Re: AspectRatio Patches [changeset]

[Ben Abbott]

References: <address@hidden> <address@hidden> <address@hidden> <address@hidden> 
<address@hidden> <address@hidden> <address@hidden> <address@hidden> 
<address@hidden> <address@hidden> <address@hidden> <address@hidden> 
<address@hidden> <address@hidden> <address@hidden> <address@hidden> 
<address@hidden> <address@hidden> <address@hidden> <address@hidden> 
<address@hidden> <address@hidden> <address@hidden> <address@hidden> 
<address@hidden> <address@hidden> <address@hidden>
To: Konstantinos Poulios <address@hidden>
X-Mailer: Apple Mail (2.1082)
Return-Path: address@hidden
X-OriginalArrivalTime: 03 Dec 2010 12:43:36.0943 (UTC) 
FILETIME=440AFF0:01CB92E7]

On Dec 3, 2010, at 4:23 AM, Konstantinos Poulios wrote:

> On Fri, Dec 3, 2010 at 12:34 AM, Ben Abbott <address@hidden> wrote:
>> On Dec 2, 2010, at 9:39 AM, Konstantinos Poulios wrote:
>> 
>>>> 
>>>> I've pushed both changesets.
>>>> 
>>>> There's been a several changes to the backend in the last few weeks. There 
>>>> are now some problems with the demos ...
>>>> 
>>>>        rundemos plot
>>>> 
>>>> ... the ones I've noticed are for "patch" and "isosurface". Demo 2 for 
>>>> "hold" also looks like there is a problem.
>>>> 
>>>> I mention this to enourage you to run the plot demos to make sure your 
>>>> changesets don't break anything (which I don't think they have).
>>>> 
>>>> Thanks
>>>> Ben
>>> 
>>> it seems that there is a small bug in the daspect and pbaspect
>>> arguments handling. E.g.:
>>> 
>>> octave:3> daspect([1,1])
>>> error: invalid value for array property "dataaspectratio"
>>> error: called from:
>>> error:   /opt/share/octave/3.3.54+/m/plot/daspect.m at line 68, column 9
>>> 
>>> I would suggest something like the fix in the attached changeset.
>>> 
>>> BR
>>> 
>>> Kostas
>>> <daspect_pbaspect_input_length_2.changeset>
>> 
>> Does it make sense to set the 3rd element to unity each time, or to preserve 
>> he existing value?
>> 
>> Specifically, how should pbaspect and daspect behave in the axes contains a 
>> 3D plot and either aspect is set using a vector of length 2?
>> 
>> If the 3rd property value is to be preserved then I can push the attached.
>> 
>> Ben
>> 
>> 
> hmm, I am not sure if it makes more sense to retain the current value.
> Since aspect ratio values are interpreted relatively to each other in
> my understanding they make sense only when they belong to the same
> set. But this is just my opinion, if you want I can provide a
> changeset that takes into account the current value for the third
> aspect value.
> 
> In the meanwhile I have noticed that ML will not accept an array of
> dimension 2 even for the 2D case, so maybe we could just eliminate the
> possibility of input arrays with length 2 in pbaspect and daspect at
> all.
> 
> Anyway this is a secondary issue, any decision will not make a big difference.
> 
> Kostas

Ok, I've pushed your change.

Ben