Testing the new MQL5 compiler for x64 platforms - 2 to 10 times faster calculations! - page 6
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Do you trade on the FORTS?
It would be interesting to hear what is the MT5 good for stock trading?
Hi honest company. Here's what I have, can anyone explain - MT5 doesn't work poke poke don't start...how to start don't know...
How do I play it back?
It is not a small thing, but the fundamental basis on which the world rests:
The point is that what matters is not the truth of the understanding of the situation, but the practicality of that understanding. According to the ideas of modern physics, all objects, and in fact the whole world, consist of emptiness. Inside each atom, particles occupy a very small space. However, it is much more practical for us to believe that objects are impenetrable. Similarly, it was much more practical for people in ancient times to believe that the earth was supported by three elephants. And in the same way, it is much more practical from a developer's point of view to look at what most people need.
The requests made here are for an extremely small number of people. I suspect that even the class browser - a very useful feature for anyone who actively uses the OOP features - is not really needed at all, at least for 10% of MQL programmers. All the rest use the old procedural-oriented style and don't even use simple classes, not to mention the features represented by encapsulation, inheritance and polymorphism.
The point is that what matters is not the truth of the understanding of the situation, but the practicality of that understanding. According to the ideas of modern physics, all objects, and, in fact, the whole world consists of emptiness. Inside each atom, particles occupy a very small space. However, it is much more practical for us to believe that objects are impenetrable. Similarly, it was much more practical for people in ancient times to believe that the earth was supported by three elephants. And in the same way, it is far more practical from a developer's point of view to look at what most people need.
The demands being made here are for an extremely small number of people. I suspect that even a class browser is a very useful feature for all those who are actively using OOP features, but now not more than 10% of MQL programmers need it. All the rest use the old procedural-oriented style and don't even use simple classes, not to mention the features represented by encapsulation, inheritance and polymorphism.
It is not more practical to consider that objects are impenetrable, just that there is no choice. As long as humanity isn't really presented with the ability to see the world as it is.
The conversation is not about classes, but about the possibilities of working with individual orders and the independent work of various experts. I don't know about anyone, but it seems to me that abandoning such opportunities is a plunge into the Middle Ages BC.
Progress should both increase opportunities and increase comfort, not only reduce the number of opportunities.
It turns out like an anecdote in one. They emptied the pool, but installed a new diving tower and issued new rubber caps.
This is a fruitless discussion.
The developers have their own concept of product development, and that is their right. I don't think they are interested in the opinion of the few who use sophisticated data analysis algorithms. To all those crying "give me this, give me that..." - just use what you need with a cash terminal. In my experience, everything that anyone may need in the analysis of financial series and trading is implemented in the R language. This language is one of the ten most popular and widely used languages. This language is implemented in their products by such mastodons as IBM, Oracle and Microsoft. The need of Big Data analysis gave rise to this offer. In our case, there is no great need of users in complex algorithms yet. So, the developers will not strain themselves. Although (speaking amateurishly of course) it seems to me that making a gateway in R is not a problem at all for such a team.
About Matlab. R implements an interface with Matlab, but it must be available (Matlab). It also costs a lot of money. I may not be familiar with it, but what is so important in Matlab that it is not implemented in R? Just for general development.
I repeat. It's a fruitless discussion. It's more a matter of preferences. Some people like Windows, while others don't recognize anything but Linux. Some (immersed in the depths of MKL5) like the fascinating words "uhar", "muhar", polymorphism", others are not interested in this level of abstraction at all.
If you know what you need, just use it instead of waiting for the developer to give you the opportunity. Maybe after you try the seemingly effective algorithm, you will change your mind about it.
Good luck
Actually my expectations from Metakvoters are simple, I hope that someday they will realize that people need profit, that they need ready-made solutions that connect to real accounts and provide guaranteed profits (even if small, but 100% guaranteed).
What will it be? Expert Advisors of Metakvoters, Signals from Metakvoters, behind which there is serious scientific work of a cohesive team .....
I see criticism of MT4/MT5 from those who have been on the forum for a long time (since MT4), the essence is clear, people have long found their solutions, developed their software ... I think most of them don't give a shit about success of MT5, what's important to them is "SHOW ME WHERE WE CAN MAKE SOME PROFIT WITHINST ACCOUNTABILITY".
The special way? So wait.... Do something that no one on an industrial scale has ever done before...create a miracle.
Am I wrong?
Very few people understand why our own secure language is important in our case. But it is a critical component.
For those who don't know - MQL4/MQL5 has the most indicators and Expert Advisors written in the world for trading platforms, while MQL4/MQL5 itself has entered the ranks of programming languages.
The next changes in Metatrader 5 will bring a large set of new features in MQL5 that, together with its current capabilities, will allow us to write powerful analytical systems right in the terminal. However much detractors may laugh, but in MQL5 you can even write a computer game inside the terminal. And not even one of them, and run it in parallel.
It's all true. Don't listen to the detractors. Listen only to well-wishers, of whom there are many on this forum.
And it's all about psychology: you see, when a person develops a trading system and it doesn't work so far, it's annoying. It's BIGGESTLY annoying. Until the trading system doesn't work properly, then the trader (or developer) is the most unhappy person in the world. There is no such thing in any business - to see for years that the chosen path does not work. Usually in business you switch to something else, instead of standing in a losing position for years. And here you have to be constantly nervous, and some of their irritation because of their trading system, spills out on your developments.
Renat, you know the feeling, too: your MT5 environment, despite some advantages, is slower than you would like it to be. And of course it must be a bit annoying. And you're forced to rethink some of your business decisions from the past, like "your quotes" in the terminal. Of course there was also a copyright on the quotes and other things, but everything is fine now, as far as I understand it.
Now to the point: the MT4-MT5 has a significant flaw. It is NOT a flaw, because other terminals don't have it either. Namely:
There is no standard means of internal processor scheduling for Expert Advisors. Like semaphores in the OS. This can be avoided by turning to DLL, but there are pitfalls from Microsoft and not everyone understands system programming.
It turns out that MT-4-MT5 has all you need for complex calculations, but it is almost impossible to implement it on 50 simultaneously open charts - currency pairs in expert-advisor, and especially in an indicator. You can test it, but it is problematic to run such a terminal with 50 simultaneously opened Expert Advisors - because the terminal will hang and each chart-flow will struggle for the processor. And if you bypass this via DLL, then such an advanced-scientific Expert Advisor has to carry around a DLL, and it can not be placed in the market because it is prohibited there by DLL.
(Note: for comparison, Ninja-trader terminal in this sense works about 50-100 times slower than MT4 terminal, and any more or less advanced indicator there will hang the terminal for a couple of minutes. And it is unreal to discuss the launch of 50 charts in one terminal with at least daily price histograms. So, Metatrader has no rivals in the field of development of scientifically based complex calculating trading systems of a reasonable price).
There is another nuance - when starting a terminal with 50 charts and pressing the "auto-trading" button, the Expert Advisors that are highly reckoned just hang the terminal up.
I personally avoided all this a long time ago and everything in MT4 works as it should, even with CUDA acceleration. But if you're half a step ahead of all other terminals in the field of C-compiler and ease of implementation of scientific calculations in the trading terminal, then such tools of a regular scheduler (or a fast interface of interaction between EAs) for heavily scientific heavy calculations wouldn't hurt.
In general, the new compiler is still raw.
It speeded up about 2-2.5 times (pattern recognition and analysis system), but some results are not adequate - most likely something is wrong with the object arrays.