April 13, 2013 at 20:27The three most hated things in computers
All of the following applies not only and not so much to experienced users, but expresses the generalized reasons for the hostility to modern computers (that is, computer interfaces , since the appearance of external computers does not cause any particular complaints) on the part of low-experienced users who do not want to spend a lot of time gaining competence .
So, the three most hated things for users: changes, surprises, a break in context .
In all my considerations, I will be based on the concept of “an interface for solving a problem,” because the only reason why a user uses a computer is because he can solve a particular problem, which I will call the “interface”.
As an alternative, I will compare several other interfaces: car, piano, typewriter.
Computers allow you to change the interface (I remind you, it's about the interface for solving the problem) very quickly and easily. The appearance of new functions puts the programmer / developer of the / etc interface in front of the problem of logical reordering of elements. For example, there were three functions in the top level, now there are seven of them, let's divide them into two logical groups (3 and 4) - everything is ok, no chaos.
From the user's point of view: previously there were three logical functions, now they have disappeared somewhere, two menus appeared in their place, with one action in the first menu, and the remaining two in the second. Hell, horror, doom, damn the programmer.
Why? Because when a person solves a problem together with a computer, changing the interface requires distracting from solving the problem and beginning to re-learn motor reflexes, instincts, behavior patterns, etc. In other words, pushes the user from the position of an experienced user to the position of "beginner user". Reducing performance, sometimes leaving with an unresolved problem (although it used to be well-solved!).
Let's see how this problem is solved with three mechanical examples. Change of driving occurs with the car. And it causes a person quite significant problems - even transfer to another car - at least get used to the brakes. And changing a machine to mechanics is a whole adventure. However, as well as back. The solution to the problem of changing the car interface is a clear understanding by the user of the volume of what he has to retrain.
A similar change on the piano - as far as I remember, the maximum that can be seen is the presence / absence of a moderator, the third pedal (compared to two). Everything else has been unchanged for centuries.
The typewriter is the most volatile of, especially in the area of auxiliary nodes. But even there, all changes are local and well-understood. “On the last typewriter, I had a tab button, but instead of it a group of buttons T1-T4 on top.” But the main functionality remains the same again.
Driving a car brings the maximum number of surprises compared to all other activities. But all these surprises are in the context of the task. Roughly speaking, a dog jumping onto the road is a surprise in the context of the driving task. But the sudden transition from front to rear-wheel drive while driving at speed - this is from the category, fortunately, fiction that has not come true. For the other two tasks, the situation is almost similar, it is also unlikely to encounter an unexpected change in the problem.
But in all modern interfaces - easily. I will not give ridiculous and boring examples, I think everyone understands that any modern application can do unexpected things. The disappearance of the menu, “started to spoil everything”, “strange signs appeared”, etc.
And all this actually means one thing: breaking the context. Suddenly, a person is pushed out of his local and cozy problem into a harsh, raging sea of abstractions, concepts, models and principles of construction, which he does not understand at all and has no resources to study.
There are so many levels of abstraction in modern computers that no one can remember and know them all. Their change, unexpected and falling out on the user, is always a break in context. Instead of his task, the user begins (forced) to think about local problems associated with something insanely distant.
In cars, this happens, by the way. And suddenly smoke started and the car stood up. A person breaks away from the process of driving and must find out the reasons (and urgently act). And, suddenly, instead of starting up, sneezes and makes strange noises. A bunch of lights lit up and blinked. Etc.
But with regard to automobiles, it is clearly known that these are all signs of poor quality (or terrible operation), that is, such breaks in the context are clearly interpreted as emergency. For cars, it is considered a sign of high quality not to do such things for the user. Likewise for a typewriter behind hooked letters, or for a piano with sliding strings tuning.
Why are context breaks considered perfectly normal for computers? Rearrange buttons in a dialog? How nefig do. Shove everything into five menus and require pointing to the menu before showing it? How nefig do, and still be proud of it. Enable mega super view mode? Elementary.
Ask a question (or just show a message) in the middle of work? It’s easy. The most striking example for me is the appearance of a message that only 15% of the battery remains on top of the navigation application, moreover, with the screen automatically turning off a few tens of seconds after the message appears and requiring you to enter a password to unlock. This is under the condition that the driver’s hands on the steering wheel should be, generally speaking, not entered a pin code.
And then, suddenly, I want to remember about the command line interface. Do you know why it is comfortable? Because there are almost no surprises. There may be problems. The situation “I don’t know how” can be. The situation "too difficult to do" - may be. However, they are all within context. Without his breaks.
coreutils (and their predecessors), that is, the ls, cp, mv, cat, xargs, head, less, etc commands, have been around for several decades. You can sit down at the console and have an interface that is immutable, that does exactly what it has been doing for the past 20 years, and that does exactly what the user will say. (yes, I understand that you can break down a terminal with evil console_codes, and forgotten quotes in can be very stressful, and Ctrl-Z / C does not always work ...). But compared to any graphical interface, the command line has only one plus - it does not change. And does not change its behavior on the next restyling of everything and everything. For this, she is loved by professionals who are less likely to fall into the "inexperienced user" mode and which allows improving motor skills without a "new level, all skills are reset."
I wonder how many years must pass before such a simple thing as a fixed interface appears in graphical interfaces as well.
So, the three most hated things for users: changes, surprises, a break in context .
In all my considerations, I will be based on the concept of “an interface for solving a problem,” because the only reason why a user uses a computer is because he can solve a particular problem, which I will call the “interface”.
As an alternative, I will compare several other interfaces: car, piano, typewriter.
Changes
Computers allow you to change the interface (I remind you, it's about the interface for solving the problem) very quickly and easily. The appearance of new functions puts the programmer / developer of the / etc interface in front of the problem of logical reordering of elements. For example, there were three functions in the top level, now there are seven of them, let's divide them into two logical groups (3 and 4) - everything is ok, no chaos.
From the user's point of view: previously there were three logical functions, now they have disappeared somewhere, two menus appeared in their place, with one action in the first menu, and the remaining two in the second. Hell, horror, doom, damn the programmer.
Why? Because when a person solves a problem together with a computer, changing the interface requires distracting from solving the problem and beginning to re-learn motor reflexes, instincts, behavior patterns, etc. In other words, pushes the user from the position of an experienced user to the position of "beginner user". Reducing performance, sometimes leaving with an unresolved problem (although it used to be well-solved!).
Let's see how this problem is solved with three mechanical examples. Change of driving occurs with the car. And it causes a person quite significant problems - even transfer to another car - at least get used to the brakes. And changing a machine to mechanics is a whole adventure. However, as well as back. The solution to the problem of changing the car interface is a clear understanding by the user of the volume of what he has to retrain.
A similar change on the piano - as far as I remember, the maximum that can be seen is the presence / absence of a moderator, the third pedal (compared to two). Everything else has been unchanged for centuries.
The typewriter is the most volatile of, especially in the area of auxiliary nodes. But even there, all changes are local and well-understood. “On the last typewriter, I had a tab button, but instead of it a group of buttons T1-T4 on top.” But the main functionality remains the same again.
Surprises
Driving a car brings the maximum number of surprises compared to all other activities. But all these surprises are in the context of the task. Roughly speaking, a dog jumping onto the road is a surprise in the context of the driving task. But the sudden transition from front to rear-wheel drive while driving at speed - this is from the category, fortunately, fiction that has not come true. For the other two tasks, the situation is almost similar, it is also unlikely to encounter an unexpected change in the problem.
But in all modern interfaces - easily. I will not give ridiculous and boring examples, I think everyone understands that any modern application can do unexpected things. The disappearance of the menu, “started to spoil everything”, “strange signs appeared”, etc.
Context break
And all this actually means one thing: breaking the context. Suddenly, a person is pushed out of his local and cozy problem into a harsh, raging sea of abstractions, concepts, models and principles of construction, which he does not understand at all and has no resources to study.
There are so many levels of abstraction in modern computers that no one can remember and know them all. Their change, unexpected and falling out on the user, is always a break in context. Instead of his task, the user begins (forced) to think about local problems associated with something insanely distant.
In cars, this happens, by the way. And suddenly smoke started and the car stood up. A person breaks away from the process of driving and must find out the reasons (and urgently act). And, suddenly, instead of starting up, sneezes and makes strange noises. A bunch of lights lit up and blinked. Etc.
But with regard to automobiles, it is clearly known that these are all signs of poor quality (or terrible operation), that is, such breaks in the context are clearly interpreted as emergency. For cars, it is considered a sign of high quality not to do such things for the user. Likewise for a typewriter behind hooked letters, or for a piano with sliding strings tuning.
Why are context breaks considered perfectly normal for computers? Rearrange buttons in a dialog? How nefig do. Shove everything into five menus and require pointing to the menu before showing it? How nefig do, and still be proud of it. Enable mega super view mode? Elementary.
Ask a question (or just show a message) in the middle of work? It’s easy. The most striking example for me is the appearance of a message that only 15% of the battery remains on top of the navigation application, moreover, with the screen automatically turning off a few tens of seconds after the message appears and requiring you to enter a password to unlock. This is under the condition that the driver’s hands on the steering wheel should be, generally speaking, not entered a pin code.
Command line
And then, suddenly, I want to remember about the command line interface. Do you know why it is comfortable? Because there are almost no surprises. There may be problems. The situation “I don’t know how” can be. The situation "too difficult to do" - may be. However, they are all within context. Without his breaks.
coreutils (and their predecessors), that is, the ls, cp, mv, cat, xargs, head, less, etc commands, have been around for several decades. You can sit down at the console and have an interface that is immutable, that does exactly what it has been doing for the past 20 years, and that does exactly what the user will say. (yes, I understand that you can break down a terminal with evil console_codes, and forgotten quotes in can be very stressful, and Ctrl-Z / C does not always work ...). But compared to any graphical interface, the command line has only one plus - it does not change. And does not change its behavior on the next restyling of everything and everything. For this, she is loved by professionals who are less likely to fall into the "inexperienced user" mode and which allows improving motor skills without a "new level, all skills are reset."
I wonder how many years must pass before such a simple thing as a fixed interface appears in graphical interfaces as well.