That's a huge stretch. CERN might be a reason we have the web that looks like it looks today, but we would have something similar with or without CERN.

We might be stuck with compuserve and AOL.

We had the internet before the web, and we had gopher. So we had an open network, the client server model and links. What the web brought was a more flexible protocol and document format. That was important, but there's no way we'd have been stuck with compuserve or AOL.

Do you remember Microsoft's Blackbird:

https://en.m.wikipedia.org/wiki/Blackbird_(online_platform)

Do you remember the time when messenger services were open to third party clients? You could actually talk to all your friends using only one app!

Those days have ended.

I could not talk to all my friends on only one app, as the vast majority of people were not online. If anything, a far larger number of people are still on open services today than were back then; there are just even more people on locked down services.

But why you brought this up in a discussion about the invention of the web, I have no idea.

Do you remember the AIM/MSN wars, where AOL was literally building buffer overflows into their protocols to prevent MSN messenger from interoperating?

There never was a golden age of interoperability on the internet, at least not after HTTP. Maybe there was a lot of cooperation when it was purely academic in the late 70s and 80s. BBS operators often were cooperative. But commercial internet has always and forever been cutthroat.

We may well be stuck with the equivalent of those again, only we'll call them Google and Facebook.

We had gopher and SGML before and independently of CERN. The internet was just starting to take off; dial-up modem BBS emigres were getting shell accounts at early ISPs around that time (I know; I had such an account and worked at such an ISP).

Just because TBL happened to work at CERN and have a need for hypermedia, doesn't mean that we wouldn't have the WWW in approximately the same shape in approximately the same time frame without CERN or TBL.

In an alternative timeline we could have had

[1] http://www.art.net/studios/Hackers/Hopkins/Don/hyperlook/ind...

[2] https://donhopkins.medium.com/hyperlook-nee-hypernews-nee-go... (with screen shots)

[3] https://libarynth.org/hyperlook (larger screenshots)

based on

[4] https://en.wikipedia.org/wiki/NeWS

running atop of whatever, maybe even

[5] https://en.wikipedia.org/wiki/TRON_project

&

[6] http://tronweb.super-nova.co.jp/homepage.html

instead of the bloat we have now.