Wishes for MQL5 - page 18

 

By the way, concerning the charts, I'm using a 1 to 1 price scale, automatic adjustment of price scale on the chart is not very convenient, but changing the price scale directly visually is very convenient, when you see the result of this action, if we talk about different periods, the density is very important, if it is attached to a period. In addition, the chart limit in MT4 ends at the reached price and time of the whole chart, it is quite inconvenient when drawing objects, because they appear hidden. The chart 1 to 1 is convenient to everyone except for the lack of change in the price density, under present conditions the vertical price scale is connected with the horizontal time scale. Besides, the automatic shift doesn't take into account the vertical price ratio on a 1-to-1 time scale. Also, it would be better to make the indentation of the current time from the end of the chart variable, because different instruments may require different indentation, especially when the screen resolution is large enough. The division of price and time scaling should be set separately or together. I would also say that the scaling itself is very limited, with only a few scaling options, and that fine tuning or adjusting these options gives a choice or complete identification of the percentages.

If you look even deeper, the use of alpha channel and antialiasing, can improve the quality of rendering the same lines using for example a tenth of a pixel, or rather a tenth of a colour, quality will catch the eye, although this is rather for gourmets graphical interface, such as me, you can do without it, but I can not do without it, I like to stamp quality interfaces (.NET 3.0 Presentation Framework):) I'm certainly not saying anything, MFC may have already implemented something similar at the moment, I haven't looked into the latest innovations.

 

I support xnsnet in terms of "there is nothing better than .NET/Visual Studio (for application software) yet".

Not everyone has the capabilities of Microsoft, and we are unlikely to see a "miracle" :)

Editor.

If we "reach" the level of C++ Builder 6/Delphi 7 it will be a miracle.

Dear developers, please do not forget about outlining/collapsing. (You promised to read!).

Language.

As for the language, frankly, if I were you (I know, I know - not everyone is up to this point :) ), I would copy C# completely.

Whatever C++ experts say, "OrderType.Buy" is more convenient than pure "OP_BUY" (100%, they also have "classic" main menu, like in Win95/Win98). Even if it is 2 times longer.

Enum in C# style is so trivial and convenient that one is surprised that it did not appear 10-15 years ago. (If it did appear somewhere,it was not widely spread!

Gentlemen developers, when saying "CLASSES", don't forget "PATTERNS"! I think, if it doesn't happen, a lot of requests on this topic will be sent at once. Of course, there are no classes/structures yet - what templates are there ... But as soon as they appear, the question will arise almost immediately. The language seems to be designed for a programmer, and templates in programming is not the highest level.

Miscellaneous.

The event window graph - something like a descendant of TWinControl/CWnd/System.Windows.Forms.Control/System::Windows::Forms::Control - oh-oh, that would be just cool!

And anyway, in between...

Metaquotes corp. doesn't aspire to use any third-party .NET or anything like that. We won't see a company diversification, some other software with label "Made by Metaquotes". Or at least "Powered by Metaquotes"...? :)

Regards, pxx

 

Templates are something in C#, though it was a long wait for this something, 2 years, and another three years for the implementation of everything that was written in frameworks, including the 3.0 specification.

In general, MQL4 is very similar to C#, and unfortunately it's not the first version.

And templates in C# are really something, I've never seen such a usage of templates except in the Doka and in my own code, but there are much more ways of implementation than one could imagine, in almost everything:

[DebuggerDisplay("Count = {Count}" )]

public abstract class ClhList<TList, TItem>: IList<TItem>

where TList: ClhList<TList, TItem>

where TItem: ClhItem<TList, TItem> {

}

Actually I'm still amazed when I look at C# specification and when I check code in debug disassembler my opinion is not the least bit distorted:)

But I'm not going to spoil the argument:) I doubt that MQ will outbid someone, although there are those who suffer, which sell the client alone as I remember for $ 500 not to mention the cost of software for DC, they certainly provide the opportunity to use .NET, but development in such price frames, oh what an easy task, especially in our country, where everyone just cream on the word free. So let's consider that the competitors will survive themselves:) For some reason here I'm on the side of MQ, probably falling for the word, and perhaps the patriot, I have not yet figured out, but something is native to them:)) So let's hope for some minimal development, implementation, though not at once:) The main thing is the base and we'll drag the rest.

 

It's already been in another thread. I'll say it again.

1. The Expert Advisors must be able to disable the investor password, because having the investor password, it is possible to duplicate the Expert Advisor's work without restrictions. This password must not be compromised by the client. Block other possibilities to clone the work of the terminal, if there are any.

2. Add a function informing the terminal version.

3. Add a function that prohibits viewing external variables.

4. Allow experts to unambiguously identify the terminal by its digital signature. Based on the hints, the terminal already has a digital signature. It must be accessible from the Expert Advisor. Then the digital signature of the terminal can be written into the expert's license.

5. Give the possibility to identify the DC by digital signature. This will increase protection against fraud.

6. Add http support to MQL.

To reduce the risk of fraud, it is desirable to form a public depository of digital signatures of servers (terminals). Ideally, each terminal should be able to turn into a signal dispenser, like a mass mailing of icq messages.

 
Improve the order setting function. So that there is a set task for its execution with an accuracy of a point. And if the order triggered, but with an accuracy of +/- 3 points, then this procedure itself should set the necessary StopLoss and TakeProfit from the actual opening price with an acceptable periodicity. For example 14 points Ticeprofit. Price + 14 points = TakeProfit. However the actual price is not equal to the requested price, but equals Price+3*points (slippage). Therefore we need to correct the Take Profit and Stop Loss after the order is triggered, correcting for the actual open price. We can do it programmatically right now, but why? If we can trust the function that is independent of the program execution. And, importantly, brokers will not mind, because you will set it up acceptably yourself.
 
pxx:

I support xnsnet in terms of "there is nothing better than .NET/Visual Studio (for application software) yet".

.....

This "wonder" is nowhere near as good as PHP.
 
Andy_Kon:
This "miracle" isn't even close to PHP.

Uh-oh, Personal Home Page Tools rule the world! )))))))))))
Or, by the new name of PHP: Hypertext Preprocessor!
Scripting languages forever!

Quote: PHP was originally created as an add-on to Perl to facilitate web page development.(https://ru.wikipedia.org/wiki/PHP).

The original "C" was originally created for a slightly different purpose. ... ))

How about calling "VirtualAlloc" or "CreateFileMapping" from PHP, anyone? :)

 

PHP as well as Perl are scripts implemented by interpreter directly from the source code, .NET is bytecode architecture for multiplatform and multifunctionality, multilanguage, on linux its counterpart Mono implementing multiplatform on almost any axis except RTOS. The differences are quite significant and the objectives are different, as well as with PHP it is hell to write a program with a user interface, as well as in C/C++ it is hell to write for example a web application, service. We need something more and it is just more like .NET, however, also Java is available, but with Java it is more complicated, although at the time it had no counterparts, Flash also was not created for nothing, and then .NET is there, eventually Silverlight appears, not to mention the XBAP. You'd better tell me where .NET isn't used now, and how competitive it is in such conditions:))) It's one step from simple to complex, and if you keep it simple, the complex becomes even simpler. Did you know that some people are still struggling to implement a Java interpreter to translate a program into C++, it would seem why, but there is such a thing. At all times people have tried to solve the problem of globalization of program code, to make it more suitable for any conditions, but at first people have an attachment to a chosen language, and secondly multiplatform, as a result .NET realizes the connection between all of this. There are dozens of languages that are ready to become adopters of .NET, there are many technologies that are available in frameworks, these are just wrappers, but it's all there and more than just wrappers are emerging. Now try to compare it all, with the narrow direction that you support trying to put in the comparison, I'm not saying that you should abandon it, you should only compare, patriotism is certainly good, but only for the sighted, who also sees the essence of another idea, so you should respect the possibility of choosing methods, otherwise one day you may be out of business :) Try to imagine MQL language in .NET architecture, can't you? Do you bring up security? Don't you use cryptography? How much effort do you spend to implement the same principles that have long been implemented in .NET? Comparison and comparison again, bugs and holes must be compared too.

Because I've gone in the direction of .NET I've lost a lot, circle of friends and acquaintances (now I simply have nothing to communicate with them about), a lot of time, but even more I've gained, the main thing is development:) In cycling, the main thing is not to invent a new one:) I apologise for the brainwashing, but I can't do without it:)

 
pxx:
Andy_Kon:
This "miracle" isn't even close to PHP.

Uh-oh, Personal Home Page Tools rule the world! )))))))))))
Or, PHP: Hypertext Preprocessor!
Scripting languages forever!

Quote: PHP was originally created as an add-on to Perl to facilitate web page development.(https://ru.wikipedia.org/wiki/PHP).

The original "C" was originally created for a slightly different purpose. ... ))

How about calling "VirtualAlloc" or "CreateFileMapping" from PHP, anyone? :)

1. Which programming language is better and worse is a rhetorical question.
2. A small request to look up the PHP features at this addresshttp://www.php.net/manual/ru/, so that you know what it's about.
3. Unlike many others, it absorbs the best (from C, Lisp, Perl and others).
4...Scripting languages - if memory serves me correctly (and it never does) the progenitors of VB, Delphi and many others were.
5. And Delphi has its roots in Pascal - now dead.
6. Regarding ...calling "VirtualAlloc" or "CreateFileMapping"... - Does Delphi have rar_close? :))
7. .NET - ask who loves the little softies?
 

7. .NET - ask who likes smallsoft?

By the way...
The main thing is that mcl5 doesn't drown in Wine ;)))