[Scilab-users] xpoly()

[Samuel Gougeon]

Hello Stefan,

Le 03/12/2014 21:01, Stefan Du Rietz a écrit :

> Hello Samuel,
> thank you for your comprehensive reply, which I have now investigated.
>
> Yet, I cannot agree with you.
>
> Do you really consider it logical that a built-in low level line 
> function (xpoly) should change a superior entity (axes) when there 
> already is a built-in plot function (plot2d) for that?

I do not consider xrpoly, xarc, xrect.. as low-level functions. It would 
be complicated to do with plot2d what xarc does.

>
> And that you can prevent (with a.auto_scale = "off") the plot function 
> but not the line function to change the axes???
I do not clearly understand what you meant in your previous message 
about that, and i did not make tests.
The autoscale status should be taken into account by all functions, 
so-said low as well as high level ones.
If the new xpoly updates bounds while autoscale=="off", this is the bug.

>
> Furthermore, if you want to draw "thousands" of lines with xpoly(), it 
> must take some time to recalculate the data_bounds every time!
You would do it within drawlatter / drawnow. You may use profiling 
functions to measure it with 5.4.1 versus 5.5.0.
>
> "when one plots something" and wants "to be able to see the plotted 
> object" one can use plot2d ...
>
> If I use a low level line function, it is because I know what I am 
> doing and don't want anything to interfere with it!

You would be kind to tell us how to match bounds when it is required 
after having drawn 1000 shapes, titled texts, tilted axes with log and 
inverted axes, etc.

Best regards
Samuel