Sadly, Yahoo did as Yahoo does, and killed the blog sometime in the 2010s. Upon realizing what the web lost, I scraped everything off the Wayback Machine in 2020, before it eventually disappeared from there as well. The scraped HTML was badly mangled, with broken layouts, missing assets, bad links, and garbled post bodies from years of archive.org's own transformations. Reassembling it by hand looked like weeks or months of effort I'd never realistically get around to. So, there it sat in a zip file collecting dust for years, waiting for a spark.

What eventually got it unstuck was the epiphany that I could treat an LLM as the cleanup crew, instead of doing it by hand. Duh. At that point, weeks of effort turned into hours. So, I fed it the archive, had the model identify and isolate actual article content from the chrome, normalize the markup into markdown, fix internal links to a consistent structure, then spot-checked a sample against the original archive snapshots rather than reviewing all 1,000+ by hand. It's not pixel-perfect to 2008-era yuiblog, but it's readable, browsable, and most importantly, the content is intact. The only new content I added was the "Over the Years" section, to commemorate the effort from the community. Github repo is at https://github.com/derek/yuiblog.

While the library itself is long dead (RIP), these explorations into "F2E" as we called it can now live on. If you remember reading this stuff in the 2000s, or you're curious what frontend engineering writing looked like before "frontend engineering" was a cool job title, take a look.