Errors, bugs, questions - page 2497

 
Сергей Таболин:


Good afternoon.

Yesterday evening, the windup update came in. I agreed to it.


Today ran the optimization.... Bummer... What was still optimised yesterday (genetics, opening prices only) for max 10 minutes, today is 265/10496 in 30 minutes.

What to do with it?

P.S. I recompiled Owl. Result is the same, deplorable.

It took me 40 mins after the question.


The reason was"Unfortunately, this update will make it impossible to run the installed activated apps fromMarket and will require a new activation".

Really, it would be nice if there was some hint of it in the magazine...
 
Сергей Таболин:

The reason was"Unfortunately, this update will make it impossible to run the installed activated apps fromMarket and will require a new activation".

Really, it wouldn't be bad if there was at least a hint of this in the logbook...

I don't quite understand - were there indicators from Market that refused to work?

 
Aleksey Vyazmikin:

I don't quite understand - were there indicators from the market that refused to work?

Exactly.

 
Сергей Таболин:

Exactly.

I see. That's too bad, the terminal should have taken care of that and downloaded the current versions from Market, while asking for confirmation of its actions.

 
Aleksey Vyazmikin:

I see, that's too bad, the terminal should have taken care of that and downloaded the current versions from the Market and asked for confirmation of its actions.

This is written about in the forum. With warning that there will not be available programs from Market after upgrade to Win 1903, and that they will need to download again. And most importantly, one activation has been added to all those who have purchased in connection with all this.

 
Artyom Trishkin:

This is written in the forum. With a warning that no programs from the market will be available after upgrading to Win 1903, and that they need to be downloaded again. And most importantly, one activation has been added for everyone who bought in connection with all this.

If I am a user of the programme, I don't necessarily need to visit the forum...

And it's more about a user-friendly approach, not about the presence or absence of formal notifications.

 
Wrote the wrong code, but it compiles
class A
{
public:
  virtual bool Func() { return(true); }
};

class B : public A
{
public:
  virtual bool Func() { return(false); }
};

class C
{
  A* obj; // Этот вариант по какой-то причине компилируется без ошибок.
//  A  obj; // Здесь правильная ошибка компиляции, но их две!
//  B  obj; // Здесь полная ерунда в ошибках и предупреждениях (expression not boolean).
  bool f() const { return(this.obj.Func()); }
};

If you use other lines in the source code, there are lefty errors.
 
2093
Similar situation from a neighboring OOP thread, with a known error.
When running in the debugger, a debugger exception is raised. Everything is fine.
But when compiling, the compiler doesn't raise any exception.
class CObj
{
   public:
   int f(){return(33);}
};

//+------------------------------------------------------------------+
void z(CObj &o)    //тут специально пропущен указатель
{
   o = new CObj();
   Print(o.f());
}

CObj * obj;

//+------------------------------------------------------------------+
void OnStart()
{      
  z(obj);
  delete(obj);
}
 
Alignment does not work
struct A pack(4)
{
  short j;
};

void OnStart()
{
  Print(sizeof(A)); // 2
}
 
fxsaber:
Alignment doesn't work.

In that case, it shouldn't. It's an alignment, not a multiple of size.