MT4 doesn't have long to live - page 25

 
MetaDriver:

No problem.

https://www.mql5.com/ru/job

Any whim for your money.

;)


:-) I'm sorry... I'm capable of a lot of things myself...
 
faa1947:

The point is a programming bias in the development of the terminal. The conversation started with the tester, which the developers proudly call a "strategy tester". I and some others in this thread claim that the tester doesn't test any strategies - it's a cool specific program debugger. As an example, I gave that the tester doesn't answer the question about stability of TS. And other examples.



Start with yourself, i.e. answer a trivial question: what do you think is a stable trading system?

 
faa1947:





Possibly. So join the methaquotv team - and make your own constructive additions.
 
Reshetov:
Start with yourself, i.e. answer the trivial question: what do you think is a stable trading system?

I answer questions if I can't find an answer in books. If you like, I can quote you, although I think I did it for you personally.

In the "Econometrics: one step ahead forecast" thread I not only answered the question, but also cited results from the field.

What is your question about?

 
IgorM:

No, I am generally satisfied with the mt5 system and I don't want the mt5 project to stop, of course there are difficulties with the historical data, but not about that

If in the Strategy Tester there was a ticker "use MT4 compatibility mode" which would set two opposite orders with 0 volume instead of closing a position at opening a counter order, then maybe the problem would disappear.00, maybe there would be no questions from inexperienced users "why lots don't exist, bring back lots!", I don't have problems with MT5, well, almost none...

The point is that while developing complex systems it's critically important to keep architecture clean, without letting crutches kill it. A small crutch in one part of a complex like a terminal starts to sprawl and grow, killing the server architecture and operations processing. Not everyone understands what it means to combine two mutually exclusive trading models in one system. However, it is enough to sit down, write down all the processing on paper, and then be horrified by the consequences.

This is why we have rewritten our systems 5 times from scratch, although it is an extremely dangerous method. The result is either to perish or to make a cool thing.

 
jelizavettka:

Possibly. So join the methaquotv team - and make your own constructive additions.
Thank you for trying to employ me.
 
sever31:

I don't care about the dick.

Why does MT5, a priori, equalise two differently directed positions, of the same volume, of two different TS? Do the developers know my MM, trading logic of systems?

If you have a trade system in place, then the same order will be placed in the same position in the same volume in the same trader.

Hmm, I don't want to explain elementary things, but all you lose in netting is only the price and volume of the opening of the lock order and the main order, replace closing a position in MT5 by placing orders with a volume of 0.00 and I suggest - in principle your TS will not lose anything at all, will be visible in the same way where the order was opened and what type of order would (Buy / Sell), the total position in the market will be equivalent to a lock, and not to change the logic of the trading servers, I suggested to organize it only on the terminal side, in fact it is just an illusion for people who do not want to analyze the history of the opening / closing orders, and do as done in lot TS - just put a lock
 
faa1947:

I answer questions if I can't find an answer in books. If you like, I can quote you, although I think I did it for you personally.

In the "Econometrics: one step ahead forecast" thread I not only answered the question, but also gave results from the field.

What is your question about?

Once again for the particularly gifted econometricians: give a definition of a stable trading system. We, illiterate traders, still do not know what it is, since we have not seen such systems in our eyes, including in the above-mentioned thread.

They have seen unstable ones, but there are always some problems with stable ones: either they have drawdowns or Kolya gives them a peek.

Therefore the question for us is not an idle one. Therefore, we ask this econometrician so well versed in books to give a clear and unambiguous answer to this very small question.

 
faa1947:
Thank you for trying to employ me.
Well, that's great. I think you'd like it.
 
faa1947: In EViews it is called "model" and in the model it is called "scenario".

Very curious. What do you think the model and scenario for the next two systems will be, at least approximately:

  • a two-ball system, working through crossovers,
  • system based on Fibo levels.