|
Hello! This is Tanaka from Gogojan.
A highly popular content series: interviews between our company’s representative, Hayakawa, and developers!
This time, to delve into the core of the MQL5 Converter developed by our company,we have with us Matsui, the Engineer Director at Gogojan.
|
|
|
| Graduate Engineer Director from University of Tokyo |
Reon MatsuiStudied mathematics and engineering in graduate school, focusing especially on graphs and functions. Driven by that curiosity, he developed an interest in chart analysis and the world of investment, leading him to Gogojan. He is proficient not only in web development languages but also in MQL4 and MQL5, and in the past has“GogoJungle LABO” as the sole official responder solved many traders’ and developers’ technical challenges. Currently, as Engineer Director, he is in charge of system development and technical supervision.
|
|
|
Gogojan Engineers appear! Also happy free tools
|
In this second installment, we’ll talk about the features and specifications of the new tool our company developed, the MQL5 Converter.
This is a tool that can convert EA and indicators created in MQL4 to MQL5. |
|
|
| Hayakawa from Gogojan |
|
| For this time as well, we have Matsui, the person responsible for developing this tool, joining us. In the first episode, we provided a handy tool as a free gift that can help during development. |
|
|
| Hayakawa from Gogojan |
|
|
| Gogojan Matsui |
|
Yes, it’s a tool that can delete comments and blank lines in code with one click!
This tool will also be very useful in the MQL5 Converter we’re introducing today, and of course can be handy in everyday coding. |
|
| Downloading requires a password, but it’s included in the first video, so please take a look. (The video is at the end of this article!) |
|
|
| Hayakawa from Gogojan |
|
|
What’s the difference from ChatGPT? The core of a MQL-focused “dedicated dictionary”
|
| People often askWhat’s the difference from general AI like ChatGPT?So, how is it really? |
|
|
| Hayakawa from Gogojan |
|
|
| Gogojan Matsui |
|
It is AI, but completely different from generic chat AIs like ChatGPT. A dedicated, proprietary dictionary specialized for MQL4/5 is built, and the MQL5 Converter is an AI that references many transformation rules. |
|
| What exactly is the dedicated dictionary? |
|
|
| Hayakawa from Gogojan |
|
|
| Gogojan Matsui |
|
For example, in MQL4 there isPointa variable, but in MQL5 you need to use_PointorPoint()functions.
By documenting these hundreds of subtle differences between MQL4 and MQL5 as replacement rules in a dictionary, the AI can refer to them and convert correctly. |
|
| General AI is based on past information and cannot follow real-time builds, as discussed last time. |
|
|
| Hayakawa from Gogojan |
|
|
| Gogojan Matsui |
|
| That’s right. Our dictionary is updated daily, so we can keep up with the latest builds in real time. |
|
|
What’s the difference from ChatGPT? The core of a MQL-focused “dedicated dictionary”
|
| Is the compilation guarantee on the purchase page meaningful or how does it work? |
|
|
| Hayakawa from Gogojan |
|
|
| Gogojan Matsui |
|
AI conversions do not always succeed on the first try. So,we automatically retry up to a maximum of 3 timesto re-convert and auto-correct based on the error content. |
|
| With ordinary chat AI you can’t know if it will compile, but that’s the difference, isn’t it? |
|
|
| Hayakawa from Gogojan |
|
|
| Gogojan Matsui |
|
Yes. The server will automatically compile, detect errors, and retry. If it doesn’t succeed after 3 attempts, that attempt does not consume usage. |
|
|
|
|
| Hayakawa from Gogojan |
|
|
| Gogojan Matsui |
|
Up to 10 uses per monthcan be used. As mentioned above,and failures are not counted, so you can try with peace of mind. |
|
|
How the conversion goes through—Compile guarantees and automatic retry
|
| Any tips to help the conversion progress smoothly? |
|
|
| Hayakawa from Gogojan |
|
|
| Matsui from Gogojan |
|
There are several! First,remove unused functions and commentsto prevent the AI from being distracted by unnecessary information. That helps improve accuracy. |
|
| I see! Any other points to note? |
|
|
| Hayakawa from Gogojan |
|
|
| Matsui from Gogojan |
|
Zero warnings on the MQL4 sideimproves accuracy. It is also recommended to resolve non-error warnings in advance. |
|
| The free tool introduced in Episode 1 should also help with comment and blank line deletion, right? |
|
|
| Hayakawa from Gogojan |
|
|
| Matsui from Gogojan |
|
|
|
Tips to Increase Conversion Success Rate—Comment Organization and Zero Warnings
|
| Regarding how to build the “dictionary,”what about security?Is that addressed? |
|
|
| Hayakawa from Gogojan |
|
|
| Matsui from Gogojan |
|
| Regarding the dictionary, we do not learn users’ code directly. Based on more than a decade of accumulated knowledge, we formalize it into knowledge and a dictionary using our own methods. Please rest assured. |
|
| I see. So it’s safe to use. Is it limited to EAs? |
|
|
| Hayakawa from Gogojan |
|
|
| Matsui from Gogojan |
|
Supports both EA and indicatorsWe aim to be an indispensable tool for those analyzing markets with MT4/5!
|
|
|
Security and learning sources—“The user code is not learned directly”
|
| Please share your thoughts on future development and prospects. |
|
|
| Hayakawa from Gogojan |
|
|
| Matsui from Gogojan |
|
| Currently we input one file at a time, but in the future we’d like to support multiple files and partial conversions. |
|
| Will the dictionary be enhanced further? |
|
|
| Hayakawa from Gogojan |
|
|
| Matsui from Gogojan |
|
Of course. We will continue updating the dictionary based on failures to improve accuracy. The more you use it, the more the dictionary levels up. |
|
| Finally, a message to the audience. |
|
|
| Hayakawa from Gogojan |
|
|
| Matsui from Gogojan |
|
Please try the MQL5 Converter at least once. We’d be happy if you could apply your assets to MT5 and use it to improve your trading skills. |
|
| Episode 1 |
“I wanted a tool like this!” A tool that can convert MT4 tools to MT5 instantly, eagerly awaited by developers!
|
|
|
| Episode 2 |
“Totally unlike ordinary generation AI!” Because Gogojan has over a decade of data assets, we could develop this miraculous conversion tool! |
|
|
|