Re: Problem with printing plots

[Ben Abbott]

On Mar 10, 2012, at 3:44 AM, Michael Goffioul wrote:

> On Sat, Mar 10, 2012 at 1:41 AM, Ben Abbott <address@hidden> wrote:
>>> The attached passed a few tests for me, but I suspect it is ugly to those 
>>> with c++ skills.
>>> 
>>> Critiques welcome :-)
>>> 
>>> Ben
>>> 
>>> <changeset.patch>
>> 
>> This change breaks the synchronization of the axes position.
>> 
>> To see the effect run the legend demos.
> 
> I think your patch is a bit too complicated and misses a few things,
> especially running automatic updaters. The easiest is to start from
> the auto-generated code (did you have a look at it?):
> 
>  void set_position (const octave_value& val)
>  {
>    if (! error_state)
>      {
>        if (position.set (val, false))
>          {
>            set_positionmode ("manual");
>            update_position ();
>            position.run_listeners (POSTSET);
>            mark_modified ();
>          }
>        else
>          set_positionmode ("manual");
>      }
>  }
> 
> As you can see, you're missing calling update_position() and
> set_positionmode(). I can;t test at the moment, but what I'd do is
> something like this:
> 
>  void set_position (const octave_value& val)
>  {
>    if (! error_state)
>      {
>        octave_value new_val (val);
> 
>        if (new_val.numel () == 2)
>          {
>            dim_vector dv (1, 3);
> 
>            new_val = new_val.resize (dv, true);
>          }
> 
>        if (position.set (new_val, false))
>          {
>            set_positionmode ("manual");
>            update_position ();
>            position.run_listeners (POSTSET);
>            mark_modified ();
>          }
>        else
>          set_positionmode ("manual");
>      }
>  }
> 
> Michael.

Nice. I hadn't noticed the resize method before, and wouldn't have realized it 
could be used to copy a 1x2 octave_value into a 1x3 octave_value.

The attached works for me. Ok to push ?

Ben

changeset.patch
Description: Binary data