Code review comment for lp:~alan-griffiths/mir/more-surface-resize

> > > 227+ * Defines an arithmetic progression of sizes starting with min_width
> > (if
> > > set, otherwise 0)
> > > would a default of 1 be better?
> >
> > I don't follow. If min_width isn't set the series is:
> >
> > 0, width_inc, 2*width_inc, ...
> >
> > You're suggesting this:
> >
> > 1, 1+width_inc, 1+2*width_inc, ...
> >
> > > That way the server wouldn't be stepping over
> > > the clients default of 0 by default. (zero seems to be a request to fix
> the
> > > surface size to min_width, min_height)
> >
> > And I don't follow this.
>
> So, I was suggesting that min_width is always set and gives the sequence:
> min_width, min_width + 1*width_inc, min_width + 2*width_inc, etc
> And, with that 1 would be a good default value, because then you get:
> min_width, min_width+1, min_width+2, etc.

OK, I think you're writing "*min_width* is always set" when you're suggesting a default for *width_inc*.

So, we're actually talking about the same behavior except, possibly, in cases where the client applies a spec to a surface that has previously had width_inc set:

What we have is a surface spec where width_inc is optional and only if that is set by the client will it update the width_inc on the server.

Are you suggesting that the surface spec should always have a width_inc (default 1) and that this should override the server setting if not set by the client? I don't think that is right.