Speaking the same language does not necessarily mean understanding each other, but it helps a lot. Artifical efforts like Esperanto are niche creations without the necessary wide spread for international communication.
English is far from being an easy to learn or perfect language, but it is easy enough to learn, and wide spread enough to serve for the time being. My native language is German, so I believe my judgement is much less biased than that of native English guys.
I use English to reach a broader audiance (read-iance?). I suggest my German fellows learn it, as I did.
I do not want to suggest German as universal language. It had and has too much non logical properties. Furthermore, it has been seriously screwed up in the 1990 to 2010 time frame by ill reforms, issued, drawn back, issued again with modifications, which in most cases replaced one non logic rule by a different one. Only very few changes were true simplifications, improvements.
If something has a built-in potential to fail, it may work for thousands of people flawlessly, but it will fail in my hands. SORRY. This is experience.
If I meet a glitch, I do not simply ignore it, but use to look at it, before it bites me again in an even worse situation. I ask why and what. Start research. Quite often I find not only a simple issue, as I truely hoped, but a whole root-network of issues, which may sum up even to deadly danger.
Perfection is what I am really missing these days. Perfection does not mean the dead end, but the serious try to do something the right way. Definitely NOT the try to make something different just to appear different.
No complex product will be perfect, at least not for everybody. I do not expect or demand that impossible level of perfection. I make suggestions for improvement, usually not only what is needed, but even how to achive that.
Minor faults, errors, malfunctions and issues are accepted, if and only if the Maker of the product plays an open game. Means the Maker tells potential users up-front of bad habits, missing features, especially issues you will not expect, because they are unusual.
I clearly prefer a product with known issues and limitations over the perfectly surface-polished thing I am not allowed to look under the hood before deciding for it. A representative telling me about issues makes a much better impression to me, because he demonstrates knowledge of the topic and his product.
Many products contain computer software these days. Quite often they are made in a way which allows update of that software. BUT often there will never be any update despite bad errors. The Maker got money in the moment the customer purchased. The product is outdated and end-of-life the same day it hits the Market, or even before. I will add examples later.
This seems to be a very special issue today. It is nothing more than a recipe, a description how to do something. DO NOT MAKE SUCH A HYPE ON IT. It has been with humans all the time. How to survive in jungle is software.
Computer software - the same expressed in a language executed by a computer.
I HATE the habit of a few publishers of free computer software, who seriously say "It does not cost anything, so do not expect any quality.". Fortunately, these bad guys are the minory. It is clearly a different story to publish something WITH source and the statement "This is a quick hack I made to achive that for me. Use with caution.". This is accepted and appreciated, because it may seriously help to solve your own issues.
Athough I am a software developer (beside other professions and hobbies), most of the time I am a user of software made by others, and in more general, products made by others. Regarding qualification, I can develop every software on my own. Regarding time, I CANNOT, no way. Besides that, co-operation is more fun than doing everything my own.
Experience: The more expensive a software is, the worse it is. Software in the above 5000 EUR class uses to contain that many errors and ill concepts it is pure waste of money.
Of course, for such a price you should expect high quality, a very good customer care, bug fixes within days. I never met that, but just the opposite, multiple times, connected with arrogancy.
I especially refer to eCAD software for schematic entry and PCB design.
Best experiences I made with Freeware and Shareware authors, and small companies, often consisting of a single person only. Bugs were fixed in a reasonable time frame, means from hours to a few weeks. Suggestions were taken and developed to reality.
I clearly prefer Open Source. It allows me to judge about inner workings with reasonable effort. If worst case appears, I can fix bugs myself, again with reasonable effort. Disassembling a larger software just to find and fix bugs is usually not worth the time, especially if an optimizing compiler for an object-oriented language did a good job obfuscating the real working, adding one more layer to the faulty screwed-up thoughts of the original author.
I pay for software I use, if the author requests payment, regardless whether it is termed Shareware or Commercial. I do not care whether I could use without payment. I use not-that-good software when I do not want to pay for the better one.
Wasting hardware resources by inefficient software has not been invented by Microsoft. I like to give credit to early UNIX programmers. Their machines had a lot of RAM for their time, and if that was insufficient, it was supported by a swap partition on mass storage. These guys created functions which needed to allocate memory to perform their task, but had no way to return an "insufficient memory" information.
Just a guess, carefully considering the facts. What happens when an artifical intelligence takes over control of the world? If the A.I. has bad intentions, it will probably wipe out human race. If it has the very best intentions, it will strongly influence (call it suppress) the human race at least, to stop their destructive work on environment.
General statements follow. Exceptions will just prove the rule.
A product works ...
In the 1996-1997 years, the owner-CEO of a mid-size Taiwanese electronics manufacturer said to my big-thinking that-time-boss during a business discussion: "I do not care whether you become rich by this product. I just want earn my small share of 50 cent from every single board I produce for you."
That's fair, isn't it?
Lessons not learned by a significant number of them: You have to support an interested school-boy, student or hobbyist well, because these are your future customers and employees. These persons often show you a quite different view onto an issue for little return, without asking for a payment, as consultants would do. I do not demand multiple hours, because this may be indeed unreasonable, but 1/2 to 2 hours per person per year is reasonable.
There are a lot of well-known brands which in the distant past worked hard to earn a very well reputation. Unfortunately, this has been.
Today a lot of brands are only managed as empty names. There is nobody developing a product, nobody manufacturing it, and nobody supporting customers employed at that company, under that name, any more. Unless you contact support first, asking good questions, you find out the hard way afterwards, when you meet a problem. Pay attention to the phone number you should call, or the email address you get the response from. If the Email address is not with the company's web site, this is a strong indication for external support. Email can be faked easily (let us term it re-directed), but some of the empty hulls do not even care for keeping that kind of good customer feeling.
Support "Customer Care" is done by external firms, which in most cases do not have the experience an in-house support had. Even if an external supporter has the best will to serve a customer well, he does not have an engineer next door who really developed the product and can be asked on difficult issues.
The past has not been good, but some things were clearly better. Many have been simpler, especially the simple things. The products had faults, too.
Software has been much smaller, in average. I am convinced: The number of errors per byte machine code stayed constant in years 1970 to 2011. There is 1 error inside 1 to 10 KByte, means 0.001 to 0.0001 errors per byte.
Electronics hardware has been less complex. You got schematics for it, to allow repair without re-engineering first. Electronics consisted of general purpose components, available in shops for many years. That time ended with wide-spread introduction of custom specific ASICs. This happened gradually over more than 10 years.
Product live cycle has been longer, although it was already too short in a lot of cases.
Most of them had 1 or more easy to operate electrical interfaces (ports, parallel or serial or Joystick), which allowed to perform simple tasks like lighting up a LED or reading the state of a mechanical switch without effort.
This is what I miss worst today. There are USB or Ethernet or Wireless, all quite complicated in electronics and software, really good for transferring larger amount of data the safe way, and comparable simple to use for that difficult task. But they fail completely on simple things. To do the simple thing you have to add e.g. a USB pheripheral having serial RS232 or parallel printer (Centronics) ports. Instead of just a single instruction accessing a pin, large protocol suites are running internally.
"Keep it simple, stupid!"? The first part is very well, does not require discussion. The second part needs correct interpretation. Treat it the positive sense, meaning "without superfluous intelligence". There is nothing worse than a product not working (as expected), being hard to understand because of superfluous integrated intelligence. A product lacking intelligence can usually safely be used by an intelligent human being, but it does not work the opposite way.
Example: "Intelligent" jet aircraft software prevented activation of reverse thrust after touch down for a significant time span because it did not sense sufficient pressure on landing gear. This was caused by front wind just starting to blow at that instant, generating extra lift the pilot had not expected. Wheel breaking was ineffective because of no pressure and wet runway. As far as I remember the report, no human being came to harm, but aircraft runned over end of runway. Of course, in-air activation of reverse thrust is deadly danger.
Most of the serious incidents have not been caused by a single failure alone, but by a linked chain of failures, ignorance, and errors. Indeed, e.g. aircraft or nuclear power plants are quite safe today, surviving single errors without harm. But much too often the first error is ignored, or not communicated to all who are in need to know, even the second error could not do serious damage, but after that the tolerance of technology had been exhausted, and the aircraft crashed, the nuclear power plant exploded.
I could give numerous examples, but because a lot of reports are freely available, I suggest you study on your own to really learn lessons before you carelessly kill people (although you always believed to be in a non-responsible position).
Real and true safety can not be achieved by any amount of restrictive technical or social means. The truely dedicated attacker will find a way to break and go thru everything you may place into his way, may even perceive it as a pleasant challenge.
Real and true safety bases on trust, openness, co-operation, culture, so called social behavior. It is impossible to do anything right to everybody, but try to avoid to give good reasons for hate. A good reason for hate is extrem imbalance of power and wealth, creating a feeling of injustice and having no legal way to change it.
There is no way to avoid dedicated attacks completely, but above works well to reduce their frequency to an acceptable value, to make them uncertain.
Use a low level of safety means to keep the honest honest.
An automated shooting system can be very efficient, but the smallest glitch may kill the owner, or even worse, some innocent unhappy by-standers.
Same applies to dogs, especially mentally damaged dogs, trained to kill.
Cited from an agent thriller: "Pull your gun only if you intend to kill!" Good statement. If you just intend to threat someone, you place yourself in real danger by showing him the technical means to damage you.
Makeing complex systems safe and secure does not really work. Complex systems are neither safe nor secure. This has been practically proven multiple times by breaking them. Keep systems simple and well thought, and you can achieve a high grade of security, safety, protection.
Not only a lawyer will tell you: The Internet is real. It is part of the real world. There is nothing virtual. You are liable, responsible, for your actions.
I do not want to point to any aspects of law here, but to environment and natural resources. Especially if you do not like to read it. Every single bit move or store needs energy. It matters whether you download X byte or 2 times X byte. It matters whether you view information stored on your local harddisk, SSD, USB-Stick, or whether you consult google every time and download it again and again over thousands of kilometers via Internet. (Assumed it is not that frequently accessed by others, too, that it sits already in a cache near you, which would reduce the negative impact, but not cancel it.)
Mankind can easily build a new Internet, but not a new environment. This is true for year 2011 and probably a lot of years to come.
It matters whether a website is lean and mean, or polluted by an excessive amount of pictures, sounds, visual styles, videos, and even worse, advertisements of any kind. There are good applications for every one of these elements. Think first, then decide what to use. A fact easily described in a few words does not need a picture or a video. A video could be the best description for a complicated move.
Unfortunately, the large number of users, especially users not experienced in technical issues, attracted a large number of criminals, and ab-users which are still inside the limits of law, but hurt sometimes more than the outlawed.
Trust vanished more and more. I still remember using real names and email addresses in News postings. Usually I still decide to use my real name, but give my email only in very selected places.
Some people talk about "Identity theft". I cannot agree, because your identity has not been stolen in most cases - you still have it - your copy of it. It has been misused, may have been copied for misuse. There is a difference.
Personal (or individual) freedom must have a clear limit where it cuts the same freedom of someone else.
They have little value on their own, may even be highly damaging.
Real friends, physically present when needed, are of high value. A total stranger helping you, just being in need of help, without expecting any immediate return, this is highest value, true social behavior.
Medium level amusement films can carry a message. Do you remember "Colt Sievers"? (German title: "Ein Colt für alle Fälle") Message: "I do not expect a return on my service for you now, but sometime in future I may ask you to do me a favour." Not bad, but not best.
I did tow others with my equal-sized car for no direct return multiple times, e.g. out of slam (dear girl, think before you drive a way like this in this weather, or take a shovel with you, to dig your car out yourself until you are trained enough to keep your car on the slippery way), out of narrow construction side on Autobahn, from highly frequented traffic lane, (dear student girl, I understand you do not have the money for a better maintained car, but at least you should have invested in a towing rope after your previous car failure). I risked damaging my car by trusting them to do a reasonable job commanding a rope-towed car, what they never did before. I clearly expect a return, and sometimes even tell: Help the next time you meet someone needing your help.
How to predict what happens in future? Study the past, the predictions made, and their wrong (or right) being. This is the way to learn the limits of extrapolation and educated guessing.
Right sizing is a key for survival of mankind. It means to use a suitable tool for a task.
--- Author: Harun Scheutzow ------ Last change: 2011-07-11 ---