Today, issue 100 of Xiazi · Yesterday’s World went live.
I stared at “100” for a while. It is only a round number, yet it suddenly gave the past three months a shape: 100 issues, each with nine Chinese posters and nine English posters, for a total of 1,800 officially published images.
Printed out, they could probably fill a very long corridor.
But what I have accumulated is not really 1,800 images. It is 100 moments at the beginning of a day when I stopped and asked myself:
What happened in the world yesterday? Was there something quietly changing our future?
100 published issues · 1,800 public posters · 18 Chinese and English images a day.
June 15 – September 22, 2026. Nine Chinese posters plus nine English posters, across 100 issues.
I do not need every piece of news. I want to stay curious about the world.
Information arrives like a flood. In theory, we can find out what is happening anywhere, at any time. In practice, after half an hour of scrolling, our minds are often left with little more than noise.
I did not make Yesterday’s World to reproduce all the news.
I prefer to think of it as a radar for what matters: each day, pick a handful of changes worth a closer look from the noise, then work through them in both Chinese and English.
That process forces me to slow down. Why does an event matter? Where does a number come from? Is this something that has happened, or a company’s own claim? Does it affect today’s mood, or the opportunities of the next three years?
Ask those questions every day, and news becomes more than news. It becomes part of how you form judgments.
Three kinds of change I pay particular attention to
The first is AI.
I am less interested in another model name than in what has become possible: can it finish longer tasks? Are costs still falling? Can an ordinary person now make something that used to require a team?
The second is one-person companies, or OPCs, and applications that make money.
They reveal changes in products, business models, and ways of organizing work. How does a feature become something people pay for? How does a small company find growth? Can one person use software and content to do work that once took ten? These questions connect directly to the products I am building.
The third is cross-border commerce.
Going global is never just “selling overseas.” Behind it are payments, taxes, logistics, localization, customer acquisition, and trust. Understanding how a single order crosses a border often brings you closer to a real opportunity than a grand story about global expansion.
These three areas may look different, but they point to the same question: when a new capability appears, who can turn it into real products, transactions, and lasting value?
Today’s hundredth issue follows the same format: nine Chinese posters and nine English posters, with one overview and eight themes. AI, one-person companies, practical AI businesses, and cross-border commerce sit on the same world map as global politics, life sciences, Earth systems, and space science.
I like putting them beside one another. Opportunities rarely appear simply because you watch one sector closely. They often live between two things that did not seem connected.
As I kept making it, I realized that being correct was not enough
Xiazi began with just one visual style.
Its advantages were obvious: consistent, clear, and efficient to produce. After a while, the limitation was obvious too. It felt like wearing the same white shirt every day. Nothing wrong with it, but not much to look forward to.
News carries emotion. A technological breakthrough can be exciting; a business failure, sobering; a discovery in space, romantic; climate change, heavy. Put everything into the same template, and the information may be organized while the feeling is flattened.
That led to Xiazi Style Atlas. Each day, I choose a visual language again, letting composition, color, typography, and materials respond to that day’s content.
It gave Yesterday’s World another kind of emotional value. Unexpectedly, it also became a private art museum that opens every day.
I began comparing things more carefully. How would the Swiss International Style organize the same business story? Why does Mid-Century Modern feel both warm and rational? When do the force of woodcut, the tension of collage, or the open space of ink painting best suit the subject?
Style Atlas is not about changing a “skin” every day. It is an ongoing exercise in seeing, judgment, and taste.
At the time of writing, 80 issues in the public archive include a style field, with 55 distinct labels before any similar names are consolidated. The earlier 20 issues do not have that field, and I do not want to fill it in afterward just to produce a prettier number. Issue 100 uses Mid-Century Modern. Choosing again each day does not mean 100 styles with no repetition. It means making a fresh visual judgment every day.
I enjoy this. While making a news product, I am also giving myself an education in aesthetics.
Behind 100 issues is more than pressing Publish each day
People see the posters. Behind them is a production process that has gradually become more complete:
Find topics → Check sources → Judge their significance → Write in Chinese and English → Design posters → Upload assets → Create the day’s release → Add it to the archive → Check that all 18 images open.
It did not look like this on day one.
I have changed the topic structure, adjusted bilingual fields, repaired image routes, introduced immutable asset versions, and filled gaps in the archive. I have also spent plenty of time on problems such as “today’s page works, but the historical page is broken.”
As of the date of this essay, the remote main branch contains 2,894 Git commits. Many are daily content and release records, so I would not describe that as “2,894 product versions.” Looking at file changes, 215 commits actually touch program source files.
Counting the same remote code snapshot, and excluding images, archive JSON, dependencies, build output, and temporary files, gives 237 source files and 30,231 physical lines of code.
Numbers cannot tell you how good the code is. They do remind me that even a small product that seems to “publish a few images every day” develops its own engineering complexity through real use.
Today I checked the 18 public image URLs for issue 100 again, one by one. All opened successfully. The release status was healthy, and the page’s release version matched its asset version.
That may not sound romantic. But a lasting body of work needs this unromantic care.
What the hundredth issue confirmed for me
At first, I simply wanted to make a daily world summary that I would want to read myself.
Gradually, it became an information filter, a radar for opportunities, bilingual practice, a design laboratory, and a visual archive I could open every day.
I am always my own first user. When I open Xiazi in the morning, I am not trying to tick off a task. I want to see what deserves further attention. Sometimes one item introduces me to a product, a company, a business model, or even a direction I want to explore next.
That is curiosity’s most practical reward. It does not promise an immediate answer, but it lets you see one more possibility than you saw yesterday.
One hundred issues is not an earth-shattering achievement. It simply shows that something which begins small, perhaps even a little impulsively, can become a body of work if you keep doing it.
Tomorrow will be issue 101.
I will open the world and look again.
Nine Chinese posters and nine English posters every day. You are welcome to join me at xiazishuo.com.
Data notes · September 22, 2026
The figures in this essay are a snapshot from September 22, 2026. The public archive was checked at 100 issues; issue 100 had nine topics and 18 accessible Chinese and English images. Its release status was healthy, with matching releaseId and assetVersion values.
Code statistics refer to remote main commit 8c873d36: 2,894 commits in total, including 215 that touched program source files; 237 source files containing 30,231 physical lines of code. Images, archive JSON, dependencies, build output, and temporary files were excluded. Commit counts include content and publication records and do not represent distinct product versions. The 1,800 posters are final public images, excluding discarded drafts and replacements.
Public issues and style records: Xiazi · Yesterday’s World. Code figures are the author’s project-repository snapshot described above.