Page 1 of 1

BUG: Wrong Barnumbers

Posted: Mar 17 2006
by Chris
TSSupport,

you have a big bug with your BarNumbers, you don't count them correctly when it comes to multiple symbols. This will definetly result in a complete mess up of studies using the BarNumber.

Both CurrentBar and BarNumber are showing wrong results.

For your convenience I have attached a screenshot taken from TS - this is how it should look like.

Thank you,

Chris

P.S.: The first study is a plot of "CurrentBar" the second of "BarNumber".

Posted: Mar 17 2006
by Chris
Another strange thing - BarNumber and CurrentBar showing different results :? .

Posted: Mar 21 2006
by Chris
Is there a workaround for this?

Posted: Mar 21 2006
by Alex Kramer
You probably have TS build never than 3060; with earlier builds when counting indicators on mixed series TS behaved differently. The difference is that the starting bar for technique calculation used to be chosen as follows:
-the first available bar for the symbol with the highest resolution on chart was found and from that bar the necessary number of BarReferences was plotted. Correspondingly, all indicators start calculation from the starting bar of its symbol.

Now TS plots the required number of bars beginning from the first bar of the underlying symbol for study if the indicator does not depend on any other symbol. This is what you saw in the picture.

MC (like the latest Omega) uses the first approach. Maybe we'll change in the future.

As to the different CurrentBar and BarNumber values - you have different number of Bar References auto-selected; if you set the BarReference value manually =1 for both, values will match.

We cannot think of a possible workaround now without changing the implementation to the TS-like one.