Is tick history available on the server? - page 2

 
But why form the history not only by timeframes, but also by several ticks? say 10, 100, 500, 1000, etc. In this case it turns out to be a time chart reduced to volume. You cannot deny that the volume is a very important indicator for analysis. It will be very clear. Besides, you can form the history by the number of ticks or by the time indicator, as you need. The accumulated data can be shared in forums - who has collected what. I don't think it is difficult to provide for this programmatically. To replenish the missing data (the terminal was shut down), you can store the tick history on the server for a small period of time (say, 1-3 days).
 
You can't deny that volume is a very important indicator for analysis. It would be very visual.


Apparently the developers just need to nail up the website in large letters in a visible place: "There are no ticks and there won't be any! Discussion of ticks as well as discussion of brokers will be removed from forums" :o))). Otherwise everything will go round and round every week. Just the words of the sufferers without any detailed evidence of the need to keep ticks...
 
Can you be more specific about which ticks we are talking about?
It would be interesting to see the tick history of real accounts.
The literature is so passionate about charts that one has to think twice before using an EA for real trading.
And I haven't seen anywhere that the broker has laid out the archives of real ticks - so I start thinking, maybe there really is something to hide?
 
Now that's where you're caught. I specifically say - prove that the intra-minute modelling has serious discrepancies. Especially by 10 points. Just take the data, do the research, publish all the findings (including historical files) and nail us to the wall. <br / translate="no">.
I argue that the margin of error within the minute modelling is within 1-2 points. And this error is perfectly acceptable and normal.

Renat,
I will not prove it as I absolutely agree with you. Moreover, I am not going to pin anyone to the wall. I declare: I am quite satisfied with ticks modelling in the tester. I do not have absolutely any complaints to it.

At the same time I repeat it once again:
The need for a tick story has nothing to do with the modelling in the tester.

And all that I could say to justify its necessity I said above.

solandr,
If users are not proactive and persistent in explaining their needs to developers, both users and developers will suffer.
If developers don't show patience, understanding, but also a strategic approach when developing their products, both users and developers will suffer.
I hope you understand that.
 
...and also, we are talking about MT5, Renat, when is it planned to be released?
will it be compatible with MQL4, or it will be something new?
...Will it be compatible with MQL4 or will it be something new?
 
Now that's where you're caught. I specifically say - prove that the intra minute modelling has serious discrepancies. ... <br / translate="no"> I argue that the error within the minute modelling is within 1-2 points. And that margin of error is perfectly acceptable and normal.

You keep talking about the PRICE aspect of the market. Here you have undoubtedly succeeded and the error in the PRICE modelling is insignificant.
However, quote flow has other parameters that are FULLY destroyed by modelling, especially during periods of news movements.
In my opinion, you need ONLY tick history from which you can get ANY timeframes and not only timeframes, but also tickframes (in a candle, for example, 10 ticks) and cagi charts and crosses and zeros.
I.e. ALL artificial restrictions related to time-frames are removed.
 
In my opinion, we need ONLY tick history from which we can get ANY timeframes and not only timeframes, but also tick-frames (in a candle, for example, 10 ticks) and cage charts and tic-tac-toe. <br / translate="no"> I.e. ALL artificial restrictions related to timeframes are removed.



Absolutely agree with that. Do the developers really need to prove that tick-frames have the same right to exist as time-frames? If there are difficulties with storing ONLY tick history because of a large amount of data for any significant timeframe, then you can at least store fixed tickframes, as it is done now with timeframes. This is my wish and has nothing to do with testers, expert advisors and other nonsense. I've been trading on a real account for about a year and I only do it manually. If there is some kind of price movement it's elementary to trade with profit, but if the price is just drifting back and forth in one place, no expert advisor will help. And let the computer make the decision for itself - God forbid.
So, I think tick-frames are a much more accurate reflection of price movement, and I personally would feel more comfortable with them. Is it really so hard to do?
 
It seems to me that the best way to store history is to store it in quality minutes.
You can build any frames from them (even timeframes, even tic-tac-toe, even
to minimise them according to some other criteria).
What is really important is evenness of axis of main MT4 charts,
which are based on data collected in constant time intervals...

But it may be decided
only in the next version of MT ... Then you can see the picture at least at once
in waves (I don't write Eliot, because there are waves, but their structure is very questionable?).
In general, I am in favour of a basic story of QUALITY minutes.
 
I think nobody will argue that time is not the best variable to set the price, the price depends more on the number and volume of operations, and the tick volume is a function of these parameters (or something close to it). This is why for researches tick-frames would be more interesting than time-frames.
In any case, if to make a simple estimation on those who have expressed in this branch, there is an absolute majority of those who would like to.
 
If I were making my own terminal =))), I would do this:
On the server I would only store ticks, and only allow them to be downloaded.
On the client I would make all necessary TimeFrames (and TickFrames) from ticks.

Thus, to obtain a graph of any (any) TF for a year, you would need 35.7 Mb of traffic (Yurixx's calculations).
Now it will require about 20.763 MB (15.71 + 3.142 + 1.047 + 0.524 + 0.262 + 0.065 + 0.011 + 0.002) to download the history for the same period of time in MT4.

Bottom line, pluses of tick history:
- The user decides for himself which TFs he wants and can always build them himself.
Cons of the tick history:
- the amount of consumed traffic increases by 1.7 times (in comparison with the downloading of all TFs);
- it increases load on the processor (for the constant construction of necessary TFs).


If I were making my own terminal, I would think carefully first....