On Thu, Sep 06, 2007 at 02:22:59PM +0200, Alfredo Braunstein wrote:
> Abdelrazak Younes wrote:
> 
> > Alfredo Braunstein wrote:
> >> Abdelrazak Younes wrote:
> >>> I think there is a misundertanding here... I agree that the baseline is
> >>> the reference and I did not change that philosophy I think. For me the
> >>> baseline is the baseline of the inset, not the baseline of the inner
> >>> Text. The fact that InsetCollapsable adjust its baseline to its inner
> >>> Text is just a special case.
> >> 
> >> Not so special (pretty much every Text related Inset) but yes.
> > 
> > Well, tabular is another special case but that's what I meant yes.
> > 
> >> 
> >>> Now, that's my turn to be unsure of the clarity of what I'm saying...
> >> 
> >> I'm not sure either... So you don't plan to switch every InsetX::draw
> >> function to accept top_y, just TextMetrics::draw?
> > 
> > No, I didn't even do that for InsetCollopsable, please note that the
> > code at the top of the draw() method now is:
> > 
> > int const top = y - tm.ascent();
> > 
> > So, this _is_ the text baseline. The only effect of my change is that I
> > factorized the substraction of the text ascent. Probably we should
> > rename top to baseline. Now I understand where you got confused...
> 
> Sorry I can't follow you. Maybe I'm just tired today. ;-)
> 
> What I read from the code is:
> 
> Insetcollapsable::draw(x, y) (and other insets):
> 
> y is the baseline of the InsetCollapsable (that happends to be the same
> baseline of the InsetText, at least in inlined situation) [-- I don't get
> why you would substract the ascent in the non-inlined situation]
> 
> y - tm.ascent() is the top of the Text, not a baseline
> 
> - InsetText::draw(x, y):
> 
> the y refers to the top of the *Text* (horrible, not even top of the
> InsetText)
> 
> - TextMetrics::draw(x, y):
> 
> the y refers to the top of the Text
> 
> 
> 
> It seems pretty much a mess to me.

whatEverDraw(x, y) should draw the baseline of the first embedded line
on y. Everything else should be changed to adhere to that scheme.

Andre'

Reply via email to