Internet Publishing - PPI
"Materials used in this course are the property of the author. These
lessons may be used only by course participants for self-study purposes.
Application for permission to use these materials for other educational purposes
such as for teaching or as a basis for teaching should be directly submitted to
the author."
There is no exercise included in this lesson.
This is the last lesson in our course: Internet Publishing. The course deals with a topic which is extremely "hot" and fast developing. In the past year, tremendous developments have taken place on the web technology.
This lesson takes more the form of a lecture than a chapter in a textbook.
Perhaps this is because distance learning leads to lecturers being removed from
the classroom. We sit and talk to ourselves. Well, at any rate, it is useful for
me to write these things down. This was the first iteration.
Recently, we have seen that content and technology have
combined with each other to the point that it is difficult to separate them.
Earlier, we had information on one hand and on the other information
carriers. Take, for example, a newspaper. The contents are the words which
lie on the printed page, the information carrier is the paper the words are
printed upon. With Web and hypertext (and Java applets) these concepts have
changed. The reader (receiver) has greater control over what he reads, how he
reads it and in which order he reads it. No reader receives the same information
in the same manner.
In data technology we have had the same development with regard to hardware
and software. Earlier, there was a clear division between what was categorized
as software and hardware. Ever since Babbage's machine, where one had hole punch
cards with data and other hole punch cards with instructions, this separation
has been fundamental: hardware (gears and wires), software (instructions of the
type: ADD #3,1), and data (10010100). However, these divisions have now been
erased. A microprocessor uses microcode in order to process macro-instructions,
an operating system can be seen as a part of the machine, etc. With hypertext
and Web and Java the document will contain program code which allows for
simulation or completion of bank transactions or whatever one wishes.
The concept meaning is in process of being distributed. Mitchel Resnick writes in his amusing book about StarLogo and large parallel systems ([Resnick95], pp. 17-18)
Traditional theories of literature assumed that meaning was created by an author and conveyed through the author's writings. According to this view, reading is a search for inherent meaning in a document, an attempt to decipher the intension of the author. But modern schools of literatry criticism - such as poststructuralism, reader-response theory, and deconstructionism - adopt a very different stance. These movements all focus on the readers (not the authors) as the main constructors of meaning. In this new view, text has little or no inherent meaning. Rather, meanings are constantly reconstructed by communities of readers through their interaction with the text. Meaning itself has become decentralized.
(This knowledge/meaning point-of-view is the basis for the constructionist's pedagogic.)
What do we call information? We are often not very precise with this concept. In every day words, we say that information is something useful, something which keeps us informed. If we become informed, we become enlightened. Information is something good. We need information. However information theory, which helps us to measure information in bits and bytes, says that information is a measure of chaos. A disconnected gossip on the telephone contains more information than a logical and stringent lecture. If there is a structure in the text, it can namely be explained in a short manner: it is faster to explain a structure than show the structure.
A structured text can be "compressed" into a shorter text. One needs fewer bits to transfer a structured text than a text which contains chatter and nonsense. Therefore, says information theory, gossip contains more information that structured discourse. This is according to the information theory introduced by Claude Shannon in 1948. Shannon was an engineer at the american telephone company AT&T's development department Bell Laboratories. He studied how much information could be transferred over telephone lines. He was, in other words, not interested in what was said. On the telephone, you pay even if you don't say anything.
(Per's comment: ASCII codes, bitmaps and files are structured information. This means that it can be compressed - and that it is used with great effect in compression programs - zip, compress?).
Simultaneously with Shannon's work, a book by Norber Wiener called "Cybernetics" was published and created the basis for cybernetics (the study of steering mechanisms and processes). The information concept is not central in Wieners cybernetics, but it is interesting (and momentous) that Wiener's concept of information is diametrically different from Shannon's. Wiener said that a signal contains information when it has a certain structure and order, a concept which is comparable to our concept of meaning. When we see a structure, we become enlightened. This difference has confused researchers for almost 50 years.
Tor Nørretranders, who writes brilliantly of these things in his book Merk Verden [Nørretranders], has introduced a new concept exformation. Exformation is an expression for information which is taken away. And, it takes a lot of work to remove information. If I have two numbers, for example, 2 and 2, and I add them together, I get the sum 4 (2+2=4). I have used energy to calculate the sum. At the same time, I have cast out information about what the addends were. When I see the sum 4, I don't know if it is the result of 1+3, 2+2 or 3+2 or perhaps 0+4 or 7-3, etc.
In order to sum up, we - as technicians - ought to be aware the fact that information is an expression for chaos. The amount of meaning in a text rests more with the exformation in the text, that is how much is taken away, than with the information, what actually is there. I shall go into two short examples.
Winston Churchill was asked to hold a speech in front of a large audience. The arrangement was to begin shortly, so Churchill was a little unsure if he would have time to prepare himself. He asked how long he should speak. When he received the answer 10 minutes, he exclaimed "Impossible! If only you had said an hour!" The story, which sounds like something I read in Reader's Digest as a boy in my grandfather's cellar, illustrates the point that it takes much more work to say something sensible in 10 minutes, that it takes to speak for an entire hour.
The other example is from music. In classical music, exformation is used quite frequently. The pauses are just as important as when one plays. (This applies to all the small pauses in a given piece. Not the long pause where people go to the bar and drink white wine.) In pop music, exformation is also important. Things are suggested???. An effective catch is to let the theme be introduced right at the end of the song. The experience of the song then becomes dependent upon your having heard the song before. Stranglers are experts at this. (In much of the commercial heavy rock - Guns 'n Roses, Metallica - exformation is unknown. Here, one is assaulted with new tones the entire time.) A lot of music assumes that the listener does much of the job. The musicians play a theme, establish a rhythm, etc. When it dances around in the listeners head, the musicians can play something else which answers or is a contrast. (And here the hobby music theorists lays aside his pen.)
Today's (and tomorrow's) technology will make it possible for the reader to become more active towards the text than he is in relation to a book. The text is constructed when the reader reads it. Programs are put together and executed, images are generated. Technology and text melt together. From this perspective, it is not difficult to see why telephone companies, which earlier was information carriers, has now become distributors of content as well. Earlier, they transferred information, now they generate exformation in addition.
It's not difficult to see this today. What is so impressive is that someone saw this 10 years ago. Nicholas Negroponte established MIT Media Lab a decade ago. "... I had no doubt that computing and content would merge together into everyday life." (Wired, 3 Nov., p. 252)
The problem with today's Web is that it contains too much information and too little exformation. If you search for information on VB and find a web page with 200 pointers to new web pages, it's not very helpful unless additional information about the contents of the other pages is also provided. If you refer to other pages, make sure that you think enough about the references that the increased information is balanced against increased exformation. That is, get enough exformation onto your pages.
Since we humans don't have time to sit and scan information to find what we are looking for, we write small programs to do the job for us. These programs are called agents, and we will talk about these in the next section.
Today's Web has some the of the same problems as today's tv. You can sit an entire evening (or an entire night as it often is when speaking of the Internet) and because you receive so much information, you sit with the illusion that you are becoming enlightened. But, you aren't becoming enlightened, you are becoming informed. Here, the agents can come to our rescue. Agents are programs you send to complete tasks for you -- searching for information, shopping for CD's, books or stocks, etc.
BargainFinder is an agent process which is developed by Anderson Consulting which searches through directories of eight on-line music stores. It finds the shop which has your desired CD at the least expensive price. This is an attempt with agents to find out how the market will behave when the customers have perfect information. Studies show that the market reacts by closing out the agent. Read more and try it for yourself at http://bf.cstar.ac.com/bf/.(Wired, 3 Oct., p. 40)
Agents are not a new idea. In science fiction literature, we have long seen such programs. The main character in the Heechee Saga, Robinette Broadhead, [Pohl], has two expert systems. One is materialized as Albert Einstein and the other as Sigmund Freud. These two expert systems are capable of finding out most everything. They accomplish this for the most part with processors which they split and send around the world. In the latest books, Broadhead operates in cyberspace alone. He has converted his personality to digital form and exists only as a being in a computer. There are thousands and thousands of such people, and they travel around and talk with each other in simulated environments. If one wishes to be in two places at once, one clones a copy of himself. If a flesh and blood person wishes to enter cyberspace, he uploads a copy of himself. There is an ethical problem when these clones, or agents, are terminated at the completion of their assigned task. They are, of course, a copy of the original. Therefore, they are reluctant to come back. They don't wish to be terminated so quickly.
Today's agent technology has in no way experienced such problems. We want a butler, but all we get today is a dog. (Fetch!)
Encryption builds upon the theory that some functions are easier to run one way than another way. For example, it is much easier to multiply than divide. Division means namely that we must guess at what the answer will be, that is try and see if an answer fits. Similarly, it is much easier to multiply than to factor. If we are talking about large numbers, factoring quickly becomes impractical: impossible to complete within a reasonable amount of time. With "reasonable amount of time" we mean, for example, the life span of the universe.
The RSA algorithm was discovered by three Americans who therewith have made their initials immortal: Rivest, Shamir and Adleman. They published their discovery in February 1978 in Communications of the ACM. It is a very simple algorithm. With mathematical formula it can be written in 4 lines. To implement it in an effective (read quick) manner is, on the other hand, something else.
The same Adleman who was in on the RSA breakthrough is also the man behind the genetic computer, a search method based on DNA technique. (Wired, 3 Aug., p. 114). He distinguishes himself in both mathmatics/informatics and genetics. Isn't it strange that when we meet crossover musicians (Arve Tellefsen, Ole E. Antonsen are Norwegian examples) it seems a little foolish, but to be a crossover in science, that a scientist goes over to another subject area, leads to groundbreaking results?
Phil Zimmermann has created a very successful implementation of RSA (Wired, 2 Nov., p.129). Unfortunately, we still live in the aftermath of the cold war, and encryption technology is defined as military technology. It is, therefore, unlawful to export encryption technology to hostile countries (that is, all countries which are not the US). However, there are no laws banning free distribution of this technology. Zimmermann's implementation can, therefore, be downloaded via the Internet from a database in Switzerland. However, Zimmermann is continually in conflict with the police for this. Zimmermann's implementation is called Pretty Good Privacy which has been shortened to PGP.
In addition to the defense's stranglehold on PGP, Rivest, Shamir and Adleman have also received patents on parts of the RSA technology. This makes it take even longer for encryption to be developed for the Internet. These patents apply only in the US. In Europe, one cannot take out patents on everything. In the US, people have managed to take out patents on surgical procedures. If one takes up a knife and makes an incision in a certain order (a certain procedure), one has to pay royalties to the patent owner.
Encryption has had a somewhat slow start on the Internet. If one wants to buy something in the US, one must send his credit/debit card number. If one encrypts the message with PGP, one can be certain that any one that snatches the message will not be able to read it. However, one cannot be sure that the organization or person receiving the message will not misuse the number. But, then again, one cannot be certain about that when one pays with a card in a shop in Europe either.
With the help of encryption technology (formula which are easier to calculate one way than the other), one can create electronic money. This would provide a distributed payment system. Instead of paper money, you would have a bunch of numbers in your electronic wallet. You could use these numbers to make payments at a store. Perhaps you will get some other numbers back as change (change numbers). The system can also be made anonymous so that you won't leave electronic footprints in the bank of all your movements.
Today's automatic payment systems (In Norway: Bank Accept) is centralized systems. If the system goes down somewhere in the country, the entire system goes down across the nation. One Saturday in the fall of 1995, a fiber optic cable was cut in Western Norway. As a result, no one in Norway could use their debit cards for several hours. One doesn't need to be a visionary to see that this situation cannot continue. We have a terrible need for new technology, and that technology is called e-cash. It is what economists call a market pull, but since today's tight centralized solution is also pushing the need because the current technology as it is is useless, we also have a little bit of technology push.
With new technology, we will of course have new problems:
Mash together the idea of agents (automated programs which run searches for you) and electronic money, and you create potential for an entirely new type of crime. When you send the agent out to buy a new pair of slippers, you send along 30 pounds for the slippers plus a little extra for taxes. Some prankster comes along and writes a virus to rob your agent, knock it senseless and take your money. Perhaps the police will maintain an army of police agents which patrol the Net hunting for such viruses. Perhaps you will get your money back, but probably not.
(Tim Barkow, Wired, 3 Nov., p. 135)
What if my Visa card is being stolen? I will have to go to the bank to take out some cash. Physically, I am not especially handicapped by this situation. It's not difficult to withdraw money. However, in the cyberspace, I will be financially crippled. :-) I can't buy anything before I get my new Visa card. A physical event such as a break-in and theft has little or not impact in the physical world, however in the virtual world (cyberspace) I am completely penniless. This is a paradox for me. Market pull. "I want something else!"???
The International PGP Home Page, The University of Oslo:
http://www.ifi.uio.no/pgp/
First published: Mon., 4 December 1995 Last updated: Tue., 5 December 1995
"Overhauled"
by Per Borgesen (5 May 1996, 19 November 1997 and 30 May 1997)