Ephemeral Internet and The Unreadable Codex

I was not blogging for some time. In the meantime, both Gfycat and Koji have been acquired, and their services shut down.

These two occurrences reminded me the ephemerality of life on the internet again. A struggle against it does require being proactive. As I have learned with a harsh lesson, Internet Archive’s fights against this decay is a cornerstone of modern internet. Even though you can rely on it to save the old web to an amber, internet technologies are a mobile target. You cannot trust it to revive the mosquito within it. As we require ruffle to replay flash games, we need tools to deal with the intricacies of the modern web.

Gfycat was a GIF hosting site. In 2014, I decided to upload my screencasts and embed the optimized GIFs (actually MP4s, HTML5 video element was in its fancy) on my blog. My prior self thought there would be tons of screencasts of ongoing game projects; let me not encumber GitHub with their artifacts. Though these projects had stalled, more so my blogging. At last, in 2020, Snapchat owner Snap acquired the Gfycat. In the end, on September 1st of 2023, the service has been discontinued. The problem was I missed the news, and the screencasts were now completely gone.

My search to restore blog content has led me to this Reddit post about a Redirector rule. While high-res versions of screencasts had now turned into dust, with its help, I could save low-resolution copies of GIFs. I have successfully updated all Gfycat embeds in posts from 2014 now.

Koji was a low-code content creation tool. I made two games to be remixed by influencers for their campaign pages for Koji in 2020 during the start of lockdowns. Linktree has acquired Koji and ceased its services by January 31st, 2024. After the news, I tried to archive my pages via the Internet Archive, but a web-based game was too complex for it to replay. Even though the page has been archived successfully with its outgoing links, trying to view it resulted in a blank page.

Sunsetting of Koji made me discover the webrecorder. The project describes itself as a suite of open source tools and packages to capture interactive websites and replay them at a later time as accurately as possible. Webrecorder encompasses archiveweb, a Chrome extension to record WARC files, and replayweb to replay them. Rescued archive is at https://c6p.netlify.app/#projects.