The World

On March 21, 1989 STD, Inc. was incorporated as a software consulting company. In the early days we worked on some interesting projects including re-writing the user interface for a sizeable electronic mail / conferencing company, editing and writing for technical magazines, some various software consulting gigs (hi GraphOn!) and not a lot more because this period only lasted several months.

Our personal backgrounds were in academic computer science, the internet, software development, and network, server, and computing facilities deployment.

The original name for the company was "Software Tool & Die", a name we still use as a d/b/a. That name was from an old, inside joke. Tool & Die companies were machine shops generally known for taking a blueprint of a tool or piece of equipment and a hunk of metal and making them both match to a high degree of precision. For example, an engine part which was no longer available, or a special purpose part for a factory, often a "die", something which might hold some raw material for automatic machining. "To Tool" was an expression in the academic technical community which meant to overwork, such as in the night before a final project was due, or on a grant proposal under deadline. The pun of "Software Tool & Die" should be obvious. To work on software with great precision and do it until it was done, or die!

We've gotten a lot of mail and sales calls from people thinking we're a tool & die shop, or see our name spelled "Software Tool & Dye".

So, why STD, Inc? The story goes that when our lawyer brought our incorporation papers in for filing the Massachusetts Secretary of State himself, Michael J. Connelly, happened to be sitting in the front office and grabbed the papers to take a look. "Software?!" he grimaced, "Software Tool & Die"? What does that mean? There are too many software this and software that companies in Massachusetts. PICK ANOTHER NAME! he ordered. Our lawyer decided better to get this done today than quibble so asked Connolly if "STD, Inc." with "Software Tool & Die" as the d/b/a ("doing business as") was acceptable? Connolly said that would be fine, so we're STD, Inc. d/b/a Software Tool & Die.

During the summer of 1989 we bought a Sun4/280 server, a 72" rack with two whopping 474MB Fujitsu Eagle II disks – yes, a total of less than a gigabyte, a lot back then. This was more computer than we really needed to do our software work so the idea came to mind to offer email and usenet online discussion group services to the general public.

One of us drove over to a computer store, "The Bit Bucket", in West Newton, MA with a credit card in hand and bought six 2400bps Microcom modems, we ordered some phone lines, arranged to connect to other UUCP servers in the area, and called the new service "The World". This was all pre-internet (for us, the internet had existed for 20 years but academic/research only), or indirectly internet, or technically "the internet" but not really. We dialed another server and exchanged data a few times per hour.

Our brilliant marketing idea was to go to the local copy shop and photocopy a few hundred flyers to hand out at local colleges because back then some people who you'd find on college campuses knew about electronic mail and maybe even usenet. We got DOZENS of customers for around $20/month! Our original price list had different hourly charges for day, evening, and late night. We had nothing to model pricing on.

At a meeting we agreed that if it paid monthly for the loan we'd taken out to get the equipment, the monthly phone bill, and electricity we'd consider this a success.

If someone wanted a file from the real internet one of us would dial into an account somewhere, download the file, and put it somewhere on The World they could access it. Yes, manually!

At that time there were two commercial internet services on the planet: UUNET and PSI. They sold "high-speed" dedicated lines (often as fast as 1.5Mb/s!) and connectivity to the internet to companies who could afford the thousands of dollars in equipment and thousand dollars or more per month for the connectivity.

One of those companies, UUNET, was founded and run by a friend who called and asked if he could locate a rack of equipment in our offices in Brookline, MA to service their Boston area customers. We said sure. He asked what we'd charge them and the answer was "all the (internet) bits we could eat!" We wanted a direct connection to the internet we could resell to our customers. Probably since that had no further direct impact on cash flow for either of us the deal was made and in October 1989 The World became the first service on the planet offering internet access to the general public for a modest fee, around $20/month.

We registered the domain std.com and gave the service the host name world.std.com. Later we would acquire world.com and theworld.com. We sold world.com in the dot-com madness and now mostly use TheWorld.com though std.com host names still exist.

But there was trouble in paradise.

The internet, at that time mostly the NSFnet run by the United States National Science Foundation, was an academic and research facility. To be connected you had to either have a research grant from a major government agency or be an accredited university. There was also the MILNET, the U.S. military's part of the net. The internet had originally been a research project funded by the pentagon's Defense Advanced Research Projects Agency, DARPA, usually shortened to just ARPA, before being transferred to NSF and opened to the general academic research community this was all known as the ARPAnet.

The powers that be at NSF decided that selling access to individuals was a problem. Partly they were worried about security because as we all know the general public is a far different kettle of fish than a few million college students who by then had access, and there was this concern that re-selling government funded services might be just plain illegal.

NSF asked us what we would do if a customer "misbehaved" in some way? For example, sent threatening email to someone or who knows what else, spam wasn't really a concept yet. They asked us to think about that and get back to them because at least with a faculty or student they could refer them to some disciplinary committee but what could we do with the general public?

We thought about this problem for a while and finally responded: Not much! We could close their account but they could probably open another account without too much trouble.

Compare and contrast with today where the internet is known to be crime-free and a place of ready accountability for transgressions. Or maybe not.

So we talked about it with the powers-that-were at NSF. And they blocked us from about 2/3 of the internet.

This started a lively discussion on the network management mailing lists. We got hate mail from individuals who thought it was just wrong to take money to resell government facilities, very wrong! Our argument was that if we had a taxi service which took you to Yellowstone Park that wouldn't be reselling Yellowstone Park (a national park), it would just be selling our transport to and from the park.

We explained the situation to our customers and promised to keep them informed of any changes, but for the time being connectivity to other sites would be hit or miss. Since there was no other choice for most of them, we were the only such service, anywhere, they bided their time. We had people dialing in from all over the world, paying international long-distance rates just to be able to send and receive e-mail and similar. AOL, for example, existed at this time but did not yet have any connectivity to the internet. Neither did other closed e-mail (only able to send between customers) such as The Well, MCImail, Compuserve, Delphi, and various home-grown "bulletin board" systems. For anyone not affiliated with an officially sanctioned organization such as a university we were all there was.

One morning the phone rang, it was NSF. There was to be a meeting about the situation, could we send an email in the next hour formally requesting connectivity to the entire net? By the end of the day we had full connectivity and the rest is history.

I've edited out the names but here was the email from NSF giving us permission to let the general public on. I think it's safe to say that the last sentence of this letter is an amazing bit of internet history:

To: bzs@world.std.com (Barry Shein)
Subject: What's wrong with this picture... 

Barry -

As the several communities' (and our) understanding of what constitutes
"Acceptable Use" matures, some ancient inequities become glaringly evident.
Odd that you should write just now, as Software Tool & Die has been on my
mind of late as a prime example.

->However, my staff and myself would be happy to pro-actively educate
->our users about the guidelines of AUP and their responsibilities to
->abide by it. I would be glad to discuss how to best do this with you.
->I have in mind paper mailing all our customers a notice about the
->rules governing such access, creating an easily accessed file on-line
->re-stating these rules and practical clarification, and pointing them
->towards this information in the login banner displayed every time they
->sign onto the system. I have no problem with discontinuing service to
->people who flagrantly violate these rules or cause other problems,
->abiding by the rules of networks accessed is already contained in our
->standard contract with customers.

What's to discuss?  That sounds fine to me.  As a value-added reseller, your
obligation - which you understand and freely accept - is to educate your
clientele and deal with violations, and if violations occur they are not
yours but your customer's.  As a user yourself, in acquiring the material
you make available, you understand our Acceptable Use plicy and will I know
observe it.  (BTW, there was a slight change to the policy in June, so I've
attached below the new version.  And of course if H.R. 5344 [the "Boucher
Amendment"] becomes law there are likely to be further changes.)

So with all the foregoing as preamble, NSF is happy to permit carriage of
your Acceptable Use-conformant traffic over NSFNET Backbone Services.