MetaTrader 5 Strategy Tester and MQL5 Cloud Network - page 10
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Don't worry, we have several areas under development in parallel and have already started to implement the test visualiser.
The problem of visualising the testing process is actually difficult, because the tester is completely outside the terminal. At the moment we decided that the visualisation will work directly in the (only local) testing agent itself, representing a player window with graphs. This means that we need to transfer to the tester a copy of the entire visualization system used in the terminal.
We move the visualisation to the tester agent so as not to burden (spoil) the architecture of the terminal itself.
Interesting. Does it mean that it will be possible to run several visualisers at the same time? Maybe it would be better to bring this part of the terminal into dll and use the dll terminal and load it to the agent during visual testing (so as not to waste resources when visualization is not required). I'm looking forward to it, as most of the development is stalled for lack of it.
This question sounds strange.
You have one terminal->tester starts in it->with it the visualization.
What visualization do you mean by several testers in one terminal?
What's so strange about it? 4 cores, 4 agents, 4 simultaneous visualisations :)
What's strange about it? 4 cores, 4 agents, 4 simultaneous visualisations :)
then that's the right thing to do:
4 cores, 4 agents, 4 terminals, 4 simultaneous visualisations
What's so strange about it? 4 cores, 4 agents, 4 simultaneous views :)
then the correct way is as follows:
4 cores, 4 agents, 4 terminals, 4 simultaneous visualisations
Single test runs on a single core, here as correctly noted by Sergeev as an option only 4 terminals will work.
PS
But instead of such an option I would personally prefer the possibility of gathering all local cores into some pool (so that they are seen as one core).
Essentially, such an implementation would give you the ability to speed up single testing considerably.
Interesting:
But instead of this feature, I would personally prefer to be able to gather all local cores into some pool (so that they are seen as a single core).
In essence, such an implementation would give the ability to speed up single testing considerably.
There will be no acceleration at all. How will the distribution of tasks and synchronization of results between cores happen, if the language has no means for this (except for some hacks)?
We have launched five MetaTrader 5 Cloud Servers in different parts of the world for load balancing:
Soon they will start showing available resources as "348 of 1456 agents available" in the "Hardware" column.
At first, the system will work in internal testing mode, and then we will start public testing. For now you can register your agents in MQL5 Cloud Network - all agents are available in your own profile on MQL5.com.
I am happy about it.
I have put about 32 cores in the pool.
However, most of the cores are the cores of laptops, which are of course not always enabled.