Copyright, Roy Brander, P.Eng. 1995. Permission to reproduce is given so long as the reference is credited.
This makes a nice point to segue into my being asked to address the specific usefulness of the Internet for civil engineers. What is it that most of us do for a living, if not to take in information of various sorts, from specs and maps and descriptions, then perform calculations and make judgements, and send back more information in the form of designs, drawings, budgets, instructions, reports, and recommendations? How much of our work could we do by E-mail when it connects to everybody and can include drawings, calcs, spreadsheets, and pictures? And what would be the total dollar figure saved by society if even 20% of business travel were replaced by electronic communications that cost less than a buck apiece? So I'm willing to give that "absurd" 1000-fold improvement in usefulness the benefit of the doubt for now. Enough with the blue-sky futures; let's have a look at what the Internet can provide to civil engineers right now. First, let's survey what kinds of basic services there are.
(Graphic 9 - Most popular Internet services)
First, of course, is E-mail, which most people already understand. It is a direct communication from the sender to an arbitrarily long list of receivers. Modern E-mail packages already allow the attachment of any kind of file, from spreadsheets to databases, sound, and pictures.
Second, a combination of E-mail and special computer programs gives us the mailing list. Suppose a thousand people worldwide want to stay in touch on the topic of wooden footbridge design. One enthusiast sets up a mailing list "server" program on his Internet node. The program has its own E-mail address. You send it an E-mail with the word "SUSCRIBE FOOTBRIDGE" anywhere in the message, and it adds you to the list, and automatically sends you a reply giving you the E-mail address of the footbridge mailing list. From then on, any message you send to that address is automatically copied out to all the other subscribers, and anything any of them sends is copied to you. Voila! Instant discussion group.
Mailing lists proved so popular that some of them had subscribers on virtually every Internet node. It made more sense to just send everything on those lists to every node so that all the users there could just read the one copy.
(Graphic 10 - USENET Group Selection)
Thus, USENET was born. USENET is the Internet version of a computer bulletin board. If you have an idea for a USENET newsgroup, and post it to the group that fields such suggestions, and the E-mail response is good - then literally a million or more Internet hosts that provide news server programs will begin recognizing messages posted to it and the discussions will be publicly visible to any Internet subscriber that cares to look. Here, I'm showing a few randomly selected pages of USENET newsgroup names. There are over 3,000 available on the Calgary servers - some enormously popular and read by millions; some by only a few thousand worldwide. This is all handled at your end by a special news reader program, and at your Internet provider's machine by a news server program. Both grew out of mailing lists but offer much more sophisticated organization of the news "posts". The page-full of newsgroup names here is a wide cross-section. "Comp" newsgroups for technical computer discussions, "Rec" groups for recreational topics, a "soc" newsgroup for arguments about social issues, and a few "alt" newsgroups for the "alternative" -frankly, that means oddball - topics. I simply stumbled across this group on urban transit; I don't know why it is under "misc" instead of "sci" for science, like "sci.engr.civil" that I'll be showing you later.
(Graphic 11 - Threads in misc.transport.urban-transit)
Shown here are a list of topics visible one day on the urban transit newsgroup. Each topic may have one or more messages posted. The number down the left shows how many replies there are to the original. I moved my cursor down to the one on "Electric Vehicles" and hit ENTER on that.
(Graphic 12 - Posts in the Electric Vehicles thread)
That jumped me to the current posting on the topic. This one comes from one Mark Shaw of Motorola, whose E-mail address is shown at the upper left. The text makes it clear he's replying to a post from Will Stewart, from which he has excerpted some text as shown by the greater-than signs at left. You don't have to do this, but we all try to follow such conventions for clarity, and the newsreader software makes it easy to do.
The other reason that people stay late at night doing this, is that if you start following USENET groups on topics of both personal and professional interest, and worse yet, begin taking part in several discussions that run over days each and involve a few dozen like-minded, articulate people from across your profession - it can start swallowing up your evenings! Some say that people who get absorbed this way need a life; I agree but if your previous life was spending just as much time in front of the TV, then this is at least some improvement. (Graphic 9 again for a moment)
Fourthly, you can use the Internet to log-in to any computer on the Internet with a terminal session, just as you could directly over the phone if you didn't mind the long-distance bills. If you aren't already involved in computers, you might not think this one useful.
(Graphic 13- U of C login screen)
But most Universities have now provided their card catalog programs as free logins to the Internet, so you can use this one program, TELNET, to search library catalogs clear across the continent, and get the book in a week by inter-library loan. Just keep up that alumni library card!
(Graphic 14 - list of libraries from Canadian Internet Handbook)
Sorry, again this is copyrighted and I can't show it. But the
two libraries on page 672 of the 1995 Handbook are of interest
to civil engineers and they are:
it a good example are:
CANMET library catalog
(Log in as "opac")
Instructions for subscribing to the
CISTI library
Besides virtually every University library on the continent, there are some others. I have here a page from the excellent book, "The Canadian Internet Handbook". They have over 20 more such pages listing library names, locations, and contents. It mentions such other libraries as the CANMET centre, and the Institute for Scientific and Technical Information. Naturally, there's a few days wait for delivery of an Inter-library loan; but, hey, before the Internet, it took a series of long-distance calls to simply find out which library had a specific book. Now, you have the library home-shopping channel. (Back to graphic 9)
Fifthly, there are various services with which I am going to do an arm-wave, and lump under the heading of "file transfers". These are the ones that let you be the library - or browse the stacks of anyone else who has decided to do the same. They simply dump whatever files they wish the world to share into a special directory they make accessible to a file-server program. It simply waits for an Internet request to come in, and automatically sends them the file. Although the user interface has improved enormously, from typing near-gibberish computer commands to today's "Internet World Wide Web Surfing" that resembles an easy video game or just TV channel zapping, all of them are really just ways to move files from a server computer to yours.
(Graphic 15 - Internet File Transfer Protocols)
The granddaddy of them all is called "UUCP", for "Unix-to-Unix Copy". It would copy a file from one Unix computer to another, but it called for a fast, perfect typist and I don't think many people miss it. I mention it mostly because it is still in use as the way those newsgroup postings are sent around. Fortunately, only the machines have to type the commands.
One that still gets a lot of use is "FTP" for "File Transfer Protocol". An FTP session is a kind of little log-in from your machine to the other one. You can list out file names, change directories, ask it to send you files, or send some of yours the other way. But without the helper programs I'll be showing you in a moment, it's still about as much fun to use as DOS on a PC - which isn't saying much. The example I have here shows what you'd type in boldface. At the end of the session, you've been sent the file "read.me" from the "pub" directory.
(Graphic 16 - Top Gopher page)
A few years ago, students at the University of Minnesota -home of the "Golden Gophers" football team- decided to make this a point-and-click effort, while keeping the file listing and selecting to an interface that didn't need a graphical screen. They invented "gopher" which now runs on tens of thousands of Internet sites.
To make files of any sort available to the world from your gopher site, you just create a file in each directory that describes all the entries in it, preferably with one line apiece. This is shown to the gopher program user who visits your site as a menu. If an entry is a file, then hitting on the entry shows the file. It can also be another directory. You work your way down through the directories to find what you want.
This is the gopher page offered by the Voice of America international news service. They continue to use gopher because it doesn't require a graphical screen, and many of their customers are in poorer countries where a 1981 IBM PC is as high-tech as it gets. Since the Internet concept continues to work - just more slowly - even over bad phone lines with old modems, and costs just a penny per page or less for text transmission, its use is growing at that same exponential rate even in some pretty poor corners of the world. On the VOA gopher, I move my cursor down to the news script item and hit enter.
(Graphic 17 - Gopher 2nd page)
That takes me to a list of the last week's broadcasts. Every day at midnight Greenwich, VOA throws out the week-old broadcast, and starts filling up that day's directory. I read this on a Wednesday, but remembered it was already Thursday GMT and avoided reading six-day old news.
(Graphic 18 - Gopher news story list)
Hitting ENTER on Thursday go into that directory. My cute Windows version of gopher is no longer showing folders beside the menu items. I am at last down to the files. Each one is a news story. VOA switches to all-capitals at this point, to make sure everyone can read them, even people with Commodore 64's that don't do small letters.
(Graphic 19 - Gopher news story)
Hitting ENTER on that one, I've at last done a true file download, and can read the story itself. They are short, as they are the scripts actually read over the air. VOA keeps right up to the minute - you can often read stories here that went out less than an hour before. Working your way back through the menus, you can also look back for previous stories on a topic like a battle for a town in Bosnia over the previous week. By comparison it's hard to see yesterday's newscast or read yesterday's paper, unless you have a subscription or head down to the library.
(Graphic 20 - NIST gopher page)
I did find one gopher site with some relation to our profession - the National Institute for Standards and Testing provides a site where you can read over their press releases that go over a year back. This page shows one of their menus, and an overlapping window with one of the stories from a few menus down. Every time you go down a level, you get a new window, which makes it easier to move back up the tree. The Windows software, incidentally, is cheap shareware. Most Internet software is either free or under fifty dollars, as most of it was developed in those same Universities that were the early beneficiaries.
Gopher was the first program to make looking around Internet sites an easy, point-and-shoot experience with no command typing. However, it is already rapidly being eclipsed by an even more sophisticated interface that has simply exploded into public interest over the last year and a half: the World Wide Web.
(Graphic 21 - HTML example) C'mon, why should I provide that? Just do a source on *this* file for an example!
With gopher, everything is a menu, until you get down to the bottom of the hierarchy and are shown a file. With the Web, everything is a document. Web documents are written in a language called HTML, for HyperText Markup Language, of which I have an example shown here. In part, HTML just does what word processors do, although the files are kept simple and human readable. These marks, called "tags", between the angle brackets, indicate how portions of the text should be interpreted. The "B" and "/B" tags incdicate boldface. "I" and "/I" would mean italics. Other tags designated headers, tables, paragraph breaks, and so forth. Graphics can be inserted into Web documents with an image tag, that indicates the name of the graphics file to insert.
Where the "hypertext" comes in is a special tag that encloses text that is an anchor, or linking spot, to another file. Click on that text, and the file is loaded in. The kicker is a file name convention that the whole Internet has adopted, which allows the designation of an individual file anywhere on the Internet, and how to get it. The first part indicates what file transfer program to use: news, ftp, gopher, or more hypertext. The second part gives the machine name, anywhere on the Internet. The last part gives you the directory and file name on that machine. In sum, that means that any word or image in a Web document can link to any file in the entire world of the Internet. Some have even claimed that the entire thing can now be regarded as one huge interlinked document. They really aren't kidding when they call it the World Wide Web.
The World Wide Web concept was invented by scientists at the European particle physics laboratory, CERN, around 1989-1990. They were mainly looking for a way to link references in one paper or document to another even if they were on different machines. By 1992 they improved their original text-only Web browser to include graphics.
The young programmers who really made it take off, however, were the inventors of the Mosiac program, funded by the National Centre for Supercomputing Applications, the NCSA. Mosaic was a universal program for reading text files, graphics, fetching FTP documents, simulating a gopher reader, and a World-Wide Web browser. These examples show how those Internet addresses work that you are now seeing mentioned in news stories and advertisements. The first word says what type of file transfer program to use - though Mosaic or one of its competitors like Netscape can do them all: ftp, gopher, or http, the "Hypertext Transfer Protocol" used by the Web. The next part specifies the server name, that is, the Internet name of the computer. The last part is the path, or the directory and file name on that computer.
With a Web browswer program like Mosaic or Netscape, and an Internet connection, you can sit at your PC and browse through any file available on any FTP, gopher, or Web server, or in the USENET newsgroups, anywhere in the world. After Mosaic was released in 1993, Metcalfe's Law hit the people with things they wanted the world to see like a ton of bricks: by using an Internet file server, you suddenly could gain an audience in the tens of millions. And the more stuff there was to see on the Web, the bigger the audience got. I spoke earlier of geometric expansions and doubling times. In that regard, the number of files accessible with a World-Wide Web browser holds the hands-down record in the entire history of technology: for over two years now it has been doubling in size every 53 days. I was asked to start writing up this talk in early September. Since I started work, the amount of material for my research base has doubled.
Now that there are commercial programs for both Windows and Macintosh that run a Web server in the background while you use your PC, and Internet service providers will let you run a phone line 24 hours a day on the Internet for $125 per month, it is affordable for every little Mom & Pop business to set up a Web site to advertise their catalog of products or services to the world. There is every reason to think it will go on doubling at that rate for another year or more before it slows down to the same rate as the Internet itself.
Next part of the speech, "Internet Materials for Civil Engineers"
