Tester supporting MG4 scripts and advisors - page 7
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
Don't try to pretend that there is no or little documentation or examples. There are so many examples on this site, that you won't find them anywhere else.
I don't pretend there is little documentation. I just said that we don't have the documentation we need. But yes, there is a lot of documentation. It's obvious. 4000 pages. 350 articles. Of course it is a lot.
You are saying too much.
Examples of poor quality implementations have to be given from time to time as I come across something. And many other people have already given them many times, and not only in the language.
All the time people swear that once again the platform has been updated in an entirely unexpected way. You're exaggerating that I'm exaggerating.
MT4 was released more than 9 years ago and without the latest upgrade we would have lost a lot. We have kept good compatibility and we still carry a second copy of the virtual machine for the old MQL4 code.
On updates: it's up to us how to pass system update files so they don't get blocked by firewalls. If you think you are so smart, then try to transfer .exe files to millions of computers and see how many of them will be removed by antivirus and firewalls without any chance.
I don't have any objections to the actual upgrades. But there is some objection to voluntarism when making upgrades, that's why people argue on forums every time it happens. But of course the upgrades themselves are NOT evil.
By the way, you still have to create executable .exe files after all. Executable files of the new version of the terminal, editor, for instance. Doesn't antivirus interfere here already?
Just when I needed to temporarily prevent the terminal from upgrading, and in the process I accidentally found such a file in the corresponding folder, which is not at all what you might think at first sight - it was remembered, in general. Turns out that's what it's for.
It's for the red word, I understand.
Pure lies.
Stop repeating nonsense and outright lying. Behind the desire to argue, don't lose your shores, please.
As for Microsoft, let it be a figure of speech, but it is my feeling developed while working with Microsoft products and MetaQuotes products in development.
And concerning the disparate documentation - this is my opinion and my feeling, not a lie. You may think otherwise, but you may also be wrong in your estimations.
Let's get closer to the reality:
And this is what simpleton is up against. Too bad he has no idea what he is criticising.
I am not against the fact that there are a lot of resources. Instead I believe that they are fragmented and that there is a lack of consistency in the basic documentation. And if this is in fact the case, then the large number of resources becomes not a blessing but an evil, because there might not be enough energy/time to go through everything to find what you need. Of course, I could be wrong. On the other hand, it can also turn out to be a good thing for MetaQuotes to prosper.
I'm not worried about it, because now I'm already comfortable with MetaQuotes products.
But, in any case, I am not against the fact that there are a lot of resources. Yes, there are a lot of them. Again, this is - obviously. It is enough to re-read your list to be convinced of that.
How does a non beginner assess the quality of documentation. So I've started studying the OOP. I've read a lot of various MQL5 documentation. Note that it is the kernel, not API, as simpleton says. The questions I had after reading the documentation became more and more numerous. Only thanks to the forum, bypassing the IP ban I was able to ask some questions. And if it wasn't for simpleton, most likely most of them would have gone unanswered. Such is the quality of the documentation.
As for the API. I still have no idea how to use it correctly to save all the computer's resources and make my code as efficient as possible. I'm ready to use CopyXXX instead of iPrice. Suppose it takes a lot of lines - when I figure it out, I'll convert it to convenient forms and use it. But it's hard to figure it out. So this is the documentation.
I have several times asked to show me how to correctly write the same indicators in OOP. I haven't seen ANY OOP indicators, although I have written them myself. The sources of indicators are horrifying. As if they were written tens of years ago.
I asked to tell the OOP-concept of the optimizer/tester - silence. As if no one has ever done the same shaggy real-time auto-optimization. As it turned out, they did, but through a sydal nerve - running testers (MT4/5) with generation of ini-files and further parsing of results. It is easier, it seems, to hit the sparrows with C-300 than to make something by yourself. In short, the simplest tasks are not solved. And this is the core of the language, not the API at all.
I ask rather simple questions here myself, to which the answer from developers, as a rule, is either silence or "you're a fool". And when you find solid arguments of their position the developers simply don't answer. I get the feeling that they are unfriendly and ignore me, even if they are not.
I am still only into algo and forex, so I don't see any advantage of MT5 over MT4 after the introduction of MQL4++. ANY tester from any developer that has the limitations of MT5 tester will be bypassed. Because the limitations are key for me.
On the other hand, I don't even fall into three sigmas of MT4/5 users, so my opinion may not be taken for granted. Marketing is, if not our everything, then very much so.
That's the impression I got from kodobase. MQL4 (without ++) - the authors of these articles sometimes squeeze the unthinkable out of the language. I can see a lot, but the language has lost its relevance.
MQL5 (MQL4++) - authors of these works sometimes have the minimum possible power of the language. Well, if only there were some exemplary codes of practical orientation. But no, I can't find any.
And I want $20K extra per month, as recently stated in an article about 10,000 market-sellers. I need to figure out what shit-TS to write and what marketing steps to take to make my signal sales go through the roof. This, by the way, is a topic that requires a separate topic... There is an opinion that a sucker lives to be mowed down. And there is such a constructor idle on the part of many. We need to master it.
I'm sitting at my desk, stretching my hand...
How can a novice judge the quality of documentation? So I started studying OOP. I have read a lot of various documentation on MQL5. Note that I have read the kernel, not the API, as simpleton says. The questions I had after reading the documentation became more and more numerous. Only thanks to the forum, bypassing the IP ban I was able to ask some questions. And if it wasn't for simpleton, most likely most of them would have gone unanswered. Such is the quality of the documentation.
As for the API. I still have no idea how to use it intelligently to save all machine resources and make the code as efficient as possible. Ready to use CopyXXX instead of iPrice. Suppose it takes a lot of lines - when I figure it out, I'll convert it to convenient forms and use it. But it's hard to figure it out. Here is such documentation.
Judging by your questions in the forum you are interested in language elements, OOP study, but not in program writing. That's why you haven't got to the API either.
The basic issues of OOP can be addressed in any book on C++. This is something that cannot be passed over in programming, it is a base and has no direct relevance to MQL5.
You communicate in a public forum, people correspond with you, they help you. That is, the task of your free personal training is being done.
And no one is obliged to answer to anyone, and everything is done purely on a voluntary basis.
MQL5 (MQL4++) - authors of these works sometimes have the minimum possible power of the language. Well, if only there were some exemplary codes of practical orientation. But no, I can't find any.
You communicate in a public forum, people correspond with you and help you. In other words, the task of your free personal training is fulfilled.
And no one is obliged to answer anyone, and everything is done solely on a voluntary basis.
For this reason I ask you to unban simpleton, because he answered the lion's share of my questions very thoroughly, almost completely teaching me the basics of OOP.
Let him go on bothering you with his remarks. But it helps me a lot, in particular. Please listen to my request. Without it, many questions (even simple ones) remain unanswered.
Trading is a very specific activity. It requires knowledge of many things, including programming.
Relying on someone else would not even be the right thing to do. We have been there.
The MQL4 language proved to be not so difficult to use, even with zero programming skills I got the hang of it very quickly.
However, I cannot move up to the next level with the addition of ++. I don't have any complaints about MQL4 and MQL5, but the fact that there is no structured and simple form of training has become a real brake. I don't think it's necessary for non-professionals to dig through 4 000 - 400 000 pages of training course. Even a few examples of programs with line-by-line commentary is more helpful, than the thickest books, and can make you more optimistic in learning. Not everyone needs to be a professional, but at user level it is a must.
This is where the developers have turned their backs on ordinary users. As frustrating as that may sound.
Until this issue is resolved, it will not be possible to drive to 5.
MQL5 has been around for quite some time, increasing the functionality. And I wasn't drawn to study it at all, because it was obvious that it wouldn't have any practical use.
But when MQL4++ (clone MQL5) appeared, it became clear that I had to eliminate a moron when reading "constructor", "operator overloading", etc. I started studying OOP through writing practical problems. I started to learn OOP by writing practical problems instead of the bare theory. Along the way, I asked questions, many of them were answered by simpleton. As a result, I began to understand the OOP practice. OOP for the sake of OOP is a sad approach.
Now I write everything in OOP only. Even small programs. It is really cool! But you can feel it only through solving practical tasks.
So I recommend to study the MQL5 kernel ( == MQL4++), and you can always decide on the platform number.
MQL5 has been around for quite some time, increasing the functionality. And I wasn't drawn to study it at all, because it was obvious that it wouldn't have any practical use.
But when MQL4++ (clone MQL5) appeared, it became clear that I had to eliminate a moron when reading "constructor", "operator overloading", etc. I started studying OOP through writing practical problems. I started to learn OOP by writing practical problems instead of the bare theory. Along the way, I asked questions, many of them were answered by simpleton. As a result, I began to understand the OOP practice. OOP for the sake of OOP is a sad approach.
Now I write everything in OOP only. Even small programs. It is really cool! But you can feel it only through solving practical tasks.
So I recommend to study the MQL5 kernel ( == MQL4++), and you can always decide on the platform number.
OOP is not a panacea. In most cases it is not necessary or even an obstacle. But there are cases when you cannot do without it. It helps make it simpler and faster
I used to think so too. And even now it seems logical to me. But I start writing some elementary stuff and I catch myself thinking that I can't do without OOP-style. I like systematization, distribution of rights, etc. That said, I'm a complete zero in network administration. Where clustering and permissions are supposed to be present. But I can't be forced to configure even a home network, not to mention others.
With OOP, even simple programs become nice and logically well-arranged. IMHO.
I used to think so too. And even now it seems logical to me. But I start writing some elementary stuff and I catch myself thinking that I can't do without OOP-style. I like systematization, distribution of rights, etc. That said, I'm a complete zero in network administration. Where clustering and permissions are supposed to be present. But I can't be forced to configure even a home network, not to mention others.
With OOP, even simple programs become nice and logically well-arranged. IMHO.