Re: Patch for bar objects

[Michael Goffioul]

On 11/19/07, David Bateman <address@hidden> wrote:
> Ok, sorry I have 4 patches unapplied at the moment and missed that this
> wasn't applied yet.. Ok, we are going to have a conflict between our
> patches as well, if John applies them separately. Therefore please
> consider the attached patch that combines both of our patches into a
> single contribution..

Fine for me, thanks.
Note however that compared to Matlab, the bar width handling is slightly
more complex in the group case: the width is than more related to individual
bars than the group itself (although there's a relation between the bar width
and the group width). For instance if the width is 0.8, inidividual bars are
not collated within a group. If the width is 1, individual bars are collated
within a group, *but* the groups are not collated (the group width is < 1).
In our current code, I think that individual bars are always collated within
a group and the specified width relates to the group, right?

Michael.