[Scilab-users] xsetetch and xgetetch incompatible definitions of "frect"

[Samuel Gougeon]

Le 26/11/2016 18:49, Samuel Gougeon a écrit :
> Le 26/11/2016 18:14, Antoine Monmayrant a écrit :
>> Other option for a SEP: could the optional parameters that we add to 
>> subplot be also added to "newaxes" ?
>> I think it could be convenient and coherent with the way "figure()" 
>> works if we could do things like:
>>
>> a= newaxes(propertyname, propertyvalue)
>>
>> example:
>>
>> a=newaxes('background', color("lightgray"), box, 'on')
>>
>> as an equivalent to :
>>
>> a=newaxes()
>> a.background=color("lightgray")
>> a.box="on"
>>   What do you think?
> .
> I think that it's a much better proposal than with subplot().
> But IMO features must not be duplicated: xsetech() features should 
> then be merged and extended only in newaxes(), not in both newaxes() 
> and subplot(). Duplicates lead to heavy useless code maintenance and 
> usages discrepancies. Very bad.

Hmmm, subplot() focuses only on .axes_bounds. So, we may add only the syntax
subplot(axes_bounds)
Now, why not allowing to fully tune the axes "viewport" by adding a 
"margins" optional argument
enabled as well for the subplot(m,n,p) and subplot(mnp) syntaxes? We 
would have:

subplot(mnp [, margins])
subplot(m,n,p [, margins])
subplot(axes_bounds [, margins])

IMO, data_bound is something else related to data scales, not to the 
viewport configuration.
replot(..) already allows to tune .data_bounds after the axes creation.
It has been improved and merged in Scilab > 6.0-b2: 
https://codereview.scilab.org/#/c/18506/

Upgrading subplot() could be done and reviewable quite quickly. But who 
will review and merge it, and when?

Samuel