MT4-Tester VS MT5-Tester - page 9

 
Petros Shatakhtsyan:

If there are about 30 or more currency pairs on MT5, they are custom quotes. And why come up with something new, especially since you can't make money on them. Who needs them ?

Since MT5 has multi-instrument testing, it will be possible to test options strategies without any extra tricks when pouring in your data.

 
Alexander Laur:

What is all this "dancing around" for?

It feels like you are left in the "stone age". Why do something with quote history when you can just connect the cloud (thousands of cores).

What is the point of plugging in the cloud and burning electricity if you don't need 90% of tick data to correctly test a strategy?

You are not testing a strategy that opens a trade at the opening of a day's candle and closes it in exactly one day, are you? What is the point of simulating the intraday price movement with the accuracy of a tick, if it will not affect the result?

fxsaber is talking about such optimization of history, and to deny its usefulness is silly. The only thing you can talk about is its uselessness for 90% of the audience.

 
Alexander Laur:

In fact, the topic is worthless, just a chat.

So pass it by. As for me the theme is certainly no worse and no more productive than 90% of the slag that is now on the forum

 
Alexander Laur:

Alexander, the topic-starter is just emphasising the usefulness of the castaway story, that's all.

And I gave an example of situations where "all ticks" are not needed. They are the reason why other modes exist. So as not to count unnecessarily, but to get adequate results.

One can simply extend these capabilities to test some other classes of strategies. Not to the detriment of what exists.


Generally, it's strange to hear nagging in response to such a constructive topic.

 
Alexander Laur:

What is all this "dancing with tambourines" for?

It feels like you are still in the "stone age". Why do something with the history of quotes when you can just connect the cloud (thousands of cores).

In general, the topic is worth NOTHING, just a chat. You can't even make a CORRECT comparison.

Aggression at things you don't understand is not surprising for a long time.

MT4 has allowed you to use custom history for at least 10 years. And to use, in particular, in the way I have shown above. That being said, you are not hindered in any way by this circumstance.

The fact is, MT5 can't yet do some of the things that MT4 can do for more than 10 years.

 
Alexander Laur:

I ask again: So what is the relevance and usefulness of the topic?

To compare two testers (not optimisers) on the same historical data. Pros and cons of each. Objective. A statement of facts. There is not a single line in the thread denigrating either product. The thread is almost a paragon of tolerance, were it not for your posts.
 
Alexander Laur:

3. The number of cores the 5 allows you to connect via the cloud will tear apart the speed of any history you convert. :)

It won't. A custom story gives you a 100x gain. That means a single run, which would take me 10 minutes, takes 6 seconds. The cloud is no help here.


But even when optimised, my 8 local free agents are equal to 800 paid cloud agents. And it's not about money (I have it), it's about rationality of spending any resources.

 
Alexander Laur:


I'm not, at all, picking on or against the castum story. There has been talk of teak and castum history since the 5's release. How many bans have been received for that talk. And as of today we have a TICK story in the tester!!! Renat has promised to do a castaway story as well. But unlike us, who are not responsible for their suggestions, the MC is responsible for all decisions made and implemented. They just need to be given time.

This method of transforming the castrum history to shorten the runtime was suggested by xrenfx a few years ago, so there is no novelty or utility in this either. No matter how you convert the history, the number of cores will still win out in speed! So what is the ACTUALITY and usefulness of the topic?

A speedup by the number of cores is not even close to an algorithmic optimization (which is what story trimming is). Even if you buy the whole world for five minutes for a hundred thousand dollars. Because iron has a ceiling.

And fxsaber is hrenfx ;)

 
Alexander Laur:

You can't judge the capabilities of a Mercedes or a six by the way their brake pumping connectors are unscrewed. That's exactly what you're doing.

That's an answer you can give in any conversation at all. Any conversation! I mean, it's such a universal way of answering without answering.
 
Alexander Laur:

On the 4 MK encountered a number of problems after this disclosure

Would you be kind enough to list this very series of problems?
Reason: