Re: waitbar function

[Michael Goffioul]

On Thu, Nov 3, 2011 at 3:57 PM, John W. Eaton <address@hidden> wrote:
> | >  http://hg.savannah.gnu.org/hgweb/octave/rev/5acb5c25e4ae
> | >  http://hg.savannah.gnu.org/hgweb/octave/rev/195ff3561152
> |
> | Could you explain why you need non-integer figure handles?
>
> It seems to be what Matlab does.  Also, I think people would expect a
> sequence like
>
>  figure (..)
>  waitbar (..)
>  figure (..)
>
> to generate figures 1 and 2, not 1 and 3.

There's actually a figure property for that: integerhandle. It's
currently unused/unimplemented in graphics.cc. It would probably
better to implement it correctly.

> | - Wouldn't it be more efficient to modify the properties of the
> | existing patch object, instead of recreating a new one?
>
> Yes, probably, but I'm not that good with the graphics stuff and it
> didn't occur to me that it would work...
>
> Also, when I try the following, the patch color disappears after the
> xdata property is updated.  Is there something else I have to do other
> than setting the xdata property?

I think it might be due to incomplete implementation of patch
properties relationship. Normally, property relationships are
implemented in subfunctions in __patch__.m. I guess the best is to
compare properties before and after changing xdata to detect any
anomaly. The patch rendering in OpenGL ias based on the face/vertex
properties, not the [x|y|z]data properties.

Michael.