A Distributed Library

To make sure not to mislead my readers for too long, in particular the programmers among them, I insist in specifying that the topic is a library made of shelves and books (remember those bunches of paper?). The idea of this essay is to explain, mostly for my own benefit but maybe for that of others, one of the interests of collecting books. Personally, I love books and, although appreciate wikipedia for what it is, I think no electronic medium can overcome the joy of possessing an extensive book collection. Once in a while, I go on a book buying spree and, although some of them have been sitting on my shelves unread, I think this habit of mine is very useful. For one thing, I read fairly slowly and, sometimes, I like to put a book back in the shelf without finishing it. For those reason, if the book is good enough, it is very beneficial to own it. On the other hand, I like to see myself as a node in a distributed library. Whenever I invite people over at my place, one of the things he or she will see is my collection of books and, sometimes, they get interested in one or two of them and bring them home to see if they like it or if they can learn something from it. It can be viewed as a means of propaganda but also as a means for education. Indeed, I think it is very convenient to be able to educate each other in an informal way and being part of a distributed library seems like an efficient way to do so. Simon Hudon July 12th, 2009 Zürich

Computer Aided ...

I made no secret that I consider E.W. Dijkstra a very wise thinker and I take many of my ideas from him. One of the things I've been hesitant to accept from him at first is his abhorrence of the use computers. It's been at least 4 months maybe as much as six months since it dawned on me that I am indeed addicted to the use of computers. There's no other ways to talk about the excessive reliance on electronic communication (e.g. looking for emails more than once an hour), my habit of browsing meaningless content on the internet and, more importantly, my habit of relying on software tools when they ofter arguably little improvements on productivity and probably some awful deterioration also. I just started reading the book "In Praise of Slow" by Carl Honoré and went through the section where he describes how in capitalist societies we made of time a new god. In other words time is more in control of our life than we are. It makes for some very nice reflections but for now, I'll come back to my main topic. In a similar way, the computer has come to possess control over our work. I can't count anymore how many times a day I get annoyed and frustrated with how my computer is working against me. It goes from simple trivia like message boxes popping up while I am typing something and forcing me to respond immediately to the use of theorem provers that relays my job to a secondary place by taking the reins passing to the use of IDE that do not work and word processors that pretend to know better than me what I want to do. In some cases, this is plain condescending but in every cases is completely harmfu to productivityl and destroys my motivation. Recently, I bought a new tool that I am pretty satisfied with. As a word processor, it does not argue with me. As a theorem prover, it allows me to keep the reins and go where I want to go. As an language editor, it does not crash and does not ship with faulty libraries. Finally, it very rarely interrupts me in my work. I bought a fountain pen and, although I still use the computer in a pretty addicted fashion, I see a change happening and I can't see how it wouldn't be for the best. In any case, the feeling is one of pleasure and satisfaction. And now, for those who wonder when it interrupts me: it's when my cartridge of ink is empty and I store two in my pen so the replacement is a fairly short operation. I am now resolute to using computer as much as possible for communicating with my loved ones which are now very far away and to question every other necessity I feel of using a computer. This is far from done but I expect some nice results. As an aside, as a software developer I think I draw an important lesson. In the same way you should never be condescending toward your readership, you should not be either towards your users. If they use the computer, they should know what they want and you should not take the control from them. For example, when forms have to be filled, you should allow the user as much as possible to fill them in one go and possibly even in the order of his choice. A good example of a violation of this principle is the setup of Windows where installation sequences are interspersed with questions so the user has to sit through the whole process. Simon Hudon 12:55 AM on July 12th Zürich during a sleepless night

On Programming Languages

Sometimes, some programming languages popularize powerful ways of thinking about a problem. It is not too rare however that a language would create popular way of obfuscating a problem. APL and C++ are an example of the former. The difference between the two is that C++ allow you to believe you actually understand the problem when you don't have a clue about it. On the other hand, I have never met anybody knowing APL that claimed to be able to read an APL program. Simon Hudon July 10th 2009 Zürich

Nice Quote and Comments about User-Friendliness

"Build a system that even a fool can use, and only a fool will want to use it." I have encountered very often the notion that user-friendliness is the quality of being intuitive or, put in Dijkstra's words, the quality of "appealing to the uneducated". It seems that Microsoft can be very active in that respect. Take Word for example, it is sold with the pretense that you don't have to learn to use it but, invariably, you'll have to get accustomed to it and get an intuition of the heuristics used to justify one or another behavior. In other word, because there is the clear criteria guiding the behavior of the system that can be written down, it is believed that you don't need to understand anything to use it. If user-friendliness is to have a non condescending meaning, I would associate it with the simplicity of the design of the system and of its interface. It is acceptable to have to learn how to use a system but the description should be as short and as precise as possible. To those who believe those to be contradictory qualities, I refer to the manual of the Algol language and conclude with a quote by Dijkstra. "About the use of [natural] language: it is vain to try to sharpen a pencil with a blunt axe. It is equally vain to try to sharpen it with ten blunt axe". E.W. Dijkstra Simon Hudon July 10th 2009 Zürich