Any questions from a PROFI to a SUPER PROFI - 1. - page 19

 
MT4 terminal in Windows-7 (64). ExpertSample.dll is compiled in Windows-7 (32) in Visual C++ 6.0. It doesn't work - error 126 is written to log. What does it need?
 

VS6 didn't have W7 in it yet. You need a new file with the system versions prescribed.

It's better to install VS 2010. It's all there.

 
I'll give it a try. But shouldn't compatibility be ensured?
 

I was able to use this dll in MT4 and MT5 on Windows-7 (64), but only in MT5 with 32 bit installation, but not with 64 bit installation.


 
Integer:

I was able to use this dll in MT4 and MT5 on Windows-7 (64), but only in MT5 with 32 bit installation, but not with 64 bit installation.

How is it possible?
 

Профи, уточните пожалуйста -

In MQL, the type double has a length of 8 bytes.

Which type will correspond to MQL-double type in 32-Bit Environment c++:

c++ double, c++ long double, or c++ float ???

From the DLL-example it looks like c++ double, but other sources show that in 32-Bit Environment c++ double type is 4 bytes long.
 
Zhunko:
How did it work?


Running the terminal with admin rights, still MSVCRTD.DLL in Windows.

 

Fellow comrades, does anyone have an implemented Huffman algorithm in MQL4/5 ? (lzw, deflate)

or commented in detail in c++ ?

I've got a bug in mine, on a big file the file turns out to be bigger than the original one. :(

 
sergeev:


Fellow comrades, does anyone have an implemented Huffman algorithm on MQL4/5 ? (lzw, deflate)

or commented in detail in c++ ?

I've got a bug in mine, on a big file it turns out to be bigger than the original one. :(

For large files, you need a dynamic Huffman so that it retroactively adjusts for variance. Deflater and LZW have windows that initialize after overflow, so they are dynamic by default.

Instead of using Huffman, it is better to use dynamic ARI, see the example in C http://compression.ru/download/sources/ar/mastrukov_arith.rar

 
Reshetov:

For large files, you need a dynamic Huffman so that it adjusts retroactively to the variance. Deflater and LZW have windows that initialize after overflow, so they are dynamic by default.

Instead of Huffman, it is better to use dynamic ARI, see example in C http://compression.ru/download/sources/ar/mastrukov_arith.rar.

I'm not sure if I did it by the book, but I think I implemented the adaptive one.

I do it in two passes.
In the first one, the file is scanned byte-by-byte, and the frequency of each byte (symbol) is determined.
Then build their tree Huffman, from it determines the bit code for each character.
And the second scan goes through the file and encode it with found bit sequences.

Reason: