Re: Does better polynomial calculations for avoid objects. (issue 4860042)

[Mike Solomon]

On Aug 23, 2011, at 6:21 AM, address@hidden wrote:

> 
> http://codereview.appspot.com/4860042/diff/8001/flower/polynomial.cc
> File flower/polynomial.cc (right):
> 
> http://codereview.appspot.com/4860042/diff/8001/flower/polynomial.cc#newcode80
> flower/polynomial.cc:80: return ret_max ? sols.back () : sols[0];
> you could return an interval instead;
> 
> Polynomial::range() ?

I'm hesitant to incorporate native Lilypond types into polynomial.cc - I'd like 
to keep the included files as "polynomial.hh" and "warn.hh" so that it could 
work stand-alone with minimal modifications.
Een beter milieu begint bij uzelf. Hergebruik!

> 
> http://codereview.appspot.com/4860042/diff/8001/lily/bezier.cc
> File lily/bezier.cc (right):
> 
> http://codereview.appspot.com/4860042/diff/8001/lily/bezier.cc#newcode223
> lily/bezier.cc:223: Bezier::minmax (Axis ax, Real l, Real r, Direction
> d, vsize lres, vsize rres) const
> you could be less stingy with letters.  "axis".
> 
> what do lres and vres mean?
> 

I've expanded these out.

> http://codereview.appspot.com/4860042/diff/8001/lily/bezier.cc#newcode258
> lily/bezier.cc:258: return p.minmax (sol[LEFT][res[LEFT]],
> sol[RIGHT][res[RIGHT]], d != LEFT);
> could return an interval here too? Bezier::extent ?
> 

It's up to you.  It is more an architecture/style question than anything else.  
If you're OK including interval.hh in polynomial.cc, then I can do this.

New patchset up.

Cheers,
MS