Public discussion of the formula for calculating the cost of resources in the MQL5 Cloud Network - page 25
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
First we will launch in the real world.
For the Championship, the Expert Advisor has been tested for about a month, maybe more (multicurrency - first in separate pairs, then together in parts, then together in parts of parts, etc.) - at this rate, it would have cost me at least $300 per month. Ow to those who are willing to pay that much. Anyone interested? It may be worth $300, but at that price, I'll do with my own resources.
That's the thing: minor intermediate tests should be done in-house, at least that's what I do (I have 13 cores available at the moment). But larger and more time-consuming tests should be transferred to the cloud.
In real life (when the basis of the Expert Advisor is ready and its parameters need to be optimized) the cloud, in my opinion, should be used only when the test has no time to be passed during a weekend.
As for the championship, it's a special case because the amount of testing will be quite large. We have to either lower the price or make the cloud for free.
We will collect the statistics for the next three months after the launch and will adjust the prices.
At first we will launch it on the real.
I am afraid that after the start of the championship the need for the cloud will decrease dramatically. Perhaps after the start of the market the use of the cloud will be needed again, but I personally doubt it for now due to the filling of the showcase with experts.
For the championship, the EA was tested for about a month, maybe more (multi-currency - first by individual pairs, then together by parts, then together by parts of parts, etc.) - with such a spread, it would have cost me at least $300 per month. Ow to those who are willing to pay that much. Anyone interested? It may be worth $300, but for that price I'll do with my own resources.
+1, but free forever :) The thing is, consumer voices are not heard at all, some "producers" say it's "cheap".
It's hard to say yet, I'll wait a month or two. But something has made me doubt whether my agents should be involved in the cloud.
Interesting material, above all. As the saying goes - Respect and respect to the creators. I have read a couple of threads in the forum "from cover to cover".
My opinion about the above and the above:
If (for now) MT5 is not so popular and in demand as MT4 - may be we should add this service for MT4? (Make up a requirements specification and take it to the right cabinet :) Then there will definitely be a demand and there will definitely be an offer. I don't think optimization is out of the question.
I myself have not moved to MT5 (yet) due to many factors (the same demand and the same suggestions).
My goal is to use all the resources I can get my hands on (32 agents / cores at the moment).
About price formation thoughts :
MQ has a sky in which a cloud lives, there are many agents in that cloud, the agents are on physical pisukes. What you can do is to take a standard expert and run it on a few/specified/all tenant agents to evaluate performance. My point is that there are EAs/experts that can compute fast and/or slow. Whoever writes it. Accordingly, the optimization time in the cloud will be different. Hence the price for renting the agents/core in the cloud must vary (even if by +/-1...xx%). For 100% we can take the same MACD, which goes as an example in all MT. After all, the agent may exist both on first generation Atoms and on more modern CPUs like i7...
One of the possible ways of using (or not using) all the power/force of the cloud can be the setting in the optimizer of the terminal. For example, if the consumer wants to save on cost - he uses his own and remote/cloud agents. If the consumer wants to quickly calculate/optimize something and not burden his system - he uses the cloud. The same choices can be found/thought out quite a lot.
My opinion on the price is adequate enough for both parties. Altruism is plentiful everywhere, but it tends to dry up over time...
p.s. I haven't said everything yet. :)
It is still difficult to understand how the cost is calculated
It is still difficult to understand how the cost is calculated
These were test runs so far + a number of bugs in the cludes have been fixed.
We'll run the cludes in real the next few days and see what happens.
These were test runs so far + a number of bugs in the cludes have been fixed.
One of these days we'll run the cludes live and see what happens.
Yeah yeah yeah I figured as much.
at the launch of course the statistics will be reset - as I understand and hope to
p.s.
Renat, I was very happy and surprised - the kernel from one of my virtual machines suddenly worked
i have i7 cores on one of my virtual machines and it counts as such (is it your fault?)
You seem to have solved the problem with the virtuals ,
but apparently not for all cases, on the other - virtual - does not allow to turn on
Yes yes, that's what I thought.
at startup, of course, the stats will be reset - as I understand and hope
p.s.
Renat, I was very happy and surprised - the cores from one of my virtual machines suddenly worked
i have i7 cores on one of my virtual machines and it counts as such (is it your fault?)
You seem to have solved the problem with the virtuals ,
but apparently not for all cases, on the other - virtual - does not allow to turn on
... And about cost - Has the question been resolved in any way?
What is the price for agents to work with the cloud?