|
Your highest priority should be to learn to use the debugger. It might sound advanced... but it isn't. You will be able to spot the course of errors like this in seconds - even without a heap of experience. Sure some errors will still be tricky... but if they are tricky with a debugger, they are pretty much impossible to solve without.
All you have to do is set a break point in the code you see a problem with (F9 in Visual Studio default key binding). Start the program with the debugger attached (F5) and reproduce the problem. Now you can step through the program line by line (F10/F11)- and you can see the values of all variables and fields at each line in the program.
Spotting the setter calling itself is trivial this way. As a beginner you might still want to ask help on how to solve a problem once you identified it - but at least you can ask a much more precise question when you know what the error is - and get help a lot easier.
So please - take the time to learn the debugger. The half hour will come back with interest of a few million percent.
|
|
|
|
|
I normally have an error message popup that tells me e exactly what is wrong (in most cases it's a spelling mistake) but this time this did not happen.
I have stepped through programs to see how they work. I could try doing that but I suspect all that would happen is the same message would pop up which does not tell me exactly what is wrong.
However a popup box with same message is better than nothing. I may need to research this message on the internet
Brian
|
|
|
|
|
Rule number 2 when troubleshooting (rule number one is "use the debugger") is not to assume anything. Assuming your debugger won't show you what is wrong is not the right approach.
What would happen if you had spend 10 seconds setting a breakpoint in the getter and setter and then kept pressing F11: You would see your code repeatedly call itself, while the callstack keeps growing one line each time it calls itself.
|
|
|
|
|
Error messages which tell you about spelling mistakes or other syntax errors are generated at compile time of your program. No executable is produced and therefore your program doesn't run.
A stack overflow condition can only occur when your program has been compiled and started running.
The debugger is the tool to be used for analyzing run time errors.
You should learn to understand the difference between these to error types.
|
|
|
|
|
There is no message you could possibly get that will tell you "exactly what is wrong" when the logic in your app is what is wrong. The error message you get can even be misleading.
Using the debugger to step through the code is THE most powerful tool you can use to find problems, not error messages. Error messages are a symptom of the problem, not the solution to it.
|
|
|
|
|
I created an error to test out the debugger.
I then pressed F5 to turn on the debugger, but when doing that the program compiles and tells me the error before I can step thought it.
Brian
|
|
|
|
|
Show the code.
Hitting F5 doesn't "turn on the debugger". It launches the result of compiling whatever is tagged as the "Startup project", and its dependencies, then the debugger is "attached" to the launched process.
There are two sets of error messages, compile-time and run-time. All of the messages that tell you "what's wrong and how to fix it" are at compile-time, before you even have an .EXE to run. These are syntax problems with your code.
The other messages show up at run-time and tell you how your code failed to execute somewhere. These are related to very specific problems with some action your code is executing, such as "Access Denied" or "File Not Found". These message never show up at compile-time and will only give you a hint of what to look for when stepping through your code and inspecting variables. These are problems with the logic in your code. Quite often you can write syntactically correct code and it will compile, but when it comes to running it, you can have all kinds of problems.
|
|
|
|
|
Seriously: Are there really people out there NOT using a debugger when developing code?
To me that is like climbing Mt.Everest blindfolded, all on your own.
There may be extreme exceptions where no traditional debugger is available or can't be used, e.g. in a certain class of embedded systems where the code is in flash memory and cannot be modified on a byte or word basis, so you can't poke debug interrupts. Or the processor is so primitive that it doesn't provide interrupts, or doesn't give sufficient external access to the address bus to run a probe on the outside. But those are very special cases. And I have never seen a system where you cannot do "printf style" debugging by setting output signals - even 8-bit embedded processors offer that (otherwise, how would they control what they are supposed to control?)
Those special cases are lightyears away from desktop development in C#. If you run a company, you most certainly can affort the investment in, say, Visual Studio. For hobby programming, there is a free version providing so near-100% functionality that you probably won't notice the difference. (There are some differences, but they are rarely relevant for hobby programming.)
Even printf debugging can work well in some cases (sometimes it is usesful as a companion to the interactive debugger, e.g. to traverse data structures and print selected values if this requires more processing than a simple watch expression).
But "debugging by cranial massage", as one University lecturer phrased it. is not sufficient.
|
|
|
|
|
Sadly yes. The number of QA posts that show the user has made no attempt to do even basic debugging, continues to amaze (and worry) me.
|
|
|
|
|
But isn't the program debugging when it compiles. I normally get an error message if the compiler can't compile the code.
Maybe your talking about some other debugging.
Brian
|
|
|
|
|
Brian_TheLion wrote: But isn't the program debugging when it compiles.
No. Errors at compile-time are just syntactic problems with your code. The compiler cannot compile the code into an executable when these show up. They have nothing at all to do with the logic of your code.
Debugging comes into play when you actually start running the code.
|
|
|
|
|
I think I understand now what you are meaning Dave.
It's when the program does run but gives you the wrong results is the other type of debugging.
Brian
|
|
|
|
|
Personally I would never call fixing compile time errors "debugging".
|
|
|
|
|
Look at your code.
public int Gold
{
get
{
return Gold;
}
...
}So now lets use it:
int g = myClassInstance.Gold; It enters the Gold getter, which returns the value of ... the Gold getter. Which returns the value of ... the Gold getter. Which returns the value of ... the Gold getter. ... until the stack blows and your app crashes.
The same happens when you try to use the setter: it infinitely recurses trying to set itself!
Create a private backing field:
private int _Gold = 0;
public int Gold
{
get
{
return _Gold;
}
set
{
if (value == 20)
{
value += 5;
_Gold = value;
}
}
}And it'll all work.
Sent from my Amstrad PC 1640
Never throw anything away, Griff
Bad command or file name. Bad, bad command! Sit! Stay! Staaaay...
AntiTwitter: @DalekDave is now a follower!
|
|
|
|
|
Tbanks Griff.
It looks like I've created a continuous loop.
I'll try what you have suggested.
I noticed that you used _Gold instead of Gold.
Brian
|
|
|
|
|
_Gold is a member variable. Gold is the property that exposes the variable. Basically, if you attempt to get a property and return the property name instead of a variable, you're going to repeatedly call the same getter. This is what causes the StackOverflowException.
This space for rent
|
|
|
|
|
Hi Griff.
Your change worked when there was a test for the value of 20 but when I changed the test value to 30 then 0 was displayed on the form for the amount of gold, then when I clicked on the Test button the gold value changed to 35.
What I'm aiming at on this test is for 20 to first to appear on the form for the Gold value after the program is run, then when I press the Test button the gold value should change to 30 and in the test have 5 added to the 30 giving a gold value of 35.
Here is my code
On the main program I have
SuperAdventure.cs
<pre lang="c#"></pre>
public partial class SuperAdventure : Form
{
private Player _player;
public SuperAdventure()
{
InitializeComponent();
Location location = new Location(1, "Home", "This is your house.");
_player = new Player(10, 10, 20, 0, 1);
Update();
}
// TEST Button on Form
private void btnTest_Click(object sender, EventArgs e)
{
_player.Gold = 30;
Update();
}
public void Update()
{
lblHitPoints.Text = _player.CurrentHitPoints.ToString();
lblGold.Text = _player.Gold.ToString();
lblExperience.Text = _player.ExperiencePoints.ToString();
lblLevel.Text = _player.Level.ToString();
}
}
}
Player.cs code
namespace Engine
{
public class Player
{
// This is needed for _player = new Player(10, 10, 20, 0, 1,);
public int CurrentHitPoints { get; set; }
public int MaximumHitPoints { get; set; }
private int _Gold = 0;
public int Gold
{
get
{
return _Gold;
}
set
{
if (value == 30)
{
value += 5;
_Gold = value;
}
}
}
public int ExperiencePoints { get; set; }
public int Level { get; set; }
//In order to use lists – we need one variable or property to hold a collection of objects that are the same class
public List<InventoryItem> Inventory { get; set; }
public List<PlayerQuest> Quests { get; set; }
// This is needed for lblHitPoints.Text = _player.CurrentHitPoints.ToString() etc;
public Player(int currentHitPoints, int maximumHitPoints, int gold, int experiencePoints, int level)// : base(currentHitPoints, maximumHitPoints)
{
CurrentHitPoints = currentHitPoints;
MaximumHitPoints = maximumHitPoints;
Gold = gold;
ExperiencePoints = experiencePoints;
Level = level;
Inventory = new List<InventoryItem>();
Quests = new List<PlayerQuest>();
}
}
}
</pre>
|
|
|
|
|
The problem you have is that you don't assign anything to _Gold unless value is 30. This is because you do _Gold = value inside the if statement. Move this line outside the if block in your setter and it should start working.
This space for rent
|
|
|
|
|
Thanks Pete.
That was easier to fix than I thought it might be.
Brian
|
|
|
|
|
Not a continuous loop, but an "unbounded recursion" - similar thing, but the later will blow the stack at some point.
Just to add to what Pete says, the naming conventions say that a backing field should have the same name as the property it is associated with, but prefixed with underscore:
private int _MyProperty;
public int MyProperty
{
get { return _MyProperty; }
set { _MyProperty = value; }
}
Sent from my Amstrad PC 1640
Never throw anything away, Griff
Bad command or file name. Bad, bad command! Sit! Stay! Staaaay...
AntiTwitter: @DalekDave is now a follower!
|
|
|
|
|
I twitch when I see a private backing field that has an uppercase letter following an initial underscore :  I'm sorry.
However, when I consider this example is still current in the MS docs [^]:
public enum Color {...}
public class Control {
public Color Color { get {...} set {...} }
}I throw up.
Remember the daze when common practice was to use: _mSomeName ?
«Where is the Life we have lost in living? Where is the wisdom we have lost in knowledge? Where is the knowledge we have lost in information?» T. S. Elliot
|
|
|
|
|
Thanks everyone who has helped me with this code problem.
I now have a way of sending a value to a class, have the class change the value (if the test passes) and have the class return the result (which is displayed on the form.
Brian
|
|
|
|
|
Do keep in mind that MS Guidelines suggest using Properties only as gateways to internal fields [^] Quote: In general, methods represent actions and properties represent data. Properties are meant to be used like fields, meaning that properties should not be computationally complex or produce side effects. When it does not violate the following guidelines, consider using a property, rather than a method, because less experienced developers find properties easier to use.
not to produce 'side-effects" by way of internal computation in the setter and/or getter ... and advocates using methods to replace any internal computation.
While I feel "relaxed" about some computation in a Property setter/getter, there are folks who have strong opinions this is bad practice. Another argument against computation in Properties is that end-users of your code may not be aware of the "cost" of access.
Of course, you can implement INotifyPropertyChanged as a "trigger" for some external action.
«Where is the Life we have lost in living? Where is the wisdom we have lost in knowledge? Where is the knowledge we have lost in information?» T. S. Elliot
modified 30-May-19 5:10am.
|
|
|
|
|
Hi Bill.
You wrote that
INotifyPropertyChanged can be used as a "trigger" for some external action.
Can you use this to auto add text to a rich text window on a form when you change the string variable in a class that the rich text window is connected to?
Brian
|
|
|
|
|
Brian_TheLion wrote: Can you use this to auto add text to a rich text window on a form when you ...
Theoretically, given the right structure, and valid references to run-time objects, the handler(s) that subscribe to the event can do anything any other method can do.
Take a look at the examples here: [^]
«Where is the Life we have lost in living? Where is the wisdom we have lost in knowledge? Where is the knowledge we have lost in information?» T. S. Elliot
modified 30-May-19 11:01am.
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
