Notes on Writing "Second Brain Is Obsidian"

I Wrote a Book for the First Time

I wrote a book for the first time. When I checked the manuscript's character count, it came to about 130,000 characters. Even if it is a practical guide, there is a certain pride in having written something this long on a single topic. A lot happened while I was writing it, so of course I had to write a retrospective.

The Publishing Offer Arrives: October 10, 2023

I got to write and publish a book for the first time. An email with a publishing offer came from Golden Rabbit↗, a publisher that specializes in programming books. I hesitated. Writing a book would take an absurd amount of time...

After a few days of deliberation, I decided to write it. The main motivation was that there was still no proper Obsidian book in Korea. I also noticed that the publisher had just put out Woowa's development book, and I figured it couldn't be a shady outfit. This was an opportunity.

So I replied to the publishing email.

A Smooth Start

After meeting with the publisher in person, we agreed that I would write a sample manuscript¹. I told them the book would not focus only on explaining Obsidian's features, but would also include material on PARA, Zettelkasten, and so on. I thought that was what would make it competitive.

For the sample manuscript, we decided I would write just 30 pages covering installing Obsidian, Markdown, and explanations of PARA and Zettelkasten. I already had plenty of notes on PARA and Zettelkasten, so I was able to write those parts quickly. The Obsidian feature explanations also came together fast. This was when I realized that screenshots are a great tool for bulking up page count in no time. ²

A Practical Guide That Doesn't Feel Like One

This is an Obsidian book, but it is not an all-out practical guide in the vein of You Can Do It! Learn Obsidian in One Day! People say Obsidian is hard, but it's just a note-taking app after all, so how hard can it be? I also thought that only explaining features would be a bit stale, so I wanted to mix in "How should you write notes?" at least once. That's because existing books explaining Notion either spent much of their length on useless parts or focused on templates.

That said, I couldn't make the content purely theoretical and difficult either. Those people learn everything they need from the internet to begin with. They'd have no reason to buy the book. So I struck a reasonable compromise on the content.

Parts 1 and 2 explain mainly Obsidian's features, and Part 3 covers PARA and Zettelkasten. That makes Part 3 the section I was most interested in while writing. In Part 4 I returned to Obsidian and packed in difficult material like HTML, CSS, and Dataview. In other words, I structured the content to target everyone evenly, from beginners to advanced users.

I know that marketing teaches you to define your target clearly and avoid vaguely blending things together. But with Obsidian, I had no choice. Obsidian itself is niche to begin with, so the content can't be niche as well. It has to be comprehensive. You need to secure the easy content and the difficult content too, so that later competing books face a hurdle. The point is to demand not just quality of feature explanations but a hurdle of experience.

Writer's Block

At first I made rapid progress, but at some point I got stuck. It felt like I had reached the culmination point of the offensive in book writing. Obsidian content should have kept spinning in my head, but it wouldn't. I started to hate thinking.

This was actually a problem not just with the book but with other work too. Drowsiness started to set in and I didn't want to do anything. In the end, I had no choice but to stop the newsletter I was working on.

I went to psychiatry diligently and took Concerta, visited a family medicine clinic, switched psychiatrists, and was told there to go to a sleep clinic. I underwent a polysomnography and a multiple sleep latency test, and was finally diagnosed with narcolepsy³. That one sentence holds all kinds of attempts and expenses. Even after a narcolepsy diagnosis, you still need an optimization process to find the right medication and dosage.

When you're stuck in writing, you need ideas and seed sentences that can develop the piece. You don't get these by agonizing over them; a natural, enjoyable mood is what matters. Just as you go out into a field and gather natural stones one by one to build a stone wall, you have to keep collecting notes.

But for the section explaining Obsidian's features, I hadn't gathered any such natural stones. Since I had no notes related to Obsidian's features, I had no choice but to write everything from scratch. There were even a few features I barely use myself, which I had to deliberately use and explain just to make the content complete. Some of the features I tried out under compulsion turned out to be more worth sticking with than I had thought. Writing a book teaches you more.

When my motivation drops, what I look back on is why I started writing this book. I wrote it because I thought, "Someone truly needs this book." I have no choice but to motivate myself this way. For the record, money was not part of my motivation. Not because I'm uninterested in money, but because, as I'll explain later, books don't make money.

Terminology

The goal of Part 3, which explains PARA and Zettelkasten, is to give readers a sense that they are learning something. That is why it even covers AI and goes as far as comparing note similarity. You put the content of your notes into an embedding model and analyze the resulting vectors with cosine similarity. People with an engineering background will mostly get it from that explanation. But the moment you try to explain it in a practical guide, it gets awkward.

Setting the embedding model aside, I needed to explain what a vector is first, because vectors were dropped from the high school curriculum. If it were in the curriculum, I could slip past it without explaining, but since it isn't, I have to explain. I couldn't describe a vector as something with magnitude and direction. The embedding model used in the book's examples outputs a 3192-dimensional vector, and you can't explain a 3192-dimensional vector in terms of magnitude and force. So the way I explained vectors was:

A vector is... a list of numbers!

"Then what's the difference from a sequence?"
"I don't know, Sian doesn't do that stuff. Let's just say a vector is a list of numbers."

Terminology must not be so difficult that you cling to the definition, and it must not be wrong. During revision with the editor, there was a line saying "HTML and CSS are programming knowledge," which I changed to development knowledge. Personally, I think that for people with no development knowledge or interest, it's actually better to explain that HTML and CSS are programming languages. The moment you say markup language or stylesheet language, it only gets more confusing. But I changed it out of fear that someone would leave a one-star review wondering, "HTML and CSS are programming languages?"

And the Curse of Knowledge

Explaining things simply depends more on the editor's ability than on mine. I'm under the curse of knowledge, so I don't really know which terms and explanations are hard. The editor has to catch that, and the editor assigned to me did a wonderful job.

They also gave active feedback on the structure. I had organized the book by topic, so everything related to viewing notes was gathered in one place. That is, search, graph view, Dataview, and so on were going to be introduced in a single chapter. But the editor suggested moving the difficult parts like HTML, CSS, and Dataview to the very end as Part 4, and I think that judgment was right.

The publisher's help didn't stop there. Once the manuscript goes through typesetting, it becomes a completely different book. I really wrote this? It looks like a book the publisher wrote, not me. Hmm... the publisher carried this one. Still, I'm told it's rare for a manuscript to come in properly, so I'll give myself a little pat on the back (?)

What Doesn't End After the Manuscript

Publishing a book isn't something you can wrap up with "It was a good experience~" as a matter of personal self-improvement. You run toward the goal of generating revenue together with people who had no connection to you before. If I stubbornly insisted on the content I wanted to talk about without caring about marketability, it would remain a good experience for me, but the publisher would be left with a deficit.

That is why sales work is a must after publication. It's a hard part for a developer like me. You have to put on a thick skin and announce that you've published a book, and be prepared for some people to turn their backs on you, saying Book peddler. Smells like money. Lost your original spirit. And being prepared for sales work doesn't settle it. Even handing out flyers like this, if you put on a thick skin and push in, you just provoke aversion. You need a way to push in appropriately.

The reason people who sell books and courses get criticized is that they come across as trying too hard to sell. Desire can be the motive and purpose behind making a product, but the moment it shows as it is, it becomes ugly. It's like how you shouldn't reveal too much that you want to meet someone, or your desire to get married, on a blind date. That isn't honesty, it's being blatant, and the person watching feels like they have become a means to an end.

To offer a bit of an excuse, it's hard for an author to make money from publishing. It takes far too much effort. If you want to earn more for the effort, you'd rather make an ebook and post it on Kmong, or self-publish. Or make card news on Instagram and build a funnel by saying "Follow and comment 'Obsidian' to get a free ebook by DM!"

For a book published in collaboration with a publisher, the author's goal is really not so much making money as branding: "Please notice me!" Royalties for practical books are said to be 8 to 12%, so what royalty rate do you think was set for Sian, who had never written a book before? And how well do practical books sell? In Korea, books don't sell well. In the end, the author's purpose in writing is to build content in a well-structured way, to avoid running off the rails by listening to feedback, and to reach the public. So if you're going to criticize someone for promoting their own book, it's more accurate to call them an attention seeker than a money-grubber.

Trauma About Length

I think writing this book was harder than writing my master's thesis in English at KAIST. That's because a thesis has a logical structure and a more or less established framework. A book has neither. Perhaps the medium isn't what matters; comparing the absolute evaluation of graduation with the relative evaluation of the market, they seem to be incomparable in difficulty.

During my master's thesis, I had a trauma about length. Oddly, I felt I had written everything I had to, yet it looked too short. In fact, I received feedback that my master's thesis was short. I felt I had said everything I wanted to say... I worried about length during the writing this time too, but I was able to resolve it with this book. So I am someone who can write long pieces after all.

On Writing a Book

If you get the chance, I think writing a book is a good thing to do. Especially while holding a job. Setting aside the extra income or branding, you get to learn about a new world and see yourself in a new situation. Writing over 100,000 characters on a single topic over a long period is an opportunity you can't easily get. Through this writing I had many experiences: writing a newsletter, a team merger, narcolepsy treatment, ObsiCon, and all sorts of other events. These eight months were the busiest eight months of my life, and they let me grow in many ways.

I've been leveling up all along, so I plan to get active once again.

Footnotes

1. A sample manuscript means writing about 30 pages in advance. It works as practice, and it is also the process by which the publisher checks, before signing a contract, "Can we trust this person enough to work with them?" You write about 30 pages. ↩

2. I only found out later, but I should have used CleanShot X on my Mac. ↩

3. Think of it as a condition that makes you sleep a lot. That doesn't mean you suddenly collapse and fall asleep like in the movies. If you suspect narcolepsy, I recommend Coslip Sleep Clinic in Seolleung. It seems to be the best place in Korea for sleep issues. ↩