What happens if physicists gain control of computer languages
Published by admin January 26th, 2008 in Doctors of PhilosophyIn Galileo’s time Doctors of Philosophy controlled secret algorithm or database and claimed to possess a secret language to access their secret database and marketed their “knowledge” as absolute true knowledge. Galileo removed Doctors’ authority and transferred it to geometry. Eventually, Doctors corrupted geometry too and established their authority on the algorithm. Today physics is a corrupt legal code with no errors.
What would happen if Doctors of Philosophy known as physicists took control of computer languages? This is not so far fetched an idea, even though it is scary. Doctors already control our natural language and freely corrupt it to gain more authority. If Doctors of Philosophy also obtained the authority to control computer languages they would corrupt them by removing any rules and errors they don’t like and bring all the world’s computers into chaos.
ringtone lie a perfectmmf index ringtone 2006i730 24 ringtone nextel for7337 ringtonesfree link alltel ringtone ringtone3585i ringtone alltel4 blondes ringtones nonphone 24 ringtones cell ctu Map
What the computer scientists do to computer languages is already bad enough. I doubt that the physicists could do worse.
The two industries are similar in a certain way, to dentistry or doctory, or the legal profession. They use control over the industry to raise barriers to the entry of competition thereby improving the value of their position. In mathematics and physics this is done by adding unnecessary complexity. The legal profession writes the tests that allow people to become lawyers, similarly for the medical fields.
In the case of the computer programmers, the tendency for a language is to continually add new featuers to it. By the time a decade has gone by, the elegant and easy to learn language has been morphed into a monster that is almost impossible for students to master. The previous generation, which grew into the various small changes, has no trouble keeping up. But the efficiency of the language is damaged and the ability of even the elders to use it is harmed.
This same process also happens in the engineering profession, and is well exampled in the evolution of the languages I am so familiar with there, Verilog and VHDL, which are now about as complicated as the US tax code (designed by accountants to make work for accountants) or legal code.
What the computer scientists do to computer languages is already bad enough. I doubt that the physicists could do worse.
I don’t know. When I read about designers of computer languages they appear to be very sensible people. I like to read the writings of Paul Graham, http://www.paulgraham.com/arc.html, for instance.
. . . . which are now about as complicated as the US tax code (designed by accountants to make work for accountants) or legal code.
I agree with everything you say here. Physics and mathematics also get more complex because physicists make their living by teaching. The more things to teach more money to be made. I was also thinking how for instance particle physics looks like it was designed by accountants for accounts.
But I am not exactly certain that physics is analogous to a computer language. It does not have the same rigor.
Pioneer1,
Yes, the people who design the original language are very good. Where the problem comes in is in what happens to the language over time. An example of this is Fortran. A link to a person who thinks that the changes are good, and I agree that some of the changes are improvements. However, what he doesn’t tell you is what happened to the Fortran manual. It ballooned from being a one page document to being a 300 page monstrosity. And his example shows the same program getting longer and longer, and using more and more concepts.
The problems started happening soon after Fortran 66. At that time, a professor could expect most of his students to pass a class teaching introductory programming in Fortran. As of now, class drop out rates of 75% are not at all uncommon. Of course the industry professes to be worried about the dropout rate, but it is politically incorrect to blame themselves for making what was once a very simple language into a monstrosity that only grad students and professors can use, LOL.
The reason I know what happened in Fortran is because I programmed with it in the old days, and then had a friend try to pass a modern Fortran class. It was amazing. They added all the features from C and Pascal to Fortran while leaving the old elements behind. The result was a monster that promoted the most horrid programming style seen, a mixture of misplaced intentions.
The same thing goes on in the programming language I use now, Java. Java itself is quite simple. Microsoft has been steadily complicating it with the (necessary) graphics utilities so it now requires a book several inches thick to describe the things that a programmer must be familiar with.
In the “hardware design languages”, which are used by electronics engineers to describe chips, they really outdid themselves with a language called “VHDL”. It was so complicated that it is almost impossible to use, but it is very general and technically sophisticated so that the idiots who created it were certain that people would be very productive with it.
However, it turns out that VHDL destroys productivity to such an extent that the primary advantage the US has over Europe is that Verilog is used here instead. The issue came to a head when the idiots arranged for a contest to see which language was superior. The result was a hilarious semi-academic fight.
The lesson is simple. If you want quick and correct coding, use a simple language where it is clear exactly and precisely what each element intends, and there are only a few elements to master. If you put every kitchen sink into a language (which is inevitably what happens as time goes on, you maintain backwards compatibility for a reason), the result will be a monstrosity that only the users who have grown with the program will be able to master.
However, what he doesn’t tell you is what happened to the Fortran manual. It ballooned from being a one page document to being a 300 page monstrosity.
This is what happens in physics. Einstein’s writings on General Relativity do not exceed 100 pages. In 50 years the same 100 pages balooned into the 1000 pages of MTW.
“What the computer scientists do to computer languages is already bad enough. I doubt that the physicists could do worse.”
Most of the people who do those things to computer languages aren’t the computer scientists. It’s the people like Stroustrup who laugh at all that “academic fiddle-diddle” that ruin a language.
Stroustrup is a computer scientist, No?
Paul Graham has an essay now talking about his new language Arc and he discusses the issue about the problem of change in a language.