Leo的恒河沙

3 Years, 4 Library Directories, 1 Million Words: Also Discussing My Obsidian Note Workflow - Minority#

#Omnivore

Highlights#

The key is the content, not the tools. ⤴️ ^3ce2c28f

3 Years, 4 Library Directories, 1 Million Words: Also Discussing My Obsidian Note Workflow

Matrix Homepage Recommendation

Matrix is the writing community of Minority, where we advocate sharing genuine product experiences and practical insights. We periodically select the highest quality articles from Matrix to showcase the most authentic experiences and viewpoints from users.

The article represents the author's personal views, and Minority only makes slight modifications to the title and layout.

748227, 102072, 68286, 66743. These are the character counts of four note libraries before I wrote this article. Including this article, I have written over one million words in Obsidian. From Youdao Note, OneNote, VNote to Obsidian, each reassessment and re-selection of tools has been a re-examination and reflection on my note workflow and my information processing pipeline.

This article is not limited to just the tool Obsidian, nor will it continue to regurgitate the almost artisan-like double-linked note method. The key is the content, not the tools: discussing how to efficiently handle explosive information input is essentially an optimization of thought. As my first article published simultaneously on Minority, I hope to share these insights with readers and also learn more from interactions to collectively improve our productivity experiences.

Information — Input and Output#

What is information?

Wikipedia states: "Information is the content that reflects (maps) events."

The essence of information lies in its flow. Static information has no meaning: events need to occur, content needs to be understood, and knowledge needs to be applied. Every information worker is essentially a mediator of input and output, and the foundation of their work is a caching system. If the carrier is the human brain, then the cache is memory; if the carrier is paper or electronic devices, then the cache is notes and documents. Information can only generate value when it flows continuously among mediators.

Therefore, an efficient information processing pipeline can be simply summarized in the following three aspects:

Input:

• Should effectively accommodate almost any type of information;
• Should temporarily store information in its most original form as much as possible;
• Should have appropriate storage locations;

Cache:

• Should have a systematic caching hierarchy: from disposable information (like daily news briefings), to short-term storage (like materials for a short-term project), to long-term or even permanent storage (like evidence of knowledge and skills);
• Should effectively manage and distill information;
• Should be easy to integrate with thinking;

Output:

• Should have convenient output modes;
• Should be able to output content in diverse information types;
• Should continuously and stably output high-quality content without resorting to copy-pasting without thought;

In the following sections, I will introduce my understanding of this system from the perspective of notes (focusing on Obsidian) and other aspects.

Notes#

Obsidian may be open longer than any non-system essential process on my computer. It is definitely the application that starts automatically when I boot up my computer every day and the last one I close before shutting down.

Note Tree

Beyond traditional notes, Obsidian is also:

My Default Notepad Application#

This title is deliberately chosen. I define it as: any application that allows for convenient text input and natural editing of longer text segments.

Notepad applications are quite useful; they form the basis of most editing. In addition to temporarily storing and editing general text, they also handle command line instructions and code, especially in a text-heavy environment like Linux.

My Browser Bookmark Storage#

After a long time syncing within the browser itself, I found that most browser folder/tree-style bookmark management and many extensions' bookmark pages are not the best way to display bookmarks. They often cannot show multi-layered categorization at once, have very low utilization of display space, and are not as suitable for diverse sharing as simple clipboard pasting. Therefore, I ultimately chose to export all bookmarks from the browser and convert them into an unordered hierarchical list in Obsidian.

Browser bookmarks. A series of personal blogs I recommend.

The Place Where I Write Most of My Output Documents and Presentations#

In addition to Markdown and the native export feature, its LaTeX editing capabilities combined with Pandoc have also provided considerable convenience for my learning. This will be mentioned later.

My Essential Template Library for Daily Work#

As a constantly open software, Obsidian stores a large number of text segments and pre-written code/scripts for my convenience in copying and pasting at any time.

My Many Work and Hobby Dashboards#

Using a bit of HTML and iframe, the note pages can display a lot of useful information. Regardless of where the information is stored, information workers are the key to facilitating the flow of information. They ensure that information can flow smoothly between event occurrence, content understanding, and knowledge application, thus creating meaning and value. Compared to the carefully assembled and advanced queries of Notion, a simple, brute-force dashboard can become a very practical tool.

Why Choose Obsidian?#

Many applications I used before Obsidian were merely transitional choices before I formed my information processing flow, which I won't elaborate on here. However, the last application I used before Obsidian, VNote, is worth mentioning.

Using VNote was the awakening period of my information management awareness. At that time, I knew nothing about note methodologies, couldn't write Markdown smoothly, and felt no thrill in recording information after trying almost all niche note software available at the time. In this context, I encountered VNote, which used Markdown and provided explanations for Markdown and notes.

To this day, these explanations are still preserved in VNote's documentation:

Markdown, as a simple markup language, differs from rich text in that it inherently carries a gap between editing and reading. There are generally three ways to address this gap:

1. As an extreme, some editors treat Markdown as plain text. Users may get lost in the messy black characters, making it difficult to trace the information context of notes.
2. Most Markdown editors use two side-by-side panels to edit and preview Markdown notes simultaneously. This simplifies things because users can preview a well-formatted layout while editing text. However, two panels can occupy the entire screen, and users must constantly move their gaze left and right, making it extremely easy to get distracted.
3. Another extreme is that some editors immediately convert Markdown elements after user input, making it feel like using shortcuts to edit rich text documents in Word. This may conflict with Markdown's design goals.

Since most editors choose the second method, many users associate Markdown with real-time previews. This could be a misunderstanding of Markdown. Positioned as a simple markup language, Markdown aims to help track text information conveniently during editing and provide beautiful formatting when converted to HTML for reading.

Four different editors, four different presentation styles.

Thus evolved my understanding of Markdown:

The design goal of Markdown's syntax is actually to allow unrendered plain text to be mentally formatted by readers who have no prior knowledge.

From this understanding, Typora clearly does not meet the standard: it is merely a "what you see is what you get" (WYSIWYG) editor that uses Markdown syntax as shortcuts, and in some sense, it is even worse than Word: the constant recognition of formats and elimination of formatted text during editing can cause significant layout shifts, increasing eye movement and operational pressure during editing. On the other hand, while proficient Markdown users can directly edit plain text to write formats, having appropriate formatting guidance during plain text editing is more conducive to effective feedback writing.

At that time, VNote was facing the threshold of a major leap to version 3.0, and many design logics I was accustomed to would change, and at this moment, the emergence of Obsidian perfectly met my needs.

For many text editors, especially code editors, they have chosen a relatively conservative approach, adding highlights to differentiate Markdown syntax. In contrast, VNote and Obsidian's "source mode" have made many further operations (such as in-place rendering of titles, bold/italic, loading images, etc.) that appropriately bring the editing experience that Markdown should have.

Therefore, Markdown source files should also be treated as partially formatted documents, considering the aesthetics of plain text browsing. For example:

• Use soft line breaks as much as possible;
  While the Markdown standard encourages hard line breaks or two spaces at the end of a line, Minority even has an article specifically discussing the issue of hard line breaks, soft line breaks align with the reading habits of most people and the logic of plain text processing pipelines, making them more suitable for files with less formatting;
• Use native syntax as much as possible;
  This particularly refers to Obsidian. For example, while Obsidian has its own syntax alternatives for image insertion, I still recommend using Markdown's native syntax as much as possible. This can improve the file's portability and aligns with Obsidian's selling point of "offline storage, never shutting down."

At the same time, do not forget the flexibility of Markdown format. Mathematical syntax, flowcharts, and even inline HTML usage should not be avoided due to the remarks of "Markdown purists" if used appropriately. Do not forget, the key is the content, not the tools.

How to Use Obsidian Correctly?#

Obsidian, along with Roam Research, Logseq, and other "EDC toys in the productivity world," is often associated with specific note methodologies, such as atomic notes, Zettelkasten, or double links. However, in my note library, files with links are few, and files under a thousand words are absolutely rare. Why am I completely uninterested in these methodologies?

Based on double-linked atomic notes and card notes, they will ultimately scatter into a disorganized mess of cards that are hard to find. Double links, in any form, carry no attributes. To understand the relationship between the information atoms connected by a link, both ends of the content need to reference each other, which is a waste of recording. At the same time, double links cannot represent any logic relationships that involve multiple transfers; a complete logical chain is broken into pieces that require step-by-step examination of logical relationships, and there may not even be any recorded arguments for logical relationships. Each time of reuse requires building understanding from scratch, which does not help in constructing a so-called "knowledge network." Furthermore, double links require constant maintenance, and the maintenance process can repeatedly fall into the decision of "should this content be placed in this card or that card," accumulating mental pressure that makes note maintenance increasingly burdensome, ultimately leading to abandonment.

Double-linked notes promote a concept of "thinking patterns close to the human brain," but this is entirely a false proposition. Experienced Obsidian user Eleanor Konik summarizes this aptly in her article titled "Obsidian is My Thinking Environment, Not My Second Brain" with five subheadings:

Obsidian doesn't do my thinking for me.
Our brains do more than record, reshuffle and regurgitate information.
If a computer isn't a second brain, nothing is.
None of my thinking is done by a tool.
My perfect auxiliary brain would not be a notetaking app.

Obsidian cannot replace my thinking.
Our brains do much more than record, reorganize, and output information.
If a computer isn't a second brain, then nothing will be.
My thinking is not done by a tool.
My perfect auxiliary brain would not be a note-taking software.

Therefore, I still adhere to a nearly "traditional" note-taking model: categorizing and hierarchically summarizing information, using double links very little, and recording a sufficiently independent information collection in large segments, like a Wiki. Over time, my notes have almost developed into a comprehensive book, and the advantage of books (especially textbooks) is that the information is clear and detailed, viewpoints are consistent and clear, and they are easy to understand and remember.

Diverse note forms. Please disregard the content.

At the same time, do not forget that the human brain is also an important component of the cache. Therefore, when recording notes, try to reassess the entire note by choosing categories and directly memorizing the titles, general content, and storage locations of the notes in your brain. Let the memory in the brain become the first index of the notes, followed by fuzzy searches, and finally folder-by-folder searches. Do not outsource this essential memory for understanding information to the double link system. Afterward, regularly maintain the note library, attempting to rewrite notes repeatedly, streamline concepts, distill information, enhance the quality of thought in notes while reinforcing memory. Inspiration cannot arise from wandering around in the double link system; thoughts must be internalized in memory to generate creativity.

Similarly, I reiterate, the key is the content, not the tools, so some flashy note-making methods, such as overly intricate visualizations, endlessly entangled formats with no thoughtful content, mind maps that only copy and paste thinking, and meticulously maintained but empty Notion pages, are fundamentally not productivity-enhancing; at most, they can be considered a peculiar hobby. Therefore, I hardly use any plugins that enhance "editing" itself and rarely fiddle with theme customization options.

Files#

However, the information received on electronic devices, besides text, also includes many other types, often existing in file form or convertible to files. The "input — cache — output" model applies here, which I call "Inbox - Queue - Outbox."

Inbox#

This is the first line of defense for information input: rough screening.
Regardless of the source of these files and how they need to be processed, all should first be collected into the Inbox folder to avoid information loss. Afterward, categorize them based on the type of information:

• If it can be processed immediately or does not require any processing, such as files that need to be forwarded or shared, then it can be sent directly from the Inbox to the required destination;
• If it needs to be temporarily stored, such as materials collected for a project, then create a related subcategory in the Queue to store it;
• If it needs to be stored long-term, such as books, documents, music, videos, etc., then send it from the Inbox to the corresponding storage location;

The benefits of having an Inbox folder are numerous, such as:

• When collecting information, one does not need to be troubled by questions like "Do I need to keep this information? Will losing it cause a loss of productivity later?" The importance and retention of a piece of information can be resolved gradually later;
• The size of the Inbox content intuitively reflects the volume of information received, effectively utilizing this metric can significantly control digital anxiety;
• Completing today's tasks today and clearing the Inbox can consistently achieve a positive psychological feedback of "work completed," and the resulting sense of accomplishment is beneficial for improving work efficiency; similarly, for emails, clearing the Inbox is also a very good habit.

Queue#

Content that needs short-term caching enters the Queue folder, categorized (usually by time-related content), and the associated matters are equivalent to being scheduled. Information in the Queue either enters notes or other long-term storage or is removed as matters are completed. As the name suggests, managing the Queue folder is best done following the First In First Out (FIFO) rule, setting appropriate deadlines for information. This method of information management can also feed back into schedule management, reducing procrastination.

Outbox#

The result of information distillation, internalization, and reorganization is output. The output of content must consider both the quality of thought and readability, aesthetics, to resonate with the recipient. This is when one should focus on achieving a beautifully crafted presentation.

Although Obsidian's built-in export is quite comprehensive and can restore the display effects in Obsidian to the greatest extent, for documents that require advanced formatting, I still choose to use Pandoc for export.

Conclusion#

This article does not attempt to introduce specific techniques on how I use Obsidian but rather focuses on my insights and personal experiences regarding note methodologies. The logic of the article may be somewhat heavy, and I hope that those who read to the end can resonate with it.
Thank you for reading.

Header image Photo by Pedro Araújo on Unsplash.

> Download Minority 2.0 Client, follow Minority WeChat Official Account, unlock a new reading experience 📰

> Practical and useful genuine software, Minority presents it for you 🚀

Lands ravaged, cities in ruins, so many lives sacrificed, and yet there was no other word for it but victory.