Organising the order cycle - page 7

 
Alexey Kozitsyn:

Then maybe we should try to select orders as quickly as possible (just select them!) and write them into an array, and then, in a separate function, check the availability of those orders + the necessary action (close/delete/modify)?

I see the solution myself, as I have stated before in the discussion that has already been closed. But in this case, it's not my own solution I'm interested in, but the way others do it.

Basically, realized that they don't bother with this topic, and if something did happen, it will wait until the next tick.

I don't think this thread is the best place to discuss it though. This thread is for features.

That's up to the moderators to make up their minds.

 

All in all another Star Wars level sci-fi. Not real, but any sci-fi becomes reality at some point, like Engineer Garin's Hyperboloid...

If that's the kind of execution a broker has, then it's not worthy of attention at all, let alone parsing possible situations...

 
Alexey Viktorov:

All in all another Star Wars level sci-fi. Not realistic, but all fiction becomes reality at some point, like Engineer Garin's Hyperboloid...

If a broker has this kind of execution, then it's not worthy of attention at all, let alone sorting out possible situations...

I don't understand why this can't happen on a perfect broker? It's not a bug, it's a market situation.

 
fxsaber:

I don't understand why this can't happen on a perfect broker? It's not a bug, it's a market situation.

In my opinion it is not a real market situation.

Well, let's say... Everything was going smoothly, calmly... How much was the change in price at the time the re-indexing took place?

And in general, what do you mean by that expression? This question is to talk about one thing. To have the same understanding of what is happening.

And what happens in the case of a gep??? What difference does it make what level the order was at? What's +1 or -1 point. I've read a lot of your run rate measurements, but never tried to figure out, much less remember which ones even exist... According to your idea, how many orders should be in a loop so that between ticks a normal broker does not have time to process all the orders? Do you have any real measurements?

What is the sense of this fuss? Well, they did not have time to do it in one cycle on one tick, they will do it on the next tick in the next cycle. Where is such a precision needed? In my opinion, only theoretically to roll out some other bug. Sorry, but the impression is that you are not trading, but looking for bugs in MT.

That's it... I'm shutting up about it.

 
Alexey Viktorov:

How many orders do you imagine there should be in a cycle so that between ticks a normal broker does not have time to process all orders?

Even one might not make it in time.

The normal speed of MT4 server processing an order is around 100ms. It might be 50 ms, but it's rare. This is without pings, pure broker speed.

And in 100ms there can come much more than 2 ticks.


Alexey Viktorov:

Where do you need that kind of accuracy?

In any strategy. To increase its profitability.

It's just like saying that 25 cents commission per lot is so small that it can be neglected. You can, of course, but at a turnover of 1000 lots it will reduce your profit by $250.

If you do not know what to do with it, you may be surprised by how much money you can earn from it. Otherwise the forum would still be discussing optimal MAs periods for finding crossovers.

 
Andrey Khatimlianskii:

Even one may not make it in time.

The normal processing speed of an MT4 order by the server is around 100ms. There might be 50 ms, but rarely. This is without ping, pure broker speed.

And in 100ms there can be a lot more than 2 ticks.

Agreed. My thoughts were more on the number of pips the price can change for missed ticks. Even the intraday gaps are not that big. If you have a profit/loss on the market, you cannot measure the difference between the price and the value of a single tick.

Andrey Khatimlianskii:

In any strategy. To increase its profit.

It's like saying that 25 cents of commission per lot is so small that it can be neglected. You can, of course, but at a turnover of 1000 lots it will reduce your profit by $250.

If you do not know what to do with it, you may be surprised by how much money you can earn from it. Otherwise, the forum would still be discussing the optimal MA MA periods for finding crossovers.

This is exactly what those who want to write an Expert Advisor for free think so...

I don't mind; I've even participated in the discussion. After all, the truth is not born in an argument, but in a dialogue.

 
Alexey Viktorov:

Agreed. My thoughts were more on the number of pips the price can change for missed ticks. Even intraday gaps are not that big.

Anything from 0 to infinity. In practice, 150 four-figure pips per second is common on the news.


Alexey Viktorov:

This is exactly the reasoning of those who need to write an EA for free...

I do not understand this at all.

What does publishing free elaborated codes and discussing subtle problems have to do with ordering a free EA?

 
Andrey Khatimlianskii:

Anything from 0 to infinity. In practice, 150 four-digit pips per second is common on the news.


I don't understand this at all.

What do the publication of free elaborated codes and discussion of subtle problems have to do with ordering a free EA?

It has everything to do with that, I've even highlighted it in your text.

It will reduce your profits by $250.

In one case we believe it is wrong to pay commission or spread and complain that it will not fatten our wallet, and in the other case, thank God it does not apply to us, they believe that paying for the writing of the EA is not desirable, because it will not fatten their wallet.

 
Alexey Viktorov:

The relationship is very direct, I even highlighted it in your text.

In one case we think it is wrong to pay commission or spread and complain that it skims our wallet, and in the other case, thank God it does not apply to us, they believe that paying for writing an EA is not desirable because it skims their wallet.

I have seen it highlighted. I can't understand it.

 
Andrey Khatimlianskii:

The highlighted has been seen. I can't understand it.

Did I make it sound so complicated? The DC earns on spreads and commissions, the programmer earns on writing code.

One of the traders says that we must save on spreads and commissions (if possible not to pay).

Another trader says it is not fair to write code for money (if possible, do not pay).

Both traders take care of their wallets. They don't seem to differ from each other.

But you, as a programmer, support the first, and argue with the second proving him wrong.