A lot of early personal websites disappeared because nobody thought the family homepage was the part of Internet history that needed saving. Greg Kennedy did the useful thing and kept one.
Visit Greg Kennedy’s preserved family webpage
Kennedy’s current project index labels it plainly: “Historic family webpage, circa 1997.” That is almost underselling it. The surviving material is a small time capsule of the period when putting your family, hobbies and hand-drawn interests on the Web was a perfectly normal reason to learn HTML.
One of the pages belongs to Erin, who introduces herself with: “I am twelve years old, and I LOVE horses!” Her horse homepage says visitors have been arriving since May 23, 1997. From there it moves through horse evolution, grooming tools, English and Western tack, riding styles and printable coloring material.
Nothing about it was designed to become archival evidence. That is why it works so well as archival evidence.
The ordinary Web is the part that vanishes first
Histories of the Internet naturally preserve major companies, browser wars, famous software and important technical documents. Personal pages are harder. They were hosted on small accounts, moved between ISPs, abandoned when families changed computers and often died when a hosting company shut down.
Kennedy’s preserved pages show the other side of the Web: children explaining hobbies, household projects, simple illustrations, informal writing and pages built because somebody wanted to share something rather than establish a brand.
The current Greg Kennedy site is still very much in that spirit. His project index jumps between retrocomputing, electronics, old software, conversion utilities, games and experiments. The 1997 family material sits among those newer projects instead of being redesigned into a modern museum exhibit.
That choice matters. A reconstruction can show what an old site contained. The old files themselves show how people actually wrote, linked, organized and decorated pages at the time.
For web-history researchers, the specific details are the useful part: the 1997 date, informal headings, visitor-counter language, hobby education, personal email links and the complete absence of a content strategy meeting.
This is the kind of surviving personal material that makes CacheRat’s 1,967 Ancient Web Domains research list useful as more than a stack of old addresses. The Web was built out of pages like these as much as it was built out of famous ones.
The family homepage did not need to change the world. It only needed somebody to remember not to delete it.