Errors, bugs, questions - page 658

New comment
 
Renat

Maybe it is time to introduce a definition of standard operators for classes

types:

class Cname
  {
public:
                     Cname(void){};
                    ~Cname(void){};
   Cname             multi(Cname *a,Cname *b){return(a*b);};// это типа перемножение, по идее компилятор тут должен ругаться
   inline Cname operator <*> (Cname *a,Cname *b){return(multi(a,b))};
  };

It would not hurt security, since the substitution takes place at the compilation stage, but it would simplify the interface of user types considerably.

Cname a=xx;
Cname b=yy;
Cname с;
// вызов функции перемножения будет такой
 c=a*b;

// вместо
 c.multi(a,b);

If we have such a possibility, we could implement complex arithmetic, matrix arithmetic with high precision.

In other words, it would solve a wide range of problems, and MQL5 could become quite an easy to use language.

 
abeiks:

I wanted to test expert that without problems worked out 2011 ATC Championship but no longer works, returns error: failed instant sell 1.00 EURUSD at 1.30898 sl: 1.33961 tp: 1.30287 [Invalid request].

I believe this is due to an update. What do I need to change in the program?

https://www.mql5.com/ru/forum/1111/page575#comment_118621

ZeroMemory we use.

 
Urain:

Maybe it is time to introduce a definition of standard operators for classes

types:

It won't hurt security, as substitution takes place at the compilation stage, but it will simplify the interface of user types considerably.

I've been waiting for it for a long time - Slava promised me this some time ago. // I already forgot when... :)

Let's start polling.

 
MetaDriver:

I've been waiting for a long time. Slava promised some time ago. // I've forgotten when... :)

Let's do a poll or something.

What's there to interrogate? It's not a mystery, the implementation has long been present in C++, so the need has been tested.

It's just a matter of MQ if they want to implement it. And what will a poll do?

SZZ Renat the beginning at the end of the previous page, you will scribble something?

 
Urain:

What's there to question? It's not a mystery, the implementation has been around for a long time in C++, so the need has been established.

It's up to MQ whether they want to implement it. And what good will it do? No matter how much you vote, they will (or won't) still write MQ.

SZZ Renat the beginning at the end of the previous page, you'll scribble something?

I am for such necessary and obligatory functionality.

Let's finish the current tasks and get on with it.

 
Renat:

I am in favour of such necessary and obligatory functionality.

Let's complete the current tasks and get on with it.

I agree, such things should be introduced with some stable build, so that the bugfix is not superimposed.

Create a proposal in the SR? Or will you remember it yourself?

 
Let's remember.
 
Karlson:

https://www.mql5.com/ru/forum/1111/page575#comment_118621

ZeroMemory we use.

I don't know, either I'm putting it in the wrong place or this option doesn't work.

 

Something terrible is happening to the bars... Killed history (thought it was corrupted on disk), re-downloaded - some segments are disappearing again (tearing in the image)... Terminal is lagging like hell (still downloading history or just bad connection to server today?). The calculated values for objects are slipping to the left by a few bars... Objects previously positioned correctly are shifting.

And in general. Is the story going to load all day or more? It's as if it doesn't come in a compact archive, but in an uncompressed form it just barely flows onto the machine. Hence the question: isn't it possible to download the history manually from the server (in an archive) and then unzip it to the desired location, and there it would be downloaded a bit at a time, as usual, from the current moment?

 
x100intraday:

And anyway. Is the story going to load all day or more?

Which access point?
1...651652653654655656657658659660661662663664665...3190
New comment