Name: The matrix
ISBN: 9781555580339

What Hacker News says

dredmorbiusJul 2025

For an excellent compendium of what early computer discussion networks (or "conferencing systems" as they were often called at the time) were like, as well as a bunch of technical background on the actual data networks and protocols of the time, I'd highly recommend John S. Quarterman's The Matrix, first published in 1989. That is, it predates the World Wide Web, and was only about six years after "The Internet" largely (weasel-word conspicuously noted) replaced "ARPANET" as the designation for the widely-used (amongst university, government, military, and some tech-company) public networking protocol and system based on TCP/IP. Late-breaking news about the effect of computer networks on notable political protests in China are included in the forward.

At the Internet Archive (and apparently from Kahle's own personal collection): <https://archive.org/details/matrixcomputernet00quar/page/n3/...>

I'd begun using Usenet at about the same time as the book was published, and can't personally attest to the information jibal's giving, but will vouch that their account is far more accurate than that to which they're responding.

thread

dredmorbiusJun 2022

For a vastly better take on the origin story of the Internet, based on the work of someone who was there at the time, documenting it at the time, I'd recommend without reservation John Quarterman's The Matrix.

First published in 1990, just as the WWW was emerging (and failing to make the book), it was revised in 1997 (the link I'm giving here). The earlier book does discuss extensive sets of networks of which ARPANET, formally decomisssioned in 1990 after spawning the then-nascent Internet, was only one instance.

The introduction includes earily prescient discussion of the role early computer networks played in the Tiananmen Square protests and massacre of 1989.

https://www.worldcat.org/title/matrix-computer-networks-and-...

thread

dredmorbiusOct 2020

For those interested in early online systems, I strongly recommend John S. Quarterman's The Matrix: Computer Networks and Conferencing Systems Worldwide:

https://archive.org/details/matrixcomputerne0000quar/mode/2u...

I lifted a copy off a friend who was thinning his book herd a decade or so back. The first edition, published in 1990, predates the commercial Internet, World Wide Web, Linux, and all of what's transpired since, though the introduction pressages much by noting the role of existing networks in the the then-fresh events of April 15 – June 4, 1989, known as the six-four incident, and culminating in the Tiananmen Square Massacre.

The book has a 15 page section describing Usenet, including user and traffic statistics from 1979--1988, mostly compiled by Gene Spafford and Brian Reid. Usenet is only one of numerous conferencing systems covered, others being BITNET, UUCP, FidoNET, VNET, and EASYnet

For April, 1988: 381 newsgroups, 57,979 messages, 130.949 MB, 4,152--7,810 hosts (depending on who counted and how), and an estimated 141,000 readers (people) from a total online population of 880,000.

Larry Wall, working at JPL, is named ... as the author of rn.

Much of the book, though, simply discusses the many many individual computer networks. The notion of a single global end-to-end interconnected inter-network was still very much a novelty.

In one section, the fact that the principle public interconnect between North America and Europe was a single 9600 Baud link is mentioned....

thread

dredmorbiusAug 2019

Interesting question.

"Replacement in what sense?" probably deserves exploration.

The underlying fundamental concepts of the Internet are 1) packet-switched (as opposed to through-circuit) communications, and inter-connected networks, via BGP.

There's a lot that's layered on top of this which is seen as fundamental, but is not entirely so, most especially end-to-end connectivity and universal point-to-point access. The Internet formed under tremendously different conditions than exist today, with only a handful of nodes through 1980, and even as late as the late 1980s only a few thousands.

Roughly each order-of-magnitude increases since then seem to have come with its own set of additional headaches and concerns, mostly regarding management and mitigation of abuse. The idea of guarded borders has long been seen as anathema. I see it as all but inevitable, and the question is whether that's done well or poorly.

There are a number of earlier networking ideas which might resurface or be adapted for new use, and a survey of history might be useful here. (John S. Quarterman's The Matrix, 1990, is a fascinating time-capsule exploration of these just at the cusp of the modern Internet -- the World Wide Web does not even make mention, though "The Web, a national Canadian nonprofit conferencing system formed in 1987, does.

I'd also look at uses of the Internet and user needs.

Fundamentally, digital networks serve as communications and control media. Whether or not these need to (or can be) segregated is an open question, but splitting off, say, the IoT from other communications, might make sense, along with SCADA and military communications -- all largely control networks.

Splitting out video and voice from text and data, likewise. Much of high-demand comes from those.

As the Internet heads from the first billion or two users to the remaining five or so, questions of what technology, interfaces, and devices are appropriate for a set of digital newcomers may be worth consideration.

Or, looking at this differently: the degree to which those already here may be interested in maintaining a separate space for themselves. Not that this is necessarily equitable, but it may well prove to be attractive, say, in particular to minimise fraud and other malicious use or attacks originating from the global poor, who are in many ways justified in wanting some of the pie that's been denied to, or taken from, them.

How the needs, wants, attractions, and/or detractors or aversions affect technology development and adoption remains to be seen.

thread

dredmorbiusMay 2024

When did you first encounter Usenet?

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...>

thread

dredmorbiusDec 2023

The principle use-case for Usenet has always been small, controlled-access networks.

That's what ARPANET and the early Internet were, where "small" meant < 1 million active participants and "controlled access" meant meeting the requirement of belonging to one of a limited number of academic institutions, tech companies, government agencies, or select military units. Total Usenet participation in April 1988 were 140,000 readers, by Eric Reid's DEC surveys, as cited in John S. Quarterman's The Matrix (1990):

<https://archive.org/details/matrixcomputerne0000quar/page/24...>

Spinning up an NNTP group for internal company communications, for a family group, or for a (sufficently tech-savvy) organisation or hobby group would be a viable option. My experience is that peak group experience tends to occur with somewhere between about 5 and maybe 1,000 participants on the high end, and that's being generous. In practice, 5--50 is probably a better sweet spot.

The technology isn't the stumbling point nearly so much as convincing people to use the service. In some industries, privacy and disclosure rules may preempt usage (e.g., healthcare, finance). And the fact that Usenet has no intrinsic authentication or attestation of identity (though these can be provided by additional mechanisms such as PGP/GPG signatures) means that publicly accessible newsgroups are likely to be nonviable given malicious actors.

What's underappreciated about much of the early Internet / online world is how small most of the "large" (as in influential) groups or movements were. Usenet, the WELL, Slashdot, and early blogging had core groups of perhaps a few hundred to a few thousand people, possibly fewer. Yes, there were additional participants in many cases, but in terms of how concentrated the bulk of activity was, small.

thread