独立开发者 Matt Birchler：从 99 个实验文件夹到年订阅制小工具
Matt Birchler’s Apps folder has 99 folders. Most were never meant to ship.
For years, Matt Birchler was known as a prolific tech blogger, YouTuber, and podcaster—not as a native iOS developer. He had a degree in television production, an early career in retail management and tech support, and one rough Swift weather app sitting at the bottom of his App Store Connect account as a permanent reminder of his first attempt.
多年来，Matt Birchler 以多产的技术博主、YouTuber 和播客主持人的身份为人所知，而非原生 iOS 开发者。他拥有电视制作学位，早期职业生涯涉及零售管理和技术支持，还有一个粗糙的 Swift 天气应用，一直躺在他的 App Store Connect 账户底部，作为他首次尝试的永久提醒。
Then he got tired of opening Sketch just to write a movie review.
后来，他厌倦了仅仅为了写一篇电影评论而打开 Sketch。
Watch on YouTube
在 YouTube 上观看
That inconvenience eventually became Quick Reviews, the first app Matt considers a serious release. It also started a run of small utilities built from the same question: if a piece of software keeps annoying you, why not fix it yourself?
这种不便最终催生了 Quick Reviews，这是 Matt 认为真正意义上的第一款应用。它也开启了一系列小型实用工具的创作，这些工具都源于同一个问题：如果软件总是让你烦恼，为什么不自己修复它呢？
A Sketch artboard had to become an app
一个 Sketch 画板必须变成一个应用
Before Quick Reviews was an iOS app, it was a manual design chore. Matt wanted to post visually appealing reviews of movies and games to his social accounts, so he built a Sketch document. Every time he wanted to post a review, he had to sit at his Mac, duplicate the artboard, swap the poster image, and retype the text.
在 Quick Reviews 成为 iOS 应用之前，它是一项手动设计任务。Matt 想在他的社交媒体账户上发布视觉上吸引人的电影和游戏评论，所以他创建了一个 Sketch 文档。每次他想发布评论时，他都必须坐在 Mac 前，复制画板，替换海报图片，并重新输入文字。
When people started asking what app he was using, he built a simple, one-page jQuery website to automate the process. It still required users to screenshot the final result, and it had no account system or review history. He wanted the tool on his phone, which meant he finally had to figure out native iOS development.
当人们开始询问他使用的是什么应用时，他构建了一个简单的单页 jQuery 网站来自动化这个过程。它仍然需要用户截图最终结果，并且没有账户系统或评论历史。他希望这个工具能在手机上使用，这意味着他最终必须掌握原生 iOS 开发。
Bad AI made him learn the code
糟糕的 AI 让他学会了代码
Matt knew CSS and HTML, but his previous attempt at building an iOS app felt like "laying out your HTML in tables." SwiftUI's declarative syntax felt familiar to someone who already knew the web. At the same time, early AI coding tools like Cursor were becoming available.
Matt 熟悉 CSS 和 HTML，但他之前构建 iOS 应用的尝试感觉像是“用表格来布局你的 HTML”。SwiftUI 的声明式语法对已经了解 Web 的人来说很熟悉。与此同时，像 Cursor 这样的早期 AI 编码工具也开始出现。
He spent about a month and a half working through the app, but the AI tools available at the time were frustratingly rudimentary.
他花了大约一个半月的时间来完成这个应用，但当时可用的 AI 工具简陋得令人沮丧。
“It was so hard. You had to tell it to do one small thing. It would screw it up. You'd have to do it again... The entire app would be in one Swift file, and you'd have to remind it, ‘you might want to do a second file.’” — Matt Birchler
“太难了。你必须告诉它做一件小事。它会搞砸。你得再试一次……整个应用会放在一个 Swift 文件中，你得提醒它，‘你可能想创建第二个文件。’”—— Matt Birchler
But that friction was exactly what he needed. Because the AI couldn't build the app perfectly on the first try, Matt was forced to read the output, understand the data structures, and learn how the views interacted. That structural knowledge paid off. Today, when he uses much better AI tools, he actually knows what to ask for and how to verify that the generated code is correct.
但正是这种摩擦是他所需要的。因为 AI 无法在第一次尝试时就完美构建应用，Matt 被迫阅读输出，理解数据结构，并学习视图如何交互。这种结构性知识得到了回报。今天，当他使用更好的 AI 工具时，他确实知道该要求什么，以及如何验证生成的代码是否正确。
The $10 subscription was downside protection
10 美元的订阅是下行保护
Quick Reviews is a free app, but it includes a $10 annual subscription from day one. That decision wasn't driven by aggressive growth targets; it was driven by API exposure.
Quick Reviews 是一款免费应用，但从第一天起就包含每年 10 美元的订阅。这一决定并非出于激进的增长目标，而是源于 API 的暴露。
The app pulls metadata, posters, and cast information from The Movie Database and uses an RSS feed to import a user's recent Letterboxd reviews. Relying on third-party services meant Matt couldn't assume every dependency would remain free forever.
该应用从 The Movie Database 获取元数据、海报和演员信息，并使用 RSS 源导入用户的近期 Letterboxd 评论。依赖第三方服务意味着 Matt 不能假设所有依赖项都会永远免费。
“I wanted to make sure all of these projects do not lose me money.” — Matt Birchler
“我想确保所有这些项目都不会让我亏钱。” — Matt Birchler
By putting the automated metadata and Letterboxd sync behind a paywall immediately, he gave the app a way to cover future service costs. Separately, the wider app catalog has become Matt's largest side-income stream, bringing in more revenue than either his blog or his YouTube channel by a significant margin.
通过立即将自动元数据和 Letterboxd 同步置于付费墙之后，他为应用提供了一种覆盖未来服务成本的方式。另外，更广泛的应用目录已成为 Matt 最大的副业收入来源，其收入远超他的博客或 YouTube 频道。
99 folders of experiments
99 个实验文件夹
Once Matt realized he could fix his own software annoyances, the development bug bit hard. He built Quick Subtitles and Chapterize to remove friction from his own podcast workflow, using them multiple times a week. When he set a goal to run 365 miles in a year and couldn't find a workout app that simply graphed his progress toward that specific target, his immediate response was, "I could make that." So he built Yearly Run Goals.
一旦 Matt 意识到他可以修复自己的软件烦恼，开发热情便一发不可收拾。他构建了 Quick Subtitles 和 Chapterize，以消除自己播客工作流程中的摩擦，每周使用多次。当他设定一年跑 365 英里的目标，却找不到一个能简单绘制他朝着该特定目标进展的锻炼应用时，他的直接反应是：“我可以做那个。”于是他构建了 Yearly Run Goals。
He names his apps clearly—using words like "Quick"—to signal that they are focused utilities, not massive platforms that require endless support. But he doesn't release everything he builds.
他给应用命名清晰——使用像“Quick”这样的词——以表明它们是专注的实用工具，而非需要无尽支持的大型平台。但他并非所有构建的东西都会发布。
In his Mac's home directory, Matt has a folder simply named "Apps." It contains 99 folders—not all of them projects he built himself, but many of them tiny tools, abandoned visualizers, and experiments that never went anywhere. He sees a friction point and wants to know whether he can fix it. Sometimes the answer is yes, and a new app ships. Sometimes the answer is no, and the experiment stays in the folder.
在 Mac 的主目录中，Matt 有一个名为“Apps”的文件夹。里面包含 99 个文件夹——并非所有都是他自己构建的项目，但其中许多是微小的工具、废弃的可视化器和从未成功的实验。他看到摩擦点，就想知道自己能否修复它。有时答案是肯定的，新应用就发布了。有时答案是否定的，实验就留在文件夹里。
Where to go next
下一步去哪里
In the full episode, Matt also discusses why actually selling software on the App Store changed his view of Apple's 30% commission, why YouTube sponsorships often create more administrative work than they are worth, and how co-hosting the Comfort Zone podcast gives him a collaborative creative outlet.
在完整剧集中，Matt 还讨论了为什么在 App Store 上实际销售软件改变了他对苹果 30% 佣金的看法，为什么 YouTube 赞助往往带来的行政工作多于其价值，以及共同主持 Comfort Zone 播客如何为他提供了协作性的创意出口。
Guest links
嘉宾链接
- Birchtree
- Quick Reviews on the App Store
- A Better Computer on YouTube
- Comfort Zone Podcast
- Birchtree
- App Store 上的 Quick Reviews
- YouTube 上的 A Better Computer
- Comfort Zone 播客
更进一步：量化金融体系
看懂新闻只是起点——沿量化金融路径，把它变成能交付的工程能力