Errors, bugs, questions - page 881

 
sergeev:
how many do you know?
Managed to add a few more classes - but then it happened again - trying to localise the problem - rather complex inheritance and overloading - will report back
 
A100:
Managed to add a few more classes - but then everything repeated itself - trying to localise the problem - rather complex inheritance and overloading - will report back
No identical names with #define functions?
 
sergeev:
c #define no identical names with functions?
The symptoms occurred in the following way: at some moment, adding any newvirtual function to the class, including the simplest type
virtual bool ABC12345() const { return ( true ); }

started to cause an error: ',' - unexpected token..., with a reference to a line in another class - much earlier than the specified addition, where there were no errors before and ex5 worked fine. At the same time, a similar addition without the virtual key word did not cause errors. Changing names, parameters - had no effect. Then I tried to add just - a simple class (without the above mentioned)

class A {};

same error. The paradox, but after adding the 7th simple class the error disappeared. And it wasn't the names of the classes that had an effect, but only their number. But when adding the example with virtual the error occurred again. From #define there was only

#define PrintNotEmpty( a, b )   _PrintNotEmpty( b, a )
Analysis revealed that within the function, using template a variable, previously declared as static const string was used , this may have been the reason - after replacement with const string the error has disappeared and is not occurring now.
 

Added code - the same error appeared again - can be fixed by adding an empty class

class B {};
Apparently there is a critical number of them
 

Last result: previous conclusions are not correct

To make the error disappear - just add any meaningful line (not a comment) anywhere in mq5, for example

#define XXXXXXXXXXXXX 15
The virtual functions in ex5 are running smoothly so far. The problematic mq5 has been saved until the next ME5 revision
 
You'd better send the problem file to cd. It would be more useful.
 
In a month :) in time for Christmas
 
sergeev:
You'd better send the problem file to cd. It would make more sense.
This error has already been detected and corrected.
 
mql5:
This error has already been detected and corrected.

 

Question: is the min (minimum required) delay between calls documented

        OrderSend( request, result );
        Sleep( n );
        bool isOK = OrderSelect( result.order );
The question has to do with the fact that when n = 0 -> isOK = false, and when n = 200 -> isOK = true
Reason: