Uncorrect behavior of error_y with grouped Y values

I've got some problems plotting an std deviation of some value Y grouped by a factor G,here's the df

df<-data.frame("X"=c(1:20),"Y"=c(1:20),"SD"=c(1:20),"G"=c(rep("A",3),rep("B",5),rep("C",4),rep("D",5),rep("A",3)))

If I plot the value and the sd without grouping them everything is OK:

p3<-plot_ly(df,
x=~X,
y=~Y,
type="scatter",
mode="markers",
marker=list(symbol="x"),
error_y =list(
    array=~SD,
    thickness=1
    )
)
p3

This yields the following plot:

But when i try to plot Y+SD grouped by G the SD appears not in the correct order. See the resulting plot below:

p3<-plot_ly(df,
    x=~X,
    y=~Y,
    color=~G,
    type="scatter",
    mode="markers",
    marker=list(symbol="x"),
    error_y =list(array=~SD,
                  thickness=1
    )
)
p3

Start by running the provided R examples with the sample data frame and compare the grouped and ungrouped plots. Trace how plot_ly handles error_y when color=~G; done means each error bar matches the SD value for its corresponding grouped observation.