Oops. wasn’t thinking. I knew that.
Thanks
<abashed>
I can also imagine scenarios where the Internet was controlled by industry from the beginning, and where people were supposed to pay from the beginning, and where it would never have taken off. The ARPAnet was heavily university oriented - I used it in 1975 at Illinois to use the Parry program at Stanford.
At the same time I also used PLATO, which was far more advanced. Besides hyperlinking, PLATO also was centered around graphics terminals with touch screens. PLATO had message boards, instant messaging, email, forums like the Dope, MUDs, interactive multi-user games, and even a smidgin of porn. It all ran from a couple of Cyber mainframes, but PLATO terminals were distributed around the world. It never took off because it got sold to CDC, who screwed it up. But it was ages before the rest of the world caught up to what PLATO had.
The Wrights invented the 3-axis control system, for which they should be given plenty of credit. But the physics of controlled flight were known since 1799, when George Cayley described the four forces of flight, methods to create stability (including dihedral and a movable tail), and there is a drawing he made in that year which shows an aircraft with wings and a movable tail for control. He built a flying glider in 1853 which had a movable tail, dihedral, and a cambered airfoil.
And of course, people had been flying gliders in the previous decades using weight shift for control. This wouldn’t work for a heavier-than-air craft of the time, because it would be so heavy that weight shift wouldn’t work. But powered hang gliders today use the same principle and can stay aloft for hours and fly hundreds of miles. Had engines with that kind of power-to-weight ratio been available in the 1800’s, we might have seen controlled, powered flight before the Wright brothers.
In addition, although the Wrights invented the 3-axis control system (and did lots of other important engineering work), they never figured out the aileron, and their wing-warp mechanism was ultimately a loser although it worked for their purposes. I seem to recall credit for the aileron going to Santos-Dumont, who was also building flying machines around that time.
I wasn’t trying to deny them credit. I’m saying that had they lived 50 years earlier, they would never have achieved controlled, powered flight because the engineering needed for the engine just wasn’t there yet, and had they been born 10 years later, someone else would have beaten them to the punch.
History is full of examples like this, where new fields of innovation lay dormant for lack of the engineerng ability to realize the innovation, and once the technical ground becomes fertile, the inventions spring up rapidly, and sometimes in multiple places at the same time.
This is taking nothing away from the Wright Brothers. They got there first. They made substantial contributions to aviation. However, the basic concepts of flight were known for some time before they came along, and a hell of a lot of people were working on the problem when they were. It was a race, and the Wrights won. Had they not existed, Louis Bleriot or Alberto Santos-Dumont or any one of a number of other aviation pioneers would have done it within a few years of the Wright’s date, and modern aviation would look substantially the same. Do you disagree?
I know all about ARPANET. I used ARPANET. I’m not talking about the original network infrastructure. “The Internet” as we know it today is the vast collection of web sites, commercial and non-profit enterprises, and related technologies (Javascript, flash, PDFs, AJAX, tunneling protocols, SOAP, ad nauseum) that are built on top of the basic infrastructure. This is where all the complexity is, and where all the value is. Almost none of that was directed by government or any central planners. It grew organically.
They got there first with an inferior solution, and locked the French people into it. As a result, they have lagged behind the rest of Europe in Internet access for a long, long time.
Exactly my point. Minitel is what a government-run internet looks like - locked into one standard, no room for the kind of innovation that comes out of a competitive market. Had the world’s governments all adopted mMnitel and set up a worldwide ‘standard’ as some were calling for, we’d still be stuck with Minitel terminals, and we’d have no idea what we lost.
And this isn’t a far-fetched fantasy. There have been all kinds of attempts to have government control the internet and IT in general. Various groups have tried to push licensing standards for programmers, licensing and government approvals for public web sites, ADA regulations extended to the internet, etc. In recent years, it’s mostly been Hollywood and the RIAA trying to get regulatory control of the 'net, but in the earlier years all sorts of special interests appealed to government to lock it down.
And yes, I do give the Clinton Administration and Al Gore credit for resisting those calls for government control. But it’s a never-ending battle.
A lot of credit needs to go the the invention of the router. Without routers there is no internet. This came out of Standford in the early eighties.
“Sugar, I can’t talk right now! My President needs me!”
Stranger
The original ARPANET IMP nodes were packet-switched routers. These pre-dated TCP/IP by a decade. And packet-switched networking technology in general existed before ARPANET was built. This was the early 1960s, not the 1980s.
Stanford was, however, one of the original nodes on the network. And the world’s biggest networking gear company, Cisco Systems, was founded by researchers who came out of Stanford.
PLATO is an interesting example, and I think you can fairly make the case you made - in this case, a single company took control of a technology and totally hosed it through bad decisions and lack of vision.
But PLATO was also hobbled in other ways. It was very hard to make courseware for it, and therefore it was extremely expensive to use. It was ahead of its time, so the computing hardware wasn’t quite up to the task (or if you got the right hardware, it was bleemin’ expensive).
PLATO makes my point as well, however - there were alternatives to the ARPANET-based internet springing up. BBS’s were evolving, and networking between them had become quite sophisticated right before the internet came along and killed them all. Fidonet and other networks were moving beyond distributed file sharing and into more sophisticated ways of sharing information.
In the 1980’s I had a company which made software for a BBS called the MajorBBS. It could handle up to 64 incoming phone lines per node and the nodes could be networked to share E-mail and files. We wrote gaming software, including a real-time multi-user dungeon and the first online poker program, as well as billing and accounting software, statistics packages for traffic monitoring, a full-text retrieval system that had a google-like textual query language, and real-time chat software. In the 1980’s, there were thousands of these multi-line BBSes running all over the world. We had our software in U.S. embassies, and they used our document retrieval to create google-like searches of government documentation.
At the same time, larger mega-BBSes like CompuServe and America Online were growing like mad, and they were starting to include multi-node networking and other internet-like features. I got addicted to message boards by CompuServe’s AVSig and other ‘sigs’.
The internet was a disruptive technology, and soon swept everything else away. But if it had not come along in the exact form it did, it would have arrived anyway in a slightly different form. Everyone was heading in that direction - including PLATO.
It’s not too difficult to imagine 2 or 3 competing networks owned by megacorporations (such as AOL and Compuserve) a couple of outliers like the Well and thousands of dialup BBSs for er specialty interests. Without big guv, I can’t see what would cause the various networks to be pooled together.
What would cause them to be pooled together? Competitive pressure. It’s what people wanted. My company worked very hard at connecting to the various other networks. If our BBS could connect to FidoNet, or to the fledgling internet, or connect to CompuServe E-mail, then I had a competitive advantage against the other BBS’s. Then they would be forced to follow suit. Or, in my software developer hat, I could see the competitive pressures on the BBS’s, and come up with ways to sell software that gave them advantages over their competitors. So did hundreds of other developers. There was an entire cottage industry of small software shops writing add-ons for the MajorBBS at that time, and just as many for the other big BBS platforms. We were all trying to come up with the best new app that the BBS’s would want. The hunger for new bbs add-ons was amazing. You could write some little utility in two days, offer it up for $100, and have 30 people phone and buy it off you in the next couple of days. New companies were springing up daily. It was like a mini Dot-Com boom - there were annual conventions for BBS wares that attracted thousands of people. There were monthly magazines covering the BBS scene. All of this energy was directed towards making it all better and more sophisticated, and more connected. The best selling apps at the time were the Fidonet and Internet connectors that were available.
Public demand was causing the various networks to be pooled together. And the ultimate pooling was the Internet, and it went that way under no direction from the government. It simply became available, and the market glommed onto it, then organized itself.
While we’re here, what about SLIP and PPP? I used to have a shell account on my 286, but, when I upgraded to my 486, I discovered my shell account also provided SLIP access by running a UNIX program. I created a script to automatically log me in in Trumpet, and my Win 3.1 machine was online and raring to go.
Then I got another account, and, although it also came with a UNIX shell, everyone was going on about PPP, which didn’t need stupid scripting. I always wondered what caused the change. As far as I know, most SLIP accounts didn’t need scripting either, so that wasn’t it.
Anyways, my friends tend to think I’m the one who experienced the early Internet, with gopher, pine, lynx, and finger. It’s nice to meet people who go even further back.
PPP and SLIP were both used for the same purpose: tunneling IP over serial lines (and modems.) But PPP had a lot of advantages: better compression, error correction, authentication, and it could configure your IP address and gateway settings from the other end, instead of having to configure this manually on the client side.
PPP is still used on broadband networks; many cable modem systems use PPP over Ethernet (PPPoE) to establish authenticated sessions on their networks (since Ethernet by itself doesn’t have any authentication mechanism; you just plug in and go.)
But nobody uses SLIP anymore.
(Cue ten Dopers jumping out of the woodwork to point out all those embedded networked weather station/nuclear core temperature monitoring robots from Khazakstan that push around gigabytes of thermal data via BitTorrent running over SLIP on top of bare copper which is simultaneously a telegraph line connecting the Kremlin with Downing Street.)
Writing good courseware was hard, but doing good anything was hard. The Tutor language (which I knew) made the mechanics of writing courseware very simple - I used it for the class on PLATO I took. It was actually much harder to write general purpose games using it, but it happened. I forget how it handled multiple users (my TUTOR manual is in a box somewhere) but it did it before most other systems could.
Terminals connected to a mainframe or bunch of mainframes was clearly not a scalable technology. But a company less interested in focusing on the one area they thought they could make money on (education) and wishing to keep the technology proprietary could have quickly sold plasma screens that would connect to PCs or workstations, and use the Internet to replace the mainframes. Before I left Illinois, in 1977, there was already a PLATO IV terminal with a microprocessor built in, which some people had programmed to cheat at games by responding faster than a user could. One of the best interactive games, Empire/Michelin, got ported to UNIX in the early 1980s. But the developers of PLATO were not network people and I don’t think there was a plausible way that PLATO would ever have evolved instead of the Internet.
BTW, PLATO response time was always pretty good, and there was enough spare capacity at night that the code that went through all the cases needed in the proof of the four color theorem ran on it.
Obviously the www today is far richer than anything we had on Minitel, but it’s absurd to say that Minitel was a failure. This is still the Straight ope, right ?
Minitel was a huge success. 10 years before the www was even a twinkle in its daddy’s eye, with Minitel we had online banking, secure payment, hookup sites etc. For purely commercial purposes, internet didn’t really catch up with Minitel until the late 90’s In some ways it still hasn’t. I was a Minitel ‘site’ manager (among other things) when the www started taking off, and quickly brought my company onto the net. At the time I was happy to tell anyone who would listen that Minitel was history, and the ‘new thing’ had come along - turned out I was jumping the gun somewhat - it was years before the net was mature enough to fully replace Minitel.
Minitel was a precursor of much of what showed up later on the net, in fact the Minitel companies have generally gone on to become internet success stories, including a small ‘porno’ provider that has gone on to become France’s largest broadband provider.
Dismissing Minitel in an attempt to support some naive and poorly thought out political “philosophy” makes as much sense as saying that television is a failure because governments imposed licensing and set the standards early in the game, or the international phone system is a failure , or that fax was a failure because it wasn’t in color and didn’t support moving pictures.
Since it hasn’t been said already, any “fact” in a Dan Brown book should be as highly regarded as any story in the Weekly World News.
‘Failed miserably’ was a poor choice of words on my part. It worked reasonably well for the functions assigned to it. But it failed in the sense that the internet very rapidly became orders of magnitude better than Minitel (although Minitel remained reasonably competitive in some niches like online bill paying), and yet France was locked into that system because they had pushed so many terminals out to the people. In addition, the widespread availability of Minitel terminals probably slowed the adoption of home computers in France.
The fact is, despite Minitel having somewhat of a head start, the internet rapidly eclipsed it. And the internet did it without any central planning, without government authorities determining its direction, and without giant government industrial research programs and ‘investments’ in the internet. It simply evolved, pushed by the forces of the market.
I wasn’t intending to disparage Minitel so much as I was trying to point out the difference between government technology programs and market-driven programs. For example, look at the evolution of Minitel - the amount of change it underwent, and compare it to the amount of change that was taking place on the internet during the same period.
I dispute the characterization of the internet as “Market-driven”. Big guv poured billions into it at various times.
…but at the end of the day AOL would have every incentive to shut out Compuserve. An analogy might be ATMs. Sure, you can dream up a story whereby foreign ATM fees don’t exist and no-fee ATM networks exist. Heck it actually happened for a while: the Plus network used to ban ATM fees. But that happy state didn’t last for long. I bring this up because it’s an empirical example of a network that had minimal governmental intervention.
Sure, competitive pressures exist. But in general there’s a tendency towards oligopoly, a characterization consistent with the privatized internet hypothetical and the actual experience with ATMs. Now a world where there are 2-5 competing internets wouldn’t be TEOTWAWKI, and I don’t doubt that something like Amazon wouldn’t be born under such circumstances. But I daresay that the outcome would be less efficient as accessing to 2-5 private networks would involve higher costs for internet vendors. OTOH, methinks Minitel would do fine under such circumstances.
When I first affiliated with Quantum Computer Services, the company that would later become AOL, they had three relatively small networks hosted on their Stratus mainframe, QLink for Commodore 64 users, AppleLink, and PCLink, whose users were obvious. What made them unique in 1989’s text-driven connectivity market was that they had GUIs, relatively simple ones but visual interfaces that were attracting a broad range of clientele. CompuServe still had about a 10:1 membership advantage over them, but their market share was growing. And as I was forced to withdraw by a combination of financial issues and what proved to be a heart attack, they had plans to merge the three services into something they were tentatively calling America Online Services. The rest, of course, is history.
Seamonkey is actually a truer descendant from Netscape. I use it and it’s like I’ve been using Netscape from day one.
The design history of the submarine is in a similar boat.