Errors, bugs, questions - page 2511
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
Dear developers, add the OnTester function to the indicator.
There is no working method to learn test completion from the indicator
Justify
Justify
In some cases it makes sense to optimise indicators in exactly the same way as EAs. The developer of the indicator knows the applied meaning of the optimization. You have made the mode of random mathematical calculations, and the indicator is actually a calculation, but with a graphical representation of the result. Of course, the indicators can be optimized and tested only by a custom value from OnTester.
I wrote a rationale in the service-desk, but since access there is no longer available, I can't provide more details. It would be logical to leave the existing SD tasks available in read-only mode - then you could link there.
In some cases it makes sense to optimise indicators in exactly the same way as EAs.
Then you can make an EA with an indicator handler.
The given recipe does not help at all. One tab is open. First I zeroed in by clicking each one three times.
Then wrote one PM, and immediately appeared supposedly six unread. Can't beat it.
Then you can make an EA with an indicator handle.
Why all these wrappers? Why not implement them in the platform itself? The tester already has the ability to test both EAs and indicators. It is absolutely illogical that the first type has the ability to return a value to the tester, while the second does not.
Why all these wrappers? Why not implement them in the platform itself? The tester already has the ability to test both Expert Advisors and indicators. It is absolutely illogical that the first type has the ability to return a value to the tester, while the second does not.
I meant that in the Tester, indicators are only run in the Visualizer. And they never execute OnDeinit and global destructor there (even after closing the Visualizer). Therefore, it is a bit troublesome to find out in an indicator that the run has stopped.
OnTester was originally designed as a custom Optimization criteria (indicators are not optimized and don't even run outside the Visualizer), not as a semaphore of the end of run. Semaphore in Expert Advisors has always been OnDeinit. There is no need to return something to the indicator in the Tester.
The point was that in the Tester, the indicators are only run in the Viewer. And OnDeinit and global destructor are never executed there(even after the Visualizer is closed). Therefore, it is a bit of a problem to find out in the indicator that the run has stopped.
OnTester was originally designed as a custom Optimization criteria (indicators are not optimized and don't even run outside the Visualizer), not as a semaphore of the end of run. Semaphore in Expert Advisors has always been OnDeinit. And there is no need to return something to the indicator in the Tester.
This boldface is the very problem, because of which I wrote the SD. Many indicators should store statistics or status when loading, but the tester does not allow to work this moment because of the failure of OnDeinit, not to mention the debugging of the code OnDeinit on the history (which is impossible now).
It's not all that clear-cut. I believe that a tester-supported program should be able to handle an OnTester event purely semantically, by definition. Visual or non-visual mode is another matter. Why can't we test the indicator in a non-visual mode is also not clear, because it would help to identify the specific problems in the indicator calculations, and in fact, this mode is still there, if we run visualization and then use the "scroll to" date button in the window.
If the indicator returned a value to the tester, it could be optimised. I think this is useful, I've encountered the need. Now the problem is solved by "dancing with tambourine".