My EAs are rarely perfect and generally go through a few revisions. Unfortunately, not every user likes to keep up to date. I'm wondering if there's a way to have an EA automatically update itself?
I was thinking of writing the EA in such a way that it downloads updates from a website and generates a new mqh file. I guess here's a better way to describe what I'm thinking:
There are two main components.
1) The EA.
2) Library (mqh) files.
The EA would have no logic in its code. Instead, it imports functions from a mqh file. As I envision things, every weekend, or every day, or whenever the user wants to update the EA calls a webpage with the full source code of the most recent mqh file. The EA then copies the webpage in 255 character strings, writes the strings to a text file and saves that file as .mqh
The EA then calls the equivalent of the start() function inside the mql file and the mql file goes ahead and does its thing.
I figure this sort of infrastructure might be best for distributing new versions of a neural network, or something along those lines.
Thoughts anyone?
I was thinking of writing the EA in such a way that it downloads updates from a website and generates a new mqh file. I guess here's a better way to describe what I'm thinking:
There are two main components.
1) The EA.
2) Library (mqh) files.
The EA would have no logic in its code. Instead, it imports functions from a mqh file. As I envision things, every weekend, or every day, or whenever the user wants to update the EA calls a webpage with the full source code of the most recent mqh file. The EA then copies the webpage in 255 character strings, writes the strings to a text file and saves that file as .mqh
The EA then calls the equivalent of the start() function inside the mql file and the mql file goes ahead and does its thing.
I figure this sort of infrastructure might be best for distributing new versions of a neural network, or something along those lines.
Thoughts anyone?