What is the future of Obsidian? The new Obsidian CEO, Stephan Ango (@kepano), talks about why he joined the Obsidian team and what he wants to do next. I’m Nicole van der Hoeven, and today I’m joined by my friend, Andy Polaine, to find out what’s on the horizon for our favourite note-taking app.

Key takeaways

10 notable points raised in the video were:

• Stephan thinks the proliferation of tools for thought is a positive thing, in that every new tool that uses Markdown makes the standard more futureproof.
• Obsidian’s three major pillars are extensibility, privacy, and longevity. Those are things Obsidian excels at.
• The things Obsidian is bad at are the consequences of those three major pillars (such as fragmentation, difficulty in sharing, and being restricted to Markdown).
• The Obsidian team consists of five full-time people, including founders Licat (Shida, dev) and Silver (Erica, product), Liam (dev of Calendar plugin), and Joethei (plugin reviews and dev ecosystem).
• Upcoming features: better metadata format than YAML (think something like Dataview inline fields), better querying of tasks
• Stephan wants to keep the Obsidian team small, and grow only when absolutely necessary.
• It’s difficult to get exact usage figures for Obsidian, but going by downloads of the app bundle, there are about 1 million users.
• The Obsidian team doesn’t want to bring on venture capital investors.
• Better sharing in Obsidian is something they’re thinking about, but they need to prioritize use cases.
• Artificial intelligence in Obsidian is likely to remain in the realm of community plugins for now.

Below is a transcript of the video.

Introductions: Nicole van der Hoeven, Stephan Ango, Andy Polaine

Nicole: Hello, everyone. I’m Nicole van der Hoeven, and today I’m joined by two special guests. And I’d like to introduce first, I guess, Stephan Ango, or you might know him better as Kepano.

Stephan: Hey, Nicole. Thank you for having me.

Nicole: You’re welcome.

Stephan: Kepano’s great. You can call me Kepano. I’m fine with that.

Nicole: So, Stephan is a designer and an engineer, and most recently, he was announced to be the CEO of Obsidian. So that’s going to be interesting. I actually know you more for the stuff that you’ve made for Obsidian, like Minimal, and also Hider and Web Clipper. But it looks like you’ve been in the industry for a while making, I’d say similar tools.

Stephan: Yeah. Well, I mean, up until now it wasn’t really in the consumer space. I was building startups in more in the B2B area, but I’ve been using a lot of these tools for a long time. So, as a user, I’m older than I look, I guess. I’ve been like, building, I think the first thing that I really got into when I was a teenager was Winamp Skins. And so like, this is like, ’90s, early 2000s, making my first website and things around that time. So that’s how, that was my first, I guess, experience, like, building things in the software world.

Nicole: Awesome. And also here to help me interview you basically, is Andy Polaine, who is a lot of things. Designer, educator, writer, design leadership coach. He’s also co-authored a book called, “Service Design: From Insight to Implementation.” And he has a podcast called, “Power of Ten.” But really, the most important thing is he’s my Dungeon Master. We play D&D every week, and he’s also a good friend. We actually met because of Obsidian. So, welcome to the channel.

Andy: Thanks. Thanks for having me on. I’m excited to be a co-host. Or co-interviewer. Co-interviewer.

Nicole: Yeah, it’s really weird for me to not take cues from you and roll dice when you tell me to, but, you know, we’ll see how that goes.

Andy: I’ve got the dice here just in case we need-

Nicole: Okay, well I’ve got mine too, so. Anytime you need me to roll a save, I’m ready.

Andy: Well, let’s just roll for initiative to see who interviews first, huh?

Nicole: Yeah. Okay, let’s do that. Rollies, what are you going to roll? Okay. Just a d20. I rolled a two, so.

Andy: I got an 11.

Nicole: You go first.

PKM history and getting started with Obsidian

Andy: Okay, all right. I will, so, really, the first question for you is, it’s kind of under a sort of batch of questions really, about why Obsidian, but how did you get started with Obsidian?

Stephan: First, can you give me any like, buffs or stats or anything as my character would?

Andy: Well, I think you get sort of plus one in design and coding, right? So, yeah.

Nicole: What class do you think Stephan would be, Andy?

Andy: Wizard, I reckon. Or an arcanist.

Nicole: Yeah, I was thinking like-

Andy: No, I reckon-

Nicole: Sorcerer?

Andy: Yeah, or a, you know, someone who kind of makes stuff as well.

Nicole: Artificer.

Andy: Artificer, yeah. That’s what I reckon he is.

Stephan: Back in the World of Warcraft days-

Andy: So high end.

Stephan: I played a lot of the paladin class. I like the hybrid class, but I think that I also in like, “Lord of the Rings” really love, you know, Gandalf and those types of guy, Radagast, you know. I’m definitely into those types of characters.

Andy: Someone who can produce something out of nothing. So. So. As you were.

Stephan: I forgot what your question was.

Andy: Yes, so we’re going to lose half the audience straightaway. So, you know, you were a customer, right? You were a user before you became CEO and pretty quickly I seem to remember seeing your, certainly your theme, I think, probably a couple of other things, seeing you in the Obsidian Discord. So, you know, what was your hook into Obsidian in the first place?

Stephan: We’re getting some comments that your mic is very crackly.

Andy: We are hearing it. This is weird, isn’t it?

Stephan: Yeah. Yeah, yeah.

Andy: I had it before.

Stephan: Some people are hearing it.

Andy: I don’t know what’s going on.

Stephan: But to answer your question, I had been using the, over the past like, 15 or so years, I’ve used a lot of different tools to kind of try to solve what for me was mostly a journaling question. So, I wanted to be able to just write kind of, my own diary notes in a tool that, over time I discovered that the Wiki type of format was what I really loved. I kind of fell in love with the wiki link kind of idea. And that idea’s been around, I’m not sure exactly who invented the notation. I think it was the Wikipedia people in the MediaWiki, the kind of like, double-bracket concept. But once I discovered that, I felt that I only wanted to use like, journaling with that. So the one that I used for a long time was TiddlyWiki, and TiddlyWiki has that same kind of bracket notation, but I found that my personal sensibilities as a user, I really wanted something that felt like, very simple and minimalistic. Like, as you may notice from the kind of like, type of themes and things that I’ve made for Obsidian. And when Obsidian came out, I found out about it almost right away, like, in, I guess April, 2020, you know, as right as the pandemic was kicking off. Like, I think a lot of people, I don’t know, around that time were staying at home, were exploring different things and I found out about it, I had spent probably a year trying to kind of customize TiddlyWiki to be a little bit more user friendly, and Obsidian, even in like, V1 or whatever, not V1, it was like, V0.001, was already better than what TiddlyWiki was doing for me in terms of, it had the quick switcher, it had the prompts, it had the file Nav in the way that I wanted. It didn’t quite look like what I wanted, but it had a lot of the kind of, nice user interface patterns that I was looking for. And so I pretty much started making, I took, I was looking back at my notes, kind of preparing for this conversation, and I pretty much went all in on Obsidian after using it just for a couple days. I imported everything, and started making the minimal theme like, immediately. But I was just making it for myself at first. I wasn’t intending to publish it, I just kind of made it to solve my own UI preferences.

Nicole: So I actually used TiddlyWiki quite a bit. The reason that I used it was a lot of, ‘cause I was a consultant at the time, and a lot of the tools that were out there required, like, they were basically on the cloud so it was a SaaS platform that I’d have to log into and not, and those sites weren’t always allowed by whatever company I happened to be working for, so that’s what I loved about TiddlyWiki, that it could just sit on like, a flash drive and it would work. And that’s part of the reason that I started using Obsidian too, is ‘cause it’s local.

Stephan: Yeah. Yeah, and that was something, you know, being able to use it on your computer when you’re offline, just having that control. I was, you know, using it on flights, and so that local first idea to me is, I don’t know, I just find it much more natural and it allows you to do things so much faster. And because I have experience, you know, I wouldn’t say that like, being an engineer is first and foremost what I do, but I do dabble in a lot of different things on the code side and using, I use Sublime Text as my main code editor, but having, being used to using like, a code editor, you just expect everything to be there for you, right away, super fast. Like, even a small delay feels like it gets in the way of my productivity in some cases. So, being in the cloud, it always feels a little slow. And I’ve used a lot of those tools in my previous, especially for work purposes like Confluence or Google Docs or those things. And the sluggishness I feel like, really gets in the way of being able to make progress on whatever you’re trying to do.

Andy: Yeah. I mean for me, it was the portability thing. I switched to Markdown for everything, I think, for my blog. Like really, really soon after Gruber kind of released it. And I used to use Notational Velocity, which I think it was probably one of the first ones I know of, of a kind of this idea of a database that just sits on top of a bunch of Markdown files. And then Brett Terpstra was putting together NvALT, and nvUltra, which he’s still working on. And you know, I’ve been pretty disloyal. I’ve used lots and lots of like, every and probably most people, every Markdown out there. And a lot of them are starting to do the kind of wiki thing too, but depending-

Stephan: And there’s so many of them and-

Andy: Yeah. Yeah.

Stephan: And I think that is such a great thing that now that the ecosystem of Markdown editing tools is so deep and broad, that I often think that Obsidian, to me what’s exciting about this whole kind of area is like, I’m very curious about how we can make digital information persist over a long period of time. Like, how will the ideas that we have right now continue to persist for hundreds of years in the future? We’ve had books for thousands of years, but we’ve only had digital files for, you know, 70 or 80 years. And I like the idea that something like, plain text could still be around for the next few hundred years. And so using something like Markdown is, like, the file in a way is way more important than the apps. Like, Obsidian will eventually become obsolete because technology changes so much. But hopefully, the plain text will continue to live forever.

Andy: Yeah. I mean a bit like you, I’m gonna ask you about Computer Show actually later, but a bit like you, I’ve been around for quite a long time and I thought, I’ve been in digital stuff for like, 30-something years and I got fed up, but one of the reasons I moved to doing everything, I do all that, still do my invoicing, you know, with Markdown and stuff because I had so many times where like, “Oh the app’s gone out of, you know, out of production,” or “I’ve got so much of my work still on like, CD-ROMs.” I keep an old laptop that’s still got a CD-ROM drive, just in order to be able to view that.

Stephan: Oh, really? You’ve got to transfer those.

Andy: It still runs, kinda like, oh, it’s a, no, but yeah, but some of it’s like, Director MacroMind, actually, Macromedia Director stuff, that that’s the only thing you can play on, so you sort of have to keep the machine for it, too. And I transferred a lot of other stuff, but I had the same thing and I started using Evernote once and then trying getting stuff out. It was horrendous. Had the same thing with Notions, of exporting out of Notion, and it kind of dumps all this other crap in there. And so, you know, yeah, I agree. I think the file format is the kind of winner.

Stephan: Yeah, the file is more important than that.

Dealing with competitors

Andy: I’m interested though more as CEO, that you’ve got, you know, there are a whole lot of, well, I kind of would perhaps a little bit rudely say Obsidian clones kind of popping up. It feels like, in the last, you know, even in the last year there’s been a lot of stuff coming up. So, you know, whether you perceive those, is that a challenge for Obsidian? Is Obsidian kind of, just take its own path that each of them have their kind of own flavors? Now is it a good thing, you see it as a good thing?

Stephan: Well, I see it as a good thing because I am still kind of like, a plain text maximalist. So I think that you have to have a diversity of clients in order for that to kind of last forever. Like, I still think the file is more important than the app. I do think that where it gets dangerous is if there’s too many app-specific syntaxes that start to emerge. I think that is kind of dangerous because it reduces interoperability. Like, I think that, you know, Markdown, for all its flaws, and I think a lot of people have frustrations with Markdown, it has kind of stood the test of time, at least for the past, however many, 12, 15 years it’s been around, I’m not exactly sure, but maybe more at this point, I’m getting old. And so, I think that that is one of the biggest challenges that we have with Obsidian is sometimes we wanna do something new. Like, recently we launched the Canvas feature and there isn’t a, you know, off-the-shelf kind of canvas type of bio format. So we, you know, created a new one that is, you know, open and the type specification for it is out there, but there’s no other client for that file format yet. Hopefully there will be, but I think that that diversity of clients is good because there’s a lot of people who have lots of different ideas about how thinking should work, and Obsidian is not particularly dogmatic, like, a big part of, you know, one of the main values that we always think about is extensibility. So how do we make it easy for individual users to customize Obsidian? And also, how do we enable developers to build on top of it? But there’s a, I think it’s very viable to create a very opinionated Markdown or tool, you know, editing tool that takes a different approach than Obsidian. And I’ve seen many people who enjoy using more than one, different one for different parts of their workflow. So yeah, I think it’s good. I think we’re living through a little bit of like, this Cambrian explosion moment because for some reason right now, in the past couple of years, like, people have realized this a really exciting problem to work on and I think we all learn a lot from each other, so, I see it as a good thing.

Nicole: So one of the things that you said, ‘cause I trolled through your blog while I was thinking of questions for this. You had a post about how great tools choose to be bad at some things. So, what do you think Obsidian has chosen to be bad at?

Stephan: Well, I think that design principles are only real design principles if you can take both sides of it and create something compelling. Like, for example, we talked about extensibility. That’s one of our, I would say there’s like, three major pillars to what Obsidian is trying to be. One of them is around privacy, one of them is around longevity of your notes, and one of them is around extensibility. So those three things are the things that we want to continue being really good at. But you can take the extensibility example as kind of one of our design principles. We wanna make Obsidian super extensible. But you could, I can see why some people would prefer an app that’s not at all extensible, that is, you know, completely designed by one small group of people who are trying to make a very cohesive experience. I think one of the big challenges with Obsidian is we’re always asking ourselves, “What should be a community plugin versus a core plugin?” We’re always asking ourselves, “How can we make community plugins feel more native to the app and more cohesive and intuitive to people? Do they make sense with the rest of the app? Because any developer can basically build anything inside of Obsidian. And so it, I think it is a common point of frustration for some new Obsidian users that they have to download a lot of plugins and some of them feel a little janky. So, you know, Obsidian has chosen to be bad at that, with the trade off that it allows a lot of freedom for people to customize the app. And you can take the same example with, you know, privacy and longevity. Those have their own trade offs as well. Andy, did you wanna say something?

Andy: I was thinking, you know, it reminded me of what you just said. It was just as you talked about the plugin thing. One of the tricks that Apple did with the iPhone, was to make the phone bit an app. And it sounds just like, a really banal thing, but up until then there were phones, smartphones, were kind of phones with sort of app bits in them, and they were really janky. And I think one of the things also, obviously not having a keyboard, having a touch screen, means that that thing where you get whatever app you’re in is completely the app. All right? So, you know, I’m in FaceTime, it’s completely FaceTime, I’m in Obsidian and it’s completely Obsidian, the whole phone turns into whatever app you’re in. And I feel that, you know, in a way that’s what the plugin mechanism can do for Obsidian. At its base it’s, you know, it’s a text editor that’s browsing as a viewer on a set of Markdown files. But you know, as soon as you start putting, particularly some of the more powerful plugins in it, it becomes something completely else. And I think, I find it really kind of powerful that you can switch that stuff on and off. I can see that the, you know, I was sharing my students the other day ‘cause they were talking about design research and how you can use Obsidian, and quite a lot of them use Notion. And I can completely see why they use Notion, right? ‘Cause it’s just, you know, zero learning curve for them. Until such time as they want to take that stuff with them, and then they’ll find out. And I think some of that stuff’s a bit like doing backups. You find out the hard way, right? But I kind of feel like that is a, it’s been a smart trade off to make with Obsidian, with the plugins. And I was just wondering at this point-

Stephan: But it has its downsides. I mean, I think the point, you know, on privacy, the way we think about that is that is having ownership over your files. They’re not in somebody else’s database. They’re on your computer, on your device. You know, if you choose to use the first-party sync, everything’s end-to-end encrypted, and so we have this kind of strong pillar around privacy, but it has the trade off that it makes certain things harder for us. For example, people ask us, would we ever see a web client for Obsidian? A like, online, you know, cloud-based Obsidian? And that’s a lot harder for us to do because we don’t have access to your files. We would have to compromise on that value of encrypting your data and kind of being able to see that information on our side in order to be able to present it back to you. And that makes things really tough. Unless, there’s some, we’ve thought about this question a lot, but it really is a very hard problem to solve. But for someone who, you know, for whatever scenario they might be in, either can’t access a local app because they, you know, there were computer restricts, you know, apps and they can’t download Obsidian or they’re on a school computer and they have to access it from the school computer, it does present a limitation. But I think strong principles force those kinds of trade-offs where you could take either side and make a compelling and legitimate app. We just, you know, have chosen our own particular set of trade off and that’s what that blog post was about. So. So there’s other things that Obsidian is bad at as well, like, relating to longevity and like, the choice of using Markdown, for example. It’s not the most intuitive to some people. I can come up with a long list of things that Obsidian is bad at, in favor of something that it wants to be really good at. And I think that’s a, making trade offs is our job.

Nicole: Well, one of the things that I hear a lot is that Obsidian is too developer focused, or some people think it’s not approachable. And I think it comes back to what we’ve been talking about. You do have to make a decision, you do sacrifice a little bit of usability in exchange for flexibility and modularity. My background is in performance engineering, so I care a lot about performance and I love this idea of making the most streamlined possible app, and then just leaving room for extensions so that other people can make their own decisions. I would rather have that, than have an app that has everything baked in and I have really no choice as to what to turn on or off because my use case is going to be different from everybody else’s.

Andy: That’s Microsoft Word, right?

Stephan: Yeah.

Andy: When you just have the sort of, bloat and they get those screenshots, sometimes, of every palette open and there’s about this much space, they actually edit some text.

Nicole: By the way, do you like my shirt?

Stephan: Oh yes. Classic.

Nicole: Where’s yours?

Andy: Yeah.

Nicole: How disappointing.

Stephan: I know, I haven’t, I’m such a minimalist-

Andy: It’s a print.

Stephan: I haven’t gotten my Obsidian merch yet, but, yeah, you’re looking good over there.

Andy: So, you know, you’ve been contributing to Obsidian, pretty much from the get go, and then you started working with them. What is it with them? It’s the, the Licat and Silver, basically-

Stephan: It’s the smart, yeah, the smart team.

Joining the Obsidian team

Andy: Two-person team. So, you know, how did that, I don’t know how much you can reveal, how did that come about and then, and what was the sort of transition? When was there this kind of moment of like, “Hey, how about becoming CEO?”

Stephan: Yeah, well, it was a very organic process because I, you know, was running a different company at the time when Obsidian first came out. And so, I had, you know, started building minimal theme, a bunch of different small plugins here and there, the Web Clipper. And I was a very vocal in the Discord and forum about some of the things that I thought would be, you know, improvements or bug fixes. And at the time, the Obsidian community was way smaller than it is today. And so, I was often interacting with Licat and Silver, the founders of Obsidian, through the Discord. And just kind of, just talking very organically about here are some things that could be improved, here are some, you know, bugs I found. Can you help me with this feature of Minimal that I am trying to, you know, build into my plugin? Can you give me some feedback on how to do this properly? And so, we started interacting, you know, talking about Obsidian for a couple of years. And, I’m not exactly sure at what point it started, and maybe it was around the time that, I’m trying to like, actually remember, but I think there was like, a Gems of the Year contest where Minimal won like, the Best Theme category, and I think I started DMing more with Silver and Licat around that time, and just kind of getting to know them just over chat in Discord. And they found out that I had been, you know, running some startups and asked me for some advice on a couple small things here and there. And so we just kind of kept that conversation going, and at some point last year, there was this kind of idea that Obsidian wanted to kinda reach the 1.0 release. And the 1.0 release, I mean I think it, you know, it’s funny that it, it was really a milestone moment, because we, that’s when I started becoming involved in the project and they asked me to basically create the new default theme. And so I was working with them part-time, kind of taking a lot of my experience on Minimal and bringing that into the core app. But also, they were asking me about some of the things that I felt were on my like, top priority list of small details that we felt were still kind of embarrassing or like, wanted to really get nailed down before we could really feel like we could call it a 1.0. And, that’s when I started working much more closely with them, as well as Liam, who is now a full-time part of the team. You might know Liam from the calendar plugin. He’s amazing. We have basically five people full-time. Joethei, you might have seen around the community who helps a lot with plugin reviews and the developer ecosystem, and then a few other folks who’ve been helping us part-time, like MG Meyers, who you might know from the Kanban plugin and style settings is working on, has been working on Canvas and kind of like, led the Canvas effort. And so, the more conversations kept going on with this project, I just realized it was so much fun and I love Obsidian so much. Like I just, I really want it to continue to be an important tool in my life, in a very selfish way, but also I think that Shida and Erica, Licat and Silver, are such mature like, entrepreneurs, they’re quite a bit younger than me. They make me feel old all the time. But it’s a great exchange of knowledge and yeah, so we started discussing this idea and they were on board. I think it’s a very, we have very complimentary skills. I’m really excited about, like, Licat’s one of the most incredible engineers that I’ve ever worked with. I’ve worked with a lot of like, amazing engineers throughout my life and I have like, so much respect for him. He’s just super, super impressive and continues to be like, they’re both gonna continue to be involved for the foreseeable future. And Erica has just an incredible mind for, you know, product and how to think about what users really care about. And so, and has been, I think, the person who’s, in a way, influenced why Obsidian has such an amazing community more than anybody else. Just kind of from the beginning, making that an important part of Obsidian, I think is a lot of credit to Erica. So, seeing, like, getting to know them just made me realize like, there’s nothing that I would rather do than help that community continue to grow, help make sure that Obsidian has the, kind of like, structure that it can last for a long time, kind of going back to that principle of longevity that matters at the app level. How do we actually make that viable at the company level so that the kind of, product and the company can be aligned? So yeah, that’s why I kind of believe that-

Nicole: I also just wanted to call out-

Stephan: Very organic.

Nicole There’s some people who are asking-

Stephan: Oh, where do we get the merchandise?

Nicole: Where do we get the merch? By the way, I bought this. And it is here.

Stephan: Yeah.

Nicole: Just so everybody can see, ‘cause there’s some awesome things on there.

Stephan: Yeah, I think it’s-

Nicole: I want I wanted to ask you-

Stephan: Yeah.

Building intentional tools for tool builders

Nicole: Yeah, it is. So I wanted to also ask you, earlier when I was introducing you, I said that you have worked in similar spaces, which I realize now is a bit odd because the companies that you had worked on before aren’t note-taking apps. But what I meant by that, is actually I’m seeing a theme in what you’ve worked on and it’s all about creating things that are minimal, intentional, well-made. And I think like, in your previous company, Lumi, you were talking about packaging, and making that a little bit more transparent and intentional as well. I know you’re also an industrial designer, or you claim not to be one anymore, but that’s also something that permeates your work there as well, doesn’t it?

Stephan: Yeah, I’m a generalist. I like to dabble in a lot of different things. And so, it has taken me in a lot of different directions, but I think the common theme is I’ve always been interested in making tools. So everything basically in my professional career has been a tool of some kind. So in the case of Lumi, it was a tool, it was actually two different tools. One was for factories, to help manufacturers kind of bring their capabilities online, and a tool that helped businesses who need to collaborate with manufacturers, find the right manufacturers and work with them. And so, I like building tools that kind of empower individuals to do something that they couldn’t do before. And so Obsidian definitely fits into that because I feel like, I mean, in my own personal life, it’s been so rewarding and exciting to use Obsidian. Like, I feel like, my thought process is clear and more reliable when I’m using Obsidian. And I feel like, that’s such an exciting superpower that I have that I want to help more people get that feeling, too. So yeah, like, throughout my life basically like, that theme just keeps coming back of like, making tools that help other people do whatever they’re trying to do.

Nicole: And it’s also helping other people make other things, right? ‘Cause like, you know, you’ve also worked on things like, that kind of pare down what is out there, because there’s an information overload that’s out there and I know you’ve made like, Tidy, which kind of streamlines what we even see on the page. What do you think, is there anything in Obsidian, ‘cause we’ve talked about things that it’s bad at. Are there any big features that you’d like to introduce to it, that don’t detract from, that don’t make it like this bloated bit of software but still are along the same lines of being intentional and helping people?

Stephan: Definitely. I think that the area we want to keep working on is the primitives. Like, what are the basic things that you can use that you can build on top of? So, we have a public roadmap that’s not always the most up to date, but one of the big things we’re working on right now is around metadata. So, a lot of, I’m a huge fan in my personal use, of DataView is one of my favorite plugins for Obsidian, and I use it for everything. And kind of asking ourselves like, “What can we do that would enable the user and developer ecosystem of Obsidian to leverage metadata even more?” So, you know, you can, there’s a couple different ways that you can add metadata to your notes with Obsidian. One of them is through YAML in the frontmatter, and one of them is through DataView has these like, custom fields. So that’s an area that, we want to, like, the author of DataView had to create this new notation with the double colon. You know, you put a key in a value and separate them with a double colon, and that notation was essentially invented by the author of DataView because it’s so painful to write YAML. And so, what can we do to clean that up a little bit and make metadata feel more like a first-class citizen from the user’s perspective, but also enable more developers to build on top of that. Yeah, API docs and developer portal. I’m looking, I can’t even remember what we have on here. Tasks. There’s a few other things that we have up our sleeve at the moment, that I’ll be excited to share in a little bit. Talking about the roadmap-

Nicole: A few questions, Stephan-

Stephan: Yeah, go ahead.

Nicole: You mentioned DataView. I know that he’s working on DataCore, right? Is that going to be something that’s going to be built into Obsidian or is it still going to be a community plugin? Honestly, I would never use Obsidian without DataView.

Stephan: You know, I haven’t really, I’m not exactly sure what the goals of DataCore are. I’m not super informed about what he’s trying to do with that. I think that our job is always to find the right balance between what feels like a core plugin, what feels like it should be part of the community. And so, I think that long term we do wanna have good ways for people to add metadata to their notes and perhaps view kind of an output of that in some way. I think that’s kind of a long-term project, but in the short term, the metadata improvements is really about making it a lot easier to add metadata to your notes and then hopefully that makes things like DataView easier to use. But maybe long term will consider some more ways that people can like, look and view their data, or the metadata from their notes. I do found that DataView, one challenge with it is it is pretty nerdy. Like, it does require, it has a syntax that’s a little bit, you know, like, SQL and I think that that’s not the most user friendly. So we’ve seen quite a few plugins pop up, like DB Folder is one that’s kind of interesting that makes it a little bit more WYSIWYG. And that’s one of the big benefits that we have as a community is that people can build plugins that optimize for different types of users. People who love writing SQL, they’re gonna love using DataView, but people who prefer something more WYSIWYG are gonna prefer a different type of UI. And we can be patient in our approach to what we actually wanna build into the core because we could just kind of watch what people like, what people make, and decide later down the road what we wanna build ourselves.

Growth of Obsidian as a team and company

Andy: Do you think there are, I mean are there any plans to then grow the team? Or is partly what makes it special, the fact that you are a very small team?

Stephan: I think that for now we wanna stay really small. We wanna be as small as we can be because, you know, one of the other things that’s really important that made me want to join Obsidian, is that from an early stage, Shida and Erica decided that they didn’t wanna bring on investors, like, venture capital investors. And I have, throughout my career, worked and built companies that have used every single funding model under the sun. Like, my first company was one of the very first Kickstarter campaigns ever, back in 2009. I’ve gone down the venture capital route, I’ve done bootstrapping, like, I’ve tried it all. And I think when it comes to the goals of Obsidian that we’ve been talking about, like, longevity in particular comes to mind, but also privacy and extensibility, I find that is not very compatible with the VC route. And, you know, if you just kind of survey the ecosystem, you’ll find that actually, almost all of the kind of big PKM apps, if you wanna call them that, have raised a lot of money, like, millions of dollars from, or hundreds of millions of dollars, even. And the problem with that is that, I think it’s misaligning the incentives in the long term for, you know, the user and how they want to kind of, store their notes. Like, basically it incentivizes certain behaviors that are kind of against our principles. And so for that reason, we’re trying to build very intentionally and also work through the extensibility part of our like, principle to enable the community to customize things more. Like, we don’t feel like it’s necessarily our job to make this like, huge app that has every single feature. We like seeing how the community kind of extends Obsidian in their own ways. So I would say that growing the team, it will be more by necessity than by a desire to like, try to do everything ourselves. We’re getting to a point now where, I would guess, we don’t actually even know how many Obsidian users there are because we’re so focused on privacy, we don’t have analytics, we just like, don’t know. But we can guess based on a few different like, things like, how many people downloaded the app bundle, that there’s probably close to a million users of Obsidian and only basically five core members of the team. And eventually, from a support standpoint are things, things might start to break and we might need to hire people to help on that front. But we’re trying to stay small if we can.

Andy: Yeah. I think you’ve been very diplomatic about the VC thing. I mean, in general it’s, you know, you’ve got, in my coaching practice, and I coach a lot of design leaders who are kind of suffering in startups or scale ups because of this constant pressure to scale, because this constant pressure for just more and you know, I want a 10 X on my, you know, investors want 10 X return on their investment in some stupid amount of time, things like that. And it definitely leads to that, you know, what Cory, Dr. Al calls sort of, an end certification, you know, this thing where then there’s a choke point and we have everyone’s data on our server and they have to use app and then you just start, it becomes a sort of process of trying to extract value. And I-

Stephan: Well, I’ve touched on the thing-

Andy: Oh, go on, go on.

Stephan: To defend VC, I will say there’s some businesses where, it kind of is necessary because especially, we’re in a privileged position, building software, because software has no cost to us. But there’s some companies, you know, that have to produce a physical product in the world-

Andy: Yeah, that’s true.

Stephan:: And they have to buy the materials and put in orders and so they need to raise money somehow, and VC is one of those paths. You can get loans, like, there’s different ways. So I think it, I’m not like, I just have a kind of, like, a more zen point of view on that where like, I understand why a lot of companies go that direction and it makes sense. I think in the case of what we’re trying to build specifically, it’s a bad fit. But it’s not necessarily a bad fit for everything.

Nicole: So-

Andy: You mentioned anything just now which is, oh sorry, Nicole. What were you gonna say?

Sharing, Obsidian Sync, Obsidian Publish

Nicole: I was gonna say I’d like to go back to, and just quickly, to what Obsidian is bad at. And you know, Stephan, I love Obsidian. It is, I’m all for it. I am possibly the biggest fan. I think one thing that Obsidian is bad at that I really would love to see improved, is sharing. Because I think that is the thing that stops most people from using it, because, you know, using it on your own in an isolated way is one thing. It excels at that. But anytime that you have other people that you wanna do that with, as an example, Andy and I play D&D. Okay, this not a work thing, right? But it is a recreational use, and honestly, it takes up a lot of, my usage of Obsidian is role playing games. And in that group, there are four of us, three and a half, who use Obsidian. We can’t share our notes. We can’t have like, one repository for all of us to collaborate on. So we have like, individual notes, and yes, we could use something like Dropbox to share those notes, but then Andy for example, can’t, or wouldn’t want to share all of his notes with us. Maybe some of them he would. So I would love to see like, a little bit more permissions around sharing.

Stephan: Yeah, I think sharing is a big, like, it’s a challenge because there’s like, three or four different use cases nested under sharing that are quite different from one another. So, we have published, so we’re doing a lot of work right now, actually, on making some publish improvements. There’s a big batch of publish improvements that just went live like, a week or two ago, and that’s more about sharing to the external world, an entire vault, and we wanna keep improving that. Then there is sync, which is like, has this concept called Shared Vaults. I don’t know if you’ve tried that out, but Sync Shared Vaults is actually what we use internally on the Obsidian team. So everyone who’s like, you know, between the five of us, are working on projects, we keep our notes in one shared vault. Then there’s also the idea of, “Hey, I’m writing this note, I just wanna send a link to someone so that they can view it.” They don’t have Obsidian or they don’t know, you know, I don’t wanna like, put them through the process of having to, you know, learn everything about Obsidian just to read this one note. And that note might be, you know, semi-private, like, it’s like, something where it’s not like, a published use case where you want the world to see it, you just want this one person to see it. And then there’s a collaboration use case where like, Google Docs or Notion makes it really easy for a few people to be like, collaboratively editing something, even if they don’t, or have Obsidian, for example. So like, those are all kind of different problems to solve in different ways, and I do think that we want to find ways to solve those problems, figuring out what the right order, that’s another challenge about being a small team is like, what do we work on now versus later? But I think all of them are problems that we would like to solve, you know, over the kind of medium to long term for sure. We’re very aware of that.

Andy: It feels like there’s a bit of a thread here, though, as well, when you’ve talked about, you know, ‘cause obviously the sharing bit brings up the privacy issue that you talked about before.

Stephan: Correct. Yeah.

Andy: And it also links to the, is there a browser-based version of Obsidian? You know, because one of the ways of obviously, easily sharing and collaborating on a document would be to have kind of a browser-based editor, even if it was just for that one document. And it’s almost like, you need a kind of, instead of a separate vault, a sort of enclave in your existing vault that you can say, “Oh this the shareable bit and all the rest of it, you know, remains encrypted,” let’s say, I don’t know if the encryption thing is an issue, but you know, at some point once you share it, you have to not do that, right.

Nicole: So I think-

Andy: And that may be at the beginning of the browser-based version.

Nicole: I think my problem is that I am able to do those use cases, but I use separate tools for each of them. I would really love to streamline in this area. Like you were talking about, being able to share a link. I use Etherpad with that, but it’s not the best. It works, though. And then when you talk about using Sync, I also subscribe to Sync, but that also requires other people to have Sync. So for those people who don’t use Obsidian enough to warrant Sync, I use Dropbox. And then at work, because we have all of our internal docs on GitHub repos, I use GitHub to bring my Obsidian notes and share them. And that brings a lot of things but also brings like, comments and being able to roll back changes, and, you know, handle merge conflicts, and that’s not something that Sync does as well for now.

Stephan: I totally agree. Yeah, I think that those are things that we have to eventually solve. And I do, going back to what feels like it should be a core thing and what feels like it should be a community thing, I think that sharing is something that belongs in core, or at least core needs to have good sharing mechanisms for all of the different use cases. Publishing, collaboration, single note sharing, is just a matter of, how do we do that? In what order do we build those features? And can we build them in a way that is consistent with our principles?

Task management in Obsidian

Nicole: Can we talk about task management? ‘Cause that was something that was on the roadmap. So is there anything you can share with us on that front?

Stephan: Our goal around that one is fairly conservative, I would say. It’s not gonna be a major, you know, complete revamp to everything. It’s more about, how do we make tasks queriable in a better way, like, and also adding better date supports? So we’ve been thinking a lot about how do due dates work with tasks? So right now, there’s ways that you can do that with community plugins, but we wanna figure out, should there be a core way of dealing with due dates? Again, thinking about the API first and how that might, you know, enable developers to come up with some new experiences on top of that. But I would say that that project’s been a little bit on the back burner for us because, we had some other ideas that we wanted to prioritize first, like Canvas, and some of the things that we’re working on right now, like metadata and publish improvements, for example.

Nicole: It does occur to me that a lot of these things, like task management, could be helped if you share a log something like DataCorp or DataView. ‘Cause the functionality is there, you just need to make it look pretty, but you’re good at that.

Stephan: Yeah, I mean the challenge there is just again, like, you know, two, three people building things. We only have so much bandwidth and kind of finding that happy medium between, we wanna make sure that we’re, you know, releasing features at a good pace. But every time we add new capabilities, we also add, you know, add to the surface area of potential bugs of potential issues that might come up or edge cases with certain plugins, we increase the number of feature requests, like, when we add something like Canvas, then we get a whole bunch of new feature requests from people who are like, “Oh, Canvas is great, but it doesn’t do X, Y or Z.” So that, I don’t know that anyone has ever solved that problem. It’s just kind of a problem of building software in the world. But, that is probably the biggest challenge that we deal with as a team is like, how to find the right balance of what to build now, how comprehensive do we want this feature to be versus, you know, starting small and kind of building on top of it. But trail locking a plugin, we tend to, we’ve done it a couple times, and when we do, we tend to rewrite it from scratch pretty much. And so that’s a pretty big effort. Like, rewriting something like DataView or Data, I haven’t looked at the code for DataCore. It’s not a small task. It tends to be a pretty big thing.

Artificial intelligence in Obsidian

Nicole: Another question here by Imaginationscene and another one by another user called Remy, “What about AI? Or integrating something like ChatGPT?” Is that something that’s on your radar?

Stephan: I’ve been thinking a lot about this. I think that right now, so this kind of like, goes back to the benefit of the extensibility of Obsidian. I’m pretty sure that Obsidian was the first app that had ChatGPT or like GPT-3 integrated into it, because someone in the community built a plugin for that called Text Generator. And now I think there’s three or four plug-ins out there that connect to the Open AI API. So, the cool thing for Obsidian is that because it’s so easy to build plug-ins, people tend to go really like, we both have the ability to be very fast and very slow, so, the community can go and build a plugin that solves that use case. And I think that they’re quite good, actually. I’ve used Text Generator a bunch, but we can also be patient about what’s the right way of building it for Obsidian. And I think that there’s a lot of companies right now rushing into adding GPT to everything. My personal preference when it comes to this, is I would like for it to all be local first, completely private. I don’t wanna send, you know, data that I don’t, like, I wanna have complete control over any data that is being sent. You know, there’s this concept of training and inference in, we’re getting like, right into the weeds of how AI works, but I would like, someday, to be able to have maybe a plugin, whether it’s a core plugin or a community plugin, that can do some training on your vault and kind of learn from your vault and be able to do the inference locally on your computer, so that it completely aligns with the principles of privacy and local first that Obsidian is known for. The challenge with that right now is that that’s quite intensive on the graphics card, and it’s quite complicated to set up locally for users, and we’re still kind of waiting to see whether a company like Stability, that they made stable diffusion and is an open-source platform for like, image generation, there’s been talks about them releasing a language model that would be open source. So we’re kind of waiting for all of those things to come together before we make a decision. We would, I think we would prefer to build something that kind of, is aligned with our principles, but in the meantime, there’s a bunch of plug-ins that you can use that solve that use case and I think they do quite a good job with them, actually.

Nicole: Stephan, have you heard of Napkin?

Stephan: No, I don’t know Napkin. Tell me more.

Nicole: It’s, yeah, it’s an awesome service that I recently have been using. And it uses AI in a really interesting way, because I think there’s a lot of focus on using AI for like, generating blocks of text or outlines or images. But what Napkin does is it looks at your vault, well, it’s not a vault because it doesn’t do it for Obsidian, it looks at your Readwise highlights, and it uses AI to tag them intelligently. So, it has this interface where you can determine, you can click on like, a thought or a highlight, and it shows you intelligently linked other highlights that you already have, that you may not have put together. And sometimes they’re like, they’re not even using the same keyword or they’re not tagged the same way, but somehow it still figures it out. That kind of thing would be invaluable for Obsidian because there’s a lot to be said for manually linking things, but getting some help would be nice.

Stephan: I think that sounds awesome. I’m seeing a lot of comments here in the chat of like, people who are suggesting some different ideas and some plugins that I’m not even aware of that are already doing some more local AI stuff. I think that’s a, I think what’s amazing about the Obsidian community is we have people kind of, at every level of the spectrum of nerdiness. Like, we have like, people who are some of the most advanced engineers that I’ve ever talked to are developing their own custom plugins, and then we have people who are complete beginners and novices, not engineers, people who are just learning about note taking. We have a lot of young students in the community. And so, I like that we have the ability to have those people who are like, all the way at the deepest end of like, understanding all of those principles, able to start working on, you know, potential plugins and extensions that could do something like that. And that can be kind of a breeding ground where we can learn, what do people actually want out of that type of tool? But I do think that if we want to continue to stay true to these kind of principles of being local and private, there’s gonna be some constraints that we have to deal with that are actual like, performance constraints and technical constraints, that may mean Obsidian will be a little slower to get to those features than apps that are completely cloud-based where they can upload that to a server and do that, you know, somewhere else.

Andy: Hm. Yeah.

Stephan: So I think it is a very tricky challenge for us to try to solve that right away. I think we’re gonna tend to watch what plugin developers do, and then start to think about that in a year or two. ‘Cause right now there’s, I mean, literally every month there’s something new going on with AI and the open source packages that are going on. So, we’re maybe a little less cutting edge than the plugin ecosystem can be, if that makes sense, on the core side it.

Andy: It feels like you could get dependent on something very quickly there as well, and just kind get really kind of bogged down. And then, you know, a couple of people have said keep that as a plugin, it’s much better that way.

Accessibility in Obsidian

Andy: There was a question, it was from Daniel Gartmann further up yeah, I don’t use a screen reader, but I think, you know, something that would need to be much more sort of, core fundamentals as obviously accessibility. I don’t know, maybe Daniel can say how, I don’t know how accessible Obsidian is to use with screen read, but read is-

Stephan: We’ve done a lot of work in that area with, yeah, I would be curious to hear a little bit more detail from Daniel about what they’re looking for, but yeah, Obsidian is already compatible with screen readers. We have done a lot of work on that side, but if there’s things that are missing in that respect, I would say let us know in the community, on the forum or on Discord. We’re definitely, I agree that building that into core is a must. I think that’s super important. We’ve added a lot of new customization features over the past year, so I’m actually kind of curious to see if there are themes that get developed that are focused on different use cases. It’s much easier today to design themes than it was a year ago in my, one of my, that was one of my biggest contributions over the past year from a code standpoint, was rewriting the way that themes can be built. And so that means that you can very easily now create like, very high-contrast themes. Themes with different font sizes, different fonts that might help people who have, you know, who need help on the vision side. But then there’s other things like, ARIA labels and other kind of web technologies that we’re able to incorporate that make it easier with screen readers for sure.

Tags vs Links

Nicole: Okay, we have a question that is a little bit controversial. It’s from Germ Germain, whom I actually, by chance, met a few weeks ago in person. He says, “What is his opinion (Well, I’m gonna ask both of you) about tags versus links?”.

Stephan: In my personal use? Or, how do you wanna clarify that?

Nicole: Yeah, in your personal use. Which one do you use more?

Stephan: So I tend to use links more than tags, personally. So, I do have quite a few tags, but I tend to use links because I like to see everything in the, all the relationships in my backlinks, I guess, is how I would say it. And I like that links, for example, like, I’ll have a top-level categories for books that I’ve read. And so I just have a category called books. And I link to it. In my template, I still use tags. I don’t actually really know why. I just feel like I should have it there for future need.

Andy: Future proofing.

Stephan: Yeah, for future proofing. Like, if I ever create a note about a book, it has both a tag for book and a link to the category books. So that redundancy helps sometimes if I’m creating like a DataView list or something I have access to both of those different ways of filtering down. But I tend to prefer links. I think links have a lot, are more powerful and cross over more different things. I think tags, where I use them the most, is for kind of very narrow meta segmentation of a very specific thing, where there’s only like, a very narrow subset of notes that needs to be grouped in a way that a link wouldn’t make sense, or I don’t really wanna have a page for that concept. Am I making sense?

Andy: Mm hm. I’m still here.

Stephan: But also, it’s not-

Andy: I’m still a fan of the folders, you know.

Stephan: Oh, I hear you.

Nicole: No. Do you really?

Andy: Oh, half the people have gone, “Oh god!”

Stephan: Yeah. I’m very-

Nicole: What are you even doing here?

Stephan: In my personal organization-

Nicole: No, I’m joking, I’m joking.

Andy: I can explain.

Stephan: This is why I feel like Obsidian being not dogmatic about this is great, because people think in different ways. Like, I love that people can use folders or tags or links or all of them. Do whatever works for you. If it works for you, that’s great.

Andy: I think partly goes back to what you were saying right at the beginning, which is, you know, where the file format is the thing, right? And so I use several other things on that group of files that I have there. And whilst, you know, it’s great if you can search the tags or you can do whatever you want with the tags or links in Obsidian when you are trying to find a file, to kind of pull into something. I use Marked, Brett Terpstra’s Marked quite a lot for Output. It’s a massive pain in the ass. Well, I mean, I use search a lot anyway on the Finder, but, you know, I much prefer to have some organization that works across other things that aren’t gonna handle tags. So that’s why.

Stephan: Well, to be fair, in our team vault, we do use folders for our like, internal grouping. But I think that that use case is very different than my personal notes.

Andy: Yeah, yeah.

Nicole: We have an answer from Daniel, actually. He says he has issues both for Windows and iOS screen reader support.

Stephan: Okay, well let’s-

Nicole: “I love the potential that Obsidian has for me as a blind person who’s already using notes in a text-based format.”

Stephan: Jump into Discord and let’s talk about it.

Andy: I don’t know how accessible Discord is, by the way.

Stephan: That’s a great question.

Andy: Because we might be saying jump into something that’s really terrible for you to use on Discord. So, yeah, on the forum, I’m guessing the forum-

Nicole: The forum.

Andy: It runs on Discourse, doesn’t it, of the forum?

Stephan: Correct, yeah.

Andy: I would imagine it’s kinda more accessible.

Stephan: If you go to obsidian.md/community, you’ll see the links there and send us your feedback. Or if you wanna just email support@obsidian.md we’ll take a look. But the best is definitely through the community and we’ll definitely take a look, yeah.

Andy: Tell us who does that email go to? Of the three of you?

Stephan: Well, right now it’s like, a kind of a shared thing where you kind of look in there, but I think that most people who use the support inbox are for kind of like, our enterprise users. We prefer to use the community as the main place because there’s more transparency around that and like, other people can jump in and add their own kind of experiences with whatever the feature request or bug might be. So in general, that’s why we kind of direct people towards the community if they wanna provide some feature requests or bug fix requests.

Andy: I was kind of hoping it triggered like, a kind of a bat signal that projected us an Obsidian sort of into the sky, the Obsidian crystal into the sky, and you all kinda come running.

Stephan: Yeah, unfortunately, it’s just one of those other things that, as we scale up, we have to find good ways of parsing what the community is looking for and how we can fit it into our roadmap.

Message to the community

Nicole: Well, I don’t, I would love to have you stay, both of you, stay on for longer. You’re both welcome back, but I also wanna be mindful of your time. So thank you for coming on to talk about the future of Obsidian, Stephan, and for helping me interview him, Andy. Stephan, is there anything that you’d like to say to the community before we end?

Stephan: Wow. Thank you. I just, I’m so thankful. I think that the thing that I love about Obsidian is, so many smart people and kind and like, kind people are part of the community. I love, you know, watching your videos, Nicole. You do such a great job of kind of making Obsidian feel more accessible. And I think that, in a way, like, beyond what we’re working on in terms of features and so on, like, the fact that we have such a friendly and collaborative and helpful community, is one of my favorite things. And so, I’m extremely thankful for that. You know, continue to be, continue doing an amazing job, being nice and helpful. Like, I think we’re still at the very beginning of people adopting these tools. Like, anyone who’s watching this is probably in the like, top 0.01% of people who knows anything about these, whatever you wanna call ’em, tools for thought, how you like, some people have been calling it integrated thinking environment. Like, we’re creating a new generation of tools that help people unlock the way that they think. And anyone who’s watching this right now is kind of, pretty much on the cutting edge of that concept. And I’m really curious what happens if that idea becomes more pervasive in society. I’m very hopeful that it can help us solve a lot of big problems. And so, if you’re one of the, you know, nerds, I say that lovingly. Geeks, dorks, that are part of this conversation, I consider myself one of those, introducing people to Obsidian and being part of, you know, helping them learn how to kind of, make their thought process more reliable is just an amazing gift. Like, I just think that’s like, and you’re doing that every day with your videos, Nicole. And so, I wanna thank you for that. So that, it’s not a specific ask, but that’s just kind of the way that I feel and I hope we can continue that for a really long time.

Nicole: Awesome. Thank you. I’ll talk to you both in a second, but we’ll say goodbye to everybody else. Thank you for joining us on this weekend, and see you next week.

Andy: Thanks. Bye.

Nicole: Bye!

Stephan: Thank you.