Solo Design & Build · 2025–2026
新闻史数据库TJNU Press-History Database
把全系历年的数据新闻作业,变成一个对外开放、可检索的线上档案。整站的技术与设计由我一人完成。目前收录 25 件作品,时间跨度从 1902 年天津《大公报》的征婚广告,到 2025 年的“苏超”。Turning years of departmental data-journalism coursework into a public, searchable archive. Designed and coded entirely by me — 25 projects spanning 1902 to 2025.
- Role
- 独立设计与开发Solo design & build
- Year
- 2025–2026
- Scale
- 25 件作品 · 1902–202525 projects · 1902–2025
- Stack
- 纯静态 · Cloudflare PagesStatic · Cloudflare Pages
- URL
- journa-graph.pages.dev
背景Context
新闻史课程每年都会产出大量数据可视化作业——数据网页、数据长图、数据海报。它们做完、评完分,然后消失在各自的网盘里。这是一种奇怪的浪费:这些作品单看是作业,合起来看是一个系的方法论年鉴。我决定把它们捞出来,搭一个所有人都能访问的档案。Every year the press-history course produces a pile of data visualizations — scrolling pages, long-form graphics, posters. They get graded, then vanish into personal cloud drives. Individually they are coursework; together they are a department's yearbook of method. I decided to fish them out and build a public archive.
问题The problem
难的不是做一个页面,而是定一套能容纳异质作品的元数据。25 件作品来自不同年级、不同载体、研究的年代跨度也完全不同。我最终定了四个字段:作品形式(数据网页 / 数据长图 / 数据海报)、研究对象的年代跨度、班级、作者组。有了这几个字段,散落的作业才第一次能被并排比较——比如能看出哪几届偏好长图、哪些历史区间被反复研究。The hard part wasn't building a page — it was designing a metadata scheme wide enough for wildly different artefacts. I settled on four fields: format, the historical span each project covers, class year, author group. Only with those fields can 25 scattered assignments be compared side by side.
两个决定Two decisions
为什么是纯静态Why static
档案站最大的敌人是“维护”。用数据库后端意味着有人要一直养着它;我毕业之后它多半会死掉。所以整站是纯静态页面,部署在 Cloudflare Pages 上——没有服务器、没有依赖、托管免费,理论上可以无人照看地活很多年。收录新作品只需要改一个数据文件。The enemy of an archive is maintenance. A database backend means someone has to keep feeding it, and it would likely die after I graduate. So the whole site is static, on Cloudflare Pages — no server, no dependencies, free hosting. Adding a project means editing one data file.
为什么按年代而不是按届展示Why organize by era, not by class year
对系里的人来说按届看最自然,但对外来访客,按“研究的历史年代”浏览才有意思——1902 年的征婚广告、1930 年代的侵华广播、1940 年代的边区摄影,连起来是一部民间视角的中国新闻史。档案的排序方式,本身就是一次编辑。Insiders think in class years; visitors care about the eras studied. Browsing by historical span — 1902 matchmaking ads, 1930s wartime radio, 1940s border-region photography — reads as a folk history of the Chinese press. How an archive sorts itself is itself an act of editing.
收录的我自己的作品My own work in the archive
- 把酒话桑麻 · 人民日报乡村振兴报道数据库(第一署名,全国特等奖)Over Wine, We Talk of Crops — rural-revitalization database (lead; Grand Prize)
- 1792次的关注 · 透视《人民日报》的乡村振兴叙事(第一署名,全国二等奖)Harvest: 1,792 Reports — the narrative sequel (lead; 2nd Prize)
- 罪恶的聒噪 · 数说日本侵华广播史(1933–1945)The Wicked Noise — Japanese wartime broadcasting, 1933–1945
- 电波与商脉 · 近代天津广播与商业空间共生可视化(1922–1928)Airwaves & Commerce — early Tianjin radio and commercial space
- 做好“经师”成为“人师” · 方汉奇先生与他的弟子们(1948–2024)Teacher of Texts, Teacher of Life — Fang Hanqi and his students
site/assets/covers/journa-graph.jpg,我把首页的图案封面换成真图。另外补两个数字:上线以来的访问量;全站代码量或开发用时——“一个人、X 行代码、X 周”这种表述对面试很有说服力。