bugs & dysfunctions - page 3

[Alain Verleyen]

graziani:

i cannot believe that you can't reproduce this failure out of first try.

just put the sensitivity to 20/50 and try to move the anchor points of trend line, fibo or any object around.

Stay calm :-)

It may be is another parameter that comes into play.

[graziani]

angevoyageur:

Stay calm :-)

It may be is another parameter that comes into play.

:)

 

[graziani]

precise time scale.

[Alain Verleyen]

graziani:

OK, it seems that we are now getting away from my idea of reporting this "dysfunction".

The point is that when you define a dynamical array with SetIndexBuffer(int, double, ENUM_INDEXBUFFER_TYPE ), it is logically to assume (as it is not specified in documentation) that by defining

it will not be possible to see the existence of the third buffer anywhere: on graph, indicator properties, or from EA using this indicator.
Especially as you are FORCED to define the nature of usage of this dynamically allocated array. 
(To repeat once more, defining number of buffers and plots through #pragmas DOESN'T mean that you will use all of them. )

But besides this not being true, it seems that this is also not obvious to experienced programmers, so i don't see any sense at the moment in reporting 'dysfunctions' like this.

So ... should i report things like this or not?

I will try once more with something similar i have in mind, but at the moment this seems like waste of (not only mine) time ;)

It is useful. Now each even inexperienced can get an opinion, from different arguments. This is a forum, so I think it is necessary to give my comments if I don't agree. Nothing personal.

If you have to report something to Service Desk about all of this, it's that documentation is not clear enough. But my opinion is still that you don't have to interpret what is not said in documentation.

Kindly.

[Simon Gniadkowski]

graziani:
precise time scale.

Yes,  that certainly makes a difference.

One thing that annoyed me for a short time was the Fixed chart position not working in the same way as it did with MT4,  but then I looked at the Help and found that it was behaving as it should. 

[Alain Verleyen]

graziani:
precise time scale.

When drawing an object on a chart that has a period other than M1, the following features of magnifying appear:

• When magnifying a point of an object to one of the price extremums (OHLC), the specific minute that extremum has been registered at is determined. Point of the object is bound to that minute, and it is correspondingly displayed in properties of the object. This kind of behavior allows keeping the correct positioning of objects when switching between timeframes.
• If the "Precise time scale" option is additionally enabled, then you may observe an effect of jumping aside of an anchor point from an extremum point. This behavior appears if the actual extremum point doesn't correspond to the extremum point of a bar.

RTFM :-D

[graziani]

angevoyageur:
RTFM :-D

i cannot believe it :)

and now that i know what 'precise time scale' means, i want to use it!!

but ... there is a bug because after you select one point, and align it to 'precise moment in time', while you try to align second point, the first aligned point automatically gets realigned to closer candle and gets aligned to a new, sometimes different value.

or is there a click to manage that also? :) 

[Alain Verleyen]

graziani:

i cannot believe it :)

and now that i know what 'precise time scale' means, i want to use it!!

but ... there is a bug because after you select one point, and align it to 'precise moment in time', while you try to align second point, the first aligned point automatically gets realigned to closer candle and gets aligned to a new, sometimes different value.

or is there a click to manage that also? :) 

You see this was not waste of time, we learned a lot today :-)

[medikament1978]

Pending orders alwys expire ate the end of the day in backtesting (independently of timeframe!) when they are sent via OrderSend with  parameters

  

   m_request.type_time   = type_time;   // type_time set to ORDER_TIME_GTC   

or

  m_request.type_time   = type_time;   // type_time set to ORDER_TIME_SPECIFIED

  m_request.expiration  =expiration;  // expiration set to a datetime in 5 or 10 days

ANY IDEAS?

[Alain Verleyen]

medikament1978:
Pending orders alwys expire ate the end of the day in backtesting (independently of timeframe!) when they are sent via OrderSend with  parameters

  

   m_request.type_time   = type_time;   // type_time set to ORDER_TIME_GTC   

or

  m_request.type_time   = type_time;   // type_time set to ORDER_TIME_SPECIFIED

  m_request.expiration  =expiration;  // expiration set to a datetime in 5 or 10 days

ANY IDEAS?

Can you post some code (a little more), logs or screenshot to demonstrate your issue ?