Transcript

00:00:00 - Speaker 1: We have to make sure that if you’re a brand new obsidian user, it feels accessible, it has infinite depth, and you can go as deep and crazy as you want, but that that surface level is intuitive and inviting to most people, and that’s a really hard thing to balance.

00:00:23 - Speaker 2: Hello and welcome to Meta Muse. Us as a tool for deep work on iPad and Mac. This podcast isn’t about me use the product, it’s about the small team and the big ideas behind it. I’m Adam Wiggins here with my colleague Mark McGranaghan. Hey, Adam. I joined today by Stefan Ago of Obsidian. How’s it going? And Stefan, you’ve got some nice recipes up on your website alongside various blog posts about tools for thought and technology. Tell me a little bit about pillowy Swedish cinnamon rolls.

00:00:55 - Speaker 1: Oh, that’s a good one. So I grew up in France.

My mom is half American, half Swedish, my dad is 100% French, and my mom is. A teacher, later in life she became a professional baker. She was always cooking and baking throughout my childhood and taught us a lot.

And she was also trying to infuse the household with her American culture and her Swedish culture as well, because I lived in France until I was 17.

And so, One of those things is a Swedish cinnamon roll. I think in the US, you know, the kind of cinnamon roll that you’re used to is probably derived from that.

It’s usually made with, you know, cinnamon or cardamom, and I came across this technique which is common in Asia called tanghong, which is a technique for making bread out of a, it’s almost like a very, I don’t know if anyone knows what a roux is, which is also a French technique, which is a mix of flour and water that you use to make gravy and other types of things.

You use basically a very Like a slurry of water and flour that you don’t darken at all, and you put part of that into the dough. And what it does is somehow, I don’t know all of the chemistry of this, but I think what it does is it encapsulates some of the moisture into the flour, and so when you mix that into the main part of the dough.

The dough stays really soft and fluffy and pillowy, and it’s just a really amazing texture.

And so I discovered this technique and I think, you know, it’s used for like milk buns and different things in Asia, but I thought it would be a good fit for the Swedish cinnamon roll that I always love to make around the holidays, you know, in December, even though I live in Los Angeles now and it’s not so cold, it’s just a kind of a nice memory. And so it turned out to be the perfect fit in an interesting fusion of two things. And so I put this recipe out. I don’t have very many recipes on my website, but it started to become a little bit of a section, and so I decided to post more of these because they’re really fun for me and very iterative. I like to incorporate techniques that I find online and get feedback from people who try it and. Iterate on them, so maybe it will become a more important section of my website. I think there’s only 2 or 3 recipes on there right now.

00:03:28 - Speaker 2: Especially like the I guess cultural mashup aspect of that, obviously drawing from your own heritage but also reaching outside of that. I always find, I guess as a person who’s an immigrant myself and I’m raising my child who has two parents from different countries and is living in a third country, so maybe not too dissimilar from your upbringing. And yeah, I think there’s just a lot of, I don’t know, interesting, you know, if we say everything is a remix now, you know, this kind of remix of fundamental cultures, I just think there’s a lot there.

00:04:02 - Speaker 1: Yeah, being able to pick and choose techniques, ideas from different cultures and like bring them together is really fun. That’s how I grew up, and so it just comes naturally to me.

00:04:13 - Speaker 2: And tell us a little bit about your background in the professional world.

00:04:18 - Speaker 1: Well, so today I’m the CEO of Obsidian. Obsidian is an app probably a lot of your listeners will know about in the tools for thought kind of space.

Before that, I’ve been an entrepreneur my whole life. I’ve run lots of companies, probably the most well known of them is called Lumi. We built a platform that helps entrepreneurs and Teams collaborate with manufacturers, particularly in the packaging space, so it’s a really interesting problem of There’s so much manufacturing capability in the world. Like there’s all these factories that are out there that can make things, but I find that the interface to access that capacity is Very confusing and difficult, and the idea behind the company was what if we could make it as easy to interface with factories as it is to interface with cloud computing.

And so that was a really fun adventure. We worked on that for about 8 years, sold the company, and then I found myself having a little bit more free time to think about things. I have been using.

A variety of different journaling and wiki type of software for a while and Obsidian came along, founded by Shia Lee and Erica Shu back in 2020. I started using it pretty much right away. It slotted into what I was doing perfectly and. I was using other tools before that and kind of had mashed up a few different things together. And Obsidian just sort of did exactly what I was trying to do by scotch taping all these different solutions together. And so I just fit like a glove right away, started using it, became close with the founders and started working on community contributions to the app. And Eventually, once I was leaving Lumi, they brought me on as the CEO and it’s a very small team. We’re only, you know, 6 people full time. So that title probably like yours, Adam, holds maybe a different meaning within our group, but it’s been really fun and I’ve been on it full time now for about 5 or 6 months, which has been really great.

00:06:32 - Speaker 2: And it must be quite a dramatic experience to come.

You’ve obviously started your own company and scaled that up and been the leader there, but coming into a tool that’s already established itself, at least within a particular niche, already has a big audience of fans, already has an existing team, plenty of culture and values and all that sort of thing. Obviously you We’re already resonant with that culture and values coming in, but to suddenly be on the inside and particularly to have this vested authority, all of a sudden, did you find that disorienting? How did that challenge compared to the challenge of starting something totally from scratch and sort of building every piece of it versus needing to like, I don’t know, bootstrap all the context or build the moral authority within the team?

00:07:15 - Speaker 1: Well, I’ve never worked at a company that I didn’t start until now, so it was surprisingly natural because I Had developed this relationship to the founders over a long period of time, very gradually, very organically just through chatting with them and reporting various bugs with the app and, you know, building some community contributions and things like that. So it was surprisingly easy and very natural. It was just really like, instead of spending, you know, a few hours here and there working on obsidian every week.

What if I was just doing that full time, and I do think it’s a At least in my mind. When I was thinking about what’s next after Lumia, my default would have been to start another company, but I couldn’t think of anything that I Thought was more exciting than obsidian.

And so that to me, at least in my own head, it says a lot. I don’t know if it says a lot to other people, but it does say a lot that I would rather kind of go and help build this thing, which I think is is such an amazing app and community than try to start another thing from scratch right now.

00:08:28 - Speaker 2: Yeah, that makes a lot of sense. And I think that entrepreneurship, you shouldn’t start with, I want to start a company. You should start with, I want to solve a problem or have an impact on the world or go to a place where I can contribute the most. Really starting a company should be your last resort. If there’s really no other way to accomplish, I agree,

00:08:47 - Speaker 2: that’s what you want to do, then you say, well, damn it, I guess I have to start a company.

00:08:51 - Speaker 1: Oh well, I have so many people come to me for entrepreneurship advice and my first advice is don’t do it. Most of the time. I’m constantly like trying to convince them not to start a company. And part of it is just me kind of probing to see how much they actually care about whatever they’re doing that, you know, they can deal with that because that’s pretty much what you’re going to get from the world, like 99% of the time is like, why does this exist? or why are you doing this? But I kind of took my own advice here and I think a lot of times it’s better to go, you know, put more wood behind fewer arrows.

00:09:26 - Speaker 2: Yeah, I think it’s also a really unique opportunity because getting the chance to, for example, I hired a CEO at Hiroku after we’ve been in business for a few years, and that was a really great opportunity to work with someone much more experienced and with knowledge about the spaces we wanted to move into.

For me, that was a new experience, yeah, always being an entrepreneur and then kind of Leading or being a co-leader in it until I’m done or don’t have anything else to say and then I just sort of leave. It’s really tricky to bring in a new leader, but it can also be an injection of new expertise, new perspective, new direction, new vision, especially because very often the kinds of people that like to start something are not the same.

People or don’t have the same skill set or just passions to want to scale it up, to want to see it grow wider, address, be available to more people, or just the natural kind of management responsibilities that go with managing a team, and existing product, a big base of customers who just seem to have an endless list of bugs and feature requests. It can take a different personality type. So when that can be done well, I think it’s really great to bring in an experienced leader at the right moment.

00:10:34 - Speaker 1: Yeah, and I was so impressed with Erica and Shea. People might know them as Silver and Leak out, which is their name online, but they’re a little bit younger than I am and have such a mature point of view on how they want to build the business of Obsidian from the start, they Made the decision not to go down the VC path, and I’ve, in my own businesses over the years have literally tried every different method of funding any business. I’ve bootstrapped companies. I’ve gone the VC route. Did Shark Tank, did Kickstarter did every different thing you can imagine, loans like if I type your name into YouTube with Shark Tank right now, will I get a

00:11:12 - Speaker 1: clip? You’re gonna find my co-founder Jessie. I was not brave enough to go on there myself, but she’s great, and you should watch that episode.

But I think they took a very mature path really thinking about the long term, which is aligned with kind of what the app is trying to do, and having the experience of going through all of those different ways of building a business.

I realized it is really hard to run a bootstrapped company and try to grow it kind of on your own. And I think that the approach that we’re taking with Obsidian is definitely hard mode in a way.

It seems surprising that you could just easily get millions of dollars, but if you have a good enough idea, there’s investors like banging at your door trying to give you money and it actually takes a lot of Fortitude to say, no, I don’t want millions of dollars.

We’re going to just do it, you know, ourselves, we’re gonna grow very carefully and organically and in an almost like selfish way, because I’m such a fan of obsidian first and foremost and a user of it, you know, pretty much every day, have it open in the background of my computer if I’m not using it actively.

I almost selfishly wanted to just kind of help ensure that obsidian continues on that independent path and continues to build kind of in a very thoughtful way. And what could I bring from a business standpoint to the table to create the structure that would enable that to continue being a priority.

00:12:45 - Speaker 2: And it’s going to lead to this later, but since we’re sort of on the topic now of the kind of the mechanics, which includes, yeah, financing, team size, but business model, obviously, it’s what I would call a prosumer model. There’s the free product you can download and use, and then there’s the sort of services, subscription services like Sync that you can sort of add on to that once you’re getting value from the product.

So I feel like prosumer is something that has like a longer ramp up, but you need to kind of like do that upfront investment, but it also doesn’t have a very good shape for venture because it doesn’t necessarily have that big kind of unicorn in 10 years shape to the graph that say like a B2BAS company might.

And then the middle ground there often ends up that companies like this basically finance it through just doing a bunch of consulting projects in the early days. I think maybe like yeah, the 37 signals folks is one example. I’ve done that with multiple businesses to the point you’re willing to reveal how does obsidian strike that balance? Have you been successful enough that you’re just able to finance on customer revenue or that early upfront investment feels like it’s got to come from somewhere?

00:13:50 - Speaker 1: And when you say prosumer, I think in my head, at least I think of prosumer as a market as a user type, but From a business model standpoint, I would say freemium is more the term that I’ve come to. Is that what you mean when you say prosumer, do you mean freemium?

00:14:06 - Speaker 2: No, because you can have a freemium B2B and you can have freemium B2C.

So music is one of the main areas like yeah, podcasters and DJs and whatever.

This is actually you’ve got people who are often hobbyists or aspiring professionals though realistically.

Maybe many of them are never going to make a living from it, but they are willing to spend hundreds or even thousands of dollars on gear on software, etc.

So classically B2C, you can’t get anyone to pay for anything. You just monetize with ads or whatever, and then B2B, you know, you get those big sort of company contracts where they’re paying thousands or tens of thousands a month.

And so I think of prosumer as being that kind of in between state of like Dropbox is a classic example, like pay $10 a month. Yeah, you get access to like, yeah, again, audio and video stuff, etc. I don’t know, maybe you don’t think of yourself that way, but that’s how I would slot it.

00:14:54 - Speaker 1: Well, I don’t, but I’m trying to kind of come around to your way of thinking of it like in that world, definitely, you know, Apple Notes and OneNote or some of these kind of apps that come shipped with the OS would probably be the consumer one like everyone just has it by default and it’s free, it’s bundled in.

Obsidian has a Freeman model so you can totally use Obsidian for free and for personal use, but it is a little more advanced, it has more complexity to it than an Apple notes.

It’s trying to, you know, give you a little bit more power user type of features, I suppose, and maybe that’s where the prosumer angle comes in.

I think for us in terms of what the kind of broader. Goals of what we’re trying to do are we’re really trying to democratize these tools like we’re trying to make it easily accessible for people to think using these tools and so we don’t feature anything behind a price, so it’s not like There’s a pro version that you pay $10 a month for. The capabilities are behind a license type. So if you are using obsidian for your business, then you need to buy a commercial license. So that’s a little bit of a unique point of view and it has to do with kind of the values that we have around really trying to democratize access to these tools.

The capabilities like sync and publish are paid add-ons, but There’s tons of free alternatives that are out there that may even be better for whatever use case you have. And so in a sense we’re competing with a bunch of free alternatives to our own services and we’re OK with that. A lot of the people who upgrade into some of those additional services, they’re doing it because they want to support Obsidian as a company as well.

00:16:51 - Speaker 2: Interesting, the commercial use kind of concept. I assume it’s to some extent is a honor system is quite the right word for it, but yeah, no, it’s an honor system,

00:17:02 - Speaker 2: right? Hard to tell if someone’s truly using something professionally and I think a lot of software again creator type software, if you think of like image editors or something like that, they might have things like water. Marks or something like that, that maybe an individual who’s just screwing around to make a meme, they’re OK with that, but a company would never put up with that.

00:17:19 - Speaker 1: So of course they’re gonna, but we don’t do any of that. Well, hopefully we don’t have to resort to any of those weird tactics.

Like I don’t want to do that. That it’s surprising how well the honor system works actually. I think that most companies, we have a lot of great organizations using obsidian that really care about privacy and so they tend to go down the obsidian path more so than some of the other like cloud-based providers that are out there and not encrypted.

And so I think that if you’re one of these organizations. You actually do care about reading the license, a lot of the software that you use, and if it says you need to pay $50 a user per year, we don’t get into that much friction, to be honest, when it comes to that. And the only friction that would come up would be everyone who’s in between who are like small, you know, couple people startups who from an honor system standpoint, they’re just probably using obsidian for free and it’s not a big deal.

00:18:21 - Speaker 2: Yeah, well, and classically, it’s been said that Fremium is almost an update of the old system which was used piracy as your sort of free version, so you pirate in Photoshop when you’re, you know, a university student that can’t afford anything and then later you have a real job at a real company and they want to be legit, so they buy you a license.

00:18:39 - Speaker 1: Exactly. Yeah, I mean, that was me. If I didn’t pirate Photoshop when I was, you know, 14 years old, I probably wouldn’t be doing the job that I’m doing now. But, you know, now today in 2023, I think it’s probably a better From a top of the funnel standpoint, piracy is not a great like method of trying to gain users. You might as well just give your app for free and then, you know, try to convince the people who can pay to come and join that tier.

00:19:10 - Speaker 2: Another example I’ve always liked that’s kind of a variation maybe on the the watermark sublime text, where when you buy a license, the only thing it does is remove the unlicensed text that’s in all caps from your title bar, which you probably don’t even notice that much in regular use. If someone’s looking over your shoulder or you’re pair programming on a screen share, it just, yeah, it looks like you’re kind of not serious about your tools and not investing or maybe just remind you of like, hey, this is a tool you rely on, it makes sense to support the creator or creators of it.

00:19:41 - Speaker 1: Yeah, so blind text is a huge inspiration to us. It’s great.

00:19:45 - Speaker 3: That reminds me sometimes you see these YouTube videos where people have the please license your windows sticker on their desktop.

00:19:52 - Speaker 1: Oh my gosh, that’s hilarious.

00:19:57 - Speaker 2: Well, I’d love to hear a little bit about how you think of obsidians fitting into, you know, we’ve talked about the tools a bit, but we self-identify, that is to say the Muse team and to some degree could switch as being part of the tools for thought community scene, whatever you want to call that, you know, your website, you call yourself a second brain.

There’s obviously the concept of note taking. You’ve already mentioned Apple Notes, for example, although you know you could argue the degree to which a very simple notes app like Apple Notes is even in the same category as a knowledge graph or a wiki. When you think of that sort of category of software, how do you think of obsidian’s place within it?

00:20:31 - Speaker 1: Yeah, I mean, I think that. What most people are familiar with when I try to explain obsidian to someone who’s never really thought about using a tool like this, I go back to Wikipedia as the touchstone because I think everyone’s been on Wikipedia. People understand that most file systems, most structural systems are hierarchical or chronological, and so they’re linear in one of those ways, but they understand. Even if it’s just sort of intuitively or emotionally that Wikipedia is not organized like that.

I mean there are some hierarchies in there, but it’s a web of links that you can click through and everything is related to something else.

And I think that is really what I mean, even the term when you say personal wiki, it’s jargon, but if you say it’s like Wikipedia, then it becomes not jargon anymore and it’s like creating your own Wikipedia is oftentimes what I use as a description for obsidian to, you know, non nerds basically. And I think that’s a powerful analogy that, you know, somehow hasn’t permeated into tools like Apple Notes though, you know, I heard recently they’re making a new journaling app, but we’ll be curious to see what they come up with there.

Um, but it is like one extra level of friction that maybe those like really basic tools are not looking to do.

And once you start to link ideas together, what can you do with that? What new structural concepts does it open up? And it sounds so simple, but at least For me and my thinking, it totally changes the way that I organize my thoughts.

There’s some people out there who are geniuses who can do this purely in their mind, but I don’t know. I just don’t have the like RAM in my brain to be able to maintain lots of different ideas at the same time. And so having this tool where I can kind of break down a problem into smaller chunks and then Remix those little chunks, however I want inside of a note is a really powerful and very basic concept. And then everything is layered up on top of that. So graph views, canvas views, you know, backlinks, like all of these different add-ons and things that can enable some new kinds of workflows, databases, like you can kind of go ad infinitum on top of that basic concept, but it comes down to. Links between notes and this kind of bottom up organizational model.

00:23:07 - Speaker 2: I think you actually perfectly teed up our topic today, which is Evergreen notes, and partially I like this term for a lot of reasons comes up, but it’s also a back reference to one of our first guests we ever had on the podcast, Andy Match. We’ll link that in the show notes, but you have a great blog post titled Evergreen Notes Turn Ideas into Objects you can manipulate.

00:23:29 - Speaker 1: Yeah, Evergreen, I think that Andy’s notes about that that he’s published are really great, and what I like about my definition is just turning an idea into a memorable chunk of text, but memorable to you, like a meme that is a meme inside of your own thinking.

How can you take An idea that you had or read and turn it into a memorable chunk.

Like sometimes I think what we love about good quotes from like famous people or from books is that they are in a way an every green note because they take a feeling or a concept and turn it into this like memorable little chunk of text.

And at least in the way that I write for my own personal thinking, having that little chunk of text, like you said, everything is a remix. That’s an evergreen note in my system.

And I can use that in the context of a sentence that might start with, because everything is a remix, you know, this thing that I found is interesting for that reason, and I use everything as a remix as a link in that sentence.

And it becomes a very natural way to compose ideas together, but I want to try to make it kind of more relatable to People who haven’t thought this way in the past, and that was the purpose of that blog post was try to explain that if you can externalize ideas and you can create your own little memes inside of your system, then you can touch those ideas, you can rotate them, you can. Manipulate them in a way that personally, I find that my brain doesn’t work that way. I don’t have the capability to just do that purely inside of my head. I have to externalize it in order to be able to manipulate it.

00:25:17 - Speaker 3: Yeah, that’s been my experience too. Or at least I think you can fool yourself into thinking that you can manipulate these things in your head, because you can hold what, 7 things in your head? It’s like, oh, look, I have 7 things in my head and I can even combine them in different ways.

But it’s sort of false because once you write down 20 or 30 things and have them as discrete objects, that’s when you have the Ability to rapidly play with new combinations. It’s one of those things that works unreasonably well, just writing it down, because it takes it out of your head and it frees up one of those 7 slots to put something in and it makes it possible for you to quickly pick up new objects to put in those 7 slots from your written down items.

00:25:56 - Speaker 1: Yeah, and if you can break your ideas down into smaller and smaller pieces, you can also build up more complex ideas that you feel have a stable foundation. Like you can build ideas on top of each other into thinking more complex thoughts than you could otherwise think, which I think is exciting. That’s really fun. Yeah.

00:26:18 - Speaker 3: This idea of being able to build up more complex thoughts because you’ve written them down, it reminds me of this idea of automation and programming, where sometimes it feels like you don’t really need to automate it because it’s basically going fine when you do it manually, which again, is true as far as it goes. But really what happens is you have some capacity to do stuff manually. So if you automate it. You can add your manual stuff on top of that, so you basically open up the ceiling to be able to do more stuff as a computer user. It kind of has the same feeling to me as this idea of writing stuff down to free up more mental space.

00:26:51 - Speaker 1: Maybe I should give an example so that people who are listening can understand what I’m talking about. I was reading this book by Murakami, I think it’s called What I Talk About When I Talk About Running or something like that.

And he likes to run marathons and write books. And so he, you know, kind of compares the two, and he has this phrase that is a very memorable phrase in the book, which is pain is inevitable, but suffering is optional. He says, basically, when you sign up for a marathon, you’re basically signing up for pain. But it’s your choice to decide whether you want to suffer that pain or not. I was going through a very painful time at the time and so it kind of resonated with me.

And then an evergreen note that I wrote in obsidian was pain is information and my thought about pain as information was I think children learn this at an early age. If you touch a hot stove, you know, that’s information that don’t touch hot things, you know, you’ll burn yourself. But in general, like pain is a signal from your body. It could be a physical pain, it could be emotional pain that gives you information. And then, There’s this phrase like knowledge is power. So, you know, if you gain enough information through pain, can you build knowledge off of that? By the transit of property is pain power? Like that that was a question that I was asking like if you are able to understand pain and synthesize pain, is it a path to power? It has a lot of connotations, but can you become more powerful by having more painful experiences? So this was kind of just like a train of thought, but like each nugget is an evergreen note. So pain is inevitable, but suffering is optional, pain is information, you know, knowledge is power, pain is power. Like you can kind of try to explore these ideas.

You don’t necessarily need to agree with them. Like there’s tons of evergreen notes in my obsidian that I don’t agree with, but I’m trying to turn them into a little meme for myself that I can come back to. And maybe I’m still trying to figure out my opinion about that thought, and I’m trying to use that little fragment of text in a sentence where I’m agreeing with it or I’m disagreeing with it. I’m trying to weigh the pros and cons of it. I’m trying to mix it with another idea. Maybe this idea is not as strong as this other variation of the idea. And I just find it a very useful, you know, way to kind of think about these things and hopefully that example makes sense.

00:29:27 - Speaker 2: The self memes, including accessing maybe yeah, books you’ve read, quotations, yeah, everything is a remix, the running book, I think of ones that I referenced with some frequency like the man in the arena quote, these serve not only to, if we do think of ideas as these little notes, which are essentially objects that we can break apart and then use to do almost like Arithmetic or something like that and explore in ways that are more complex and interesting than the ideas would be on their own, but also can build up that pyramid and then encapsulate actually a lot of kind of sub items under one so everything is a remix or across the chasm or something like that actually contains a lot to it and you could read a whole book in many cases or whole section or think many deep thoughts on that, but eventually if you bubble it up into that meme, it almost reminds me of like a scientific citation where if I’m writing a paper about a complex topic and I need to reference another complex topic, I don’t need to go and restate that whole complex topic. I just referenced the paper and for a very small number of tokens I get essentially all of that brought in if you already know the paper or you know the meme or you know the quote. That basically can just serve as a reference to all that.

If you don’t know that well, then you can go and explore, go down the rabbit trail there, which again, I guess does bring us back to that kind of Wikipedia linked knowledge, knowledge graphs, scientific papers as citations like, yeah, these are things that exist in other forms, but this version of it for the personal environment, the personal notes tool, personal knowledge base that has this highly manipulable component, I guess that’s what it feels like is truly new with tools. Likesidian.

00:31:09 - Speaker 1: And maybe the difference there between what you’re talking about and Wikipedia itself is that Wikipedia will reference specific books and places and, you know, concepts or terms, but it doesn’t really have like memes inside of Wikipedia because those are very personal kind of interpretations of an idea.

One school of thought could have many different like sub ideas within it and those ideas don’t really tend to make their way.

Into Wikipedia in that same way, but the concept of being able to kind of like manipulate them is similar.

Am I making sense? Like Wikipedia doesn’t really contain interpretations of an idea, because it’s not trying to do that, it’s trying to be an encyclopedia, so it’s trying to be objective and not subjective, but these evergreen notes are intentionally subjective.

00:32:00 - Speaker 3: The evergreen note examples that we’ve been talking about have been very granular, aphorisms of a few words. Do you also have evergreen notes that are huge sprawling pages in which you’re creating stuff over time, or do you really prefer the granular style?

00:32:16 - Speaker 1: Um, I like really granular, I mean this is just me personally, you know, I’m not dogmatic about this.

People can do whatever they want. I tend to have small fragments that I can compose into bigger fragments, so.

Yeah, I can’t really think of, you know, really huge evergreen notes that I have.

What those turn into is journal entries or stream of consciousness type of things where it’s like playing with Lego blocks. It’s like I’ve got these Lego blocks which are my evergreen notes, and then I have a session where I’m going to think about these like 10 different evergreen notes and combine them together. But that thing is not an evergreen note. It’s just a stream of consciousness, a thought process in my system, it lives as a date stamp with a name, and it’s just like on this day, I had these thoughts about these evergreen notes, but the evergreen notes are not time stamped, they hopefully have longevity.

00:33:19 - Speaker 2: Now longevity, I also feel like it’s an interesting fork to explore here. Some Mark and I have talked about as some of the listeners will know and talking about software longevity and sort of digital preservation and the challenge of bit ros and how quickly files and applications and whole systems sort of cease to be accessible. He talking about kind of your own personal knowledge systems and obviously I know that this is a big part of what. built on, which is just a folder full of mark down files and that’s plain text and now marked down as an extension of that is something that has really stood the test of time in a way that almost any other format you can think of hasn’t. How do you think about evergreen notes, durability, and especially in the context of your personal notes and how long those need to last.

00:34:08 - Speaker 1: It’s a very high priority. I would say that we are kind of plain text maximalists, like even more so than markdown. Markdown is definitely kind of this system that seems to have permeated enough and has lasted long enough that, you know, we feel comfortable using it as the kind of default markup in obsidian, but I think that we’re in this era that’s a very Unusual time because Digital files have only been around for 70 or 80 years.

And that’s not very long relative to time. People have been writing things down for thousands of years and so we’ve started generating a huge amount of digital data.

How much of that digital data is going to still exist 1000 years from now? It seems like on the one hand, we’re able to capture a lot more than we ever have been able to, but how much will be retained is the question.

And my framework for this is just the Lindi effect. I just want to think about what has existed for a long time and can we use that as a proxy for, you know, hopefully something that will last a long time.

And my gut feeling is that if computers are still around in 1000 years, plain text will probably still work, you know, maybe some other dramatic thing will happen where computers are not still around, but we’re trying to make decisions within the context that we know about right now.

And so that’s also why sometimes I say like files are much more important than apps. We care about the file that you create in Obsidian much more so than the app. The app is ephemeral. Like the app is not gonna last forever.

I think it’s a fallacy to think that you’re gonna design a tool that’s gonna last forever.

Maybe like a chisel can last forever or something like that, but a software app is probably. not operating systems change, users change, things are changing so quickly. I don’t really care what kind of chisel someone used to, you know, inscribe hieroglyphics on a pyramid or something like that, but they were able to communicate some information that has stood the test of time.

And so that’s why Obsidian is writing to plain text files that for now in terms of what we have for digital information is the you could open. An obsidian file on a computer from the 60s, which means that hopefully it will also work for a computer from, you know, 200 years from now.

00:36:38 - Speaker 2: Yeah, I love that and it touches on a few things of very much of interest to me.

I mean, one is, I think that from a user perspective, putting aside how long of a duration you expect or want or would be desirable in your data, your work, the things you’ve created, ultimately, I think, especially here talking about creators, people using tools or productivity software to make.

Things you really care about your work, not the tool, and obviously people can get excited about the hot new tool and they do, and that’s a lot of fun, but ultimately I care about when I’m using a piece of video editing software to edit a video or I’m using a word processor to write my PhD dissertation. I care about what I’m creating way, way more than the tool itself. As software creators, as tool makers, it’s very easy to have a certain kind of egocentrism, which is the tool is the important part, or maybe this even just comes from programmers where we think, well, the program is a complicated, interesting, important part and all of those bits we write to disk on behalf of the user, that’s kind of a secondary thing, but I think the user perspective is really the inverse of that.

00:37:40 - Speaker 1: Yeah, and I think That’s why the term tools for thought kind of rubs me the wrong way sometimes because it’s putting the tool at a higher level of importance than the thought in a weird way.

And I think that the question that I’m wondering about is like, what are we doing on a civilizational level when I say we, like everyone who’s involved in Making and using tools for thought right now. What’s happening right now? Because it does feel like there’s something brewing, there’s like something that’s happening right now in this area that hasn’t, for some reason it wasn’t happening 10 years ago or 20 years ago. It seems to be happening right now.

And I do think we’re inventing some interesting new tools, and we’re making some interesting decisions about society or humanity in some way. And I feel like the things that We’ve been kind of talking about in this conversation are the things that we’re doing.

We’re trying to unlock a way that people can have thoughts that they haven’t had before. Like maybe some of these tools can open up ideas and allow people to think more complex thoughts or accelerate their progress towards some sort of creative output that they wouldn’t have otherwise been able to get to without these tools.

So that seems kind of cool and important.

And then the other part is How did those ideas or creations or whatever you made the outputs of the tool last for a long time, hopefully. I mean, maybe you make something very ephemeral and it’s not meant to last for a long time and that’s fine, but if you want them to be able to, they should. And I think that we’re at a turning point, like the printing press or something where we have the opportunity to kind of design these tools to hopefully pursue at least one or both of those goals.

00:39:27 - Speaker 2: Part of what I find so interesting about this, yeah, tools for thought scene, whatever you wanna call it, is just caring about the ability to Use software and computers first of all, as a thinking tool, which I think has only very rarely been something that’s on people’s minds.

We’re usually thinking much more pragmatically. Here’s a calendar, here’s email, here’s a to do list, something of that nature. And I’m not sure that all the things that these different tools are trying, whether it’s sort of canvas-based tools, more tech orient. tools, things using space repetition, all that sort of thing, that those things were impossible to do with computers 10 or 15 or 20 years ago, but for some reason, people just got really interested in it right now and there’s some excitement around it and some sexiness around it and maybe some commercial opportunity around it as well, and that just has a bunch of people thinking about it.

And regardless of the specifics of any individual product or project, I just love that there’s so many people thinking about the problem from that perspective. How can we use computers to help us all think thoughts we didn’t have before, like you said, be able to do more with our thoughts, be able to do more with our productive philosophical and creative efforts.

00:40:41 - Speaker 1: Yeah, and I think what’s new about computers in this respect is that every paper-based physical based system for thinking was kind of bound by the limitations of physics and physical objects, making it more linear, like a book, you know, is linear in nature.

I’m always amazed when people get into the settle cast and like slip box concept that this guy literally did what I’m talking about with Evergreen notes, but just like have these little pieces of paper that were cross referencing itself.

It’s a nightmare. It’s so cool that someone tried to do that, but literally, you know, one person did that because it was that complicated.

And so we have the opportunity now to do things like canvas or graphs or things that have infinite levels of depth and nonlinear, non-hierarchical structures because we’re not bound by the three dimensional space when we’re, you know, working with these digital files and so that’s a really cool thing.

How does the output of whatever you came up with. Like, it’s a means to an end still, like the canvas view, for example, in Obsidian or Muse, like, to me it’s in service of creating something at the other side of that, that is probably not a canvas in itself, like the canvas is not the output, the canvas is the kind of playground to arrive at an output.

00:42:11 - Speaker 2: Yeah, and that has been, I think a challenge in just marketing a product to a wider audience that is fundamentally for thinking, which is the output is the epiphany.

The output is the the idea that you wouldn’t have had, as you said, and there may be pragmatic. something where again you’re working on your PhD dissertation or your grand idea or your software product or whatever it is, and the thinking tools help you to achieve that and maybe you’re copy pasting some things out of it, but in a way to me it’s almost a feature that Whatever it is that plain text file that canvas or whatever form the thinking space is taking, or even going to the physical world, right, the whiteboard or the sketchbook, it’s sort of a feature that my sketchbook I can’t turn that into the finished artifact because the sketchbook is the place to have loose thoughts. In a way that’s open ended, that’s safe and private, that is just messy and combinatorial, and then when I feel like, OK, I’ve had the aha moment, this is the thing I need to say or do, now I’m going to move to those more kind of production tools, but that can seem confusing, I think in some cases because it sort of seems like you’re doing extra work and why. And I think of it as you’re doing the work you’re already doing in your head, but you’re doing it through this externalized form, as you said earlier, and that that is a help, even though it may in the sense of like what it looks like to an external person, look like I’m doing more work, but you have to do the thinking either way with or without the help, with or without the aid. And then there’s also that, how do you turn this into like a more Production thing for consumption by other people or execute that idea and that should just be a separate set of tools.

00:43:48 - Speaker 1: It is hard to be messy in a digital form, and I think that’s kind of we’re trying to make that gradient between messy to finished smoother in a way, at least with the obsidian, it’s kind of a Implicit goal of trying to bridge those two things in a way that feels like continuous.

00:44:09 - Speaker 2: Yeah, I know this is something you’ve you’ve touched on with some regularity mark is the idea of like not having there be too discreet of a transition from, OK, now I need to take something that’s kind of transcribe it or take it from my sketchbook and move it to, for example, a digital form.

And what’s in there, I’m taking it somewhere else.

I think the reality is most production pipelines, if you want to call it that, do have multiple steps, right? I write a script for my movie and the script is in a different tool in a completely different format from shooting on film and then editing that down and then how I’m actually going to distribute that to my end audience is also, you know, uses a different tool, but making those steps less jarring is, I think, very desirable.

00:44:52 - Speaker 3: Yeah, I think you correctly point out that in practice, you’re almost certainly gonna have separate discrete tools, and now that I’m thinking about it in terms of explaining the products and marketing can be quite difficult because, OK, in the ideal case, you have one Uber tool that like magically morphs from a messy idea sandbox into a finalized, you know, edited movie or something. OK, sure. We know that in practice you’re gonna have to have a discrete step there where you say, Develop an outline and use, for example, and that eventually goes into the film production process.

That is kind of explainable, like you do need to introduce this extra step that people often skip, unfortunately, but it’s in fact even harder than that because in many cases, what you’re doing with something like muse or obsidian, in my opinion, is you’re basically rewiring your brain, you’re introducing new thoughts into your head and you can actually throw away that artifacts, but it’s what’s in your head. So now you gotta explain to someone, oh, you know, it’s actually just that your neural net weights have been updated. And then furthermore, it’s often the case that these are not in your waking conscious mind. You’re updating the weights in your unconscious mind and explaining that is very, very difficult, you know, source, trust me, bro.

00:45:58 - Speaker 2: Now I’m curious, you mentioned trying to make that process of starting from the raw and unfinished and messy and moving to the more sorted out and organized, smoother, what sorts of things in practice has that looked like for your product?

00:46:13 - Speaker 1: I think for us it’s being nonprescriptive about how the tool works and really working on the primitives. So, Obsidian has a point of view on malleability and extensibility that I think is pretty unique. We try to get the basic things right like text entry, just like even that problem is actually really, really hard just actually making an editor that feels fast and responsive. I think a lot of people get frustrated with other tools that just don’t give them the feeling of they can type as fast as they want anytime and that basic problem is one that we Retain as like one of the most important things about obsidian.

And so there’s a handful of these things that are kind of like the primitives like we think links are really important and being able to quickly link between files is really important. We added a new primitive with Canvas, which is like spatial relationships. But all of these different aspects of obsidian, you know, the next priority is extensibility. So the first priority is like let’s get the basic thing. The basic experience as good as possible for 90% of the use cases, but then everything else is a really long tail that is quite unpredictable and very different from person to person depending on how they think, depending on the kind of work that they do. Are they academic, are they creative? Are they using PDFs all day long? Like what are they doing in their actual workflow and how does the tool adapt to that? And so. We just accept that we’re not going to be able to, you know, put all those features into the app and instead what we’re gonna do is just make it really, really extensible so that people can build those things on top of Obsidian and take them in all the different directions that they want and assume that basically there’s going to be very little overlap between which specific plugins any given person is going to be using.

I think that I personally have an inclination towards making, you know, these really well designed, opinionated tools that have like a way to use them. Like that’s what I’ve done a lot of my life. But obsidian has challenged me to really think the opposite way and say like, what is core, what is something that everyone needs, and then what is everything else? And recently, like AI has also been this kind of big interesting topic that I think a lot of us have been playing around with these different tools for chat GBT and so on to kind of use them inside of our tools. Like right now there’s probably like 15 different AI plug-ins for Obsidian, and they each have their own like little different take, whereas if you look at some of the other products in the tools for Though space. They’re each implementing it kind of as a first party thing into their app with an opinionated point of view. We don’t have an AI like first party AI integration right now partially because it kind of conflicts with our privacy and values, but if you want to use one of those tools, you can use any of the 15 different open AI integrations that exist. And actually one of the things that I’ve been enjoying is using Chad GPT to make plugins. So the other day I had an idea for a plug-in inspired by someone on Twitter had created this really nifty kind of prototype demo of a stream of consciousness writing experience where the words fade out as you’re typing them, and it just like is very immersive way of writing just like basically one word at a time. And I thought that was so cool. I thought it would be an interesting experiment to ask Chat GBT to make me an obsidian plugin that does this, and it’s not perfect, but I was able to get to something that basically replicated that, you know, within an hour. And so that really speaks to the malleability of the software. Like, can you take this thing and shape it to, you know, what you need as a tool for your process, knowing that maybe your process is even going to change over time, like over the next 5 to 10 years as you evolve as a person.

00:50:14 - Speaker 3: Yeah, it’s been really cool to observe the obsidian community with the plugins. There’s so much cool stuff that people have been trying. I’d be curious, is it your vision for plug-ins that they’re mostly kind of content oriented versus behavior oriented? So for content, I think of, you know, rendering basically and the MySpace backgrounds and that sort of customization, whereas workflow behavior, I could imagine something like every day you take all the Check items from your to do list that haven’t been done and move into a new document that you create with today’s date. Do you see both those as being in play for plug-ins?

00:50:48 - Speaker 1: Oh yeah, for sure. I mean, there’s literally 1000, I think we just crossed 1000 plugins for obsidian, and they do everything under the sun.

If they don’t do something, if they’re trying to do something that we like don’t have an API for, that’s something that we should add an API for.

Um, so I would say most plug-in, at least half of the plugins do what you’re describing. They’re workflow oriented, they allow. You know, syncing from other systems, they allow, you know, pushing out to other things to do lists like different view modalities for like Kanban or dates or like all kinds of different things. So we try to make it as open ended as possible as far as what plug-ins can do.

00:51:31 - Speaker 2: Yeah, and malleability is a big topic for the researchers at I and Switch, and certainly something I know from a lot of that research is that you do always have this trade-off as an extensible as possible system like you’re describing.

If there isn’t an API for something, you should make it, but inevitably that does mean that people are gonna try and do everything to the point that the more extensible it is, the more easy it is to end up either completely shooting yourself in the foot.

But perhaps a little less dramatically than that, deleting your data or something like that is more like just this conflict between you install these 5 different plug-ins, the whole interface gets really weird and janky. Now they’re writing in with bug reports about like this button doesn’t work right, but it’s because they installed some plug-in that messed with it in some unexpected way, right? Famously, this is like the difference between like Apple and Android, right? The iOS world is very locked down. You want to customize it, we’ll let you change the background or something, but that’s kind of it, so that they can make sure that experience is really curated and always kind of, you know, just works, so to speak, and then there’s the Android world which is more open-ended, but then is more famous for being basically a little bit janky. How do you think about that trade-off?

00:52:43 - Speaker 1: It’s definitely a trade-off. It definitely has its downsides, but I think that There are for the most part too few tools like obsidian in other spaces that give you that kind of freedom.

So I think it’s also a reaction to what most tools that people are used to don’t have that malleability to begin with. So we’re OK with carving out this section of the market where people can have a little bit more freedom at the cost of, yes, you can shoot yourself in the foot for sure.

The only thing that we can do is kind of provide the guide rails like what we try to do, for example, we just recently completely revamped the developer documentation. I mean, there’s basic things like how do you capitalize the text on a button, you know, like because we’re open-ended to plugins, like there’s different guidelines on Apple OS, the interface guidelines say that you should always title case your buttons, whereas on the Google interface guidelines it says it should be sentence case, and those are. Just two different opinions that these two different valid and viable and large platforms are choosing. So if you’re a plug-in developer, which one do you use? Like there’s all these like basic decisions you wouldn’t think about, but they make the app feel less cohesive overall when you’re using it if people are capitalizing their buttons in different ways. And so, There’s an effort that we need to make, which is to kind of help developers have really good defaults when they’re kind of building on top of obsidian and guide them towards things that are going to feel intuitive and cohesive to obsidian users while not limiting their freedom.

Similarly, we have to also make sure that if you’re a brand new obsidian user and you’ve never used it before, that the kind of like. Top level, it feels accessible that you can understand how this app works, even, you know, with no plug-ins, it makes sense to you, but then it has infinite depth and you can go as deep and crazy as you want, but that that surface level is intuitive and inviting and accessible to most people, and that’s a really, really hard thing to balance.

00:54:59 - Speaker 3: Yeah, I think this is such a rich area, like plug-ins, extensibility, programmability, cause there’s so many variables that you’re dealing with, and at least with our current technology, there’s no way to satisfy fully all deerra at the same time. So you gotta explore the trade-off space and do the best you can and try to push out the frontier. So yeah, I’m very excited to see that you all are giving that a shot.

00:55:21 - Speaker 2: Yeah, I definitely like the idea of the core app that’s sort of very cohesively designed and it has more of a top, top down’s quite the right word for it, but it’s designed by one organization that can, you know, for example, use the standard capitalization on its buttons, but then there is a wider world or a deeper world that you can go out to, but you kind of know when you’re crossing from the relative safety into a bit more of the frontier.

We talked about this a little bit with. The creators of Raycast and their plugin system and a lot of what they do is because there are so many of their users and certainly obviously the developers are developers and they can kind of work with them on the pull request and their review process is less, telling them what they can and can’t do and more saying, well, look, it’d be a bit more idiomatic or it fit in better or I think our users would appreciate it more if you made these changes, even potentially working with them on the code.

And then of course, obviously you’ve got something like Apple that has this, you know, very heavy-handed and opaque review process, maybe there’s something like the browser vendors with their browser extension reviews are a little bit looser, but how do you think about that curation of that official List that you can submit your lugging into, is there really kind of strict guidelines? Is it more of a, you know, just what you feel is best for the community or you really just trying to filter out true malicious actors and it’s fine if it’s kind of heterogeneous in terms of the style and approach.

00:56:45 - Speaker 1: Yes, so we do have some strict guidelines that have to do with like security, for example, we will obviously like take plugins out of the directory if they have malicious code or anything like that. And so those guidelines are part of the submission process. So we do check for that before we Include a plug-in in the directory.

All of the plugins for obsidian 99.99% are open source. And so there’s a little bit of kind of community validation around that as well.

There’s certain principles that are important to obsidian, for example, around privacy, so we don’t allow plugins to collect any telemetry data within the app because we think that that is just part of the kind of Set of values that we care about, you know, a different platform may not care as much about that while still carrying a lot of extensibility, but to feel native to Obsidian, it really needs to work offline and have some of these principles built in. So those are the things that we don’t compromise on where we’re saying no, you know, your plug-in cannot be part of that directory. Actually, it’s almost never that we encounter bad actors. I mean, maybe it’s just because we’re small or the community that we’ve built is friendly and so on, but we don’t really encounter that many plug-ins that really stray outside of those guidelines. The next level is like more of a, I guess, of an editorial question which is what stuff bubbles up to the surface if you’re someone who’s brand new and is kind of asking the question, what can I do with obsidian? That’s where things, I think we could do a much better job and those are things that I’d like to work on over the next few years is. How do we surface the realm of possibilities that is there and right now, you know, the most like simple heuristic is just looking at the number of downloads, like you can see that some plug-ins have been downloaded way more than others and that gives you some sort of filter on which ones are better than others. But I think over time, we probably will get involved in curating a little bit more from an editorial standpoint, not from a who’s allowed in, but rather what do we promote, what do we want to showcase some of the interesting things that people have built and the stories behind them and how they connect to other different parts of the app. So I’m not sure if I’m answering your question, but I think that that’s kind of like trying to find that happy medium.

00:59:12 - Speaker 2: Yeah, well, certainly being relatively niche, having a relatively tight knit and friendly community is just a benefit of if you’re either early on or just a, you know, an independent software company that isn’t necessarily going for the mass market, you know, now, nevertheless, I do think that those values like for example, being able to work offline, that is something that is, you know, fairly unique in this industry right now where cloud software is the norm. And presumably someone doesn’t want to get in there and write a plugin for a piece of software if they don’t have some, you know, vibe with the core principles there, but I could still easily imagine as you grow, having people come in and just say, well, yeah, of course I’m going to access this. Cloud service because I can make a useful plug-in with that and then you come back and say, well, actually this isn’t consistent with our values.

00:59:59 - Speaker 1: Well, by the way, that’s not a requirement. There’s lots of plugins like the open AI plugins, like there’s maps plugins, there’s all kinds of plugins that communicate with cloud providers like so we don’t prevent that, but we do prevent telemetry data and we do require plugins to disclose when they’re using network data and for what purpose. And so, The guidelines are fairly open ended. Like people can really build things that heavily modify obsidian, and we’re OK with that, but we do get involved in, you know, making suggestions on the code when we review a plugin or developers will often ask in our Discord channel, what’s the right way of implementing something and you know, if it’s not already in our developer documentation, we’ll work with them or try to improve the documentation.

And then there’s kind of a more subtle question that I think we want to work towards, which is what feels obsidian, like what is going to make a plug-in intuitive to a user? It partially is the interface, partially their expectations coming into what obsidian is like as a platform as an ecosystem, and we definitely want to encourage.

You know, some developers, for example, are amazing at writing back end software or dealing with like really complex data problems, but they might not be as used to doing the front end part. And so when they go about implementing the GUI part of their plugin, they just need a little bit of extra help with the CSS or something, or what is the right kind of approach to designing this UI for this plug-in. So, Those are things that, I mean, we’re still early on, I would say, and how do we kind of strive towards like making things cohesive um while still giving plugin developers as much freedom as possible, but we definitely skew more towards the freedom than the cohesiveness.

01:01:51 - Speaker 2: I’d love also to ask you about a feature you launched pretty recently called Obsidian Canvas.

Obviously this one caught our attention since Canvas is a kind of fundamental document type or infinite canvas or spatial canvas, different terms for it, something that is sort of a predicate or a foundation of our company, the idea that that’s actually a useful style of a space for thinking, you know, I think it’s well established in design tools, illustrator, Figma, etc. but that for.

Yeah, sort of a freeform exploration of ideas that a canvas actually is a good space and indeed I think as we talked about with Steve Ruiz in our episode on infinite canvases, I think that actually this document type has potential application in a lot of places and software now that sort of computers are sort of good enough to be able to render stuff on canvases fairly reliably, but I’m curious what drew you to want to create this feature and especially how you think of it as fitting in with the otherwise spatial and Yeah, I’m not sure, abstract and dimensional graph of linking that is the text-based knowledge graph that is the core product for obsidian.

01:03:00 - Speaker 1: Yeah, I almost want to go in the opposite direction, kind of answering the latter question first, which is there’s a little bit of a why now, like why are canvas type of Environments suddenly feeling so important and I think FIMA should take a lot of credit for that because they found a way to make something that did exist in Illustrator and other tools for a long time and make it a little bit more accessible to everyone and I think part of that is they figured out how to, you know, make a browser based infinite canvas that was performing and Maybe it’s just the Confluence of these different things happening at the same time that computers are performing enough that we have, you know, technologies like in browser rendering engines that are capable of displaying that, that it all kind of came together at this point.

Where people who are not designers could interact with this type of interface and get some value out of it, being able to pan around and kind of understand this new way of working with files, even basic things like, I think pretty much every candidate. This tool now has like hold space bar to move around, you know, and that just is something that none of these tools even bother to try to teach you anymore, but like when you go into it, if you don’t know how to do that, you don’t even know how to get around.

But now it’s becoming part of the muscle memory as a society that this exists in the same way that like. Pin to zoom or something was this gesture that no one was ever doing, you know, before the iPhone came out and now everybody’s like pinch to zooming things all the time, or all the text editor affordances right?

01:04:39 - Speaker 1: that we’ve all been copying and moving the cursor. Around like all of these things somehow they make their way into like the knowledge and with kids, a lot of my friends are having like little babies playing around with these phones like it’s such a fascinating thing to see how like, what an early age kids can start to adopt these software features. It’s crazy. And so I think there’s something about coming back to that idea of like messiness versus polished kind of finished product. We need more. Ways in the software world to be messy and canvases can be a place to be messy and to be collaborative, and I agree with the notion that in the same way that there are many different text input boxes all over your computer everywhere, like so many things that you interact with are text entry.

Boxes you send messages, you send emails, you write in Google Docs, you’re writing all over the place, but then there’s this other thing of moving stuff around that for some reason has been less prominent in user interfaces up until this point. So just the concept of taking a thing and moving it anywhere on the screen. Seems pretty basic, but up until now we haven’t had that concept be particularly pervasive, like everything feels a little bit more locked in and like what you’re doing on a really basic level is giving people a text form, text box that they can move around and do stuff with or make a different color change in different ways. You’re giving people a Two dimensional plane to work with and what they do with that is pretty open ended.

I love working on these like very basic super super basic things because the open-endedness of it creates the freedom that lets people take it in all kinds of different directions.

As far as how it came to be with Obsidian, I think, you know, there was this kind of demand emerging from our users that they were looking for something like this, and I think that. It speaks to the variety of how people like to think. Some people are very visual thinkers and some people aren’t, and Because we want to provide this kind of breadth of primitives inside of the obsidian ecosystem, we felt that it made sense to build that feature in, and we came at it from kind of a sense-making point of view. So I think a lot of the canvas experiences that people have had in the past came from like UI design with Figma or something like Mirro where they’re trying to diagram or mind map.

But we started from the point of view that you had these notes already that had links within them and their text documents that sometimes can be very long and have many references and many things that they’re pointing out to. And what can you do with that? And how can you Extend that with other types of content, like one of the features that we added early on to Canvas was you can drag and drop a YouTube link in there and it will do a nice embed for you.

It’s not just gonna give you like the YouTube web page, it’s gonna give you like a video version of it that you can more easily set side by side with a note that you have so that you can write in there. So that canvas now becomes not just a place to Describe relationships visually, but also to like create a little desk for yourself inside of like a virtual space where you can make things small or big, and left or right or up or down around what you’re trying to create in the same way that When you’re working on a problem on your physical desk, you might lay out different pieces of paper or, you know, pin them on a pin board or draw them on a whiteboard and having that ability to go in different directions, kind of on the X Y axis gives you a lot of freedom to maintain multiple ideas at the same time.

What we didn’t expect was that it would be first of all, so popular, we didn’t expect that people would love it so much and we also didn’t expect that people would come to Obsidian first for the canvas, came for the canvas and then started writing notes and then started to appreciate, you know, Obsidian as a text editor too. So that was quite surprising because We were designing it from the other way around from uh your writing text, how do you take that text and extend it into a two dimensional space, but the other way around also seems to be very attractive to some people. How do you start with this messy environment and arrive at, you know, maybe some finished text or some ideas that you can then turn into something more polished. Which makes sense, we should probably have thought about that all along, but it was a little bit more of a surprise than we expected.

01:09:27 - Speaker 3: Now to our conversation earlier about plug-ins, was canvases written as an extension, an expansion of core, or could it, or was it even written as a plug-in slash if someone wanted to write their own canvas today, could they do that with the current plug-in API?

01:09:43 - Speaker 1: So Canvas is a plug-in. Actually, most features of obsidians are plug-ins, so you can turn off like a lot of the features of Obsidian. You can turn off backlinks if you want.

Basically everything, you know, it just looks like a plain text with not that much going on if you turn off all the plugins. You can even turn off the file navigator, so you can really have like a very, very limited in shell of an editor inside of obsidian.

So Canvas is a plug-in.

And there were a couple of different canvas-like things that existed. For example, there’s a very popular plugin called Excala Draw. It’s an implementation of Excala Draw for obsidian, so that is more of a, I would say of a drawing tool than a canvas because It’s not meant to embed media, but the author of the Excalaro plug-in also made one called Excali Brain, which is kind of canvas-like, so you could definitely do it, but this is where we did feel like it was important to have the kind of canonical like first party version because it’s such a fundamental concept.

And then, you know, because we are so focused on plain text and open file formats, one of the things that we struggled with was there isn’t an open file format that has mass adoption for Canvas files, so we created one and opened it up and It’s still very early because we only launched Canvas, you know, less than 6 months ago, but we’ve made the type specification open. We’re encouraging developers to build on top of it, and we even would love for more tools to adopt this file format if they go on and make their own Canvas and a very basic JSON kind of file structure that hopefully will be intuitive to any developers who are trying to build their own canvas type of experiences.

01:11:30 - Speaker 2: And that part of it caught my attention because we actually have a Muse bundle export format that’s almost the same thing. I was quite struck by reading over it and because it’s almost obvious. If you have a canvas, there’s these objects that have an XY position, there’s the assets that are associated with it.

JSO is a natural choice, but yeah, at the moment there is really no data interoperability between the canvases, even something simple like, yeah, copy pasting a couple of boxes and a couple arrows out of a mirror into a figma or vice versa usually doesn’t work at all or if it does work, the data is just so.

Now, text editors, especially rich text editors are famous for not being all that great at copy pasting between them, but at least there is some sense that I can bring some bullet points, some headers, a chunk of text from one kind of text entry application or box somewhere to another one.

The canvases really have none of that, but now they are exactly as you said, kind of having a bit of a moment in the last year or so. I’m hoping that that is something that can become a thing and certainly things like file format, compatibility for import export. I think copy paste would be another one.

But also just even the UI primitives, as you mentioned, hold space to move things around so that somehow the parts of the canvas that are sort of the core of the primitive can be as similar and as transposable as possible between these different products and these different domains.

01:12:58 - Speaker 1: And I think interoperability is something that is such a geeky term, and I think that the geekiness of the term has in a way limited its potential as something that we should strive for, and it’s in opposition. This is where something like us being independent and not VC backed allows us to make this like strong decision about interoperability, but it’s almost aspirational because I think that.

The average person is so used to the lock in factor. Like you just expect every tool to lock you in, which is so sad when you think about it that you go into any other piece of software and expect that you cannot open that same file in another tool. But you know, a lot of obsidian users like to use obsidian alongside with it, you know, completely different tool to edit the exact same file. You know, you can have obsidian open and IA writer open at the same time and edit the same file and see the same updates at the same time in both places.

And I think when it comes to something like Canvas, the possibilities are so much broader about the types of things that you might want to do in a tool that maybe doesn’t even visualize things in the canvas way, but edits the position and content of a canvas like. A database viewer, but then like behind the scenes, like stuff is happening on a canvas and obsidian is where you look at the output, but the manipulation of the data inside of it is in something that looks more like a spreadsheet.

I don’t know, but people could do whatever they want, but because the Expectation from users is still kind of living in this world where everything is siloed. I think it has limited the potential of what kind of interfaces and tools people have built up until now.

Hopefully, like having this emphasis on interoperability could open up some other tools that live completely outside of Obsidian that are not plug-ins, but that are extensions to try and do things that obsidian is never going to be good at.

01:15:00 - Speaker 2: So there’s a place to end. I’d love to hear about where you personally and the obsidian team are thinking about the next coming years. I know you have a public roadmap, but thinking a little bit longer term. You’ve achieved a lot. People love the tool. You’ve got the canvas feature now. What’s, you know, what’s your bigger aspirations?

01:15:20 - Speaker 1: Well, you know, we’re 5 or 6 people. We want to stay small. That’s one of the big priorities that we have as a company.

So being small, one of the challenges is like we can’t do 10 different things at the same time, so we have to pick our projects very carefully.

We’re right now working on a bunch of metadata related things and I’m pretty excited about that, but I think we’re gonna stay focused on continuing to build on these primitives and just sort of.

There’s still quite a few areas that we want to improve when it comes to the primitives and how different types of information, so like tabular type of data, you know, facts and things like that, I think have been less of an emphasis for us.

Up until now, we’ve been a little bit more focused on pros, I would say. So how do we bring that into the tool as an area that we’ve been thinking about lately. And so there’s quite a few things like that, and that’s what you’ll probably find on the roadmap.

The other side of it is what I was describing earlier when I was asking the question, what do we need to do to make these tools inviting for New Year’s. I think that if we’re on this arc of And when I say we not just the obsidian team, but everyone you involved who are building these tools, how do we invite people into this realm and make these tools feel intuitive, interesting, worthwhile to invest in? I feel like we’re at such an early stage in that. I think that over the next 10 to 20 years, I hope that. You know, hundreds of millions if not billions of people will start to adopt these kinds of tools.

What does it mean to make these tools useful for the wider population while still continuing to be able to provide the depth of experiences and workflows that people want in the diversity of thinking processes that people want to have.

So that’s probably the big challenge, which is Obsidian has something like a million users today. What would it mean to have many more than that? What would Obsidian need to look like to be able to do that, while at the same time trying to managed to stay very small and like true to our principles. So that’s the core challenge that we’re gonna be working on for the foreseeable future.

01:17:31 - Speaker 2: That sounds like a very worthy challenge indeed. Well, let’s wrap it there. Thanks everyone for listening. You can join us in Discord to discuss this episode with me, Mark, and our community. A links in the show notes. And Stefan, thanks for helping us all to think better thoughts through Evergreen notes and breaking our ideas down into objects, but especially just to try to see if we can use these computers to help us all be better thinkers.

01:17:56 - Speaker 1: Thank you so much for having me. It was really fun.