Hacker News Reader

Comment by hyperman1 | Hacker News Reader

One thing I learned from Cliff Stoll's the cuckoo's egg is that the internet was far from the only computer network. Every country was building one or more networks. There was some kind of cambrian explosion of network technologys going on.

What the internet did was more subtle, but far more important: It standardized the technology, connecting these networks. Instead of tons of little private network, we got 1 global net owned by nobody.

I have no problem believing chile had one of these little private networks.

View on HN

Isn't that the point? The Internet is an "inter-network", a network of networks.

But it wasn't even "The Internet", singular, to begin with. Any network of networks is "an internet". It's just when all the networks (or internetworks) are connected together (because Metcalfe's Law - the value of a telecommunications network is proportional to the square of the number of connected users) that it becomes "The Internet".

A believe a large portion of this is thanks to Tim Berners-Lee, CERN, and the web. That was the free technology that made it immediately apparent how valuable a global internet is.

(Unfortunately, with social networks and chat systems and such opting out of federation and defaulting to closed ecosystems, I think we are moving toward independent corporate-sponsored networks, this time built on top of the global one.)

True. The web was built on top of the internet, which gave it a huge boost. Today, it's hard to separate the www and the internet, they seem almost 1 combined entity.

I wasn't there, but the internet presumably had value before the www, by having email, ftp and usenet.

It also demonstrates how futile nationalistic feelings are on computers. We invented the internet says the USA. We invented the www, says Europe. Both are irrelevant to the point they are almost wrong. The tech got valuable only because a planet worked together.

> I wasn't there, but the internet presumably had value before the www, by having email, ftp and usenet.

Don’t forget telnet and multi—player games. The Internet was a pretty cool place before ad networks conquered it.

They did but the cracks are showing. China with their great firewall. Even Western countries censoring stuff. Russia making sure they can function independently because they'll probably be sanctioned one day.

It's far from one happy family.

> I wasn't there, but the internet presumably had value before the www, by having email, ftp and usenet.

As someone who was barely there, it was largely a standardized client problem. That's why NSCA Mosaic and later Netscape Navigator were so important.

The Internet provided transport... but for what?

To GP's standardization point, the Internet wouldn't have been successful if every server required you to have a different client to talk a different protocol. (The landscape was very different than that in which mobile apps now live)

Web protocols + a client browser that understood them = palatable UX

That's true. However, before the web, it was absolutely common to anonftp to some site, download the software, type ./configure and make, and off you went. That's how software got distributed. And Internet standards.

Not the same as the web, of course, and your grandmother wouldn't have been doing it.

one of the bigger problems was telneting between VAX and UNIX systems.

there was content back then, but no curation. we would telnet from one machine to another using CLI.

the reward was a wall of ASCII highly formatted with extended fonts.

you could lurk in the login areas trying unix-fu or find a mud or a BBS.

there was always tin for reading usenet as a method of content search.

there wasnt so much content fiefdom, you could share links or post content with little oversight except numbers of attendance.

i was there.

  one night on a dumpster dive i found a minitel terminal [i thought it was "trash" but it was perfectly working.]

it dialed out no problem, there was a queue of phone numbers

I was launched into a world of freenets and BBSs

Wait, Chobits?

im not the only one that found that path, i think Akira may have had a similar beginning.

More specifically, Vint Cerf was one of the main guys who actually standardised the interconnectivity that made them all one internet in the first place

https://en.m.wikipedia.org/wiki/Vint_Cerf

Indeed, the web was the "killer app" of the technology

There were a lot of protocols at the beginning. Google worked hard to kill some of them: nntp, ftp (archie), http.

Sesh, people blame google for everything.

Usenet died due to spam problems.

Ftp has been on the way out for a long long time.

Neither of these had anything to do with google.

Usenet is still "dead", eh?

Yes. It is.

Its ghost might still technically exist, but it is a pale ghost of what it was before.

They didn't, though. They had a bunch of telex machines that were old even at the time. The title of this link is hyperbole to serve an ideological agenda. It's Teen Vogue tech history.

Before the Internet became widespread, people were using BBSs, and several large (for the time) networks of interconnections between them. Fidonet, and many more regional ones like Z-Netz. There were also networks for businesses and research institutions that were not the Internet. None of these were just a bunch of Telex machines, even if the network in the article was.

Also Telenet when I was a kid.

> Before the Internet became widespread

This is looking at things with blinders and assuming the public at large determines the significance of the Internet. The People were not using BBS, etc. A tiny vanishingly small minority of intrepid computer geeks were using BBS, etc. in the late 1970's, but Arpanet was already very well established by 1973.[1] Just because it was mainly limited to academics and corporations, this still vastly outmatched BBS use in the 1970's. BBS only came into popularity by the mid-1980's, when the Internet was already huge and stayed big, and BBS were effectively extinct within a few years of the invention of WWW. BBS use itself was never "widespread," unless we define widespread as a fraction of a fraction of a percent of the population. Not to diminish the nostalgia for the unique fun of logging into a BBS, in reality, hardly anyone ever used them, which would be maybe tens of thousands at most among hundreds of millions. The number of academics at one large university attached to the Internet in 1978 could match the entire population of BBS users for it's entire existence.

[1] https://www.computerhistory.org/internethistory/1970s/

If you say so. I don't have any hard numbers (I would love to see them, either way), but I think there was a brief moment around 1990 when BBS usage was more widespread than the Internet. I wouldn't (and didn't) say BBS usage was widespread in absolute terms. Nor did I say anything about the significance of anything.

I don't have hard numbers on BBS, but on the internet-history@ mailing list, someone estimated there were around a million people worldwide on the Internet in 1990. This is a pretty informed estimate. You think BBS was more than that?

Apparently, there were close to 7000 Fidonet nodes in 1990[1]. What's a reasonable average for distinct user accounts per node? I don't know. If it's 100, that's 0.7 million already.

Compuserve had 380000 users in 1987[2], when it was basically US-only. Now, I don't want to argue that Compuserve is a BBS (though I suppose you could). But it's evidence people were using their modems to connect to things that weren't the internet. And Compuserve in 1987 apparently had no ability to exchange information with the Internet proper; when I used it as a child, a couple of years later, I used Compuserve -- its groups, pages, whatever they were called, and did not use it for Internet access.

[1] https://de.wikipedia.org/wiki/FidoNet#/media/Datei:Fidonodes... almost 40000 at its peak in 1996

[2] https://en.wikipedia.org/wiki/CompuServe#Market_share

I wouldn't dispute any of your numbers. It sounds like they were roughly comparable. At least that's the best we can say 34 years later.

Yes, much more. The internet didn’t go mainstream until the mid-90’s. According to https://en.m.wikipedia.org/wiki/Bulletin_board_system there “were 60,000 BBSes serving 17 million users in the United States alone in 1994.”

First of all, "in the US alone," is misleading because the vast majority of BBS and their users were actually in the US. In 1990, the Internet was composed of 350,000 hosts being accessed by over a million computers (aka users). 60K BBSs means a capacity of exactly 60K users simultaneously, while the Internet capacity at the time was... more that a million simultaneous users... IOW all of them. The Internet (called ARPANET) was very well established by 1973, but experienced exponential growth from the advent of WWW.

Your citation has no citation of its own. Curious how there could be 17M BBS users in 1994 when the US had only 11M modem-equipped computers.[1] IOW, the wiki is very wrong, and the claim is unsupportable.

[1] https://www.pewresearch.org/politics/1995/10/16/americans-go...

Even if it’s only 11 million, most of those users were not on the internet in 1994. They were on local BBSes, CompuServe, GEnie, The Source, and other online services. Do you remember that time? I do. I’ve started BBSing in the late 80’s and first got on the Internet in 1991. It was mostly academics, defense contractors, and early ISPs… which generally started as BBSes.

I don’t dispute the early internet could’ve supported many more users than BBSes. They simply weren’t there yet.

You are vastly overestimating the size of that population of BBS users, and vastly underestimating the population of academics and students at hundreds of universities as well as engineers and computer scientists at hundreds of companies. It wasn't everyone logging into BBS, wasn't most people, and if you were logging into BBS in 1990, you were pretty special. In 1983, when I was using BBS, I never met another soul until 1989 that had any idea what I was doing. If you were at university using the Internet, this was commonplace by 1990, and really it was pretty common by 1980. But since you're apparently unaware that a network of networks of hundreds of universities and hundreds of corporations spanning the globe existed prior to your awareness of BBS, it makes a lot of sense that you might believe that.

I think you are vastly overestimating the number of university users that actually used the Internet in those early days. For example, just because a school "had access" did not mean all their students actually used it. Around here, even as late as 1993, at a state school, you had to sign up for a CS class to even get an email address.

I stand by my statement that the total number of BBS users (including online services, like CompuServe, Prodigy, GEnie, pre-Internet AOL, etc.) out numbered the users on the early Internet in 1990.

Let's break it down. CompuServe, the oldest of the big 3 dialup services, never had more than 250K subscribers, and it was bought by AOL in 1997. When AOL went public in 1992, it had fewer than 200K subscribers.[1] At its peak, in 2002, AOL had 25M subscribers.[2] The rest of the dialup services never totaled more than 500K subscribers combined, total for their entire existance.

Again, in 1990, the Internet had 350K hosts worldwide and over a million users. The total number of dialup subscribers in 1990 doesn't begin to approach that number, and it doesn't matter if you pile on all BBS users, because BBS at its peak, low hundreds of them. Within 2 years of WWW arriving, there were only 20 left.[3]

So take your 10K BBS users, take all the CompuServe users ever, 250K, take the 200K AOL users in 1990, and all the other dialup subscribers for the entirety of their existence, 500K, and with this screwed up cheating math, you might get a million to match the over a million Internet users in 1990.

[1] https://time.com/3857628/aol-1985-history/

[2] https://upload.wikimedia.org/wikipedia/commons/6/61/Aol_subs...

[3] https://www.theatlantic.com/technology/archive/2016/11/the-l...

I will admit it difficult to get accurate numbers, but this Wikipedia article claims CompuServe had 600K and Prodigy had 450K+ subscribers in 1990: https://en.wikipedia.org/wiki/Prodigy_(online_service) It references NY Times articles that are pay walled.

This graph shows roughly 12,000 BBSes existing in 1990: http://bbslist.textfiles.com/support/statistics.html How many users do you think each one had? 50? 100? Of course there is some overlap and it's impossible to figure out actual unique users. Regardless, it is way more than 10K. One large BBS alone (Rusty and Edie's, which got busted for piracy in a big scandal) had over 14,000 subscribers: https://en.wikipedia.org/wiki/Rusty_n_Edie%27s_BBS

And that last article from The Atlantic says there are 20 dialup BBSes existing when it was written in 2016, not within 2 years of the WWW arriving: "Did any direct-access, telephone-dial-up BBSes survive the internet’s proverbial asteroid? Sure enough, there are about 20 known dial-up BBSes in North America"

The Internet killed BBSing, that's for sure. But that didn't happen until the mid 90's when it went mainstream.

I saw the reference, and my explanation is CompuServe, and especially Prodigy, lied, precisely for the same reasons Facebook lies, about the number of subscribers. You think any BBS users ever patronized more than one BBS? I wonder what the record is for the most logins on one BBS in a day was, but I expect the average to be close to 10 logins a day, because the day is so short, starts after work ends at bedtime, and because the bandwidth was so small, every execution took forever, so logging in and tying up a BBS for 2 hours was probably a very short session. Also, rhetorically, of the 14K alleged BBS that supposedly existed, how many of them could be considered popular among BBS users? I wonder how many would go days or weeks without a login.

If you've ever browsed old newsgroups, you'd get an idea of just how many people were online. By 1983, thousands of people participated from more than 500 hosts, mostly universities and Bell Labs sites but also a growing number of Unix-related companies; the number of hosts nearly doubled to 940 in 1984.[1] The number of newsgroups grew from more than 100 as of 1983 to more than 110,000... some newsgroups receive fewer than a dozen posts per year while the most popular can get several thousand in under an hour.[2] Presumably, each newsgroup had a creator, and my hope is to give you better perspective on how many were using the Internet daily in the early 1980's, which would have only increased massively by 1990. The more than a million Internet users by 1990 figure is extremely soft. The reason the WWW exploded is probably due to Berners-Lee announcing it on Usenet. Just sayin' every Usenet user was using the Internet.

[1] https://en.wikipedia.org/wiki/Usenet#Network

[2] https://en.wikipedia.org/wiki/Usenet_newsgroup#Types

It's possible they exaggerated their numbers, certainly, but we can only rely on the public data.

I was very active on BBSes in the early 90's. I would call generally 5 or 6 a day myself. I also ran my own, starting around 1992. I don't have any exact figures, but I remember I had roughly 300 users registered, of which about 20 to 30 called a day (I only had a single phone line.) I was a teenager at the time, and most of my users were also teens, logging in after school, from around 3 PM to midnight. There were larger systems in that area that had 8 to 10 lines, and 1000's of registered users.

I remember Usenet quite well. I was active on it, starting around 1989, through a local system that had a UUCP feed. Also, not every "Usenet" user was necessarily an "Internet" (TCP/IP) user. You could receive mail and news through UUCP (batch, over dialup), without being connected to Internet with IP. More ancient history, but UUCP feeds were very common for smaller sites.

Let's just remember that most Internet services nowadays measure their traffic as "7-day actives" and sometimes "30-day actives." (Maybe for Twitter it's "daily actives"?)

That's a way of distinguishing people who are just signed up from those who've actually done something in the last N days. I doubt that either the Internet or the BBS's have those numbers for 1990 or thereabouts.

First of all, "in the US alone," is misleading because the vast majority of BBS and their users were actually in the US.

Another vague assertion stated confidently, but without sources. How do you know this? What percentage is a vast majority?

60K BBSs means a capacity of exactly 60K users simultaneously

Again, stated like a fact, but there were BBSs with multiple lines. What percentage? I don't know. Do you? But "capacity" is a red herring anyway, just like "significance".

Multi-line BBSes were few and far between. As a very active BBS user in 1990, there were about 100 systems in my local calling area. Of those, I only remember 3 or 4 having multiple lines. A system like that was a serious investment: you had the $$$ for the phone lines every month, plus all the equipment for the system itself.

We were talking about 1990. 1994 was much, much different. Not that that isn't also interesting.

If you believe this, Prodigy and CompuServe combined had over a million users in 1990: https://en.m.wikipedia.org/wiki/Prodigy_(online_service) Probably the vast majority of those users also called local BBSes.

> Probably the vast majority of those users also called local BBSes.

This is an absurd notion. The opposite is probably true, that most if not every single BBS user also used dialup services, but most users of dialup services probably were entirely unaware of any BBS.

btw, CompuServe never had more than 250K subscribers total. Prodigy had far less. So it is likely Facebook math being performed here with subscriptions that existed but were never used.

The reality is most people could not afford online services. This is anecdotal, but most local BBS users I knew did not use CompuServe, Prodigy, or any of the others. The hourly fees were expensive.

This is all ancient history, but most articles I look at showed CompuServe around 500K by 1990: https://www.researchgate.net/figure/tracks-the-number-of-hou...

AFAIK the French minitel system was used by The People over there. It was semi obligatory for e.g subscribing to (some?) higher education, and maybe interacting with some governemental services. The People, once introduced to minitel, had some commercial services available, including primitive porn.

I consider minitel a fork in the road to the www, with different architectural choices leading to a different net culture. Minitel had a clear consumer vs producer split and 1 central governer, compared to the www free-for-all peer to peer vision. It probably would not have grown so fast as the www. But it had payment infra from day 1 and no 100%free connectivity, so it wouldn't become the privacy invasive ad infestation the www has become.

Reply on news.ycombinator.com