The problem of transferring from MT4 to MT5. Or, more precisely, the inability to execute some algorithms in MT5 without'err. - page 2

[Vladimir Karputov]

Uladzimir Izerski:

If that's a question for me?

Here's a simple example.

D1.

Go to H4

Go to H1

There will be a clear chart on a small TF.

Objects are in the lists, but they are not on the chart.

They are just pictures. Please describe, step by step, how you create, where you create the object, what are the object properties. In general, how to reproduce the situation.

[Uladzimir Izerski]

Manually inserted the "trendline" object on the chart. In properties/display/show on all TFs. This is on Win7. I don't have 10 at hand at the moment. But it is the same. The pictures show on what TF the object was created. When shifting to smaller TFs, the trend lines are not drawn. This concerns all objects that were drawn on higher TFs but disappear at shifts to small TFs. I assume that there is not enough data for the far point of the object.

You can reproduce it yourself.

[Vladimir Karputov]

Uladzimir Izerski:

Manually inserted the "trendline" object on the chart. In properties/display/show on all TFs. This is on Win7. I don't have 10 at hand at the moment. But it is the same. The pictures show on what TF the object was created. When shifting to smaller TFs, the trend lines are not drawn. This concerns all objects that were drawn on higher TFs but disappear at shifts to small TFs. I assume that there is not enough data for the far point of the object.

You may try to reproduce it yourself.

What is in the terminal settings: how many bars to show?

[aleger]

Uladzimir Izerski:

Manually inserted the "trendline" object on the chart. In properties/display/show on all TFs. This is on Win7. I don't have 10 at hand at the moment. But it is the same. The pictures show on what TF the object was created. When shifting to smaller TFs, the trend lines are not drawn. This concerns all objects that were drawn on higher TFs but disappear at shifts to small TFs. I assume that there is not enough data for the far point of the object.

You can reproduce it yourself.

Alternatively, you can use the ZigZag lines as trend lines

[Uladzimir Izerski]

Vladimir Karputov:

What is in the terminal settings: how many bars to show?

50 000.

Set it to 500,000. It's changed.

It is clear then. I have to set it to the maximum. But... I see.

In 4, it didn't depend on it.

[Eugeni Neumoin]

Renat Fatkhullin:
And can you reproducibly demonstrate the inaccessibility of the data?

You assert without reproducible evidence.

Well, first of all, all my reasoning started with a quote from a language reference book. I'll repeat it again.

Spavotnik's rubric Organisation of data access.

"...Data accessibility.

The availability of data in HCC format, or even in ready-to-use HC format, does not always mean the unconditional availability of these data for display on the chart or for use in mql5-programs.

When accessing price data or indicator values from mql5-programs, one should keep in mind that it is not guaranteed to be available at a certain point in time, or from a certain point in time...".

It directly follows from this that the situation I have described is possible.

Now about the logs. In post 6 -https://www.mql5.com/ru/forum/318246#edit_form look at the first picture:

Let me decode this picture. There is also a log in it.

Moving cursor to W1 means that the forks were created on cf W1. Look at the tooltip:

time1 - shows the time of anchoring the pitchfork to the extremum number 3.

m20-MN1 - shows boundaries of fractal filtering. These two items are fine.

But bar t1 - should display the bar number for time1. And as we can see it is equal to -1. Thus, the history for timeframe W1 has been formed. This can be seen on the chart. The zigzag has been built on this history.

But we have received an error when calculating the number of the bar.

I will show you a picture of how it should be.

The tooltip at shows bar t1 = 271.

Now let's look into the code:

        ObjectSetString(0,name_obj_tfcreate,OBJPROP_TOOLTIP,"time1 : "+TimeToString(arr_ap[numObject[2]].t1,TIME_DATE|TIME_MINUTES)+
        "\r"+"bar t1 : "+(string)iBarShift(_Symbol,_Period,arr_ap[numObject[2]].t1,true)+
        "\r"+((arr_ap[numObject[2]].tfMin==arr_ap[numObject[2]].tfMax) ? ("Only "+TF_txt[arr_ap[numObject[2]].tfMin]) : (TF_txt[arr_ap[numObject[2]].tfMin]+" -:- "+TF_txt[arr_ap[numObject[2]].tfMax])));

Здесь формирование строки с bar t1:

"bar t1 : "+(string)iBarShift(_Symbol,_Period,arr_ap[numObject[1]].t1,true)

The bar t1 search by time is exactly what we get the error.

This is an illustration of that very log. And the cited quote from the language reference. And everything that was said in the first post of this thread.

History is generated. But there is no access to the generated history.

The Language Guide says that in such cases you have to wait until the next time. But put yourself in the user's shoes. He/she wants to display a pitchfork. He moves the cursor to the extremum number 1. Clicks on number 1. The pitchfork will appear. In my first post I described a situation when there was no access to timeframes older than the current one. It really happened. The upper limit of fractal filtration equaling the current timeframe has been formed. When moving to higherframes the fractal filtering does not allow access to the chart. The user is shocked.

It turns out that when timeframes are formed on all TFs, access to timeframes is lost at uncontrollable moments.

I am not inventing anything here. I read the Handbook and see confirmation of what is written there. And that, in my opinion, is the most significant disadvantage of MT5.

[Renat Fatkhullin]

No reproducible code = no proof.

Conclusions on derivatives are not accepted.

[Eugeni Neumoin]

Renat Fatkhullin:
Can you reproducibly demonstrate the unavailability of data?

You assert without reproducible evidence.

I have tried to answer the question. The data is not available.

The log is in the tooltip. The code for the tooltip is given. The timeseries is generated. A graph with W1 timeseries is built. Dostap is absent. TheiBarShift function generates an error. There is no such error in MT4.

[Artyom Trishkin]

Eugeni Neumoin:

I have tried to answer the question. The data is not available.

The log is in the tooltip. The code for the tooltip is given. The timeseries has been generated. A graph with W1 timeseries is built. Dostap is absent. The iBarShift function generates an error. There is no such error in MT4.

From the pictures, did you try to reproduce the behaviour claimed by the other person?

This is a technical forum and you need to reproduce the code describing the problem.

Until there is one, the problem you're talking about won't be solved either.

[Eugeni Neumoin]

Artyom Trishkin:

Have you tried to reproduce the behaviour claimed by another person from the pictures yourself?

This is a technical forum, and you need code that reproduces the problem you're describing.

Until he is gone - the problem you are talking about won't be solved either.

Artem, you are the one who deleted my post in the article discussion thread. Where I was replying to a forum member's request?

I was just touching upon the subject brought up in this thread. And I was angered by this cavalier attitude. So I created this forum thread.

======

I've had disagreements on this forum before. You don't want to discuss here. There are many forums where you can say a great deal without censorship.

Back in 2005 a moderator (HQ) did something similar on the Alpari forum. And then almost the whole forum moved from Alpari to Onyx.