Errors, bugs, questions - page 683
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
When the market takes a tumble on some news then it's too abrupt and cool, then on M1 and M5 one has to sit down with a sniffer and catch it. This is where the expansion of consciousness begins, when you understand what the slightest inaccuracy means...
Enough with the self-delusion.
There is so much computer capacity now that there is no question of a shortage of rltime triggers from pre-prepared conditions. Any good software contains a lot of caches rather than working head-on.
We have brought the level of technical support for developers in MQL5 to a level never dreamt of in self-written or competitive software. So don't go overboard with your demands and claims.
In the first place, it's the arrogance.
Secondly, you have moved away from the issue and gone into accusations in a completely different area.
Third, you're talking about extremum binding.
Our work in MT5 on automatic identification of extrema during binding of points at higher timeframes on M1 details has been done perfectly well.
Provide a clear example when the binding did not work under normal conditions, for example on H1. And prove that the M1 history was present at that moment and was not deliberately smothered by the Max Bars parameter in the window. This is to prevent the cheating in the mode of "I adjust the minimum bars, shove off into the depths, and then complain that the search of real extremum on M1 did not work at Daily".
Besides, it is the trader's own fault when he/she builds fixing on higher periods, and then does not check its positioning accuracy on lower periods. And it's not saying that the higher is the period, the more chances to get a situation of multiple extrema, where only man can destroy it.
Magnetisation works correctly - by price proximity and you know this very well.
Fourthly, you are complaining that you are not able to find the necessary extremum with a simple function CopyRates(string symbol_name,ENUM_TIMEFRAMES timeframe,datetime start_time,datetime stop_time,MqlRates rates_array[]) on М1 timeframe.
1. You're confusing the insolent "I want it all!" - in self-interest with the well-founded audacity to make a suggestion and improve the product for all users, not for yourself. It's no longer a matter of insolence, even, but an old and tired question of a particular application. If you bend to your demagoguery about impudence and specificity, your ServiceDesk shouldn't function at all, because every such "insolent" tries to get in there with his/her own particular needs. And there doesn't seem to be any voting there, it's no place to collect other people's votes. You don't wait for the opinions of others, if you personally, the developers, the idea seems inconvenient; you simply close the application, even if nothing has been done on it. It's strange that the world has come to a head with me, and publicly. If you responded to everyone in ServiceDesk in this manner, you would not have any interaction with users, you would stew in your own juice and get bored with reverse-engineering competitors' products.
To once and for all close the subject of particular, I'll say this: yes, I'm a private developer, no one helps me (even with money), I originally write for myself, but... I was going to put the release to free access, as you will be aware, Renat (with some formal conditions). But the release date is postponed again and again indefinitely for the same reasons... The frequency of use of a product is determined by the number of downloads, for example. That would show whether it is a private or public need. It is short-sighted to judge ahead of time. What about caches, pre-calculations, optimization... I bet that even for crude and resource-intensive markup indicator, there is already an army of grateful hand-to-hand traders (unless I have invented a wheel that I will get poked with a feather in the first few days).
2. Sabject: precise positioning of graphical objects at extrema. But since I'm well aware of the fact, that this is just a private specificity, it will not be interesting for everyone; therefore, being well aware of this narrowness, I started with hyperonym - question about giving exact time to hi & low bars, which is interesting to a greater number of people sympathizing with this problem... and then you can set other tasks, other than the task of positioning objects, on this...
3. About fixing - there was no bullshit. Forced to clarify... Now this 5% could not be reproduced. And let me say at once: there's probably never a problem with the actual positioning point, apart from the left-hand snap instead of the more logical right-hand one. (If I happen to notice it, I'll write here right away.) However, I can easily demonstrate how some of the already anchored points of the same object jump off during positioning of not yet anchored ones:
NZDUSD, D1, Daily Equidistant Channel:
This is what has now come to hand. But do not think that there are no binds on H1. They do occur on practically all medium and high TFs. The essence of the problem is the following: after the accurate stretching of the base by two points, we proceed to positioning of the third - the upper - point. As soon as we have put it on hi, the left bottom point bounces and has to be re-positioned. This does not happen in all cases, but in some. There should be no issues with the M1-history depth, there is nothing wrong with it, check with yourself and see the essence of the problem. (The picture shows exact and final positioning of the channel after all manual adjustments). The number of bars in the terminal is Unlimited.
4. a.) "Besides, the trader is guilty himself when he builds bindings on higher periods, and then does not check their exact positioning on lower periods". The only thing the unfortunate user can be guilty of is not guessing about the existence of a boundary separating the zone of real minute bars and fake ones... or simply not knowing where it is. Ah, well, yes, about limiting the bars in the terminal - also true. But in this case the experiment satisfies the initial conditions from the previous point. The function of precise auto-positioning of objects along extrema is clearly stated in the forum of the "annals" of MT5 builds. If it is stated, then it must be done, without shifting the blame onto the trader who has absolutely different objectives and knowledge in his/her mind.
4. b.) "...The higher the period, the more chances to catch a situation of multiple extrema, where only a human can solve it" - Renat, I didn't expect to hear this from you as a platform developer. Everything, that "can be handled only by a human", turns out sooner or later to be automated, what we all are doing here, in fact, for many years. Once upon a time, man could not think of any other trade than manual... The thing that you have so obediently tried to solve is more or less elementary, which has been implemented in my indicator. There's nothing to solve, it's a schoolboy level problem. Do you want to dig into the template with the objects generated by the indicator? - I am ready to provide it. Everything there is exactly right-handed, exactly as intended. But it doesn't cost anything to resharpen it for left-handed extremum-twins.
5. AboutCopyRates - thanks for the tip, I'll have a look... although it's not certain that it will give a performance gain.
I also understand perfectly well: it's exciting to watch the native boat being rocked, but it hasn't been private for a long time, it's full of people who don't want to sink or stand still. Or did you expect not to be rocked?
And about the radius... - they told me about it at school. It still brings a tear to my eye.
Renat:
...Also, it's the trader's own fault when he builds bindings on higher periods and then doesn't check their positioning accuracy on lower periods.
...
Thanks for the explanation.
Maybe it makes sense to add a command on the chart "move to date/time"? It's kind of slow to scroll.
Would it make sense to add a "jump to date/time" command to the chart? It's a bit slow.
Look at the old quick navigation feature - you can specify dates and periods.
It's about 8 years old, I think.
Look at the long-standing quick navigation function - you can enter dates and periods in it too.
It's about eight years old, I think.
My experience clearly indicates that filling in the blanks is nonsense and self-delusion, which will immediately reveal itself once you get that filled in story.
This issue has been raised many times over the last 10 years.
Renat, twenty-five again.... OK, I'll say it again. The problems you're talking about haven't gone anywhere in those same ten years. I will list them (that I remember) so that some of us don't fall into self-delusion.
1) Distortion of indicator readings. For most indicators the insertion of missed values into a calculation period will result in different output values. And the correct values are obtained when the buffer is full, and not when it is empty. Examples need to be given, or is it obvious? There are exceptions, like a zigzag, in which the point of extremum is unlikely to change even in time, because at the extrema trading usually lifts up, not hangs. (Although there are exceptions even from this exception!) Now tell me how to get the indicator calculation in a simple way with a self-filled buffer, and how, when applied to a chart, it will be synchronized with this chart?The answer -no way! You don't give me access to the history, so I can correct it once, and after that I can display it filled in the "empty" places. There is only one, unusually crooked choice: after the calculation of correct values - to remove the added bars again (from the buffer of the now calculated indicator!), in order to synchronize it visually with the readings of kotir and other indicators in the same chart. Wonderful technology, isn't it? Can you offer better technology ?????
2) If in the previous paragraph I wrote about all indicators, in this paragraph I will focus. So, the multi-currency indicator will be correct even in case of full time synchronization of input quotes. In other words - if we catch a missed bar for a usual single-currency stochastic, we have distortions of readings within the period length of this stochastic. On later (and earlier) bars it will read correctly. It's not so with multicurrency - if the history of entry quotes is asynchronously distorted (that is the medical fact of 99.9999%), then the indicator is unable to show anything useful.In other words, if we can accept a leaky quote for the single currency trading (referring to the "insignificance" of the difference in indicator readings), then for the multi-currency trading the creation of any indicator necessarily requires synchronization of the time of input bars on all input symbols. Otherwise the glitches will be catastrophic - the indicator becomes not only "inaccurate", it becomes inoperable. To display the only multi-currency indicator on the chart, you need to do the following: (1) synchronize the mql5 history (unpleasant and slow procedure, because the hardly detectable glitch is easy to catch). (2) Read our indicator. (3) Remove some of its readings from the buffer - in general - it depends on which particular hole in the chart we are going to display it.......... That's how it is done on the coolest multicurrency terminal called MT5, on the development of which a group of brilliant developers (without irony) worked for eleven years (if other versions are considered the same terminal). Bravo, right, Renat?Right, bravissimo. So, at this very point it is appropriate for you to admit that in your terminal(except for the tester - there's no denying that), there are no services to facilitate multicurrency trading. Not only that - there are enormous obstacles. Which are not limited to the topic of indicators, by the way.
3) Glitchy chart structures on the chart. Do I need to explain? If we restore the "empty" bars, even the slope of the trend line will change, it's obvious. I personally do not perform visual analysis of quotes using graphic terminal tools, but I do not deny that this approach can be successfully applied in trading. In my case the trend slope calculation is performed in an indicator buffer and therefore it is reduced to the first point of this prolonged post.
--
And now explain me, dear (sincerely!) Renat, in which of the above three (enough for now) points I made a mistake, I screwed up and fooled myself.
So that I am forever healed about the missed bars.
Those ones, in which the price was real, and there weren't just market deals causing price changes, which were enough reason for you to remove them from the history. The wonderful motto "no ticks - no bars" looks doubtful to me, when I cannot use my usual tools to display (and calculate) quotes for a symbol at any time during a trading session. I don't request (god forbid. In this whole post I don't request anything!) and I don't even ask to store them on a disk, there's no need for that. And there's no need to transfer them over the internet either. Your ancient motto "no ticks, no bars" perfectly describes the technology of economical storage of quotes on disk. But for calculations I need access to the full price history, not "thrifty".
And now explain to me, respected (sincerely!) Renat, in which of the above three (enough for now) points I have glitched, pounced and self-deceived.
May I be eternally healed about the missed bars.
You tell such platitudes as if someone doesn't know them.
I'll explain, since you ask.
You are just spoilt by the liquidity of forex and don't look at the general principles of bar construction. Look at the jagged and sparse quotes of illiquid instruments. The market and others do not personally care about your idea of a proper bar time scale and its continuity. The general scheme of operation in all (roughly) analytical constructions is "no prices - no bars".
To close the problem of missing some of the minute bars, my advice is to move to higher timeframes. At least M5 or higher, since you are using them for your technical analysis.
It's ridiculous to sit on M1, build trends on it and ask others "well, where am I fooling myself?".
Since this crutch has been known for a long time, even on MT4 there are technical solutions for brokers (used) in which there are exactly as many minute bars as there are minutes in the opening/closing interval of a trade.
Doing bar synchronization before multicurrency analysis is not a problem, if once it is thoroughly solved. But it would be better to spend computational resources not on bypassing blind stubbornness.
The crutch is not in the holes but in the synchronization of different FIs. A new bar should be formed using a time model, not a tick model.
P.S. I think many will agree with me, Renat's answer is outrageously illiterate. He has a very primitive understanding of all the stages of writing and using trading strategies. You should be ashamed of yourself.
P.S. I think many would agree that Renat's answer is utterly illiterate. A primitive understanding of all the stages of writing and using trading strategies. You should be ashamed of yourself.
I got used to it that traders do not want to think beyond a one-way street.
"Unwillingness to change" is not "ignorance of the problem", it is "a repeatedly thoughtful decision to leave the standard model as changing it will bring many times more problems".
ps: A young politician beautifully stigmatizes the system, throws slogans one-way, and on coming to power realizes how naive and stupid he was (we believe that the politician is honest, from the people). Then all the same compromises and decisions, incomprehensible to others, continue. Such is life.
... A new bar should be formed using a time model, not a tick model.
P.S. I think many would agree, Renat's answer is extremely illiterate. He has a very primitive understanding of all the stages of writing and using trading strategies. You should be ashamed of yourself.