Which broker should I choose? - page 5

New comment
 
C-4:

Stops are implemented by pauses. Unfortunately, TakeProfits are virtual.

In my model both SL and TP are implemented.

Without the CCA, it is fundamentally impossible to use TP and SL at the same time.

Without a CCA we cannot speak about the reliability of even a single SL in the form of a pending order (part of a lot from a netting position).
so any similar systems put business logic at risk beforehand.

Stops were chosen as the most important element in most systems.

this is a strange approach, TA is just as important.

You can only use one of them, but not both at the same time.

there are no pitfalls or difficulties in not using both at the same time.

In the beginning I thought so too, but then it became more and more complicated. Binding orders, which is obvious to everyone, is a trivial task.

it stays that way.

I'll open up my engine a bit more ;)

The principle behind my model is simple, and the abstraction and OOP allows you to do wonders like:

#include <MQL45™.mqh>

class _EA : public CMQL45™
{

старый код MQL4

};

And the programmer does not need to change anything in the old MQL4 code. Even the indicators remain in the form in which they live in the MQL4 model.

Martingales, averaging, chaebols, simple indicator EAs - all successfully tested and fully identical to MT4 and MT5 inputs of tests and real use of EAs.

In other words, get the full-fledged porting of MQL4 code in MQL5 using the MQL45™ engine. Or use the MQL4™ library to develop MT4 order systems in MT5.

 
sergeev:

Opening up my engine a bit more ;)

Wow...
 
sergeev:

...

Opening my engine a bit more ;)

...

Get full-fledged porting of MQL4 code to MQL5 via the MQL45™ engine. Or you can use the MQL4™ library to develop MT4 order systems in MT5.

Something even made my brain cringe trying to imagine all this. Cool! ))
 

There is just one problem...

All this stuff is interesting only for our brother-progamer, theorist-researcher.

Traders (freelance customers) are far from technology, they need a solution, and they are not interested in the content of the code. all the contrivances, OOP, intricacies, assumptions that should work in some way not always right - they are not satisfied.

That's why MQL45 is sitting on the shelf waiting for a unique consumer :)

 
sergeev:

There is just one problem...

All this stuff is interesting only for our brother-progamer, theorist-researcher.

Traders (freelance customers) are far from technology, they need a solution, and they are not interested in the content of the code. all the contrivances, OOP, intricacies, assumptions that should work in some way not always right - they are not satisfied.

That's why MQL45 is sitting on the shelf waiting for a unique consumer :)

Maybe it makes sense to keep this technology not on the shelf, but in the Marketplace?
 
sergeev:

so the MQL45 is lying on the shelf waiting for a unique consumer :)

What about the reliability of the order system?
 
sergeev:

In my model both SL and TP are implemented. quite real ones with additional "MT4 order" integrity control.
Without a CCA we cannot speak about reliability of even a single SL in the form of a pending order (part of a lot from a netting position).
therefore any similar systems put business logic at risk beforehand.

this is a strange approach, TA is just as important.
there are no pitfalls or difficulties in not using both at the same time.
it stays that way.

I'll open up my engine a bit more ;)

The principle behind my model is simple, and the abstraction and OOP allows you to do wonders like:

And the programmer does not need to touch anything in his old MQL4 code. Even indicators remain in the form in which they live in MQL4 model.

Martingales, averaging, chaebols, simple indicator EAs - all successfully tested and fully identical to MT4 and MT5 inputs of tests and real use of EAs.

In other words, get the full-fledged porting of MQL4 code in MQL5 using the MQL45™ engine. Or use the MQL4™ library to develop MT4 order systems in MT5.

Cool. But again, let's postpone this argument to a later date. I see this could be an interesting conversation: solving one problem with two different approaches. It's just that until the specs are published, there's nothing to discuss.
 
sergeev:

There's just one problem...

All this nonsense is of interest only to our brother, the theoretical researcher.

Traders (freelance customers) are far from technology, they need a solution, and they are not interested in the content of the code. all the contrivances, OOP, intricacies, assumptions that should work not always right - they are not satisfied.

That's why MQL45 is sitting on the shelf waiting for a unique customer :)

Honestly, I don't really understand it. Sell your customers a ready-made solution in a box. And they don't need to know what's inside.

tol64:
Maybe it makes sense to store this technology not on the shelf, but in the Marketplace?
I completely agree with your question. It's easier to post it on the Market than to give a unique service on a freelancer. The engine code is 100% MQL based, no third party libraries used?
 
TheXpert:
How is the reliability of the order system?

There is no CCA on the server - there is no guarantee that the business logic will not fail.

After all, we are modelling based on those deals and orders that are in the terminal history. Please note that the Expert Advisor will not fail when maintaining its model.
But, as pointed out by C-4, a trade history may fail when the broker has changed it (who knows when it happens, but there is a chance).

the reliability of the order system (and its ironclad logic) is based on just three identifications

1. DEAL_ORDER
2. DEAL_MAGIC / ORDER_MAGIC
3. orderState

- The notion of a position is neutralized as such. The positions are not taken into account in the MT4 model at all.
- The identification of order filling by trades(DEAL_ORDER/OrderState) + if a SL/TP is triggered, the order filling of these SL/TP orders is done.
- The SL and TP actually set in the pending order - after it has been filled (volume is controlled) - are replaced by a stop and limit order with a corresponding ORDER_MAGIC linking the pending order with the initial order.

This is the scheme of general analysis and control of the orders

then - just a matter of technics - necessary OrderSend in the required direction + a bit of charting showing which order is placed where.
The only thing you need to consider is to close the market order from the MT4 terminal as a position, otherwise it would be like opening a new opposite one.

 
C-4:

It's easier to post on the marketplace than to give a unique service in freelance.

no demand.
my MQL45™ project is over three years old.
I think I started at the same time as https://www.mql5.com/ru/forum/3693/page2. The project was implemented back then, but it didn't go further than theory.

so far, you are the only one in the right direction :) - publish an article and get paid at least for openness and a variant of the implementation of this idea.

The engine code is 100% MQL-based, no third-party libraries used?

No.

123456789
New comment