Errors, bugs, questions - page 372

 
drm1:
It's understandable that you can't, but when will these developers add a SIMPLE function!
write a request to servicedesk.
Общайтесь с разработчиками через Сервисдеск!
Общайтесь с разработчиками через Сервисдеск!
  • www.mql5.com
Ваше сообщение сразу станет доступно нашим отделам тестирования, технической поддержки и разработчикам торговой платформы.
 
sergeev:
write a request to servicedesk.
ok
 
Wrote an EA which is very slow to test (slower than real time). I suspect there are a few pieces of code that may be slowing down execution. Is there any way to find out how much time it takes to execute a certain piece of code?
 
molotkovsm:
Wrote an EA which is very slow to test (slower than real time). I suspect there are a few pieces of code that may be slowing down execution. Is there any way to find out how much time it takes to execute a certain piece of code?
Sure, you can. Try to insert GetTickCount() at the beginning and at the end of the code fragment being checked and calculate their difference
 
MONTEGRO:

I'm looking at it... and I don't see critical errors in them, for example, such that forward falls off and no matter what.
About money: MetaTrader5 is designed for what? No one will buy a pig in a poke, neither the brokerage companies nor other users (even if they would sell it)
and the developer company is primarily interested in the release of stable builds. But what happens in reality - the build is released,
and no basic functionality tests are performed. The question is, why did you release it in the first place?

1. It has already been hinted at - testing is underway and no one guarantees absolute availability of the shopping complex.

2. At the expense of testing builds developers

And who told you (where did you get it personally) that such testing is not going?

Here's just wondering how and when the next build will be released, if the developers will each time a full performance testing of the entire trading complex?

An additional question - but why do they need us as independent beta testers?

3. Now imagine how many people work there and how many lines of code they have already written.

Simply to eliminate errors without destroying the serviceability of the whole project becomes problematic, while we make changes in the terminal (at our request as well).

As usual, we forgot about the server side. Traders need a terminal every week (or better every day), and we are not interested in the server side...

 
MONTEGRO:

There is, but in MT5 the same errors appear regularly (forward crashed for the second time) and you have to wait weeks for a fix... is this "prompt" ?

1. Probably it will fail 10 times. So what of it?

2) Builds are released once every 2 weeks on average, and it seems to be a practice that is "accepted" by most of those that are present on this forum.

If the developers can release more often, then they will, but I don't think they'll adjust for you (or me).

I don't think they will even tune in to DCs and banks (as many believe the true customers of the "music").

MONTEGRO:

Judging by the number of reactions to its absence in the 390th build in this thread - it needs it only a few system designers. The rest probably do not know what it is and what it is for.
No comment from me about your figures in 20K code, which 100% will not earn anything either on the virtual, or on the real ))))

Agree, it's not quite correct to state without proof that this or that project can't make money. And even so categorically...
 
Yedelkin:
Sure, you can. Try to insert GetTickCount() at the beginning and at the end of the code fragment being checked and calculate their difference
Thanks for the tip, I tried it. Doesn't work in the tester, apparently every tick generates a certain GetTickCount() value and does not change, the difference is always 0. I will try it on demo account tomorrow.
 
molotkovsm:
Thanks for the tip, tried it. Doesn't work in the tester, apparently every tick generates a certain GetTickCount() value and does not change, the difference is always 0. I will try it on a demo account tomorrow. Of course, I want to have something in the tester to measure the execution time of a code fragment without waiting for ticks and certain market conditions.
Perhaps, I should write to SD concerning the tester. Let them emulate at least from the beginning of testing...
 
molotkovsm:
Thanks for the tip, tried it. Doesn't work in the tester, apparently every tick generates a certain GetTickCount() value and does not change, the difference is always 0. I will try it on a demo account tomorrow.
I was wrong, it is the same in the tester and in the demo. Apparently the lag is not at all where I thought it would be.
 
You've had a clear chart and markup going around at the start of every weekly trading session since time immemorial! XP SP2 32-bit. I would have filmed the whole thing, but I'm having trouble with that at the moment. Hasn't anyone noticed? The jerking seems to disappear in the morning, I haven't checked it specially, it can be fine in the afternoon. Restarting the terminal doesn't help, I have to wait for daylight hours. Either data downloading or unpacking of quote history goes flaky or the chart window starts freaking out due to automatic output cut-out... There is "wobbling" back and forth for a few bars with frequency of 1 every few seconds. Anyway, it's time for a fix.