Re: Caesura, n-th time :)

[Maximilian Albert]

Werner LEMBERG schrieb:

> Applied.  I've modified the shape of the straight caesura to be more
> in sync with other feta glyphs; I've simplified and `metafontisized'
> the code additionally.

Thanks! I was aware that the code I sent over was slightly complicated 
(as Han-Wen correctly pointed out). But my first attempt back in March 
or April had the problem that adjacent curved corners *looked* as if one 
was larger than the other although mathematically they had the same 
"size" (i.e., radius). This was due to a different amount of blackness 
caused by the slant. Therefore I manually tweaked the code this time so 
as to achieve more balanced glyphs by having roughly the same area 
filled by each rounded corner.

I'm surprised that you managed to do it with this much simpler approach 
because I'm sure I tried this myself. The "blackness" is not 100% the 
same for all corners, but it's hardly noticeably at all and I guess it 
doesn't have to be anyway. Thanks again for reworking and incorporating 
this.

BTW, regarding your comment in an earlier email about asymmetrical corners:

> Regarding `noteheads.s[012]slash', this is not true.  If you use
> mf2pt1 to generate the feta fonts you can see that.  I rather suspect
> an artifact of the current glyph generation with mftrace.

After seizing your suggestion and playing a bit with FontForge (not 
much, though), I still believe that the mentioned glyphs have rather 
unbalanced corners - unless I'm missing something obvious. I don't mind 
in any way, just thought I'd mention it because Han-Wen complained about 
the same aspect of the first caesura draft from spring (where he was 
absolutely right).

  Max