I'd had my first tastes in the late 1980s, where some uniformity of tools, platforms, and cultures (tin, Unix, and largely uni-based participants) tended to cohere. I've cited on HN Brian Reid's Usenet activity surveys (from John Quarterman's The Matrix,[1] an early exploration of what we'd now consider social networks) several times on HN[2]. As of 1988, there were 381 newsgrous, 1,933 articles/day, 4.4 MiB/day of traffic, 7,800 hosts, and 141,000 readers. That's ... tiny by contemporary standards.
Cultural norms broke down rapidly as Usenet spread first to corporate networks (Lotus Notes email formats remain a massive annoyance etched in my brain), and then the general public. By the mid-1990s, Usenet though far more active and reaching far more people was a pale shadow of its former self in terms of culture and relevance.
The fundamental technical presentation along with the original posting culture was fairly effective. That unfortunately didn't scale. One of my concerns as I look at decentralised networks with multiple clients and server implementations (e.g., the Fediverse) is that a diversity of tools will inevitably result in a broken set of standards and practices. In the case of the Fediverse, the baseline is fairly low, though there are implementations and/or instances which offer some fairly narrowly-supported capabilities, notably raw HTML or Markdown formatting, and equations support (as with ColinWright's Mathstodon).
I'm not arguing against your experience, FWIW. I am arguing that your experience is probably highly time-dependent on when you participated in Usenet, and the era from 1979--1992 is markedly different from that of 1993 onward, though cracks were already starting to show.
I've also noted that HN has survived longer than Usenet's golden age, and with remarkably stable quality. It's not all it could be, but it's not the worst of what's online either, and by a long shot. Despite some frustrations, I still find it useful.
I'd had my first tastes in the late 1980s, where some uniformity of tools, platforms, and cultures (tin, Unix, and largely uni-based participants) tended to cohere. I've cited on HN Brian Reid's Usenet activity surveys (from John Quarterman's The Matrix,[1] an early exploration of what we'd now consider social networks) several times on HN[2]. As of 1988, there were 381 newsgrous, 1,933 articles/day, 4.4 MiB/day of traffic, 7,800 hosts, and 141,000 readers. That's ... tiny by contemporary standards.
Cultural norms broke down rapidly as Usenet spread first to corporate networks (Lotus Notes email formats remain a massive annoyance etched in my brain), and then the general public. By the mid-1990s, Usenet though far more active and reaching far more people was a pale shadow of its former self in terms of culture and relevance.
The fundamental technical presentation along with the original posting culture was fairly effective. That unfortunately didn't scale. One of my concerns as I look at decentralised networks with multiple clients and server implementations (e.g., the Fediverse) is that a diversity of tools will inevitably result in a broken set of standards and practices. In the case of the Fediverse, the baseline is fairly low, though there are implementations and/or instances which offer some fairly narrowly-supported capabilities, notably raw HTML or Markdown formatting, and equations support (as with ColinWright's Mathstodon).
I'm not arguing against your experience, FWIW. I am arguing that your experience is probably highly time-dependent on when you participated in Usenet, and the era from 1979--1992 is markedly different from that of 1993 onward, though cracks were already starting to show.
I've also noted that HN has survived longer than Usenet's golden age, and with remarkably stable quality. It's not all it could be, but it's not the worst of what's online either, and by a long shot. Despite some frustrations, I still find it useful.
________________________________
Notes:
1. <https://archive.org/details/matrixcomputerne0000quar/page/24...>
2. Search shows most, excepting where I've brainfarted "Eric" for "Brian": <https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...>