In March 1997, Kushal Dave stopped trying to keep his page about emerging Internet technology current. He left behind one excellent consolation prize: a note suggesting it might someday be useful to “Internet historians or something.”
Visit Kushal Dave’s net.tech page
Correct.
The page is a frozen survey of what looked important while the commercial Web was still deciding what it wanted to become. Dave had started it because he wanted something more useful than another link page. He collected new software, hardware, protocols, services, browser technologies, and network ideas, then organized links and commentary around them.
The resulting list is a wonderful mixture of winners, dead ends, ancestors, and technologies that survived in forms nobody in 1997 could fully predict.
There is Java. JavaScript. VRML. QuickTime VR. RealAudio. Shockwave. PDF. Internet telephony. Streaming video. Netscape plug-ins. PNG. Electronic cash. Conferencing systems. Whiteboards. Browser helpers. Offline browsing. The MBone. RTP. Cable modems. ISDN. Satellite Internet. Even schemes for mixing television broadcasts with related Web content.
The page also preserves period assumptions that make the technology easier to understand. A copied 1996 comparison chart puts a 28.8 kbps modem at the baseline, with ISDN at 128 kbps, a T1 at 1.54 Mbps, and ADSL still described as two or three years away. Cable modems are treated as an exciting near-future development.
The obsolete parts are now the useful parts
Dave eventually admits defeat near the top of the page. The field was growing too quickly, and professional publications were now paying people to track exactly what he had been trying to follow as a student project.
That abandoned state is what makes the page unusually valuable now.
A continually updated article would have replaced yesterday’s guesses with today’s facts. Dave’s page did not. It still shows technologies competing before the outcomes were obvious. RealAudio sits beside other streaming approaches. Browser plug-ins still look like a major architectural direction. Electronic cash is discussed as an emerging category rather than something with decades of later baggage attached to it.
Even the broken guestbook note belongs in the record.
For researchers, this is a primary-source snapshot of the mid-1990s Internet looking forward. For webmasters, it is a reminder that today’s supposedly inevitable stack will eventually look just as contingent.
CacheRat’s 1,967 Ancient Web Domains research list was built to surface pages like this. Search engines are good at returning current explanations of old technologies. They are less reliable at returning the old explanation written while the technology was still new.
Kushal Dave accidentally preserved the second kind.
And he called the audience correctly 29 years early.