[nobody9999]

>There was the better part of the decade after the CDA passed that the tech community was still focused on protocols that worked this way.

Which protocols? I was designing and implementing networks throughout the 90s and aughts and I really don't know what you're talking about. Perhaps I wasn't in the right place at the right time?

Email mailing lists? IRC? Instant Messaging? NNTP? Those all would have been vulnerable to frivolous and malicious lawsuits without Section 230.

Honestly, I'm at a loss here. Please do enlighten me as to which protocols you're referring.

>I then want all of these messages stitched together to form a cohesive presentation on each person's computer, by software that represents their interests.

Sounds like you want personal ActivityPub platforms. I'm all for that. But nothing even approximating that existed in the 1990s. In fact, there's nothing like that now that a non-technical person can host for themselves.

[mindslight]

I agree regarding listserv and NNTP. It's questionable whether IRC and IM would be treated as "publishers" without sec 230.

Perhaps your coming up was a little earlier than mine? My perspective included things like gnutella and edonkey. There was a general feeling of building new application protocols to support new types of applications. Hard problems that needed to be figured out, for sure. But also background baseline values of people running software they choose on their own computers.

For protocols, there was also websites themselves. Someone with something to say would host their own. And some rough solutions for distributed discovery there like webrings.

Then web 2.0 came along and swept that all away in favor of the old centralized-mainframe dumb-terminal model (3270->browser, rs232->http, 80x24->html).

> there's nothing like that now that a non-technical person can host for themselves.

Yes. All of the high-cost productization/advertising work to make software palatable to normies doesn't get done, because investment money heads towards technical architectures that are more capable of exfiltrating value from end users. So any software still based around representing the interests of its users gets relegated to developers scratching their own itch.

[nobody9999]

>I agree regarding listserv and NNTP. It's questionable whether IRC and IM would be treated as "publishers" without sec 230.

Firstly, the concept of "publisher" is irrelevant to Section 230, then and now. IRC and IM (at least chat rooms) require servers to host the back and forth. As such, the issues were exactly the same as with email or usenet. Section 230 protects the hosts of any platform that allows third-party content. Full stop. This whole "publisher vs. platform" thing is a canard and a malicious attempt to muddy the waters. The law itself does not make such a distinction, nor does the case law surrounding it.

>Perhaps your coming up was a little earlier than mine? My perspective included things like gnutella and edonkey. There was a general feeling of building new application protocols to support new types of applications. Hard problems that needed to be figured out, for sure. But also background baseline values of people running software they choose on their own computers.

Sure, I was aware of gnutella and edonkey and other peer to peer file sharing tools. And yes, you're correct that there was much discussion of peer to peer applications for, well, almost everything. And even before that, there was KA9Q[0] which I ran on my PC/XT back in 1990. But none of that really went anywhere once NCSA-Mosaic[1] was released and the web (as you mention below) was born.

>For protocols, there was also websites themselves. Someone with something to say would host their own. And some rough solutions for distributed discovery there like webrings.

Right, and Section 230 protected (and still does!) the hosts of those sites too, while Mark Zuckerberg was in middle school.

>Yes. All of the high-cost productization/advertising work to make software palatable to normies doesn't get done, because investment money heads towards technical architectures that are more capable of exfiltrating value from end users. So any software still based around representing the interests of its users gets relegated to developers scratching their own itch.

On that I kind of disagree. It's not so much that the normies aren't interested. They certainly would be if the could click to download and then follow an install script to set it up, and it just works.

Going all the way back to Diaspora[2], to pixelfed, mastodon and it's offshoots, etc., I've set up a variety of open source platforms that tried to fulfill that dream of personal ownership/possession of one's content.

As a technical person, most of them were installable with significant complexities, but none were simple to install for the non-technical user.

And that is/was because the developers didn't make it that way, not any sort of malicious conspiracy. In fact, I recall some discussion around Diaspora, with the developers saying they preferred to focus on functionality rather than ease of installation.

The developers of Fediverse projects have continued in that vein.

tl;dr, I think we're mostly in agreement here, but you seem to be a little confused about how Section 230 works/worked. No matter. It's all good. I certainly appreciate the discussion and your perspective. Thanks!

[0] https://en.wikipedia.org/wiki/KA9Q

[1] https://en.wikipedia.org/wiki/NCSA_Mosaic

[2] https://en.wikipedia.org/wiki/Diaspora_(social_network)