The 100 oldest domains on the internet

Wait... it wasn't?

Back in the mid-1980's IBM was still riding the System 390 mainframe cash cow. They were more concerned with connecting their stuff to their other stuff (AS/400's, PC's, RS/6000's, etc) using their own Systems Application Architecture over Token Ring networks.

Laf

When I left Symbolics I went to a company (Alpharel), that was using the 4341 mid sized mainframe as a central processor/database manager for a set of periphriels that laser scanned Aperture cards (punch cards with micofilm embedded), had a 300 dpi graphics monitor, and large scale engineering document printers. It was a lot of fun as well and the AT&T guys that were working the system had no clue about how to make an IBM mainframe sing as they all were from the Unix world.

Lol.. good one.

Yep, in the 1980's IBM had boatloads of 1401 systems on lease to the IRS and Navy, the things sere running Hollirith Card readers and had paid for themselves about five times over. The game was to keep any installed base growing and not communicating with any one else. Lots of tape machines doing nightly backups too.

Data General, Burroughs, Honeywell were all huge cluge makers that had the political clout and Big Eight consultants behind them. That's where the leasing business began to take on the creative funding models.

If there's one thing the names on the 100 list all know, it's how best to partner with government to profit. The Defense contractors like Lockheed and Boeing as well as the communications and research groups all make sense as early listings.

Maybe the Omnipotent GE should fall to the wired in loop. Heaven knows they've had their own contract fraud division at the Pentagon since the 1980's maybe they should be on every communication list to help find the billions missing from taxpayers. No, they're busy hunting enemies, no time to remove pcb's or correct fraudulent spending when there's mischief afoot in a remote place, but I digress...these few jumped out for big oil and chemicals but it made me wonder, who and where are the rest??

21. 05-Aug-1986 GE.COM
40. 05-Nov-1986 ALCOA.COM
55. 11-Dec-1986 PEREGRINE.COM
67. 04-Apr-1987 PHILIPS.COM
81. 27-Jul-1987 DUPONT.COM

55. 11-Dec-1986 PEREGRINE.COM

More than a few don't go to anywhere with a www added to them but peregrine goes to Hewlett Packard and it just made me wonder if anyone knows where the rest are now??

Now I should point out an inaccuracy in what I myself said, "domain names existed long before DNS was put into place." The term "domain name" probably only came to be along with DNS. So the host names that existed before then are not technically domain names. But my point is that the order of companies on the list isn't really that significant. IBM, for example, was probably on the net before the date shown in the list.

It is my memory that the .mil and .edu "endings" were in existence as far back as the summer of 1984 when I was working at Symbolics. Everything was in software and the .edu and .mil addresses were just text aliases so that you did not have to remember the hard ip addresses (your IP address was hard coded into your ethernet card).

I was a test engineer at Symbolic over the summer of 1984 and had a LOT of fun fussing around on ARPANet at the time. There were only about 250 nodes on the ARPANet in the summer of 84. I am a double idiot in that the same summer I saw hypertext used for the first time on an Apple computer in Hypercard and knew that something big was going to happen but was too interested in skiing and backpacking to care.

Thanks for that addition vas, it made me wonder because my son went to Stuyvesant and had a professor for computer science who was involved with Berkley and the early development. My son had told me that because of whatever his name was, Stuyvesant was one of the first with a web connection and still had some special domain from the creation days and there were no academic listings, only commercial ones.

Those universities that were early adopters have an incredible asset that very few people know about. They have a Class "B" license that essentially gives them either 128,000 Ip addresses or 256,000 addresses. You can't buy a class B license today for love nor money (or at least without many zeros on the check).

I think that advantage will soon become redundant as IPv6 gets adopted through which there will be enough addresses for every grain of sand on the planet. That is some 340 trillion, trillion, trillion.

Source

I only make reference to IPv6 in passing in my recent article titled 'Virtual Worlds and Where they are going' but it is a key element in my argument.

That's because all the short, succinct ones are used up now.

That's because all the short, succinct ones are used up now.

yep every 3 character combination is taken.

Wasn't me Ms! Vanessa seeded; I just provided the lift:-)

LOL - and I thought it was the effect I was having on you:-)

Yep DEC was one of the ones there in the beginning, although DEC was pushing their overlay on the ARPANet called DECNet. I had a DECNet address for several years at the university of Alabama in Huntsville and there were some really cool features such as instant messaging to any host that you knew the name and address of someone to chat with!

I think that DECNet died soon after Mosaic destroyed any competing system. I also had an email address at MIT's Athena host, which was a major early Usenet host.

Ahhh the early days when the biggest porn server on the Internet was the EE Department at Delft University in the Netherlands! (I wonder if the horny geek squad data still exists somewhere)

That's because digital had all the scientific and intelligence customers and they all ran UNIX. If you couldn't do UNIX it was considered a business application or worse, word processing!

Pamela

Now that was funny in a way that only true geeks who were there in the early days can appreciate.

My areas of specialty at DEC did not involve DECNet, so I can't speak about its good and bad points with much authority. There were several popular networking protocol suites at the time, and ultimately ARPA's IP and its companion level 4 protocols (TCP, UDP, et al.) prevailed over all DECNet and all the rest to become the software foundation of the Internet.

I don't recall ARPA being interested in DECNet, which is not surprising because ARPA developers were implementing their own protocols for ARPA purposes. In the late 1970s, a substantial number of ARPANET hosts were DEC mainframes (hosts labeled PDP-10, DEC-1080, DEC-1090, DEC-2040, DEC-2050 in the 1977 network map) which ran DEC operating systems or academic oddities such as the Incompatible Timesharing System, aka "ITS" of the Artificial Intelligence Lab at MIT. (ITS was a hoot - a geek's system through and through. Its top-level command language, for example, was of all things the DDT assembly language debugger !)

The ARPA community had a substantial investment in DEC mainframes then, and when the time came to consider moving to more contemporary hardware - the microprocessor had yet to arrive on the scene - DEC's VAX caught the eye of many. Planners however questioned the wisdom of using the VAX in combination with its proprietary operating system VMS, and ultimately decided to fund development to enhance Bell Labs' UNIX/32V to be a viable base for the project's goals. Bill Joy and his crew at the computer science department at UC Berkeley got the job and executed it superbly. This was a huge shot in the arm for UNIX, which quickly took the world by storm and in doing so established the viability of a portable operating system. In the 1980s, companies such as Sequent sprang up offering proprietary hardware products that ran UNIX. As Pamela noted, computing sharpies of the day found in UNIX a solid base for their work. You could even do word processing, although to produce a professional-level document such as a user's manual, you had to be very competent with using troff, which offered no WYSIWYG assistance.

Back in those days the porn servers were all located at Universities where there was a lot of storage.

Trust me, their problem was that they were so proprietary that their little group of geeks at MIT were soon eclipsed by the larger world.

I quit because the LISP machine that directly executed LISP statements was considered so proprietary that they would not give the test techs or test engineers the schematics to troubleshoot the boards! There were a lot of mods to the boards as well and the only way to troubleshoot was to take the netlists for the boards and try and trace out the circuit paths. I got peeved and took the netlists and reverse engineered a schematic, which was promptly taken away from me. They had millions of dollars worth of bad boards (these were big 18" tall boards!) that just kept piling up in inventory. A really crappy way of running a company and the VP of Engineering was an idiot. The company went out of business because they just wallowed in their own crap boards and did not keep up with the times.

LISP was interesting and was able to do a lot of cool stuff back in the day but their near religious zealotry that LISP would lead to artificial intelligence was just a load of hooey.

CAR( oldest_domains ) = symbolics.com