[changeset] - improve clf() compatibility

[Ben Abbott]

On Oct 20, 2008, at 11:07 PM, John W. Eaton wrote:

> On 20-Oct-2008, Ben Abbott wrote:
>
> |
> | On Oct 10, 2008, at 1:49 PM, Ben Abbott wrote:
> |
> | > On Friday, October 10, 2008, at 01:31PM, "John W. Eaton" <jwe at bevo.che.wisc.edu
> | > > wrote:
> | >> On 10-Oct-2008, Ben Abbott wrote:
> | >>
> | >> |
> | >> | On Oct 10, 2008, at 7:54 AM, Michael Goffioul wrote:
> | >> |
> | >> | > On Fri, Oct 10, 2008 at 1:46 PM, Ben Abbott  
> <bpabbott at mac.com>
> | >> wrote:
> | >> | >>> What's missing here is probably "reset" function.
> | >> | >>>
> | >> | >>> Michael.
> | >> | >>
> | >> | >>
> | >> | >> Agreed.
> | >> | >>
> | >> | >> If I had a method available to inquire as to what the  
> defaults
> | >> | >> were, that
> | >> | >> would be easy enough to impement. Do you imply that I am  
> already
> | >> | >> able to
> | >> | >> determine the default property values?
> | >> | >
> | >> | > get(0, 'defaultfigurecolor') for instance
> | >> | >
> | >> | > However, implementing this in m-code will probably be rather
> | >> | > inefficient.
> | >> | > I guess a C++ implementation (maybe with some autogenerated  
> code)
> | >> | > would be better suited.
> | >> | >
> | >> | > Michael.
> | >> |
> | >> | ok, that does work in Matlab, but not in Octave's 3.1.51+  
> branch.
> | >> Does
> | >> | this functionality exist somewhere else?
> | >>
> | >> However, setting default properties does work.  For example
> | >>
> | >> figure (1);
> | >> get (1, "color")
> | >>   ==> [1, 1, 1]
> | >>
> | >> close ("all");
> | >>
> | >> set (0, "defaultfigurecolor", [0.1, 0.2, 0.3]);
> | >> figure (1);
> | >> get (1, "color")
> | >>   ==> [0.1, 0.2, 0.3]
> | >>
> | >>
> | >> I think getting defaults is just not completely implemented yet.
> | >> Some
> | >> things do work.  For example, I see
> | >>
> | >> set (0, "defaultfigurecolor", [0.1, 0.2, 0.3]);
> | >> get (0, "default")
> | >> ans =
> | >> {
> | >>   defaultfigurecolor =
> | >>
> | >>      0.20000   0.20000   0.20000
> | >>
> | >> }
> | >>
> | >> get (0, "defaultfigurecolor")
> | >> ans =
> | >>
> | >>    0.20000   0.20000   0.20000
> | >>
> | >>
> | >> but as you noted, this fails when the default has not been  
> explicitly
> | >> set.  I guess that's just something that still needs to be done.
> | >>
> | >> jwe
> | >
> | > heck, it appears all of the defaults could be handled by the
> | > functions that create them. Is that what should be done, or should
> | > the defaults be handled at a lower level.
> | >
> | > Ben
> |
> | My inference regarding access to default property values was wrong.
> | What is needed it a reset() function (likely built-in). I've  
> requested
> | such on another thread.
> |
> | Additionally, when writing the changeset I submitted, I confused the
> | "visible" property to be an indicator of handle visibility, rather
> | than the object's visibility.
> |
> | If I understand the visibility of handles correctly, allchild(h)
> | returns all children, including those with hidden handles, and
> | get(h,"children") returns the children with visible handles. Please
> | correct me if I'm wrong.
> |
> | If it is permissible to accept a changeset for clf() without the
> | capability to reset the default propery values, please commit this
> | one. Note that while I've included a call to reset(), it is not used
> | unless requested.
>
> Before we apply this change, I think we need to fix the title, xlabel,
> ylabel, and zlabel properties to be hidden children of the axes
> object.  My best guess at how to do that is below, but it does not
> quite work.  I don't have any more time to work on this problem, but I
> thought I would post the diffs anyway to give anyone else who wants to
> work on this a starting point.
>
> jwe
>

Do I infer correctly that the x/y/zlabels are not presently hidden via  
set(hlabel,'HandleVisibility','off')?

... but by some alternative means?

Ben