_ __ __ _
(_)__ / /_________ / /________ _______(_)
/ / _ \/ __/ ___/ _ \/ __/ ___/ / / / ___/ /
/ / __/ /_(__ ) __/ /_/ / / /_/ / / / /
__/ /\___/\__/____/\___/\__/_/ \__,_/_/ /_/
/___/
(Originally posted to Cohost on Thu, Nov 2, 2023, 12:18 PM)
Yo, everyone! I hope you've all had a sp00ktacular Halloween.
This was a blog post I was writing last week, but I delayed it after being uncertain whether or not it was a fit for this blog. Here it goes.
There exist tons of files that are still out there somewhere, but haven't seen the light of day. Old software, old videos, old music, old images... everything. Some old websites that weren't included in the Wayback Machine (or excluded for whatever reason), and other digital forms of human creativity that were present possibly since the dawn of read-write storage. Or experiences that were tied to online services, but the original source files are nowhere to be found.
It's easy to forget about them in a world where software, videos, music and art is so widespread, easily shared and easily downloaded. A world where, with mainstream search engines like Google, "newer" results often feel like they're pushed higher than older ones.
The fortunate, popular ones at least had an active effort to archive them. For Macromedia Flash, this resulted in the Ruffle emulator and repositories like Flashpoint and the Internet Archive. For Geocities websites, this resulted in several different backup mirrors, including one on the ProtoWeb.
But in my net diving, I've come across some that weren't so fortunate to be saved:
(Slightly NSFW warning) There was this Japanese webcomic from 2002 by the name of "Nyan Nyan Police" (or "ニャンニャン☆ポリス"), as pictured above. The only signs I have of its existence are two Insert Credit mentions, an article from a French site, an article from a Japanese site, wallpapers and scans that were saved onto image boorus, and low quality thumbnails of the comics from a Japanese GameSpot article I've yet to find again. As far as I'm aware, the comic itself is completely lost.
Also discovered on Insert Credit is a 2002 Korean MMORPG called Espiritu Online (or "에스피리드 온라인"). There exist no publicly available gameplay footage other than screenshots on several different websites, and a song from the game's soundtrack that was uploaded on YouTube in November of 2020. There's also an official website for the game that's not available anymore, with some major snapshots from November 3rd, 2002 and June 16th, 2004. June 7th, 2005 is the earliest we see the single dead link titled "server_term.jpg", and the domain expires as early as November 7th, 2005.
".hack//チャット" (".hack//Chat", better known by English fans as ".hack//Kyat") was an online game based on the .hack series. Originally released in Japan with its web page dating to June 5th, 2004 at the earliest, it was a 2D virtual world with a .hack theme, using "MetaGateX" as its backend. It was somewhat reminiscent of older platforms like Habbo Hotel and some new ones like Pony Town. There was no objective other than to hang out and chat in this world. It had a fanbase of English players from as early as 2004 to as late as 2008. All players were able to represent themselves with custom sprite-based avatars. Screenshots and gameplay are available. The official server was closed down in April 30th, 2008 according to the wiki, but a fan reboot (kyat.us) made its debut on July 1st, 2008. For a long time, kyat.us was in a state of limbo, where the game's homepage displayed the chat server as being offline but the map being online, permanently stuck in a Christmas "event", and due to its dependency on ActiveX the only way to play the game was to use Internet Explorer, which I didn't get too far with. Sometime in 2022, the homepage for kyat.us was shut down for good. I always wished there was an open-source, self-hosted "game" that was just like this...
Literally any personal website for artwork. My favorite of which being Deadpanda, the home of an artist (Hayden Scott-Baron, known by his pseudonym "Dock") who I discovered thanks to his temporary contributions to Minecraft in the form of characters like Rana and a planned logo and player model overhaul. Since the website isn't a complete archive, some of his creations are publicly lost, or could only be found on printed media that he published or contributed to. This is unfortunately common when visiting lesser-known websites from the 90s-early 2000s through the Wayback Machine. After some downtime, the Deadpanda domain was brought back up earlier this decade (and redirects to docky.net), but seems to serve a more formal purpose?
This doesn't count as an internet story, but still relates to seemingly-lost digital data. Cotton 2 was a video game developed by a Japanese company named SUCCESS about an impulsive, bratty witch girl named Cotton (and her fairy companion Silk) on her pursuit for magical candy. A sort of physical collector's item called Cotton 2 Tenkomori was available for Windows 95/98 users and was a CD that gave them a desktop theme, desktop accessories, minigames, concept art, and assets from the video game (such as sprites and voice clips), all the materials they'd need to make their own fangame. Fortunately, unlike the previous stories, this one had a happy ending. It was announced that someone had dumped it to the Internet Archive, and personally, that moment was the most excited I've ever been in such a long time.
In case you want to get into digital archaeology or look for more things like this, I'll share a few tips and tricks that I apply in my travels:
Marginalia Search is very useful for "stumbling" across smaller and older websites that are still up, but pushed down by mainstream search engines like Google. Searching with Marginalia is similar to early Google and other old-school search engines, and tips on how to use it are found at the bottom of the page. Compare Google and Marginalia results for something like "how to draw anime" or "pork chop recipe" and you may be surprised.
Internet Archive's Wayback Machine has its own search engine for old websites (the same bar used for entering URLs), but it only seems to search the URLs, headers and descriptions instead of what's in the page itself. This is useful for typing in an umbrella term like "art" or "wallpapers" and seeing what you can uncover, but has a limit of showing only 50 sites.
The Old Net is catered to older computers, but similarly has the option to search archives. It's a lot easier to search a variety of "personal" sites this way (see here), but sometimes the sites they lead to may not work. A perfect time to paste them into the Wayback Machine and scavenge what you can.
GeoCities archives. Pretty self-explanatory if you know just how big of a role GeoCities played in the early days of the internet.
Don't just lurk the surface. Venture deeper into the rabbit hole. Websites often link to other websites, which is how I've made some of these discoveries.