi there! Sorry it's our first time doing a license like this.

If I understand correctly, licenses are usually written more strictly for legal purposes, but in my opinion your use case sounds like it should belong to personal use.

If anyone has pointers for us to make the license text more clear, please let me know!


I would not write my own license. I'd really recommend you get a lawyer look over the text you have (that shouldn't be too expensive, it's a one time cost), or look at existing ones that are actually written by lawyers. For a public license (to restrict commercial use), I'd recommend both LicenseZero Prosperity ( https://prosperitylicense.com/versions/3.0.0 ) and the Polyform licenses ( https://polyformproject.org/ ).

For the private license, you could use/adapt the LicenseZero Private license ( https://licensezero.com/licenses/private ).

Those are more geared towards developer tools, so I'm not sure whether they'd be a good fit for your product. But all of them are written in simple to understand language so you should be able to figure that out by youself, and at the very least get a few pointers. If not, that is considered a bug with the license, and both projects are very open to feedback.


Obsidian folks: please don't use any of the aforementioned licenses. The reality is that they're _not_ a good fit (although the author would probably love to get you as a client to handle your EULA and convince you to use one of the others for your eventually-open-source plans, and that would be an even worse idea).


Why are they not a good fit? From a shallow reading, I thought Prosperity and the Polyform licenses would be fairly similar in intent to the licenses they have on their page. I can't imagine the authors (not author) of Polyform/Prosperity would care either way, and since nobody ever claimed those are open source licenses I don't understand your other ill-made point either. Maybe you refer to the Parity license? That one would make no sense in this context at all, and I doubt anybody would suggest it here.

Also, since your post is fairly negative and unsubstantiated overall, would you care to suggest an alternative?


Give them an alternative?


Licenses are written by people versed in the appropriate laws. That’s why they come across as “more strict”, for precision’s sake. However, one should be very careful about trying to emulate legal precision just by virtue of stricter language.


> licenses are usually written more strictly for legal purposes

Well written legal documents are precise (not strict). The trouble is that you haven't unambiguously defined what qualifies as personal use, and there are a rather large number of obvious edge cases. This is a good example of something that should be drafted by a qualified attorney.


As a feedback from me: the license doesn't leave any room for evaluation on the commercial side.

90% of the notes I generate are for my business, and probably ~10% of my notes are for personal stuff. The volume of personal notes isn't enough to be worth having my notes separated out while I trial Obsidian. Since I can't put any of my commercial notes in it without paying for it, I'm just giving up trying it.

If Obsidian works well for me, I'd definitely consider putting down $50/year for it. The price seems a little high to me, but if it works well, then I'd probably do it. But the license doesn't let me evaluate it, so I won't be able to find out if it works well for me. $50/year is way too high for me to put down to see if I like it.

A 14-day or 30-day evaluation period in the commercial license text would be really helpful to letting me see if Obsidian works for me.


The simple answer is: don't write your own software license, because if you're not a lawyer (or if you're a lawyer not specialized in this area of law), you'll get it wrong.


I will never again build any kind of workflow for my knowledge on top of anything that isn’t open source. 5y is a long time and 10y is an eternity for these sorts of products but its just a fraction of my working life.

My own personal setup is a bunch of markdown files and it’s great, so I like that approach, but I’m very cautious about investing time for something so important in a product i don’t control.


> I will never again build any kind of workflow for my knowledge on top of anything that isn’t open source.

100% this. The peace of mind I get knowing that all of my data is under my control is worth it, after scrambling to archive content from failed or pivoting services, removing my data from businesses that try to exploit it or trying to migrate my data from one old app to a different newer app.

There are many different open-source and self-hosted wikis, note taking apps and mind-mapping tools. Some of them are listed here[1], just Ctrl-F for "wiki", "notes" and "knowledge"

[1] https://github.com/awesome-selfhosted/awesome-selfhosted


Just a quick notw, I think everythink (need to archive content, ease of migrating data etc) is solved simply by Obsidian being offline-first and markdown-only. I.e. you don’t need the app to be open-source in the end.


We'll likely make a formal guarantee to open source in the case of us shutting down. Didn't realize that was a possibility, thanks for the suggestion!


You may want to look into the Gitlab [1] model of open source monetization, as it has strong resemblance of what you are trying to achieve with Obsidian.

[1] https://about.gitlab.com/pricing/#self-managed


Standard Notes is open source (and free software) today, while still in business with a business model similar to yours.

It also allows people to read, audit, and trust the crypto code, which your model does not. For people like me, that’s a hard pass.


That’s pretty cool and it’s good of you to think about the scenario.

The problem is the bankruptcy court/buyer/ex-spouse/whatever may have a different preference. So you may want to set up something legally binding now when everyone is the same agreement.

Could you also solve this by GPL+Commercial license for those that pay?


"Open source" is both a license and a development model. You're talking about a license, but someone interested in workflow longevity will probably not be satisfied with a source dump from a dead company. They're looking for a community which knows how to maintain it.

As JWZ once said: "You can't take a dying project, sprinkle it with the magic pixie dust of 'open source,' and have everything magically work out." Has there ever been a case where a company as its last act released their software as open-source, and a community formed and picked it up?


I can recall a few on the top of my head, but those are the exception rather than the norm: Blender & Mozilla, though that last one wasn't really a last act. I think I saw a few others, but can't recall which ones, I'd be interested in such a list.

Now, there are interesting communities that form around games, that tend to be a "dead product" after release, and for which there was a source dump (or not, actually): doom, openMW, openRA, etc.


You do realize communities can die out as well?


By definition, that's what happens when its developer community dies off down to size n=0. A source dump without a community is essentially fast-forwarding a (potential) developer community to its death. It should be clear why this is strictly worse.


Parse?


that sounds good in principle but I still wouldn't feel comfortable with it unless I was aware of similar setups that worked out well for the users.

Not saying such examples aren't out there. I just am unaware of any.


If anyone is wonder what this comment is in reply to, this comment is a dupe of https://news.ycombinator.com/item?id=23325105, replying to https://news.ycombinator.com/item?id=23325035.


I don’t follow. This app would work on top of your existing markdown files, and if they go out of business there is nothing preventing you from either keep using the app, or take your markdown files and take them elsewhere.


That's really a nice thing, to be honest, and probably the only reason I'm somewhat interested, even though I mostly run open-source software on my computers, except when such software is a one-off comodity, like videogames.

But for something I'm going to invest some time learning, and will rely on, that's a very difficult pill to swallow. I don't want to depend on proprietary software for important parts of my workflow. What if I want to switch to a pinebook, for instance? No ARM build seems to be available yet. And that's just a single example... (a WASM/LLVM IR build could sidestep that issue).

That said, I like the idea. Is there any open-source equivalent?


vimwiki is pretty close but this adds some nice touches on top.


You could build a knowledge base in https://joplinapp.org/ and store/sync it the way you want.

I don't see any restrictions regarding corporate usage.


I was actually using Joplin for the past year after migrating to evernote. I really liked it but am now migrating to something with a little more flexibility around structuring.

Thinking to structure it around hugo.


I'm checking out Joplin on this suggestion. Couple of questions - In what way did you use it? And what were your problems with it?


As the other commenter noted Joplin is a desktop application that syncs (over E2EE) if you like to other machines via gdrive/dropbox/nextcloud etc.

I was using it via nextcloud and it worked perfectly for what it was. The web clipper also did a great job of snatching simplified versions of web pages.

What killed it for me was just my expanding desires. I want to things like

- view my knowledge base from my work machine without downloading the whole knowledge base to it. - Be able to search my notes AND all the books i've read. I tried using pandoc to go epub to md but it was too clunky and the searchability within notes wasn't great.

If I was still using it as a drop in replacement for evernote I think it would be great, but to go beyond that starts to stretch the seams a bit.

I also second looking at github's awesome-selfhosted.


Joplin isn't like your typical web app, and you need to either carry a copy of your notes around on every device, or you need to sync them using a cloud service. For that reason, I went with other note-taking and wiki software.


I'm in agreement with you on this. I've been playing with a package for Atom that would simplify a bunch of stuff in my process, but in general I don't like the risk of proprietary software/stack for my workflow.


that’s a funny idea, putting the logic in the editor


Joplin is the state of the art in terms of open source note taking in my opinion. It too uses Markdown, and supports a whole heap of other features on top of it. But it's been around for a long time, has mobile apps, cloud backups, the lot - and is all open source on Github.


This is what attracted me to org-roam


What actually happened? I have a really hard time imagining


I've been using Obsidian for the past week. Here are my thoughts:

I love that this is basically just markdown with wiki links. I'm not too concerned about Obsidian going out of business and my notes not being useful anymore. Contrast that with Notion (which is great in many other ways), where I store my data on their cloud and even though I can export to markdown not everything is actually markdown (e.g. tables) so they just put a link to their site instead.

I also love that since this is locally stored, I have control of how I store my data. I have a private git repo, and just occasionally commit and push my changes.

I wish it were open source, that way I could feel better about the wiki links being useless if they go out of business. It's not a blocker for me though, since my notes are fairly straightforward and I mostly just use the links for table of contents files.

The downside of being locally stored is that it isn't cross-device capable. This limits me when I'm taking notes on my iPad. That said, I don't take many notes on my iPad so if I need to I can just manually transfer my notes when I get back to my computer. I can understand though that this would a be a show stopper for some people.

Overall, this solution works for me. I have high hopes that it will continue to improve and become even better than it is now. Thanks to the team for the hard work!


I store the notes in Dropbox and use iA on my mobile. Works well enough.


I can confirm that both iA writer and 1Writer work great as mobile solutions. 1Writer allows for following links. Both allow you to search through all your notes quickly, though there's probably a tiny bit less friction with iA than 1Writer.


You can also keep notes in git, sync them using Working Copy and use IA Writer.


This looks great! Good timing too, since Roam has recently closed its doors to new signups. I've been looking for a new notetaking / PKM / Markdown app recently, but unfortunately all apps out there fall short in at least one of the following criteria:

* bidirectional [[wiki links]]

* support for both inline and display math

* customizable themes / CSS

* rich formatting beyond markdown (e.g. wrap content in <div> tags with custom formatting--useful for e.g. placing a box around text, etc.)

* WYSIWYG (crucial for documents with tons of math and rich formatting as above) including wysiwyg editing of tables

Tools I've tried that come close but aren't quite good enough:

* Jupyter Lab / Notebooks

* Typora

* Roam

* OneNote (just let me write $\LaTeX$ math!)

It's actually been quite frustrating. because some apps are soooo close to what I need, but they're closed-source so I'm powerless to make the small improvements I need. My current workflow for notetaking uses a pretty suboptimal combination of Overleaf, Typora, and OneNote. I'd really like to be able to replace all three with a single tool. At the moment, the only thing that comes close is a Chrome tab with document.designMode="on".

To the Obsidian team: Please add inline math support and consider WYSWIG!


Since roam was lamented, obligatory boost for org-roam:

https://github.com/org-roam/org-roam


Apart from the learning curve org-roam is the answer.

I’m on week 2 with it and I can hardly sleep I’m so excited by how it’s exactly what I’ve been looking for.


I love this, and used Deft and org-mode for a long while, but formatting and inline images aren't something that's non-hackily possible to do in Emacs. Would love to be wrong about this.


A quick search suggests it's doable https://duckduckgo.com/html?q=org%20mode%20pictures


Yep, I did use inline images in my days with org-mode, but I found it to be rather hacky. It's possible, I just found it unpleasant.


May I suggest contacting the emacs org-mode maintainers and chat. The emacs guys are very receptive I know, I imagine the org-mode guys will be just as good.


Resizing image is a pain, and inserting quick screenshots takes some hacking around.


Inline math is coming very soon! WYSIWYG unfortunately will take a while to get right, but we'll get there.

Give Obsidian a try and let us know what you think! :)


You didn't ask for advice, so feel free not to take it - but you should consider using a pre-baked editing solution like Quill.js or something similar. Rich text serialization is its own problem and you shouldn't waste your development time on it :)


As someone who was experimenting with text editing in a browser recently: I could not find any open source (non-commercial expensive license) text editors that could actually handle workloads I cared about. (and which are not also specialized on code, such as Ace Editor) While Quill (and friends) definitely provide some very nice functionality, the moment I tried to do something like shove a full novel into one, not only did it take quite a while to actually load that text, but it took 500ms+ to insert single characters. There is a serious lack of an open source text editing library for Javascript that can handle any amount of characters without completely bogging down.

I was actually seriously impressed that I was able to take a huge amount of text and shove it into Obsidian and the UI stayed completely smooth the entire time. Props to the Obsidian devs for that!


Have you looked at Trilium Notes?

It appears to match all of your requirements, and is scriptable. It's also strongly FOSS, in the AGPL sense. It also has the ability to self-host a sync server.

https://github.com/zadam/trilium


Have you looked into Zettlr(.com)? It ticks most of your boxes. Plain old markdown. And it's actually open source and a pretty decent community, so there's a good chance to get features in.


Looks very promising, thanks! I'm surprised I've never come across Zettlr before in all my research about editors. The features page and docs really are a testament to potential of OSS.


I felt the same way so have switched completely to taking nodes using https://github.com/rust-lang/mdBook

I back everything up in a private git repo and then use various plugins built on top of mdbook to get all the functionality I want:

* rendering graphviz: https://github.com/dylanowen/mdbook-graphviz

* adding tags to the pages: https://github.com/dylanowen/mdbook-tag

* setting up sections to only render locally and not get published: https://github.com/dylanowen/mdbook-access

The plugin system is super easy to use and you don't need to code everything in rust. You can use any language.

I'm also working on a better renderer instead of just running the mdbook server all the time: https://github.com/dylanowen/mdnotes

I believe with plugins you can meet most of the requirements you have: * bidirectional: definitely could be a plugin

* math support: potentially this feature: https://github.com/rust-lang/mdBook/blob/master/book-example...

* customizable themes: you can completely overwrite the wrapping html and CSS

* rich formatting beyond markdown: mdbook supports inline html/js as it's based on commonmark

* WYSIWYG: this is the main missing piece, my strategy is to just run my mdnotes application next to my text editor.


I too have been looking for a better note-taking app. I also signed up for Roam but I'm still on the waitlist. I've been trying out all kinds of programs and coincidentally enough tried Obsidian this last week.

I feel the same exact way as you in regards to all these apps. I currently use Joplin but I wish it was more "roam like" but looked like https://bear.app/.

So far I've settled on using Typora with Joplin.


Have you tried https://wiki.js.org/? Curious if it fits your needs, given they are open source.

P.S.

I use GitHub for all my personal writing and only now realized that a big part of it is due to its open source nature and data portability. I use emacs org-mode for my todo lists for the same reason. You're so right about how critical these features are.


I'm pretty sure org-mode can do just about all of those (except maybe for the custom formatting bit). Inline LaTEX was something I found super nice to have in org-mode.


I was about to spend $25 for a one-time/lifetime license, until I read your license. No commercial or professional use?

I'm not paying $50/yr for that sorry. A yearly fee for this is unrealistic and is way more expensive than other products like Quiver or SimpleMind which offers a lifetime product


Agreed.

Devs, please consider the Jetbrains model - make the current version (up to v1 while in beta) have a lifetime license for a one-time fee, even commercial, and keep subscribers updated.

Since it's locally based, if I only need the basic features, I'm not sure what I'll be paying for with a subscription.


And make those licences able to be used for personal commercial projects. Then have a separate org license that can be transferred between users.

I want to switch companies without the hassle of requesting the application to be expensed every time.


It is quite an odd choice. It makes sense to tier based on features, but tiers based on how or when people use an app is bizarre. Am I supposed to monitor myself so I don't read a note in Obsidian while I'm doing commercial writing work?

(Also, their license page could do with some proofreading)


The idea is that if you're creating value with Obsidian, your employer should expense that, ideally it shouldn't impact you financially. If you're working for a non-profit for example, it's all free.

> (Also, their license page could do with some proofreading)

Interested to hear more! We did a few rounds of proofreading, did you find any typos?


> your employer should expense that, ideally it shouldn't impact you financially.

You're really limiting your userbase (customer pool) here by expecting only enterprise users. Many people here are freelance designers, students with part-time jobs, researchers writing for-profit books, or they work for themselves. Most employers don't allow expenses for note-taking apps. I would encourage you to adopt this pricing model: https://simplemind.eu/features-pricing/

It looks cool, but I won't buy it until it's a lifetime license.

Edit: Sorry that this came out overly-negative. It's a beautiful product, congrats on launching it!


I confirm, I could be interested in such a tool for my professional activity, but I work in a large company and there’s 0 chance they’d let me expense whatever software I want. I’d be paying from my own pocket.


For what it's worth: JetBrains has a model where they sell cheaper individual licenses that people can buy from their own money (company credit cards/expenses are not allowed), and those can be used for work. And more expensive per-seat licenses for companies/organizations. Also, that also includes a perpetual license for the version at the time of purchase (and discounts for people who upgrade). I think that is overall a fair deal, and keeps the incentives for both developers and customers as aligned as is possible under such circumstances, IMHO. Yearly licensing like this doesn't make much sense for such a product, I think.


I think you are a bit too optimistic on what employers are willing to expense. A note-taking app falls far outside what most workplaces that I know of in IT would accept. In reality, I am afraid that you are excluding a large group of users.

Personally, I would pay 25$ out of my own pocket for this app, but not if I have to ensure that the notes are not related to my work, only to my personal life. And with my personal projects I use similar technologies as I do at work. Should I then not look at the notes when at work? It is just to complicated to live up to the terms of the license.


I can't even get my employer to expense texts that are necessary references for my job. They're not going to pay for my subscription for your app. I don't think your fee is unreasonable. But there are probably a lot of people with employers similar to mine.


This is really murky to me. As a professional, I keep notes as I learn things which may or may not relate to my current employment. The notes are my property and part of my knowledge base that I bring to any job. But I suppose I indirectly receive compensation for knowing things, so I need a commercial license?

Or what if I start taking notes on some random subject for fun, and eventually I write and sell a book based on them? All of a sudden I need to start paying a yearly fee?

It's not a prohibitive amount of money, but I'd be much happier with paying a one-time fee for the current version, then buying future versions if I choose to.


Generally the way you extract value out of an enterprise customer is to sell it to their enterprise as an upsell from the personal plan. Check out Evernote's pricing for example:

https://evernote.com/compare-plans

You can use Evernote for personal or work uses. They have two tiers for that. Or you can get your company to buy it for you and at least one other person, with extra features that benefit the enterprise, not the user.


That doesn’t work for freelancers.


> I have never worked for any employer that would expanse a note taking app. Expansing anything is usually akin to pulling teeth. How am I supposed to get my employer to expanse it?


This seems reasonable to me.


Found and fix the mistakes, thanks for the heads-up!


Can somebody help me understand something about Markdown and Obsidian?

Why would I ever want an edit mode and a preview mode while taking notes? Don't I want to be able to both see my note and edit it at any time?

I mean, I guess I could just ignore preview mode if the edit mode actually allowed me to see all of the content of my note, but it doesn't. Images are only visible in preview mode. So all the talk about transclusion seems dumb to me. I can't see jack unless I decide to stop working and start viewing. I just don't get it. Is this really what people do? Switch back and forth between editing and viewing all the time?

Or perhaps they just decide to put all the content of the note on-screen twice? But why would I ever want to look at the same content twice? We're not talking about folding, so I can see two different places in the same note. We're talking viewing the exact same content twice, cutting my usable screen real estate in half for displaying the content of my note. So I have 50% of the context I could have... And don't even get me started about how synchronized scrolling is impossible in this view if you use images in your note.

I feel like I must be taking crazy pills, because everybody seems to love using markdown, and I really want to like it, too. I know the benefits of plaintext. But I just don't get making that sort of usability sacrifice.


We're working on WYSIWYG but it's not an easy problem. Eventually we want editing to feel comfortable to everyone! :)


That's good news for me, but apparently a lot of people don't feel the need for it. I'm still confused why. In general, are people switching modes or doing side-by-side or what?


Writting and browsing are often different mental modes. You are not always working with your notes, but mostly reading them. Thus have a plain edit-mode used 10% the time, and a rich viewing-mode for the remaining 90% of the usage-time is good enough for many people.


I think most people are used to seeing the source and can basically "imagine" what it looks like. So they really only switch for double checking before sending something out I would say.

Many people in our community said things about "staying in the editor" so I guess that's a common workflow.


I think I could get used to that in notes where I don't embed images.


For me, I just don’t look at the preview at all. Markdown is simple enough that I know how it will be rendered.

To be fair, I do enjoy the Typora model where my images get automatically included in the edit mode, but I don’t specifically need it.


I honestly don't think people have ever used good WYSIWYG editors.


i feel the same about the dual window (edit and preview) concept of markdown. such a hackish way and distracts the writing. rooting for a nice wysiwyg markdown editor for vscode and light weight program for windows.


I also feel like hot keys, VIM key bindings, and org-mode key bindings are a big part of what make it good to use for a power user. Bindings are learn once, benefit forever, whereas no key bindings may make it easy initially but are an infinite drag on productivity.


It depends on the editor, but esp for markdown pure text, WYSIWYG is a thorny topic - see slack's recent problems with their IME. Typora has a nice feel and switches from editor to preview as you move to the next area in a document.

Obsidian is trying this a bit with their headings, but seemingly small features like this are very difficult to get to feel natural, esp for devs comfortable with ASCII.


> I just don't get it. Is this really what people do? Switch back and forth between editing and viewing all the time?

Yes, I do this all the time. In particular I want to be editing my markdown in a fixed-width font (because it includes a lot of code, typically), but also able to view it prettily.


Can't comment on the merits of the product itself.

That in the first five minutes of this post being online, several users more or less claim that this has changed their lives in their first-ever comment on HN strikes me as a little odd however.

(Yes I'm aware that this is also my first-ever comment on HN.)


This got shared on their discord, but I s2g that it really is the case, given that I use a similar tool called Roam, which recently had such a big spike of attention (in part due to Thomas Frank's video on it: https://www.youtube.com/watch?v=vxOffM_tVHI) and somewhat popularized what Obsidian is kind of aiming for here (so much that Roam had to close registrations due to massive traffic influx). Plenty of ppl just swear by Roam at this point, including its creator who has big vision for a thing. Here is his white paper on a thing: https://roamresearch.com/#/app/help/page/Vu1MmjinS.

I can attest to that as well, and given that Roam is unavailable, would suggest to try out this alternative that is pretty decent for what it does (even though Conaw (Roam's creator) does think it doesn't exactly align with what Roam really seeks to do, yk the thing)

Edit: to elaborate, I am not raving about either Roam or Obsidian, I am raving about they approaches they provide, and after nearly a month of using Roam it honestly feels like a game changer to me. Solved my art block a bit as well.


We have been in private beta for a few months already, with 5000+ beta testers and a community. We did let them know that we're doing a "Show HN" today.

Hope that explains things!


Thanks for taking the time to reply! I guess it does explain things ...

Best of luck with the product, also +1 to the suggestion of an open source client from a sibling thread!


Yes! We really want people to be able to leave their knowledge bases as a legacy to their grandchildren, so we'll do everything it takes.


I've been using Obsidian for a short while after seeing a comment here on HN. I've used a lot of desktop notetaking apps, and this one really is different. It's the app HN commenters would write themselves. I expect it to be very successful in the next few years. It hasn't changed my life, because I'm not going to trust it with anything important right now, but they do have an ethusiastic community.

There's a new video out today if you want an overview:

https://youtu.be/cFYaWC_86W0


I've been using Obsidian for about a week, and it may be the first time I've ever felt compelled to comment in an app's user forum. I even installed Discord so that I could subscribe to the Obsidian channel and watch its (rapid) progress. I am quite excited by this tool.


It does seem suspicious but personally the only reason I jumped in was because I’m genuinely so happy with this software and I believe that it should be tried out by as many people as possible!


I've been beta testing Obsidian, and only rarely comment on HN. But I'm excited enough about it that I wanted to share my thoughts.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: