Libraries: Calendar - page 9

New comment
 
Aleksey Vyazmikin #:

What if it's a different country's time? I know that different countries have different times for different dates.

I checked it out as best I could. I'm exhausted. That's why I asked you to let me know if there is a bug.

So far, I have little understanding of who uses MQL-calendar or other automated solutions without corrections.

 
fxsaber #:

Checked it out as best I could. I'm exhausted. That's why I asked you to let me know if you see a bug.

I have little understanding of who uses MQL-calendar or other automated solutions.

I understand - just sharing my thoughts. I can't do historical accounting of clock translation in different countries - there is no consolidated information and my hands are down.

You are good. I don't use MT5 calendar yet - I need big historical statistics for analysis. I want to work with news (more precisely, I experimented and found a number of regularities), but I can't manage to sit down and do it.

 

During the DST Europe/USA disync, ALL calendar events are shifted by one hour on the servers that are synchronised by Europe.

 
fxsaber #:

During the DST Europe/USA disynchronisation, ALL calendar events are shifted by one hour on the servers that are synchronised by Europe.

All events? Or only events occurring in USA/USA time?

 
Mikola_2 #:

All events? Or only events taking place in the USA/in USA time?

Absolutely all of them.

 
fxsaber #:

Absolutely everyone.

For all events two weeks out of sync, force an hourly correction?

But it's so crutchy and not really effective... that you don't even have to trade for a fortnight...

 
Mikola_2 #:

To all events two weeks out of sync introduce a forced clock correction?

But it's so crutchy and not really effective... that you don't even have to trade those two weeks...

It's solved with one line.

Calendar += 3600.

I didn't make an automatic additional correction invisible to the user.

 
fxsaber #:

This is now what seems to be the correct way to use the calendar for backtests.

On the events I checked, in backtests (MQ-Demo) everything is perfect all year round.

 
fxsaber #:

This is solved with a single line.

That's understandable. But what if MQ decide (quietly) to add their own time correction?

My point is that it is not a fact that the events marked in the calendar during the disynchronisation period happen exactly at the declared time.

The calendar has the same time, and the response on the chart is +- an hour.

However, I do not use the calendar, I am bound to the hour of the day. Therefore, I'll stop floundering.

 
Mikola_2 #:

That's understandable. But what if MQ decide to (quietly) add their own time correction?

I don't see the point in considering all hypothetical situations.

My point is that it is not a fact that the events marked in the calendar during the disynchronisation period happen exactly at the declared time.

The calendar has the same time, and the response on the chart is +- an hour.

I didn't find any discrepancies. Everywhere I looked, it matches.

123456789101112131415
New comment