On 2/6/06, Han-Wen Nienhuys <address@hidden> wrote:
Some comments,
(tie2.patch)
+ if (abs (pos) < 3 * Staff_symbol_referencer::staff_radius
(details_.staff_symbol_referencer_))
why the 3 ? I take it you mean
<= 2 * ...
Hmmm, I think you're right.
Maybe we need an inside_staff() method?
Absolutely. There are already too many magical numbers floating around. That would make the code more readable. I could add one in my next shot at the tie code?
Why are you centering ties in spaces outside the staff in the else case?
As I wrote, most of the ties outside the staff are placed way to high (or low). It just got better results with centering them (just look at the pngs :-)).
I would expect them to look good by default, especially if
sign(pos) == tie_direction
Shouldn't you just leave them in the else case?
Hmmm... I'll need to experiment a bit more with this.
(tie1.patch)
Why this change, i.e. why should you center between the staff lines
outside the staff as well?
Same as tie2.patch: Just to avoid the huge gaps between the noteheads and the ties.
(tie3.patch)
the old code tries to put the top of the tie in the center of a space,
should it touch a staff line. Your patch makes it do
center_tie_vertically, which will put the y-center of the tie at the
position_ of the Tie_configuration. Since this is (in the case of a
larger tie) often more than
0.5 staff-space, this is not the right
solution (since the Tie_configuration with the next or previous
position_ should fit the requirements better)
Hmmm, I guess you're right. I'll have another try at the code.
In short, while I agree that the PNGs you send look better, it seems to
me that the patch uses the wrong approach to reach the target. Maybe you
can have another look?
Will do.
Also, it would help me if you could show the effect of each patch
individually in a PNG file.
Shure!
Regards,
Erlend Aasland