Time to convert libraries to MQL5 - page 4

New comment
 
falkov:

I would venture to say that those who write software for sale, as well as everyone else who writes software, need a usable editor first and foremost, not the cheesy MetaEditor.

Couldn't it be made even remotely similar to Visual Studio? I'm very annoyed when switching from VS to ME.

The MetaEditor 5 is already very good and we are developing it further.

It has intellisense, feature navigation, huge help, snippets, subs (have you seen the diffs?), integration with CodeBase/Market and native communication with the terminal. It's just all taken for granted.

The project navigator will be there, showing values in the hovering debugger too - work on the editor is going on daily.


There is also a reason to be proud - the editor with the compiler takes only 3 mb (10 mb with help files and translations), everything works very fast.
 
Renat:

You seem to be operating solely on your own scale of perception of the environment. Therefore, you do not understand why "it is not easier to demand ...." is not suitable for the real mass market.

We operate in terms of "what effect does this or that solution have within N million users? For example, in March 2012 alone over a million MetaTrader 4 and MetaTrader 5 client terminals have actually been installed. In one month alone. That's a low estimate as there are a huge number of old non-web installers.

The reason for supporting C++ -> DLL compilation is to allow real exchange of DLL sources instead of dealing with very dangerous unverified DLL files. On a market scale, this feature would provide increased security for thousands of traders who would go into "give me the source, I don't want an unverified DLL" mode.

Of course, this is just an additional solution, our main goal is to translate maximum code into native and secure MQL5.

Most of all, I wonder how many people who want to pass the source dll were deprived of this opportunity before?

How much this solution is in demand?

I, for example, could easily distribute source dlls among my friends and easily sell compiled dlls among strangers.

No one bothered that I wrote the source code in Dev-Cpp and they compiled it in Visual Studio.

The first and so far the only problem I encountered was when I was asked to connect C# code to mql, before that everything was fine.

 
Urain:

Most of all, I wonder how many people wishing to transfer the source dll were deprived of this opportunity before?

How much demand for this solution is there?

I, for example, easily hand out source dlls to people I know and easily sell compiled dlls to strangers.

And no one was confused by the fact that I wrote the sources in Dev-Cpp and they compiled them in Visual Studio.

The first and so far the only problem I encountered, when I was asked to connect code in C# to mql, and before that all was without problems.

You don't hear, and other programmers don't either.

You speak of units, while I repeat MARKETING, MILLIONS OF TRADERS. The mass market is served by automated and scalable solutions, embedded at the working platform level.

MetaEditor's built-in ability to compile C++ code into DLLs will encourage thousands of users to demand source code, and programmers to provide and post this source code to the public in advance. Our goal is for each of our solutions/ideas to have a market leverage of 1,000 to 1,000,000.


I recommend that all developers get out of the trap of a narrow view of the world as soon as possible. Forget about yourself, think and evaluate the behaviour of the masses - it gives a completely different level of understanding of the processes. The clues are everywhere - you just have to take them in.

 
Renat:

You do not hear, and other programmers do not either.

You are talking about units, and I repeat MASS MARKET, MILLIONS OF TRADERS. The mass market is served by automated and scalable solutions built in at the working platform level.

MetaEditor's built-in ability to compile C++ code into DLLs will encourage thousands of users to request source code, and programmers will provide and share this source code with the public in advance. Our goal is for each of our solutions/ideas to have a market leverage of 1,000 to 1,000,000.


I recommend that all developers get out of the trap of a narrow view of the world as soon as possible. Forget about yourself, think and evaluate the behaviour of the masses - it gives a completely different level of understanding of the processes.

What will change if the questions are put in the plural, we wondered or we gave away, I think the point is that the mass programmer won't give away the source code to the mass public. And if some solutions are handed out, there's nothing to hold them back even now.

And if the masses demand it is their public's personal problem, let them demand it to shit, and a river of codes will not flow to the masses for free, and opening a code automatically means free.

 
Urain:

What will change if the questions are put in the plural, we wondered or we gave out, I think nothing is the point that a mass programmer will not give out the source code to the mass public. And if some solutions are distributed, nothing restrains them even now.

And if the masses are demanding - it is their own problems, let them demand till they die, and the river of free codes will not flow to the masses, and to open the code automatically means free of charge.

The practice shows that around the mass of open source of different systems (in any language) is growing at a great pace.

So the trend is quite clear and we actively support it in our solutions. Soon we will make some serious changes in CodeBase allowing users to access the huge source code base on the net right from their terminals.

Our goal is to provide as much source code as possible for software developers who will be able to sell their solutions to end users through the MQL5 Market.

Those who do not want to buy ready-made solutions should be able to see other people's examples, figure them out and write their own programs.

And support for compiling C++ source code into DLLs fits nicely into our ideology.

 
OK, I admit that I have a narrow view of MT and ME. But you have to admit that general opinion is made up of such narrow views.

I will express my narrow opinion with hindsight:

I don't use debugger, because I need debugging not just anywhere, but at certain parts of the market, debugger does not have it, the rest is debugged anyway.

I don't use the visualizer because it doesn't work with a shitload of functionality.

I don't use the cloud, because I don't trust the values it returns.

I don't use watchman because I've never figured out how to use it (shame, I haven't), either it's crooked or I've got my fingers sharpened for x.

I don't use the marketplace because you can't really see what a product is until you buy it.

I still don't have folding, which is a big deterrent to writing code. Without holding, projects over 1000 lines begin to slow down a lot.

Now look at the world from my point of view, none of what you're fanfare is useful to me for some reason :(

My world is at the level of beta-testing of MT5, no brokers, the only advantage against MT4 is OOP.

You are somewhere over the horizon in your dreams, but until my (plural) vision comes close to yours, the reality of MT will be stuck in place.
 

I suggest you relax and start enjoying the features you do not use as a matter of principle.

The narrow view is not just of the terminal, but more of the market. To rise above the level, you need to change your mentality and break your old settings. Break it and you will see new opportunities for development. This is advice to all developers.

And you should take my word for it - I have much more information and knowledge about all our systems and the market around them.

 
Urain:

I don't use debugger because I need to debug not just anywhere, but at certain parts of the market, which is not available in debugger, the rest is debugged as it is.

Hmmm. Internal validation plus a debugger is a filter for most bugs.

I don't use the visualizer because it doesn't work with a shitload of functionality.

So you don't check the indices to make sure they're working? At least?

I don't use watchman, because I've never figured out how to use it (shame, I haven't), either it's wrong or, my fingers are sharpened for x.

Not by keeper, by repository :) By the way, subversion is pretty obsolete already.

I still don't have folding, which is a strong deterrent to writing code. Without folding projects over 1000 lines start to slow down a lot.

I don't use folding at all. Not anywhere. And it doesn't bother me in any way.

 
TheXpert:

Um, jeez. Internal validation plus a debugger is a filter for most errors.

So you don't check the indices to make sure they're working? At least?


Of course, in the last year, I've looked into the debugger a couple of times. But to be honest, I can write 200-line code without once pressing F7. Well, not everyone can, so debugger is needed, but agree that it is missing a very important part of the "debugging on selected section". And this is unrealized because you chose the wrong development strategy. If we had originally divided the tester into a debug-tester and an optimaz-tester, things would have been different.

I debug indices in realtime, because that is where they have to be run, while all other environments do not guarantee the correctness of their work. This is why I almost never use the tester because some indices do not always work as they do in reality. By and large I agree with the brokers who say that MT5 is still raw.

 
Renat:
I would like to start a list of potential projects. Help with links and brief description, please.

You need everything you can. At least just auxiliary libraries, e.g. for working with strings, arrays, files, for this you can take PHP as a basis, there are functions for any case.

Also, here is an interesting collection: http://musicdsp.org/archive.php.

1234567891011...14
New comment