MT5 is for programmers, not traders - page 8

New comment
 
Alexander Puzanov:

:)

1. CopyHIgh may return an error - this has to be checked manually and handled. At least 3 lines

2. CopyHIgh may not return all values you give it - this needs to be checked manually and processed. At least 3 strings.

3) To use CopyHIgh you have to prepare an array where Copy will do it. At least 1 line

4. To realize CopyHIgh benefits we need another stack of strings. With manual error checking, of course.

---

I envy who has only one line of complication.

Is it okay that errors are possible in mql4 and they should be handled in the same way?

Return value

The value of maximal price of a bar (specified by the shift parameter) of a corresponding chart or 0 in case of an error. To get more information about an error, call GetLastError().

And the fact that you have to declare a variable to store this value doesn't count either?

And the fact that in mql4 you have to write as many strings as you need to get several values does not count either??? So show me what are the advantages of iHigh mql4 compared to CopyHigh, if we need to process the last hundred bars. Array again? Another loop? Or a primitive hundred variables?

And functions such as 

int  ArrayCopySeries( 
   void&  array[],           // массив, переданный по ссылке 
   int    series_index,      // идентификатор массива-таймсерии 
   string symbol=NULL,       // инструмент 
   int    timeframe=0        // таймфрейм 
   );

и

int  ArrayCopyRates( 
   void&     dest_array[][],    // массив, переданный по ссылке 
   string    symbol=NULL,       // инструмент 
   int       timeframe=0        // таймфрейм 
   );
have you ever seen? I think this is the first time you've seen one.
 
Sergey Vradiy:
There is a class of software called expert adviser generator. You can build any trading system, put together an algorithm by visual bricks (condition check, variant branching), choose ready-made indicators, generate code, modify it, etc. You can analyse trade statistics (Sharpe ratio, expectation, etc.). There are programs that allow you to approximate the TS by neural networks to a ready-made set of manual trades. There is everything. You don't have to be lazy to look for it.

Thank you, lots of interesting stuff!

 
Sergey Vradiy:
There is a class of software called expert adviser generator. You can construct any trading system, put together an algorithm using visual bricks (condition checking, variant branching), choose ready-made indicators, generate code, modify it, etc. You can analyse trade statistics (Sharpe ratio, expectation, etc.). There are programs that allow you to approximate the TS by neural networks to a ready-made set of manual trades. There is everything. You don't have to be lazy to look for it.

What I've seen is nothing more than generating a template for later refinement

have you seen real diamonds from this class? so that you could generate them and not be ashamed to go straight to market ))

 
Alexey Volchanskiy:

What I've seen is nothing more than generating a template for later refinement

have you seen real diamonds from this class? to generate and not be ashamed to go straight to market ))


All this from the fact that there are a lot of utopians here and believe in all sorts of nonsense.

 
Alexey Viktorov:

...


I won't answer you, no offence.

---

fxsaber:

Try me.

Thanks, but I'm not an 'adept' :) Plus, I'm too often required to write human-understandable code with comments on who went where. And with your constructions you can't even send me to the reference.

As far as I understand, the standard library was supposed to be an "automatic" version of MQL5, intended purely for traders. So that e.g. finding an extremum (which is what TS wants) would require 2 operators. If it was done in time, there would be much less creaking when moving to 5. But it looks like this idea is dead or never existed.

 
Alexander Puzanov:

Thanks, but I'm not an 'adept' :) Plus, I'm too often required to write human-readable code with comments on who went where. And with your constructions you can't even send me to the help.

As far as I understand, the standard library was supposed to be an "automatic" version of MQL5, intended purely for traders. So that e.g. finding an extremum (which is what TS wants) requires 2 operators. If it was done in time, there would be much less creaking when moving to 5. But it seems this idea has stalled altogether or did not exist at all.

The SB implementation is a black box. To use it, there is no need to take apart how it is implemented. My code is just an example that MQL4-style is technically feasible and can be implemented in an mqh-file that does not need understanding at all. One indicator and it will work as in MQL4. So, there is no need to talk about complexity. Transition from "complex" to "simple" is solved by one line.

 
fxsaber:

The SB implementation is a black box. To use it, there is no need to parse how it is implemented. My code is just an example that technically MQL4-style is implementable and can be designed in an mqh-file, which does not need to be understood at all. One inluder and it works just like in MQL4. So, there is no need to talk about complexity. The transition from "complex" to "simple" is solved by one line.


Here are two points:

1. 1) "pure" transition is not possible with only one mqh - at least the indicators call should be changed.

2. The way I see it, if the standard library was less prevalent on the forum, the understanding of mql5 would have been easier and faster.

I actually don't understand how one can say "there is nothing complicated in mastering mql5", if at the same time, in every corner and from different belfries, the standard library is being hammered into the masses -- tons of examples in KB with errors, with questionable code - but with the proud look and words "the standard library only, as a matter of principle".

How can anyone explain, explain and communicate anything on a "black box" at all.

 
Alexander Puzanov:

I won't answer you, don't be offended.

---

Thanks, but I'm not an 'adept' :) Plus, I'm too often required to write human-understandable code with comments on who went where. And with your constructions you can't even send me to the help.

As far as I understand, the standard library was supposed to be an "automatic" version of MQL5, intended purely for traders. So that e.g. finding an extremum (which is what TS wants) would require 2 operators. If it was done in time, there would be much less creaking when moving to 5. But it seems this idea has stalled altogether or did not exist at all.


Take the example of petr who does not like oop
 
Alexander Puzanov:

I won't answer you, don't be offended.

---

Thank you for your candor. You'll excuse me for being emotional, too.

 
Andrey F. Zelinsky:

two points:

1. The "pure" transition with only one mqh will not work - at least the indicator call must be corrected.

By the word in quotes I meant the MQL4-problem when writing or rewriting MT5-programs. Of course, MQ4 -> MQ5 will not work through copy-paste. I think simplicity has been discussed. It's technically achievable for a long time, but for some reason it's not being implemented.

2. The way I see it -- if the standard library was less hype on this forum, the understanding of mql5 would be more accessible and faster.

I actually don't understand how one can say "there's nothing complicated about mql5", if at the same time, at all corners and from different benches, one could hawk the masses about the standard library - a lot of examples in KB with errors, with questionable code - but with the proud look and words "only the standard library, fundamentally".

How can you on a "black box" explain, explain and convey anything to anyone at all.

I agree, SB trading (at least) part is extremely unfortunate and I started watching it only after studying MQL5. Learning MQL5 from it is one of the main demotivators. However, SB is forcibly injected in both documentation and kodobase and on the forum.

123456789101112131415...23
New comment