Why are there so few experts in the MQL5 database? - page 3

 
sergey1294:

I have an expert written without classes, but I can't show it to you as it's written for commercial distribution. but it's possible to write without classes, but it's not possible to use standard structures such as

and some more

A structure is the same class but without functions. I told you that you cannot do it without classes.

 
sergey1294:
... I don't want to put my MQL5 Expert Advisors in Code Base and I don't think many people who use it now want to.
+1
 
Yedelkin:
+1
Well then the question is why those who write on MT4 have such a desire, and those who are already writing on MT5 do not have such a desire?
 
joo:

A structure is the same as a class, but without functions. I'm telling you, you can't do it without classes.


Maybe I'm not familiar with classes yet, so it turns out that you can't write indicators and scripts without classes either.
 

Figar0:

In order to promote the platform (which is undoubtedly part of MQ's plans), it needs to be popularised among traders. The main advantage of MQ's development is the availability of autotrading, and if there is no autotrading available, there will be no popularity and nothing else. Who is the easiest to promote it to? Certainly among people already familiar with the medium. And most of them are still friends with MQL4 precisely because of the complexity of the transition... The use of custom programmers is not an option (I can assert it, because of my rather limited experience with MT4).


And the platform is not popular among traders, may be autotrading banned?

If many people are not aware of it, I would like to inform that the developers of this new version of the platform had two main goals: Introducing MT to new markets (the main goal) and increasing the abilities of MQL language.

Both of these goals are being achieved.

Figar0:

Z.U. How hard is the transition? More than difficult so far... My example, technical background, I used to program in perl, asm, foxpro, html, php and a little bit in C, java, VB. When I got interested, I quickly and easily mastered MQ4 (I wrote my first Expert Advisor on the first day of the acquaintance). But this is the fifth approach, and I've progressed very little.... But maybe I'm not so lucky yet:)

How difficult is the transition? And you have to decide whether you personally need it or just stay with the MT4.

My experience is also "amateurish", I know all languages I've mentioned above, and I can add Pascal as well.

Unlike many other EAs, I'm not writing 100-line DEMANDED EAs, but for already six months I've been developing a code library, which I will use later as a constructor in development of EAs of any complexity. Most of the code in this library is devoted to the issue of migration from MQL4 to MQL5. I'll tell you right away, to avoid questions why it takes so long - I develop everything from scratch and don't use ready-made solutions (classes, structures, etc.).

The idea of MQL5 was described in the article, and I want to share my experience with all my friends, and assure them that the transition from MQL4 to MQL5 is not more complicated, than from Pascal 7.0 to Delphi 7.0 (which, in the opinion of many, was the most successful implementation of the Classic Edition).

The point is to find a good description of what is: Modular (Unity) Layout, what are objects and classes, what are data structures and finally, what are events and methods (all this is of course well described in the help, but who is not enough can easily find it all from other sources).


PS

I'd like to add another comment to my previous posts about SIMPLE and maximally accurate rewritten MQL4 Expert Advisors...

The time when you could write an EA trading with two МАшиши, and all MM of which was based on lot size selection, has passed. All who have not understood it sooner or later will be removed from THIS MARKET (good if not at the cost of a TRADING DEPOSIT).

I am still amazed how one could talk about effective EAs (especially with neural networks), when often their size was 100-300 lines of code (and there was no effective system of multicurrency operation and testing)...

:)

 
Figar0:
That's why there are so many indicators which don't differ much from 4 Indicators, while experts and scripts are very few and far between?

What will programmers eat if everything is in the public domain? When August comes (maybe even later), the SHOP will open, then come (especially the impatient can use the WORK)...

:)

PS

As to examples of MQ Expert Advisors - there are my posts with suggestions, those who are interested can find and see what and how ...

 
Figar0:
Why those who write in MT4 have such a desire, and those who are already writing in MT5 do not?
If you're an expert, you need time to develop it, and in addition to that the programmer has some other things to do and also his/her private life. i have just counted it, in 2006 15 MQL4 Expert Advisors and 60 indicators were posted in total for six months. moreover Expert Advisors were posted starting from 2005.11.09 and indicators from 2005.09.15, two months later the first script was posted in 2005.09.16 and about 10 articles were posted in total for six months
 
Figar0:
Well, then the question is, why do those who write on MT4 have this desire, and those who already write on MT5 do not?

I've never written on MT4, so I can't compare the desire and the reasons for it.

 
sergey1294:
Maybe, I'm not familiar with classes yet, but according to you it's impossible to write indicators and scripts without classes as well.

I've written before:

joo:

Without classes, you can do almost any complexity:

scripts - yes

indicators - yes

However, you cannot make an Expert Advisor without classes. Maybe that is why there are so few of them in the code base? Most people do not know OOP very well. If I am wrong, please show me an expert without classes.
 

joo:

Asked for an example of an expert without classes. Can you give an example? I would be very grateful.


You can easily do it yourself if:

1. Dig around on the adjacent forum for a not very complex Expert Advisor;

2. Forget about how to explicitly apply the contents of directories: Arrays, ChartObjects, Charts and others (which make our life easier);

3. Forbid yourself to use the Timer, and other charms of life appearing in MQL5.

4. Forbid to create classes and structures that are not explicitly needed in the Expert Advisor and not included in the core language (for example, MqlTradeRequest and MqlTradeResult).


PS

Of course, I could give a ready example of an Expert Advisor that does not include all of the above, but I would not do it, because I don't think it is necessary (besides, in this case I would have to implement everything without using my experience).

And I don't have such Expert Advisors in pure form now (it's been a couple of months), because I use OWN CLASSES and STRUCTURES in them (I'm not a complete idiot, so I can't refuse benefits of MQL5)...