Errors, bugs, questions - page 905

 

Hello!

1 I have this problem - when I try to use a symbol in an indicator, not the one on the chart the indicator is attached to, it doesn't draw and the 4806 error comes out... if the symbols coincide, it draws everything normally... what's the problem? .... I just started studying the language...

 
vlad_123:

The picture has been tweaked.

Here are the calculations:

It shows that the optimizer is fooling people :-)

So, manual calculation shows that the result of the second line is greater than expected. But the optimizer shows that the result of the second line is lower.

And the optimizer shows a different result value than you expect when calculating manually.

Are you sure that "1.90" and "2.03" are exactly the same, and not the result of rounding to the second digit when displayed in the report?

The point is that these are rounded values for the report and cannot be the basis until you simply check the formula.

 
Yedelkin:
Exactly right. Not in the Help or in either of the two User Guides. Thanks for the link.

The reference points to a copy of the certificate. The reference has had it all since the beginning:


 
x100intraday:

But still.

I'm interested directly from developers: what is algorithm of my further actions? I have to do something about "MemoryException 180772428 bytes not available" (Log tab) and "out of memory" (Experts tab). Should I make an application to the SR?

Increase memory size, switch to 64 bits, reduce the amount of bars on the chart, or rewrite the code to a more frugal form.

If the program says there is not enough memory, then there is.

 
Renat: The reference points to a copy of the reference. It was and is all in the help from the beginning:

I asked in Russian :"What is MetaAssist and where can I read about it? Roche replied that "This term can be understood as everything described in the section...". But neither you nor Roche has answered what MetaAssist is. So don't refer to "Intelligent control" if there is no such term as "MetaAssist" in the Russian references. I could have found the "Intelligent control" section without your help.

In other words, there was and is no such term as "MetaAssist" in the reference materials, which completely refutes your statement that "It was and is all in the reference from the beginning".

 
Renat:

Are you sure that the numbers "1.90" and "2.03" are exactly that and not the result of rounding to the second digit when shown in the report?

The point is that these are rounded values for the report and they cannot be the basis until we simply check the formula.

I agree that for the first line of the screenshot, the optimizer's result 439639 is comparable to the manual calculation 444801 and the difference can be explained by rounding.

My question is about the second line where the optimizer calculated 373049, while the manual one adds 504383.

 
Renat:
This is a temporary bug, already fixed and will be available in the official build this week. Only the beta version on our demo server has been released so far.
Fixed in build 742. Thank you.
 
Renat:

Increase the amount of memory, go to 64 bits, reduce the amount of bars on the chart or rewrite the code to be more economical.

If the program says there's not enough memory, that's what it is.

How come,Renat?! With 1GB of memory not available about 100MB. I added 3 GB (but as XP x32 sees only 3 GB, I will consider that only 2) - the result didn't change. I increased swap file from 1.5GB to 10GB (let's walk around so walk around) - again no result, same error. How many more ruddy hussar GB do I need to cram in to make up the small hole of 100 MB? There must be a reasonable logical explanation for this circumstance. I don't need to know any details about MT5 code, but you must admit that you have made some changes in one of the builds, which now eats the memory, and that's a very fair move.

For a guide to action, my sincere thanks. That's all very well, of course,Renat. But it's too obvious and trivial. All I have to do is to unilaterally optimise something that was working carefree some six months ago. But one-sided recommendations are still not enough. Do I look like a novice programmer screaming because I can't make my first "Hello, world! Not at all. I am not the one who ruined something that once worked and did not need fixing.

But where, I wonder, are the notifications that the new build is eating more? I, as a good Indian, every time I update the terminal, go to the forum and thoroughly study every point of innovation and fixes. But I don't recall there being a single hint of "Isn't it time for a new computing power, buddy?...and preferably a server at once! I mean, come on, not everyone has that option. I've already got up to my ears and got four identical (as recommended in motherboard manuals) obsolete and expensive modules of Gigabyte each, and I've reached the architectural limit, there's nothing more to add. And in half a year we'll wait for another surprise from MQ?

For us, MQL-programmers, it's much more difficult to adapt to your new builds of MT5: you know what you modify, delete and add, while we don't know it (not counting explicit innovations in commands, functions, libraries, etc., which we must study to keep up with the terminal and the language). It turns out that your tool is logic, and we are left with intuitive searching, adapting at random and shoveling out stale code. Somehow that's not fair or anything.

Anyway, on x64 I'll definitely test it soon and report back, but something tells me...

 
x100intraday:

How come,Renat?! With 1 GB of memory not available about 100 MB. Added 3GB

Once you become a programmer, you need to understand that it is you who starts consuming resources with your requests. Calling expensive functions by no means means means that you can disconnect from the question of "how resources are actually handled".

The best way is to post the full code here in the forum and the problem area will be detected immediately. If you can't do it here, you can do it in service Desk (the code will be removed after checks).

ps: we never have questions when the OS says not enough memory and we never blame Microsoft for that.

 

after a quiet update to the latest build, deleted agents started to fall off:

FD      0       Tester  23:27:14        expert file added: Experts\grider1.1.ex5. 18867 bytes loaded
KR      3       Network 23:27:14        wrong optimized parameter 7
OG      2       Tester  23:27:14        requested data synchronization error
DR      0       Tester  23:27:14        tester agent shutdown
Reason: