bug#37563: 27.0.50; fit-frame-to-buffer does not account for line-spacing

[martin rudalics]

> thanks for the quick reply!  The fix is working for me.  The separate
> window-*-height functions are much better than what I sent a few
> minutes ago, I missed all kinds of scenarios, of course.

For consistency, I would use these functions in 'fit-window-to-buffer'
as well.  But I haven't looked into all consequences yet.  There's
also Bug#14825 still sitting around the corner, awaiting a proper
solution.  It's somehow troubling that all these substitute canonical
character height with real line height fixes are inherently backward
incompatible.  What if someone did mean to use the canonical character
height there?

> There's one thing from my patch, however, that I think is missing in
> yours:

I think you're right but I need to see your patch first.  It's not
here yet.

> In ll 8831ff, the height is rounded if frame-resize-pixelwise is
> nil.  That also uses char-height, and I think it should be line-height
> there as well.  In that case, char-height is no longer used in the
> function.

Which certainly would be an asset.

Thanks, martin